research study on cervical cancer

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
• Chin J Cancer Res
• v.32(6); 2020 Dec 31

Cervical cancer: Epidemiology, risk factors and screening

Shaokai zhang.

1 Department of Cancer Epidemiology, Affiliated Cancer Hospital of Zhengzhou University/Henan Cancer Hospital, Henan Engineering Research Center of Cancer Prevention and Control, Henan International Joint Laboratory of Cancer Prevention, Zhengzhou 450008, China

Luyao Zhang

Youlin qiao.

2 Department of Epidemiology, National Cancer Center, Chinese Academy of Medical Sciences, School of Population Medicine and Public Health, Peking Union Medical College, Beijing 100021, China

Cervical cancer is one of the leading causes of cancer death among females worldwide and its behavior epidemiologically likes a venereal disease of low infectiousness. Early age at first intercourse and multiple sexual partners have been shown to exert strong effects on risk. The wide differences in the incidence among different countries also influenced by the introduction of screening. Although the general picture remains one of decreasing incidence and mortality, there are signs of an increasing cervical cancer risk probably due to changes in sexual behavior. Smoking and human papillomavirus (HPV) 16/18 are currently important issues in a concept of multifactorial, stepwise carcinogenesis at the cervix uteri. Therefore, society-based preventive and control measures, screening activities and HPV vaccination are recommended. Cervical cancer screening methods have evolved from cell morphology observation to molecular testing. High-risk HPV genotyping and liquid-based cytology are common methods which have been widely recommended and used worldwide. In future, accurate, cheap, fast and easy-to-use methods would be more popular. Artificial intelligence also shows to be promising in cervical cancer screening by integrating image recognition with big data technology. Meanwhile, China has achieved numerous breakthroughs in cervical cancer prevention and control which could be a great demonstration for other developing and resource-limited areas. In conclusion, although cervical cancer threatens female health, it could be the first cancer that would be eliminated by human beings with comprehensive preventive and control strategy.

Introduction

Cervical cancer is the second common female malignant tumor globally which seriously threatens female’s health. Persistent infection of high-risk human papillomavirus (HPV) has been clarified to be the necessary cause of cervical cancer ( 1 , 2 ). The clear etiology accelerated the establishment and implementation of comprehensive prevention and control system of cervical cancer. In May 2018, the World Health Organization (WHO) issued a call for the elimination of cervical cancer globally, and more than 70 countries and international academic societies acted positively immediately ( 3 - 6 ). Thereafter, in November 17, 2020, WHO released the global strategy to accelerate the elimination of cervical cancer as a public health problem to light the road of cervical cancer prevention and control in future which mean that 194 countries promise together to eliminate cervical cancer for the first time ( 7 ). At this milestone time point, we reviewed the update progress of cervical cancer prevention and control in epidemiology, risk factors and screening, in order to pave the way of cervical cancer elimination.

Epidemiology for cervical cancer

Cervical cancer is one of the leading causes of cancer death among women ( 8 ). Over the past 30 years, the increasing proportion of young women affected by cervical cancer has ranged from 10% to 40% ( 9 ). According to the WHO and International Agency for Research on Cancer (IARC) estimates, the year 2008 saw 529,000 new cases of cervical cancer globally. In developing countries, the number of new cases of cervical cancer was 452,000 and ranked second among malignancies in female patients ( 10 ). Conversely, the number of new cases of cervical cancer was 77,000 in developed countries and ranked tenth among female malignancies.

In 2018 worldwide with an estimated 570,000 cases and 311,000 deaths, cervical cancer ranks as the fourth most frequently diagnosed cancer and the fourth leading cause of cancer death in women ( 11 ). However, approximately 85% of the worldwide deaths from cervical cancer occur in underdeveloped or developing countries, and the death rate is 18 times higher in low-income and middle-income countries compared with wealthier countries ( 12 ). Cervical cancer ranks second in incidence and mortality behind breast cancer in lower Human Development Index (HDI) settings; however, it is the most commonly diagnosed cancer in 28 countries and the leading cause of cancer death in 42 countries, the vast majority of which are in Sub-Saharan Africa and South Eastern Asia ( 13 ). The highest regional incidence and mortality rates are seen in Africa ( 14 ). In relative terms, the rates are 7−10 times lower in North America, Australia/New Zealand, and Western Asia (Saudi Arabia and Iraq) ( 15 ).

In China, cervical cancer is the second largest female malignant tumor ( 11 ). According to the data from National Cancer Center in 2015, there were 98,900 new cases and 30,500 deaths of cervical cancer ( 16 ). In the past 20 years, the incidence and mortality of cervical cancer have been increasing gradually in China ( 17 ).

Between 2004 and 2007, the Chinese scientific research team, cooperated with WHO/IARC and the Cleveland Medical Center in the United States in 8 rural and urban areas (Xiangyuan county of Shanxi Province, Yangcheng county of Shanxi Province, Xinmi county of Henan Province, Hotan Prefecture of Xinjiang Uygur Autonomous Region, Shanghai City, Beijing City, Shenzhen City of Guangdong Province, and Shenyang City of Liaoning Province), carried out a population-based multicenter HPV type distribution study among females aged 15−59 years old, clarifying the dominant HPV types of rural and urban populations in China, as well as female HPV infection status and age distribution ( 18 ). Studies have confirmed that persistent infection of high-risk HPV is closely related to the occurrence of cervical cancer. There are 14 types of high-risk HPV, namely HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68 and 73. A multi-center cross-sectional survey study showed that the infection rate of high-risk HPV in China is about 14.3%, and the dominant types are HPV16 (2.9%), HPV52 (1.7%), HPV58 (1.5%), HPV33 (1%) and HPV18 (0.8%), and showed double peaks during adolescence and perimenopause ( 19 ). Globally, HPV16 has the highest infection rate, HPV18 is the second most common type, while HPV 33 is common in Asia, and HPV52 and HPV58 have relatively low infection rates. This shows that compared with the global HPV epidemiology, HPV epidemiology in China has both similarities and differences.

Subsequently, the Chinese scientific research team conducted a cross-sectional multi-center cervical cancer and precancerous HPV genotyping study based on 19 hospitals in 7 geographic regions (Northeast China, North China, Northwest China, Central China, East China, Southwest, and South China). Through the pathological laboratory procedures of strict quality control, it was found that the dominant HPV types in cervical cancer tissue were HPV16, 18, 31, 52 and 58, respectively, and that HPV16 and 18 were the most carcinogenic, which could cause more than 84.5% of cervical cancer ( 20 ). The above research on HPV dominant types from different perspectives provides solid scientific evidence and support for the future research and application of preventive HPV vaccine and in vitro diagnostic technology, epidemiological research and health economics research in the Chinese population.

Risk factors for cervical cancer

A number of risk factors for cervical cancer are linked to exposure to the HPV ( 21 , 22 ). Invasive cancer development process could prolong up to 20 years from the precursor lesion caused by sexually transmitted HPV ( 23 ). However, there are also other numerous risk factors (such as reproductive and sexual factors, behavioral factors, etc) for cervical cancers which include sexual intercourse at a young age (<16 years old), multiple sexual partners, smoking, high parity and low socio-economic level ( 24 , 25 ).

Sexually transmitted infections (STI)

The primary cause of pre-cancerous and cancerous cervical lesions is infection with a high-risk or oncogenic HPV types. Most cases of cervical cancer occur as a result of infection with HPV16 and 18. High-risk types, especially HPV16, are found to be highly prevalent in human populations ( 22 ). The infection is usually transmitted by sexual contact, causing squamous intraepithelial lesions. Most lesions disappear after 6−12 months due to immunological intervention. However, a small percentage of these lesions remain and can cause cancer.

The results of a meta-analysis showed that the highest prevalence of HPV occurs at the age of 25 years, which could be related to changes in sexual behavior ( 26 ). In a meta-analysis study, the bimodal distribution of cervical cancer in some regions has been studied. In this distribution, immediately after sexual intercourse, an outbreak of HPV can be observed, which is followed by a plateau at adult age; the second peak again is observed after 45 years old ( 27 ). Permanent infection with one of the high-risk types of HPV over time leads to the development of cervical intraepithelial neoplasia (CIN). The major mechanisms through which HPV contributes to carcinogenesis involve the activity of two viral oncoproteins, E6 and E7, which interfere with major tumor suppressor genes, P53 and retinoblastoma. In addition, E6 and E7 are associated with changes in host DNA and virus DNA methylation. Interactions of E6 and E7 with cellular proteins and DNA methylation modifications are associated with changes in key cellular pathways that regulate genetic integrity, cell adhesion, immune response, apoptosis, and cellular control ( 28 ).

Human immunodeficiency virus (HIV)

The risk of developing infection from high-risk HPV types is higher in women with HIV ( 29 ). The results of the studies on the relationship between HIV and cervical cancer suggested a higher rate of persistent HPV infection with multiple oncogene viruses, more abnormal Papanicolau (Pap) smears, and higher incidence of CIN and invasive cervix carcinoma among people with HIV ( 23 ). Women infected with HIV are at increased risk of HPV infection at an early age (13−18 years) and are at high risk of cervical cancer. Compared with non-infected women, HIV positive patients with cervical cancer are diagnosed at an earlier age (15−49 years old) ( 30 ).

Reproductive and sexual factors

Sexual partners.

Factors relating to sexual behavior have also been linked to cervical cancer. One study found that an increased risk of cervical cancer is observed in people with multiple sexual partners ( 31 ). Moreover, many studies have also suggested that women with multiple sexual partners are at high risk for HPV acquisition and cervical cancer ( 32 , 33 ). From the meta-analysis, a significant increased risk of cervical diseases was observed in individuals with multiple sexual partners compared to individuals with few partners, both in non-malignant cervical disease and in cervical cancer ( 34 ). The association remained exist even after controlling for the status of HPV infection, which is a major cause of cervical cancer. Also, early age at first intercourse is a risk factor for cervical cancer ( 35 ).

Oral contraceptive (OC) pills

OC pills are known to be a risk factor for cervical cancer. In an international collaborative epidemiological study of cervical cancer, the relative risk in current users increased with an increase in the duration of OC use. It has been reported that the use of OC for 5 years or more can double the risk of cancer ( 36 ). And in a multi-center case-control study, among women who tested positive for HPV DNA, the risk of cervical cancer increased by 3 times if they have used OC pills for 5 years or more ( 37 ). In addition, a recent systematic review & meta-analysis also suggested that OC pills use had a definite associated risk for developing cervical cancer especially for adenocarcinoma. This study concluded that use of OC pills is an independent risk factor in causing cervical cancer ( 38 ).

Cervical cancer screening

With the background of cervical cancer elimination worldwide, cervical cancer screening plays an increased role in the comprehensive prevention and control besides HPV vaccination, especially those methods that demonstrated excellent clinical performance.

Overview of cervical cancer screening methods

The screening methods for cervical cancer are mainly as following: traditional Pap smear, visual inspection with acetic acid & Lugol’s iodine (VIA/VILI), liquid-based cytology (LBC) and HPV testing. The disease burden of cervical cancer has been significantly reduced in developed countries by Pap smear, mainly in the United States, since 1950s. However, the accuracy of traditional Pap smear could be easily affected by following factors: the level of cytological room, professional technicians, sampling method, slide quality, dyeing skills, and cytological personnel experience. In developed countries with high standard experimental conditions and technical level, the sensitivity of cytology is as high as 80%−90%, in contrast, in resource-limited regions, it could be as low as 30%−40%. To overcome the limitations of traditional Pap smear in cervical cancer screening, LBC was developed and approved by Food and Drug Administration (FDA) in 1996 for clinical-use purpose. Compared with the traditional Pap smear, the sensitivity of LBC was significantly improved. Meanwhile, organized and practicable LBC screening program has also been established in developed countries which could ensure cervical cancer screening strategy is carried out continuously and effectively.

Cervical cancer screening has been facilitated since the cause clarified. HPV-based testing is a pivotal part for cervical cancer screening besides cytology-based tests.

The detection of high-risk HPV in cervical lesion biopsies and exfoliated cells has evolved from restriction endonuclease cleavage patterns and hybridization techniques to polymerase chain reaction (PCR)-based system ( 39 ) and most recently next-generation sequencing (NGS) assays ( 40 ). Currently, HPV genotyping is primarily based on the detection of individual types by various methods that utilizing the highly conserved L1 gene and PCR-based methods. These PCR methods employed consensus primers that could target and amplify different sized fragments such as 455 bp with the MY09/11|PGMY system ( 41 ), 150 bp with the GP5+/6+ system ( 42 ), or <100 bp with SPF10 ( 43 ). And another point that is worth noting is that all these techniques remained the most validated methodology to identify and characterize clinically relevant HPV ( 44 - 46 ).

Additionally, the type-specific probes are always to be used to achieve HPV genotyping, besides DNA sequencing ( 46 , 47 ). Other types of assays may be type-specific with immediate discrimination and quantitation of specific HPV types in an “onetube” assay. These methods employ real-time (RT)-PCR techniques, coupled with beta-globin detection for internal quality control utilizing specialized detection systems ( 48 ).

Cervical cancer malignant pathways are tightly correlated to the viral E6 and E7 oncoprotein activities which could also contribute to the accumulation of cellular genomic mutations and viral integration ( 47 ). Therefore, identification of HPV E6/E7 mRNA has been shown to be promising in cervical cancer screening. And most of the assays utilized reverse transcriptase PCR or nucleic acid sequence-based amplification to identify E6/E7 genome fragments ( 49 ).

Recently, the correlation between increased HPV CpG site methylation levels and high-grade cervical lesions has also been demonstrated in numerous studies and has facilitated the development of quantitative assays targeted CpG methylation ( 50 , 51 ). . Studies indicate that NGS assays can provide single-molecule CpG methylation levels to help unravel the mechanism of methylation in cervical cancer development ( 39 , 50 ).

The application of HPV detection has accelerated the transition of cervical cancer screening from morphology to molecular biology. HPV testing was initially used as a triage method for the reflex triage of population with atypical squamous cells of undetermined significance (ASC-US). In 2014, FDA approved HPV detection for the use in cervical screening. Thereafter, HPV detection plays an increasingly important role in the practice of cervical cancer screening. At present, more than 425 HPV testing has been developed worldwide, of which more than 150 is from China. To restrict and standardize HPV testing market, China released guidelines for the clinical performance evaluation for HPV testing against clinical endpoints in 2015. In other countries, it is also necessary to set similar regulations in consideration that 59.7% of HPV tests on the global market without a single peer-reviewed publication ( 49 ). To improve the coverage of cervical cancer screening, HPV testing that is rapid, simple, inexpensive could be more popular and can further promote the application in practice. In 2008, care HPV was developed in China, which demonstrated excellent performance in screening, although it was easy to use, cheap, fast and friendly to the laboratory requirements ( 52 , 53 ). In 2018, the care HPV achieved the pre-qualification certification issued by WHO, which was expected to benefit more people in developing countries and resource-poor areas such as Africa and Southeast Asia ( 54 ). In addition, the cost-effective reflex triage, referral of women, and management strategies appropriate to various resource level areas were also in evaluation ( 55 - 58 ).

In recent years, with the rapid development of science and technology, the application of artificial intelligence (AI) based products is booming. In cervical cancer prevention and control, AI also showed to be promising in cytology-based screening and colposcopy examination based on the image pattern recognition ( 59 , 60 ). These AI-based technology or system can intelligently identify lesions and assist medical staff in clinical examination and diagnosis which could alleviate difficulties in diagnosis in primary clinics.

