drug development research paper

An overview of drug discovery and development

Affiliation.

• 1 Department of biomedical Science, Nazarbayev University School of Medicine, Nur-Sultan 010000, Kazakhstan.
• PMID: 32270704
• DOI: 10.4155/fmc-2019-0307

A new medicine will take an average of 10-15 years and more than US$2 billion before it can reach the pharmacy shelf. Traditionally, drug discovery relied on natural products as the main source of new drug entities, but was later shifted toward high-throughput synthesis and combinatorial chemistry-based development. New technologies such as ultra-high-throughput drug screening and artificial intelligence are being heavily employed to reduce the cost and the time of early drug discovery, but they remain relatively unchanged. However, are there other potentially faster and cheaper means of drug discovery? Is drug repurposing a viable alternative? In this review, we discuss the different means of drug discovery including their advantages and disadvantages.

Keywords: drug repurposing; high throughput; natural sources; small molecule.

Publication types

• Artificial Intelligence
• Drug Development*
• Drug Evaluation, Preclinical

• Reference Manager
• Simple TEXT file

People also looked at

Mini review article, drug discovery and development: introduction to the general public and patient groups.

• 1 Evotec SE, Molecular Architects, Integrated Drug Discovery, Campus Curie, Toulouse, France
• 2 NeuroDiderot Department, Inserm UMR 1141, Robert-Debré Hospital, Université Paris Cité, Paris, France
• 3 Onco-Dermatology and Therapies, Inserm UMRS976, Hôpital Saint Louis, France Institut de Recherche Saint Louis, Université Paris Cité, Paris, France
• 4 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, United States
• 5 Akttyva Therapeutics, Inc., Mansfield, MA, United States

Finding new drugs usually consists of five main stages: 1) a pre-discovery stage in which basic research is performed to try to understand the mechanisms leading to diseases and propose possible targets (e.g., proteins); 2) the drug discovery stage, during which scientists search for molecules (two main large families, small molecules and biologics) or other therapeutic strategies that interfere or cure the investigated disease or at least alleviate the symptoms; 3) the preclinical development stage that focuses on clarifying the mode of action of the drug candidates, investigates potential toxicity, validates efficacy on various in vitro and in vivo models, and starts evaluate formulation; 4) the clinical stage that investigates the drug candidate in humans; 5) the reviewing, approval and post-market monitoring stage during which the drug is approved or not. In practice, finding new treatments is very challenging. Despite advances in the understanding of biological systems and the development of cutting-edge technologies, the process is still long, costly with a high attrition rate. New approaches, such as artificial intelligence and novel in vitro technologies, are being used in an attempt to rationalize R&D and bring new drugs to patients faster, but several obstacles remain. Our hope is that one day, it becomes possible to rapidly design inexpensive, more specific, more effective, non-toxic, and personalized drugs. This is a goal towards which all authors of this article have devoted most of their careers.

Introduction

Drug discovery has a long history and dates back to the early days of human civilization. In those ancient times, treatments were often discovered by chance or resulted from observation of nature, typically but not exclusively, using ingredients extracted from plants/animals, and not just used for physical remedy but also for spiritual healing. Modern drug discovery research started to being performed around the early 1900s. Nowadays, the development of a new medicine usually starts when basic research, often performed in academia, identifies a macromolecule (i.e., a molecule with a large molecular weight like genes/proteins), or a dysfunctional signaling pathway or a molecular mechanism apparently linked to a disease condition (pre-discovery stage) ( Figure 1 ; Table 1 ) ( Hefti, 2008 ; Hughes et al., 2011 ; Mohs and Greig, 2017 ; Villoutreix, 2021 ). In general, at this stage, research teams attempt to identify the so-called therapeutic targets (often a protein) that are linked to the disease state ( Gashaw et al., 2012 ). To be nominated therapeutic target, scientists will also have to find therapeutic agents that modify the function of the perturbed target and restore health or alleviate symptoms. Finding the right target is however extremely challenging. Further, drugs are efficient in humans because of specific actions on the intended therapeutic target but also due to interactions with other, unintended (often unknown) targets! The process continues with the search of therapeutic agents followed by a preclinical phase, during which potential drugs are tested in a battery of animal models, to demonstrate safety and select drug candidates (novel strategies to avoid animal testing are being developed, see below). Clinical studies in humans can then get started to establish safety and efficacy of the drugs in patients with the highest benefit-to-risk ratio ( Kandi and Vadakedath, 2023 ). The studies are then submitted to regulatory agencies, which review the documents and decide about market approval. If the review is positive, the drug can then be released to the market and be administrated to patients. Once a drug has been approved, investigations continue to monitor putative side effects that could be caused, over time, by the new treatment. This last step is often referred to as pharmacovigilance studies (or real-world evidence), generally dubbed “phase 4” clinical trial. The entire drug discovery and development process involves many disciplines, years of efforts and is very expensive. It also implies the generation and use of vast amount of data usually obtained via different types of high-throughput technologies. Many of these experiments and the analysis of the results can be automated via computer-assisted methods to speed-up some steps of the process, gain knowledge and reduce mistakes.

FIGURE 1 . Drug discovery and development. The main stages are represented in a highly simplified manner. The process varies depending on the molecular mechanisms expected to be linked to the disease and the type of therapeutic agents that needs to be developed. The approximate cost is around US $2.8 billion and the time needed to complete the entire process is around 12–15 years.

TABLE 1 . Glossary.

As mentioned above, to act on a disease, the problematic target(s) have to be modulated by a therapeutic agent (or several). There is a wide variety of agents that traditionally fits into two major classes, the so-called “small molecules” (small chemical compounds, some modified short peptides…) and the “biologics” (typically macromolecules such as recombinant proteins, antibodies, siRNAs, long peptides, cells, genes … and vaccines). There are major differences between biologics and small molecules ( Figure 2 ; Table 2 ) and we will essentially focus here on small molecules. It is also important to note that gene therapy is different from the other types of therapeutic agents because it is a technique that modifies a person’s genes to treat or cure a disease. In this case, the target is a disease-causing gene which has to be modified with a healthy copy of the gene, or the disease causing gene could be inactivated. Thus, beside technical issues, there are a number of ethical questions surrounding gene therapy and genome editing strategies that are not easy to answer. Further, some therapeutic agents are not acceptable to some parts of the population, as seen during the COVID-19 crisis and vaccine hesitancy. This is often due to misunderstanding of the biological processes and misinformation, resulting in fears, but yet this has to be considered. Also, about 5%–10% of the population are non-responders and have to receive other medications than vaccines. The division into small molecules and biologics is far from being perfect as some therapeutic agents combine a small molecule grafted onto a biologic (e.g., tisotumab vedotin is an antibody-drug conjugate used to treat cervical cancer). Therapeutic agents can be administrated to patients via different routes, called “routes of administration”. Small molecules can in general be administrated orally (the most convenient route for patients), while biologics usually need to be injected. The choice of a route of administration is also governed by the patient’s condition, for instance, in acute situations in hospitals, drugs are most often given intravenously. Other critical medical interventions that will not be discussed here are surgery, radiotherapy and psychological support.

FIGURE 2 . Small molecules, peptides and biologics. The properties and sizes of the therapeutic agents vary greatly. Three molecules are presented at the same scale, these involve rivaroxaban, a small chemical molecule used to treat thrombosis and pulmonary embolism, cyclosporine, a short immunosuppressive cyclic peptide (11 amino-acids, a biologic that still resembles to a certain extent to a small molecule) used to treat post-transplant organ rejection and a biologic, pembrolizumab (antibody, over 1300 amino-acids), used to treat various types of cancer.

TABLE 2 . General characteristics of small molecule drugs and biologics.

Drug discovery and development: overview of the process

There are several stages in the drug discovery process that require numerous skills and the use of various advanced technological platforms (often a combination of computational and experimental approaches) to validate targets and search for therapeutic agents. When initial experimental compounds have been sufficiently optimized to be selective, potent and safe in preliminary in vitro experiments and animal models, they can be nominated as drug candidates. At this stage, the project focus shifts from drug discovery to drug development to enable human clinical trials. If the therapeutic agent is successful in all three phases of the clinical trials, it goes through regulatory registration and the drug can be marketed ( Hefti, 2008 ; Hughes et al., 2011 ; Mohs and Greig, 2017 ).

Now, we will take a closer look at the process with the discovery of small molecules as an example. The process usually begins by focusing on a disease and the search of possible targets, often proteins, that can be modulated by small compounds ( Hughes et al., 2011 ) ( Figure 1 ). These compounds are expected to interfere or prevent the disease or at least limit the development of symptoms. These targets can be identified using cellular assays, genomic studies, proteomic studies, among many others. Then, thousands (to millions or even billions when using computer-aided drug design approaches prior to vitro assays) of small molecules have to be tested in various types of assays and a few promising molecules are then evaluated in animal models (and in alternative in vitro models) of human diseases. It is worth mentioning here that animal models can be misleading (e.g., a drug found toxic in animal models may not be toxic to humans or the opposite) ( Pognan et al., 2023 ). At the same time, absorption, distribution and elimination studies (ADME) are conducted. After years of research, a few compounds will hopefully be safe and effective enough to take forward to trials in patients. The different stages can have different names in the scientific literature, often they are referred to as: the pre-discovery and basic research stage (around 5–6 years) in which targets and modifying small molecules are searched in silico (i.e., using a computer), in vitro (i.e., in the test tube), ex vivo (e.g., on tissues or organs) and in vivo using simple animal models (i.e., in a living organism, typically rats or mice) and a preclinical stage (2–3 years) during which the best small molecules are selected using various in silico , in vitro and in vivo experiments. In general, after all these steps, only a few compounds progress to the next stage. Toxicity is investigated further on at least two animal models [one rodent (e.g., rat) and one non-rodent (e.g., dogs, mini-pigs)] often using different administration routes before they become nominated clinical candidates and get a regulatory permission to proceed to human clinical trials. Prior to starting clinical trials, a so-called Investigational New Drug (IND) application is submitted to regulatory agencies (e.g., the Food and Drug Administration in the United States). Such documents, at least up to now (see below), usually include animal efficacy data and toxicity (Good Laboratory Practice (GLP)-compliant animal toxicology data are performed supporting the dose, dosing schedule, administration), manufacturing information, clinical protocols (e.g., patient population, number of patients, duration of the study) proposed for the clinical trials and information about the investigators of the study.

If the IND is approved, then clinical trials start (4–7 years) ( Kandi and Vadakedath, 2023 ). In some specific cases such as cancer, a so-called phase 0 may get started, which involves the use of very small doses of the new drug in a limited number of people and sometimes in patients. This is an exploratory study with the goal of quickly exploring if and how the drug may work. In Phase I, the safety, and tolerability of the therapeutic agent (usually a single dose at first and then short-term multi-dose studies) is tested in a small number of healthy individuals (e.g., 20–80 people). Other parameters are investigated including the dose. Phase II typically involves 100–500 patients and the study can take place in several hospitals located in different countries. The study is designed to determine whether or not the therapeutic agent provides the desired therapeutic effect. Safety studies continue through the phase II trials. In the first part of phase II, referred to as phase IIa, the goal is to further refine the dose required to provide the desired therapeutic impact or monitored endpoints for the clinical candidate. Once the proper dose levels are determined, phase IIb studies can be initiated. The goal of the phase IIb is to determine the overall efficacy of the candidate drugs in a limited population of subjects. Numerous drug candidates fail in phase II due to safety issues or lack of efficacy. In phase III, the efficacy of the drug candidate is evaluated in a larger patient population. These studies are typically randomized and involve 1,000–5,000 patients at multiple clinical trial centers and are designed to determine the efficacy of the candidate compound relative to the current standard of care or a placebo, possible interactions with other medications and re-assess different doses (optimal dose is important for medication effectiveness). When neither the clinicians nor the patients know which of the treatments the patient is getting, the study is said to be double-blind. The cost and time associated with this phase can vary dramatically depending on the disease and the clinical endpoint under investigation. Phase III clinical trials are the most expensive part of drug discovery and development as it has a complex design and requires a large number of patients. Last but not least, formulation and stability studies are performed during the development stage to characterize the impurities present (either in batches or during storage conditions worldwide), and to determine the best formulation. Upon completion of the phase III trial, a New Drug Application (NDA) is submitted to the regulatory agencies to demonstrate drug safety and efficacy. Regulatory reviews can lead to requests for additional information, or even additional clinical trials to further establish either safety or efficacy. Ideally, these reviews lead to regulatory approval, including labelling requirements, and approval to market (review and approval ∼1–2 years). For approval, the drug must have adequate pharmaceutical quality, therapeutic effectiveness, and safety. It has to have a favorable “risk-benefit ratio”. Drugs offering important advances in treatment of a condition are given priority. Approval of regulatory bodies does not, however, signal the end of clinical trials. In many cases, regulatory agencies will require additional follow-up studies, often referred to as phase IV or post-marketing surveillance (“real-world evidence” trials) with infinite duration. In general, these studies are designed to detect rare adverse effects across a much larger population of patients or long-term adverse effects. The impact of phase IV studies can include alterations to labelling based on safety observations, contraindications for use of the new drug in combination with other medications, or even the withdrawal of marketing approval if the findings are severe enough.

Drug repurposing: challenges and opportunities

Drug repurposing or repositioning aims to take a drug (approved or in advanced clinical stages or even a drug that has been withdrawn from the market, most of the time it involves small molecules but biologics like antibodies are also explored), thus a molecule that has undergone extensive safety and efficacy testing, and use it for an additional or unrelated indication ( van den Berg et al., 2021 ; Roessler et al., 2021 ; Schipper et al., 2022 ). In some situations, even a withdrawn drug can be repurposed like thalidomide, originally intended as a sedative and then used for treating a wide range of other conditions, including morning sickness in pregnant women. Thalidomide was then withdrawn due to causing birth defects but then was approved to treat leprosy (in 1998) and multiple myeloma (in 2006) ( Begley et al., 2021 ). Drug repurposing approach can be very valuable in most cases including emergency situation like a pandemic, for rare and neglected diseases [for which specific drug developments are in general missing in pharmaceutical companies ( Scherman and Fetro, 2020 ; Roessler et al., 2021 )]. This strategy is promoted as a cost- and time-effective approach for providing novel medicines. It is often claimed that repurposing drugs can be faster, more economical, less risky, and carry higher success rates as compared to traditional approaches, primarily because it is in theory possible to bypass early stages of development such as establishing drug safety. Other benefits that come with this approach include readily available products and manufacturing supply chains. Drug repurposing can be very profitable as in the case of fenfluramine (in 2022, acquisition of Zogenix by UCB Pharma for about US$ 1.9 billion, https://www.ucb.com/stories-media/Press-Releases/article/UCB-Completes-Acquisition-of-Zogenix-Inc ), a drug initially developed for weight loss, withdrawn and now used in several countries for the treatment of some forms of epilepsy ( Odi et al., 2021 ). Yet, despite advantages, drug repurposing suffers from several issues. One problem is that there are no possibilities for optimization of the therapeutic molecule without losing the repurposing potential because any small change in the structure of the therapeutic agent means a new full manufacture process validation and preclinical safety development. Identifying an optimal dosage and formulation for the new disease indication can also be time consuming and requires novel investigations while side effects can indeed arise due to the new indication or in cases doses need to be changed. Also, assessing the patent status of the drug to repurpose requires very specific skills. The molecules that are investigated for repurposing are either patented or off-patent, and in some cases the intellectual property protection for the new indication may not be strong enough to engage in such project. Overall, while drug repurposing is intuitively attractive as it offers shorter routes to the clinic, challenges throughout the entire process are usually substantial. Investigating molecular mechanisms behind repurposing can however be very valuable as it can help identifying novel targets and as the repurposed drugs could be considered as starting point for the development of novel compounds (e.g., lenalidomide and pomalidomide are superior molecules derived from thalidomide) and as such emerge as breakthrough innovation in a reduced amount of time and still reduced cost compared to starting from scratch. It could also be of interest to combine several approved drugs (in some cases with a newer drug) to increase effectiveness.

Artificial intelligence: trust, but verify

Providing efficient and safe drug to patients is a long and complex process. The amount of data generated during this process or that can be collected from various sources is massive. It is thus necessary to integrate as much as possible quality data so as to be able to make decision in real time. Artificial Intelligence (AI or indeed, most of the time, machine learning) can definitely contribute here as it involves the use of powerful computers and efficient program algorithms to integrate large volume of data to train expert systems to perform a complex task ( Brogi and Calderone, 2021 ; Ruffolo et al., 2021 ; Jayatunga et al., 2022 ; Sadybekov and Katritch, 2023 ). During the early discovery phases, AI is used to rationalize processes, and to assist in project management (e.g., definition of a target product profile that allows to locate each compound with regard to the expected final drug specifications in a complex multi-dimensional space), to summarize information, to understand better complex biological systems (e.g., using for instance system biology and chemogenomics approaches), or to propose original compounds or biologics (e.g., small molecules, peptides) generated by the machine under various types of constraints (e.g., ADMET constraints or affinity to the target) ( Lambert, 2010 ; Gupta et al., 2021 ; Paul et al., 2021 ; Kontoyianni, 2022 ; Vijayan, et al., 2022 ). Most of the well-known success stories of AI have been in image recognition (e.g., in the early days, the approach was trained to for instance recognize cat and dog images, but today the method can be used to analyze biopsies or guide surgery) while also advertised in reducing time to reach phase I clinical trial. In the latter case, one can site the story of compound DSP-1181, developed by Exscientia and Sumitomo Dainippon Pharma, intended to treat obsessive compulsive disorder where time from first screening to the development stage was 4 time faster than using a conventional approach (although, unfortunately, the molecule failed in phase I, for numerous reasons including a difficult target while it was also observed that the molecules generated by AI were not novel) ( Santa Maria Jr et al., 2023 ) ( https://www.science.org/content/blog-post/another-ai-generated-drug ; https://www.cas.org/resources/cas-insights/drug-discovery/ai-designed-drug-candidates ). Similar observations have been posted by hundreds of financial analysts and research scientists about results obtained by other AI companies. In other words, the AI predictions are not perfect and indeed cannot be perfect at present ( Bajorath, 2021 ; Bender and Cortés-Ciriano, 2021 ). This situation reflects the dependency of AI/machine learning to quality, size and diversity of the data used to train the mathematical models. There are millions of compounds (most will never be a drug) tested via standard experiments available in various databases, but there are only a few thousand approved in humans that are annotated on which to learn from, highlighting the so-called data gap (i.e., there are billions of pictures of dogs and cats to learn from, but a limited amount of quality data is available in the field of drug discovery despite the use of numerous the high-throughput approaches). The predictions can thus be misleading, because we do not have enough quality data as input and/or because we do not understand enough the complexity of the biological systems ( Moingeon et al., 2022 ). During the drug development phases, in human, AI is associated to data-mining to for instance model some properties (e.g., PB/PK, PK/PD or population-based simulations and analysis, prediction of drug-drug interactions …). At this stage, these computer approaches can also be used to select the most informative population profile to be included in clinical trials or to explain the variability of effects, or provide « virtual » patients or populations, and applied to, for example, pediatric formulation using as input data collected on adults ( Lang et al., 2021 ). Related to these, the concept of digital twins (which has been around for a while in other areas of research), now starts to be explored in the context of drug discovery and development. The overall idea would be to collect data about a particular disease, how it progresses, about the current treatments, about specific patients, and about a whole population, encapsulate all these data into a computer model so as to create a digital representation of a biological system or of a person and be able to simulate, for example, what might happen if one were to take a novel drug. While the concept is attractive, there are still major challenges and obstacles ahead but progresses are being made ( An and Cockrell, 2022 ). Overall, AI, in the field of drug discovery and development, is still in the infancy stage and it will take time to fully integrate the technology into the R&D process ( Hillisch et al., 2015 ). AI-discovered drugs do not guarantee success in clinical trials. The understanding of the data used as well as the critical mind of the scientists are key points that lead to the success or failure of AI-assisted drug research and development processes. The technology, in some circumstances, can make the process faster and more cost-effective, however, AI needs quality data to produce meaningful results and still today requires significant experimental validation. As such, it is important to trust AI, but verify the predictions ( Schneider et al., 2020 ; Bajorath, 2021 ).

Rising cost: from drug discovery to new treatments

Analyses across all therapeutic areas indicate that the development of a new medicine, from target identification through approval for marketing, takes around 12–15 years and often longer. The cost to develop a new drug is very high, in part because failure is endemic in drug discovery, and success is rare. While various numbers have been reported, the latest formal assessment is around US $2.8 billion ( DiMasi, 2020 ). There are many factors that contribute to this situation: the lack of understanding of what causes the disease can lead to the selection of the wrong therapeutic target; the impossibility of reaching the target with a sufficient concentration of drug in vivo without leading to adverse effects; no formulation compatible with the use of the drug in human; the therapeutic agent developed during years is found in phase III to have very low efficacy; the therapeutic agents or a metabolite (e.g., case of a small molecule) can interacts, specifically or not, with other drugs or with hundreds of molecules in the body, these interactions are usually not known in details and can lead to numerous adverse effects; animal experiments that are used to evaluate potency, selectivity, and toxicity during the different stages of the process can be highly misleading; stricter regulatory guidelines; duration of patents; the identified therapeutic molecule can be toxic in some patients but this could not be anticipated during the clinical trials due to the relatively small number of patients treated. Next and related to the cost of R&D, comes the cost of the treatments. Although there is a very complex protocol to determine the price tag of a drug (it varies from country to country, it can consider the insurance system, whether the drug is curative and represents a major advance to both patients and the health system or it has a minor effect on the disease), but in the end, biologics are generally much more expensive than small molecules, in part due to the complex manufacturing process. Studies suggest that on average, the daily dose of biologics costs 22 times more than a small molecule ( Makurvet, 2021 ). It is important to keep in mind that the healthcare systems, in many countries, are about to collapse and that about half of the world population cannot get access to basic treatments ( Ozawa et al., 2019 ). Biologics have been here for several decades already and are becoming increasingly important in several therapeutic areas. For example, cancer checkpoint inhibitors (e.g., the antibody ipilimumab and about 4–5 others at the time of writing) have received considerable and broad interest because of their ability to generate responses in many hitherto intractable malignant tumors. Yet, many recent studies suggest that such molecules lead to responses in less than 10%–15% of patients with cancer. Clearly, such molecules offer hope but also rise many questions ( Fojo et al., 2014 ; Kantarjian and Rajkumar, 2015 ). That is, in some cases, biologics are real innovative breakthroughs, but in other situations, the strategy is pursued only for commercial reasons and alternative molecules such as small molecules are not even considered. These questions are, in theory, investigated by regulatory agencies [The United States Food and Drug Administration (FDA), European Medicines Agency (EMA), Pharmaceuticals and Medical Devices Agency (PMDA)] so as to try to avoid speculative drugs but more transparent processes would certainly be beneficial to patients and the general population. Although finding new treatments is very difficult, it is a profitable market, with global drug sales expected to grow to US$ 1.9 trillion by 2027 ( Mullard, 2023 ).

Innovation in regulatory science and methodologies

It is important to note that, in step with the scientific progress in human tissue models research in the past decades, in the US, new medicines may not have to be tested in animals, according to legislation signed by the President Joe Biden in late December 2022 (“Text–S.5002–117th Congress (2021–2022): FDA Modernization Act 2.0.” 29 September 2022. https://www.congress.gov/bill/117th-congress/senate-bill/5002/text ). Accordingly, US FDA is already accepting data from in vitro studies as part of the formal submission to the Agency ( Wadman, 2023 ). Additionally, at the same time, following the leadership of some academic researchers (e.g., Guzelian et al., 2005 ; Hoffmann and Hartung, 2006 ), major European and US agencies started using evidence-based methodologies, such as systematic reviews and systematic maps, in toxicological assessment. These methodologies were developed and tested over the last 40 years in clinical research, spearheaded by Cochrane Collaboration ( www.Cochrane.org ) to compare the effectiveness of treatments, and have been applied to toxicological assessment of data-rich substances by the European Food Safety Authority ( EFSA, 2017 ) and US Environmental Protection Agency’s (US EPA, https://cfpub.epa.gov/ncea/iris_drafts/recordisplay.cfm?deid=356370 ). While some of the aspects of these methodologies are not entirely applicable to drug-discovery because of the proprietary nature of the work, the main principles of evidence-based approaches, which encourage pre-publishing the methodologies before the research is conducted, comprehensiveness and transparency in data selection, minimization of bias (or systematic error), are in line with basic principles of the scientific method, and are applicable to drug discovery. Programs and drug candidates are all too frequently selected based on a biased opinion of a few scientists who are bound by similar training, scientific methodologies and beliefs. Opening-up drug discovery to scrutiny by other scientists with different training and opinions may lead to more failures in the earlier discovery stage, but less failures in the clinic, resulting in enhanced efficiency and more successes, benefiting the patients who need new treatments, first and foremost.

Concluding remarks

Drug discovery and development is a long and difficult endeavor; all novel ideas and strategies that can improve the process are valuable to explore. It is interesting to note that despite the steady increase in research and development expenditure, and major scientific advances in proteomics and genomics, the discovery of new drugs either seems to be drying-up some years or to remain essentially stable ( Laermann-Nguyen and Backfisch, 2021 ). This situation has various origins (e.g., many diseases with no treatment are extremely difficult to study), while, certainly, industry scientists would benefit from greater exposure to new ideas from public research and public researchers would benefit from the private sector to move beyond exploration of molecular mechanisms towards the end goal of efficient development of candidate therapeutic agents. Along these lines, some countries like the United States and United Kingdom have been working extensively at improving academic drug discovery (e.g., all the skills and platforms connected via open research networks with rational protocols) but in the others, the process is fragmented (no coordination, no intent, duplication of efforts and inefficient investments …) and, thus, not capable of producing desired results compared to the time, energy and money spent. A first step could be to develop strong academic drug discovery networks in countries where this type of activity is not coordinated or not considered. Strong collaborations between the private sector, academia and not-for-profit institutions are clearly of major importance and have led to some successes in the past but such partnerships can be difficult to maintain over a long period of time ( Yildirim et al., 2016 ; Takebe et al., 2018 ). The rationale being that open interconnections between the different scientific disciplines involved in drug research allow a “cross-fertilization”, each of them benefiting from the advances of the other fields. Obviously, such collaborations tend to be easier when academic and private research teams are located on the same campus, with possibilities of sharing ideas or technologies. Other types of collaboration imply building consortia, often for around 4–5 years, with research teams located in different cities or countries (unfortunately, most of the time, when the consortia have been built, they function as closed systems not allowing new scientists or novel research teams to join). Therefore, novel strategies need to be pursued, and among the novel public-private models that are being investigated, open science partnerships, could be of interest, if correctly implemented (e.g., the system must be open to all interested scientists, teams and relevant disciplines) ( Gold and Edwards, 2022 ). Open science projects ( Chodera et al., 2020 ), like the consortia models discussed above, are built on the differential expertise of the various partners, with generally academic and governmental partners taking on a larger role in the earlier stages and big pharmas leading in the later stages (e.g., advanced preclinical investigations, product development, manufacturing, and distribution). But in open science projects, results, publications, data, tools, and materials are open without regard for intellectual property. At some points, the various partners are free to use the results and develop their own proprietary products if deemed appropriate.

Next, novel technologies including AI could be a game changer in the years to come, even more so once we get past the hype stage. Novel approaches to replace animal models by more efficient, ethical, human-biology-based in vitro approaches could also play a significant role this next decade. Indeed, new tools and understanding, in, for instance, the area of investigative toxicology, are continually being implemented to reduce safety-related attrition in drug development ( Aleo et al., 2020 ). Combining all these strategies, methods and know-how should definitively facilitate the design of more specific, effective, non-toxic, and patient-tailored drugs, thereby, providing a more optimistic outlook to the field. As a last note, we encourage the general public and patients to become more curious about the process of finding novel therapies, from the pre-discovery to the post-marketing stages. Further, crowd-funded citizen science initiatives are emerging in various areas of drug discovery and development (e.g., https://www.clinicaltrialsarena.com/news/citizen-science-as-an-open-trials-tool-for-post-marketing-and-drug-repurposing-5909331-2/; see also the CTSA program at NIH ), these projects are definitively valuable to the field.

Author contributions

NS, PV, and BV conceptualized the topic and drafted the first version of the manuscript. All authors contributed to the article and approved the submitted version.

Support from the INSERM institute is appreciated. This article is based upon work supported by the National Science Foundation under Grant No. 2136307.

Acknowledgments

We thank Stephane Auvin, Epileptologist and Child Neurologist, Head of the Pediatric Neurology Department, Robert Debré Hospital, for interesting discussions about the use of fenfluramine for the treatment of some forms of epilepsy.

Conflict of interest

Authors KT and BV were employed by the company Akttyva Therapeutics, Inc. Author NS is employed by Evotech Se.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Aleo, M. D., Shah, F., Allen, S., Barton, H. A., Costales, C., Lazzaro, S., et al. (2020). Moving beyond binary predictions of human drug-induced liver injury (DILI) toward contrasting relative risk potential. Chem. Res. Toxicol. 33, 223–238. doi:10.1021/acs.chemrestox.9b00262

PubMed Abstract | CrossRef Full Text | Google Scholar

An, G., and Cockrell, C. (2022). Drug development digital twins for drug discovery, testing and repurposing: A schema for requirements and development. Front. Syst. Biol. 2, 928387. doi:10.3389/fsysb.2022.928387

Bajorath, J. (2021). State-of-the-art of artificial intelligence in medicinal chemistry. Future Sci. oa. 7, FSO702. doi:10.2144/fsoa-2021-0030

Begley, C. G., Ashton, M., Baell, J., Bettess, M., Brown, M. P., Carter, B., et al. (2021). Drug repurposing: Misconceptions, challenges, and opportunities for academic researchers. Sci. Transl. Med. 13, eabd5524. doi:10.1126/scitranslmed.abd5524

Bender, A., and Cortés-Ciriano, I. (2021). Artificial intelligence in drug discovery: What is realistic, what are illusions? Drug Discov. Today. 26, 511–524. doi:10.1016/j.drudis.2020.12.009

Brogi, S., and Calderone, V. (2021). Artificial intelligence in translational medicine. Int. J. Transl. Med. 1, 223–285. doi:10.3390/ijtm1030016

CrossRef Full Text | Google Scholar

Chodera, J., Lee, A. A., London, N., and von Delft, F. (2020). Crowdsourcing drug discovery for pandemics. Nat. Chem. 12, 581. doi:10.1038/s41557-020-0496-2

DiMasi, J. A. (2020). Research and development costs of new drugs. JAMA 324, 517. doi:10.1001/jama.2020.8648

EFSA (2017). Protocol for a systematic review on health outcomes related to the age of introduction of complementary food for the scientific assessment of the appropriate age of introduction of complementary feeding into an infant's diet. EFSA J. 15, e04969. doi:10.2903/j.efsa.2017.4969

Fojo, T., Mailankody, S., and Lo, A. (2014). Unintended consequences of expensive cancer therapeutics—the pursuit of marginal indications and a me-too mentality that stifles innovation and creativity: The john conley lecture. JAMA Otolaryngol. Head. Neck Surg. 140, 1225–1236. doi:10.1001/jamaoto.2014.1570

Gashaw, I., Ellinghaus, P., Sommer, A., and Asadullah, K. (2012). What makes a good drug target? Drug Discov. Today Suppl, S24–S30. doi:10.1016/j.drudis.2011.12.008

Gold, E. R., and Edwards, A. M. (2022). Overcoming market failures in pandemic drug discovery through open science: A Canadian solution. Front. Drug. Discov. 2. doi:10.3389/fddsv.2022.898654

Gupta, R., Srivastava, D., Sahu, M., Tiwari, S., Ambasta, R., and Kumar, P. (2021). Artificial intelligence to deep learning: Machine intelligence approach for drug discovery. Mol. Divers. 25, 1315–1360. doi:10.1007/s11030-021-10217-3

Guzelian, P. S., Victoroff, M. S., Halmes, N. C., James, R. C., and Guzelian, C. P. (2005). Evidence-based toxicology: A comprehensive framework for causation. Hum. Exp. Toxicol. 24, 161–201. doi:10.1191/0960327105ht517oa

Hefti, F. F. (2008). Requirements for a lead compound to become a clinical candidate. BMC Neurosci. 9, S7. doi:10.1186/1471-2202-9-S3-S7

Hillisch, A., Heinrich, N., and Wild, H. (2015). Computational chemistry in the pharmaceutical industry: From childhood to adolescence. ChemMedChem 10, 1958–1962. doi:10.1002/cmdc.201500346

Hoffmann, S., and Hartung, T. (2006). Toward an evidence-based toxicology. Hum. Exp. Toxicol. 25, 497–513. doi:10.1191/0960327106het648oa

Hughes, J. P., Rees, S., Kalindjian, S. B., and Philpott, K. L. (2011). Principles of early drug discovery. Br. J. Pharmacol. 162, 1239–1249. doi:10.1111/j.1476-5381.2010.01127.x

Jayatunga, M. K. P., Xie, W., Ruder, L., Schulze, U., and Meier, C. (2022). AI in small-molecule drug discovery: A coming wave? Nat. Rev. Drug Discov. 21, 175–176. doi:10.1038/d41573-022-00025-1

Kandi, V., and Vadakedath, S. (2023). Clinical trials and clinical research: A comprehensive review. Cureus 15, e35077. doi:10.7759/cureus.35077

Kantarjian, H., and Rajkumar, S. V. (2015). Why are cancer drugs so expensive in the United States, and what are the solutions? Mayo Clin. Proc. 90, 500–504. doi:10.1016/j.mayocp.2015.01.014

Kontoyianni, M. (2022). Library size in virtual screening: Is it truly a number's game? Expert Opin. Drug Discov. 17, 1177–1179. doi:10.1080/17460441.2022.2130244

Laermann-Nguyen, U., and Backfisch, M. (2021). Innovation crisis in the pharmaceutical industry? A survey. SN Bus. Econ. 1, 164. doi:10.1007/s43546-021-00163-5