Screening practice in China

In China, cervical cancer screening started since 1990s, although late compared with Western countries, China still achieved great breakthroughs. Common screening methods were introduced into China for the first time after clinical performance evaluation in high-risk areas which included HPV DNA detection (Hybrid Capture II, HC2), LBC and visual inspection with VIA/VILI ( 61 - 63 ). At the same time, these studies also further made it clear that “one or more HPV tests in a lifetime for cervical cancer screening could be feasible in developing countries” which had important impact on the clinical practice of cervical cancer screening in China and even in the world.

In July 2019, the State Council issued the “Healthy China Action (2019−2030)” plan, emphasizing the need to move forward the diagnosis and treatment and optimize the allocation of medical resources, from the treatment-centered to the health-centered, and to improve health level of the whole people. The program also clearly points out that cervical cancer screening coverage rate needs to reach more than 80% by 2030 ( 64 ), indicating the importance and severity of cervical cancer prevention and control.

Finally, the achievements of scientific research should be able to be developed into products and applied in practice. Based on the experience and study findings, two “National Demonstration Base for Early Diagnosis and Treatment of Cervical Cancer” were set up in Shenzhen Maternal and Child Health Hospital (City type) and Xiangyuan Maternal and Child Health Hospital (Rural type) in Shanxi Province in February 2005 ( 65 ). Thereafter, National Health and Family Planning Commission of China and China Women’s Federation launched cervical cancer and breast cancer screening program for women aged 35−64 years old in rural areas in 2009 ( 66 ), which was also one of the major public health service projects in China organized by national government. Different screening and management strategies have been set up for various resource-level regions. Up to 2017, the project has offered cervical cancer screening for 73.99 million women. Currently, the project has covered 1,501 counties ( 67 ). Meanwhile, China has developed effective cervical cancer prevention and control network which covered screening, diagnosis to treatment, follow-up and rehabilitation step by integrating government support and leadership, multi-sectors’ cooperation, professional personnel support and whole society participation. In 2017, Chinese Preventive Medicine Association released the “Guideline for Comprehensive Prevention and Control of Cervical Cancer” to further promote the standardized and development of cervical cancer prevention and control in China ( 68 ).

The priority of public health measures for cancer prevention and control reflects the government and society’s attention to public’s health, especially in resource-limited areas, and also reflects the civilization and progress of a country and society.

A large number of studies around the world have confirmed that cervical cancer could be prevented and controlled well by screening and early treatment. And it has been widely recognized if only considering the effect of cancer screening. However, the screening methods or solutions with the best effect may be not the best one. In the case of limited health resources, it is necessary to analyze and compare the input and output of different programs from the perspective of health economics which included how to scientifically determine the initial age of screening and time interval, select appropriate screening programs according to local health resources, and focus on cancer intervention in order to maximize the use of limited health resources. And then, we could determine the screening solution that not only has a good effect of disease prevention and control, but also is in line with the principle of cost-effectiveness.

Conclusions

The disease burden of cervical cancer has decreased significantly in developed countries and regions in last decades, however it is still serious in less developed countries and regions, and effective preventive measures in these areas still face serious challenges. At present, there are various available prevention and control measures that are cost-effective and scientific evidence-based to meet the needs of areas with different economic levels. It is gratifying to note that the globe has achieved a strategic consensus on the elimination of cervical cancer and also has developed and released the global strategy to accelerate the elimination of cervical cancer. Although the global elimination of cervical cancer has a long way to go, it is believed that through large-scale continuous promotion and widely use of existing effective prevention and control measures, cervical cancer will become the first cancer eliminated by human beings.

Acknowledgements

This study was supported by grants from the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (No. 2017-I2M-B&R-03 and No. 2016-I2M-1-019).

Conflicts of Interest : The authors have no conflicts of interest to declare.

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
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 23 January 2017

Integrated genomic and molecular characterization of cervical cancer

The cancer genome atlas research network.

Nature volume  543 ,  pages 378–384 ( 2017 ) Cite this article

108k Accesses

1004 Citations

246 Altmetric

Metrics details

• Cervical cancer
• Systems analysis

This article has been updated

Cervical cancer remains one of the leading causes of cancer-related deaths worldwide. Here we report the extensive molecular characterization of 228 primary cervical cancers, one of the largest comprehensive genomic studies of cervical cancer to date. We observed notable APOBEC mutagenesis patterns and identified SHKBP1 , ERBB3 , CASP8 , HLA-A and TGFBR2 as novel significantly mutated genes in cervical cancer. We also discovered amplifications in immune targets CD274 (also known as PD-L1 ) and PDCD1LG2 (also known as PD-L2 ), and the BCAR4 long non-coding RNA, which has been associated with response to lapatinib. Integration of human papilloma virus (HPV) was observed in all HPV18-related samples and 76% of HPV16-related samples, and was associated with structural aberrations and increased target-gene expression. We identified a unique set of endometrial-like cervical cancers, comprised predominantly of HPV-negative tumours with relatively high frequencies of KRAS , ARID1A and PTEN mutations. Integrative clustering of 178 samples identified keratin-low squamous, keratin-high squamous and adenocarcinoma-rich subgroups. These molecular analyses reveal new potential therapeutic targets for cervical cancers.

Similar content being viewed by others

Human papillomavirus integration perspective in small cell cervical carcinoma

Genomic alterations associated with mutational signatures, DNA damage repair and chromatin remodeling pathways in cervical carcinoma

Multi-omics mapping of human papillomavirus integration sites illuminates novel cervical cancer target genes

Cervical cancer accounts for 528,000 new cases and 266,000 deaths worldwide each year, more than any other gynaecological tumour 1 . Ninety-five per cent of cases are caused by persistent infections with carcinogenic HPVs 2 . Effective prophylactic vaccines against the most important carcinogenic HPV types are available, but the number of people receiving the vaccine remains low. Although early cervical cancer can be treated with surgery or radiation, metastatic cervical cancer is incurable and new therapeutic approaches are needed 3 .

While most HPV infections are cleared within months, some persist and express viral oncogenes that inactivate p53 and RB, leading to increased genomic instability, accumulation of somatic mutations, and in some cases, integration of HPV into the host genome 4 . The association with cancer risk and histological subtypes varies substantially among carcinogenic HPV types, but the reasons for these differences are poorly understood. Furthermore, clinically relevant subgroups of cervical cancer patients have yet to be identified. Here we present a comprehensive study of invasive cervical cancer conducted as part of The Cancer Genome Atlas (TCGA) project, with a focus on identifying novel clinical and molecular associations as well as functionally altered signalling pathways that may drive tumorigenesis and serve as prognostic or therapeutic markers.

Samples and clinical data

Primary frozen tumour tissue and blood were obtained from women with cervical cancer who had not received prior chemotherapy or radiotherapy ( Supplementary Information 1 and Supplementary Tables 1, 2 ). DNA, RNA and protein were processed as previously described 5 ( Supplementary Information 1, 3, 5 and 8 ). Mutations were called for 192 samples (the extended set), while all other platform (aside from protein) and integrated analyses were performed on a subset of 178 samples (the core set). Protein levels were measured on 155 samples (119 samples from both the core and extended sets plus 36 additional samples). The total number of non-overlapping samples in these three sets was 228 ( Extended Data Fig. 1a ). Of the 178 core-set samples, surgery was the primary treatment in 121 cases, median follow-up time was 17 months, and 145 patients were alive at the time of last follow-up. A committee of expert gynaecological pathologists reviewed most cases ( Supplementary Information 1 and Extended Data Fig. 1b–g ). The core set included 144 squamous cell carcinomas, 31 adenocarcinomas and 3 adenosquamous cancers.

Somatic genomic alterations

Whole-exome sequencing was performed on 192 extended-set tumour–blood pairs. All samples had at least 32 Mb of target exons covered with a median depth of 49× (range: 7–351×) for tumour samples and 47× (range: 9–341×) for normal samples. Collectively, the samples contained 43,324 somatic mutations, including 24,551 missense, 2,470 nonsense, 9,260 silent, 5,841 non-coding, 535 splice-site, 74 non-stop mutations, 475 frameshift insertions and deletions (indels) and 118 in-frame indels. Eleven tumours with outlier mutation frequencies (>600 per sample) were classified as ‘hypermutant’. The aggregate mutation density was 4.04 mutations per Mb across all tumours, and 2.53 when the hypermutant tumours were excluded.

Fourteen genes that are significantly mutated (SMGs) with false-discovery rates (FDR) < 0.1 were found using the MutSig2CV 6 algorithm ( Supplementary Table 4 ). We identified SHKBP1 , ERBB3 , CASP8 , HLA-A and TGFBR2 as novel SMGs in cervical cancer, and confirmed that PIK3CA , EP300 , FBXW7 , HLA-B , PTEN , NFE2L2 , ARID1A , KRAS and MAPK1 are SMGs, as previously reported 7 , 8 ( Fig. 1 , Extended Data Fig. 2a–g and Supplementary Fig. 6 ). Supplementary Table 4 shows the comparison of SMGs identified in the current TCGA set and a previously published dataset 8 . Mutations in 7 of the 14 SMGs in the TCGA set were present in at least one squamous cell carcinoma and one adenocarcinoma; however, mutations in HLA-A , HLA-B , NFE2L2 , MAPK1 , CASP8 , SHKBP1 and TGFBR2 were found exclusively in squamous tumours.

a – d , Cervical carcinoma samples ordered by histology and mutation frequency ( a ), clinical and molecular platform features ( b ), SMGs ( c ), and select somatic copy number alterations ( d ) are presented. SMGs are ordered by the overall mutation frequency and colour-coded by mutation type. Novel SMGs identified in squamous cell carcinomas are labelled in turquoise text. The number of APOBEC signature mutations (red) and other mutations (blue) present in every SMG is plotted to the right of the SMG panel and the number of gene-level somatic copy number alterations across all genes is plotted as gain (red) and loss (blue) to the right of the somatic copy number alteration panel. CN, copy number; SCNAs, somatic copy number alterations; Adeno., adenocarcinomas; Adenosq., adenosquamous cancers; Squamous, squamous cell carcinomas.

PowerPoint slide

PIK3CA had mostly activating helical-domain E542K and E545K mutations, with a marked relative decrease in mutations elsewhere in the gene ( Extended Data Fig. 2g ). This observation resembles findings in bladder cancer 9 and HPV-positive head and neck squamous cell cancers (HNSCs) 10 , but it differs from observations in breast and most other cancers 11 . The underlying nucleotide substitution pattern in the E542K and E545K mutations is associated with mutagenesis by a subclass of APOBEC cytidine deaminases 8 , 12 , 13 , 14 , 15 , with 150 out of 192 exomes displaying significant ( q  < 0.05) enrichment (up to sixfold) for the APOBEC signature. Further, the APOBEC mutation load correlated strongly with the total number of mutations per sample ( Extended Data Fig. 2h ), suggesting that APOBEC mutagenesis is the predominant source of mutations in cervical cancers.

We found an average of 88 somatic copy number alterations per tumour, fewer than in HNSC, ovarian and serous endometrial carcinomas, but more than in endometrioid endometrial carcinomas 10 , 16 , 17 . GISTIC2.0 analysis (with a threshold of q  < 0.25) revealed 26 focal amplifications and 37 focal deletions along with 23 recurrently altered whole arms ( Extended Data Fig. 3c and Supplementary Table 7 ). Novel recurrent focal amplification events were identified (in genomic order) at 7p11.2 ( EGFR , 17%), 9p24.1 ( CD274 , PDCD1LG2 , 21%), 13q22.1 ( KLF5 , 18%) and 16p13.13 ( BCAR4 , 20%). Other previously reported amplification events occurred at 3q26.31 ( TERC , MECOM , 78%), 3q28 ( TP63 , 77%), 8q24.21 ( MYC , PVT1 , 42%), 11q22.1 ( YAP1, BIRC2 , BIRC3 , 17%), and 17q12 ( ERBB2 , 17%). Novel recurrent deletions were identified at 3p24.1 ( TGFBR2 , 36%) and 18q21.2 ( SMAD4 , 28%), in addition to previously identified deletions at 4q35.2 ( FAT1 , 36%) and 10q23.31 ( PTEN , 31%). A cluster with high copy number alterations mostly contained squamous tumours with amplification events involving 11q22 ( YAP1 , BIRC2 , BIRC3 ) and 7p11.2 ( EGFR ), whereas the cluster containing low copy number variations included most adenocarcinomas and was enriched for tumours with deletions in TGFBR2 and SMAD4 , and gains in ERBB2 and KLF5 ( Extended Data Fig. 3a, b ). Notably, both groups had amplifications involving CD274 (PD-L1) and PDCD1LG2 (PD-L2) that correlated significantly ( P  < 0.0001) with expression of two key immune cytolytic effector genes, granzyme A and perforin 18 ( Extended Data Fig. 3d ). This highlights the potential of immunotherapeutic strategies for a subset of cervical cancers.

Structural rearrangements were identified by analysis of RNA sequencing (RNA-seq) (core set, n  = 178) and whole-genome sequencing (WGS) data with low-pass ( n  = 50) and deep ( n  = 19) coverage. Both RNA-seq and WGS detected 22 putative structural rearrangements in 14 patients ( Supplementary Table 8 ). In total, 26 recurrent fusions were found ( Supplementary Table 9 , with examples in Extended Data Fig. 4d ). RNA-seq analysis revealed four samples with 16p13 ZC3H7A – BCAR4 gene fusions, whereby exon 1 of ZC3H7A was linked to the last exon of BCAR4 . WGS revealed tandem duplication and copy number gain of BCAR4 on chromosome 16p13.13 ( Extended Data Fig. 4c ). BCAR4 is a metastasis-promoting long non-coding RNA that enhances cell proliferation in oestrogen-resistant breast cancer by activating the HER2/HER3 pathway. Lapatinib, an EGFR/HER2 inhibitor, counteracts BCAR4 -driven tumour growth in vitro , and warrants evaluation as a possible therapeutic agent in BCAR4 -positive cervical cancer 19 .

Integrated analysis of molecular subgroups

Integration of copy number, methylation, mRNA and microRNA (miRNA) data using iCluster 20 highlighted the molecular heterogeneity of cervical carcinomas. Three clusters were identified that largely corresponded to mRNA clusters ( Supplementary Fig. 9 ): a squamous cluster with high expression of keratin gene family members (keratin-high), another squamous cluster with lower expression of keratin genes (keratin-low), and an adenocarcinoma-rich cluster (adenocarcinoma). Keratin-high and keratin-low clusters included 133 out of 144 squamous cell carcinomas and the adenocarcinoma cluster contained 29 out of 31 adenocarcinomas ( Fig. 2 ). KRAS ( P  = 9.7 × 10 −5 ), ERBB3 ( P  = 2.6 × 10 −3 ) and HLA-A ( P  = 0.03) mutations were significantly associated with clusters, whereby KRAS mutations were absent from the keratin-high cluster and HLA-A mutations were absent from the adenocarcinoma cluster ( Fig. 2 ). Members of the SPRR and TMPRSS cornification gene families and the SMGs ARID1A ( P  = 0.02), NFE2L2 ( P  = 6.9 × 10 −6 ) and PIK3CA ( P  = 0.01) were differentially expressed between keratin-low and keratin-high clusters ( Extended Data Fig. 4b ).

a , Integrative clustering of 178 core-set cervical cancer samples using mRNA, methylation, miRNA and copy number variation (CNV) data identifies two squamous-carcinoma-enriched groups (keratin-low and keratin-high) and one adenocarcinoma-enriched group, as shown in the feature bars (top). Features presented include histology, HPV clade, HPV integration status, UCEC-like status, APOBEC mutagenesis level, mRNA EMT score, tumour purity and three SMGs ( KRAS , ERBB3 and HLA-A ) that are significantly associated across the three clusters identified with iCluster ( ERBB2 is presented for comparison purposes with its family member ERBB3 ). b , The cluster of clusters panel displays subtypes defined independently by mRNA, miRNA, methylation, reverse phase protein array (RPPA), CNV and PARADIGM data. C1–C6 indicate clusters. Black, sample is not represented in the cluster; red, sample is represented in the cluster; grey, data not available. c , The heatmaps show select mRNAs, miRNAs, proteins and CNVs that are either significantly associated with iCluster groups or have been identified as markers in other analyses. The heatmap colour scale bar represents the scale for the features presented in the heatmaps with a breakpoint of zero represented by white. APOBEC mut., APOBEC mutagenesis; inter., intermediate.

Unsupervised hierarchical clustering of variable DNA-methylation probes produced three groups ( Extended Data Fig. 5a ), including a small ‘CpG island hypermethylated’ (CIMP-high) cluster, a CIMP-intermediate cluster and a CIMP-low cluster that were associated with an epithelial–mesenchymal transition (EMT) mRNA score 10 , 21 ( Extended Data Fig. 5b ). Most of the samples in the adenocarcinoma cluster were CIMP-high, whereas the other iCluster groups contained a mixture of CIMP-intermediate and CIMP-low samples ( Fig. 2 ). Comparing all cervical carcinomas to 120 normal samples drawn from 12 TCGA projects, we identified 1,026 epigenetically silenced genes that were methylated to a greater extent in cancers than in normal tissues, including several zinc-finger (ZNF), protease (ADAM, ADAMTS), and collagen (COL) genes ( Supplementary Tables 11 and 12 ).

Unsupervised clustering resulted in six miRNA clusters that were associated with the iCluster groups ( P  = 1.7 × 10 −19 ) ( Extended Data Fig. 6a ). Samples from the adenocarcinoma cluster almost exclusively overlapped with miRNA cluster 5, and were characterized by high expression of miR-375 and low expression of miR-205-5p and miR-944 ( Supplementary Table 31 ). Expression levels of tumour suppressors miR-99a-5p and miR-203a were significantly higher in samples from the keratin-high cluster than samples from the keratin-low cluster ( Supplementary Table 31 ; P  = 0.01 and P  = 0.008, respectively). Among miRNAs with significant and functionally validated gene and protein anti-correlations 22 , one large subnetwork involved the miR-200 family and other miRNAs with expression patterns that anti-correlated with those of the EMT-related transcription factors ZEB1 , ZEB2 and SNAI2 , the Hippo and p73 transcriptional co-factor YAP1 , the receptor tyrosine kinases (RTKs) ERBB2 , ERBB3 and AXL , and the hormone receptor ESR1 ( Extended Data Fig. 6b , Supplementary Figs 17 , 18 and Supplementary Table 15 ).

Reverse phase protein array (RPPA) analysis of 155 samples with 192 antibodies ( Extended Data Fig. 1a and Supplementary Table 17 ) identified three clusters significantly associated with the iCluster groups ( P  = 1.8 × 10 −4 ) and EMT mRNA score ( Fig. 3a, c, d and Supplementary Table 16 ). Samples from the EMT cluster were enriched in the keratin-low cluster, whereas PI3K–AKT and hormone cluster samples were enriched in the keratin-high and adenocarcinoma clusters, respectively, suggesting distinct pathway activation across integrated cervical cancer subtypes. Differential expression levels of phosphorylated (p)-MAPK, p-EGFR (Y1068), p-SRC (Y416), IGFBP2 and TIGAR between keratin-high and keratin-low clusters suggest diverse activation patterns of RTK, MAPK, PI3K and metabolic signalling pathways that may underlie the molecular diversity of cervical squamous cancers ( Fig. 2 ).

a , Clustered heatmap of samples (columns) and 192 antibodies (rows) for 155 samples (112 overlap with the core set of 178; see Extended Data Fig. 1a ). Clusters presented from left to right include hormone (dark blue), EMT (red) and PI3K–AKT (green). A subset of proteins differentially expressed between the clusters is highlighted. Tracks for clinical and molecular features are shown for features that were significantly associated with RPPA clusters ( P  < 0.05). Correlation between RPPA clusters and other categorical variables were detected by χ 2 test, whereas correlations with continuous variables were analysed using the non-parametric Kruskal–Wallis test. In the heatmap, blue represents downregulated expression, red represents upregulated expression and white represents no change in expression. NA, data not available. b , Five-year Kaplan–Meier survival curves and log-rank test P value ( P  = 6.1 × 10 −4 ) comparing overall survival (OS) across all RPPA clusters using 115 silhouette width core samples (silhouette core; see Supplementary Information 8 ). c , EMT mRNA score levels were calculated for all samples and compared across RPPA clusters. P  = 0.001 (one-way ANOVA). d , Pathway scores for EMT, hormone-receptor and PI3K–AKT signalling pathways are presented for all RPPA clusters ( x axis); Kruskal–Wallis test used to identify significant pathway score differences between the clusters.

The core members of each RPPA cluster with the highest silhouette width (>0.02, n  = 115) were associated with five-year survival ( Fig. 3b ; P  = 6.1 × 10 −4 ), with the EMT group exhibiting worse outcome. Notably, this was the only platform where clusters associated with outcomes ( Supplementary Figs 8 , 9 , 12 and 22 ; Supplementary Information 6 ). Samples in the EMT cluster exhibited high ‘reactive’ pathway scores 11 ( Supplementary Fig. 20 ), illustrating for the first time in cervical cancer the presence of a subset of stromal reactive tumours that have high expression of caveolin-1, MYH11 and RAB11, a subset which also appears in other diseases 23 ( Supplementary Table 16 ). YAP was the most significantly differentially expressed protein distinguishing samples from the EMT cluster from all others ( Supplementary Table 18 ; P  = 1.7 × 10 −15 ) and YAP1 was significantly amplified in the samples from the EMT cluster compared to the hormone ( P  = 1.1 × 10 −5 ) and PI3K–AKT cluster ( P  = 6.4 × 10 −4 ) samples. Regulation of the EMT-related molecules YAP and ZEB1 (refs 24 , 25 , 26 ) may also be driven by significantly lower expression levels of miR-200a-3p in the samples from the EMT cluster compared to samples from the other RPPA clusters ( Extended Data Figs 6b , 7a ; P  = 3.8 × 10 −3 ). These results highlight potential roles for YAP and reactive stroma in EMT-regulated progression of cervical cancers.

The mutual exclusivity modules in cancer (MEMo) algorithm 27 uses somatic-mutation and copy number data to identify oncogenic networks with mutually exclusive genomic alterations. Because miR-200a and miR-200b (miR-200a/b) expression was negatively correlated with EMT mRNA scores ( Extended Data Fig. 7b, d ), we used MEMo to examine alterations in miR-200a/b and EMT gene networks and found a potential link between the TGFβ pathway and miR-200a/b alterations in regulating EMT 28 , 29 . Deletions and mutations affecting the receptor gene TGFBR2 , the modulating genes CREBBP and EP300 , and the transcription factor SMAD4 probably all affect growth-suppressive and pro-apoptotic functions driven by TGFβ ( Fig. 4c ) and were observed in 30% of squamous cell carcinomas ( Fig. 4d ). Tumours with both hypermethylation and downregulation of miR-200a/b (referred to as altered) were restricted to squamous cell carcinomas, were enriched in the keratin-low cluster ( Fig. 4d and Extended Data Fig. 8 ; P  = 0.001 for both miR-200a and miR-200b), showed significant upregulation of both ZEB1 and ZEB2 ( Extended Data Fig. 9a–d ), and were mutually exclusive with alterations in the TGFβ signalling pathway ( Fig. 4d ). Notably, samples with altered miR-200a/b exhibited higher EMT mRNA scores than unaltered samples, whereas no significant difference was found between samples with or without TGFβ-pathway alterations ( Fig. 4d and Extended Data Fig. 7c, e ). These findings highlight potential treatment approaches for this subgroup of cervical cancer patients, as targeting EMT may render tumours more sensitive to small-molecule inhibitors and cytotoxic chemotherapy 21 , 30 , 31 .

a , Multiple alterations affect RTK, AKT and MAPK signalling in both squamous cell carcinomas and adenocarcinomas. A schematic diagram of the pathways is shown for altered genes along with the percentage of alteration in squamous cell carcinomas and adenocarcinomas. Significant P values ( P  < 0.05, Student’s t -test) for alteration frequency differences between squamous cell carcinomas and adenocarcinomas are listed at the gene level, with significantly different genes marked with an asterisk. b , Distinct types of alterations (amplification, deletion, missense mutation and truncating mutation) affect genes (rows) in these pathways in each sample (columns). c , TGFβ signalling is frequently altered in cervical tumours. Alterations in this pathway are divided between those probably affecting TGFβ-tumour-suppressive functions and those affecting the TGFβ-driven EMT program. d , Samples with alterations targeting TGFβ-tumour-suppressive functions do not show significantly different EMT scores compared with all other samples; however, samples with low expression/high methylation of miR-200a/b have significantly higher EMT scores than all other samples. miR-down, samples met double threshold of methylated and downregulated as described in Methods. NS, not significant. Percentages in b and d , indicate per cent of the total histological subgroup population.

MEMo analysis also showed differences in therapeutically relevant alterations in RTK, PI3K and MAPK pathways across cervical cancers. MEMo identified mutual exclusivity modules involving alterations within both the PI3K and MAPK pathways ( Supplementary Table 27 ; adjusted P  = 0.06); however, there was a strong tendency for co-occurrence of ERBB2 and ERBB3 alterations within adenocarcinomas ( P  < 0.001, log odds ratio > 3), indicating that a subset of these tumours may exhibit aberrant HER3 signalling through interactions between mutant HER3 and activated HER2 and therefore could potentially benefit from HER2- and HER3-targeted therapies 32 ( Fig. 4a, b ). Although not statistically significant, aberrations in PIK3CA also tended to co-occur with PTEN somatic mutations and deletions ( P  = 0.078, log odds ratio = 0.71), which is similar to endometrial tumours with few copy number alterations and suggests potential therapeutic benefit from PI3K-pathway-targeting agents 17 .

PARADIGM 33 , 34 , which integrates copy number, RNA-seq and pathway-interaction data, showed markedly different pathway activation profiles between squamous carcinomas and adenocarcinomas ( Extended Data Fig. 10 and Supplementary Fig. 48 ). PARADIGM identified higher inferred activation of p53, p63, p73, AP-1, MYC, HIF1A, FGFR3 and MAPK signalling as key distinguishing features of squamous cell carcinomas, similar to other squamous cancers 35 . By contrast, adenocarcinomas exhibited higher inferred activation of ERα, FOXA1, FOXA2 and FGFR1 pathways ( Extended Data Fig. 10 , Supplementary Figs 25 , 48 and Supplementary Table 18 ). Possible underlying mechanisms for ERα upregulation may stem from the expression of miR-193b-3p, a direct regulator of ESR1 that was significantly downregulated in adenocarcinomas compared to squamous carcinomas ( Fig. 2 , Extended Data Fig. 6 and Supplementary Table 14 ; P  = 0.04), or from oestrogen signalling in stromal cells 36 .

Cross-cancer analysis

We next evaluated the relationship of cervical cancer subtypes with endometrial cancer, an adjacent cancer site with hormone-related carcinogenesis, and HNSC, a subset of which is caused by HPV. For this, hierarchical clustering of cervical, uterine corpus endometrial (UCEC) 17 , and HNSC 10 mRNA-expression data was performed. Three major groups were observed, with cluster 1 including all UCEC samples and most cervical adenocarcinomas and characterized by overexpression of hormone-receptor genes ESR1 and PGR ( Extended Data Fig. 4a ). Cluster 2 included predominantly squamous cervical carcinomas and 23 out of 27 HPV-positive HNSC samples. Cluster 3 included few cervical cancers and the remaining HNSC cancers, which were mostly HPV-negative. This highlights the similarity of HPV-related squamous cancers at different anatomical sites.

Since a subset of cervical cancers clustered with endometrial samples, a gene-expression classifier was developed to predict whether carcinomas were cervical or endometrial ( Supplementary Information 5 ). We classified 8 out of 178 (4.5%) cervical cancer samples as endometrial-like (UCEC-like) cancers, which were confirmed to be cervical cancers by study pathologists ( Extended Data Fig. 1f, g ). These tumours included 7 out of 9 HPV-negative cancers and 5 of the 8 were adenocarcinomas. Six UCEC-like cancers were in the adenocarcinoma cluster and two were in the keratin-low cluster. Despite their low number, the UCEC-like tumours accounted for 33%, 27% and 20% of mutations in ARID1A , KRAS and PTEN , respectively. They were associated with the RPPA hormone and miRNA C6 clusters, and all but one sample was CIMP-low and copy number-low ( Supplementary Table 1 ).

HPV genotypes, variants and integration

Of the 178 core-set tumours, 169 (95%) were HPV-positive, 120 (67%) had alpha-9 (A9) types (103 HPV16), 45 (25%) had alpha-7 (A7) types (27 HPV18), and 9 (5%) were HPV-negative ( Supplementary Table 3 ). HPV variants were predominantly European (137 out of 169, 81% A variants), and there was a significant association of non-European HPV16 variants with cervical adenocarcinomas ( Supplementary Table 3 ; odds ratio = 5.3, P  = 3 × 10 –3 ). All HPV-positive cancers had detectable expression of HPV E6- and E7-oncogene mRNAs, which encode proteins that inhibit p53 and RB function, respectively 37 , 38 . Notably, HPV18 cancers had significantly higher ratios of unspliced to spliced transcripts encoding the active E6 oncoprotein than the HPV16 cancers ( Extended Data Fig. 11a ; P  = 2 × 10 –10 ), suggesting different functional implications of E6 and E7 in cancers associated with different HPV genotypes.

HPV A7 types were enriched in the keratin-low and adenocarcinoma clusters ( P  = 5 × 10 –4 ). Most HPV clade A7 tumours were CIMP-low, and HPV-negative tumours formed a distinct subgroup within the CIMP-low cluster with a significantly lower mean promoter-methylation level than other samples in that cluster ( Extended Data Fig. 5a ; P  = 5 × 10 −3 ). Samples with the highest rate of gene silencing were HPV-positive adenocarcinomas, particularly those related to A9 types ( t -test P  < 0.001). Functional epigenetic module ( Supplementary Information 13 ) analysis 39 , which integrates DNA-methylation and gene-expression data using protein–protein interaction networks, identified inverse correlations between methylation and gene expression in HPV-positive versus HPV-negative cervical cancers and HPV-positive ( n  = 36) versus HPV-negative ( n  = 243) HNSCs. The analysis revealed 12 statistically significant subnetworks for cervical cancer and 11 for HNSCs, with one common subnetwork centred around Forkhead Box A2 ( FOXA2 ) ( Supplementary Table 19 and Supplementary Fig. 32 ). miR-944, miR-767-5p and miR-105-5p were the most differentially expressed miRNAs between HPV-positive and HPV-negative samples ( Supplementary Fig. 14e ). miR-944 expression was also significantly higher, whereas miR-375 expression was significantly lower in HPV16-positive squamous cancers compared to HPV18-positive squamous cancers ( Supplementary Fig. 14d ). Notably, HPV-negative cancers had a significantly higher EMT mRNA score and a lower frequency of the APOBEC mutagenesis signature compared with HPV-positive tumours ( Extended Data Fig. 11b and Supplementary Fig. 27 ; P  = 0.02 and P  = 0.004, respectively).