Lambert, W. (2010). Considerations in developing a target product profile for parenteral pharmaceutical products. AAPS Pharm. Sci. Tech. 11, 1476–1481. doi:10.1208/s12249-010-9521-x

Lang, J., Vincent, L., Chenel, M., Ogungbenro, K., and Galetin, A. (2021). Impact of hepatic CYP3A4 ontogeny functions on drug-drug interaction risk in pediatric physiologically-based pharmacokinetic/pharmacodynamic modeling: Critical literature review and ivabradine case study. Clin. Pharmacol. Ther. 109, 1618–1630. doi:10.1002/cpt.2134

Makurvet, F. D. (2021). Biologics vs. small molecules: Drug costs and patient access. Med. Drug Discov. 9, 100075. doi:10.1016/j.medidd.2020.100075

Mohs, R. C., and Greig, N. H. (2017). Drug discovery and development: Role of basic biological research. Alzheimers Dement. (NY) 3, 651–657. doi:10.1016/j.trci.2017.10.005

Moingeon, P., Kuenemann, M., and Guedj, M. (2022). Artificial intelligence-enhanced drug design and development: Toward a computational precision medicine. Drug Discov. Today. 27, 215–222. doi:10.1016/j.drudis.2021.09.006

Mullard, A. (2023). Drug sales to reach $1.9 trillion within 5 years? Nat. Rev. Drug. Discov. 22, 172. doi:10.1038/d41573-023-00026-8

Odi, R., Invernizzi, R. W., Gallily, T., Bialer, M., and Perucca, E. (2021). Fenfluramine repurposing from weight loss to epilepsy: What we do and do not know. Pharmacol. Ther. 226, 107866. doi:10.1016/j.pharmthera.2021.107866

Ozawa, S., Shankar, R., Leopold, C., and Orubu, S. (2019). Access to medicines through health systems in low- and middle-income countries. Health Policy Plan. 34, iii1–iii3. doi:10.1093/heapol/czz119

Paul, D., Sanap, G., Shenoy, S., Kalyane, D., Kalia, K., and Tekade, R. (2021). Artificial intelligence in drug discovery and development. Drug Discov. Today. 26, 80–93. doi:10.1016/j.drudis.2020.10.010

Pognan, F., Beilmann, M., Boonen, H. C. M., Czich, A., Dear, G., Hewitt, P., et al. (2023). The evolving role of investigative toxicology in the pharmaceutical industry. Nat. Rev. Drug. Discov. 1, 317–335. doi:10.1038/s41573-022-00633-x

Roessler, H. I., Knoers, N. V. A. M., van Haelst, M. M., and van Haaften, G. (2021). Drug repurposing for rare diseases. Trends Pharmacol. Sci. 42, 255–267. doi:10.1016/j.tips.2021.01.003

Ruffolo, J. A., Sulam, J., and Gray, J. J. (2021). Antibody structure prediction using interpretable deep learning. Patterns (N Y). 3, 100406. doi:10.1016/j.patter.2021.100406

Sadybekov, A. V., and Katritch, V. (2023). Computational approaches streamlining drug discovery. Nature 616, 673–685. doi:10.1038/s41586-023-05905-z

Santa Maria, J. P., Wang, Y., and Camargo, L. M. (2023). Perspective on the challenges and opportunities of accelerating drug discovery with artificial intelligence. Front. Bioinform. 3, 1121591. doi:10.3389/fbinf.2023.1121591

Scherman, D., and Fetro, C. (2020). Drug repositioning for rare diseases: Knowledge-based success stories. Therapie 75, 161–167. doi:10.1016/j.therap.2020.02.007

Schipper, L. J., Zeverijn, L. J., Garnett, M. J., and Voest, E. E. (2022). Can drug repurposing accelerate precision oncology? Cancer Discov. 12, 1634–1641. doi:10.1158/2159-8290.CD-21-0612

Schneider, P., Walters, W. P., Plowright, A. T., Sieroka, N., Listgarten, J., Goodnow, R. A., et al. (2020). Rethinking drug design in the artificial intelligence era. Nat. Rev. Drug Discov. 19, 353–364. doi:10.1038/s41573-019-0050-3

Takebe, T., Imai, R., and Ono, S. (2018). The current status of drug discovery and development as originated in United States academia: The influence of industrial and academic collaboration on drug discovery and development. Clin. Transl. Sci. 11, 597–606. doi:10.1111/cts.12577

van den Berg, S., de Visser, S., Leufkens, H. G. M., and Hollak, C. E. M. (2021). Drug repurposing for rare diseases: A role for academia. Front. Pharmacol. 12, 746987. doi:10.3389/fphar.2021.746987

Vijayan, R. S. K., Kihlberg, J., Cross, J. B., and Poongavanam, V. (2022). Enhancing preclinical drug discovery with artificial intelligence. Drug Discov. Today. 27, 967–984. doi:10.1016/j.drudis.2021.11.023

Villoutreix, B., O. (2021). Post-pandemic drug discovery and development: Facing present and future challenges. Front. Drug. Discov. 1. doi:10.3389/fddsv.2021.728469

Wadman, M. (2023). FDA no longer has to require animal testing for new drugs. Science 379, 127–128. doi:10.1126/science.adg6276

Yildirim, O., Gottwald, M., Schüler, P., and Michel, M. C. (2016). Opportunities and challenges for drug development: Public-private partnerships, adaptive designs and big data. Front. Pharmacol. 7, 461. doi:10.3389/fphar.2016.00461

Keywords: drug discovery, drug development, therapeutic agent, biologics, small molecules, artificial intelligence (AI)

Citation: Singh N, Vayer P, Tanwar S, Poyet J-L, Tsaioun K and Villoutreix BO (2023) Drug discovery and development: introduction to the general public and patient groups. Front. Drug Discov. 3:1201419. doi: 10.3389/fddsv.2023.1201419

Received: 06 April 2023; Accepted: 08 May 2023; Published: 24 May 2023.

Reviewed by:

Copyright © 2023 Singh, Vayer, Tanwar, Poyet, Tsaioun and Villoutreix. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Bruno O. Villoutreix, [email protected]

† Present address: Natesh Singh, Evotec SE, Molecular Architects, Integrated Drug Discovery, Campus Curie, 195 Rte d’Espagne, 31100 Toulouse, France

This article is part of the Research Topic

Drug Discovery and Development Explained: Introductory Notes for the General Public

• Technical advance
• Open access
• Published: 14 November 2021

Advancing data science in drug development through an innovative computational framework for data sharing and statistical analysis

• Ann-Marie Mallon   ORCID: orcid.org/0000-0003-4047-4019 1 ,
• Dieter A. Häring 2 ,
• Frank Dahlke 2 ,
• Piet Aarden 2 ,
• Soroosh Afyouni 3 ,
• Daniel Delbarre 1 ,
• Khaled El Emam 4 ,
• Habib Ganjgahi 5 ,
• Stephen Gardiner 1 ,
• Chun Hei Kwok 1 ,
• Dominique M. West 1 ,
• Ewan Straiton 1 ,
• Sibylle Haemmerle 2 ,
• Adam Huffman 3 ,
• Tom Hofmann 2 ,
• Luke J. Kelly 3 , 5 ,
• Peter Krusche 2 ,
• Marie-Claude Laramee 2 ,
• Karine Lheritier 2 ,
• Greg Ligozio 6 ,
• Aimee Readie 6 ,
• Luis Santos 1 ,
• Thomas E. Nichols 3 ,
• Janice Branson 2 &
• Chris Holmes 5  

BMC Medical Research Methodology volume  21 , Article number:  250 ( 2021 ) Cite this article

6634 Accesses

7 Citations

3 Altmetric

Metrics details

Novartis and the University of Oxford’s Big Data Institute (BDI) have established a research alliance with the aim to improve health care and drug development by making it more efficient and targeted. Using a combination of the latest statistical machine learning technology with an innovative IT platform developed to manage large volumes of anonymised data from numerous data sources and types we plan to identify novel patterns with clinical relevance which cannot be detected by humans alone to identify phenotypes and early predictors of patient disease activity and progression.

The collaboration focuses on highly complex autoimmune diseases and develops a computational framework to assemble a research-ready dataset across numerous modalities. For the Multiple Sclerosis (MS) project, the collaboration has anonymised and integrated phase II to phase IV clinical and imaging trial data from ≈35,000 patients across all clinical phenotypes and collected in more than 2200 centres worldwide. For the “IL-17” project, the collaboration has anonymised and integrated clinical and imaging data from over 30 phase II and III Cosentyx clinical trials including more than 15,000 patients, suffering from four autoimmune disorders (Psoriasis, Axial Spondyloarthritis, Psoriatic arthritis (PsA) and Rheumatoid arthritis (RA)).

A fundamental component of successful data analysis and the collaborative development of novel machine learning methods on these rich data sets has been the construction of a research informatics framework that can capture the data at regular intervals where images could be anonymised and integrated with the de-identified clinical data, quality controlled and compiled into a research-ready relational database which would then be available to multi-disciplinary analysts. The collaborative development from a group of software developers, data wranglers, statisticians, clinicians, and domain scientists across both organisations has been key. This framework is innovative, as it facilitates collaborative data management and makes a complicated clinical trial data set from a pharmaceutical company available to academic researchers who become associated with the project.

Conclusions

An informatics framework has been developed to capture clinical trial data into a pipeline of anonymisation, quality control, data exploration, and subsequent integration into a database. Establishing this framework has been integral to the development of analytical tools.

Peer Review reports

Novartis and the University of Oxford’s Big Data Institute (BDI) have established a research alliance with the aim to improve health care and drug development by making it more efficient and targeted. Working with multi-disciplinary teams and using artificial intelligence (AI) and advanced analytics, the collaboration expects to transform how multiple large multidimensional (clinical, imaging, omics, biomarkers) datasets are combined, analysed and interpreted to identify phenotypes and early predictors of patient disease activity and progression, and to improve prognosis for patients. The collaboration currently focuses on two projects; one on Multiple Sclerosis (MS) and one on several autoimmune diseases treated with the -IL-17 antibody Cosentyx (secukinumab). The collaboration will also be making use of anonymised data from approximately 5 million patients from both the UK and international partner organisations, together with anonymised data captured from relevant Novartis clinical trials - in a total of approximately 50,000 patients. Using the BDI’s latest statistical machine learning technology and experience in data analysis, combined with Novartis’ clinical expertise and high-quality clinical trial data, the collaboration expects to better understand the underlying disease and early predictors of disease activity to improve prognosis for patients.

Here we describe the data anonymisation and the development of an innovative IT environment and AI technology, through which the alliance is working collaboratively to identify patterns, often across multiple data sources and types, which cannot be detected by humans alone. Novartis and the BDI expect to gain insights into the characteristics of specific, complex diseases and their pathways to understand what drives disease progression, and to understand commonalities between diseases.

A large-scale high-dimensional dataset

The collaboration currently focuses on Multiple Sclerosis (MS) and other inflammatory diseases in dermatology and rheumatology which are major areas of drug development. A computational framework and data management process have been established to facilitate the collaboration and enabling all data scientists to work together on the data. Throughout the next section we will describe the breadth and scope of the data contributed in both the Multiple Sclerosis and IL-17 projects, highlighting the variety of data types and modalities, the global nature of the data generation and the longitudinal aspects which increase the complexity of data management, curation and integration to enable us to produce research ready datasets of the highest utility to the statisticians and analysts.

Clinical trial data from patients

The core data used in the collaboration stem from Novartis clinical trials. All trials were conducted in accordance with the provisions of the International Conference on Harmonisation guidelines for Good Clinical Practice and the principles of the Declaration of Helsinki. The trial populations were defined by eligibility criteria (i.e. inclusion and exclusion criteria), all trial procedures followed trial protocols, specifying the purpose of the experiment, the medical objectives, the endpoints, the assessments and assessment frequencies, as well as the statistical analysis methods to address the key-objectives of the trials. All trial protocols were approved by an institutional review board or ethics committee and all patients or their legal representatives gave written informed consent before any trial-related procedures were performed. In general, data from clinical trials can only be used if the usage is covered by the informed consent. Usage beyond the informed consent defined scope requires the data to be anonymised in order that they are no longer personal data. To maintain data privacy, all data have been anonymised before use for analyses by the collaboration. The collaboration uses dose finding studies (phase II), confirmatory clinical trials (phase III) which are designed to confirm the efficacy and safety of a new treatment option versus a control treatment to seek regulatory approval for the new treatment, and phase IIIb and IV studies which are typically open-label studies after the approval of a drug. For illustration purposes, the design of a typical confirmatory (phase III) clinical trial is presented in Fig.  1 .

A schematic of a typical randomised clinical trial. A typical confirmatory phase III clinical trial, followed by an open-label extension study. During a screening period, and after signing informed consent, eligibility of the patient for the trial is assessed. Eligible patients may then be randomised to one or several test and control (placebo or active control) treatments. In a double-blind study, patients, physicians and study personnel are blinded to the patient’s treatment assignment until the core experiment has been completed and all assessments have been collected. Then the database is locked, the treatment allocation unblinded and the data analysed (core analysis). Often patients who complete a core study are offered to continue in an open-label study, for instance on the newly tested treatment until the new treatment option becomes available on the market. All assessments and study procedures are defined by a study protocol. EOS represents the end of the study

Multiple sclerosis (MS) project

Multiple Sclerosis (MS) is a chronic, immune-mediated disease of the central nervous system (CNS) characterized by inflammation, demyelination, and axonal/neuronal destruction, ultimately leading to severe disability. MS is the most common autoimmune demyelinating disorder of the CNS, affecting approximately 2.3 million individuals worldwide ( https://www.msif.org/about-us/who-we-are-and-what-we-do/advocacy/atlas/ ).

MS typically affects young adults (mean age at onset 30 years) and women are affected more often than men. Reflecting the current understanding of MS, the disease course of MS can be grouped into 2 corresponding main MS categories [ 1 ]:

relapsing MS (RMS): clinically isolated syndrome (CIS), relapsing-remitting MS (RRMS), active secondary progressive MS (SPMS)

progressive MS: secondary progressive Multiple Sclerosis (SPMS) and primary progressive MS (PPMS)

For Multiple Sclerosis (MS), the collaboration has integrated data from a total of approximately 35,000 MS patients collected in more than 2200 centres across 57 countries in 34 clinical trials (4 phase II trials of which 3 had extensions, 13 phase III trials of which 5 had extensions and 9 phase IV trials) which were conducted between years 2003 and now (2019 and ongoing). Over 88,000 patient-years of data is on record with individual patients being followed up for up to 15 years. Studies include randomised controlled clinical trials from three major drug development programs including patients from placebo- and active control arms, but also open-label and real-world studies which used other disease modifying therapies, in all phenotypes of MS patients:

Relapsing Remitting MS (N ~ 32,000)

Primary- and secondary progressive MS ( N  > 2800)

The database covers the entire spectrum of MS phenotypes including paediatric ( N  = 235), treatment naive MS patients ( N  = 5445), but also patients who had the disease for > 25 and up to 50 years ( N  = 1624).

The dataset also includes a wealth of different data modalities, including detailed information on demography and baseline characteristics, medical history, MS relapses (> 16,000 relapses recorded) with symptoms and recovery grades, physical disability assessments (> 238,000 neurological assessments of the Expanded Disability Status Scale [EDSS]) covering all levels of disability (from EDSS = 0 normal neurological exam with no disability to EDSS = 10 death due to MS), 25-ft walking test (walking ability), 9-hole peg test (hand coordination), cognitive assessments (PASAT, SDMT), laboratory values, ECGs, vital signs, concomitant medication, detailed treatment information, and adverse event data. More than 13,000 patients have clinical records that also include MRI summary features. For > 11,000 of these patients, we also have the raw MRI images available for analyses and feature extraction, totalling to > 230,000 MRI scans, with T1-weighted (pre and post gadolinium contrast enhancement), T2-weighted, proton density, FLAIR, magnetization transfer, and diffusion-weighted sequences, longitudinal data being available for up to 12 years in individual patients. The collation and anonymization of these raw MRI images (see results) has been established in this collaboration to ensure they can be integrated into the overall research ready database.

Interleukin-17 inhibitor project

Cosentyx (secukinumab) is a high-affinity recombinant, fully human monoclonal Interleukin-17A (IL-17A) antibody. By binding to human IL-17A, Cosentyx neutralizes the bioactivity of this cytokine. IL-17A is the central lymphokine of a defined subset of inflammatory T cells, which appear to be pivotal in several autoimmune and inflammatory processes. The collaboration has integrated data from a total of 16,576 randomised patients from over 21 phase II and 44 phase III Cosentyx clinical trials targeting four autoimmune disorders in dermatology and rheumatology. They are:

Psoriasis (PsO)

Axial spondyloarthritis (axSpA), including ankylosing spondylitis and non-radiographic axial spondyloarthritis

Psoriatic arthritis (PsA)

Rheumatoid arthritis (RA)

Each study records a range of measurements on each subject at multiple time points throughout the trial, with the duration of collection of patient data ranging from 12 weeks to 5 years. The dataset includes different data modalities ranging from demography and patient history to laboratory data and imaging. A number of assessments are collected as a standard in all clinical studies and are therefore available across indications, such as demography and baseline characteristics, medical history, ECGs, vital signs, concomitant medication, detailed treatment information, and adverse event data, Quality of Life questionnaires, and laboratory measurements from serum and whole blood samples. Genomic, proteomic, and transcriptomic data, as well as imaging data from MRI and X-ray scans are also available for some patients depending on the trial design. Overall, the diseases under study all have a commonality of inflammation in different regions of the body, therefore measurements of inflammation are a common assessment as well. Other datasets are collected in some of the indications due to common disease pattern. One such example is skin assessment, commonly done in PsO and PsA or joint assessments done in PsA, axSpA and RA. Additionally, MRI and X-ray images of affected body locations were taken. In axSpA, which manifests predominantly in the axial region, MRI and X-rays of the spine and sacroiliac joints are taken in order to monitor treatment response. In PsA, X-rays of the hands, wrists and feet are collected. This dataset will be used to model patient disease trajectories, as well as interpret and predict of multivariate longitudinal response to Cosentyx , in order to improve the clinical outcome of patients across the four autoimmune disorders.

A fundamental component for the successful data analysis and the collaborative development of novel machine learning methods on this rich data set has been the construction of an innovative research informatics framework that can capture the data at regular intervals from Novartis into a secure IT infrastructure at the BDI where data could be integrated, quality controlled and compiled in to a research ready relational database which would then be available to analysts. The development of this framework (Fig.  2 ) in the BDI and the successful capture of the data (described previously in the Methods) into a versioned research ready dataset is described below.

Innovative and robust informatics framework for high-dimensional clinical trial data. Raw clinical data is anonymised by Privacy Analytics, Inc. (PAI), followed by data wrangling that involves harmonisation and pooling of data. Data is then integrated into a relational database (DB) whereby users of the DB are able to obtain analysis-ready datasets through querying. Data within the DB can be used for data analysis and visualisation purposes

A secure and collaborative research infrastructure

A critical part of the research collaboration was to develop an IT infrastructure and corresponding information security architecture that was technologically feasible, business viable and foremost user desirable for the project. The design of the information security architecture and controls was governed by the need for the processing of a large amount of clinical data shared between two organisations. In order to design and implement proper information security, it was vital to understand the various stages in the process, the states of the data, the information security goals and the overall risks. After collecting and assessing all of this data, a custom information security architecture was defined and security controls and guidelines applied.

The overall results of the above methodology were that confidentiality was identified as the overall and most relevant information security goal within the project. This decision defined the fundamental principle that governed the design and implementation of the security controls.

Isolation from the existing Oxford BDI infrastructure whenever possible, at least for every activity involving non-anonymised data.

Encrypted data transfer via dedicated channels between both organisations to ensure confidentiality and integrity.

User access to the environment is proxied via a demilitarized zone (DMZ), which contains a certain set of jump hosts that serve as portals to the full environment.

Identity and access management is realized within the environment, providing authentication and authorisation services.

The log management is done at a central place, together with security information and event management.

Two separate environments were created for anonymisation and analytics work, and both of these were instantiated within a dedicated OpenStack private cloud, as two separate tenants. This ensures network, compute and storage isolation enforced at the hypervisor level. For data processing virtual clusters were created within each tenant, including an instance (virtual machine) with direct access to GPUs for accelerated work. The virtual resources within each tenant were defined and provisioned by means of Ansible roles and playbooks, for consistency and repeatability. Encrypted backups are made to an S3 object store. In short, we have produced a unique research computing infrastructure that provides high levels of security while providing a shared environment where both academic and industrial researchers can jointly work.

Clinical data anonymisation

This section describes the anonymisation of the clinical trial data and the specific methods developed to anonymise MRI data to ensure data privacy.

Clinical trial data – basic principles

The process for anonymising the clinical trial data was intended to ensure that the risk of re-identifying participants in the dataset was below a pre-defined critical threshold. There are three key concepts in this risk-based anonymisation approach:

The risk of re-identification can be measured quantitatively. Various models of adversaries and re-identification attacks have been developed and have demonstrated robustness in practice [ 2 ]. Metrics quantifying the probability of a successful re-identification have been developed based on these models. The specific metrics that we used are based on strict average risk models. These capture the average risk while ensuring that there are no population unique individuals in the anonymised dataset (i.e., in the context of the General Data Protection Regulation (GDPR), the likelihood of individuals being “singled out” is very small [ 3 ].

The overall risk measurement takes into account the context of data processing as is illustrated in Fig.  3 . For example, if the anonymised dataset will be analysed in a secure compared to a less secure environment then less modification of the data is required to bring the risk of re-identifying a patient to below the targeted threshold. Checklists have been developed and validated to capture this context risk [ 2 ].

A specific threshold needs to be defined to determine what an acceptably low risk is. There are many precedents for what is deemed to be an acceptable threshold, including from regulators (see the review in [ 2 ]). The choice of a specific threshold from the precedent range takes into account the sensitivity of the data and the potential harm if there is a re-identification.

Clinical Trial Data Anonymisation. The overall risk of re-identification is a function of both the data risk and the context risk. The context risk is assessed by examining three re-identification attacks on a dataset: (a) a deliberate attack by an adversary, (b) an inadvertent re-identification by a data analyst where they recognize someone they know, and (c) a data breach occurring. The success of the three attacks is affected by the controls that are in place. The context consists of first the contractual controls which reduce the context risk. The residual risk is managed by security and privacy controls, which are also part of the context. The extent of these controls reduces the overall risk further. Then any residual risk is managed by perturbing or transforming the data

Once a threshold is defined and the re-identification risk is computed, taking into account the context, transformation may be required until the risk is below the defined threshold. The transformations can be performed to the data itself (e.g. by modifying variables that may lead to re-identification such as a patient’s age) or to the context (e.g. by modifying the security of the IT system). After each transformation the overall risk can be re-computed until it is below the threshold.

Justification for threshold

The European Medicines Agency (EMA) has established a policy on the publication of clinical data for medicinal products [ 4 ] which requires applicants/sponsors to openly share clinical trial data. The guidelines accompanying the policy recommend a maximum risk threshold of 0.09. Health Canada implemented the same threshold for the sharing of clinical trial data [ 5 ]. This is the threshold that is used for the anonymization of the clinical data.

Calculation of risk

The risk of re-identification is calculated only on the quasi-identifiers. The quasi-identifiers are variables that are knowable by an adversary. There are two general types of quasi-identifiers. The first are those which are in the public domain and can be collected from registries such as voter registration lists [ 6 ] and lien registries [ 7 ]. Examples of these include date of birth and ZIP/postal codes. The second are acquaintance quasi-identifiers, which are known by adversaries who are also acquaintances, such as neighbours, relatives, and co-workers. Acquaintance quasi-identifiers include the public ones as well as things like medical history and key events and dates. Once the quasi-identifiers are determined in a dataset, the probability of re-identification can be calculated.

The calculation of re-identification risk considers three potential attacks on the data, which we shall call T1, T2, and T3.

The first attack, T1, assumes that an adversary deliberately attempts to re-identify individuals in the dataset [ 8 ]. This means that the probability of re-identification is conditional on an attempted attack:

The first term captures the risk in the data and the second term captures the risk from the context. There are multiple estimators that can be used to evaluate data risk which vary in accuracy and scalability [ 2 , 9 , 10 , 11 , 12 , 13 , 14 ].

Context risk has three components: security controls, privacy controls, and contractual controls. The strength of these controls as they were implemented at the BDI were assessed using a checklist. The checklist is reproduced elsewhere [ 2 ]. The responses to the checklist are converted into a conservative subjective probability. This means that the exact probability value is not known, but the modeled value is convincingly conservative (over estimates the context risk) but still allows us to model the controls that are in place and account for the benefits of stronger controls.

The premise of the controls for attack T1 is that the existence of stronger security controls (e.g., audit logs that are checked, analyst screening, and limited access), privacy controls (e.g., regular privacy training and a privacy officer), and contractual controls (e.g., all analysts have to sign a confidentiality agreement when working with the data) act as deterrents for an attempted attack and make it more difficult.

A T2 attack pertains to an inadvertent re-identification. This is when an analyst inadvertently or spontaneously recognizes someone that they know in the dataset as they are working on it. This type of risk is given by:

An inadvertent re-identification is contingent on an analyst knowing someone in the data. In our case this means that an analyst would know someone who has participated in a trial in this therapeutic area. This is estimated as: p ( acquaintance ) = 1 − (1 −  v ) 150 : where v is the proportion of patients in the current studies compared to all studies in this therapeutic area over the same period and geography, which can be computed by gathering target recruitment data from https://clinicaltrials.gov/ . The 150 value is the Dunbar number, which provides us with an estimate of the average number of individuals that an analyst would know. Dunbar’s has proven to be robust across multiple studies (for a literature review see [ 2 ]).

The third attack is when there is a data breach and the dataset is accessed by an adversary. This is modeled as follows:

The probability of a breach is computed from published reports on health data breaches and their likelihood that are produced on a regular basis by security companies.

After computing the risk values for the three types of attack, the maximum across them is then taken to reflect the overall risk in the data. If this maximum risk is below the 0.09 threshold, then the dataset is deemed to have an acceptably low risk of re-identification. The same approach is applied to analyse the risk in clinical trial data and in the header information in DICOM files.

Strict average risk

The risk calculation described above gives us the average risk (averaged across all patients). The strict average conditions this on no records in the dataset being unique in the population. The population is defined as all patients who have participated in clinical trials in the same therapeutic area over the same period and geography. There are a number of estimators that can be used for estimating population uniqueness, with a specific one recommended based on a comparative assessment [ 15 ].

Application of the basic principles

The basic principles have been operationalized for the anonymisation of clinical trial data as a series of default anonymisation practices, which can then be adjusted to account for study-specific data issues. Patient identifiers (typically consisting of a clinical centre number and the patient’s randomization number with which a patient is identified in a clinical trial) is replaced by an anonymised identifier, a new number specifically and uniquely generated for the use of the data in the context of the collaboration. The link file that connects the original patient identifier from the trial with the new anonymised identifier is securely protected and only accessible to a very small independent team who are working exclusively on the anonymisation of the data but who are not otherwise involved in the collaboration or the subsequent research. The link file is used for the sole purpose of assigning the same anonymised patient ID to both the patient’s clinical data and MRI images, so that the corresponding imaging and clinical data remain together after the anonymization is completed, for downstream analyses. By default, event dates in the dataset are offset into relative dates as defined in the PhUSE standard [ 16 ]. Also, variables like age are typically generalized to, for example, five year ranges, or modified by adding uniform noise. The SiteID is suppressed so that the geographic location of a site cannot be determined by looking up recruitment information in public registries. Other variables that may contain information that could lead to the re-identification, such as a patient’s medical history, can be generalised or suppressed. The decision as to which variables are transformed takes the intended research purpose into account to preserve the data as much as possible where critical for the research while still bringing the risk of re-identification below the defined threshold. A detailed report is produced documenting the anonymisation methodology, how it was operationalised for each dataset, and a summary of the anonymisation outcomes (e.g., which variables were transformed and how). This detailed report is crucial for the data wrangling and downstream analysis. All data in the final relational database could be linked to the report which described the steps taken.

Magnetic resonance imaging (MRI) data

In clinical trials MRI images are commonly obtained at hospitals in Digital Imaging and Communications (DICOM) format and provided to Clinical Research Organisations (CROs) who specialise in imaging analysis. In the MS project the DICOM images were transferred from the CROs to the isolated anonymisation computational environment. Within this environment the image data went into a three-stage process of image conversion, defacing and data curation. Each DICOM file represents one slice of the entire scan; one MRI session will generate multiple scans. The DICOMs were re-assembled into a single 3D volume (as per the research needs), using the DICOM conversion software HeuDiConv [ 17 ]. The resulting output of this process is a set of files in a different format (JSON – JavaScript Object Notation, and NIfTI – Neuroimaging Informatics Technology Initiative) that exactly preserve original DICOM data values as well as their associated non-identifying DICOM metadata (i.e. meta-data that could contribute to the identification of patients was stripped out during conversion), but organised in a research ready format, developed and used extensively within the Neuroimaging research community, called Brain Imaging Data Structure (BIDS - https://bids.neuroimaging.io/ ). In addition to a controlled, standardised file-structure, BIDS provides a file naming convention with the same characteristics, including adding scan-type details for ease of processing. During the initial conversion process, scans that failed to convert, or converted with errors, were put aside and evaluated to see if they could be successfully converted. Overall, we were able to convert scans for over 99% of subjects.

Once the data have been converted to NIfTI and is in BIDS format, they were run through a processing pipeline, simply called ‘defacing’. This pipeline has several steps that aim to achieve two key objectives:

Remove identifiable facial features (nose, mouth, front of the eyes, ears)

Remove identifiable metadata from the scan’s associated JSON

For privacy/security reasons the identifiable facial features were removed (defaced). The facial identifiable elements were selected according to the anonymisation principles used in the UK BioBank project [ 18 ], and removed using defacing software from the FSL software library [ 19 ] . To ensure the successful anonymisation, all defacing results are visually checked via multiple 3D surface renderings, confirming the removal of facial features and the retention of brain and meninges. Scans were QC checked and classified as either ‘passed’, or as one of four subclasses of defacing issues. Due to this being a multi-site, longitudinal dataset, MRI scans were of variable quality, and initial defacing failure rates were up to 40% in some studies. Scans that failed QC checks were put through additional rounds of re-defacing and subsequent QC checks, where custom defacing parameters – derived from the type of previous QC classifications and scan modality – were applied to the scans, allowing us to achieve high rates of successful defacing (96%). In total over 230,000 MRIs were defaced and manually checked before entering the research ready dataset. The anonymised data is stripped of all metadata except non-identifiable acquisition parameters. Additional checks were also undertaken to ensure that identifiable details had not been erroneously inserted (during acquisition) into the retained metadata fields. Once all QC checks have been completed, this data is copied via a dedicated and automated mechanism from the anonymisation environment to the analytics environment. Additional safeguards have been implemented to ensure confidentiality and integrity of the data.

Data exploration, quality control and integration – research-ready dataset

Non-imaging clinical datasets are anonymised by a third party (Privacy Analytics, Inc) and transferred to the analytics environment at the BDI to begin the data wrangling process. This process has the ultimate goal of providing all data in a relational database, from where streamlined, research ready datasets for the analytics team can be retrieved. A detailed tracking system was developed that is shared across the collaboration to transparently convey the status of each data set within the pipeline. Due to the large number of steps, transformations, and transactions that each dataset goes through, it is essential to track each data point received.

The initial stage in the extract, transform and load (ETL) of data from Novartis to the BDI, was the capture of the clinical trial data as Statistical Analysis Software (SAS) files. For both the MS and IL-17 projects, the BDI team worked closely with the clinical teams at Novartis to ensure full understanding of the data to be downloaded and all related documentation. Each study was downloaded separately in an average of 30 to 50 different tables. These tables contained the primary raw data and study-specific information as described in methods. Each table contained hundreds of thousands of measurements, across hundreds of variables.

Once the data was received, the next critical collaborative step in the process was the data exploration by a dedicated data wrangler to validate that the data received matched the expected data and the protocol documents. This step was extensive and performed in collaboration with the team at Novartis who could review identified queries by exploring the primary data which has not been anonymised. At this stage the tracking of data and related queries between the data generators (Novartis) and the BDI was paramount to ensure all downstream analysis would be reproducible and linkable to a well-defined set of data.

Once the data was agreed to be correct and valid, the data wrangling team developed codebooks (data specifications) which described the structure of the data in a computationally readable format. These codebooks are then utilised by a bespoke software pipeline to homogenise the data into a generic relational database structure. This task had many challenges due to varying trial designs, inconsistencies in data capture between trials, changes in technology throughout history, subjective evaluations, changes in data standards, and anonymisation. Upon successful completion of this part of the pipeline individual data files were imported into the relational database. In conjunction with the import of clinical trial data, the pipeline imports metadata about the relevant additional datasets provided e.g. imaging or omics. The key remit for this level of integration is to ensure that relevant data slices can be provided downstream to the analysts and also to ensure the data outputs from the analytics can be integrated back into the overall architecture. This work was deemed to be important, as it was setup in a way that it can be reproduced as new datasets come onboard and was developed in a manner to manage any clinical trial data not just the current data from this project.

Once the data dictionaries were completed across the projects, the data wrangling team began the overall data quality control process in parallel with the aim of identifying any data quality issues through the data life cycle. The data validation and QC process are innovative as they have been created in a sustainable manner to ensure data is tracked and checked throughout the lifetime of the project and that data provenance is managed at the level of individual data points. The quality of the data from Novartis to the analytical teams was deemed critically important, to ensure data analysts did not have to perform this level of exploratory work and that the results they identified were reproducible. A robust QC pipeline was therefore developed through intensive collaboration that assessed the data at many levels. The quality control pipeline was developed to perform both validation and verification at different stages. For example, source validation ensured the data received matched what was expected and global validation checked the merged data. Validation and verification encompass a large list of checks, from structural checks, assessing levels of missingness, the effects of the anonymisation, and visual checks for potential data anomalies. The anonymisation reports were key to check whether data was missing because it was not captured in the first place, or if it was suppressed due to anonymisation.