PARADIGM was used to evaluate molecular pathways differentially activated in squamous samples with A7- and A9-HPV infections. We observed higher inferred activation of p53 and p63 signalling and lower FOXA1 signalling in tumours infected with A9 types ( Fig. 5a and Supplementary Fig. 23a ). Higher SFN pathway activation was also observed for A9-positive tumours, which is consistent with the low methylation and high gene-expression patterns of SFN found in functional epigenetic module analysis ( Fig. 5a and Supplementary Table 19 ). Notably, the SFN -encoded stratifin (also known as 14-3-3σ) adaptor protein has previously been associated with epithelial immortalization and squamous cell cancers 40 , 41 , altered p53-pathway activation 42 , and Wnt-mediated β-catenin signalling 43 .

a , Cytoscape display of the largest interconnected regulatory network of PARADIGM integrated pathway level (IPL) features showing differential inferred activation between HPV A9 and A7 squamous carcinomas ( n  = 101 and n  = 35, respectively). Node colour and intensity reflect the level of differential activation. Node size represents level of significance. SFN is within a subnetwork identified by functional epigenetic module analysis ( Supplementary Information 13 ) as disrupted between HPV A9 and A7 squamous cell carcinomas, and is highlighted using a bold black outline. rRNA, ribosomal RNA. DST , DST isoform 3. b , Circos plot showing frequency (0–100%) of gains and losses for regions of each chromosome (outer circle). Lines within the inner circle indicate integration breakpoints from the HPV genome ( L1 , L2 , E1 , E2 , E4 , E5 , E6 and E7 genes) to the human genome as defined in Methods, Supplementary Information 2 , and Supplementary Table 3 . Lines are colour coded by HPV clade.

Viral–cellular fusion transcripts indicating integration of HPV into the host genome were observed in 141 out of 169 (83%) HPV-positive cancers, including all HPV18-positive cancers. Of these 141 samples, 90 (64%) had a single HPV integration event, 35 had two events, and 16 had three or more events (totalling 220 unique integration events) ( Supplementary Table 3 ). HPV integration events affected all chromosomes, including some previously described hotspots such as 3q28 and 8q24 (ref. 44 ) ( Fig. 5b ). Genomic loci affected by integration were characterized by increased somatic copy number alterations ( P  = 6.9 × 10 −13 for HPV16 and P  = 0.058 for HPV18) and increased gene expression ( P  = 1.6 × 10 −11 for HPV16 and P  = 0.011 for HPV18) ( Extended Data Fig. 11c, d ). In addition, 153 (70%) fusion transcripts included known or predicted genes, whereas the remainder included intergenic regions ( Fig. 5b and Supplementary Table 3 ).

Through comprehensive molecular and integrative profiling, we identified novel genomic and proteomic characteristics that subclassify cervical cancers. Integrated clustering identified keratin-low squamous, keratin-high squamous, and adenocarcinoma-rich clusters defined by different HPV and molecular features ( Extended Data Fig. 8 ). ERBB3 , CASP8 , HLA-A , SHKBP1 and TGFBR2 were identified as SMGs for the first time in cervical cancer, with ERBB3 (HER3) immediately applicable as a therapeutic target. For the first time in cancer, we report amplifications and fusion events involving the BCAR4 gene, which can be targeted indirectly by lapatinib. Further, we identified amplifications in CD274 and PDCD1LG2 , two genes that encode well-known immunotherapy targets. A set of endometrial-like cervical cancers comprised predominantly of HPV-negative tumours and characterized by mutations in KRAS , ARID1A and PTEN was discovered, with PTEN and potentially ARID1A proteins serving as therapeutic targets. Importantly, over 70% of cervical cancers exhibited genomic alterations in either one or both of the PI3K–MAPK and TGFβ signalling pathways ( Extended Data Fig. 9e ), illustrating the potential clinical significance of therapeutic agents targeting members of these pathways. For the first time, we report distinct molecular pathways activated in cervical carcinomas caused by different HPV types, highlighting the biological diversity of HPV effects.

Together, these findings provide insight into the molecular subtypes of cervical cancers and rationales for developing clinical trials to treat populations of cervical cancer patients with distinct therapies.

Data reporting

No statistical methods were used to predetermine sample size. The experiments were not randomized and the investigators were not blinded to allocation during experiments and outcome assessment.

Samples and data freeze

Each tissue source site provided documentation that their IRBs either: a) approved their participation specifically in the TCGA project, through an approved protocol, amendment, exemption, or waiver, and the documentation must include specific mention of TCGA; or b) provided documentation that the IRB does not consider participation in TCGA to constitute ‘human subjects research’, and therefore does not have purview.

The Core Data Freeze (core set) included 178 samples from cervical carcinoma batches 88, 114, 127, 148, 169, 179, 200, 217, 236, 256, 280, 297, 335 and 350 ( Supplementary Table 1 ). This is a standard data freeze whereby the case list was ‘frozen’ and analyses used the one set even though other samples came through the pipeline. Samples in the core set had mRNA-seq, whole exome DNA-seq (WES), miRNA-seq, methylation, SNP6 copy number and clinical data available. Additional samples that had multicentre mutation calls and/or RPPA data included 67 samples from cervical carcinoma batches 88, 114, 127, 148, 169, 179, 200, 217, 236, 256, 280, 297, 335, 350, 361, 373, 380, 394 and 420 ( Supplementary Table 2 ). Of these samples, 14 had mutations called and 60 had RPPA data available; however, RPPA data for 17 samples was excluded owing to low protein content within the samples ( Supplementary Table 2 ). Mutations were called for 192 samples (extended set), while all other platform and integrated analyses (aside from protein) were performed on the subset of 178 core-set samples. Protein levels were measured on 155 samples, which included 119 samples from both the core and extended sets as well as 36 samples outside of these sets. The total number of non-overlapping samples across core, extended and RPPA datasets is 228 ( Extended Data Fig. 1a ).

HPV detection, variant calling and transcript analysis

HPV status was determined using consensus results from MassArray and RNA-seq ( Supplementary Information 2 ). MassArray uses real-time competitive polymerase chain reaction and matrix-assisted laser desorption/ionization–time-of-flight mass spectroscopy with separation of products on a matrix-loaded silicon chip array, similar to the work described in ref. 45 . Two approaches for pathogen detection from RNA-seq data were used. The first used the microbial detection pipeline at the British Columbia Cancer Agency’s Genome Sciences Centre (BC), which is based on BioBloom Tools (BBT, v1.2.4b1) 46 . The second used the PathSeq algorithm 47 at the Broad Institute (BI) to perform computational subtraction of human reads followed by alignment of residual reads to a combined database of human reference genomes and microbial reference genomes including HPV. In 97% of samples, complete agreement between MassArray and both RNA-seq approaches was observed. The remaining discrepant samples were resolved by majority decision, assigning the genotype called by at least two of the methods. RNA-seq data in FASTA format was used to identify HPV variants ( Supplementary Fig. 1 ). Unaligned reads were taken from the PathSeq analysis and aligned to HPV reference genomes using TopHat 48 with default parameters 49 . The HPV variant lineages/sublineages were assigned based on the phylogenetic topology and confirmed visually using the SNP patterns 50 . HPV splice junctions from RNA-seq were determined using TopHat. Two transcript types were distinguished for HPV16 and HPV18: transcripts that included evidence of an unspliced sequence of E6, and transcripts spliced at the E6 splice donor site (position 226 for HPV16 and position 233 for HPV18) ( Supplementary Fig. 2 ). Read counts for unspliced, spliced, as well as the ratio of unspliced/spliced transcripts were categorized into quartiles separately for HPV16 and HPV18.

HPV integration analysis

Using RNA-seq data, concordance of integration events based on alignments of contigs from de novo transcriptome assembly (BC) and read alignments (BI) was evaluated ( Supplementary Fig. 3 ). We identified method-specific integration events by assigning all sites within a 500-kb sliding window to a single integration event located at the median coordinate of that assigned sites for that event. An integration event was labelled as ‘confident’ when the total read support for each of its supporting integration sites passed centre-specific read evidence thresholds. To take advantage of differences between the two integration methods (that is, contig and read), for the concordance analysis we used all method-specific integration events (both confident and non-confident events). We labelled an integration event as ‘concordant’ when both methods reported an integration event within 500 kb in the same patient’s sample. For some concordant events, both methods reported a confident event. An integration event was labelled as ‘discordant’ when only one centre reported a confident integration event within 500 kb ( Supplementary Figs 4 and 5 ). For both intragenic and intergenic concordant events, we reported a range of coordinates that extends from the most proximal to the most distal supported integration site. We assessed gene-level expression relative to somatic copy number and structural-variant data for genes into which we had mapped viral–human junctions from RNA sequencing data and for genes that were associated with enhancers into which we had mapped RNA junctions.

DNA sequencing and mutation calling

Detailed methods for library hybrid capture, read alignments and somatic variant calling are documented in Supplementary Information 3 . MutSig2CV 6 was used to identify significantly mutated genes (SMGs) within the cervical cancer exome sequencing data. Mutations were analysed for the core set plus 14 samples for a total of 192 extended-set samples. Eleven samples were identified to exhibit greater than average mutations rates and were termed hypermutants (somatic mutations > 600). These 11 samples were excluded from the analysis for identifying SMGs. All three sample subsets (all samples, squamous carcinomas only, adenocarcinomas only) without hypermutants ( Supplementary Table 4 ) were analysed using an FDR cut-off of 0.1. FDR values are shown in Supplementary Table 4 . SMG analysis using the entire sample cohort in from ref. 8 was performed as described previously 8 .

Copy number analysis

DNA from each tumour or germline sample was hybridized to Affymetrix SNP 6.0 arrays using protocols at the Genome Analysis Platform of the Broad Institute as previously described 51 . Briefly, Birdseed was used to infer a preliminary copy number at each probe locus from raw .cel files 52 . For each tumour, genome-wide copy number estimates were refined using tangent normalization, in which tumour signal intensities are divided by signal intensities from the linear combination of all normal samples that are most similar to the tumour 16 . Individual copy number estimates then underwent segmentation using circular binary segmentation 53 , and segmented copy number profiles for tumour and matched control DNAs were analysed using Ziggurat Deconstruction 54 . Significance of copy number alterations were assessed from the segmented data using GISTIC2.0 (version 2.0.22) 54 . For the purpose of this analysis, an arm-level event was defined as any event spanning more than 50% of a chromosome arm. For copy number-based clustering, tumours were clustered based on copy number at regions using GISTIC analysis. Clustering was done in R on the basis of Euclidean distance using Ward’s method. Allelic and integer copy number, tumour purity and tumour ploidy were calculated using the ABSOLUTE algorithm 55 .

Detecting structural variants from RNA-seq and WGS data

Integrative analysis was performed to identify putative driver fusions using both WGS (low-pass and high-coverage) and RNA-seq data. RNA-seq data for 178 core-set samples were analysed using the TopHat-Fusion and BreakFusion, PRADA and MapSplice algorithms. To identify structural variations in WGS data, 50 low-pass WGS and 19 high-pass WGS samples were analysed. Detection of structural variations in low-pass WGS data was performed using two algorithms, BreakDancer 56 and Meerkat 57 , with a requirement for at least two discordant read pairs supporting each event and at least one read covering the breakpoint junction. High-pass WGS data were analysed to detect somatic structural variations using two runs of BreakDancer and one run of SquareDancer ( https://github.com/ding-lab/squaredancer ). The gene fusion lists generated by all methods and platforms were integrated (see Supplementary Tables 8–10 ).

APOBEC mutagenesis analysis

Analysis is based on previous findings that APOBECs deaminate cytidines predominantly in a tCw motif and that the APOBEC mutagenesis signature is composed of approximately equal numbers of two kinds of changes in this motif: tCw→tTw and tCw→tGw mutations (flanking nucleotides are shown in small letters; w = A or T). Using mutation data from all 192 extended-set samples, we calculated on a per-sample basis the enrichment of the APOBEC mutation signature among all mutated cytosines in comparison to the fraction of cytosines that occur in the tCw motif among the ±20 nucleotides surrounding each mutated cytosine (APOBEC_enrich column in data files). The minimum estimate of the number of APOBEC-induced mutations in a sample (APOBEC_MutLoad_MinEstimate) was calculated using the formula: [tCw→G + tCw→T]×[(APOBEC_enrich−1)/APOBEC_enrich], which allows estimation of the number of APOBEC signature mutations in excess of what would be expected by random mutagenesis. APOBEC_MutLoad_MinEstimate was calculated only for samples passing the threshold of FDR < 0.05 for APOBEC enrichment ([BH_Fisher_p-value_tCw] < 0.05). Samples with a BH_Fisher_p-value_tCw > 0.05 were given a value of 0. The APOBEC_MutLoad_MinEstimate value shows high correlation (0.9–0.95) with all other parameters used to characterize the APOBEC mutagenesis pattern, such as APOBEC enrichment, and absolute and relative APOBEC mutation loads. For some analyses and figures, the APOBEC_MutLoad_MinEstimate parameter was converted into categorical values as follows: no, APOBEC_MutLoad_MinEstimate = 0; low, 0 < APOBEC_MutLoad_MinEstimate > median of non-zero values; high, APOBEC_MutLoad_MinEstimate > median of non-zero values. The median of non-zero values in the extended set = 33.

Methylation analysis

The Illumina Infinium HM450 array 58 was used to evaluate DNA methylation in the core set of samples from cervical cancer patients. Unsupervised consensus clustering was performed with Euclidean distance and partitioning around medoids (PAM) using the most variable 1% of CpG-island promoter probes. Epigenetically silenced genes were identified as previously described 59 . A total of 120 normal samples were used for this analysis by selecting 10 samples at random from the 12 TCGA projects that included normal samples.

RNA-seq analysis

RNA was extracted, converted into mRNA libraries, and paired-end sequenced (paired 50 nucleotide reads) on Illumina HiSeq 2000 Genome Analyzers as previously described 5 . RNA reads were aligned to the hg19 genome assembly using Mapsplice version 12_07 60 . Gene expression was quantified for the transcript models corresponding to the TCGA GAF2.1 ( https://gdc-api.nci.nih.gov/v0/data/a0bb9765-3f03-485b-839d-7dce4a9bcfeb ) using RSEM4 (ref. 61 ) and normalized within a sample to a fixed upper quartile. To predict whether a cancer sample was from the cervix or the uterus, the data matrix of normalized gene-level RSEM values from 170 UCEC samples was merged with the data matrix from the core set ( n  = 178) of cervical cancers. This merged dataset was then randomly split into a training set (87 cervical carcinoma samples; 86 UCEC samples) and a test set (91 cervical carcinoma samples; 84 UCEC samples). A sample was predicted to be cervical carcinoma if the t -statistic versus UCEC was significant ( P  < 0.05), but was not significantly different from the cervical carcinoma mean (and vice versa for the UCEC prediction). A data matrix of RSEM values from 178 cervical carcinoma, 170 UCEC and 279 HNSC samples was used to identify expression patterns across the 3 cancer types. The gene expression matrix was further filtered to only include the top 25% most variable genes by mean absolute deviation ( n  = 4,039 genes).

EMT mRNA score analysis

The EMT score was computed as previously described 10 , 21 . Briefly, the EMT score was the value resulting from the difference between the average expression of mesenchymal (M) genes minus the average expression of epithelial (E) genes. All values for unavailable data (NA) were removed from the calculation. Two-sample t -test and ANOVA were applied to each comparison accordingly.

miRNA sequencing and analysis

MicroRNA-sequencing (miRNA-seq) data was generated for the core set of tumour samples using methods described previously 11 . We identified miRNAs that have been associated with EMT 62 , 63 , 64 , 65 , 66 and then calculated Spearman correlations between the EMT scores and normalized expression (reads per million, RPM) for 5p and 3p mature strands for each of the miRNAs using MatrixEQTL and filtering by FDR < 0.05. An miRNA was considered to be epigenetically controlled if the BH-corrected P values were less than 0.01 for both (i) a Spearman correlation of miRNA abundance (RPM) to beta for probes in promoter regions associated with the miRNAs, and for (ii) a t -test of RPM between unmethylated ( β  < 0.1) and methylated ( β  > 0.3) samples (an epigenetically controlled pattern). We assessed potential miRNA targeting for all 178 samples and then separately for the 144 squamous samples by calculating miRNA–mRNA and miRNA–protein (RPPA) Spearman correlations with MatrixEQTL v2.1.1 using gene-level normalized abundance RNA-seq (RSEM) data and normalized RPPA data. Correlations were calculated with a P value threshold of 0.05, and then the anti-correlations were filtered at FDR < 0.05. We extracted miRNA–gene pairs that were functionally validated in publications reported by miRTarBase v4.5 (ref. 22 ). For miRNA–RPPA anti-correlations, all gene names that were associated with each antibody were used. Results were displayed with Cytoscape v2.8.3.