The final output of the ETL process was to generate snapshots of data as tracked and versioned data releases for the analysts. A data release is a snapshot of merged datasets available as a relational database or in a data format that can be inputted to analytical tools (e.g. API, table structure etc.). The analytical teams can therefor develop methods which are attributed to the correct version ensuring transparency and reproducibility. As the data analytical methods are being developed, the ETL pipeline is being expanded to ensure that data outputted from new methods can be integrated back into data releases. This is key in a data project.

Novartis and the University of Oxford’s Big Data Institute (BDI) have established a research alliance which has developed an innovative IT platform to manage large volumes of anonymised data. The IT infrastructure that has been developed for this project has enabled the alliance to successfully capture, anonymise, quality control, integrate, and explore data from a large collection of Novartis clinical trials in one research ready environment. This research ready database is now available to a highly multi-disciplinary team of researchers who are analysing and interpreting the data to gain insights about the diseases. The data integrated in this project has not been compiled before, and therefore the technology developed here is allowing data analysts an unprecedented opportunity to develop methods to gain insights across different data modalities (imaging, omics, clinical and biological) and to identify novel patterns with clinical relevance which cannot be detected by humans alone to identify phenotypes, and early predictors of patient disease activity and progression and to improve prognosis for patients. The collaboration currently focuses on Multiple Sclerosis (MS) and other inflammatory diseases in dermatology and rheumatology which are major areas of drug development, but may extend the scope to other disease areas at a later time point.

A milestone achievement of the collaboration is the development of a data anonymisation pipeline for multi-modality data (including clinical and imaging data) which ensures data privacy while preserving the essential clinically relevant pattern in the data. Data anonymization at this scale has the potential to create datasets which are unusable for analysis, so a key step in this project was that specific adjustments were incorporated into the overall anonymisation process that ensured the analytical questions could be addressed. A fundamental component of successful data analysis and the collaborative development of novel machine learning methods on this rich data sets has been the construction of a research informatics framework that can capture the data at regular intervals from Novartis into an IT infrastructure at the Big Data Institute (BDI) where images could be anonymised and integrated with the de-identified clinical data, quality controlled and compiled into a research-ready relational database which would then be available to multi-disciplinary analysts. The collaborative development from a group of software developers, data wranglers, statisticians, clinicians and domain scientists across both organisations has been key. The project has proactively engaged a number of external academic researchers in a number of fields to work with the consortium, get access to the data and contribute to the overall strategic vision. This framework is innovative, as it facilitates collaborative data management and makes a complicated clinical trial data set from a pharmaceutical company available to academic researchers in a secure, granular and robust way. The level of data tracking and data provenance incorporated will ensuring reproducibility and transparency.

The research alliance has developed an informatics framework to capture multi-dimensional clinical trial data into a pipeline of anonymisation, quality control, data exploration and subsequent integration into a research-ready database. With an emphasis on ensuring data privacy while allowing the development of analytical tools to be conducted, the framework can extend to research on other disease areas, and its principles can be transversally applied into other data settings, especially ones with data privacy concerns.

Availability of data and materials

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

Abbreviations

Oxford’s Big Data Institute

Multiple sclerosis

Psoriatic arthritis

Rheumatoid arthritis

Axial spondyloarthritis

Central nervous system

Relapsing multiple sclerosis

Clinically isolated syndrome

Relapsing-remitting multiple sclerosis

Secondary progressive multiple sclerosis

Primary progressive multiple sclerosis

Expanded Disability Status Scale

Fluid-attenuated inversion recovery

Magnetic Resonance Imaging

Demilitarized zone

General Data Protection Regulation

European Medicines Agency

Digital Imaging and Communications Format

Clinical Research Organisations

JavaScript Object Notation

Neuroimaging Informatics Technology Initiations

Brain Imaging Data Structure

Extract, Transform and Load

Lublin, et al. Defining the clinical course of multiple sclerosis The 2013 Revisions. Neurology. 2014;83(3):278–86.

Article   Google Scholar  

El Emam K. Guide to the De-Identification of Personal Health Information. CRC Press (Auerbach), 2013.

Article 29 Data Protection Working Party. Opinion 05/2014 on Anonymization Techniques. (2014).

European Medicines Agency. European Medicines Agency policy on publication of data for medicinal products for human use: Policy 0070.”= Oct. 02, 2014, [Online]. Available: http://www.ema.europa.eu/docs/en_GB/document_library/Other/2014/10/WC500174796.pdf .

Health Canada. Guidance document on public release of clinical information, Apr. 01, 2019. https://www.canada.ca/en/health-canada/services/drug-health-product-review-approval/profile-public-release-clinical-information-guidance.html .

Benitez K, Malin B. Evaluating re-identification risks with respect to the HIPAA privacy rule. J Am Med Inform Assoc. 2010;17(2):169–77. https://doi.org/10.1136/jamia.2009.000026 .

Article   PubMed   PubMed Central   Google Scholar  

El Emam K et al. Pan-Canadian De-identification guidelines for personal health information. 2007. http://www.ehealthinformation.ca/documents/OPCReportv11.pdf .

Google Scholar  

Marsh C, et al. The case for samples of anonymized records from the 1991 census. Journal of the Royal Statistical Society, Series A (Statistics in Society). 1991;154(2):305–40.

Article   CAS   Google Scholar  

Hundepool A, et al. Statistical Disclosure Control: Wiley; 2012.

Book   Google Scholar  

Hundepool A, et al. Handbook on statistical disclosure control: ESSNet SDC; 2010.

Duncan G, Elliot M, Salazar G. Statistical confidentiality - principles and practice: Springer; 2011.

Matthias Templ. Statistical disclosure control for microdata - methods and applications in R. Aug. 24, 2018. https://www.springer.com/us/book/9783319502700 . Accessed Aug. 24, 2018).

Willenborg L, de Waal T. Statistical disclosure control in practice. New York: Springer-Verlag; 1996.

Willenborg L, de Waal T. Elements of statistical disclosure control. New York: Springer-Verlag; 2001.

Dankar F, El Emam K, Neisa A, Roffey T. Estimating the re-identification risk of clinical data sets. BMC Medical Informatics and Decision Making. 2012;12:66.

PhUSE De-Identification Working Group . De- Identification Standards for CDISC SDTM 3.2. (2015).

Halchenko Y, Goncalves M, Visconti di Ollegio Catello M, Ghosh S, Hanke M, Dae, … , Carlin J (2019). https://github.com/nipy/heudiconv .

Alfaro-Almagro F, Jenkinson M, Bangerter NK, Andersson JLR, Griffanti L, Douaud G, et al. Image processing and quality control for the first 10,000 brain imaging datasets from UK Biobank. NeuroImage. 2018;166(April 2017):400–24. https://doi.org/10.1016/j.neuroimage.2017.10.034 .

Article   PubMed   Google Scholar  

Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. FSL. Neuroimage. 2012;62:782–90.

Download references

Acknowledgements

This collaboration was made possible with the subsequent access to clinical trial data from Novartis. We are grateful for the assistance and facilitation from Prof. Gil McVean for leading the project, Ms. Joanna Stoneham, Mr. Byron Jones and Dr. Anna Zalevski for managing the project, Prof. Lars Fugger, Prof. Mark Jenkinson and Dr. George Nicholson as advisors for data analysis. We are also very grateful to Mr. Henrik Westerberg and Mr. Aaron McCoy for their impactful contribution to key aspects of the MRI anonymisation pipeline; Dr. Bartek Papiez, Dr. Soroosh Afyouni, Dr. Angelos Armen, Dr. Brieuc Lehmann, Mr. Amit Knanna, Mr. Karan Rajesh, Mr. Gordon Graham, Mr. Frank Freischlaeger, Mr. Stephen Robertson and Mr. Costantino Catuogno for data preparation and input on analysis on the MS project; Mr. Shephard Mpofu, Mr. Brian O. Porter, Mr. Hanno Richards, Ms. Luminata Pricop, Ms. Ana de Vera, Ms. Yanli Chang, Ms. Stephanie Danetz, Mr. Michael Beste, Mr. Thibaud Coroller, Mr. Matthias Kormaksson, Ms. Tingting Zhuang, Ms. Xuan Zhu, Mr. Albert Widmer, Ms. Ruvie Martin and Mr. Chengeng Tian for data preparation and input on analysis on the IL17 project; Ms. Natasa Hadjistephanou for supporting data wrangling; Mr. Timor Kadir and Mr. Amir Jamaludin for input on spinal MRI and image segmentation; Mr. Robert Esnouf, Mr. Richard Urbanski, Mr. Tim O’Sullivan and Mr. Arnold Ingo for IT infrastructural and security support.

This paper is the output from the Novartis funded alliance with Oxford Big Data Institute. Novartis funded the design of the study and collection, analysis and interpretation of data, and in writing the manuscript.

Author information

Authors and affiliations.

MRC Harwell Institute, Harwell Campus, Oxfordshire, OX11 0RD, UK

Ann-Marie Mallon, Daniel Delbarre, Stephen Gardiner, Chun Hei Kwok, Dominique M. West, Ewan Straiton & Luis Santos

Novartis Pharma AG, Basel, Switzerland

Dieter A. Häring, Frank Dahlke, Piet Aarden, Sibylle Haemmerle, Tom Hofmann, Peter Krusche, Marie-Claude Laramee, Karine Lheritier & Janice Branson

Big Data Institute, University of Oxford Li Ka Shing Centre for Health Information and Discovery, Old Road Campus, Oxford, OX3 7LF, UK

Soroosh Afyouni, Adam Huffman, Luke J. Kelly & Thomas E. Nichols

Children’s Hospital of Eastern Ontario Research Institute, 401 Smyth Road, Ottawa, Ontario, K1J 8 L1, Canada

Khaled El Emam

Department of Statistics, University of Oxford, 24-29 St Giles’, OX1 3LB, Oxford, UK

Habib Ganjgahi, Luke J. Kelly & Chris Holmes

Novartis Pharma AG, East Hanover, NJ, USA

Greg Ligozio & Aimee Readie

You can also search for this author in PubMed   Google Scholar

Contributions

AMM, DH, SH, KE and TN lead the drafting of the paper, with all other authors helping to contribute to the drafting, reading and final approval of the manuscript. AMM, DD, SG, CHK, DMW, ES and LS contributed to data exploration, data wrangling, quality control, anonymisation of MRI images and development of databases. AH, TH and MCL performed the design, development and maintenance of the IT infrastructure and information security architecture. DH, FD, PA, SA, HG and PK performed data analysis on the MS project. SH, LK, GL and AR performed data analysis on the IL-17 project. KE lead the anonymisation of clinical data. TN, KL, JB and CH oversee and manage both projects within the research collaboration.

Corresponding author

Correspondence to Ann-Marie Mallon .

Ethics declarations

Ethics approval and consent to participate.

The core data used in the collaboration stem from Novartis clinical trials, approved by institutional review boards or ethics committees. All trials were conducted in accordance with the provisions of the International Conference on Harmonisation guidelines for Good Clinical Practice and the principles of the Declaration of Helsinki. No human data or analyses from human data are published in this article.

Consent for publication

Not applicable.

Competing interests

Some of the work for this article was performed while Khaled El Emam was with IQVIA where he led the Privacy Analytics business, before he joined the University of Ottawa / CHEO Research Institute. All other authors declare they have no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Mallon, AM., Häring, D.A., Dahlke, F. et al. Advancing data science in drug development through an innovative computational framework for data sharing and statistical analysis. BMC Med Res Methodol 21 , 250 (2021). https://doi.org/10.1186/s12874-021-01409-4

Download citation

Received : 27 July 2020

Accepted : 22 September 2021

Published : 14 November 2021

DOI : https://doi.org/10.1186/s12874-021-01409-4

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Machine learning
• Data management
• Data anonymisation
• Clinical trial

BMC Medical Research Methodology

ISSN: 1471-2288

• General enquiries: [email protected]

Advertisement

Machine Learning and Artificial Intelligence in Pharmaceutical Research and Development: a Review

• Review Article
• Theme: Celebrating Women in the Pharmaceutical Sciences
• Published: 04 January 2022
• Volume 24 , article number  19 , ( 2022 )

Cite this article

• Sheela Kolluri 1 ,
• Jianchang Lin 2 ,
• Rachael Liu 2 ,
• Yanwei Zhang 2 &
• Wenwen Zhang 2  

26k Accesses

70 Citations

44 Altmetric

Explore all metrics

Over the past decade, artificial intelligence (AI) and machine learning (ML) have become the breakthrough technology most anticipated to have a transformative effect on pharmaceutical research and development (R&D). This is partially driven by revolutionary advances in computational technology and the parallel dissipation of previous constraints to the collection/processing of large volumes of data. Meanwhile, the cost of bringing new drugs to market and to patients has become prohibitively expensive. Recognizing these headwinds, AI/ML techniques are appealing to the pharmaceutical industry due to their automated nature, predictive capabilities, and the consequent expected increase in efficiency. ML approaches have been used in drug discovery over the past 15–20 years with increasing sophistication. The most recent aspect of drug development where positive disruption from AI/ML is starting to occur, is in clinical trial design, conduct, and analysis. The COVID-19 pandemic may further accelerate utilization of AI/ML in clinical trials due to an increased reliance on digital technology in clinical trial conduct. As we move towards a world where there is a growing integration of AI/ML into R&D, it is critical to get past the related buzz-words and noise. It is equally important to recognize that the scientific method is not obsolete when making inferences about data. Doing so will help in separating hope from hype and lead to informed decision-making on the optimal use of AI/ML in drug development. This manuscript aims to demystify key concepts, present use-cases and finally offer insights and a balanced view on the optimal use of AI/ML methods in R&D.

Graphical abstract

Similar content being viewed by others

Artificial intelligence to deep learning: machine intelligence approach for drug discovery

Special FDA designations for drug development: orphan, fast track, accelerated approval, priority review, and breakthrough therapy

Deep learning in drug discovery: an integrative review and future challenges

Avoid common mistakes on your manuscript.

Introduction

Artificial intelligence (AI) and machine learning (ML) have flourished in the past decade, driven by revolutionary advances in computational technology. This has led to transformative improvements in the ability to collect and process large volumes of data. Meanwhile, the cost of bringing new drugs to market and to patients has become prohibitively expensive. In the remainder of this paper, we use “R&D” to generally describe the research, science, and processes associated with drug development, starting with drug discovery to clinical development and conduct, and finally the life-cycle management stage.

Developing a new drug is a long and expensive process with a low success rate as evidenced by the following estimates: average R&D investment is $1.3 billion per drug [ 1 ]; median development time for each drug ranges from 5.9 to 7.2 years for non-oncology and 13.1 years for oncology; and proportion of all drug-development programs that eventually lead to approval is 13.8% [ 2 ]. Recognizing these headwinds, AI/ML techniques are appealing to the drug-development industry, due to their automated nature, predictive capabilities, and the consequent expected increase in efficiency. There is clearly a need, from a patient and a business perspective, to make drug development more efficient and thereby reduce cost, shorten the development time and increase the probability of success (POS). ML methods have been used in drug discovery for the past 15–20 years with increasing sophistication. The most recent aspect of drug development where a positive disruption from AI/ML is starting to occur, is in clinical trial design, operations, and analysis. The COVID-19 pandemic may further accelerate utilization of AI/ML in clinical trials due to increased reliance on digital technology in patient data collection. With this paper, we attempt a general review of the current status of AI/ML in drug development and also present new areas where there might be potential for a significant impact. We hope that this paper will offer a balanced perspective, help in separating hope from hype, and finally inform and promote the optimal use of AI/ML.

We begin with an overview of the basic concepts and terminology related to AI/ML. We then attempt to provide insights on when, where, and how AI/ML techniques can be optimally used in R&D, highlighting clinical trial data analysis where we compare it to traditional inference-based statistical approaches. This is followed by a summary of the current status of AI/ML in R&D with use-case illustrations including ongoing efforts in clinical trial operations. Finally, we present future perspectives and challenges.

AI And ML: Key Concepts And Terminology

In this section, we present an overview of key concepts and terminology related to AI and ML and their interdependency (see Fig.  1  and Table I ). AI is a technique used to create systems with human-like behavior. ML is an application of AI, where AI is achieved by using algorithms that are trained with data. Deep learning (DL) is a type of ML vaguely inspired by the structure of the human brain, referred to as artificial neural networks.

Chronology of AI and ML

Human intelligence is related to the ability of the human brain to observe, understand, and react to an ever-changing external environment. The field of AI not only tries to understand how the human brain works but also tries to build intelligent systems that can react to an ever-changing external environment in a safe and effective way (see Fig. 2 for a brief overview of AI [ 3 ]). Researchers have pursued different versions of AI by focusing on either fidelity to human behavior or rationality (doing the right thing) in both thought and action. Subfields of AI can be either general focusing on perception, learning, reasoning, or specific such as playing chess. A multitude of disciplines have contributed to the creation of AI technology, including philosophy, mathematics, and neuroscience. ML, an application of AI, uses statistical methods to find patterns in data, where data can be text, images, or anything that is digitally stored. ML methods are typically classified as supervised learning, unsupervised learning, and reinforcement learning. (See Fig.  3  for a brief overview of supervised and unsupervised learning.)

Brief overview of AI

Brief overview of supervised and unsupervised learning

Current Status

AI/ML techniques have the potential to increase the likelihood of success in drug development by bringing significant improvements in multiple areas of R&D including: novel target identification, understanding of target-disease associations, drug candidate selection, protein structure predictions, molecular compound design and optimization, understanding of disease mechanisms, development of new prognostic and predictive biomarkers, biometrics data analysis from wearable devices, imaging, precision medicine, and more recently clinical trial design, conduct, and analysis. The impact of the COVID-19 pandemic on clinical trial execution will potentially accelerate the use of AI and ML in clinical trial execution due to an increased reliance on digital technology for data collection and site monitoring.

In the pre-clinical space, natural language processing (NLP) is used to help extract scientific insights from biomedical literature, unstructured electronic medical records (EMR) and insurance claims to ultimately help identify novel targets; predictive modeling is used to predict protein structures and facilitate molecular compound design and optimization for enabling selection of drug candidates with a higher probability of success. The increasing volume of high-dimensional data from genomics, imaging, and the use of digital wearable devices, has led to rapid advancements in ML methods to handle the “Large p, Small n” problem where the number of variables (“p”) is greater than the number of samples (“n”). Such methods also offer benefits to research in the post-marketing stage with the use of “big data” from real-world data sources to (i) enrich the understanding of a drug’s benefit-risk profile; (ii) better understand treatment sequence patterns; and (iii) identify subgroups of patients who may benefit more from one treatment compared with others (precision medicine).

While AI/ML have been widely used in drug discovery, translational research and the pre-clinical phase with increasing sophistication over the past two decades, their utilization in clinical trial operations and data analysis has been slower. We use “clinical trial operations” to refer to the processes involved in the execution and conduct of the clinical trials, including site selection, patient recruitment, trial monitoring, and data collection. Clinical trial data analysis refers to data management, statistical programming, and statistical analysis of participant clinical data collected from a trial. On the trial operations end, patient recruitment has been particularly challenging with an estimated 80% of trials not meeting enrollment timelines and approximately 30% of phase 3 trials terminating early due to enrollment challenges [ 4 ]. Trial site monitoring (involving in-person travel to sites) is an important and expensive quality control step mandated by regulators. Furthermore, with multi-center global trials, clinical trial monitoring has become labor-intensive, time-consuming, and costly. In addition, the duration from the “last subject last visit” trial milestone for the last phase 3 trial to the submission of the data package for regulatory approval, has been largely unchanged over the past two decades and presents a huge opportunity for positive disruption by AI/ML. Shortening this duration will have a dramatic impact on our ability to get drugs to patients faster while reducing cost. The steps in-between include cleaning and locking the trial database, generating the last phase 3 trial analysis results (frequently involving hundreds of summary tables, data listings, and figures), writing the clinical study report, completing the integrated summary of efficacy and safety, and finally creation of the data submission package. The impact of COVID-19 may further accelerate the push to integrate AI/ML into clinical trial operations due to an increased push toward 100% or partially virtual (or “decentralized”) trials and the increased use of digital technology to collect patient data. AI/ML methods can be used to enhance patient recruitment and project enrollment and also to allow real-time automated and “smart” monitoring for clinical data quality and trial site performance monitoring. We believe AI/ML hold potential to have a transformative effect on clinical trial operations and clinical trial data analyses particularly in the areas of trial data analysis, creation of clinical study reports, and regulatory submission data packages.

Case Studies

Below, we offer a few use cases to illustrate how AI/ML methods have been used or are in the process of improving existing approaches in R&D.

Case Study 1 (Drug Discovery)—DL for Protein Structure Prediction and Drug Repurposing

A protein’s biological mechanism is determined by its three-dimensional (3D) structure that is encoded in its one-dimensional (1D) string of amino acid sequence. Knowledge about protein structures is applied to understand their biological mechanisms and help discover new therapies that can inhibit or activate the proteins to treat target diseases. Protein misfolding has been known to be important in many diseases, including type II diabetes, as well as neurodegenerative diseases such as Alzheimer’s, Parkinson’s, Huntington’s, and amyotrophic lateral sclerosis [ 5 ]. Given the knowledge gap between a proteins’1D string of amino acid sequence and its 3D structure, there is significant value in developing methods that can accurately predict 3D protein structures to assist new drug discovery and an understanding of protein-folding diseases. AlphaFold [ 6 , 7 ] developed by DeepMind (Google) is an AI network used to determine a protein’s 3D shape based on its amino acid sequence. It applied a DL approach to predict the structure of the protein using its sequence. The central component of AlphaFold is a convolutional neural network that was trained on the Protein Data Bank structures to predict the distances between every pair of residues in a protein sequence, giving a probabilistic estimate of a 64 × 64 region of the distance map. These regions are then tiled together to produce distance predictions for the entire protein for generating the protein structure that conforms to the distance predictions. In 2020, AlphaFold released the structure predictions of five understudied SARS-CoV-2 targets including SARS-CoV-2 membrane protein, Nsp2, Nsp4, Nsp6, and Papain-like proteinase (C terminal domain), which will hopefully deepen the understanding of under-studied biological systems [ 8 ].

Beck et al. [ 9 ] developed a deep learning–based drug-target interaction prediction model, called Molecule Transformer-Drug Target Interaction (MT-DTI), to predict binding affinities based on chemical sequences and amino acid sequences of a target protein, without their structural information, which can be used to identify potent FDA-approved drugs that may inhibit the functions of SARS-CoV-2’s core proteins. Beck et al. computationally identified several known antiviral drugs, such as atazanavir, remdesivir, efavirenz, ritonavir, and dolutegravir, which are predicted to show an inhibitory potency against SARS-CoV-2 3C–like proteinase and can be potentially repurposed as candidate treatments of SARS-CoV-2 infection in clinical trials.

Case Study 2 (Translational Research/Precision Medicine)—Machine Learning for Developing Predictive Biomarkers

Several successful case studies have now been published to show that the biomarkers derived by the ML predictive models were used to stratify patients in clinical development. Predictive models were developed [ 10 ] to test whether the models derived from cell line screen data could be used to predict patient response to erlotinib (treatment for non-small cell lung cancer and pancreatic cancer) and sorafenib (treatment for kidney, liver, and thyroid cancer), respectively. The predictive models have IC50 values as dependent variables and gene expression data from untreated cells as independent variables. The whole-cell line panel was used as the training dataset and the gene expression data generated from tumor samples of patients treated with the same drug was used as the testing dataset. No information from the testing dataset was used in training the drug sensitivity predictive models. The BATTLE clinical trial data was used as an independent testing dataset to evaluate the performance of the drug sensitivity predictive models trained by cell line data. The best models were selected and used to predict IC50s that define the model-predicted drug-sensitive and drug-resistant groups.

Li et al. [ 10 ] applied the predictive model to stratify patients in the erlotinib arm from the BATTLE trial. The median progression-free survival (PFS) for the model-predicted erlotinib-sensitive patient group was 3.84 months while the PFS for model-predicted erlotinib-resistant patients was 1.84 months, which suggests that the erlotinib-sensitive patients predicted by the model had more than doubled PFS benefit relative to erlotinib-resistant patients. Similarly, the model-predicted sorafenib-sensitive group had a median PFS benefit of 2.66 months over the sorafenib-resistant group with a p -value of 0.006 and a hazard ratio of 0.32 (95%CI, 0.15 to 0.72). The median PFS was 4.53 and 1.87 months, for model-predicted sorafenib-sensitive and model-predicted sorafenib-resistant groups, respectively.

Case Study 3—Nonparametric Bayesian Learning for Clinical Trial Design and Analysis

Many of the existing ML methods are focused on learning a set of parameters within a class of models using the appropriate training data, which is often referred to as model selection. However, an important issue encountered in practice is the potential model over-fitting or under-fitting, as well as the discovery of an underlying data structure and related causes [ 11 ]. Examples include but are not limited to the following: selecting the number of clusters in clustering problem, the number of hidden states in a hidden Markov model, the number of latent variables in a latent variable model, or the complexity of features used in nonlinear regression. Thus, it is important to appropriately train ML methods to perform reliably under real-world conditions with trustworthy predictions. Cross-validation is commonly used as an efficient way to evaluate how well the ML methods perform in the selection of tuning parameters.

Nonparametric Bayesian learning has emerged as a powerful tool in modern ML framework due to its flexibility, providing a Bayesian framework for model selection using a nonparametric approach. More specifically, a Bayesian nonparametric model allows us to use an infinite-dimensional parameter space and involve only a finite subset of the available parameters on the given sample set. Among them, the Dirichlet process is currently a commonly used Bayesian nonparametric model, particularly in Dirichlet process mixture models (also known as infinite mixture models). Dirichlet process mixtures provide a nonparametric approach to model densities and identify latent clusters within the observed variables without pre-specification of the number of components in a mixture model. With advances in Markov Chain Monte Carlo (MCMC) techniques, sampling from infinite mixtures can be done directly or using finite truncations.

There are many applications of such Bayesian nonparametric models in clinical trial design. For example, in oncology dose-finding clinical trials, nonparametric Bayesian learning can offer efficient and effective dose selection. In oncology first in human trials, it is common to enroll patients with multiple types of cancers which causes heterogeneity. Such issues can be more prominent in immuno-oncology and cell therapies. Designs that ignore the heterogeneity of safety or efficacy profiles across various tumor types could lead to imprecise dose selection and inefficient identification of future target populations. Li et al. [ 12 ] proposed nonparametric Bayesian learning–based designs for adaptive dose finding with multiple populations. These designs based on the Bayesian logistic regression model (BLRM) allow data-driven borrowing of information, across multiple populations, while accounting for heterogeneity, thus improving the efficiency of the dose search and also the accuracy of estimation of the optimal dose level. Liu et al. [ 13 ] extended another commonly used dose-finding design, modified toxicity probability interval (mTPI) designs to BNP-mTPI and fBNP-mTPI, by utilizing Bayesian nonparametric learning across different indications. These designs use the Dirichlet process, which is more flexible in prior approximation, and can automatically group patients into similar clusters based on the learning from the emerging data.

Nonparametric Bayesian learning can also be applied in master protocols including basket, umbrella, and platform trials, which allow investigation of multiple therapies, multiple diseases, or both within a single trial [ 14 , 15 , 16 ]. With the use of nonparametric Bayesian learning, these trials have an enhanced potential to accelerate the generation of efficacy and safety data through adaptive decision-making. This can affect a reduction in the drug development timeline in an area of significant unmet medical need. For example, in the evaluation of potential COVID-19 therapies, adaptive platform trials have quickly emerged as a critical tool, e.g ., the clinical benefits of remdesivir and dexamethasone have been demonstrated using such approaches in the Adaptive COVID-19 Treatment Trial (ACTT) and the RECOVERY [ 17 ] trial.

One of the key questions in master protocols is whether borrowing across various treatments or indications is appropriate. For example, ideally, each tumor subtype in a basket trial should be tested separately; however, it is often infeasible given the rare genetic mutations. There is potential bias due to the small sample size and variability as well as the inflated type I error if there is a naïve pooling of subgroup information. Different Bayesian hierarchical models (BHMs) have been developed to overcome the limitation of using either independent testing or naïve pooling approaches, e.g ., Bayesian hierarchical mixture model (BHMM) and exchangeability-non-exchangeability (EXNEX) model. However, all these models are highly dependent on the pre-specified mixture parameters. When there is limited prior information on the heterogeneity across different disease subtypes, the misspecification of parameters can be a concern. To overcome the potential limitation of existing parametric borrowing methods, Bayesian nonparametric learning is emerging as a powerful tool to allow flexible shrinkage modeling for heterogeneity between individual subgroups and for automatically capturing the additional clustering. Bunn et al. [ 18 ] show that such models require fewer assumptions than other more commonly used methods and allow more reliable data-driven decision-making in basket trials. Hupf et al. [ 19 ] further extend these flexible Bayesian borrowing strategies to incorporate historical or real-world data.

Case Study 4—Precision Medicine with Machine Learning

Based on recent estimates, among phase 3 trials with novel therapeutics, 54% failed in clinical development, with 57% of those failures due to inadequate efficacy [ 20 ]. A major contributing factor is failure in identification of the appropriate target patient population with the right dose regimen including the right dose levels and combination partners. Thus, precision medicine has become a priority in pharmaceutical industry for drug development. One approach could be a systematic model utilizing ML applied to (a) build a probabilistic model to predict probability of success; and (b) identify subgroups of patients with a higher probability of therapeutic benefit. This will enable the optimal match of patients with the right therapy and maximize the resources and patient benefit. The training datasets can include all ongoing early-phase data, published data, and real-world evidence but are limited to the same class of drugs.

One major challenge to establish the probabilistic model is defining endpoints that can best measure therapeutic effect. Early-phase clinical trials (particularly in oncology) frequently adopt different primary efficacy endpoints compared with confirmatory pivotal trials due to a relatively shorter follow-up time and need for faster decision-making. For example, common oncology endpoints are overall response rate or complete response rate in phase I/II and progression-free survival (PFS) and/or overall survival (both measure long-term benefit) in pivotal phase III trials. In oncology, it is also common that phase I/II trials use single-arm settings to establish the proof of concept and generate the hypothesis of treatment benefit, while in pivotal trials, especially in randomized phase III trial with a control arm, the purpose is to demonstrate superior treatment benefit over available therapy. This change in the targeted endpoints from the early phase to late phase makes the prediction of POS in the pivotal trial, using early-phase data, quite challenging. Training datasets using previous trials for drugs with a similar mechanism and/or indications can help establish the relationship between the short-term endpoints and long-term endpoints, which ultimately determines the success of drug development.

Additionally, the clustering of patients can be done using unsupervised learning. For example, nonparametric Bayesian hierarchical models using the Dirichlet process enables patient grouping (without pre-specified number of clusters) with key predictive or prognostic factors, to represent various levels of treatment benefit. This DL approach will bring efficiency in patient selection for precision medicine clinical development.

Case Study 5—AI/ML-assisted Tool for Clinical Trial Oversight

Monitoring of trials by a sponsor is a critical quality control measure mandated by regulators to ensure the scientific integrity of trials and safety of subjects. With increasing complexity of data collection (increased volume, variety, and velocity), and the use of contract research organizations (CROs)/vendors, sponsor oversight of trial site performance and trial clinical data has become challenging, time-consuming, and extremely expensive. Across all study phases (excluding estimated site overhead costs and costs for sponsors to monitor the study), trial site monitoring is among the top three cost drivers of clinical trial expenditures (9–14% of total cost) [ 21 ].

For monitoring of trial site performance, risk-based monitoring (RBM) has recently emerged as a potential cost-saving and efficient alternative to traditional monitoring (where sponsors sent study monitors to visit sites for 100% source-data verification (SDV) according to a pre-specified schedule). While RBM improves on traditional monitoring, inconsistent RBM approaches used by CROs and the current prospective nature of the operational/clinical trial data reviews—has meant that sponsor’s ability to detect critical issues with site performance, may be delayed or compromised (particularly in lean organizations where CRO oversight is challenging due to limited resources).

For monitoring of trial data quality, current commonly used approaches largely rely on review of traditional subject and/or aggregate data listings and summary statistics based on known risk factors. The lack of real-time data and widely available automated tools limit the sponsor’s ability for prospective risk mitigation. This delayed review can have a significant impact on the outcome of a trial, e.g ., in an acute setting where the primary endpoint uses ePRO data—monthly transfers may be too late to prevent incomplete or incorrect data entry. The larger impact is a systemic gap in study team oversight that could result in critical data quality issues.

One potential solution is the use of AI/ML-assisted tools for monitoring trial site performance and trial data quality. Such a tool could offer an umbrella framework, overlaid on top of the CRO systems, for monitoring trial data quality and sites. With the assistance of AI/ML, study teams may be able to use an advanced form of RBM (improved prediction of risk and thresholds for signal detection) and real-time clinical data monitoring with increased efficiency/quality and reduced cost in a lean resourced environment. Such a tool could apply ML and predictive analytics to current RBM and data quality monitoring—effectively moving current study monitoring to the next generation of RBM. The use of accumulating data from the ongoing trial and available data from similar trials, to continuously improve on the data quality and site performance checks, could have a transformative effect on sponsor’s ability to protect patient safety, reduce trial duration, and trial cost.