PARADIGM analysis

Integration of copy number, RNA-seq and pathway interaction data was performed on the core set of samples using PARADIGM 33 , 34 . Briefly, PARADIGM infers integrated pathway levels (IPLs) for genes, complexes and processes using pathway interactions, genomic and functional genomic data from each patient sample. One was added to all expression values, which were then log 2 -transformed and median-centred across samples for each gene. The log 2 -transformed, median-centred mRNA data were rank-transformed based on the global ranking across all samples and all genes and discretized (+1 for values with ranks in the highest tertile, −1 for values with ranks in the lowest tertile and 0 otherwise) before PARADIGM analysis.

Pathways were obtained in BioPax level 3 format, and included the NCIPID and BioCarta databases from http://pid.nci.nih.gov and the Reactome database from http://reactome.org . Gene identifiers were unified by UniProt ID and then converted to Human Genome Nomenclature Committee’s HUGO symbols using mappings provided by HGNC ( http://www.genenames.org/ ). Altogether, 1,524 pathways were obtained. Interactions from all of these sources were then combined into a merged superimposed pathway (SuperPathway). Genes, complexes and abstract processes (for example, cell cycle and apoptosis) were retained and henceforth referred to collectively as pathway features. The resulting pathway structure contained a total of 19,504 features, representing 7,369 protein-coding genes, 9,354 complexes, 2,092 families, 82 RNAs, 15 miRNAs and 592 abstract processes.

The PARADIGM algorithm infers an IPL for each pathway element that reflects the log likelihood that contrasts the probability of activity against inactivity. An initial minimum variation filter (at least 1 sample with absolute activity >0.05) was applied, resulting in 15,502 concepts (5,898 protein-coding genes, 7,307 complexes, 1,916 families, 12 RNAs, 15 miRNAs and 354 abstract processes) with relative activities showing distinguishable variation across tumours.

iCluster analysis

Integrative clustering of RNA-seq, methylation, copy number and miRNA data was performed using the R package iCluster 20 . The core set of samples was used since all samples in this set had data available across these four platforms. RNA-seq, methylation, copy number and mature-strand miRNA datasets had 20,531, 395,552, 23,109 and 1,213 features, respectively. The 500 most variable features based on the standard deviation from each dataset were selected for the integrative clustering analyses. For analysis involving the RNA-seq and miRNA datasets, a log[ x  + 1] transformation was used in order to deal with skewness in the data 67 . Methylation data was logit transformed to make it closer to normal distribution. The CNV data included the regions determined from GISTIC2.0, with CNVs treated as a continuous measurement based on the segmentation mean value for the region.

MEMo analysis

High DNA-methylation levels upstream of miR-200a and miR-200b corresponded to transcriptional downregulation of the miRNAs ( Extended Data Fig. 9a ). For a sample to be called altered for either miR-200a or miR-200b (or both), we required both high DNA-methylation level upstream of the miRNA ( β  > 0.3) and low miRNA expression (log 2 [RPM] < 9.3 for miR-200a and log 2 [RPM] < 9 for miR-200b). Binary calls were given to altered and unaltered samples based on this double threshold (1 = altered, 0 = unaltered).

The mutual exclusivity modules in cancer (MEMo) algorithm 27 was run on all core-set samples. MEMo was first run on 27 regions of recurrent copy number gain, 36 regions of copy number loss and 22 recurrently mutated genes. In order to include alterations for miR-200a and miR-200b in the MEMo analysis, a custom network was designed where each miRNA was connected to its known and validated targets (see above). Second, this network was merged with the comprehensive pathway network used by MEMo to search for modules of altered genes that include at least one of the miRNAs. Extracted modules were tested for mutual exclusivity using statistical framework of MEMo ( Supplementary Table 27 ). A Student’s t -test was performed for comparison of the EMT mRNA scores between groups.

Data availability

The primary and processed data used in analyses can be downloaded by registered users from https://gdc-portal.nci.nih.gov/ and the TCGA publication page ( https://tcga-data.nci.nih.gov/docs/publications/cesc_2016/ ).

Change history

16 march 2017.

The Creative Commons licence information was added to the HTML and PDF.

Ferlay, J. et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int. J. Cancer 136 , E359–E386 (2015)

Article   CAS   PubMed   Google Scholar  

Schiffman, M. et al. Human papillomavirus testing in the prevention of cervical cancer. J. Natl. Cancer Inst. 103 , 368–383 (2011)

Article   PubMed   PubMed Central   Google Scholar  

Uyar, D. & Rader, J. Genomics of cervical cancer and the role of human papillomavirus pathobiology. Clin. Chem. 60 , 144–146 (2014)

Moody, C. A. & Laimins, L. A. Human papillomavirus oncoproteins: pathways to transformation. Nat. Rev. Cancer 10 , 550–560 (2010)

Cancer Genome Atlas Research Network. Comprehensive genomic characterization of squamous cell lung cancers. Nature 489 , 519–525 (2012)

Lawrence, M. S. et al. Mutational heterogeneity in cancer and the search for new cancer-associated genes. Nature 499 , 214–218 (2013)

Article   CAS   ADS   PubMed   PubMed Central   Google Scholar  

Chung, T. K. H. et al. Genomic aberrations in cervical adenocarcinomas in Hong Kong Chinese women. Int. J. Cancer 137 , 776–783 (2015)

Ojesina, A. I. et al. Landscape of genomic alterations in cervical carcinomas. Nature 506 , 371–375 (2014)

Article   CAS   ADS   PubMed   Google Scholar  

Cancer Genome Atlas Network. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature 507 , 315–322 (2014)

Cancer Genome Atlas Network. Comprehensive genomic characterization of head and neck squamous cell carcinomas. Nature 517 , 576–582 (2015)

Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature 490 , 61–70 (2012)

Alexandrov, L. B. et al. Signatures of mutational processes in human cancer. Nature 500 , 415–421 (2013)

Article   CAS   PubMed   PubMed Central   Google Scholar  

Burns, M. B., Temiz, N. A. & Harris, R. S. Evidence for APOBEC3B mutagenesis in multiple human cancers. Nat. Genet. 45 , 977–983 (2013)

Henderson, S., Chakravarthy, A., Su, X., Boshoff, C. & Fenton, T. R. APOBEC-mediated cytosine deamination links PIK3CA helical domain mutations to human papillomavirus-driven tumor development. Cell Reports 7 , 1833–1841 (2014)

Roberts, S. A. et al. An APOBEC cytidine deaminase mutagenesis pattern is widespread in human cancers. Nat. Genet. 45 , 970–976 (2013)

Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature 474 , 609–615 (2011)

The Cancer Genome Atlas Research Network. Integrated genomic characterization of endometrial carcinoma. Nature 497 , 67–73 (2013)

Rooney, M. S., Shukla, S. A., Wu, C. J., Getz, G. & Hacohen, N. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell 160 , 48–61 (2015)

Godinho, M. F. E. et al. BCAR4 induces antioestrogen resistance but sensitises breast cancer to lapatinib. Br. J. Cancer 107 , 947–955 (2012)

Shen, R., Olshen, A. B. & Ladanyi, M. Integrative clustering of multiple genomic data types using a joint latent variable model with application to breast and lung cancer subtype analysis. Bioinformatics 25 , 2906–2912 (2009)

Byers, L. A. et al. An epithelial–mesenchymal transition gene signature predicts resistance to EGFR and PI3K inhibitors and identifies Axl as a therapeutic target for overcoming EGFR inhibitor resistance. Clin. Cancer Res. 19 , 279–290 (2013)

Hsu, S.-D. et al. miRTarBase update 2014: an information resource for experimentally validated miRNA-target interactions. Nucleic Acids Res. 42 , D78–D85 (2014)

Akbani, R. et al. A pan-cancer proteomic perspective on The Cancer Genome Atlas. Nat. Commun. 5 , 3887 (2014)

Seton-Rogers, S. Oncogenes: all eyes on YAP1. Nat. Rev. Cancer 14 , 514–515 (2014)

Shao, D. D. et al. KRAS and YAP1 converge to regulate EMT and tumor survival. Cell 158 , 171–184 (2014)

Vandewalle, C., Van Roy, F. & Berx, G. The role of the ZEB family of transcription factors in development and disease. Cell. Mol. Life Sci. 66 , 773–787 (2009)

Ciriello, G., Cerami, E., Sander, C. & Schultz, N. Mutual exclusivity analysis identifies oncogenic network modules. Genome Res. 22 , 398–406 (2012)

Gregory, P. A. et al. The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1. Nat. Cell Biol. 10 , 593–601 (2008)

Massagué, J. TGFβ signalling in context. Nat. Rev. Mol. Cell Biol. 13 , 616–630 (2012)

Haslehurst, A. M. et al. EMT transcription factors snail and slug directly contribute to cisplatin resistance in ovarian cancer. BMC Cancer 12 , 91 (2012)

Taube, J. H. et al. Core epithelial-to-mesenchymal transition interactome gene-expression signature is associated with claudin-low and metaplastic breast cancer subtypes. Proc. Natl Acad. Sci. USA 107 , 15449–15454 (2010)

Jaiswal, B. S. et al. Oncogenic ERBB3 mutations in human cancers. Cancer Cell 23 , 603–617 (2013)

Sedgewick, A. J., Benz, S. C., Rabizadeh, S., Soon-Shiong, P. & Vaske, C. J. Learning subgroup-specific regulatory interactions and regulator independence with PARADIGM. Bioinformatics 29 , i62–i70 (2013)

Vaske, C. J. et al. Inference of patient-specific pathway activities from multi-dimensional cancer genomics data using PARADIGM. Bioinformatics 26 , i237–i245 (2010)

Hoadley, K. A. et al. Multiplatform analysis of 12 cancer types reveals molecular classification within and across tissues of origin. Cell 158 , 929–944 (2014)

den Boon, J. A. et al. Molecular transitions from papillomavirus infection to cervical precancer and cancer: role of stromal estrogen receptor signaling. Proc. Natl Acad. Sci. USA 112 , E3255–E3264 (2015)

Roman, A. & Munger, K. The papillomavirus E7 proteins. Virology 445 , 138–168 (2013)

Vande Pol, S. B. & Klingelhutz, A. J. Papillomavirus E6 oncoproteins. Virology 445 , 115–137 (2013)

Jiao, Y., Widschwendter, M. & Teschendorff, A. E. A systems-level integrative framework for genome-wide DNA methylation and gene expression data identifies differential gene expression modules under epigenetic control. Bioinformatics 30 , 2360–2366 (2014)

Dellambra, E. et al. Downregulation of 14-3-3σ prevents clonal evolution and leads to immortalization of primary human keratinocytes. J. Cell Biol. 149 , 1117–1130 (2000)

Moreira, J. M. A., Gromov, P. & Celis, J. E. Expression of the tumor suppressor protein 14-3-3σ is down-regulated in invasive transitional cell carcinomas of the urinary bladder undergoing epithelial-to-mesenchymal transition. Mol. Cell. Proteomics 3 , 410–419 (2004)

Hermeking, H. et al. 14-3-3σ is a p53-regulated inhibitor of G2/M progression. Mol. Cell 1 , 3–11 (1997)

Chang, T.-C. et al. 14-3-3σ regulates β-catenin-mediated mouse embryonic stem cell proliferation by sequestering GSK-3β. PLoS One 7 , e40193 (2012)

Wentzensen, N., Vinokurova, S. & von Knebel Doeberitz, M. Systematic review of genomic integration sites of human papillomavirus genomes in epithelial dysplasia and invasive cancer of the female lower genital tract. Cancer Res. 64 , 3878–3884 (2004)

Tang, A. L. et al. UM-SCC-104: a new human papillomavirus-16-positive cancer stem cell-containing head and neck squamous cell carcinoma cell line. Head Neck 34 , 1480–1491 (2012)

Article   PubMed   Google Scholar  

Chu, J. et al. BioBloom tools: fast, accurate and memory-efficient host species sequence screening using bloom filters. Bioinformatics 30 , 3402–3404 (2014)

Kostic, A. D. et al. PathSeq: software to identify or discover microbes by deep sequencing of human tissue. Nat. Biotechnol. 29 , 393–396 (2011)

Trapnell, C., Pachter, L. & Salzberg, S. L. TopHat: discovering splice junctions with RNA-seq. Bioinformatics 25 , 1105–1111 (2009)

Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10 , R25 (2009)

Schiffman, M. et al. A population-based prospective study of carcinogenic human papillomavirus variant lineages, viral persistence, and cervical neoplasia. Cancer Res. 70 , 3159–3169 (2010)

McCarroll, S. A. et al. Integrated detection and population-genetic analysis of SNPs and copy number variation. Nat. Genet. 40 , 1166–1174 (2008)

Korn, J. M. et al. Integrated genotype calling and association analysis of SNPs, common copy number polymorphisms and rare CNVs. Nat. Genet. 40 , 1253–1260 (2008)

Olshen, A. B., Venkatraman, E. S., Lucito, R. & Wigler, M. Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics 5 , 557–572 (2004)

Article   PubMed   MATH   Google Scholar  

Mermel, C. H. et al. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. 12 , R41 (2011)

Carter, S. L. et al. Absolute quantification of somatic DNA alterations in human cancer. Nat. Biotechnol. 30 , 413–421 (2012)

Chen, K. et al. BreakDancer: an algorithm for high-resolution mapping of genomic structural variation. Nat. Methods 6 , 677–681 (2009)

Yang, L. et al. Diverse mechanisms of somatic structural variations in human cancer genomes. Cell 153 , 919–929 (2013)

Bibikova, M. et al. High density DNA methylation array with single CpG site resolution. Genomics 98 , 288–295 (2011)

Cancer Genome Atlas Research Network. Integrated genomic characterization of papillary thyroid carcinoma. Cell 159 , 676–690 (2014)

Wang, K. et al. MapSplice: accurate mapping of RNA-seq reads for splice junction discovery. Nucleic Acids Res. 38 , e178 (2010)

Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-seq data with or without a reference genome. BMC Bioinformatics 12 , 323 (2011)

Carstens, J. L., Lovisa, S. & Kalluri, R. Microenvironment-dependent cues trigger miRNA-regulated feedback loop to facilitate the EMT/MET switch. J. Clin. Invest. 124 , 1458–1460 (2014)

Ceppi, P. & Peter, M. E. MicroRNAs regulate both epithelial-to-mesenchymal transition and cancer stem cells. Oncogene 33 , 269–278 (2014)

Díaz-Martín, J. et al. A core microRNA signature associated with inducers of the epithelial-to-mesenchymal transition. J. Pathol. 232 , 319–329 (2014)

Kiesslich, T., Pichler, M. & Neureiter, D. Epigenetic control of epithelial–mesenchymal-transition in human cancer. Mol. Clin. Oncol. 1 , 3–11 (2013)

Tam, W. L. & Weinberg, R. A. The epigenetics of epithelial–mesenchymal plasticity in cancer. Nat. Med. 19 , 1438–1449 (2013)

Zwiener, I., Frisch, B. & Binder, H. Transforming RNA-seq data to improve the performance of prognostic gene signatures. PLoS One 9 , e85150 (2014)

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Download references

Acknowledgements

We would like to acknowledge the late H. Salvesen (the University of Bergen), who provided critical clinical and translational insight, and we dedicate this manuscript to her memory. We also acknowledge L. Gaffney (The Broad Institute) for her work in preparing some of the figures. In addition, this study was supported by National Institutes of Health (NIH) grants U54 HG003273, U54 HG003067, U54 HG003079, U24 CA143799, U24 CA143835, U24 CA143840, U24 CA143843, U24 CA143845, U24 CA143848, U24 CA143858, U24 CA143866, U24 CA143867, U24 CA143882, U24 CA143883, U24 CA144025 and P30 CA016672.