In terms of data quality reviews, data fields, and components contributing to the key endpoints that impact the outcome of the trial would be identified by the study team. For trial data monitoring, an AI/ML-assisted tool can make use of predictive analytics and R Shiny visualization for cross-database checks and real-time “smart monitoring” of clinical data quality. By “smart monitoring,” we mean the use of AI/ML techniques that continuously learn from accumulating trial data and improve on the data quality checks, including edit checks. Similarly, for trial site performance, monitoring an AI/ML tool could begin with the Transcelerate (a non-profit cross-pharma consortium) library of key risk indicators (KRIs) and team-specified thresholds to identify problem sites based on operational data. In addition, the “smart” feature of an AI/ML tool could use accumulating data to continuously improve on the precision of the targeted site monitoring that makes up RBM. The authors of this manuscript are currently collaborating with a research team at MIT to advance research in Bayesian probabilistic programming approaches that could aid the development of an AI/ML tool with the features described above for clinical trial oversight of trial data quality and trial site performance.

AI/ML as a field has tremendous growth potential in R&D. As with most technological advances, this presents both challenges and hope. With modern-day data collection, the magnitude and dimensionality of data will continue to increase dramatically because of the use of digital technology. This will increase the opportunities for AI/ML techniques to deepen understanding of biological systems, to repurpose drugs for new indications, and also to inform study design and analysis of clinical trials in drug development.

Although the development of recent ML/AI methods represents major technological advances, the conclusions made could be misleading if we are not able to tease out the confounding factors, use reliable algorithms, look at the right data, and fully understand the clinical questions behind the endpoints and data collection. It is imperative to train ML algorithms properly to have trustworthy performance in practice using various data scenarios. Additionally, not every research question can be answered utilizing AI/ML, particularly if there is high variability, limited data, poor quality of the data collection, under-represented patient populations, or flawed trial design. The issue of under-represented patient populations is particularly concerning as it could lead to a systematic bias. Furthermore, in line with the emerging concerns in other spaces where AI/ML have been used, care and caution needs to be exercised to address patient privacy and bioethical considerations.

It is also important to be aware when DL/AI vs . ML vs . traditional inference-based statistical methods are most effective in R&D. In Fig.  4 below, we attempt to provide a recommendation based on the dimensionality of the dataset. In Fig.  5 , we attempt to provide a similar recommendation, this time based on different aspects of drug development. Although many ML algorithms are able to handle high-dimensional data with the “Large p, Small n” problem, the increased number of variables/predictors, especially those not related to the response, continues to be a challenge. As the number of irrelevant variables/predictors increases, the volume of the noise becomes greater, resulting in the reduced predictive performance of most ML algorithms.

Application of ML/AI based on the dimensionality of the data

Application of ML/AI based on various aspects of drug development

In summary, a combination of appropriate understanding of both R&D and advanced ML/AI techniques can offer huge benefits to drug development and patients. The implementation and visualization of AI/ML tools can offer user-friendly platforms to maximize efficiency and promote the use of breakthrough techniques in R&D. However, a sound understanding of the difference between causation and correlation is vital, as is the recognition that the evolution of sophisticated prediction capabilities does not render the scientific method to be obsolete. Credible inference still requires sound statistical judgment and this is particularly critical in drug development, given the direct impact on patient health and safety. This further underscores that a well-rounded understanding of ML/AI techniques along with adequate domain-specific knowledge in R&D is paramount for their optimal use in drug development.

DiMasi JA, Grabowski HG, Hansen RW. Innovation in the pharmaceutical industry: new estimates of R&D costs. J Health Econ. 2016;47:20–33. https://doi.org/10.1016/j.jhealeco.2016.01.012 .

Article   PubMed   Google Scholar  

Wong CH, Siah KW, Lo AW. Estimation of clinical trial success rates and related parameters. Biostatistics. 2019;20(2):273–86. https://doi.org/10.1093/biostatistics/kxx069 .

Russell S, Norvig P. Artificial intelligence: a modern approach (4th edition), 2021; Pearson Series in Artificial Intelligence.

Mitchell A, Sharma Y, Ramanathan S, Sethuraman V. Is data science the treatment for inefficiencies in clinical trial operations? White paper. https://www.zs.com/insights/is-data-science-the-treatment-for-inefficiencies-in-clinical-trial-operations .

Dill KA and MacCallum JL. The protein-folding problem, 50 years on. Science. 2012;338(6110):1042–6. https://doi.org/10.1126/science.1219021 .

Senior A, Evans R, Jumper J, Kirkpatrick J, Sifre L, Green T, Qin C, Žídek A, Nelson AWR, Bridgland A, Penedones H, Petersen S, Simonyan K, Crossan S, Kohli P, Jones DT, Silver D, Kavukcuoglu K, Hassabis D. Protein structure prediction using multiple deep neural networks in the 13th Critical Assessment of Protein Structure Prediction (CASP13). Proteins. 2019;87:1141–8. https://doi.org/10.1002/prot.25834 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Senior A, Evans R, Jumper J, Kirkpatrick J, Sifre L, Green T, Qin C, Žídek A, Nelson AWR, Bridgland A, Penedones H, Petersen S, Simonyan K, Crossan S, Kohli P, Jones D. T, Silver D, Kavukcuoglu K, Hassabis. Improved protein structure prediction using potentials from deep learning. Nature. 2020;577. https://doi.org/10.1038/s41586-019-1923-7 .

John Jumper, Kathryn Tunyasuvunakool, Pushmeet Kohli, Demis Hassabis, and the AlphaFold Team, Computational predictions of protein structures associated with COVID-19, Version 3, DeepMind website, 4 August 2020, https://deepmind.com/research/open-source/computational-predictions-of-protein-structures-associated-with-COVID-19 .

Beck BR, Shin B, Choi Y, Park S, Kang K. Predicting commercially available antiviral drugs that may act on the novel coronavirus (SARS-CoV-2) through a drug-target interaction deep learning model. Comput Struct Biotechnol J. 2020;18:784–90.

Article   CAS   Google Scholar  

Li B, Shin H, Gulbekyan G, Pustovalova O, Nikolsky Y, Hope A, Bessarabova M, Schu M, Kolpakova-Hart E, Merberg D, Dorner A, Trepicchio WL. Development of a drug-response modeling framework to identify cell line derived translational biomarkers that can predict treatment outcome to erlotinib or sorafenib. PLoS ONE. 2015;10(6):e0130700. https://doi.org/10.1371/journal.pone.0130700 .

Teh YW. Dirichlet process. In Sammut C, Webb GI (Eds) Encyclopedia of Machine Learning. 2011; pp. 280–287, Springer. https://doi.org/10.1007/978-0-387-30164-8_219 .

Li M, Liu R, Lin J, Bunn V, Zhao H. Bayesian semi-parametric design (BSD) for adaptive dose-finding with multiple strata. J Biopharm Stat. 2020;30(5):806–20. https://doi.org/10.1080/10543406.2020.1730870 .

Liu R, Lin J, Li P. Design considerations for phase I/II dose finding clinical trials in immuno-oncology and cell therapy. Contemporary Clinical Trials. 2020;96:106083. https://doi.org/10.1016/j.cct.2020.106083 .

Saville BR, Berry SM. Efficiencies of platform clinical trials: a vision of the future. Clin Trials. 2016;13(3):358–66. https://doi.org/10.1177/1740774515626362 .

Woodcock J, LaVange L. Master protocols to study multiple therapies, multiple diseases, or both. N Engl J Med. 2017;377:62–70. https://doi.org/10.1056/NEJMra1510062 .

Article   CAS   PubMed   Google Scholar  

Lin J, Lin L, Bunn V, and Liu R. Adaptive randomization for master protocols in precision medicine. In Contemporary Biostatistics with Biopharmaceutical Application, 2019; Springer, 251–270.

NIAID (2020). Adaptive COVID-19 treatment trial (ACTT). https://clinicaltrials.gov/ct2/show/study/NCT04280705 .

Bunn V, Liu R, Lin J, Lin J. Flexible Bayesian subgroup analysis in early and confirmatory trials. Contemp Clin Trials. 2020;98:106149. https://doi.org/10.1016/j.cct.2020.106149 .

Hupf B, Bunn V, Lin J, Dong C. Bayesian semiparametric meta-analytic-predictive prior for historical control borrowing in clinical trials. Stat Med (accepted). 2021. https://doi.org/10.1002/sim.8970 .

Article   Google Scholar  

Hwang TJ, Carpenter D, Lauffenburger JC. Failure of investigational drugs in late-stage clinical development and publication of trial results. JAMA Intern Med. 2016;176(12):1826–33. https://doi.org/10.1001/jamainternmed.2016.6008 .

Sertkaya A, Wong H, Jessup A, Beleche T. Key cost drivers of pharmaceutical clinical trials in the United States. Clin Trials. 2016;13(2):117–26. https://doi.org/10.1177/1740774515625964 .

Download references

Author information

Authors and affiliations.

Global Clinical & Real World Evidence Statistics, Global Biometrics, Teva Pharmaceuticals, 145 Brandywine Pkwy, PA, 19380, West Chester, USA

Sheela Kolluri

Statistical and Quantitative Science, Data Sciences Institute, Takeda Pharmaceutical Co. Limited, 300 Mass Ave, West Chester, PA, 19380, USA

Jianchang Lin, Rachael Liu, Yanwei Zhang & Wenwen Zhang

You can also search for this author in PubMed   Google Scholar

Contributions

SK, JL, RL, YZ, and WZ contributed to the ideas, implementation, and interpretation of the research topic, and to the writing of the manuscript.

Corresponding author

Correspondence to Sheela Kolluri .

Ethics declarations

Conflict of interest.

Sheela K. was previously employed by Takeda Pharmaceuticals and is currently employed by Teva Pharmaceuticals (West Chester PA USA) during the development and revision of this manuscript. All other authors are employed by Takeda Pharmaceuticals during the development and revision of this manuscript.

Additional information

Guest Editors: Diane Burgess, Marilyn Morris and Meena Subramanyam

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Kolluri, S., Lin, J., Liu, R. et al. Machine Learning and Artificial Intelligence in Pharmaceutical Research and Development: a Review. AAPS J 24 , 19 (2022). https://doi.org/10.1208/s12248-021-00644-3

Download citation

Received : 30 April 2021

Accepted : 26 August 2021

Published : 04 January 2022

DOI : https://doi.org/10.1208/s12248-021-00644-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Artificial intelligence
• Machine learning
• Drug development
• Precision medicine
• Probability of success
• Clinical trial design
• Risk-based monitoring
• Predictive modeling
• Find a journal
• Publish with us
• Track your research

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• BIOBUSINESS BRIEFS
• 08 May 2024

Development of Chinese innovative drugs in the USA

• Qianwei Ge 0 ,
• Xuan Zhang 1 ,
• Kenneth I. Kaitin 2 &
• Liming Shao 3

School of Pharmacy, Fudan University, Pudong, Shanghai, China.

You can also search for this author in PubMed   Google Scholar

Tufts Center for the Study of Drug Development (CSDD), Tufts University School of Medicine, Boston, MA, USA.

School of Pharmacy, Fudan University, Pudong, Shanghai, China; Shanghai Center for Drug Discovery & Development, Pudong, Shanghai, China.

Over the past decade, research advances, infrastructure and ecosystem improvements and drug regulatory system reforms have boosted the level of innovative drug research and development (R&D) in China. Furthermore, Chinese pharmaceutical companies have accelerated their integration into the global drug development enterprise by conducting clinical trials outside of China. Since February 2007, when the Chinese company Tasly Pharmaceuticals announced its decision to conduct a phase II trial in the USA for dantonic (T89), the first Chinese innovative drug to undergo clinical trials in the United States, the number of drugs of Chinese origin tested in the USA has grown steadily.

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

24,99 € / 30 days

cancel any time

Subscribe to this journal

Receive 12 print issues and online access

195,33 € per year

only 16,28 € per issue

Rent or buy this article

Prices vary by article type

Prices may be subject to local taxes which are calculated during checkout

doi: https://doi.org/10.1038/d41573-024-00079-3

Supplementary Information

• Supplementary information

Competing Interests

The authors declare no competing interests.

Faculty Positions& Postdoctoral Research Fellow, School of Optical and Electronic Information, HUST

Job Opportunities: Leading talents, young talents, overseas outstanding young scholars, postdoctoral researchers.

Wuhan, Hubei, China

School of Optical and Electronic Information, Huazhong University of Science and Technology

Postdoc in CRISPR Meta-Analytics and AI for Therapeutic Target Discovery and Priotisation (OT Grant)

APPLICATION CLOSING DATE: 14/06/2024 Human Technopole (HT) is a new interdisciplinary life science research institute created and supported by the...

Human Technopole

Research Associate - Metabolism

Houston, Texas (US)

Baylor College of Medicine (BCM)

Postdoc Fellowships

Train with world-renowned cancer researchers at NIH? Consider joining the Center for Cancer Research (CCR) at the National Cancer Institute

Bethesda, Maryland

NIH National Cancer Institute (NCI)

Faculty Recruitment, Westlake University School of Medicine

Faculty positions are open at four distinct ranks: Assistant Professor, Associate Professor, Full Professor, and Chair Professor.

Hangzhou, Zhejiang, China

Westlake University

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Pharmacological Reviews

New Aspects of Diabetes Research and Therapeutic Development

Both type 1 and type 2 diabetes mellitus are advancing at exponential rates, placing significant burdens on health care networks worldwide. Although traditional pharmacologic therapies such as insulin and oral antidiabetic stalwarts like metformin and the sulfonylureas continue to be used, newer drugs are now on the market targeting novel blood glucose–lowering pathways. Furthermore, exciting new developments in the understanding of beta cell and islet biology are driving the potential for treatments targeting incretin action, islet transplantation with new methods for immunologic protection, and the generation of functional beta cells from stem cells. Here we discuss the mechanistic details underlying past, present, and future diabetes therapies and evaluate their potential to treat and possibly reverse type 1 and 2 diabetes in humans.

Significance Statement

Diabetes mellitus has reached epidemic proportions in the developed and developing world alike. As the last several years have seen many new developments in the field, a new and up to date review of these advances and their careful evaluation will help both clinical and research diabetologists to better understand where the field is currently heading.

I. Introduction

Diabetes mellitus, a metabolic disease defined by elevated fasting blood glucose levels due to insufficient insulin production, has reached epidemic proportions worldwide (World Health Organization, 2020 ). Type 1 and type 2 diabetes (T1D and T2D, respectively) make up the majority of diabetes cases with T1D characterized by autoimmune destruction of the insulin-producing pancreatic beta cells. The much more prevalent T2D arises in conjunction with peripheral tissue insulin resistance and beta cell failure and is estimated to increase to 21%–33% of the US population by the year 2050 (Boyle et al., 2010 ). To combat this growing health threat and its cardiac, renal, and neurologic comorbidities, new and more effective diabetes drugs and treatments are essential. As the last several years have seen many new developments in the field of diabetes pharmacology and therapy, we determined that a new and up to date review of these advances was in order. Our aim is to provide a careful evaluation of both old and new therapies ( Fig. 1 ) in a manner that we hope will be of interest to both clinical and bench diabetologists. Instead of the usual encyclopedic approach to this topic, we provide here a targeted and selective consideration of the underlying issues, promising new treatments, and a re-examination of more traditional approaches. Thus, we do not discuss less frequently used diabetes agents, such as alpha-glucosidase inhibitors; these were discussed in other recent reviews (Hedrington and Davis, 2019 ; Lebovitz, 2019 ).

Pharmacologic targeting of numerous organ systems for the treatment of diabetes. Treatment of diabetes involves targeting of various organ systems, including the kidney by SGLT2 inhibitors; the liver, gut, and adipose tissue by metformin; and direct actions upon the pancreatic beta cell. Beta cell compounds aim to increase secretion or mass and/or to protect from autoimmunity destruction. Ultimately, insulin therapy remains the final line of diabetes treatment with new technologies under development to more tightly regulate blood glucose levels similar to healthy beta cells. hESC, human embryonic stem cell.

II. Diabetes Therapies

A. metformin.

Metformin is a biguanide originally based on the natural product galegine, which was extracted from the French lilac (Bailey, 1992 ; Rojas and Gomes, 2013 ; Witters, 2001 ). A closely related biguanide, phenformin, was also used initially for its hypoglycemic actions. Based on its successful track record as a safe, effective, and inexpensive oral medication, metformin has become the most widely prescribed oral agent in the world in treating T2D (Rojas and Gomes, 2013 ; He and Wondisford, 2015 ; Witters, 2001 ), whereas phenformin has been largely bypassed due to its unacceptably high association with lactic acidosis (Misbin, 2004 ). Unlike sulfonylureas, metformin lowers blood glucose without provoking hypoglycemia and improves insulin sensitivity (Bailey, 1992 ). Despite these well known beneficial metabolic actions, metformin’s mechanism of action and even its main target organ remain controversial. In fact, metformin has multiple mechanisms of action at the organ as well as the cellular level, which has hindered our understanding of its most important molecular effects on glucose metabolism (Witters, 2001 ). Adding to this, a specific receptor for metformin has never been identified. Metformin has actions on several tissues, although the primary foci of most studies have been the liver, skeletal muscle, and the intestine (Foretz et al., 2014 ; Rena et al., 2017 ). Metformin and phenformin clearly suppress hepatic glucose production and gluconeogenesis, and they improve insulin sensitivity in the liver and elsewhere (Bailey, 1992 ). The hepatic actions of metformin have been the most exhaustively studied to date, and there is little doubt that these actions are of some importance. However, several of the studies remain highly controversial, and there are still open questions.

One of the first reported specific molecular targets of metformin was mitochondrial complex I of the electron transport chain. Inhibition of this complex results in reduced oxidative phosphorylation and consequently decreased hepatic ATP production (El-Mir et al., 2008 ; Evans et al., 2005 ; Owen et al., 2000 ). As is the case in many other studies of metformin, however, high concentrations of the drug were found to be necessary to depress metabolism at this site (El-Mir et al., 2000 ; He and Wondisford, 2015 ; Owen et al., 2000 ). Also controversial is whether metformin works by activating 5′ AMP-activated protein kinase (AMPK), a molecular energy sensor that is known to be a major metabolic sensor in cells, or if not AMPK directly, then one of its upstream regulators such as liver kinase B2 (Zhou et al., 2001 ). Although metformin was shown to activate AMPK in several excellent studies, other studies directly contradicted the AMPK hypothesis. Most dramatic were studies showing that metformin’s actions to suppress hepatic gluconeogenesis persisted despite genetic deletion of the AMPK’s catalytic domain (Foretz et al., 2010 ). More recent studies identified additional or alternative targets, such as cAMP signaling in the liver (Miller et al., 2013 ) or glycogen synthase kinase-3 (Link, 2003 ). Other work showed that the phosphorylation of acetyl-CoA carboxylase and acetyl-CoA carboxylase 2 are involved in regulating lipid homeostasis and improving insulin sensitivity after exposure to metformin (Fullerton et al., 2013 ).

Although there are strong data to support each of these pathways, it is not entirely clear which signaling pathway(s) is most essential to the actions of metformin in hepatocytes. Metformin clearly inhibits complex I and concomitantly decreases ATP and increases AMP. The latter results in AMPK activation, reduced fatty acid synthesis, and improved insulin receptor activation, and increased AMP has been shown to inhibit adenylate cyclase to reduce cAMP and thus protein kinase A activation. Downstream, this reduces the expression of phosphoenolpyruvate carboxykinase and glucose 6-phosphatase via decreased cAMP response element-binding protein, the cAMP-sensitive transcription factor. Decreased PKA also promotes ATP-dependent 6-phosphofructokinase, liver type activity via fructose 2,6-bisphosphate and reduces gluconeogenesis, as fructose-bisphosphatase 1 is inhibited by fructose 2,6-bisphosphate, along with other mechanisms (Rena et al., 2017 ; Pernicova and Korbonits, 2014 ).

More recent work has shown that metformin at pharmacological rather than suprapharmacological doses increases mitochondrial respiration and complex 1 activity and also increases mitochondrial fission, now thought to be critical for maintaining proper mitochondrial density in hepatocytes and other cells. This improvement in respiratory activity occurs via AMPK activation (Wang et al., 2019 ).

Although the liver has historically been the major suspected site of metformin action, recent studies have suggested that the gut instead of the liver is a major target, a concept supported by the increased efficacy of extended-release formulations of metformin that reside for a longer duration in the gut after their administration (Buse et al., 2016 ). An older, but in our view an important observation, is that the intravenous administration of metformin has little or no effect on blood glucose, whereas, in contrast, orally administered metformin is much more effective (Bonora et al., 1984 ). Recent imaging studies using labeled glucose have shown directly that metformin stimulates glucose uptake by the gut in patients with T2D to reduce plasma glucose concentrations (Koffert et al., 2017 ; Massollo et al., 2013 ). Additionally, it is possible that metformin may exert its effect in the gut by inducing intestinal glucagon-like peptide-1 (GLP-1) release (Mulherin et al., 2011 ; Preiss et al., 2017) to potentiate beta cell insulin secretion and by stimulating the central nervous system (CNS) to exert control over both blood glucose and liver function. Indeed, CNS effects produced by metformin have been proposed to occur via the local release of GLP-1 to activate intestinal nerve endings of ascending nerve pathways that are involved in CNS glucose regulation (Duca et al., 2015 ). Lastly, several papers have now implicated that metformin may act by altering the gut microbiome, suggesting that changes in gut flora may be critical for metformin’s actions (McCreight et al., 2016 ; Wu et al., 2017 ; Devaraj et al., 2016 ). A new study proposed that activation of the intestinal farnesoid X receptor may be the means by which microbiota alter hyperglycemia (Sun et al., 2018 ). However, these studies will require more mechanistic detail and confirmation before they can be fully accepted by the field. In addition to the action of metformin on gut flora, the production of imidazole propionate by gut microbes in turn has been shown to interfere with metformin action through a p38-dependent mechanism and AMPK inhibition. Levels of imidazole propionate are especially higher in patients with T2D who are treated with metformin (Koh et al., 2020 ).

In summary, the combined contribution of these various effects of metformin on multiple cellular targets residing in many tissues may be key to the benefits of metformin treatment on lowering blood glucose in patients with type 2 diabetes (Foretz et al., 2019 ). In contrast, exciting new work showing metformin leads to weight loss by increasing circulating levels of the peptide hormone growth differentiation factor 15 and activation of brainstem glial cell-derived neurotropic factor family receptor alpha like receptors to reduce food intake and energy expenditure works independently of metformin’s glucose-lowering effect (Coll et al., 2020 ).

B. Sulfonylureas and Beta Cell Burnout

The class of compounds known as sulfonylureas includes one of the oldest oral antidiabetic drugs in the pharmacopoeia: tolbutamide. Tolbutamide is a “first generation” oral sulfonylurea secretagogue whose clinical usefulness is due to its prompt stimulation of insulin release from pancreatic beta cells. “Second generation” sulfonylureas include drugs such as glyburide, gliclazide, and glipizide. Sulfonylureas act by binding to a high affinity sulfonylurea binding site, the sulfonylurea receptor 1 subunit of the K(ATP) channel, which closes the channel. These drugs mimic the physiologic effects of glucose, which closes the K(ATP) channel by raising cytosolic ATP/ADP. This in turn provokes beta cell depolarization, resulting in increased Ca 2+ influx into the beta cell (Ozanne et al., 1995 ; Ashcroft and Rorsman, 1989 ; Nichols, 2006 ). Importantly, sulfonylureas, and all drugs that directly increase insulin secretion, are associated with hypoglycemia, which can be severe, and which limits their widespread use in the clinic (Yu et al., 2018 ). Meglitinides are another class of oral insulin secretagogues that, like the sulfonylureas, bind to sulfonylurea receptor 1 and inhibit K(ATP) channel activity (although at a different site of action). The rapid kinetics of the meglitinides enable them to effectively blunt the postprandial glycemic excursions that are a hallmark (along with elevated fasting glucose) of T2D (Rosenstock et al., 2004). However, the need for their frequent dosing (e.g., administration before each meal) has limited their appeal to patients.

The efficacy of sulfonylureas is known to decrease over time, leading to failure of the class for effective long-term treatment of T2D (Harrower, 1991 ). More broadly, it is now widely accepted that the number of functional beta cells in humans declines during the progression of T2D. Thus, one would expect that due to this decline, all manner of oral agents intended to target the beta cell and increase its cell function (and especially insulin secretion) will fail over time (RISE Consortium, 2019 ), a process referred to as “beta cell failure” (Prentki and Nolan, 2006 ). Currently, treatments that can expand beta cell mass or improve beta cell function or survival over time are not yet available for use in the clinic. As a result, treatments that may be able to help patients cope with beta cell burnout such as islet cell transplantation, insulin pumps, or stem cell therapy are alternatives that will be discussed below.

C. Ca 2+ Channel Blockers and Type 1 Diabetes

Strategies to treat and prevent T1D have historically focused on ameliorating the toxic consequences of immune dysregulation resulting in autoimmune destruction of pancreatic beta cells. More recently, a concerted focus on alleviating the intrinsic beta cell defects (Sims et al., 2020 ; Soleimanpour and Stoffers, 2013 ) that also contribute to T1D pathogenesis have been gaining traction at both the bench and the bedside. Several recent preclinical studies suggest that Ca 2+ -induced metabolic overload induces beta cell failure (Osipovich et al., 2020 ; Stancill et al., 2017 ; Xu et al., 2012 ), with the potential that excitotoxicity contributes to beta cell demise in both T1D and T2D, similar to the well known connection between excitotoxicity and, concomitantly, increased Ca 2+ loading of the cells and neuronal dysfunction. Indeed, the use of the phenylalkylamine Ca 2+ channel blocker verapamil has been successful in ameliorating beta cell dysfunction in preclinical models of both T1D and T2D (Stancill et al., 2017 ; Xu et al., 2012 ). Verapamil is a well known blocker of L-type Ca 2+ channels, and, in normally activated beta cells, it limits Ca 2+ entry into the beta cell (Ohnishi and Endo, 1981 ; Vasseur et al., 1987 ). This would be expected to, in turn, alter the expression of many Ca 2+ influx–dependent beta cell genes (Stancill et al., 2017 ), and the evidence to date suggests it is likely that verapamil preserves beta cell function in diabetes models by repressing thioredoxin-interacting protein (TXNIP) expression and thus protecting the beta cell. This is somewhat surprising given the physiologic role of Ca 2+ is to acutely trigger insulin secretion; this process would be expected to be inhibited by L-type Ca 2+ channel blockers (Ashcroft and Rorsman, 1989 ; Satin et al., 1995 ).

Hyperglycemia is a well known inducer of TXNIP expression, and a lack of TXNIP has been shown to protect against beta cell apoptosis after inflammatory stress (Chen et al., 2008a ; Shalev et al., 2002 ; Chen et al., 2008b ). Excitingly, the use of verapamil in patients with recent-onset T1D improved beta cell function and improved glycemic control for up to 12 months after the initiation of therapy, suggesting there is indeed promise for targeting calcium and TXNIP activation in T1D. Use of verapamil for a repurposed indication in the preservation of beta cell function in T1D is attractive due its well known safety profile as well as its cardiac benefits (Chen et al., 2009 ). Although the long-term efficacy of verapamil to maintain beta cell function in vivo is unclear, a recently described TXNIP inhibitor may also show promise in suppressing the hyperglucagonemia that also contributes to glucose intolerance in T2D (Thielen et al., 2020 ). As there is a clear need for increased Ca 2+ influx into the beta cell to trigger and maintain glucose-dependent insulin secretion (Ashcroft and Rorsman, 1990 ; Satin et al., 1995 ), it remains to be seen how well regulated insulin secretion is preserved in the presence of L-type Ca 2+ channel blockers like verapamil in the system. One might speculate that reducing but not fully eliminating beta cell Ca 2+ influx might reduce TXNIP levels while preserving enough influx to maintain glucose-stimulated insulin release. Alternatively, these two phenomena may operate on entirely different time scales. At present, these issues clearly will require further investigation.

D. GLP-1 and the Incretins

Studies dating back to the 1960s revealed that administering glucose in equal amounts via the peripheral circulation versus the gastrointestinal tract led to dramatically different amounts of glucose-induced insulin secretion (Elrick et al., 1964 ; McIntyre et al., 1964 ; Perley and Kipnis, 1967 ). Gastrointestinal glucose administration greatly increased insulin secretion versus intravenous glucose, and this came to be known as the “incretin effect” (Nauck et al., 1986a ; Nauck et al., 1986b ). Subsequent work showed that release of the gut hormone GLP-1 mediated this effect such that food ingestion induced intestinal cell hormone secretion. GLP-1 so released would then circulate to the pancreas via the blood to prime beta cells to secrete more insulin when glucose became elevated because these hormones stimulated beta cell cAMP formation (Drucker et al., 1987 ). The discovery that a natural peptide corresponding to GLP-1 could be found in the saliva of the Gila monster, a desert lizard, hastened progress in the field, and ample in vitro studies subsequently confirmed that GLP-1 potentiated insulin secretion in a glucose-dependent manner. GLP-1 has little or no significant action on insulin secretion in the absence of elevated glucose (such as might typically correspond to the postprandial case or during fasting), thus minimizing the likelihood of hypoglycemia provoked by GLP-1 in treated patients (Kreymann et al., 1987 ). Although not completely understood, the glucose dependence of GLP-1 likely reflects the requirement for adenine nucleotides to close glucose-inhibited K(ATP) channels and thus subsequently activate Ca 2+ influx–dependent insulin exocytosis. Besides potentiating GSIS at the level of the beta cell, glucagon-like peptide-1 receptor (GLP-1R) agonists also decrease glucagon secretion from pancreatic islet alpha cells, reduce gastric emptying, and may also increase beta cell proliferation, among other cellular actions (reviewed in Drucker, 2018 ; Muller et al., 2019).

Intense interest in the incretins by basic scientists, clinicians, and the pharma community led to the rapid development of new drugs for treating primarily T2D. These drugs include a range of GLP-1R agonists and inhibitors of the incretin hormone degrading enzyme dipeptidyl peptidase 4 (DPP4), whose targeting increases the half-lives of GLP-1 and gastric inhibitory polypeptide (GIP) and thereby increases protein hormone levels in plasma. GLP-1R agonists have been associated with not only a lowering of plasma glucose but also weight loss, decreased appetite, reduced risk of cardiovascular events, and other favorable outcomes (Gerstein et al., 2019; Hernandez et al., 2018; Husain et al., 2019; Marso et al., 2016a; Marso et al., 2016b ; Buse et al., 2004). Regarding their untoward actions, although hypoglycemia is not a major concern, there have been reports of pancreatitis and pancreatic cancer from use of GLP-1R agonists. However, a recent meta-analysis covering four large-scale clinical trials and over 33,000 participants noted no significantly increased risk for pancreatitis/pancreatic cancer in patients using GLP-1R agonists (Bethel et al., 2018).

Ongoing and future developments in the use of proglucagon-derived peptides such as GLP-1 and glucagon include the use of combined GLP-1/GIP, glucagon/GLP-1, and agents targeting all three peptides in combination (reviewed in Alexiadou and Tan, 2020 ). Although short-term infusions of GLP-1 with GIP failed to yield metabolic benefits beyond those seen with GLP-1 alone (Bergmann et al., 2019 ), several GLP-1/GIP dual agonists are currently in development and have shown promising metabolic results in clinical trials (Frias et al., 2017 ; Frias et al., 2020 ; Frias et al., 2018 ). At the level of the pancreatic islet, beneficial effects of dual GLP-1/GIP agonists may be related to imbalanced and biased preferences of these agonists for the gastric inhibitory polypeptide receptor over the GLP-1R (Willard et al., 2020 ) and possibly were not simply to dual hormone agonism in parallel. Dual glucagon/GLP-1 agonist therapy has also been shown to have promising metabolic effects in humans (Ambery et al., 2018 ; Tillner et al., 2019 ). Oxyntomodulin is a natural dual glucagon/GLP-1 receptor agonist and proglucagon cleavage product that is also secreted from intestinal enteroendocrine cells, which has beneficial effects on insulin secretion, appetite regulation, and body weight in both humans and rodents (Cohen et al., 2003 ; Dakin et al., 2001 ; Dakin et al., 2002 ; Shankar et al., 2018 ; Wynne et al., 2005 ). Interestingly, alpha cell crosstalk to beta cells through the combined effects of glucagon and GLP-1 is necessary to obtain optimal glycemic control, suggesting a potential pathway for therapeutic dual glucagon/GLP-1 agonism within the islets of patients with T2D (Capozzi et al., 2019a ; Capozzi et al., 2019b ). Although the early results appear promising, more studies will be necessary to better understand the mechanistic and clinical impacts of these multiagonist agents.

E. DPP4 Inhibitors

Inhibition of DPP4, the incretin hormone degrading enzyme, is one of the most common T2D treatments to increase GLP-1 and GIP plasma hormone levels. These DPP4 inhibitors or “gliptins” are generally used in conjunction with other T2D drugs such as metformin or sulfonylureas to obtain the positive benefits discussed above (Lambeir et al., 2008 ). DPP4 is a primarily membrane-bound peptidase belonging to the serine peptidase/prolyl oligopeptidase gene family, which cleaves a large number of substrates in addition to the incretin hormones (Makrilakis, 2019 ). DPP4 inhibitors provide glucose-lowering benefits while being generally well tolerated, and the variety of available drugs (including sitagliptin, saxagliptin, vildagliptin, alogliptin, and linagliptin) with slightly different dosing frequency, half-life, and mode of excretion/metabolism allows for use in multiple patient populations (Makrilakis, 2019 ). This includes the elderly and individuals with renal or hepatic insufficiency (Makrilakis, 2019 ).

Although hypoglycemia is not a concern for DPP4 inhibitor use, other considerations should be made. DPP4 inhibitors tend to be more expensive than metformin or other second-line oral drugs in addition to having more modest glycemic effects than GLP-1R agonists (Munir and Lamos, 2017 ). Finally, meta-analysis of randomized and observational studies concluded that heart failure in patients with T2D was not associated with use of DPP4 inhibitors; however, this study was limited by the short follow-up and lack of high-quality data (Li et al., 2016 ). Thus, the US Food and Drug Administration (FDA) did recommend assessing risk of heart failure hospitalization in patients with pre-existing cardiovascular disease, prior heart failure, and chronic kidney disease when using saxagliptin and alogliptin (Munir and Lamos, 2017 ).

F. Sodium Glucose Cotransporter 2 Inhibitors

A recent development in the field of T2D drugs are sodium glucose cotransporter 2 (SGLT2) inhibitors, which have an interesting and very different mechanism of action. Within the proximal tubule of the nephron, SGLT2 transports ingested glucose into the lumen of the proximal tubule between the epithelial layers, thereby reclaiming glucose by this reabsorption process (reviewed in Vallon, 2015 ). SGLT2 inhibitors target this transporter and increase glucose in the tubular fluid and ultimately increase it in the urine. In patients with diabetes, SGLT2 inhibition results in a lowering of plasma glucose with urine glucose content rising substantially (Adachi et al., 2000 ; Vallon, 2015 ). These drugs, although they are relatively new, have become an area of great interest for not only patients with T2D (Grempler et al., 2012 ; Imamura et al., 2012 ; Meng et al., 2008 ; Nomura et al., 2010 ) but also for patients with T1D (Luippold et al., 2012 ; Mudaliar et al., 2012 ). Part of their appeal also rests on reports that their use can lead to a statistically significant decline in cardiac events that are known to occur secondarily to diabetes, possibly independently of plasma glucose regulation (reviewed in Kurosaki and Ogasawara, 2013 ). Although the long-term consequences of their clinical use cannot yet be determined, raising the glucose content of the urogenital tract leads to an increased risk of urinary tract infections and other related infections in some patients (Kurosaki and Ogasawara, 2013 ).

Another recent concern about the use of SGLT2 inhibitors has been the development of normoglycemic diabetic ketoacidosis (DKA). Despite the efficacy of SGLT2 inhibitors, observations of hyperglucagonemia in patients with euglycemic DKA has led to a number of recent studies focused on SGLT2 actions on pancreatic islets. Initial studies of isolated human islets treated with small interfering RNA directed against SGLT2 and/or SGLT2 inhibitors demonstrated increased glucagon release. These studies were complemented by the finding of elevations in glucagon release in mice that were administered SGLT2 inhibitors in vivo (Bonner et al., 2015 ). Insights into the possible mechanistic links between SGLT2 inhibition, DKA frequency, and glucagon secretion in humans may relate to the observation of heterogeneity in SGLT2 expression, as SGLT2 expression appears to have a high frequency of interdonor and intradonor variability (Saponaro et al., 2020 ). More recently, both insulin and GLP-1 have been demonstrated to modulate SGLT2-dependent glucagon release through effects on somatostatin release from delta cells (Vergari et al., 2019 ; Saponaro et al., 2019 ), suggesting potentially complex paracrine effects that may affect the efficacy of these compounds.

On the other hand, several recent studies question that the development of euglycemic DKA after SGLT2 inhibitor therapy may be through alpha cell–dependent mechanisms. Three recent studies found no effect of SGLT2 inhibitors to promote glucagon secretion in mouse and/or rat models and could not detect SGLT2 expression in human alpha cells (Chae et al., 2020 ; Kuhre et al., 2019 ; Suga et al., 2019 ). A fourth study demonstrated only a brief transient effect of SGLT2 inhibition to raise circulating glucagon concentrations in immunodeficient mice transplanted with human islets, which returned to baseline levels after longer exposures to SGLT2 inhibitors (Dai et al., 2020 ). Furthermore, SGLT2 protein levels were again undetectable in human islets (Dai et al., 2020 ). These results could suggest alternative islet-independent mechanisms by which patients develop DKA, including alterations in ketone generation and/or clearance, which underscore the additional need for further studies both in molecular models and at the bedside. Nevertheless, SGLT2 inhibitors continue to hold promise as a valuable therapy for T2D, especially in the large segment of patients who also have superimposed cardiovascular risk (McMurray et al., 2019; Wiviott et al., 2019; Zinman et al., 2015).

G. Thiazolidinediones

Once among the most commonly used oral agents in the armamentarium to treat T2D, thiazolidinediones (TZDs) were clinically popular in their utilization to act specifically as insulin sensitizers. TZDs improve peripheral insulin sensitivity through their action as peroxisome proliferator-activated receptor (PPAR) γ agonists, but their clinical use fell sharply after studies suggested a connection between cardiovascular toxicity with rosiglitazone and bladder cancer risk with pioglitazone (Lebovitz, 2019 ). Importantly, an FDA panel eventually removed restrictions related to cardiovascular risk with rosiglitazone in 2013 (Hiatt et al., 2013 ). Similarly, concerns regarding use of bladder cancer risk with pioglitazone were later abated after a series of large clinical studies found that pioglitazone did not increase bladder cancer (Lewis et al., 2015 ; Schwartz et al., 2015 ). However, usage of TZDs had already substantially decreased and has not since recovered.

Although concerns regarding edema, congestive heart failure, and fractures persist with TZD use, there have been several studies suggesting that TZDs protect beta cell function. In the ADOPT study, use of rosiglitazone monotherapy in patients newly diagnosed with T2D led to improved glycemic control compared with metformin or sulfonylureas (Kahn et al., 2006). Later analyses revealed that TZD-treated subjects had a slower deterioration of beta cell function than metformin- or sulfonylurea-treated subjects (Kahn et al., 2011). Furthermore, pioglitazone use improved beta cell function in the prevention of T2D in the ACT NOW study (Defronzo et al., 2013; Kahn et al., 2011). Mechanistically, it is unclear if TZDs lead to beneficial beta cell function through direct effects or through indirect effects of reduced beta cell demand due to enhanced peripheral insulin sensitivity. Indeed, a beta cell–specific knockout of PPAR γ did not impair glucose homeostasis, nor did it impair the antidiabetic effects of TZD use in mice (Rosen et al., 2003 ). However, other reports demonstrated PPAR-responsive elements within the promoters of both glucose transporter 2 and glucokinase that enhance beta cell glucose sensing and function, which could explain beta cell–specific benefits for TZDs (Kim et al., 2002 ; Kim et al., 2000 ). Furthermore, TZDs have been shown to improve beta cell function by upregulating cholesterol transport (Brunham et al., 2007 ; Sturek et al., 2010 ). Additionally, use of TZDs in the nonobese diabetic (NOD) mouse model of T1D augmented the beta cell unfolded protein response and prevented beta cell death, suggesting potential benefits for TZDs in both T1D and T2D (Evans-Molina et al., 2009 ; Maganti et al., 2016 ). With a now refined knowledge of demographics in which to avoid TZD treatment due to adverse effects, together with genetic approaches to identify candidates more likely to respond effectively to TZD therapy (Hu et al., 2019 ; Soccio et al., 2015 ), it remains to be seen if TZD therapy will return to more prominent use in the treatment of diabetes.

H. Insulin and Beyond: The Use of “Smart” Insulin and Closed Loop Systems in Diabetes Treatment

Due to recombinant DNA technology, numerous insulin analogs are now available in various forms ranging from fast acting crystalline insulin to insulin glargine; all of these analogs exhibit equally effective insulin receptor binding. Most are generated by altering amino acids in the B26–B30 region of the molecule (Kurtzhals et al., 2000 ). The American Diabetes Association delineates these insulins by their 1) onset or time before insulin reaches the blood stream, 2) peak time or duration of maximum blood glucose–lowering efficacy, and 3) the duration of blood glucose–lowering time. Insulin administration is independent of the residuum of surviving and/or functioning beta cells in the patient and remains the principal pharmacological treatment of both T1D and T2D. The availability of multiple types of delivery methods, i.e., insulin pens, syringes, pumps, and inhalants, provides clinicians with a solid and varied tool kit with which to treat diabetes. The downsides, however, are that 1) hypoglycemia is a constant threat, 2) proper insulin doses are not trivial to calculate, 3) compliance can vary especially in children and young adults, and 4) there can be side effects of a variety of types. Nonetheless, insulin therapy remains a mainstay treatment of diabetes.

To eliminate the downsides of insulin therapy, research in the past several decades has worked toward generating glucose-sensitive or “smart” insulin molecules. These molecules change insulin bioavailability and become active only upon high blood glucose using glucose-binding proteins such as concanavalin A, glucose oxidase to alter pH sensitivity, and phenylboronic acid (PBA), which forms reversible ester linkages with diol-containing molecules including glucose itself (reviewed in Rege et al., 2017 ). Indeed, promising recent studies included various PBA moieties covalently bonded to an acylated insulin analog (insulin detemir, which contains myristic acid coupled to Lys B29 ). The detemir allows for binding to serum albumin to prolong insulin’s half-life in the circulation, and PBA provided reversible glucose binding (Chou et al., 2015 ). The most promising of the PBA-modified conjugates showed higher potency and responsiveness in lowering blood glucose levels compared with native insulin in diabetic mouse models and decreased hypoglycemia in healthy mice, although the molecular mechanisms have not yet been determined (Chou et al., 2015 ).

An additional active area of research includes structurally defining the interaction between insulin and the insulin receptor ectodomain. Importantly, a major conformational change was discovered that may be exploited to impair insulin receptor binding under hypoglycemic conditions (Menting et al., 2013 ; Rege et al., 2017 ). Challenges in the design, testing, and execution of glucose-responsive insulins may be overcome by the adaptation of novel modeling approaches (Yang et al., 2020 ), which may allow for more rapid screening of candidate compounds.

Technologies have also progressed in the field of artificial pancreas design and development. Currently two “closed loop” systems are now available: Minimed 670G from Medtronic and Control-IQ from Tandem Diabetes Care. Both systems use a continuous glucose monitor, insulin pump, and computer algorithm to predict correct insulin doses and administer them in real time. Such algorithm systems also take into account insulin potency, the rate of blood glucose increase, and the patient’s heart rate and temperature to adjust insulin delivery levels during exercise and after a meal. In addition, so-called “artificial pancreas” systems have also been clinically tested, which use both insulin and glucagon and as such result in fewer reports of hypoglycemic episodes (El-Khatib et al., 2017 ). These types of systems will continue to become more popular as the development of room temperature–stable glucagon analogs continue, such as GVOKE by Xeris Pharmaceuticals (currently available in an injectable syringe) and Baqsimi, a nasally administered glucagon from Eli Lilly.

I. Present and Future Therapies: Beta Cell Transplantation, Replication, and Immune Protection

1. islet transplantation.

The idea to use pancreatic allo/xenografts to treat diabetes remarkably dates back to the late 1800s (Minkowski, 1892 ; Pybus, 1924 ; Williams, 1894 ). Before proceeding to the discovery of insulin (together with Best, MacLeod, and Collip), Frederick Banting also postulated the potential for transplantation of pancreatic tissue emulsions to treat diabetes in dog models in a notebook entry in 1921 (Bliss, 1982 ). Decades later, Paul Lacy, David Scharp, and colleagues successfully isolated intact functional pancreatic islets and transplanted them into rodent models (Kemp et al., 1973 ). These studies led to the initial proof of concept studies for humans, with the first successful islet transplant in a patient with T1D occurring in 1977 (Sutherland et al., 1978 ). A rapid expansion of islet transplantation, inspired by these original studies led to key observations of successfully prolonged islet engraftment by the “Edmonton protocol” whereby corticosteroid-sparing immunosuppression was applied, and islets from at least two allogeneic donors were used to achieve insulin independence (Shapiro et al., 2000 ). More recent work has focused on improving upon the efficiency and long-term engraftment of allogeneic transplants leading to more prolonged graft function (to the 5-year mark) and successful transplantation from a single islet donor (Hering et al., 2016; Hering et al., 2005 ; Rickels et al., 2013 ). Critical to these efforts to improve the success rate was the recognition that the earlier generation of immunosuppressive agents to counter tissue rejection was toxic to islets (Delaunay et al., 1997 ; Paty et al., 2002 ; Soleimanpour et al., 2010 ) and that more appropriate and less toxic agents were needed (Hirshberg et al., 2003 ; Soleimanpour et al., 2012 ).

Certainly, islet transplantation as a therapeutic approach for patients with T1D has been scrutinized due to several challenges, including (but not limited to) the lack of available donor supply to contend with demand, limited long-term functional efficacy of islet allografts, the potential for re-emergence of autoimmune islet destruction and/or metabolic overload-induced islet failure, and significant adverse effects of prolonged immunosuppression (Harlan, 2016 ). Furthermore, although islet transplantation is not currently available for individuals with T2D, simultaneous pancreas-kidney transplantation in T2D had similar favorable outcomes to simultaneous pancreas-kidney transplantation in T1D; therefore, islet-kidney transplantation may eventually be a feasible option to treat T2D, as patients will already be on immunosuppressors (Sampaio et al., 2011 ; Westerman et al., 1983 ). An additional significant obstacle is the tremendous expense associated with islet transplantation therapy. Indeed, the maintenance, operation, and utilization of an FDA-approved and Good Manufacturing Practice–compliant islet laboratory can lead to operating costs at nearly $150,000 per islet transplant, which is not cost effective for the vast majority of patients with T1D (Naftanel and Harlan, 2004 ; Wallner et al., 2016 ). At present, the focus has been to obtain FDA approval for islet allo-transplantation as a therapy for T1D to allow for insurance compensation (Hering et al., 2016; Rickels and Robertson, 2019 ). In the interim, the islet biology, stem cell, immunology, and bioengineering communities have continued the development of cell-based therapies for T1D by other approaches to overcome the challenges identified during the islet transplantation boom of the 1990s and 2000s.

2. Pharmacologic Induction of Beta Cell Replication

Besides transplantation, progress in islet cell biology and especially in developmental biology of beta cells over several decades raised the additional possibility that beta cell mass reduction in diabetes might be countered by increasing beta cell number through mitogenic means. A key method to expand pancreatic beta cell mass is through the enhancement of beta cell replication. Although the study of pancreatic beta cell replication has been an area of intense focus in the beta cell biology field for several decades, only recently has this seemed truly feasible. Seminal studies identified that human beta cells are essentially postmitotic, with a rapid phase of growth occurring in the prenatal period that dramatically tapers off shortly thereafter (Gregg et al., 2012 ; Meier et al., 2008 ). The plasticity of rodent beta cells is considerably higher than that of human beta cells (Dai et al., 2016 ), which has led to a renewed focus on validation of pharmacologic agents to enhance rodent beta cell replication using isolated and/or engrafted human islets (Bernal-Mizrachi et al., 2014 ; Kulkarni et al., 2012 ; Stewart et al., 2015 ). Indeed, a large percentage of agents that were successful when applied to rodent systems were largely unsuccessful at inducing replication in human beta cells (Bernal-Mizrachi et al., 2014 ; Kulkarni et al., 2012 ; Stewart et al., 2015 ). However, several recent studies have begun to make significant progress on successfully pushing human beta cells to replicate.

Several groups have reported successful human beta cell proliferation, both in vitro and in vivo, in response to inhibitors of the dual specificity tyrosine phosphorylation-regulated kinase 1A (DYRK1A). These inhibitors include harmine, INDY, GNF4877, 5-iodotubericidin, leucettine-42, TG003, AZ191, CC-401, and more specific, recently developed DYRK1A inhibitors (Ackeifi et al., 2020 ). Although DYRK1A is conclusively established as the important mediator of human beta cell proliferation, comprehensively determining other cellular targets and if additional gene inhibition amplifies the proliferative response is still in process. New evidence from Wang and Stewart shows dual specificity tyrosine phosphorylation-regulated kinase 1B to be an additional mitogenic target and also describes variability in the range of activated kinases within cells and/or levels of inhibition for the many DYRK1A inhibitors listed above (Ackeifi et al., 2020 ). Interestingly, opposite to these human studies, earlier mouse studies from the Scharfmann group demonstrated that Dyrk1a haploinsufficiency leads to decreased proliferation and loss of beta cell mass (Rachdi et al., 2014b ). In addition, overexpression of Dyrk1a in mice led to beta cell mass expansion with increased glucose tolerance (Rachdi et al., 2014a ).

Although important differences in beta cell proliferative capacity have been shown between human and rodent species, there are also significant differences in the mitogenic capacity of beta cells from juvenile, adult, and pregnant individuals. This demonstrates that proliferative stimuli appear to act within the complex islet, pancreas, and whole-body environments unique to each time point. For example, the administration of the hormones platelet-derived growth factor alpha or GLP-1 result in enhanced proliferation in juvenile human beta cells yet are ineffective in adult human beta cells (Chen et al., 2011 ; Dai et al., 2017 ). This has been shown to be due to a loss of platelet-derived growth factor alpha receptor expression as beta cells age but appears to be unrelated to GLP-1 receptor expression levels (Chen et al., 2011 ). Indeed, the GLP-1 receptor is highly expressed in adult beta cells, and GLP-1 secretion increases insulin secretion, as detailed previously; however, the induction of proliferative factors such as nuclear factor of activated T cells, cytoplasmic 1; forkhead box protein 1; and cyclin A1 is only seen in juvenile islets (Dai et al., 2017 ). Human studies using cadaveric pancreata from pregnant donors also showed increased beta cell mass, yet lactogenic hormones from the pituitary or placenta (prolactin, placental lactogen, or growth hormone) are unable to stimulate proliferation in human beta cells despite their ability to produce robust proliferation in mouse beta cells (reviewed in Baeyens et al., 2016 ). Experiments overexpressing mouse versus human signal transducer and activator of transcription 5, the final signaling factor inducing beta cell adaptation, in human beta cells allows for prolactin-mediated proliferation revealing fundamental differences in prolactin pathway competency in human (Chen et al., 2015 ). Overcoming the barrier of recapitulating human pregnancy’s effect on beta cells through isolating placental cells or blood serum during pregnancy may result in the discovery of a factor(s) that facilitates the increase in beta cell mass observed during human pregnancy.

Mechanisms that stimulate beta cell proliferation have also been discovered from studying genetic mutations that result in insulinomas, spontaneous insulin-producing beta cell adenomas. The most common hereditary mutation occurs in the multiple endocrine neoplasia type 1 (MEN1) gene. Indeed, administration of a MEN1 inhibitor in addition to a GLP-1 agonist (which cannot induce proliferation alone) is able to increase beta cell proliferation in isolated human islets through synergistic activation of KRAS proto-oncogene, GTPase downstream signals (Chamberlain et al., 2014 ). Interestingly, MEN1 mutations are uncommon in sporadic insulinomas, yet assaying genomic and epigenetic changes in a large cohort of non-MEN1 insulinomas found alterations in trithorax and polycomb chromatin modifying genes that were functionally related to MEN1 (Wang et al., 2017 ). Stewart and colleagues hypothesized that changes in histone 3 lysine 27 and histone 3 lysine 4 methylation status led to increased enhancer of zeste homolog 2 and lysine demethylase 6A, decreased cyclin-dependent kinase inhibitor 1C, and thereby increased beta cell proliferation, among other phenotypes. They also proposed that these findings help to explain why increased proliferation always occurs despite broad heterogeneity of mutations found between individual insulinomas (Wang et al., 2017 ).

Although factors that induce proliferation are continuing to be discovered, there are drawbacks that still limit their clinical application. Harmine and other DYRK1A inhibitors are not beta cell specific, nor have all their cellular targets been determined (Ackeifi et al., 2020 ). Targeting other pathways to induce human beta cell proliferation such as modulation of prostaglandin E2 receptors (i.e., inhibition of prostaglandin E receptor 3 alone or in combination with prostaglandin E receptor 4 activation) showed promising increases in proliferative rate yet suffers from the same lack of specificity (Carboneau et al., 2017 ). Induction of proliferation may also come at the expense of glucose sensing as in insulinomas, which have an increased expression of “disallowed genes” and alterations in glucose transporter and hexokinase expression (Wang et al., 2017 ). A further untoward consequence that must be avoided is the production of cancerous cells through unchecked proliferation. Finally, increasing beta cell mass through low rates of proliferation may increase the pool of functional insulin-secreting cells in T2D, but without additional measures, these beta cells will still ultimately be targeted for immune cell destruction in T1D.

3. Beta Cell Stress Relieving Therapies

Metabolic, inflammatory, and endoplasmic reticulum (ER) stress contribute to beta cell dysfunction and failure in both T1D and T2D. Although reduction of metabolic overload of beta cells by early exogenous insulin therapy or insulin sensitizers can temporarily reduce loss of beta cell mass/function early in diabetes, a focus on relieving ER and inflammatory stress is also of interest to preserve beta cell health.

ER stress is a well known contributor to beta cell demise both in T1D and T2D (Laybutt et al., 2007 ; Marchetti et al., 2007 ; Marhfour et al., 2012 ; Tersey et al., 2012 ) and a target of interest in the prevention of beta cell loss in both diseases. Preclinical studies suggest that the use of chemical chaperones, including 4-phenylbutyric acid and tauroursodeoxycholic acid (TUDCA), to alleviate ER stress improves beta cell function and insulin sensitivity in mouse models of T2D (Cnop et al., 2017 ; Ozcan et al., 2006 ). Furthermore, TUDCA has been shown to preserve beta cell mass and reduce ER stress in mouse models of T1D (Engin et al., 2013 ). Interestingly, TUDCA has shown promise at improving insulin action in obese nondiabetic human subjects, yet beta cell function and insulin secretion were not assessed (Kars et al., 2010 ). A clinical trial regarding the use of TUDCA for humans with new-onset T1D is also ongoing ( {"type":"clinical-trial","attrs":{"text":"NCT02218619","term_id":"NCT02218619"}} NCT02218619 ). However, a note of caution regarding use of ER chaperones is that they may prevent low level ER stress necessary to potentiate beta cell replication during states of increased insulin demand (Sharma et al., 2015 ), suggesting that the broad use of ER chaperone therapies should be carefully considered.

The blockade of inflammatory stress has long been an area of interest for treatments of both T1D and T2D (Donath et al., 2019 ; Eguchi and Nagai, 2017 ). Indeed, use of nonsteroidal anti-inflammatory drugs (NSAIDs), which block cyclooxygenase, have been observed to improve metabolic control in patients with diabetes since the turn of the 20th century (Williamson, 1901 ). Salicylates have been shown to improve insulin secretion and beta cell function in both obese human subjects and those with T2D (Fernandez-Real et al., 2008; Giugliano et al., 1985 ). However, another NSAID, salsalate, has not been shown to improve beta cell function while improving other metabolic outcomes (Kim et al., 2014 ; Penesova et al., 2015 ), possibly suggesting distinct mechanisms of action for anti-inflammatory compounds. The regular use of NSAIDs to enhance metabolic outcomes is also often limited to the tolerability of long-term use of these agents due to adverse effects. Recently, golilumab, a monoclonal antibody against the proinflammatory cytokine tumor necrosis factor alpha, was demonstrated to improve beta cell function in new-onset T1D, suggesting that targeting the underlying inflammatory milieu may have benefits to preserve beta cell mass and function in T1D (Quattrin et al., 2020). Taken together, both new and old approaches to target beta cell stressors still remain of long-term interest to improve beta cell viability and function in both T1D and T2D.

3. New Players to Induce Islet Immune Protection

Countless researchers have expended intense industry to determine T1D disease etiology and treatments focused on immunotherapy and tolerogenic methods. Multiple, highly comprehensive reviews are available describing these efforts (Goudy and Tisch, 2005 ; Rewers and Gottlieb, 2009 ; Stojanovic et al., 2017 ). Here we will focus on the protection of beta cells through programmed cell death protein-1 ligand (PD-L1) overexpression, major histocompatibility complex class I, A, B, C (HLA-A,B,C) mutated human embryonic stem cell–derived beta cells, and islet encapsulation methods.

Cancer immunotherapies that block immune checkpoints are beneficial for treating advanced stage cancers, yet induction of autoimmune diseases, including T1D, remains a potential side effect (Stamatouli et al., 2018 ; Perdigoto et al., 2019 ). A subset of these drugs target either the programmed cell death-1 protein on the surface of activated T lymphocytes or its receptor PD-L1 (Stamatouli et al., 2018 ; Perdigoto et al., 2019 ). PD-L1 expression was found in insulin-positive beta cells from T1D but not insulin-negative islets or nondiabetic islets, leading to the hypothesis that PD-L1 is upregulated in an attempt to drive immune cell attenuation (Osum et al., 2018 ; Colli et al., 2018 ). Adenoviral overexpression of PD-L1 specifically in beta cells rescued hyperglycemia in the NOD mouse model of T1D, but these animals eventually succumbed to diabetes by the study’s termination (El Khatib et al., 2015 ). A more promising report from Ben Nasr et al. ( 2017 ) demonstrated that pharmacologically or genetically induced overexpression of PD-L1 in hematopoietic stem and progenitor cells inhibited beta cell autoimmunity in the NOD mouse as well as in vitro using human hematopoietic stem and progenitor cells from patients with T1D.

As mentioned above, islet transplantation to treat T1D is limited by islet availability, cost, and the requirement for continuous immunosuppression. Islet cells generated by differentiating embryonic or induced pluripotent stem (iPS) cells could circumvent these limitations. Ideally, iPS-derived beta cells could be manipulated to eliminate the expression of polymorphic HLA-A,B,C molecules, which were found to be upregulated in T1D beta cells (Bottazzo et al., 1985 ; Richardson et al., 2016 ). These molecules allow peptide presentation to CD8+ T cells or cytotoxic T lymphocytes and may lead to beta cell removal. Interestingly, remaining insulin-positive cells in T1D donor pancreas are not HLA-A,B,C positive (Nejentsev et al., 2007; Rodriguez-Calvo et al., 2015 ). However, current differentiation protocols are still limited in their ability to produce fully glucose-responsive beta cells without transplantation into animal models to induce mature characteristics. Additionally, use of iPS-derived beta cells will still lead to concerns regarding DNA mutagenesis resulting from the methods used to obtain pluripotency or teratoma formation from cells that have escaped differentiation.

Encapsulation devices would protect islets or stem cells from immune cell infiltration while allowing for the proper exchange of nutrients and hormones. Macroencapsulation uses removable devices that would help assuage fears surrounding mutation or tumor formation; indeed, the first human trial using encapsulated hESC-derived beta cells will be completed in January 2021 ( {"type":"clinical-trial","attrs":{"text":"NCT02239354","term_id":"NCT02239354"}} NCT02239354 ). Macroencapsulation of islets prior to transplantation using various alginate-based hydrogels has historically been impeded by a strong in vivo foreign body immune response (Desai and Shea, 2017 ; Doloff et al., 2017 ; Pueyo et al., 1993 ). More recently, chemically modified forms of alginate that avoid macrophage recognition and fibrous deposition have been successfully used in rodents and for up to 6 months in nonhuman primates (Vegas et al., 2016 ). Indeed, Bochenek et al. ( 2018 ) successfully transplanted alginate protected islets for 4 months without immunosuppression in the bursa omentalis of nonhuman primates demonstrating the feasibility for this approach to be extended to humans. It remains to be seen if these devices will be successful for long-term use, perhaps decades, in patients with diabetes.

III. Summary

Although existing drug therapies using classic oral antidiabetic drugs like sulfonylureas and metformin or injected insulin remain mainstays of diabetes treatment, newer drugs based on incretin hormone actions or SGLT2 inhibitors have increased the pharmacological armamentarium available to diabetologists ( Fig. 1 ). However, the explosion of progress in beta cell biology has identified potential avenues that can increase beta cell mass in sophisticated ways by employing stem cell differentiation or enhancement of beta cell proliferation. Taken together, there should be optimism that the increased incidence of both T1D and T2D is being matched by the creativity and hard work of the diabetes research community.

Abbreviations

Authorship contributions.

Wrote and contributed to the writing of the manuscript: Satin, Soleimanpour, Walker

This work was supported by the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) [Grant R01-DK46409] (to L.S.S.), [Grant R01-DK108921] (to S.A.S.), and [Grant P30-DK020572 pilot and feasibility grant] (to S.A.S.), the Juvenile Diabetes Research Foundation (JDRF) [Grant CDA-2016-189] (to L.S.S. and S.A.S.), [Grant SRA-2018-539] (to S.A.S.), and [Grant COE-2019-861] (to S.A.S.), and the US Department of Veterans Affairs [Grant I01 BX004444] (to S.A.S.). The JDRF Career Development Award to S.A.S. is partly supported by the Danish Diabetes Academy and the Novo Nordisk Foundation.