Author information

Helga B. Salvesen: Deceased.

Authors and Affiliations

Albert Einstein College of Medicine, Bronx, New York, 10461, New York, USA

Robert D. Burk & Zigui Chen

Analytical Biological Services, Inc., Wilmington, 19801, Delaware, USA

Charles Saller & Katherine Tarvin

Barretos Cancer Hospital, Barretos, Sao Paulo, Brazil

Andre L. Carvalho, Cristovam Scapulatempo-Neto, Henrique C. Silveira & José H. Fregnani

Baylor College of Medicine, Houston, 77030, Texas, USA

Chad J. Creighton, Matthew L. Anderson & Patricia Castro

Beckman Research Institute of City of Hope, Duarte, 91010, California, USA

• Sophia S. Wang

Buck Institute for Research on Aging, Novato, 94945, California, USA

Christina Yau & Christopher Benz

Canada’s Michael Smith Genome Sciences Centre, BC Cancer Agency, Vancouver, V5Z 4S6, British Columbia, Canada

A. Gordon Robertson, Karen Mungall, Lynette Lim, Reanne Bowlby, Sara Sadeghi, Denise Brooks, Payal Sipahimalani, Richard Mar, Adrian Ally, Amanda Clarke, Andrew J. Mungall, Angela Tam, Darlene Lee, Eric Chuah, Jacqueline E. Schein, Kane Tse, Katayoon Kasaian, Yussanne Ma, Marco A. Marra, Michael Mayo, Miruna Balasundaram, Nina Thiessen, Noreen Dhalla, Rebecca Carlsen, Richard A. Moore, Robert A. Holt, Steven J. M. Jones & Tina Wong

Harvard Medical School, Boston, 02115, Massachusetts, USA

Angeliki Pantazi, Michael Parfenov, Raju Kucherlapati, Angela Hadjipanayis, Jonathan Seidman, Melanie Kucherlapati, Xiaojia Ren, Andrew W. Xu, Lixing Yang, Peter J. Park & Semin Lee

Helen F. Graham Cancer Center and Research Institute at Christiana Care Health Services, Inc., Newark, 19713, Delaware, USA

Brenda Rabeno, Lori Huelsenbeck-Dill, Mark Borowsky, Mark Cadungog, Mary Iacocca, Nicholas Petrelli & Patricia Swanson

HudsonAlpha Institute for Biotechnology, Huntsville, 35806, Alabama, USA

Akinyemi I. Ojesina, Akinyemi I. Ojesina & Akinyemi I. Ojesina

The Eli and Edythe L. Broad Institute of Massachusetts Institute of Technology and Harvard University, Cambridge, 02142, Massachusetts, USA

Akinyemi I. Ojesina, Akinyemi I. Ojesina, Bradley A. Murray, Hailei Zhang, Andrew D. Cherniack, Carrie Sougnez, Chandra Sekhar Pedamallu, Lee Lichtenstein, Matthew Meyerson, Michael S. Noble, David I. Heiman, Doug Voet, Gad Getz, Gordon Saksena, Jaegil Kim, Juliann Shih, Juok Cho, Michael S. Lawrence, Nils Gehlenborg, Pei Lin, Rameen Beroukhim, Scott Frazer, Stacey B. Gabriel, Steven E. Schumacher & Akinyemi I. Ojesina

University of Alabama at Birmingham, Birmingham, 35294, Alabama, USA

ILSbio, LLC, Chestertown, 21620, Maryland, USA

Indiana University School of Medicine, Indianapolis, 46202, Indiana, USA

• George Sandusky

Institute of Human Virology, Nigeria, Abuja, Nigeria

Sally N. Adebamowo, Teniola Akeredolu & Clement Adebamowo

Institute for Systems Biology, Seattle, 98109, Washington, USA

Sheila M. Reynolds & Ilya Shmulevich

International Genomics Consortium, Phoenix, 85004, Arizona, USA

Candace Shelton, Daniel Crain, David Mallery, Erin Curley, Johanna Gardner, Robert Penny, Scott Morris & Troy Shelton

Leidos Biomedical, Rockville, 20850, Maryland, USA

Jia Liu, Laxmi Lolla, Sudha Chudamani & Ye Wu

Massachusetts General Hospital, Boston, 02114, Massachusetts, USA

• Michael Birrer

McDonnell Genome Institute at Washington University, St Louis, 63108, Missouri, USA

Michael D. McLellan, Matthew H. Bailey, Christopher A. Miller, Matthew A. Wyczalkowski, Robert S. Fulton, Catrina C. Fronick, Charles Lu, Elaine R. Mardis, Elizabeth L. Appelbaum, Heather K. Schmidt, Lucinda A. Fulton, Matthew G. Cordes, Tiandao Li, Li Ding & Richard K. Wilson

Medical College of Wisconsin, Milwaukee, 53226, Wisconsin, USA

Janet S. Rader, Behnaz Behmaram, Denise Uyar & William Bradley

Medical University of South Carolina, Charleston, 29425, South Carolina, USA

• John Wrangle

Memorial Sloan Kettering Cancer Center, New York, 10065, New York, USA

Alessandro Pastore, Douglas A. Levine, Fanny Dao, Jianjiong Gao, Nikolaus Schultz, Chris Sander & Marc Ladanyi

Montefiore Medical Center, Bronx, New York, 10461, New York, USA

Mark Einstein & Randall Teeter

NantOmics, Santa Cruz, 95060, California, USA

• Stephen Benz

National Cancer Institute, Bethesda, 20892, Maryland, USA

Nicolas Wentzensen, Ina Felau, Jean C. Zenklusen, Clara Bodelon, John A. Demchok, Liming Yang, Margi Sheth, Martin L. Ferguson, Roy Tarnuzzer, Hannah Yang, Mark Schiffman, Jiashan Zhang, Zhining Wang & Tanja Davidsen

National Hospital, Abuja, Nigeria

• Olayinka Olaniyan

National Human Genome Research Institute, Bethesda, 20892, Maryland, USA

Carolyn M. Hutter & Heidi J. Sofia

National Institute of Environmental Health Sciences, Durham, 27709, North Carolina, USA

Dmitry A. Gordenin, Kin Chan, Steven A. Roberts & Leszek J. Klimczak

National Institute on Deafness and Other Communication Disorders, Bethesda, 20892, Maryland, USA

Carter Van Waes, Zhong Chen, Anthony D. Saleh & Hui Cheng

Ontario Tumour Bank, London Health Sciences Centre, London, N6A 5A5, Ontario, Canada

• Jeremy Parfitt

Ontario Tumour Bank, Ontario Institute for Cancer Research, Toronto, M5G 0A3, Ontario, Canada

John Bartlett & Monique Albert

Ontario Tumour Bank, The Ottawa Hospital, Ottawa, K1H 8L6, Ontario, Canada

Angel Arnaout, Harman Sekhon & Sebastien Gilbert

Oregon Health and Science University, Portland, 97201, Oregon, USA

Penrose-St Francis Health Services, Colorado Springs, 80906, Colorado, USA

Jerome Myers, Jodi Harr, John Eckman, Julie Bergsten, Kelinda Tucker & Leigh Anne Zach

Samuel Oschin Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, 90048, California, USA

Beth Y. Karlan, Jenny Lester & Sandra Orsulic

SRA International, Fairfax, 22033, Virginia, USA

Qiang Sun, Rashi Naresh, Todd Pihl & Yunhu Wan

St Joseph’s Candler Health System, Savannah, 31406, Georgia, USA

Howard Zaren, Jennifer Sapp, Judy Miller & Paul Drwiega

The Research Institute at Nationwide Children’s Hospital, Columbus, 43205, Ohio, USA

Kristen M. Leraas, Tara M. Lichtenberg, Erik Zmuda, Jay Bowen, Jessica Frick, Julie M. Gastier-Foster, Lisa Wise, Mark Gerken & Nilsa C. Ramirez

The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University, Baltimore, 21287, Maryland, USA

Ludmila Danilova, Leslie Cope & Stephen B. Baylin

The University of Bergen, Bergen, Norway

• Helga B. Salvesen

The University of Texas MD Anderson Cancer Center, Houston, 77030, Texas, USA

Christopher P. Vellano, Zhenlin Ju, Lixia Diao, Hao Zhao, Zechen Chong, Michael C. Ryan, Emmanuel Martinez-Ledesma, Roeland G. Verhaak, Lauren Averett Byers, Yuan Yuan, Ken Chen, Shiyun Ling, Gordon B. Mills, Yiling Lu, Rehan Akbani, Sahil Seth, Han Liang, Jing Wang, Leng Han, John N. Weinstein, Christopher A. Bristow, Wei Zhang, Harshad S. Mahadeshwar, Huandong Sun, Jiabin Tang, Jianhua Zhang, Xingzhi Song, Alexei Protopopov, Kenna R. Mills Shaw & Lynda Chin

University of Abuja Teaching Hospital, Gwagwalada, Abuja, Nigeria

• Oluwole Olabode

University of California, Irvine, Orange, 92668, California, USA

• Philip DiSaia

University of California Santa Cruz, Santa Cruz, 95064, California, USA

Amie Radenbaugh, David Haussler, Jingchun Zhu & Josh Stuart

University of Kansas Medical Center, Kansas City, 66160, Kansas, USA

Prabhakar Chalise, Devin Koestler, Brooke L. Fridley, Andrew K. Godwin & Rashna Madan

University of Lausanne, Lausanne, Switzerland

• Giovanni Ciriello

University of New Mexico Health Sciences Center, Albuquerque, 87131, New Mexico, USA

Cathleen Martinez, Kelly Higgins & Therese Bocklage

University of North Carolina at Chapel Hill, Chapel Hill, 27599, North Carolina, USA

J. Todd Auman, Charles M. Perou, Donghui Tan, Joel S. Parker, Katherine A. Hoadley, Matthew D. Wilkerson, Piotr A. Mieczkowski, Tara Skelly, Umadevi Veluvolu, D. Neil Hayes, W. Kimryn Rathmell, Alan P. Hoyle, Janae V. Simons, Junyuan Wu, Lisle E. Mose, Matthew G. Soloway, Saianand Balu, Shaowu Meng, Stuart R. Jefferys, Tom Bodenheimer, Yan Shi, Jeffrey Roach, Leigh B. Thorne, Lori Boice, Mei Huang & Corbin D. Jones

University of Oklahoma Health Sciences Center, Oklahoma City, 73104, Oklahoma, USA

Rosemary Zuna, Joan Walker, Camille Gunderson, Carie Snowbarger, David Brown, Katherine Moxley, Kathleen Moore, Kelsi Andrade, Lisa Landrum, Robert Mannel, Scott McMeekin, Starla Johnson & Tina Nelson

University of Pittsburgh, Pittsburgh, 15213, Pennsylvania, USA

Esther Elishaev, Rajiv Dhir, Robert Edwards & Rohit Bhargava

University of São Paulo, Ribeirão Preto Medical School, Ribeirão Preto, 14049-900, São Paulo, Brazil

Daniel G. Tiezzi, Jurandyr M. Andrade, Houtan Noushmehr, Carlos Gilberto Carlotti Jr & Daniela Pretti da Cunha Tirapelli

University of Southern California, Los Angeles, 90033, California, USA

Daniel J. Weisenberger, David J. Van Den Berg, Dennis T. Maglinte, Moiz S. Bootwalla, Phillip H. Lai & Timothy Triche Jr

University of Washington, Seattle, 981095, Washington, USA

Elizabeth M. Swisher & Kathy J. Agnew

University of Wisconsin School of Medicine and Public Health, Madison, 53705, Wisconsin, USA

• Carl Simon Shelley

Van Andel Research Institute, Grand Rapids, 49503, Michigan, USA

• Peter W. Laird

Washington University in St Louis, St Louis, 63110, Missouri, USA

Julie Schwarz, Perry Grigsby & David Mutch

Albert Einstein College of Medicine

• Robert D. Burk
•  & Zigui Chen

Analytical Biological Services

• Charles Saller
•  & Katherine Tarvin

Barretos Cancer Hospital

• Andre L. Carvalho
• , Cristovam Scapulatempo-Neto
• , Henrique C. Silveira
•  & José H. Fregnani

Baylor College of Medicine

• Chad J. Creighton
• , Matthew L. Anderson
•  & Patricia Castro

Beckman Research Institute of City of Hope

Buck institute for research on aging.

• Christina Yau
•  & Christopher Benz

Canada's Michael Smith Genome Sciences Centre

• A. Gordon Robertson
• , Karen Mungall
• , Lynette Lim
• , Reanne Bowlby
• , Sara Sadeghi
• , Denise Brooks
• , Payal Sipahimalani
• , Richard Mar
• , Adrian Ally
• , Amanda Clarke
• , Andrew J. Mungall
• , Angela Tam
• , Darlene Lee
• , Eric Chuah
• , Jacqueline E. Schein
• , Katayoon Kasaian
• , Yussanne Ma
• , Marco A. Marra
• , Michael Mayo
• , Miruna Balasundaram
• , Nina Thiessen
• , Noreen Dhalla
• , Rebecca Carlsen
• , Richard A. Moore
• , Robert A. Holt
• , Steven J. M. Jones
•  & Tina Wong

Harvard Medical School

• Angeliki Pantazi
• , Michael Parfenov
• , Raju Kucherlapati
• , Angela Hadjipanayis
• , Jonathan Seidman
• , Melanie Kucherlapati
• , Xiaojia Ren
• , Andrew W. Xu
• , Lixing Yang
• , Peter J. Park
•  & Semin Lee

Helen F. Graham Cancer Center & Research Institute at Christiana Care Health Services

• Brenda Rabeno
• , Lori Huelsenbeck-Dill
• , Mark Borowsky
• , Mark Cadungog
• , Mary Iacocca
• , Nicholas Petrelli
•  & Patricia Swanson

HudsonAlpha Institute for Biotechnology

• Akinyemi I. Ojesina

ILSbio, LLC

Indiana university school of medicine, institute of human virology.

• Sally N. Adebamowo
• , Teniola Akeredolu
•  & Clement Adebamowo

Institute for Systems Biology

• Sheila M. Reynolds
•  & Ilya Shmulevich

International Genomics Consortium

• Candace Shelton
• , Daniel Crain
• , David Mallery
• , Erin Curley
• , Johanna Gardner
• , Robert Penny
• , Scott Morris
•  & Troy Shelton

Leidos Biomedical

• , Laxmi Lolla
• , Sudha Chudamani
•  & Ye Wu

Massachusetts General Hospital

Mcdonnell genome institute at washington university.

• Michael D. McLellan
• , Matthew H. Bailey
• , Christopher A. Miller
• , Matthew A. Wyczalkowski
• , Robert S. Fulton
• , Catrina C. Fronick
• , Charles Lu
• , Elaine R. Mardis
• , Elizabeth L. Appelbaum
• , Heather K. Schmidt
• , Lucinda A. Fulton
• , Matthew G. Cordes
• , Tiandao Li
•  & Richard K. Wilson

Medical College of Wisconsin

• Janet S. Rader
• , Behnaz Behmaram
• , Denise Uyar
•  & William Bradley

Medical University of South Carolina

Memorial sloan kettering cancer center.