https://doi.org/10.1124/pharmrev.120.000160

• Ackeifi C, Swartz E, Kumar K, Liu H, Chalada S, Karakose E, Scott DK, Garcia-Ocaña A, Sanchez R, DeVita RJ, et al. (2020) Pharmacologic and genetic approaches define human pancreatic β cell mitogenic targets of DYRK1A inhibitors . JCI Insight 5 :e132594. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Adachi T, Yasuda K, Okamoto Y, Shihara N, Oku A, Ueta K, Kitamura K, Saito A, Iwakura I, Yamada Y, et al. (2000) T-1095, a renal Na+-glucose transporter inhibitor, improves hyperglycemia in streptozotocin-induced diabetic rats . Metabolism 49 :990–995. [ PubMed ] [ Google Scholar ]
• Alexiadou K, Tan TM (2020) Gastrointestinal peptides as therapeutic targets to mitigate obesity and metabolic syndrome . Curr Diab Rep 20 :26. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Ambery P, Parker VE, Stumvoll M, Posch MG, Heise T, Plum-Moerschel L, Tsai LF, Robertson D, Jain M, Petrone M, et al. (2018) MEDI0382, a GLP-1 and glucagon receptor dual agonist, in obese or overweight patients with type 2 diabetes: a randomised, controlled, double-blind, ascending dose and phase 2a study . Lancet 391 :2607–2618. [ PubMed ] [ Google Scholar ]
• Ashcroft FM, Rorsman P (1989) Electrophysiology of the pancreatic beta-cell . Prog Biophys Mol Biol 54 :87–143. [ PubMed ] [ Google Scholar ]
• Ashcroft FM, Rorsman P (1990) ATP-sensitive K+ channels: a link between B-cell metabolism and insulin secretion . Biochem Soc Trans 18 :109–111. [ PubMed ] [ Google Scholar ]
• Baeyens L, Hindi S, Sorenson RL, German MS (2016) β-Cell adaptation in pregnancy . Diabetes Obes Metab 18 ( Suppl 1 ):63–70. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bailey CJ (1992) Biguanides and NIDDM . Diabetes Care 15 :755–772. [ PubMed ] [ Google Scholar ]
• Ben Nasr M, Tezza S, D’Addio F, Mameli C, Usuelli V, Maestroni A, Corradi D, Belletti S, Albarello L, Becchi G, et al. (2017) PD-L1 genetic overexpression or pharmacological restoration in hematopoietic stem and progenitor cells reverses autoimmune diabetes . Sci Transl Med 9 :eaam7543. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bergmann NC, Lund A, Gasbjerg LS, Meessen ECE, Andersen MM, Bergmann S, Hartmann B, Holst JJ, Jessen L, Christensen MB, et al. (2019) Effects of combined GIP and GLP-1 infusion on energy intake, appetite and energy expenditure in overweight/obese individuals: a randomised, crossover study . Diabetologia 62 :665–675. [ PubMed ] [ Google Scholar ]
• Bernal-Mizrachi E, Kulkarni RN, Scott DK, Mauvais-Jarvis F, Stewart AF, Garcia-Ocaña A (2014) Human β-cell proliferation and intracellular signaling part 2: still driving in the dark without a road map . Diabetes 63 :819–831. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bethel MA, Patel RA, Merrill P, Lokhnygina Y, Buse JB, Mentz RJ, Pagidipati NJ, Chan JC, Gustavson SM, Iqbal N, et al.; EXSCEL Study Group (2018) Cardiovascular outcomes with glucagon-like peptide-1 receptor agonists in patients with type 2 diabetes: a meta-analysis . Lancet Diabetes Endocrinol 6 :105–113. [ PubMed ] [ Google Scholar ]
• Bliss M (1982) Banting’s, Best’s, and Collip’s accounts of the discovery of insulin . Bull Hist Med 56 :554–568. [ PubMed ] [ Google Scholar ]
• Bochenek MA, Veiseh O, Vegas AJ, McGarrigle JJ, Qi M, Marchese E, Omami M, Doloff JC, Mendoza-Elias J, Nourmohammadzadeh M, et al. (2018) Alginate encapsulation as long-term immune protection of allogeneic pancreatic islet cells transplanted into the omental bursa of macaques . Nat Biomed Eng 2 :810–821. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bonner C, Kerr-Conte J, Gmyr V, Queniat G, Moerman E, Thévenet J, Beaucamps C, Delalleau N, Popescu I, Malaisse WJ, et al. (2015) Inhibition of the glucose transporter SGLT2 with dapagliflozin in pancreatic alpha cells triggers glucagon secretion . Nat Med 21 :512–517. [ PubMed ] [ Google Scholar ]
• Bonora E, Cigolini M, Bosello O, Zancanaro C, Capretti L, Zavaroni I, Coscelli C, Butturini U (1984) Lack of effect of intravenous metformin on plasma concentrations of glucose, insulin, C-peptide, glucagon and growth hormone in non-diabetic subjects . Curr Med Res Opin 9 :47–51. [ PubMed ] [ Google Scholar ]
• Bottazzo GF, Dean BM, McNally JM, MacKay EH, Swift PG, Gamble DR (1985) In situ characterization of autoimmune phenomena and expression of HLA molecules in the pancreas in diabetic insulitis . N Engl J Med 313 :353–360. [ PubMed ] [ Google Scholar ]
• Boyle JP, Thompson TJ, Gregg EW, Barker LE, Williamson DF (2010) Projection of the year 2050 burden of diabetes in the US adult population: dynamic modeling of incidence, mortality, and prediabetes prevalence . Popul Health Metr 8 :29. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Brunham LR, Kruit JK, Pape TD, Timmins JM, Reuwer AQ, Vasanji Z, Marsh BJ, Rodrigues B, Johnson JD, Parks JS, et al. (2007) Beta-cell ABCA1 influences insulin secretion, glucose homeostasis and response to thiazolidinedione treatment . Nat Med 13 :340–347. [ PubMed ] [ Google Scholar ]
• Buse JB, DeFronzo RA, Rosenstock J, Kim T, Burns C, Skare S, Baron A, Fineman M (2016) The primary glucose-lowering effect of metformin resides in the gut, not the circulation: results from short-term pharmacokinetic and 12-week dose-ranging studies . Diabetes Care 39 :198–205. [ PubMed ] [ Google Scholar ]
• Buse JB, Henry RR, Han J, Kim DD, Fineman MS, Baron AD; Exenatide-113 Clinical Study Group (2004) Effects of exenatide (exendin-4) on glycemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes . Diabetes Care 27 :2628–2635. [ PubMed ] [ Google Scholar ]
• Capozzi ME, Svendsen B, Encisco SE, Lewandowski SL, Martin MD, Lin H, Jaffe JL, Coch RW, Haldeman JM, MacDonald PE, et al. (2019a) β Cell tone is defined by proglucagon peptides through cAMP signaling . JCI Insight 4 :e126742. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Capozzi ME, Wait JB, Koech J, Gordon AN, Coch RW, Svendsen B, Finan B, D’Alessio DA, Campbell JE (2019b) Glucagon lowers glycemia when β-cells are active . JCI Insight 5 :e129954. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Carboneau BA, Allan JA, Townsend SE, Kimple ME, Breyer RM, Gannon M (2017) Opposing effects of prostaglandin E 2 receptors EP3 and EP4 on mouse and human β-cell survival and proliferation . Mol Metab 6 :548–559. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chae H, Augustin R, Gatineau E, Mayoux E, Bensellam M, Antoine N, Khattab F, Lai BK, Brusa D, Stierstorfer B, et al. (2020) SGLT2 is not expressed in pancreatic α- and β-cells, and its inhibition does not directly affect glucagon and insulin secretion in rodents and humans . Mol Metab 42 :101071. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chamberlain CE, Scheel DW, McGlynn K, Kim H, Miyatsuka T, Wang J, Nguyen V, Zhao S, Mavropoulos A, Abraham AG, et al. (2014) Menin determines K-RAS proliferative outputs in endocrine cells . J Clin Invest 124 :4093–4101. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chen H, Gu X, Liu Y, Wang J, Wirt SE, Bottino R, Schorle H, Sage J, Kim SK (2011) PDGF signalling controls age-dependent proliferation in pancreatic β-cells . Nature 478 :349–355. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chen H, Kleinberger JW, Takane KK, Salim F, Fiaschi-Taesch N, Pappas K, Parsons R, Jiang J, Zhang Y, Liu H, et al. (2015) Augmented Stat5 signaling bypasses multiple impediments to lactogen-mediated proliferation in human β-cells . Diabetes 64 :3784–3797. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chen J, Cha-Molstad H, Szabo A, Shalev A (2009) Diabetes induces and calcium channel blockers prevent cardiac expression of proapoptotic thioredoxin-interacting protein . Am J Physiol Endocrinol Metab 296 :E1133–E1139. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chen J, Hui ST, Couto FM, Mungrue IN, Davis DB, Attie AD, Lusis AJ, Davis RA, Shalev A (2008a) Thioredoxin-interacting protein deficiency induces Akt/Bcl-xL signaling and pancreatic beta-cell mass and protects against diabetes . FASEB J 22 :3581–3594. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chen J, Saxena G, Mungrue IN, Lusis AJ, Shalev A (2008b) Thioredoxin-interacting protein: a critical link between glucose toxicity and beta-cell apoptosis . Diabetes 57 :938–944. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Chou DH, Webber MJ, Tang BC, Lin AB, Thapa LS, Deng D, Truong JV, Cortinas AB, Langer R, Anderson DG (2015) Glucose-responsive insulin activity by covalent modification with aliphatic phenylboronic acid conjugates . Proc Natl Acad Sci USA 112 :2401–2406. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Cnop M, Toivonen S, Igoillo-Esteve M, Salpea P (2017) Endoplasmic reticulum stress and eIF2α phosphorylation: the Achilles heel of pancreatic β cells . Mol Metab 6 :1024–1039. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Cohen MA, Ellis SM, Le Roux CW, Batterham RL, Park A, Patterson M, Frost GS, Ghatei MA, Bloom SR (2003) Oxyntomodulin suppresses appetite and reduces food intake in humans . J Clin Endocrinol Metab 88 :4696–4701. [ PubMed ] [ Google Scholar ]
• Coll AP, Chen M, Taskar P, Rimmington D, Patel S, Tadross JA, Cimino I, Yang M, Welsh P, Virtue S, et al. (2020) GDF15 mediates the effects of metformin on body weight and energy balance . Nature 578 :444–448. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Colli ML, Hill JLE, Marroquí L, Chaffey J, Dos Santos RS, Leete P, Coomans de Brachène A, Paula FMM, Op de Beeck A, Castela A, et al. (2018) PDL1 is expressed in the islets of people with type 1 diabetes and is up-regulated by interferons-α and-γ via IRF1 induction . EBioMedicine 36 :367–375. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• RISE Consortium (2019) Lack of durable improvements in β-cell function following withdrawal of pharmacological interventions in adults with impaired glucose tolerance or recently diagnosed type 2 diabetes . Diabetes Care 42 :1742–1751. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Dai C, Hang Y, Shostak A, Poffenberger G, Hart N, Prasad N, Phillips N, Levy SE, Greiner DL, Shultz LD, et al. (2017) Age-dependent human β cell proliferation induced by glucagon-like peptide 1 and calcineurin signaling . J Clin Invest 127 :3835–3844. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Dai C, Kayton NS, Shostak A, Poffenberger G, Cyphert HA, Aramandla R, Thompson C, Papagiannis IG, Emfinger C, Shiota M, et al. (2016) Stress-impaired transcription factor expression and insulin secretion in transplanted human islets . J Clin Invest 126 :1857–1870. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Dai C, Walker JT, Shostak A, Bouchi Y, Poffenberger G, Hart NJ, Jacobson DA, Calcutt MW, Bottino R, Greiner DL, et al. (2020) Dapagliflozin does not directly affect human α or β cells . Endocrinology 161 :bqaa080. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Dakin CL, Gunn I, Small CJ, Edwards CM, Hay DL, Smith DM, Ghatei MA, Bloom SR (2001) Oxyntomodulin inhibits food intake in the rat . Endocrinology 142 :4244–4250. [ PubMed ] [ Google Scholar ]
• Dakin CL, Small CJ, Park AJ, Seth A, Ghatei MA, Bloom SR (2002) Repeated ICV administration of oxyntomodulin causes a greater reduction in body weight gain than in pair-fed rats . Am J Physiol Endocrinol Metab 283 :E1173–E1177. [ PubMed ] [ Google Scholar ]
• Defronzo RA, Tripathy D, Schwenke DC, Banerji M, Bray GA, Buchanan TA, Clement SC, Gastaldelli A, Henry RR, Kitabchi AE, et al.; ACT NOW Study (2013) Prevention of diabetes with pioglitazone in ACT NOW: physiologic correlates . Diabetes 62 :3920–3926. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Delaunay F, Khan A, Cintra A, Davani B, Ling ZC, Andersson A, Ostenson CG, Gustafsson J, Efendic S, Okret S (1997) Pancreatic beta cells are important targets for the diabetogenic effects of glucocorticoids . J Clin Invest 100 :2094–2098. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Desai T, Shea LD (2017) Advances in islet encapsulation technologies . Nat Rev Drug Discov 16 :338–350. [ PubMed ] [ Google Scholar ]
• Devaraj S, Venkatachalam A, Chen X (2016) Metformin and the gut microbiome in diabetes . Clin Chem 62 :1554–1555. [ PubMed ] [ Google Scholar ]
• Doloff JC, Veiseh O, Vegas AJ, Tam HH, Farah S, Ma M, Li J, Bader A, Chiu A, Sadraei A, et al. (2017) Colony stimulating factor-1 receptor is a central component of the foreign body response to biomaterial implants in rodents and non-human primates . Nat Mater 16 :671–680. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Donath MY, Dinarello CA, Mandrup-Poulsen T (2019) Targeting innate immune mediators in type 1 and type 2 diabetes . Nat Rev Immunol 19 :734–746. [ PubMed ] [ Google Scholar ]
• Drucker DJ (2018) Mechanisms of action and therapeutic application of glucagon-like peptide-1 . Cell Metab 27 :740–756. [ PubMed ] [ Google Scholar ]
• Drucker DJ, Philippe J, Mojsov S, Chick WL, Habener JF (1987) Glucagon-like peptide I stimulates insulin gene expression and increases cyclic AMP levels in a rat islet cell line . Proc Natl Acad Sci USA 84 :3434–3438. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Duca FA, Côté CD, Rasmussen BA, Zadeh-Tahmasebi M, Rutter GA, Filippi BM, Lam TK (2015) Metformin activates a duodenal Ampk-dependent pathway to lower hepatic glucose production in rats . Nat Med 21 :506–511. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Eguchi K, Nagai R (2017) Islet inflammation in type 2 diabetes and physiology . J Clin Invest 127 :14–23. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• El Khatib MM, Sakuma T, Tonne JM, Mohamed MS, Holditch SJ, Lu B, Kudva YC, Ikeda Y (2015) β-Cell-targeted blockage of PD1 and CTLA4 pathways prevents development of autoimmune diabetes and acute allogeneic islets rejection . Gene Ther 22 :430–438. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• El-Khatib FH, Balliro C, Hillard MA, Magyar KL, Ekhlaspour L, Sinha M, Mondesir D, Esmaeili A, Hartigan C, Thompson MJ, et al. (2017) Home use of a bihormonal bionic pancreas versus insulin pump therapy in adults with type 1 diabetes: a multicentre randomised crossover trial . Lancet 389 :369–380. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• El-Mir MY, Detaille D, R-Villanueva G, Delgado-Esteban M, Guigas B, Attia S, Fontaine E, Almeida A, Leverve X (2008) Neuroprotective role of antidiabetic drug metformin against apoptotic cell death in primary cortical neurons . J Mol Neurosci 34 :77–87. [ PubMed ] [ Google Scholar ]
• El-Mir MY, Nogueira V, Fontaine E, Avéret N, Rigoulet M, Leverve X (2000) Dimethylbiguanide inhibits cell respiration via an indirect effect targeted on the respiratory chain complex I . J Biol Chem 275 :223–228. [ PubMed ] [ Google Scholar ]
• Elrick H, Stimmler L, Hlad CJ Jr, Arai Y (1964) Plasma Insulin Response to Oral and Intravenous Glucose Administration . J Clin Endocrinol Metab 24 :1076–1082. [ PubMed ] [ Google Scholar ]
• Engin F, Yermalovich A, Nguyen T, Hummasti S, Fu W, Eizirik DL, Mathis D, Hotamisligil GS (2013) Restoration of the unfolded protein response in pancreatic β cells protects mice against type 1 diabetes [published correction appears in Sci Transl Med (2013) 5 :214er11] . Sci Transl Med 5 :211ra156. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Evans JM, Donnelly LA, Emslie-Smith AM, Alessi DR, Morris AD (2005) Metformin and reduced risk of cancer in diabetic patients . BMJ 330 :1304–1305. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Evans-Molina C, Robbins RD, Kono T, Tersey SA, Vestermark GL, Nunemaker CS, Garmey JC, Deering TG, Keller SR, Maier B, et al. (2009) Peroxisome proliferator-activated receptor gamma activation restores islet function in diabetic mice through reduction of endoplasmic reticulum stress and maintenance of euchromatin structure . Mol Cell Biol 29 :2053–2067. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Fernández-Real JM, López-Bermejo A, Ropero AB, Piquer S, Nadal A, Bassols J, Casamitjana R, Gomis R, Arnaiz E, Pérez I, et al. (2008) Salicylates increase insulin secretion in healthy obese subjects . J Clin Endocrinol Metab 93 :2523–2530. [ PubMed ] [ Google Scholar ]
• Foretz M, Guigas B, Bertrand L, Pollak M, Viollet B (2014) Metformin: from mechanisms of action to therapies . Cell Metab 20 :953–966. [ PubMed ] [ Google Scholar ]
• Foretz M, Guigas B, Viollet B (2019) Understanding the glucoregulatory mechanisms of metformin in type 2 diabetes mellitus . Nat Rev Endocrinol 15 :569–589. [ PubMed ] [ Google Scholar ]
• Foretz M, Hébrard S, Leclerc J, Zarrinpashneh E, Soty M, Mithieux G, Sakamoto K, Andreelli F, Viollet B (2010) Metformin inhibits hepatic gluconeogenesis in mice independently of the LKB1/AMPK pathway via a decrease in hepatic energy state . J Clin Invest 120 :2355–2369. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Frias JPBastyr EJ 3rd, Vignati L, Tschöp MH, Schmitt C, Owen K, Christensen RHDiMarchi RD (2017) The sustained effects of a dual GIP/GLP-1 receptor agonist, NNC0090-2746, in patients with type 2 diabetes . Cell Metab 26 :343–352.e2. [ PubMed ] [ Google Scholar ]
• Frias JP, Nauck MA, Van J, Benson C, Bray R, Cui X, Milicevic Z, Urva S, Haupt A, Robins DA (2020) Efficacy and tolerability of tirzepatide, a dual glucose-dependent insulinotropic peptide and glucagon-like peptide-1 receptor agonist in patients with type 2 diabetes: A 12-week, randomized, double-blind, placebo-controlled study to evaluate different dose-escalation regimens . Diabetes Obes Metab 22 :938–946. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Frias JP, Nauck MA, Van J, Kutner ME, Cui X, Benson C, Urva S, Gimeno RE, Milicevic Z, Robins D, et al. (2018) Efficacy and safety of LY3298176, a novel dual GIP and GLP-1 receptor agonist, in patients with type 2 diabetes: a randomised, placebo-controlled and active comparator-controlled phase 2 trial . Lancet 392 :2180–2193. [ PubMed ] [ Google Scholar ]
• Fullerton MD, Galic S, Marcinko K, Sikkema S, Pulinilkunnil T, Chen ZP, O’Neill HM, Ford RJ, Palanivel R, O’Brien M, et al. (2013) Single phosphorylation sites in Acc1 and Acc2 regulate lipid homeostasis and the insulin-sensitizing effects of metformin . Nat Med 19 :1649–1654. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Gerstein HC, Colhoun HM, Dagenais GR, Diaz R, Lakshmanan M, Pais P, Probstfield J, Riesmeyer JS, Riddle MC, Rydén L, et al.; REWIND Investigators (2019) Dulaglutide and cardiovascular outcomes in type 2 diabetes (REWIND): a double-blind, randomised placebo-controlled trial . Lancet 394 :121–130. [ PubMed ] [ Google Scholar ]
• Giugliano D, Ceriello A, Saccomanno F, Quatraro A, Paolisso G, D’Onofrio F (1985) Effects of salicylate, tolbutamide, and prostaglandin E2 on insulin responses to glucose in noninsulin-dependent diabetes mellitus . J Clin Endocrinol Metab 61 :160–166. [ PubMed ] [ Google Scholar ]
• Goudy KS, Tisch R (2005) Immunotherapy for the prevention and treatment of type 1 diabetes . Int Rev Immunol 24 :307–326. [ PubMed ] [ Google Scholar ]
• Gregg BE, Moore PC, Demozay D, Hall BA, Li M, Husain A, Wright AJ, Atkinson MA, Rhodes CJ (2012) Formation of a human β-cell population within pancreatic islets is set early in life . J Clin Endocrinol Metab 97 :3197–3206. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Grempler R, Thomas L, Eckhardt M, Himmelsbach F, Sauer A, Sharp DE, Bakker RA, Mark M, Klein T, Eickelmann P (2012) Empagliflozin, a novel selective sodium glucose cotransporter-2 (SGLT-2) inhibitor: characterisation and comparison with other SGLT-2 inhibitors . Diabetes Obes Metab 14 :83–90. [ PubMed ] [ Google Scholar ]
• Harlan DM (2016) Islet transplantation for hypoglycemia unawareness/severe hypoglycemia: caveat emptor . Diabetes Care 39 :1072–1074. [ PubMed ] [ Google Scholar ]
• Harrower AD (1991) Efficacy of gliclazide in comparison with other sulphonylureas in the treatment of NIDDM . Diabetes Res Clin Pract 14 ( Suppl 2 ):S65–S67. [ PubMed ] [ Google Scholar ]
• He L, Wondisford FE (2015) Metformin action: concentrations matter . Cell Metab 21 :159–162. [ PubMed ] [ Google Scholar ]
• Hedrington MS, Davis SN (2019) Considerations when using alpha-glucosidase inhibitors in the treatment of type 2 diabetes . Expert Opin Pharmacother 20 :2229–2235. [ PubMed ] [ Google Scholar ]
• Hering BJ, Clarke WR, Bridges ND, Eggerman TL, Alejandro R, Bellin MD, Chaloner K, Czarniecki CW, Goldstein JS, Hunsicker LG, et al.; Clinical Islet Transplantation Consortium (2016) Phase 3 trial of transplantation of human islets in type 1 diabetes complicated by severe hypoglycemia . Diabetes Care 39 :1230–1240. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hering BJ, Kandaswamy R, Ansite JD, Eckman PM, Nakano M, Sawada T, Matsumoto I, Ihm SH, Zhang HJ, Parkey J, et al. (2005) Single-donor, marginal-dose islet transplantation in patients with type 1 diabetes . JAMA 293 :830–835. [ PubMed ] [ Google Scholar ]
• Hernandez AFGreen JBJanmohamed SD’Agostino RB Sr , Granger CB, Jones NP, Leiter LA, Rosenberg AE, Sigmon KN, Somerville MCet al.; Harmony Outcomes committees and investigators (2018) Albiglutide and cardiovascular outcomes in patients with type 2 diabetes and cardiovascular disease (Harmony Outcomes): a double-blind, randomised placebo-controlled trial . Lancet 392 :1519–1529. [ PubMed ] [ Google Scholar ]
• Hiatt WR, Kaul S, Smith RJ (2013) The cardiovascular safety of diabetes drugs--insights from the rosiglitazone experience . N Engl J Med 369 :1285–1287. [ PubMed ] [ Google Scholar ]
• Hirshberg B, Preston EH, Xu H, Tal MG, Neeman Z, Bunnell D, Soleimanpour S, Hale DA, Kirk AD, Harlan DM (2003) Rabbit antithymocyte globulin induction and sirolimus monotherapy supports prolonged islet allograft function in a nonhuman primate islet transplantation model . Transplantation 76 :55–60. [ PubMed ] [ Google Scholar ]
• Hu W, Jiang C, Guan D, Dierickx P, Zhang R, Moscati A, Nadkarni GN, Steger DJ, Loos RJF, Hu C, et al. (2019) Patient adipose stem cell-derived adipocytes reveal genetic variation that predicts antidiabetic drug response . Cell Stem Cell 24 :299–308.e6. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Husain M, Birkenfeld AL, Donsmark M, Dungan K, Eliaschewitz FG, Franco DR, Jeppesen OK, Lingvay I, Mosenzon O, Pedersen SD, et al.; PIONEER 6 Investigators (2019) Oral semaglutide and cardiovascular outcomes in patients with type 2 diabetes . N Engl J Med 381 :841–851. [ PubMed ] [ Google Scholar ]
• Imamura M, Nakanishi K, Suzuki T, Ikegai K, Shiraki R, Ogiyama T, Murakami T, Kurosaki E, Noda A, Kobayashi Y, et al. (2012) Discovery of Ipragliflozin (ASP1941): a novel C-glucoside with benzothiophene structure as a potent and selective sodium glucose co-transporter 2 (SGLT2) inhibitor for the treatment of type 2 diabetes mellitus . Bioorg Med Chem 20 :3263–3279. [ PubMed ] [ Google Scholar ]
• Kahn SE, Haffner SM, Heise MA, Herman WH, Holman RR, Jones NP, Kravitz BG, Lachin JM, O’Neill MC, Zinman B, et al.; ADOPT Study Group (2006) Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy . N Engl J Med 355 :2427–2443. [ PubMed ] [ Google Scholar ]
• Kahn SE, Lachin JM, Zinman B, Haffner SM, Aftring RP, Paul G, Kravitz BG, Herman WH, Viberti G, Holman RR; ADOPT Study Group (2011) Effects of rosiglitazone, glyburide, and metformin on β-cell function and insulin sensitivity in ADOPT . Diabetes 60 :1552–1560. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kars M, Yang L, Gregor MF, Mohammed BS, Pietka TA, Finck BN, Patterson BW, Horton JD, Mittendorfer B, Hotamisligil GS, et al. (2010) Tauroursodeoxycholic acid may improve liver and muscle but not adipose tissue insulin sensitivity in obese men and women . Diabetes 59 :1899–1905. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kemp CB, Knight MJ, Scharp DW, Lacy PE, Ballinger WF (1973) Transplantation of isolated pancreatic islets into the portal vein of diabetic rats . Nature 244 :447. [ PubMed ] [ Google Scholar ]
• Kim HI, Cha JY, Kim SY, Kim JW, Roh KJ, Seong JK, Lee NT, Choi KY, Kim KS, Ahn YH (2002) Peroxisomal proliferator-activated receptor-gamma upregulates glucokinase gene expression in beta-cells . Diabetes 51 :676–685. [ PubMed ] [ Google Scholar ]
• Kim HI, Kim JW, Kim SH, Cha JY, Kim KS, Ahn YH (2000) Identification and functional characterization of the peroxisomal proliferator response element in rat GLUT2 promoter . Diabetes 49 :1517–1524. [ PubMed ] [ Google Scholar ]
• Kim SH, Liu A, Ariel D, Abbasi F, Lamendola C, Grove K, Tomasso V, Ochoa H, Reaven G (2014) Effect of salsalate on insulin action, secretion, and clearance in nondiabetic, insulin-resistant individuals: a randomized, placebo-controlled study . Diabetes Care 37 :1944–1950. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Koffert JP, Mikkola K, Virtanen KA, Andersson AD, Faxius L, Hällsten K, Heglind M, Guiducci L, Pham T, Silvola JMU, et al. (2017) Metformin treatment significantly enhances intestinal glucose uptake in patients with type 2 diabetes: Results from a randomized clinical trial . Diabetes Res Clin Pract 131 :208–216. [ PubMed ] [ Google Scholar ]
• Koh A, Mannerås-Holm L, Yunn NO, Nilsson PM, Ryu SH, Molinaro A, Perkins R, Smith JG, Bäckhed F (2020) Microbial imidazole propionate affects responses to metformin through p38γ-dependent inhibitory AMPK phosphorylation . Cell Metab 32 :643–653.e4. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kreymann B, Williams G, Ghatei MA, Bloom SR (1987) Glucagon-like peptide-1 7-36: a physiological incretin in man . Lancet 2 :1300–1304. [ PubMed ] [ Google Scholar ]
• Kuhre RE, Ghiasi SM, Adriaenssens AE, Wewer Albrechtsen NJ, Andersen DB, Aivazidis A, Chen L, Mandrup-Poulsen T, Ørskov C, Gribble FM, et al. (2019) No direct effect of SGLT2 activity on glucagon secretion . Diabetologia 62 :1011–1023. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kulkarni RN, Mizrachi EB, Ocana AG, Stewart AF (2012) Human β-cell proliferation and intracellular signaling: driving in the dark without a road map . Diabetes 61 :2205–2213. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kurosaki E, Ogasawara H (2013) Ipragliflozin and other sodium-glucose cotransporter-2 (SGLT2) inhibitors in the treatment of type 2 diabetes: preclinical and clinical data . Pharmacol Ther 139 :51–59. [ PubMed ] [ Google Scholar ]
• Kurtzhals P, Schäffer L, Sørensen A, Kristensen C, Jonassen I, Schmid C, Trüb T (2000) Correlations of receptor binding and metabolic and mitogenic potencies of insulin analogs designed for clinical use . Diabetes 49 :999–1005. [ PubMed ] [ Google Scholar ]
• Lambeir AM, Scharpé S, De Meester I (2008) DPP4 inhibitors for diabetes--what next? Biochem Pharmacol 76 :1637–1643. [ PubMed ] [ Google Scholar ]
• Laybutt DR, Preston AM, Akerfeldt MC, Kench JG, Busch AK, Biankin AV, Biden TJ (2007) Endoplasmic reticulum stress contributes to beta cell apoptosis in type 2 diabetes . Diabetologia 50 :752–763. [ PubMed ] [ Google Scholar ]
• Lebovitz HE (2019) Thiazolidinediones: the forgotten diabetes medications . Curr Diab Rep 19 :151. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lewis JD, Habel LA, Quesenberry CP, Strom BL, Peng T, Hedderson MM, Ehrlich SF, Mamtani R, Bilker W, Vaughn DJ, et al. (2015) Pioglitazone use and risk of bladder cancer and other common cancers in persons with diabetes . JAMA 314 :265–277. [ PubMed ] [ Google Scholar ]
• Li L, Li S, Deng K, Liu J, Vandvik PO, Zhao P, Zhang L, Shen J, Bala MM, Sohani ZN, et al. (2016) Dipeptidyl peptidase-4 inhibitors and risk of heart failure in type 2 diabetes: systematic review and meta-analysis of randomised and observational studies . BMJ 352 :i610. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Link JT (2003) Pharmacological regulation of hepatic glucose production . Curr Opin Investig Drugs 4 :421–429. [ PubMed ] [ Google Scholar ]
• Luippold G, Klein T, Mark M, Grempler R (2012) Empagliflozin, a novel potent and selective SGLT-2 inhibitor, improves glycaemic control alone and in combination with insulin in streptozotocin-induced diabetic rats, a model of type 1 diabetes mellitus . Diabetes Obes Metab 14 :601–607. [ PubMed ] [ Google Scholar ]
• Maganti AV, Tersey SA, Syed F, Nelson JB, Colvin SC, Maier B, Mirmira RG (2016) Peroxisome proliferator-activated receptor-γ activation augments the β-cell unfolded protein response and rescues early glycemic deterioration and β cell death in non-obese diabetic mice . J Biol Chem 291 :22524–22533. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Makrilakis K (2019) The role of DPP-4 inhibitors in the treatment algorithm of type 2 diabetes mellitus: when to select, what to expect . Int J Environ Res Public Health 16 :2720. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Marchetti P, Bugliani M, Lupi R, Marselli L, Masini M, Boggi U, Filipponi F, Weir GC, Eizirik DL, Cnop M (2007) The endoplasmic reticulum in pancreatic beta cells of type 2 diabetes patients . Diabetologia 50 :2486–2494. [ PubMed ] [ Google Scholar ]
• Marhfour I, Lopez XM, Lefkaditis D, Salmon I, Allagnat F, Richardson SJ, Morgan NG, Eizirik DL (2012) Expression of endoplasmic reticulum stress markers in the islets of patients with type 1 diabetes . Diabetologia 55 :2417–2420. [ PubMed ] [ Google Scholar ]
• Marso SP, Bain SC, Consoli A, Eliaschewitz FG, Jódar E, Leiter LA, Lingvay I, Rosenstock J, Seufert J, Warren ML, et al.; SUSTAIN-6 Investigators (2016a) Semaglutide and cardiovascular outcomes in patients with type 2 diabetes . N Engl J Med 375 :1834–1844. [ PubMed ] [ Google Scholar ]
• Marso SP, Daniels GH, Brown-Frandsen K, Kristensen P, Mann JF, Nauck MA, Nissen SE, Pocock S, Poulter NR, Ravn LS, et al.; LEADER Steering Committee; LEADER Trial Investigators (2016b) Liraglutide and cardiovascular outcomes in type 2 diabetes . N Engl J Med 375 :311–322. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Massollo M, Marini C, Brignone M, Emionite L, Salani B, Riondato M, Capitanio S, Fiz F, Democrito A, Amaro A, et al. (2013) Metformin temporal and localized effects on gut glucose metabolism assessed using 18F-FDG PET in mice . J Nucl Med 54 :259–266. [ PubMed ] [ Google Scholar ]
• McCreight LJ, Bailey CJ, Pearson ER (2016) Metformin and the gastrointestinal tract . Diabetologia 59 :426–435. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• McIntyre N, Holdsworth CD, Turner DS (1964) New interpretation of oral glucose tolerance . Lancet 2 :20–21. [ PubMed ] [ Google Scholar ]
• McMurray JJV, Solomon SD, Inzucchi SE, Køber L, Kosiborod MN, Martinez FA, Ponikowski P, Sabatine MS, Anand IS, Bělohlávek J, et al.; DAPA-HF Trial Committees and Investigators (2019) Dapagliflozin in patients with heart failure and reduced ejection fraction . N Engl J Med 381 :1995–2008. [ PubMed ] [ Google Scholar ]
• Meier JJ, Butler AE, Saisho Y, Monchamp T, Galasso R, Bhushan A, Rizza RA, Butler PC (2008) Beta-cell replication is the primary mechanism subserving the postnatal expansion of beta-cell mass in humans . Diabetes 57 :1584–1594. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Meng W, Ellsworth BA, Nirschl AA, McCann PJ, Patel M, Girotra RN, Wu G, Sher PM, Morrison EP, Biller SA, et al. (2008) Discovery of dapagliflozin: a potent, selective renal sodium-dependent glucose cotransporter 2 (SGLT2) inhibitor for the treatment of type 2 diabetes . J Med Chem 51 :1145–1149. [ PubMed ] [ Google Scholar ]
• Menting JG, Whittaker J, Margetts MB, Whittaker LJ, Kong GK, Smith BJ, Watson CJ, Záková L, Kletvíková E, Jiráček J, et al. (2013) How insulin engages its primary binding site on the insulin receptor . Nature 493 :241–245. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Miller RA, Chu Q, Xie J, Foretz M, Viollet B, Birnbaum MJ (2013) Biguanides suppress hepatic glucagon signalling by decreasing production of cyclic AMP . Nature 494 :256–260. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Minkowski O (1892) Weitere Mitteilungen über den Diabetes mellitus nach Extirpation des Pankreas . Berliner Klinische Wochenschrift 29 :90–93. [ Google Scholar ]
• Misbin RI (2004) The phantom of lactic acidosis due to metformin in patients with diabetes . Diabetes Care 27 :1791–1793. [ PubMed ] [ Google Scholar ]
• Mudaliar S, Armstrong DA, Mavian AA, O’Connor-Semmes R, Mydlow PK, Ye J, Hussey EK, Nunez DJ, Henry RR, Dobbins RL (2012) Remogliflozin etabonate, a selective inhibitor of the sodium-glucose transporter 2, improves serum glucose profiles in type 1 diabetes . Diabetes Care 35 :2198–2200. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Mulherin AJ, Oh AH, Kim H, Grieco A, Lauffer LM, Brubaker PL (2011) Mechanisms underlying metformin-induced secretion of glucagon-like peptide-1 from the intestinal L cell . Endocrinology 152 :4610–4619. [ PubMed ] [ Google Scholar ]
• Müller TD, Finan B, Bloom SR, D’Alessio D, Drucker DJ, Flatt PR, Fritsche A, Gribble F, Grill HJ, Habener JF, et al. (2019) Glucagon-like peptide 1 (GLP-1) . Mol Metab 30 :72–130. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Munir KM, Lamos EM (2017) Diabetes type 2 management: what are the differences between DPP-4 inhibitors and how do you choose? Expert Opin Pharmacother 18 :839–841. [ PubMed ] [ Google Scholar ]
• Naftanel MA, Harlan DM (2004) Pancreatic islet transplantation . PLoS Med 1 :e58. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Nauck M, Stöckmann F, Ebert R, Creutzfeldt W (1986a) Reduced incretin effect in type 2 (non-insulin-dependent) diabetes . Diabetologia 29 :46–52. [ PubMed ] [ Google Scholar ]
• Nauck MA, Homberger E, Siegel EG, Allen RC, Eaton RP, Ebert R, Creutzfeldt W (1986b) Incretin effects of increasing glucose loads in man calculated from venous insulin and C-peptide responses . J Clin Endocrinol Metab 63 :492–498. [ PubMed ] [ Google Scholar ]
• Nejentsev S, Howson JM, Walker NM, Szeszko J, Field SF, Stevens HE, Reynolds P, Hardy M, King E, Masters J, et al.; Wellcome Trust Case Control Consortium (2007) Localization of type 1 diabetes susceptibility to the MHC class I genes HLA-B and HLA-A . Nature 450 :887–892. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Nichols CG (2006) KATP channels as molecular sensors of cellular metabolism . Nature 440 :470–476. [ PubMed ] [ Google Scholar ]
• Nomura S, Sakamaki S, Hongu M, Kawanishi E, Koga Y, Sakamoto T, Yamamoto Y, Ueta K, Kimata H, Nakayama K, et al. (2010) Discovery of canagliflozin, a novel C-glucoside with thiophene ring, as sodium-dependent glucose cotransporter 2 inhibitor for the treatment of type 2 diabetes mellitus . J Med Chem 53 :6355–6360. [ PubMed ] [ Google Scholar ]
• Ohnishi ST, Endo M, editors. (1981) The Mechanism of Gated Calcium Transport Across Biological Membranes , Academic Press, New York. [ Google Scholar ]
• Osipovich AB, Stancill JS, Cartailler JP, Dudek KD, Magnuson MA (2020) Excitotoxicity and overnutrition additively impair metabolic function and identity of pancreatic β-cells . Diabetes 69 :1476–1491. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Osum KC, Burrack AL, Martinov T, Sahli NL, Mitchell JS, Tucker CG, Pauken KE, Papas K, Appakalai B, Spanier JA, et al. (2018) Interferon-gamma drives programmed death-ligand 1 expression on islet β cells to limit T cell function during autoimmune diabetes . Sci Rep 8 :8295. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Owen MR, Doran E, Halestrap AP (2000) Evidence that metformin exerts its anti-diabetic effects through inhibition of complex 1 of the mitochondrial respiratory chain . Biochem J 348 :607–614. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Ozanne SE, Guest PC, Hutton JC, Hales CN (1995) Intracellular localization and molecular heterogeneity of the sulphonylurea receptor in insulin-secreting cells . Diabetologia 38 :277–282. [ PubMed ] [ Google Scholar ]
• Ozcan U, Yilmaz E, Ozcan L, Furuhashi M, Vaillancourt E, Smith RO, Görgün CZ, Hotamisligil GS (2006) Chemical chaperones reduce ER stress and restore glucose homeostasis in a mouse model of type 2 diabetes . Science 313 :1137–1140. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Paty BW, Harmon JS, Marsh CL, Robertson RP (2002) Inhibitory effects of immunosuppressive drugs on insulin secretion from HIT-T15 cells and Wistar rat islets . Transplantation 73 :353–357. [ PubMed ] [ Google Scholar ]
• Penesova A, Koska J, Ortega E, Bunt JC, Bogardus C, de Courten B (2015) Salsalate has no effect on insulin secretion but decreases insulin clearance: a randomized, placebo-controlled trial in subjects without diabetes . Diabetes Obes Metab 17 :608–612. [ PubMed ] [ Google Scholar ]
• Perdigoto AL, Quandt Z, Anderson M, Herold KC (2019) Checkpoint inhibitor-induced insulin-dependent diabetes: an emerging syndrome . Lancet Diabetes Endocrinol 7 :421–423. [ PubMed ] [ Google Scholar ]
• Perley MJ, Kipnis DM (1967) Plasma insulin responses to oral and intravenous glucose: studies in normal and diabetic subjects . J Clin Invest 46 :1954–1962. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Pernicova I, Korbonits M (2014) Metformin--mode of action and clinical implications for diabetes and cancer . Nat Rev Endocrinol 10 :143–156. [ PubMed ] [ Google Scholar ]
• Preiss D, Dawed A, Welsh P, Heggie A, Jones AG, Dekker J, Koivula R, Hansen TH, Stewart C, Holman RR, et al.; DIRECT consortium group (2017) Sustained influence of metformin therapy on circulating glucagon-like peptide-1 levels in individuals with and without type 2 diabetes . Diabetes Obes Metab 19 :356–363. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Prentki M, Nolan CJ (2006) Islet beta cell failure in type 2 diabetes . J Clin Invest 116 :1802–1812. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Pueyo ME, Darquy S, Capron F, Reach G (1993) In vitro activation of human macrophages by alginate-polylysine microcapsules . J Biomater Sci Polym Ed 5 :197–203. [ PubMed ] [ Google Scholar ]
• Pybus F (1924) Notes on suprarenal and pancreatic grafting . Lancet 204 :550–551. [ Google Scholar ]
• Quattrin T, Haller MJ, Steck AK, Felner EI, Li Y, Xia Y, Leu JH, Zoka R, Hedrick JA, Rigby MR, et al.; T1GER Study Investigators (2020) Golimumab and Beta-Cell Function in Youth with New-Onset Type 1 Diabetes . N Engl J Med 383 :2007–2017. [ PubMed ] [ Google Scholar ]
• Rachdi L, Kariyawasam D, Aïello V, Herault Y, Janel N, Delabar JM, Polak M, Scharfmann R (2014a) Dyrk1A induces pancreatic β cell mass expansion and improves glucose tolerance . Cell Cycle 13 :2221–2229. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rachdi L, Kariyawasam D, Guez F, Aïello V, Arbonés ML, Janel N, Delabar JM, Polak M, Scharfmann R (2014b) Dyrk1a haploinsufficiency induces diabetes in mice through decreased pancreatic beta cell mass . Diabetologia 57 :960–969. [ PubMed ] [ Google Scholar ]
• Rege NK, Phillips NFB, Weiss MA (2017) Development of glucose-responsive ‘smart’ insulin systems . Curr Opin Endocrinol Diabetes Obes 24 :267–278. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rena G, Hardie DG, Pearson ER (2017) The mechanisms of action of metformin . Diabetologia 60 :1577–1585. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rewers M, Gottlieb P (2009) Immunotherapy for the prevention and treatment of type 1 diabetes: human trials and a look into the future . Diabetes Care 32 :1769–1782. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Richardson SJ, Rodriguez-Calvo T, Gerling IC, Mathews CE, Kaddis JS, Russell MA, Zeissler M, Leete P, Krogvold L, Dahl-Jørgensen K, et al. (2016) Islet cell hyperexpression of HLA class I antigens: a defining feature in type 1 diabetes . Diabetologia 59 :2448–2458. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rickels MR, Liu C, Shlansky-Goldberg RD, Soleimanpour SA, Vivek K, Kamoun M, Min Z, Markmann E, Palangian M, Dalton-Bakes C, et al. (2013) Improvement in β-cell secretory capacity after human islet transplantation according to the c7 protocol . Diabetes 62 :2890–2897. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rickels MR, Robertson RP (2019) Pancreatic islet transplantation in humans: recent progress and future directions . Endocr Rev 40 :631–668. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rodriguez-Calvo T, Suwandi JS, Amirian N, Zapardiel-Gonzalo J, Anquetil F, Sabouri S, von Herrath MG (2015) Heterogeneity and lobularity of pancreatic pathology in type 1 diabetes during the prediabetic phase . J Histochem Cytochem 63 :626–636. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rojas LB, Gomes MB (2013) Metformin: an old but still the best treatment for type 2 diabetes . Diabetol Metab Syndr 5 :6. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rosen ED, Kulkarni RN, Sarraf P, Ozcan U, Okada T, Hsu CH, Eisenman D, Magnuson MA, Gonzalez FJ, Kahn CR, et al. (2003) Targeted elimination of peroxisome proliferator-activated receptor gamma in beta cells leads to abnormalities in islet mass without compromising glucose homeostasis . Mol Cell Biol 23 :7222–7229. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rosenstock J, Hassman DR, Madder RD, Brazinsky SA, Farrell J, Khutoryansky N, Hale PM; Repaglinide Versus Nateglinide Comparison Study Group (2004) Repaglinide versus nateglinide monotherapy: a randomized, multicenter study . Diabetes Care 27 :1265–1270. [ PubMed ] [ Google Scholar ]
• Sampaio MS, Kuo HT, Bunnapradist S (2011) Outcomes of simultaneous pancreas-kidney transplantation in type 2 diabetic recipients . Clin J Am Soc Nephrol 6 :1198–1206. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Saponaro C, Gmyr V, Thévenet J, Moerman E, Delalleau N, Pasquetti G, Coddeville A, Quenon A, Daoudi M, Hubert T, et al. (2019) The GLP1R agonist liraglutide reduces hyperglucagonemia induced by the SGLT2 inhibitor dapagliflozin via somatostatin release . Cell Rep 28 :1447–1454.e4. [ PubMed ] [ Google Scholar ]
• Saponaro C, Mühlemann M, Acosta-Montalvo A, Piron A, Gmyr V, Delalleau N, Moerman E, Thévenet J, Pasquetti G, Coddeville A, et al. (2020) Interindividual heterogeneity of SGLT2 expression and function in human pancreatic islets . Diabetes 69 :902–914. [ PubMed ] [ Google Scholar ]
• Satin LS, Tavalin SJ, Kinard TA, Teague J (1995) Contribution of L- and non-L-type calcium channels to voltage-gated calcium current and glucose-dependent insulin secretion in HIT-T15 cells . Endocrinology 136 :4589–4601. [ PubMed ] [ Google Scholar ]
• Schwartz AV, Chen H, Ambrosius WT, Sood A, Josse RG, Bonds DE, Schnall AM, Vittinghoff E, Bauer DC, Banerji MA, et al. (2015) Effects of TZD use and discontinuation on fracture rates in ACCORD Bone Study . J Clin Endocrinol Metab 100 :4059–4066. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Shalev A, Pise-Masison CA, Radonovich M, Hoffmann SC, Hirshberg B, Brady JN, Harlan DM (2002) Oligonucleotide microarray analysis of intact human pancreatic islets: identification of glucose-responsive genes and a highly regulated TGFbeta signaling pathway . Endocrinology 143 :3695–3698. [ PubMed ] [ Google Scholar ]
• Shankar SS, Shankar RR, Mixson LA, Miller DL, Pramanik B, O’Dowd AK, Williams DM, Frederick CB, Beals CR, Stoch SA, et al. (2018) Native oxyntomodulin has significant glucoregulatory effects independent of weight loss in obese humans with and without type 2 diabetes . Diabetes 67 :1105–1112. [ PubMed ] [ Google Scholar ]
• Shapiro AM, Lakey JR, Ryan EA, Korbutt GS, Toth E, Warnock GL, Kneteman NM, Rajotte RV (2000) Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen . N Engl J Med 343 :230–238. [ PubMed ] [ Google Scholar ]
• Sharma RB, O’Donnell AC, Stamateris RE, Ha B, McCloskey KM, Reynolds PR, Arvan P, Alonso LC (2015) Insulin demand regulates β cell number via the unfolded protein response . J Clin Invest 125 :3831–3846. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sims EK, Mirmira RG, Evans-Molina C (2020) The role of beta-cell dysfunction in early type 1 diabetes . Curr Opin Endocrinol Diabetes Obes 27 :215–224. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Soccio RE, Chen ER, Rajapurkar SR, Safabakhsh P, Marinis JM, Dispirito JR, Emmett MJ, Briggs ER, Fang B, Everett LJ, et al. (2015) Genetic variation determines PPARγ function and anti-diabetic drug response in vivo . Cell 162 :33–44. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Soleimanpour SA, Crutchlow MF, Ferrari AM, Raum JC, Groff DN, Rankin MM, Liu C, De León DD, Naji A, Kushner JA, et al. (2010) Calcineurin signaling regulates human islet beta-cell survival . J Biol Chem 285 :40050–40059. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Soleimanpour SA, Hirshberg B, Bunnell DJ, Sumner AE, Ader M, Remaley AT, Rother KI, Rickels MR, Harlan DM (2012) Metabolic function of a suboptimal transplanted islet mass in nonhuman primates on rapamycin monotherapy . Cell Transplant 21 :1297–1304. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Soleimanpour SA, Stoffers DA (2013) The pancreatic β cell and type 1 diabetes: innocent bystander or active participant? Trends Endocrinol Metab 24 :324–331. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Stamatouli AM, Quandt Z, Perdigoto AL, Clark PL, Kluger H, Weiss SA, Gettinger S, Sznol M, Young A, Rushakoff R, et al. (2018) Collateral damage: insulin-dependent diabetes induced with checkpoint inhibitors . Diabetes 67 :1471–1480. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Stancill JS, Cartailler JP, Clayton HW, O’Connor JT, Dickerson MT, Dadi PK, Osipovich AB, Jacobson DA, Magnuson MA (2017) Chronic β-cell depolarization impairs β-cell identity by disrupting a network of Ca 2+ -regulated genes . Diabetes 66 :2175–2187. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Stewart AF, Hussain MA, García-Ocaña A, Vasavada RC, Bhushan A, Bernal-Mizrachi E, Kulkarni RN (2015) Human β-cell proliferation and intracellular signaling: part 3 . Diabetes 64 :1872–1885. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Stojanovic I, Dimitrijevic M, Vives-Pi M, Mansilla MJ, Pujol-Autonell I, Rodríguez-Fernandez S, Palova-Jelínkova L, Funda DP, Gruden-Movsesijan A, Sofronic-Milosavljevic L, et al. (2017) Cell-based tolerogenic therapy, experience from animal models of multiple sclerosis, type 1 diabetes and rheumatoid arthritis . Curr Pharm Des 23 :2623–2643. [ PubMed ] [ Google Scholar ]
• Sturek JM, Castle JD, Trace AP, Page LC, Castle AM, Evans-Molina C, Parks JS, Mirmira RG, Hedrick CC (2010) An intracellular role for ABCG1-mediated cholesterol transport in the regulated secretory pathway of mouse pancreatic beta cells . J Clin Invest 120 :2575–2589. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Suga T, Kikuchi O, Kobayashi M, Matsui S, Yokota-Hashimoto H, Wada E, Kohno D, Sasaki T, Takeuchi K, Kakizaki S, et al. (2019) SGLT1 in pancreatic α cells regulates glucagon secretion in mice, possibly explaining the distinct effects of SGLT2 inhibitors on plasma glucagon levels . Mol Metab 19 :1–12. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sun L, Xie C, Wang G, Wu Y, Wu Q, Wang X, Liu J, Deng Y, Xia J, Chen B, et al. (2018) Gut microbiota and intestinal FXR mediate the clinical benefits of metformin . Nat Med 24 :1919–1929. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sutherland DE, Matas AJ, Najarian JS (1978) Pancreatic islet cell transplantation . Surg Clin North Am 58 :365–382. [ PubMed ] [ Google Scholar ]
• Tersey SA, Nishiki Y, Templin AT, Cabrera SM, Stull ND, Colvin SC, Evans-Molina C, Rickus JL, Maier B, Mirmira RG (2012) Islet β-cell endoplasmic reticulum stress precedes the onset of type 1 diabetes in the nonobese diabetic mouse model . Diabetes 61 :818–827. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Thielen LA, Chen J, Jing G, Moukha-Chafiq O, Xu G, Jo S, Grayson TB, Lu B, Li P, Augelli-Szafran CE, et al. (2020) Identification of an anti-diabetic, orally available small molecule that regulates TXNIP expression and glucagon action . Cell Metab 32 :353–365.e8. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Tillner J, Posch MG, Wagner F, Teichert L, Hijazi Y, Einig C, Keil S, Haack T, Wagner M, Bossart M, et al. (2019) A novel dual glucagon-like peptide and glucagon receptor agonist SAR425899: Results of randomized, placebo-controlled first-in-human and first-in-patient trials . Diabetes Obes Metab 21 :120–128. [ PubMed ] [ Google Scholar ]
• Vallon V (2015) The mechanisms and therapeutic potential of SGLT2 inhibitors in diabetes mellitus . Annu Rev Med 66 :255–270. [ PubMed ] [ Google Scholar ]
• Vasseur M, Debuyser A, Joffre M (1987) Sensitivity of pancreatic beta cell to calcium channel blockers. An electrophysiologic study of verapamil and nifedipine . Fundam Clin Pharmacol 1 :95–113. [ PubMed ] [ Google Scholar ]
• Vegas AJ, Veiseh O, Doloff JC, Ma M, Tam HH, Bratlie K, Li J, Bader AR, Langan E, Olejnik K, et al. (2016) Combinatorial hydrogel library enables identification of materials that mitigate the foreign body response in primates . Nat Biotechnol 34 :345–352. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Vergari E, Knudsen JG, Ramracheya R, Salehi A, Zhang Q, Adam J, Asterholm IW, Benrick A, Briant LJB, Chibalina MV, et al. (2019) Insulin inhibits glucagon release by SGLT2-induced stimulation of somatostatin secretion . Nat Commun 10 :139. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Wallner K, Shapiro AM, Senior PA, McCabe C (2016) Cost effectiveness and value of information analyses of islet cell transplantation in the management of ‘unstable’ type 1 diabetes mellitus . BMC Endocr Disord 16 :17. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Wang H, Bender A, Wang P, Karakose E, Inabnet WB, Libutti SK, Arnold A, Lambertini L, Stang M, Chen H, et al. (2017) Insights into beta cell regeneration for diabetes via integration of molecular landscapes in human insulinomas . Nat Commun 8 :767. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Wang Y, An H, Liu T, Qin C, Sesaki H, Guo S, Radovick S, Hussain M, Maheshwari A, Wondisford FE, O’Rourke B, He L (2019) Metformin improves mitochondrial respiratory activity through activation of AMPK . Cell Rep 29 :1511–1523.e5. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Westerman J, Wirtz KW, Berkhout T, van Deenen LL, Radhakrishnan R, Khorana HG (1983) Identification of the lipid-binding site of phosphatidylcholine-transfer protein with phosphatidylcholine analogs containing photoactivable carbene precursors . Eur J Biochem 132 :441–449. [ PubMed ] [ Google Scholar ]
• World Health Organization (2020) World Health Organization Diabetes Fact Sheet . [ Google Scholar ]
• Willard FS, Douros JD, Gabe MB, Showalter AD, Wainscott DB, Suter TM, Capozzi ME, van der Velden WJ, Stutsman C, Cardona GR, et al. (2020) Tirzepatide is an imbalanced and biased dual GIP and GLP-1 receptor agonist . JCI Insight 5 :e140532. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Williams P (1894) Notes on diabetes treated with extract and by grafts of sheep’s pancreas . BMJ 2 :1303–1304. [ Google Scholar ]
• Williamson RT (1901) On the treatment of glycosuria and diabetes mellitus with sodium salicylate . BMJ 1 :760–762. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Witters LA (2001) The blooming of the French lilac . J Clin Invest 108 :1105–1107. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Wiviott SD, Raz I, Bonaca MP, Mosenzon O, Kato ET, Cahn A, Silverman MG, Zelniker TA, Kuder JF, Murphy SA, et al.; DECLARE–TIMI 58 Investigators (2019) Dapagliflozin and cardiovascular outcomes in type 2 diabetes . N Engl J Med 380 :347–357. [ PubMed ] [ Google Scholar ]
• Wu H, Esteve E, Tremaroli V, Khan MT, Caesar R, Mannerås-Holm L, Ståhlman M, Olsson LM, Serino M, Planas-Fèlix M, et al. (2017) Metformin alters the gut microbiome of individuals with treatment-naive type 2 diabetes, contributing to the therapeutic effects of the drug . Nat Med 23 :850–858. [ PubMed ] [ Google Scholar ]
• Wynne K, Park AJ, Small CJ, Patterson M, Ellis SM, Murphy KG, Wren AM, Frost GS, Meeran K, Ghatei MA, et al. (2005) Subcutaneous oxyntomodulin reduces body weight in overweight and obese subjects: a double-blind, randomized, controlled trial . Diabetes 54 :2390–2395. [ PubMed ] [ Google Scholar ]
• Xu G, Chen J, Jing G, Shalev A (2012) Preventing β-cell loss and diabetes with calcium channel blockers . Diabetes 61 :848–856. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Yang JF, Gong X, Bakh NA, Carr K, Phillips NFB, Ismail-Beigi F, Weiss MA, Strano MS (2020) Connecting rodent and human pharmacokinetic models for the design and translation of glucose-responsive insulin . Diabetes 69 :1815–1826. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Yu O, Azoulay L, Yin H, Filion KB, Suissa S (2018) Sulfonylureas as initial treatment for type 2 diabetes and the risk of severe hypoglycemia . Am J Med 131 :317.e11–317.e22. [ PubMed ] [ Google Scholar ]
• Zhou G, Myers R, Li Y, Chen Y, Shen X, Fenyk-Melody J, Wu M, Ventre J, Doebber T, Fujii N, et al. (2001) Role of AMP-activated protein kinase in mechanism of metformin action . J Clin Invest 108 :1167–1174. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Zinman B, Wanner C, Lachin JM, Fitchett D, Bluhmki E, Hantel S, Mattheus M, Devins T, Johansen OE, Woerle HJ, et al.; EMPA-REG OUTCOME Investigators (2015) Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes . N Engl J Med 373 :2117–2128. [ PubMed ] [ Google Scholar ]

Cultural Relativity and Acceptance of Embryonic Stem Cell Research

Article sidebar.

Main Article Content

There is a debate about the ethical implications of using human embryos in stem cell research, which can be influenced by cultural, moral, and social values. This paper argues for an adaptable framework to accommodate diverse cultural and religious perspectives. By using an adaptive ethics model, research protections can reflect various populations and foster growth in stem cell research possibilities.

INTRODUCTION

Stem cell research combines biology, medicine, and technology, promising to alter health care and the understanding of human development. Yet, ethical contention exists because of individuals’ perceptions of using human embryos based on their various cultural, moral, and social values. While these disagreements concerning policy, use, and general acceptance have prompted the development of an international ethics policy, such a uniform approach can overlook the nuanced ethical landscapes between cultures. With diverse viewpoints in public health, a single global policy, especially one reflecting Western ethics or the ethics prevalent in high-income countries, is impractical. This paper argues for a culturally sensitive, adaptable framework for the use of embryonic stem cells. Stem cell policy should accommodate varying ethical viewpoints and promote an effective global dialogue. With an extension of an ethics model that can adapt to various cultures, we recommend localized guidelines that reflect the moral views of the people those guidelines serve.

Stem cells, characterized by their unique ability to differentiate into various cell types, enable the repair or replacement of damaged tissues. Two primary types of stem cells are somatic stem cells (adult stem cells) and embryonic stem cells. Adult stem cells exist in developed tissues and maintain the body’s repair processes. [1] Embryonic stem cells (ESC) are remarkably pluripotent or versatile, making them valuable in research. [2] However, the use of ESCs has sparked ethics debates. Considering the potential of embryonic stem cells, research guidelines are essential. The International Society for Stem Cell Research (ISSCR) provides international stem cell research guidelines. They call for “public conversations touching on the scientific significance as well as the societal and ethical issues raised by ESC research.” [3] The ISSCR also publishes updates about culturing human embryos 14 days post fertilization, suggesting local policies and regulations should continue to evolve as ESC research develops. [4]  Like the ISSCR, which calls for local law and policy to adapt to developing stem cell research given cultural acceptance, this paper highlights the importance of local social factors such as religion and culture.

I.     Global Cultural Perspective of Embryonic Stem Cells

Views on ESCs vary throughout the world. Some countries readily embrace stem cell research and therapies, while others have stricter regulations due to ethical concerns surrounding embryonic stem cells and when an embryo becomes entitled to moral consideration. The philosophical issue of when the “someone” begins to be a human after fertilization, in the morally relevant sense, [5] impacts when an embryo becomes not just worthy of protection but morally entitled to it. The process of creating embryonic stem cell lines involves the destruction of the embryos for research. [6] Consequently, global engagement in ESC research depends on social-cultural acceptability.

a.     US and Rights-Based Cultures

In the United States, attitudes toward stem cell therapies are diverse. The ethics and social approaches, which value individualism, [7] trigger debates regarding the destruction of human embryos, creating a complex regulatory environment. For example, the 1996 Dickey-Wicker Amendment prohibited federal funding for the creation of embryos for research and the destruction of embryos for “more than allowed for research on fetuses in utero.” [8] Following suit, in 2001, the Bush Administration heavily restricted stem cell lines for research. However, the Stem Cell Research Enhancement Act of 2005 was proposed to help develop ESC research but was ultimately vetoed. [9] Under the Obama administration, in 2009, an executive order lifted restrictions allowing for more development in this field. [10] The flux of research capacity and funding parallels the different cultural perceptions of human dignity of the embryo and how it is socially presented within the country’s research culture. [11]

b.     Ubuntu and Collective Cultures

African bioethics differs from Western individualism because of the different traditions and values. African traditions, as described by individuals from South Africa and supported by some studies in other African countries, including Ghana and Kenya, follow the African moral philosophies of Ubuntu or Botho and Ukama , which “advocates for a form of wholeness that comes through one’s relationship and connectedness with other people in the society,” [12] making autonomy a socially collective concept. In this context, for the community to act autonomously, individuals would come together to decide what is best for the collective. Thus, stem cell research would require examining the value of the research to society as a whole and the use of the embryos as a collective societal resource. If society views the source as part of the collective whole, and opposes using stem cells, compromising the cultural values to pursue research may cause social detachment and stunt research growth. [13] Based on local culture and moral philosophy, the permissibility of stem cell research depends on how embryo, stem cell, and cell line therapies relate to the community as a whole . Ubuntu is the expression of humanness, with the person’s identity drawn from the “’I am because we are’” value. [14] The decision in a collectivistic culture becomes one born of cultural context, and individual decisions give deference to others in the society.

Consent differs in cultures where thought and moral philosophy are based on a collective paradigm. So, applying Western bioethical concepts is unrealistic. For one, Africa is a diverse continent with many countries with different belief systems, access to health care, and reliance on traditional or Western medicines. Where traditional medicine is the primary treatment, the “’restrictive focus on biomedically-related bioethics’” [is] problematic in African contexts because it neglects bioethical issues raised by traditional systems.” [15] No single approach applies in all areas or contexts. Rather than evaluating the permissibility of ESC research according to Western concepts such as the four principles approach, different ethics approaches should prevail.

Another consideration is the socio-economic standing of countries. In parts of South Africa, researchers have not focused heavily on contributing to the stem cell discourse, either because it is not considered health care or a health science priority or because resources are unavailable. [16] Each country’s priorities differ given different social, political, and economic factors. In South Africa, for instance, areas such as maternal mortality, non-communicable diseases, telemedicine, and the strength of health systems need improvement and require more focus. [17] Stem cell research could benefit the population, but it also could divert resources from basic medical care. Researchers in South Africa adhere to the National Health Act and Medicines Control Act in South Africa and international guidelines; however, the Act is not strictly enforced, and there is no clear legislation for research conduct or ethical guidelines. [18]

Some parts of Africa condemn stem cell research. For example, 98.2 percent of the Tunisian population is Muslim. [19] Tunisia does not permit stem cell research because of moral conflict with a Fatwa. Religion heavily saturates the regulation and direction of research. [20] Stem cell use became permissible for reproductive purposes only recently, with tight restrictions preventing cells from being used in any research other than procedures concerning ART/IVF.  Their use is conditioned on consent, and available only to married couples. [21] The community's receptiveness to stem cell research depends on including communitarian African ethics.

c.     Asia

Some Asian countries also have a collective model of ethics and decision making. [22] In China, the ethics model promotes a sincere respect for life or human dignity, [23] based on protective medicine. This model, influenced by Traditional Chinese Medicine (TCM), [24] recognizes Qi as the vital energy delivered via the meridians of the body; it connects illness to body systems, the body’s entire constitution, and the universe for a holistic bond of nature, health, and quality of life. [25] Following a protective ethics model, and traditional customs of wholeness, investment in stem cell research is heavily desired for its applications in regenerative therapies, disease modeling, and protective medicines. In a survey of medical students and healthcare practitioners, 30.8 percent considered stem cell research morally unacceptable while 63.5 percent accepted medical research using human embryonic stem cells. Of these individuals, 89.9 percent supported increased funding for stem cell research. [26] The scientific community might not reflect the overall population. From 1997 to 2019, China spent a total of $576 million (USD) on stem cell research at 8,050 stem cell programs, increased published presence from 0.6 percent to 14.01 percent of total global stem cell publications as of 2014, and made significant strides in cell-based therapies for various medical conditions. [27] However, while China has made substantial investments in stem cell research and achieved notable progress in clinical applications, concerns linger regarding ethical oversight and transparency. [28] For example, the China Biosecurity Law, promoted by the National Health Commission and China Hospital Association, attempted to mitigate risks by introducing an institutional review board (IRB) in the regulatory bodies. 5800 IRBs registered with the Chinese Clinical Trial Registry since 2021. [29] However, issues still need to be addressed in implementing effective IRB review and approval procedures.

The substantial government funding and focus on scientific advancement have sometimes overshadowed considerations of regional cultures, ethnic minorities, and individual perspectives, particularly evident during the one-child policy era. As government policy adapts to promote public stability, such as the change from the one-child to the two-child policy, [30] research ethics should also adapt to ensure respect for the values of its represented peoples.

Japan is also relatively supportive of stem cell research and therapies. Japan has a more transparent regulatory framework, allowing for faster approval of regenerative medicine products, which has led to several advanced clinical trials and therapies. [31] South Korea is also actively engaged in stem cell research and has a history of breakthroughs in cloning and embryonic stem cells. [32] However, the field is controversial, and there are issues of scientific integrity. For example, the Korean FDA fast-tracked products for approval, [33] and in another instance, the oocyte source was unclear and possibly violated ethical standards. [34] Trust is important in research, as it builds collaborative foundations between colleagues, trial participant comfort, open-mindedness for complicated and sensitive discussions, and supports regulatory procedures for stakeholders. There is a need to respect the culture’s interest, engagement, and for research and clinical trials to be transparent and have ethical oversight to promote global research discourse and trust.

d.     Middle East

Countries in the Middle East have varying degrees of acceptance of or restrictions to policies related to using embryonic stem cells due to cultural and religious influences. Saudi Arabia has made significant contributions to stem cell research, and conducts research based on international guidelines for ethical conduct and under strict adherence to guidelines in accordance with Islamic principles. Specifically, the Saudi government and people require ESC research to adhere to Sharia law. In addition to umbilical and placental stem cells, [35] Saudi Arabia permits the use of embryonic stem cells as long as they come from miscarriages, therapeutic abortions permissible by Sharia law, or are left over from in vitro fertilization and donated to research. [36] Laws and ethical guidelines for stem cell research allow the development of research institutions such as the King Abdullah International Medical Research Center, which has a cord blood bank and a stem cell registry with nearly 10,000 donors. [37] Such volume and acceptance are due to the ethical ‘permissibility’ of the donor sources, which do not conflict with religious pillars. However, some researchers err on the side of caution, choosing not to use embryos or fetal tissue as they feel it is unethical to do so. [38]

Jordan has a positive research ethics culture. [39] However, there is a significant issue of lack of trust in researchers, with 45.23 percent (38.66 percent agreeing and 6.57 percent strongly agreeing) of Jordanians holding a low level of trust in researchers, compared to 81.34 percent of Jordanians agreeing that they feel safe to participate in a research trial. [40] Safety testifies to the feeling of confidence that adequate measures are in place to protect participants from harm, whereas trust in researchers could represent the confidence in researchers to act in the participants’ best interests, adhere to ethical guidelines, provide accurate information, and respect participants’ rights and dignity. One method to improve trust would be to address communication issues relevant to ESC. Legislation surrounding stem cell research has adopted specific language, especially concerning clarification “between ‘stem cells’ and ‘embryonic stem cells’” in translation. [41] Furthermore, legislation “mandates the creation of a national committee… laying out specific regulations for stem-cell banking in accordance with international standards.” [42] This broad regulation opens the door for future global engagement and maintains transparency. However, these regulations may also constrain the influence of research direction, pace, and accessibility of research outcomes.

e.     Europe

In the European Union (EU), ethics is also principle-based, but the principles of autonomy, dignity, integrity, and vulnerability are interconnected. [43] As such, the opportunity for cohesion and concessions between individuals’ thoughts and ideals allows for a more adaptable ethics model due to the flexible principles that relate to the human experience The EU has put forth a framework in its Convention for the Protection of Human Rights and Dignity of the Human Being allowing member states to take different approaches. Each European state applies these principles to its specific conventions, leading to or reflecting different acceptance levels of stem cell research. [44]

For example, in Germany, Lebenzusammenhang , or the coherence of life, references integrity in the unity of human culture. Namely, the personal sphere “should not be subject to external intervention.” [45]  Stem cell interventions could affect this concept of bodily completeness, leading to heavy restrictions. Under the Grundgesetz, human dignity and the right to life with physical integrity are paramount. [46] The Embryo Protection Act of 1991 made producing cell lines illegal. Cell lines can be imported if approved by the Central Ethics Commission for Stem Cell Research only if they were derived before May 2007. [47] Stem cell research respects the integrity of life for the embryo with heavy specifications and intense oversight. This is vastly different in Finland, where the regulatory bodies find research more permissible in IVF excess, but only up to 14 days after fertilization. [48] Spain’s approach differs still, with a comprehensive regulatory framework. [49] Thus, research regulation can be culture-specific due to variations in applied principles. Diverse cultures call for various approaches to ethical permissibility. [50] Only an adaptive-deliberative model can address the cultural constructions of self and achieve positive, culturally sensitive stem cell research practices. [51]

II.     Religious Perspectives on ESC

Embryonic stem cell sources are the main consideration within religious contexts. While individuals may not regard their own religious texts as authoritative or factual, religion can shape their foundations or perspectives.

The Qur'an states:

“And indeed We created man from a quintessence of clay. Then We placed within him a small quantity of nutfa (sperm to fertilize) in a safe place. Then We have fashioned the nutfa into an ‘alaqa (clinging clot or cell cluster), then We developed the ‘alaqa into mudgha (a lump of flesh), and We made mudgha into bones, and clothed the bones with flesh, then We brought it into being as a new creation. So Blessed is Allah, the Best of Creators.” [52]

Many scholars of Islam estimate the time of soul installment, marked by the angel breathing in the soul to bring the individual into creation, as 120 days from conception. [53] Personhood begins at this point, and the value of life would prohibit research or experimentation that could harm the individual. If the fetus is more than 120 days old, the time ensoulment is interpreted to occur according to Islamic law, abortion is no longer permissible. [54] There are a few opposing opinions about early embryos in Islamic traditions. According to some Islamic theologians, there is no ensoulment of the early embryo, which is the source of stem cells for ESC research. [55]

In Buddhism, the stance on stem cell research is not settled. The main tenets, the prohibition against harming or destroying others (ahimsa) and the pursuit of knowledge (prajña) and compassion (karuna), leave Buddhist scholars and communities divided. [56] Some scholars argue stem cell research is in accordance with the Buddhist tenet of seeking knowledge and ending human suffering. Others feel it violates the principle of not harming others. Finding the balance between these two points relies on the karmic burden of Buddhist morality. In trying to prevent ahimsa towards the embryo, Buddhist scholars suggest that to comply with Buddhist tenets, research cannot be done as the embryo has personhood at the moment of conception and would reincarnate immediately, harming the individual's ability to build their karmic burden. [57] On the other hand, the Bodhisattvas, those considered to be on the path to enlightenment or Nirvana, have given organs and flesh to others to help alleviate grieving and to benefit all. [58] Acceptance varies on applied beliefs and interpretations.