• Alessandro Pastore
• , Douglas A. Levine
• , Fanny Dao
• , Jianjiong Gao
• , Nikolaus Schultz
• , Chris Sander
•  & Marc Ladanyi

Montefiore Medical Center

• Mark Einstein
•  & Randall Teeter

National Cancer Institute

• Nicolas Wentzensen
• , Ina Felau
• , Jean C. Zenklusen
• , Clara Bodelon
• , John A. Demchok
• , Liming Yang
• , Margi Sheth
• , Martin L. Ferguson
• , Roy Tarnuzzer
• , Hannah Yang
• , Mark Schiffman
• , Jiashan Zhang
• , Zhining Wang
•  & Tanja Davidsen

National Human Genome Research Institute

• Carolyn M. Hutter
•  & Heidi J. Sofia

National Institute of Environmental Health Sciences

• Dmitry A. Gordenin
• , Steven A. Roberts
•  & Leszek J. Klimczak

National Institute on Deafness & Other Communication Disorders

• Carter Van Waes
• , Zhong Chen
• , Anthony D. Saleh
•  & Hui Cheng

Ontario Tumour Bank, London Health Sciences Centre

Ontario tumour bank, ontario institute for cancer research.

• John Bartlett
•  & Monique Albert

Ontario Tumour Bank, The Ottawa Hospital

• Angel Arnaout
• , Harman Sekhon
•  & Sebastien Gilbert

Oregon Health & Science University

• , Jerome Myers
• , Jodi Harr
• , John Eckman
• , Julie Bergsten
• , Kelinda Tucker
•  & Leigh Anne Zach

Samuel Oschin Comprehensive Cancer Institute, Cedars-Sinai Medical Center

• Beth Y. Karlan
• , Jenny Lester
•  & Sandra Orsulic

SRA International

• , Rashi Naresh
• , Todd Pihl
•  & Yunhu Wan

St Joseph's Candler Health System

• Howard Zaren
• , Jennifer Sapp
• , Judy Miller
•  & Paul Drwiega

The Eli & Edythe L. Broad Institute of Massachusetts Institute of Technology & Harvard University

• , Bradley A. Murray
• , Hailei Zhang
• , Andrew D. Cherniack
• , Carrie Sougnez
• , Chandra Sekhar Pedamallu
• , Lee Lichtenstein
• , Matthew Meyerson
• , Michael S. Noble
• , David I. Heiman
• , Doug Voet
• , Gordon Saksena
• , Jaegil Kim
• , Juliann Shih
• , Michael S. Lawrence
• , Nils Gehlenborg
• , Rameen Beroukhim
• , Scott Frazer
• , Stacey B. Gabriel
•  & Steven E. Schumacher

The Research Institute at Nationwide Children's Hospital

• Kristen M. Leraas
• , Tara M. Lichtenberg
• , Erik Zmuda
• , Jay Bowen
• , Jessica Frick
• , Julie M. Gastier-Foster
• , Lisa Wise
• , Mark Gerken
•  & Nilsa C. Ramirez

The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University

• Ludmila Danilova
• , Leslie Cope
•  & Stephen B. Baylin

The University of Bergen

The university of texas md anderson cancer center.

• Christopher P. Vellano
• , Zhenlin Ju
• , Lixia Diao
• , Zechen Chong
• , Michael C. Ryan
• , Emmanuel Martinez-Ledesma
• , Roeland G. Verhaak
• , Lauren Averett Byers
• , Yuan Yuan
• , Shiyun Ling
• , Gordon B. Mills
• , Yiling Lu
• , Rehan Akbani
• , Sahil Seth
• , Han Liang
• , Jing Wang
• , John N. Weinstein
• , Christopher A. Bristow
• , Wei Zhang
• , Harshad S. Mahadeshwar
• , Huandong Sun
• , Jiabin Tang
• , Jianhua Zhang
• , Xingzhi Song
• , Alexei Protopopov
• , Kenna R. Mills Shaw
•  & Lynda Chin

University of Abuja Teaching Hospital

University of alabama at birmingham, university of california, irvine, university of california santa cruz.

• Amie Radenbaugh
• , David Haussler
• , Jingchun Zhu
•  & Josh Stuart

University of Kansas Medical Center

• Prabhakar Chalise
• , Devin Koestler
• , Brooke L. Fridley
• , Andrew K. Godwin
•  & Rashna Madan

University of Lausanne

University of new mexico health sciences center.

• Cathleen Martinez
• , Kelly Higgins
•  & Therese Bocklage

University of North Carolina at Chapel Hill

• J. Todd Auman
• , Charles M. Perou
• , Donghui Tan
• , Joel S. Parker
• , Katherine A. Hoadley
• , Matthew D. Wilkerson
• , Piotr A. Mieczkowski
• , Tara Skelly
• , Umadevi Veluvolu
• , D. Neil Hayes
• , W. Kimryn Rathmell
• , Alan P. Hoyle
• , Janae V. Simons
• , Junyuan Wu
• , Lisle E. Mose
• , Matthew G. Soloway
• , Saianand Balu
• , Shaowu Meng
• , Stuart R. Jefferys
• , Tom Bodenheimer
• , Jeffrey Roach
• , Leigh B. Thorne
• , Lori Boice
• , Mei Huang
•  & Corbin D. Jones

University of Oklahoma Health Sciences Center

• Rosemary Zuna
• , Joan Walker
• , Camille Gunderson
• , Carie Snowbarger
• , David Brown
• , Katherine Moxley
• , Kathleen Moore
• , Kelsi Andrade
• , Lisa Landrum
• , Robert Mannel
• , Scott McMeekin
• , Starla Johnson
•  & Tina Nelson

University of Pittsburgh

• Esther Elishaev
• , Rajiv Dhir
• , Robert Edwards
•  & Rohit Bhargava

University of São Paulo, Ribeir ão Preto Medical School

• Daniel G. Tiezzi
• , Jurandyr M. Andrade
• , Houtan Noushmehr
• , Carlos Gilberto Carlotti Jr
•  & Daniela Pretti da Cunha Tirapelli

University of Southern California

• Daniel J. Weisenberger
• , David J. Van Den Berg
• , Dennis T. Maglinte
• , Moiz S. Bootwalla
• , Phillip H. Lai
•  & Timothy Triche Jr

University of Washington

• Elizabeth M. Swisher
•  & Kathy J. Agnew

University of Wisconsin School of Medicine & Public Health

Van andel research institute, washington university in st louis.

• Julie Schwarz
• , Perry Grigsby
•  & David Mutch

Contributions

The Cancer Genome Atlas research network contributed collectively to this work. Biospecimens were collected at the tissue source sites (TSSs) and processed by the biospecimen core resource (BCR). Data was generated by the genome sequencing and genome data analysis centres, with analyses performed by members across the network. Data were stored and released through the data coordinating centre (DCC). The NCI project coordinator was I. Felau and the overall analysis coordinator and data coordinator was C. P. Vellano. Special thanks also go out to TCGA network members who made substantial contributions to this work: C. P. Vellano (analysis coordinator, data coordinator, co-manuscript coordinator, RPPA analysis), N. Wentzensen (co-manuscript coordinator, HPV-analysis subgroup co-leader), A. I. Ojesina (co-manuscript coordinator, HPV-analysis subgroup co-leader, somatic-alteration analysis), A. G. Robertson (miRNA analysis, HPV analysis), M. D. McLellan (mutation calling), L. Danilova (methylation analysis), B. A. Murray (copy number and ABSOLUTE analysis), Z. Ju (RPPA analysis), J. T. Auman (mRNA-sequencing analysis, fusion analysis), P. Chalise (iCluster analysis), C. Yau (PARADIGM pathway analysis), G. Ciriello (MEMo pathway analysis), D. A. Gordenin (APOBEC analysis), R. Zuna (pathologist), H. Zhang (mutation analysis, Firehose), A. Pantazi (structural-variant-analysis subgroup leader, low-pass sequencing), M. H. Bailey (mutation analysis), L. Diao (EMT analysis), D. Koestler (methylation data processing, functional epigenetic module analysis), K. Mungall (HPV analysis), L. Lim (HPV analysis), R. Bowlby (miRNA analysis), S. Sadeghi (HPV analysis), D. Brooks (miRNA analysis), C. Sekhar Pedamallu (HPV analysis), K. Chen (fusion analysis), H. Zhao (fusion analysis), Z. Chong (fusion analysis), E. Martinez-Ledesma (fusion analysis), R. G. Verhaak (fusion analysis), K. M. Leraas (BCR), T. M. Lichtenberg (BCR), D. G. Tiezzi (immune-response gene analysis), M. C. Ryan (splicing analysis), S. M. Reynolds (regulome explorer analysis), G. B. Mills (project co-chair) and J. S. Rader (project co-chair).

Corresponding authors

Correspondence to Akinyemi I. Ojesina , Janet S. Rader , Nicolas Wentzensen or Christopher P. Vellano .

Ethics declarations

Competing interests.

The author declare no competing financial interests.

Additional information

The Cancer Genome Atlas Research Network (participants are arranged by institution)

Extended data figures and tables

Extended data figure 1 sample sets and histological patterns of cervical cancer..

a , Summary of sample numbers and degree of overlap between the core, extended and RPPA datasets. b , Example of a large-cell non-keratinizing squamous cell carcinoma. Tongues of highly atypical polygonal neoplastic squamous cells infiltrate through a fibrotic stroma. The cells show abundant eosinophilic cytoplasm with pleomorphic nuclei and prominent mitotic figures. Although the tumour cells contain abundant cytokeratin filaments, this tumour has traditionally been termed non-keratinizing because of the absence of characteristic keratin pearls. c , An example of a large-cell keratinizing squamous cell carcinoma. Nests of atypical squamous cells infiltrate through a fibrotic stroma. In addition, this tumour shows highly eosinophilic keratin pearls with small, inky dark nuclei that imperfectly mimic the normal keratinization that is found in the epidermis. This differentiation pattern is aberrant in the cervix in which the squamous epithelium is normally a non-keratinizing squamous mucosa. d , An example of an endocervical adenocarcinoma (well differentiated). Closely set, atypical glands with enlarged nuclei and scattered mitotic figures infiltrate through the connective tissue of the cervix. The tall columnar tumour cells show basally placed, crowded, enlarged nuclei that show frequent mitotic figures. Compared with normal endocervical cells, the tumour cells show relative loss of intra-cytoplasmic mucin and are frequently called mucin-depleted, although most, but not all endocervical adenocarcinomas show varying amounts of intracytoplasmic mucin at least focally. e , Adenosquamous carcinoma of cervix. This tumour shows both nests of non-keratinizing squamous cell carcinoma and glands composed of tall columnar adenocarcinoma reflecting the origin of most cervical cancers in the transformation zone of the cervix in which both squamous and glandular cells normally differentiate. Despite this biphasic differentiation potential, adenosquamous carcinomas are relatively uncommon in the cervix. f , UCEC-like HPV-negative endocervical adenocarcinoma from a radical hysterectomy specimen. The endometrium in the uterus was benign. g , UCEC-like HPV-positive endocervical adenocarcinoma from a radical hysterectomy specimen. The endometrium in the uterus was benign. All samples were stained with haematoxylin and eosin (20×). Scale bar, 100 μm.

Extended Data Figure 2 SMGs and the role of APOBEC in cervical cancer mutagenesis.

a – f , High-confidence somatic mutations in SMGs among 192 exome-sequenced samples in the extended case set are shown. Domains are labelled according to Gencode 19, corresponding to Ensembl 74. Mutations at canonical intronic splice acceptor (e−1 and e−2) are labelled based on proximity to the nearest coding exon. Panels display somatic mutations detected in novel cervical cancer SMGs, with HLA-B included for comparison with its family member HLA-A . Each axis is the protein-coding portion of a gene and each highlighted section represents the UniProt functional domain. Vertical lines indicate the boundaries of multiple annotation sources within common domain annotations as outlined in Supplementary Table 5 . Horizontal lines distinguish overlapping domains. Circles represent a single mutation and are coloured based on mutation type. Mutations present in squamous cell carcinomas are black, whereas those present in adenocarcinomas are pink. g , PIK3CA mutations and recurrence are shown in a stacked circle plot, as above. Additionally, lolliplot sticks are coloured red if the mutation type coincides with patterns of APOBEC mutagenesis. h , The minimal estimated number of APOBEC-induced mutations (APOBEC_MutLoad_MinEstimate column in Supplementary Table 1 ) strongly correlates with total number of mutations in a sample, as well as with the number of single-nucleotide variants (SNVs) in G:C pairs that are the exclusive substrate for mutagenesis by APOBEC cytidine deaminases. Although correlation with mutagenesis in A:T base pairs, which cannot be mutated by APOBEC enzymes, is statistically significant (two-tailed P  = 0.047), it is very weak. Pearson correlation and R 2 were calculated for all 192 exome-sequenced samples, including samples with zero values. Only samples with non-zero values of APOBEC_MutLoad_MinEstimate are presented.

Extended Data Figure 3 Copy number alterations in cervical cancer.

a , A log 2 -centred heatmap of somatic copy number alterations across 178 core-set cervical tumours. The x axis includes samples that have been ordered based on the cluster assignment. The y axis is based on genomic position, from 1p to Xq. Features associated with copy number clusters are annotated with asterisks; * P  < 0.05; ** P  < 0.01. b , GISTIC2.0 amplification and deletion plots within copy number clusters. Chromosomal locations for peaks of significantly recurrent focal amplifications (red) and deletions (blue) are plotted by −log 10 q value for the high (CN High) and low (CN Low) copy number clusters. Peaks are annotated with cytoband and candidate driver genes. The total number of genes in the peak region is indicated in parentheses. Peaks with more than 30 genes in the peak region are excluded. Any genes annotated have a significant positive correlation with mRNA expression. c , Chromosomal locations for peaks of significantly recurrent focal amplifications (red) and deletions (blue) are plotted by −log 10 q value for all core set samples. Peaks are annotated with cytoband and candidate driver genes. The total number of genes in the peak region is indicated in parentheses. Peaks consisting of more than 30 genes in the peak region are excluded. Annotated genes have a significant positive correlation with mRNA expression. d , Cytolytic activity (CYT) associations with PD-L1 and/or PD-L2 amplification. Each bar represents a single tumour and the height of that bar represents the z score of the cytolytic activity of that tumour compared to the rest of the cohort. Bars are coloured according to their PD-L1 and/or PD-L2 amplification status and sorted from the highest to the lowest z score.

Extended Data Figure 4 Gene-expression patterns and fusion genes found in cervical cancer.

a , Hierarchical clustering (uncentred correlation with centroid linkage as the clustering method) was performed on 4,039 expressed and highly variable genes across samples from 178 cervical, 170 endometrial and 279 head and neck cancer patients. Normalized gene-level RSEM values were median-centred before clustering and relative increased expression values are indicated in red and relative decreased expression values are indicated in blue. Samples from patients with cervical (CESC, light blue), endometrial (UCEC, purple) and head and neck (HNSC, orange) cancer are categorized by different colours as indicated. Also included are indications of HPV status, histology of cervical and endometrial cancers, and tissue site for head and neck cancer samples. Select genes are noted to the right of their locations on the heatmap. b , Box plots of the three differentially expressed SMGs and top six significantly differentially expressed non-SMGs across the iCluster groups using Kruskal–Wallis test. All genes are significantly different between the keratin-low and keratin-high clusters. Significant P values across keratin-low and keratin-high clusters are presented. c , A schematic of BCAR4 tandem duplication in one case (C5-A3HF), detected by analysis of somatic copy number (top) and structural variation (middle). Split reads and genomic breakpoints indicating the tandem duplication are shown. At the RNA level (bottom) the last exon of BCAR4 forms a fusion gene with the first exon of ZC3H7A (red bars indicate the location of mRNA breakpoints; NR_024049 shown as BCAR4 representative transcript). d , Schematic of recurrent fusions ( CPSF6 – C9orf3 , ARL8B–ITPR1 and MYH9–TXN2 ) or fusions with known occurrences in other cancer types ( FGFR3–TACC3 ), detected by at least two RNA-seq fusion callers in 178 samples. Red bars indicate the mRNA breakpoints.