Catholicism does not support embryonic stem cell research, as it entails creation or destruction of human embryos. This destruction conflicts with the belief in the sanctity of life. For example, in the Old Testament, Genesis describes humanity as being created in God’s image and multiplying on the Earth, referencing the sacred rights to human conception and the purpose of development and life. In the Ten Commandments, the tenet that one should not kill has numerous interpretations where killing could mean murder or shedding of the sanctity of life, demonstrating the high value of human personhood. In other books, the theological conception of when life begins is interpreted as in utero, [59] highlighting the inviolability of life and its formation in vivo to make a religious point for accepting such research as relatively limited, if at all. [60] The Vatican has released ethical directives to help apply a theological basis to modern-day conflicts. The Magisterium of the Church states that “unless there is a moral certainty of not causing harm,” experimentation on fetuses, fertilized cells, stem cells, or embryos constitutes a crime. [61] Such procedures would not respect the human person who exists at these stages, according to Catholicism. Damages to the embryo are considered gravely immoral and illicit. [62] Although the Catholic Church officially opposes abortion, surveys demonstrate that many Catholic people hold pro-choice views, whether due to the context of conception, stage of pregnancy, threat to the mother’s life, or for other reasons, demonstrating that practicing members can also accept some but not all tenets. [63]

Some major Jewish denominations, such as the Reform, Conservative, and Reconstructionist movements, are open to supporting ESC use or research as long as it is for saving a life. [64] Within Judaism, the Talmud, or study, gives personhood to the child at birth and emphasizes that life does not begin at conception: [65]

“If she is found pregnant, until the fortieth day it is mere fluid,” [66]

Whereas most religions prioritize the status of human embryos, the Halakah (Jewish religious law) states that to save one life, most other religious laws can be ignored because it is in pursuit of preservation. [67] Stem cell research is accepted due to application of these religious laws.

We recognize that all religions contain subsets and sects. The variety of environmental and cultural differences within religious groups requires further analysis to respect the flexibility of religious thoughts and practices. We make no presumptions that all cultures require notions of autonomy or morality as under the common morality theory , which asserts a set of universal moral norms that all individuals share provides moral reasoning and guides ethical decisions. [68] We only wish to show that the interaction with morality varies between cultures and countries.

III.     A Flexible Ethical Approach

The plurality of different moral approaches described above demonstrates that there can be no universally acceptable uniform law for ESC on a global scale. Instead of developing one standard, flexible ethical applications must be continued. We recommend local guidelines that incorporate important cultural and ethical priorities.

While the Declaration of Helsinki is more relevant to people in clinical trials receiving ESC products, in keeping with the tradition of protections for research subjects, consent of the donor is an ethical requirement for ESC donation in many jurisdictions including the US, Canada, and Europe. [69] The Declaration of Helsinki provides a reference point for regulatory standards and could potentially be used as a universal baseline for obtaining consent prior to gamete or embryo donation.

For instance, in Columbia University’s egg donor program for stem cell research, donors followed standard screening protocols and “underwent counseling sessions that included information as to the purpose of oocyte donation for research, what the oocytes would be used for, the risks and benefits of donation, and process of oocyte stimulation” to ensure transparency for consent. [70] The program helped advance stem cell research and provided clear and safe research methods with paid participants. Though paid participation or covering costs of incidental expenses may not be socially acceptable in every culture or context, [71] and creating embryos for ESC research is illegal in many jurisdictions, Columbia’s program was effective because of the clear and honest communications with donors, IRBs, and related stakeholders.  This example demonstrates that cultural acceptance of scientific research and of the idea that an egg or embryo does not have personhood is likely behind societal acceptance of donating eggs for ESC research. As noted, many countries do not permit the creation of embryos for research.

Proper communication and education regarding the process and purpose of stem cell research may bolster comprehension and garner more acceptance. “Given the sensitive subject material, a complete consent process can support voluntary participation through trust, understanding, and ethical norms from the cultures and morals participants value. This can be hard for researchers entering countries of different socioeconomic stability, with different languages and different societal values. [72]

An adequate moral foundation in medical ethics is derived from the cultural and religious basis that informs knowledge and actions. [73] Understanding local cultural and religious values and their impact on research could help researchers develop humility and promote inclusion.

IV.     Concerns

Some may argue that if researchers all adhere to one ethics standard, protection will be satisfied across all borders, and the global public will trust researchers. However, defining what needs to be protected and how to define such research standards is very specific to the people to which standards are applied. We suggest that applying one uniform guide cannot accurately protect each individual because we all possess our own perceptions and interpretations of social values. [74] Therefore, the issue of not adjusting to the moral pluralism between peoples in applying one standard of ethics can be resolved by building out ethics models that can be adapted to different cultures and religions.

Other concerns include medical tourism, which may promote health inequities. [75] Some countries may develop and approve products derived from ESC research before others, compromising research ethics or drug approval processes. There are also concerns about the sale of unauthorized stem cell treatments, for example, those without FDA approval in the United States. Countries with robust research infrastructures may be tempted to attract medical tourists, and some customers will have false hopes based on aggressive publicity of unproven treatments. [76]

For example, in China, stem cell clinics can market to foreign clients who are not protected under the regulatory regimes. Companies employ a marketing strategy of “ethically friendly” therapies. Specifically, in the case of Beike, China’s leading stem cell tourism company and sprouting network, ethical oversight of administrators or health bureaus at one site has “the unintended consequence of shifting questionable activities to another node in Beike's diffuse network.” [77] In contrast, Jordan is aware of stem cell research’s potential abuse and its own status as a “health-care hub.” Jordan’s expanded regulations include preserving the interests of individuals in clinical trials and banning private companies from ESC research to preserve transparency and the integrity of research practices. [78]

The social priorities of the community are also a concern. The ISSCR explicitly states that guidelines “should be periodically revised to accommodate scientific advances, new challenges, and evolving social priorities.” [79] The adaptable ethics model extends this consideration further by addressing whether research is warranted given the varying degrees of socioeconomic conditions, political stability, and healthcare accessibilities and limitations. An ethical approach would require discussion about resource allocation and appropriate distribution of funds. [80]

While some religions emphasize the sanctity of life from conception, which may lead to public opposition to ESC research, others encourage ESC research due to its potential for healing and alleviating human pain. Many countries have special regulations that balance local views on embryonic personhood, the benefits of research as individual or societal goods, and the protection of human research subjects. To foster understanding and constructive dialogue, global policy frameworks should prioritize the protection of universal human rights, transparency, and informed consent. In addition to these foundational global policies, we recommend tailoring local guidelines to reflect the diverse cultural and religious perspectives of the populations they govern. Ethics models should be adapted to local populations to effectively establish research protections, growth, and possibilities of stem cell research.

For example, in countries with strong beliefs in the moral sanctity of embryos or heavy religious restrictions, an adaptive model can allow for discussion instead of immediate rejection. In countries with limited individual rights and voice in science policy, an adaptive model ensures cultural, moral, and religious views are taken into consideration, thereby building social inclusion. While this ethical consideration by the government may not give a complete voice to every individual, it will help balance policies and maintain the diverse perspectives of those it affects. Embracing an adaptive ethics model of ESC research promotes open-minded dialogue and respect for the importance of human belief and tradition. By actively engaging with cultural and religious values, researchers can better handle disagreements and promote ethical research practices that benefit each society.

This brief exploration of the religious and cultural differences that impact ESC research reveals the nuances of relative ethics and highlights a need for local policymakers to apply a more intense adaptive model.

[1] Poliwoda, S., Noor, N., Downs, E., Schaaf, A., Cantwell, A., Ganti, L., Kaye, A. D., Mosel, L. I., Carroll, C. B., Viswanath, O., & Urits, I. (2022). Stem cells: a comprehensive review of origins and emerging clinical roles in medical practice.  Orthopedic reviews ,  14 (3), 37498. https://doi.org/10.52965/001c.37498

[2] Poliwoda, S., Noor, N., Downs, E., Schaaf, A., Cantwell, A., Ganti, L., Kaye, A. D., Mosel, L. I., Carroll, C. B., Viswanath, O., & Urits, I. (2022). Stem cells: a comprehensive review of origins and emerging clinical roles in medical practice.  Orthopedic reviews ,  14 (3), 37498. https://doi.org/10.52965/001c.37498

[3] International Society for Stem Cell Research. (2023). Laboratory-based human embryonic stem cell research, embryo research, and related research activities . International Society for Stem Cell Research. https://www.isscr.org/guidelines/blog-post-title-one-ed2td-6fcdk ; Kimmelman, J., Hyun, I., Benvenisty, N.  et al.  Policy: Global standards for stem-cell research.  Nature   533 , 311–313 (2016). https://doi.org/10.1038/533311a

[4] International Society for Stem Cell Research. (2023). Laboratory-based human embryonic stem cell research, embryo research, and related research activities . International Society for Stem Cell Research. https://www.isscr.org/guidelines/blog-post-title-one-ed2td-6fcdk

[5] Concerning the moral philosophies of stem cell research, our paper does not posit a personal moral stance nor delve into the “when” of human life begins. To read further about the philosophical debate, consider the following sources:

Sandel M. J. (2004). Embryo ethics--the moral logic of stem-cell research.  The New England journal of medicine ,  351 (3), 207–209. https://doi.org/10.1056/NEJMp048145 ; George, R. P., & Lee, P. (2020, September 26). Acorns and Embryos . The New Atlantis. https://www.thenewatlantis.com/publications/acorns-and-embryos ; Sagan, A., & Singer, P. (2007). The moral status of stem cells. Metaphilosophy , 38 (2/3), 264–284. http://www.jstor.org/stable/24439776 ; McHugh P. R. (2004). Zygote and "clonote"--the ethical use of embryonic stem cells.  The New England journal of medicine ,  351 (3), 209–211. https://doi.org/10.1056/NEJMp048147 ; Kurjak, A., & Tripalo, A. (2004). The facts and doubts about beginning of the human life and personality.  Bosnian journal of basic medical sciences ,  4 (1), 5–14. https://doi.org/10.17305/bjbms.2004.3453

[6] Vazin, T., & Freed, W. J. (2010). Human embryonic stem cells: derivation, culture, and differentiation: a review.  Restorative neurology and neuroscience ,  28 (4), 589–603. https://doi.org/10.3233/RNN-2010-0543

[7] Socially, at its core, the Western approach to ethics is widely principle-based, autonomy being one of the key factors to ensure a fundamental respect for persons within research. For information regarding autonomy in research, see: Department of Health, Education, and Welfare, & National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (1978). The Belmont Report. Ethical principles and guidelines for the protection of human subjects of research.; For a more in-depth review of autonomy within the US, see: Beauchamp, T. L., & Childress, J. F. (1994). Principles of Biomedical Ethics . Oxford University Press.

[8] Sherley v. Sebelius , 644 F.3d 388 (D.C. Cir. 2011), citing 45 C.F.R. 46.204(b) and [42 U.S.C. § 289g(b)]. https://www.cadc.uscourts.gov/internet/opinions.nsf/6c690438a9b43dd685257a64004ebf99/$file/11-5241-1391178.pdf

[9] Stem Cell Research Enhancement Act of 2005, H. R. 810, 109 th Cong. (2001). https://www.govtrack.us/congress/bills/109/hr810/text ; Bush, G. W. (2006, July 19). Message to the House of Representatives . National Archives and Records Administration. https://georgewbush-whitehouse.archives.gov/news/releases/2006/07/20060719-5.html

[10] National Archives and Records Administration. (2009, March 9). Executive order 13505 -- removing barriers to responsible scientific research involving human stem cells . National Archives and Records Administration. https://obamawhitehouse.archives.gov/the-press-office/removing-barriers-responsible-scientific-research-involving-human-stem-cells

[11] Hurlbut, W. B. (2006). Science, Religion, and the Politics of Stem Cells.  Social Research ,  73 (3), 819–834. http://www.jstor.org/stable/40971854

[12] Akpa-Inyang, Francis & Chima, Sylvester. (2021). South African traditional values and beliefs regarding informed consent and limitations of the principle of respect for autonomy in African communities: a cross-cultural qualitative study. BMC Medical Ethics . 22. 10.1186/s12910-021-00678-4.

[13] Source for further reading: Tangwa G. B. (2007). Moral status of embryonic stem cells: perspective of an African villager. Bioethics , 21(8), 449–457. https://doi.org/10.1111/j.1467-8519.2007.00582.x , see also Mnisi, F. M. (2020). An African analysis based on ethics of Ubuntu - are human embryonic stem cell patents morally justifiable? African Insight , 49 (4).

[14] Jecker, N. S., & Atuire, C. (2021). Bioethics in Africa: A contextually enlightened analysis of three cases. Developing World Bioethics , 22 (2), 112–122. https://doi.org/10.1111/dewb.12324

[15] Jecker, N. S., & Atuire, C. (2021). Bioethics in Africa: A contextually enlightened analysis of three cases. Developing World Bioethics, 22(2), 112–122. https://doi.org/10.1111/dewb.12324

[16] Jackson, C.S., Pepper, M.S. Opportunities and barriers to establishing a cell therapy programme in South Africa.  Stem Cell Res Ther   4 , 54 (2013). https://doi.org/10.1186/scrt204 ; Pew Research Center. (2014, May 1). Public health a major priority in African nations . Pew Research Center’s Global Attitudes Project. https://www.pewresearch.org/global/2014/05/01/public-health-a-major-priority-in-african-nations/

[17] Department of Health Republic of South Africa. (2021). Health Research Priorities (revised) for South Africa 2021-2024 . National Health Research Strategy. https://www.health.gov.za/wp-content/uploads/2022/05/National-Health-Research-Priorities-2021-2024.pdf

[18] Oosthuizen, H. (2013). Legal and Ethical Issues in Stem Cell Research in South Africa. In: Beran, R. (eds) Legal and Forensic Medicine. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-32338-6_80 , see also: Gaobotse G (2018) Stem Cell Research in Africa: Legislation and Challenges. J Regen Med 7:1. doi: 10.4172/2325-9620.1000142

[19] United States Bureau of Citizenship and Immigration Services. (1998). Tunisia: Information on the status of Christian conversions in Tunisia . UNHCR Web Archive. https://webarchive.archive.unhcr.org/20230522142618/https://www.refworld.org/docid/3df0be9a2.html

[20] Gaobotse, G. (2018) Stem Cell Research in Africa: Legislation and Challenges. J Regen Med 7:1. doi: 10.4172/2325-9620.1000142

[21] Kooli, C. Review of assisted reproduction techniques, laws, and regulations in Muslim countries.  Middle East Fertil Soc J   24 , 8 (2020). https://doi.org/10.1186/s43043-019-0011-0 ; Gaobotse, G. (2018) Stem Cell Research in Africa: Legislation and Challenges. J Regen Med 7:1. doi: 10.4172/2325-9620.1000142

[22] Pang M. C. (1999). Protective truthfulness: the Chinese way of safeguarding patients in informed treatment decisions. Journal of medical ethics , 25(3), 247–253. https://doi.org/10.1136/jme.25.3.247

[23] Wang, L., Wang, F., & Zhang, W. (2021). Bioethics in China’s biosecurity law: Forms, effects, and unsettled issues. Journal of law and the biosciences , 8(1).  https://doi.org/10.1093/jlb/lsab019 https://academic.oup.com/jlb/article/8/1/lsab019/6299199

[24] Wang, Y., Xue, Y., & Guo, H. D. (2022). Intervention effects of traditional Chinese medicine on stem cell therapy of myocardial infarction.  Frontiers in pharmacology ,  13 , 1013740. https://doi.org/10.3389/fphar.2022.1013740

[25] Li, X.-T., & Zhao, J. (2012). Chapter 4: An Approach to the Nature of Qi in TCM- Qi and Bioenergy. In Recent Advances in Theories and Practice of Chinese Medicine (p. 79). InTech.

[26] Luo, D., Xu, Z., Wang, Z., & Ran, W. (2021). China's Stem Cell Research and Knowledge Levels of Medical Practitioners and Students.  Stem cells international ,  2021 , 6667743. https://doi.org/10.1155/2021/6667743

[27] Luo, D., Xu, Z., Wang, Z., & Ran, W. (2021). China's Stem Cell Research and Knowledge Levels of Medical Practitioners and Students.  Stem cells international ,  2021 , 6667743. https://doi.org/10.1155/2021/6667743

[28] Zhang, J. Y. (2017). Lost in translation? accountability and governance of Clinical Stem Cell Research in China. Regenerative Medicine , 12 (6), 647–656. https://doi.org/10.2217/rme-2017-0035

[29] Wang, L., Wang, F., & Zhang, W. (2021). Bioethics in China’s biosecurity law: Forms, effects, and unsettled issues. Journal of law and the biosciences , 8(1).  https://doi.org/10.1093/jlb/lsab019 https://academic.oup.com/jlb/article/8/1/lsab019/6299199

[30] Chen, H., Wei, T., Wang, H.  et al.  Association of China’s two-child policy with changes in number of births and birth defects rate, 2008–2017.  BMC Public Health   22 , 434 (2022). https://doi.org/10.1186/s12889-022-12839-0

[31] Azuma, K. Regulatory Landscape of Regenerative Medicine in Japan.  Curr Stem Cell Rep   1 , 118–128 (2015). https://doi.org/10.1007/s40778-015-0012-6

[32] Harris, R. (2005, May 19). Researchers Report Advance in Stem Cell Production . NPR. https://www.npr.org/2005/05/19/4658967/researchers-report-advance-in-stem-cell-production

[33] Park, S. (2012). South Korea steps up stem-cell work.  Nature . https://doi.org/10.1038/nature.2012.10565

[34] Resnik, D. B., Shamoo, A. E., & Krimsky, S. (2006). Fraudulent human embryonic stem cell research in South Korea: lessons learned.  Accountability in research ,  13 (1), 101–109. https://doi.org/10.1080/08989620600634193 .

[35] Alahmad, G., Aljohani, S., & Najjar, M. F. (2020). Ethical challenges regarding the use of stem cells: interviews with researchers from Saudi Arabia. BMC medical ethics, 21(1), 35. https://doi.org/10.1186/s12910-020-00482-6

[36] Association for the Advancement of Blood and Biotherapies.  https://www.aabb.org/regulatory-and-advocacy/regulatory-affairs/regulatory-for-cellular-therapies/international-competent-authorities/saudi-arabia

[37] Alahmad, G., Aljohani, S., & Najjar, M. F. (2020). Ethical challenges regarding the use of stem cells: Interviews with researchers from Saudi Arabia.  BMC medical ethics ,  21 (1), 35. https://doi.org/10.1186/s12910-020-00482-6

[38] Alahmad, G., Aljohani, S., & Najjar, M. F. (2020). Ethical challenges regarding the use of stem cells: Interviews with researchers from Saudi Arabia. BMC medical ethics , 21(1), 35. https://doi.org/10.1186/s12910-020-00482-6

Culturally, autonomy practices follow a relational autonomy approach based on a paternalistic deontological health care model. The adherence to strict international research policies and religious pillars within the regulatory environment is a great foundation for research ethics. However, there is a need to develop locally targeted ethics approaches for research (as called for in Alahmad, G., Aljohani, S., & Najjar, M. F. (2020). Ethical challenges regarding the use of stem cells: interviews with researchers from Saudi Arabia. BMC medical ethics, 21(1), 35. https://doi.org/10.1186/s12910-020-00482-6), this decision-making approach may help advise a research decision model. For more on the clinical cultural autonomy approaches, see: Alabdullah, Y. Y., Alzaid, E., Alsaad, S., Alamri, T., Alolayan, S. W., Bah, S., & Aljoudi, A. S. (2022). Autonomy and paternalism in Shared decision‐making in a Saudi Arabian tertiary hospital: A cross‐sectional study. Developing World Bioethics , 23 (3), 260–268. https://doi.org/10.1111/dewb.12355 ; Bukhari, A. A. (2017). Universal Principles of Bioethics and Patient Rights in Saudi Arabia (Doctoral dissertation, Duquesne University). https://dsc.duq.edu/etd/124; Ladha, S., Nakshawani, S. A., Alzaidy, A., & Tarab, B. (2023, October 26). Islam and Bioethics: What We All Need to Know . Columbia University School of Professional Studies. https://sps.columbia.edu/events/islam-and-bioethics-what-we-all-need-know

[39] Ababneh, M. A., Al-Azzam, S. I., Alzoubi, K., Rababa’h, A., & Al Demour, S. (2021). Understanding and attitudes of the Jordanian public about clinical research ethics.  Research Ethics ,  17 (2), 228-241.  https://doi.org/10.1177/1747016120966779

[40] Ababneh, M. A., Al-Azzam, S. I., Alzoubi, K., Rababa’h, A., & Al Demour, S. (2021). Understanding and attitudes of the Jordanian public about clinical research ethics.  Research Ethics ,  17 (2), 228-241.  https://doi.org/10.1177/1747016120966779

[41] Dajani, R. (2014). Jordan’s stem-cell law can guide the Middle East.  Nature  510, 189. https://doi.org/10.1038/510189a

[42] Dajani, R. (2014). Jordan’s stem-cell law can guide the Middle East.  Nature  510, 189. https://doi.org/10.1038/510189a

[43] The EU’s definition of autonomy relates to the capacity for creating ideas, moral insight, decisions, and actions without constraint, personal responsibility, and informed consent. However, the EU views autonomy as not completely able to protect individuals and depends on other principles, such as dignity, which “expresses the intrinsic worth and fundamental equality of all human beings.” Rendtorff, J.D., Kemp, P. (2019). Four Ethical Principles in European Bioethics and Biolaw: Autonomy, Dignity, Integrity and Vulnerability. In: Valdés, E., Lecaros, J. (eds) Biolaw and Policy in the Twenty-First Century. International Library of Ethics, Law, and the New Medicine, vol 78. Springer, Cham. https://doi.org/10.1007/978-3-030-05903-3_3

[44] Council of Europe. Convention for the protection of Human Rights and Dignity of the Human Being with regard to the Application of Biology and Medicine: Convention on Human Rights and Biomedicine (ETS No. 164) https://www.coe.int/en/web/conventions/full-list?module=treaty-detail&treatynum=164 (forbidding the creation of embryos for research purposes only, and suggests embryos in vitro have protections.); Also see Drabiak-Syed B. K. (2013). New President, New Human Embryonic Stem Cell Research Policy: Comparative International Perspectives and Embryonic Stem Cell Research Laws in France.  Biotechnology Law Report ,  32 (6), 349–356. https://doi.org/10.1089/blr.2013.9865

[45] Rendtorff, J.D., Kemp, P. (2019). Four Ethical Principles in European Bioethics and Biolaw: Autonomy, Dignity, Integrity and Vulnerability. In: Valdés, E., Lecaros, J. (eds) Biolaw and Policy in the Twenty-First Century. International Library of Ethics, Law, and the New Medicine, vol 78. Springer, Cham. https://doi.org/10.1007/978-3-030-05903-3_3

[46] Tomuschat, C., Currie, D. P., Kommers, D. P., & Kerr, R. (Trans.). (1949, May 23). Basic law for the Federal Republic of Germany. https://www.btg-bestellservice.de/pdf/80201000.pdf

[47] Regulation of Stem Cell Research in Germany . Eurostemcell. (2017, April 26). https://www.eurostemcell.org/regulation-stem-cell-research-germany

[48] Regulation of Stem Cell Research in Finland . Eurostemcell. (2017, April 26). https://www.eurostemcell.org/regulation-stem-cell-research-finland

[49] Regulation of Stem Cell Research in Spain . Eurostemcell. (2017, April 26). https://www.eurostemcell.org/regulation-stem-cell-research-spain

[50] Some sources to consider regarding ethics models or regulatory oversights of other cultures not covered:

Kara MA. Applicability of the principle of respect for autonomy: the perspective of Turkey. J Med Ethics. 2007 Nov;33(11):627-30. doi: 10.1136/jme.2006.017400. PMID: 17971462; PMCID: PMC2598110.

Ugarte, O. N., & Acioly, M. A. (2014). The principle of autonomy in Brazil: one needs to discuss it ...  Revista do Colegio Brasileiro de Cirurgioes ,  41 (5), 374–377. https://doi.org/10.1590/0100-69912014005013

Bharadwaj, A., & Glasner, P. E. (2012). Local cells, global science: The rise of embryonic stem cell research in India . Routledge.

For further research on specific European countries regarding ethical and regulatory framework, we recommend this database: Regulation of Stem Cell Research in Europe . Eurostemcell. (2017, April 26). https://www.eurostemcell.org/regulation-stem-cell-research-europe   

[51] Klitzman, R. (2006). Complications of culture in obtaining informed consent. The American Journal of Bioethics, 6(1), 20–21. https://doi.org/10.1080/15265160500394671 see also: Ekmekci, P. E., & Arda, B. (2017). Interculturalism and Informed Consent: Respecting Cultural Differences without Breaching Human Rights.  Cultura (Iasi, Romania) ,  14 (2), 159–172.; For why trust is important in research, see also: Gray, B., Hilder, J., Macdonald, L., Tester, R., Dowell, A., & Stubbe, M. (2017). Are research ethics guidelines culturally competent?  Research Ethics ,  13 (1), 23-41.  https://doi.org/10.1177/1747016116650235

[52] The Qur'an  (M. Khattab, Trans.). (1965). Al-Mu’minun, 23: 12-14. https://quran.com/23

[53] Lenfest, Y. (2017, December 8). Islam and the beginning of human life . Bill of Health. https://blog.petrieflom.law.harvard.edu/2017/12/08/islam-and-the-beginning-of-human-life/

[54] Aksoy, S. (2005). Making regulations and drawing up legislation in Islamic countries under conditions of uncertainty, with special reference to embryonic stem cell research. Journal of Medical Ethics , 31: 399-403.; see also: Mahmoud, Azza. "Islamic Bioethics: National Regulations and Guidelines of Human Stem Cell Research in the Muslim World." Master's thesis, Chapman University, 2022. https://doi.org/10.36837/ chapman.000386

[55] Rashid, R. (2022). When does Ensoulment occur in the Human Foetus. Journal of the British Islamic Medical Association , 12 (4). ISSN 2634 8071. https://www.jbima.com/wp-content/uploads/2023/01/2-Ethics-3_-Ensoulment_Rafaqat.pdf.

[56] Sivaraman, M. & Noor, S. (2017). Ethics of embryonic stem cell research according to Buddhist, Hindu, Catholic, and Islamic religions: perspective from Malaysia. Asian Biomedicine,8(1) 43-52.  https://doi.org/10.5372/1905-7415.0801.260

[57] Jafari, M., Elahi, F., Ozyurt, S. & Wrigley, T. (2007). 4. Religious Perspectives on Embryonic Stem Cell Research. In K. Monroe, R. Miller & J. Tobis (Ed.),  Fundamentals of the Stem Cell Debate: The Scientific, Religious, Ethical, and Political Issues  (pp. 79-94). Berkeley: University of California Press.  https://escholarship.org/content/qt9rj0k7s3/qt9rj0k7s3_noSplash_f9aca2e02c3777c7fb76ea768ba458f0.pdf https://doi.org/10.1525/9780520940994-005

[58] Lecso, P. A. (1991). The Bodhisattva Ideal and Organ Transplantation.  Journal of Religion and Health ,  30 (1), 35–41. http://www.jstor.org/stable/27510629 ; Bodhisattva, S. (n.d.). The Key of Becoming a Bodhisattva . A Guide to the Bodhisattva Way of Life. http://www.buddhism.org/Sutras/2/BodhisattvaWay.htm

[59] There is no explicit religious reference to when life begins or how to conduct research that interacts with the concept of life. However, these are relevant verses pertaining to how the fetus is viewed. (( King James Bible . (1999). Oxford University Press. (original work published 1769))

Jerimiah 1: 5 “Before I formed thee in the belly I knew thee; and before thou camest forth out of the womb I sanctified thee…”

In prophet Jerimiah’s insight, God set him apart as a person known before childbirth, a theme carried within the Psalm of David.

Psalm 139: 13-14 “…Thou hast covered me in my mother's womb. I will praise thee; for I am fearfully and wonderfully made…”

These verses demonstrate David’s respect for God as an entity that would know of all man’s thoughts and doings even before birth.

[60] It should be noted that abortion is not supported as well.

[61] The Vatican. (1987, February 22). Instruction on Respect for Human Life in Its Origin and on the Dignity of Procreation Replies to Certain Questions of the Day . Congregation For the Doctrine of the Faith. https://www.vatican.va/roman_curia/congregations/cfaith/documents/rc_con_cfaith_doc_19870222_respect-for-human-life_en.html

[62] The Vatican. (2000, August 25). Declaration On the Production and the Scientific and Therapeutic Use of Human Embryonic Stem Cells . Pontifical Academy for Life. https://www.vatican.va/roman_curia/pontifical_academies/acdlife/documents/rc_pa_acdlife_doc_20000824_cellule-staminali_en.html ; Ohara, N. (2003). Ethical Consideration of Experimentation Using Living Human Embryos: The Catholic Church’s Position on Human Embryonic Stem Cell Research and Human Cloning. Department of Obstetrics and Gynecology . Retrieved from https://article.imrpress.com/journal/CEOG/30/2-3/pii/2003018/77-81.pdf.

[63] Smith, G. A. (2022, May 23). Like Americans overall, Catholics vary in their abortion views, with regular mass attenders most opposed . Pew Research Center. https://www.pewresearch.org/short-reads/2022/05/23/like-americans-overall-catholics-vary-in-their-abortion-views-with-regular-mass-attenders-most-opposed/

[64] Rosner, F., & Reichman, E. (2002). Embryonic stem cell research in Jewish law. Journal of halacha and contemporary society , (43), 49–68.; Jafari, M., Elahi, F., Ozyurt, S. & Wrigley, T. (2007). 4. Religious Perspectives on Embryonic Stem Cell Research. In K. Monroe, R. Miller & J. Tobis (Ed.),  Fundamentals of the Stem Cell Debate: The Scientific, Religious, Ethical, and Political Issues  (pp. 79-94). Berkeley: University of California Press.  https://escholarship.org/content/qt9rj0k7s3/qt9rj0k7s3_noSplash_f9aca2e02c3777c7fb76ea768ba458f0.pdf https://doi.org/10.1525/9780520940994-005

[65] Schenker J. G. (2008). The beginning of human life: status of embryo. Perspectives in Halakha (Jewish Religious Law).  Journal of assisted reproduction and genetics ,  25 (6), 271–276. https://doi.org/10.1007/s10815-008-9221-6

[66] Ruttenberg, D. (2020, May 5). The Torah of Abortion Justice (annotated source sheet) . Sefaria. https://www.sefaria.org/sheets/234926.7?lang=bi&with=all&lang2=en

[67] Jafari, M., Elahi, F., Ozyurt, S. & Wrigley, T. (2007). 4. Religious Perspectives on Embryonic Stem Cell Research. In K. Monroe, R. Miller & J. Tobis (Ed.),  Fundamentals of the Stem Cell Debate: The Scientific, Religious, Ethical, and Political Issues  (pp. 79-94). Berkeley: University of California Press.  https://escholarship.org/content/qt9rj0k7s3/qt9rj0k7s3_noSplash_f9aca2e02c3777c7fb76ea768ba458f0.pdf https://doi.org/10.1525/9780520940994-005

[68] Gert, B. (2007). Common morality: Deciding what to do . Oxford Univ. Press.

[69] World Medical Association (2013). World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA , 310(20), 2191–2194. https://doi.org/10.1001/jama.2013.281053 Declaration of Helsinki – WMA – The World Medical Association .; see also: National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979).  The Belmont report: Ethical principles and guidelines for the protection of human subjects of research . U.S. Department of Health and Human Services.  https://www.hhs.gov/ohrp/regulations-and-policy/belmont-report/read-the-belmont-report/index.html

[70] Zakarin Safier, L., Gumer, A., Kline, M., Egli, D., & Sauer, M. V. (2018). Compensating human subjects providing oocytes for stem cell research: 9-year experience and outcomes.  Journal of assisted reproduction and genetics ,  35 (7), 1219–1225. https://doi.org/10.1007/s10815-018-1171-z https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6063839/ see also: Riordan, N. H., & Paz Rodríguez, J. (2021). Addressing concerns regarding associated costs, transparency, and integrity of research in recent stem cell trial. Stem Cells Translational Medicine , 10 (12), 1715–1716. https://doi.org/10.1002/sctm.21-0234

[71] Klitzman, R., & Sauer, M. V. (2009). Payment of egg donors in stem cell research in the USA.  Reproductive biomedicine online ,  18 (5), 603–608. https://doi.org/10.1016/s1472-6483(10)60002-8

[72] Krosin, M. T., Klitzman, R., Levin, B., Cheng, J., & Ranney, M. L. (2006). Problems in comprehension of informed consent in rural and peri-urban Mali, West Africa.  Clinical trials (London, England) ,  3 (3), 306–313. https://doi.org/10.1191/1740774506cn150oa

[73] Veatch, Robert M.  Hippocratic, Religious, and Secular Medical Ethics: The Points of Conflict . Georgetown University Press, 2012.

[74] Msoroka, M. S., & Amundsen, D. (2018). One size fits not quite all: Universal research ethics with diversity.  Research Ethics ,  14 (3), 1-17.  https://doi.org/10.1177/1747016117739939

[75] Pirzada, N. (2022). The Expansion of Turkey’s Medical Tourism Industry.  Voices in Bioethics ,  8 . https://doi.org/10.52214/vib.v8i.9894

[76] Stem Cell Tourism: False Hope for Real Money . Harvard Stem Cell Institute (HSCI). (2023). https://hsci.harvard.edu/stem-cell-tourism , See also: Bissassar, M. (2017). Transnational Stem Cell Tourism: An ethical analysis.  Voices in Bioethics ,  3 . https://doi.org/10.7916/vib.v3i.6027

[77] Song, P. (2011) The proliferation of stem cell therapies in post-Mao China: problematizing ethical regulation,  New Genetics and Society , 30:2, 141-153, DOI:  10.1080/14636778.2011.574375

[78] Dajani, R. (2014). Jordan’s stem-cell law can guide the Middle East.  Nature  510, 189. https://doi.org/10.1038/510189a

[79] International Society for Stem Cell Research. (2024). Standards in stem cell research . International Society for Stem Cell Research. https://www.isscr.org/guidelines/5-standards-in-stem-cell-research

[80] Benjamin, R. (2013). People’s science bodies and rights on the Stem Cell Frontier . Stanford University Press.

Mifrah Hayath

SM Candidate Harvard Medical School, MS Biotechnology Johns Hopkins University

Olivia Bowers

MS Bioethics Columbia University (Disclosure: affiliated with Voices in Bioethics)

Article Details

This work is licensed under a Creative Commons Attribution 4.0 International License .