Extended Data Figure 5 Unsupervised clusters of DNA methylation data.

a , Heatmap showing β values of 178 core-set samples ordered by CIMP clusters. Samples are presented in columns and the CpG island promoter CpG loci are presented in rows. An annotation panel on the right of the heatmap indicates CpG loci that are differentially methylated within a particular feature (see Supplementary Table 13 ). All features (marked with an asterisk) are significantly associated with DNA-methylation clusters (Fisher’s Exact test P  < 0.01), except APOBEC mutagenesis level, UCEC-like status and HPV integration status. b , Box plots of the EMT mRNA score and tumour purity by CIMP clusters. * P  < 0.05; ** P  < 0.01 (Student’s t -test).

Extended Data Figure 6 miRNA clusters and miRNA-gene/protein anti-correlations in cervical cancer.

a , Unsupervised clustering for miRNA profiles across 178 core-set tumour samples. Top to bottom: a normalized abundance heatmap for the fifty 5p or 3p strands that were highly ranked as differentially abundant by a SAMseq multiclass analysis; a silhouette width profile calculated from the consensus membership matrix; covariates with associated P values; and a summary table of the number of samples in each cluster. The scale bar shows row-scaled log 10 [RPM + 1] normalized abundances. Blue triangles mark selected cancer-associated miRs that were both differentially abundant across the subtypes and abundant in at least one subtype. b , Subnetworks of potential targeting relationships for a subset of miRNAs, as significance-thresholded (FDR < 0.05) miRNA–mRNA and miRNA–protein anti-correlations that are supported by functionally validated publications. For genes (nodes), colour distinguishes those that are only present in mRNA data (grey) from those that are present in both mRNA and RPPA data (green). Edges represent anti-correlations, and colour distinguishes anti-correlations between miRNA and mRNA (blue); and miRNA and unphosphorylated protein (green). In the n  = 178 core-set cohort, no correlations satisfying FDR < 0.05 were reported between miRNA and phosphorylated protein.

Extended Data Figure 7 EMT-associated miRNAs and their relationship to miRNA clusters and TGFβR2 somatic alterations.

a , Normalized miR-200a-3p abundance (RPM) across RPPA clusters for all 112 (top) and 92 squamous (bottom) samples of the core set for which RPPA data are available. P values presented are from two-sided Kolmogorov–Smirnov tests for RPPA-based EMT cluster versus non-EMT cluster samples. For n  = 112 samples, median miR-200a-3p RPM = 296.4 within the EMT cluster ( n  = 29) and 410.0 ( n  = 83) in non-EMT cluster samples. For squamous samples, median miR-200a-3p RPM = 296.4 ( n  = 29) within the EMT cluster and 393.4 ( n  = 63) in non-EMT cluster samples. EK-A2R7, which is in the hormone RPPA cluster, has an RPM value of 4,267 and is not shown. Results are not presented for adenocarcinoma samples separately owing to limited sample numbers ( n  = 18 from the core set with RPPA data available). b , Negative and positive Spearman correlation coefficients (FDR < 0.05) between EMT mRNA score and normalized abundance (RPM) for miRNA mature strands ( n  = 178). miRNAs that have been reported as associated with EMT (see Methods) are highlighted by blue bars. c , Normalized abundance heatmap of miRNAs most strongly negatively and positively correlated with EMT mRNA scores, with samples grouped by miRNA cluster and sorted by EMT score within each cluster. Somatic mutations (MUT) and deletions (HOMDEL) are shown for TGFBR2 , CREBBP , EP300 and SMAD4 . Methylation and concomitant downregulated expression alterations (ALT) as defined in Methods for miR-200a/b are also shown. miRNAs in blue represent those highlighted by blue bars in b . d , e , Same as b , c , for the n  = 144 squamous tumour samples.

Extended Data Figure 8 Distinguishing features of cervical cancer integrated molecular subtypes.

a , Integrative clustering of 178 core-set samples from patients with cervical cancer using mRNA, methylation, miRNA and copy number data identified three iCluster groups: keratin-low, keratin-high and adenocarcinoma-rich (adenocarcinoma). Relative frequencies of various cervical cancer classifications defined by histology, HPV clade, CNVs, methylation, miRNA and RPPA are plotted. The key for each feature is shown at the bottom. For each category, the statistically significantly enriched features in each cluster are highlighted with asterisks and include the name of the enriched feature. Each category was significantly associated with the clusters ( χ 2 test; P  < 0.05). The width of each plot is scaled according to the number of samples within each cluster. b , The frequencies of somatic alterations and additional novel features that distinguish the clusters, specifically those that do not occur in all three clusters, are plotted. The ‘somatic mutations’ panel shows the presence or absence of mutations for 7 of the identified SMGs. The ‘copy number alterations’ panel shows select copy number alterations (high-level amplifications and focal deletions) that are differentially present across the iCluster groups. The ‘additional features’ panel highlights miscellaneous features that also distinguish the clusters, including the presence of miR-200a/b alterations, UCEC-like samples and BCAR4 -fusion events. The key for each feature is shown on the right.

Extended Data Figure 9 miR-200a/b associations with EMT-regulating genes and somatic alterations within RTK, PI3K, MAPK and TGFβR2 pathways in cervical cancer.

a , Expression levels for miR-200a and miR-200b compared to DNA-methylation level at their promoter. Samples were called altered if the miRNAs were concurrently hypermethylated ( β  > 0.3) and downregulated (red). b , mRNA expression levels for ZEB2 , a target of both miR-200a and miR-200b, in subsets of miR-200a/b altered samples. ZEB2 is upregulated in samples with concurrent hypermethylation and downregulation of the miRNAs. c , mRNA expression levels of both ZEB1 and ZEB2 in miR-200a/b hypermethylated/downregulated (altered) and all other (WT) samples. d , Correlations of miR-200a and miR-200b expression with multiple genes involved in EMT signalling across squamous cell carcinomas and adenocarcinomas. e , Extent of genetic alterations and miRNA downregulation in the RTK, PI3K, MAPK and TGFβ pathways across all cervical tumours.

Extended Data Figure 10 Pathway biomarkers differentiating squamous cell carcinomas and adenocarcinomas.

a , Cytoscape display of the largest interconnected regulatory network of PARADIGM pathway features that are differentially activated between squamous cell carcinomas and adenocarcinomas connected through hubs with ≥10 downstream targets. Hubs with ≥10 downstream targets are labelled. Genes showing mRNA–miRNA expression anti-correlation with strong supporting evidence are highlighted with a thicker black outline and are labelled. Top differentially expressed genes relating to immune function are also labelled. Node size is proportional to significance of differential activation. b , Zoom-in display of the p63 sub-network neighbourhood. First neighbours (upstream or downstream) of four p63 complexes (bold text) are displayed in this view.

Extended Data Figure 11 HPV integration and molecular characteristics in cervical cancer.

a , E6 unspliced/spliced ratio for HPV16 and HPV18 intragenic, enhancer and intergenic sites. HPV16, median = 0.44 ( n  = 102); HPV18, median = 0.93 ( n  = 40). The P value is from a two-sided Kolmogorov–Smirnov test. b , Distribution of RNA-seq-based EMT score for HPV-negative (HPV−) and HPV-positive (HPV+) samples ( n  = 178). The P value was calculated as in a . c , Distributions of somatic copy number alterations and mRNA abundance ranks (left) and distribution functions for somatic copy number alterations and mRNA abundance ranks with 500 random samples shown close to the diagonals (grey) (right) for genomic loci with integrated HPV16. d , Distributions as in c for genomic loci with integrated HPV18. Benjamini–Hochberg-corrected P values for the somatic copy number alteration and mRNA abundance ranks are medians of the P values from Kolmogorov–Smirnov tests for all random samples.

Supplementary information

Supplementary.

This file contains Supplementary Methods, Results and associated references for each core and integrated analysis. It also contains Supplementary Figures 1-48. (PDF 26248 kb)

Supplementary Tables

This zipped file contains Supplementary Tables 1-31 and a Supplementary Table guide. (ZIP 4107 kb)

PowerPoint slides

Powerpoint slide for fig. 1, powerpoint slide for fig. 2, powerpoint slide for fig. 3, powerpoint slide for fig. 4, powerpoint slide for fig. 5, rights and permissions.

This work is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) licence. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons licence, users will need to obtain permission from the licence holder to reproduce the material. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

The Cancer Genome Atlas Research Network. Integrated genomic and molecular characterization of cervical cancer. Nature 543 , 378–384 (2017). https://doi.org/10.1038/nature21386

Download citation

Received : 07 December 2015

Accepted : 14 January 2017

Published : 23 January 2017

Issue Date : 16 March 2017

DOI : https://doi.org/10.1038/nature21386

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

This article is cited by

Rgs20 promotes non-small cell lung carcinoma proliferation via autophagy activation and inhibition of the pka-hippo signaling pathway.

• Xiaoyan Ding
• Robert Chunhua Zhao

Cancer Cell International (2024)

HPV16-miRNAs exert oncogenic effects through enhancers in human cervical cancer

• Yunuan Wang
• Xueying Wang

LncRNA WWTR1-AS1 upregulates Notch3 through miR-136 to increase cancer cell stemness in cervical squamous cell carcinoma

• Xiaofeng Zhou

BMC Women's Health (2024)

Long-read sequencing reveals the structural complexity of genomic integration of HPV DNA in cervical cancer cell lines

• Zhijie Wang
• Xiaoyuan Huang

BMC Genomics (2024)

Decoding the basis of histological variation in human cancer

• Masayuki Fujii
• Shigeki Sekine
• Toshiro Sato

Nature Reviews Cancer (2024)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

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

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

• Article Information
• USPSTF Recommendation: Screening for Cervical Cancer JAMA US Preventive Services Task Force August 21, 2018 This 2018 Recommendation Statement from the US Preventive Services Task Force makes recommendations regarding use of cervical cytology alone, high-risk HPV testing alone, and both in combination to screen for cervical cancer. US Preventive Services Task Force; Susan J. Curry, PhD; Alex H. Krist, MD, MPH; Douglas K. Owens, MD, MS; Michael J. Barry, MD; Aaron B. Caughey, MD, PhD; Karina W. Davidson, PhD, MASc; Chyke A. Doubeni, MD, MPH; John W. Epling Jr, MD, MSEd; Alex R. Kemper, MD, MPH, MS; Martha Kubik, PhD, RN; C. Seth Landefeld, MD; Carol M. Mangione, MD, MSPH; Maureen G. Phipps, MD, MPH; Michael Silverstein, MD, MPH; Melissa A. Simon, MD, MPH; Chien-Wen Tseng, MD, MPH, MSEE; John B. Wong, MD
• Screening for Cervical Cancer JAMA JAMA Patient Page August 21, 2018 This JAMA Patient Page describes the US Preventive Services Task Force’s recommendations on screening for cervical cancer. Jill Jin, MD, MPH
• Cervical Cancer Screening: More Choices in 2019 JAMA JAMA Insights May 28, 2019 This JAMA Clinical Insights review summarizes recent cervical cancer screening guidelines from the USPSTF, ACOG, and ACS, and proposes an algorithm for responding to abnormal screening results in women at average risk in age group 25 to 29 and older than 30. George F. Sawaya, MD; Karen Smith-McCune, MD, PhD; Miriam Kuppermann, PhD, MPH
• Managing Minimally Abnormal Cervical Cancer Screening Results JAMA JAMA Insights October 20, 2020 This JAMA Insights review summarizes 2020 consensus guidelines on management of minimally abnormal cervical cancer screening results, including no abnormal pathology, atypical squamous cells of undetermined significance (ASCUS), and low-grade squamous intraepithelial lesion (LSIL) with and without positive high-risk HPV testing. George F. Sawaya, MD; Robyn Lamar, MD, MPH; Rebecca B. Perkins, MD, MSc
• Cervical Cancer Screening Guideline for Individuals at Average Risk JAMA JAMA Clinical Guidelines Synopsis December 7, 2021 This JAMA Clinical Guidelines Synopsis summarizes guidance from the American Cancer Society regarding cervical cancer screening, including when to initiate screening, the frequency of testing, and which modality to use. Julie Chor, MD, MPH; Andrew M. Davis, MD, MPH; Jennifer M. Rusiecki, MD
• Patient Information: Cervical Cancer Screening JAMA JAMA Patient Page November 28, 2023 This JAMA Patient Page discusses screening for cervical cancer, including who should be screened, recommendations for handling positive results, and the prevention of cervical cancer. Rebecca A. Voelker, MSJ
• Incidence of Cervical Cancer in Puerto Rico, 2001-2017 JAMA Oncology Research Letter March 1, 2021 This cohort study examines recent trajectories in the incidence of cervical cancer in Puerto Rico by age and among birth cohorts. Ana Patricia Ortiz, PhD, MPH; Karen J. Ortiz-Ortiz, DrPH, MPH; Vivian Colón-López, PhD, MPH; Guillermo Tortolero-Luna, MD, PhD; Carlos R. Torres-Cintrón, MPH; Chi-Fang Wu, PhD; Ashish A. Deshmukh, PhD, MPH
• Association of Socioeconomic Status With Cervical Cancer Incidence in New York City JAMA Oncology Research Letter January 1, 2022 This cross-sectional study uses Agency for Healthcare Research and Quality data to compare incidence of cervical cancer in the lowest–socioeconomic status neighborhoods with that in higher–socioeconomic status neighborhoods of New York City from 2012 through 2016. Stephanie Cham, MD; Alicia Li, BA; J. Alejandro Rauh-Hain, MD, MPH; Ana I. Tergas, MD, MPH; Dawn L. Hershman, MD, MS; Jason D. Wright, MD; Alexander Melamed, MD, MPH
• Human Papillomavirus Vaccination and Trends in Cervical Cancer Incidence and Mortality in the US JAMA Pediatrics Research Letter March 1, 2022 This cohort study examines the association of human papillomvirus vaccination with cervical cancer incidence and mortality rates in the US, comparing age groups understood to have differing levels of vaccination. Tara Tabibi, BA; Justin M. Barnes, MD, MS; Aneri Shah, BS; Nosayaba Osazuwa-Peters, PhD, BDS, MPH, CHES; Kimberly J. Johnson, PhD, MPH; Derek S. Brown, PhD

See More About

Select your interests.

Customize your JAMA Network experience by selecting one or more topics from the list below.

• Academic Medicine
• Acid Base, Electrolytes, Fluids
• Allergy and Clinical Immunology
• American Indian or Alaska Natives
• Anesthesiology
• Anticoagulation
• Art and Images in Psychiatry
• Artificial Intelligence
• Assisted Reproduction
• Bleeding and Transfusion
• Caring for the Critically Ill Patient
• Challenges in Clinical Electrocardiography
• Climate and Health
• Climate Change
• Clinical Challenge
• Clinical Decision Support
• Clinical Implications of Basic Neuroscience
• Clinical Pharmacy and Pharmacology
• Complementary and Alternative Medicine
• Consensus Statements
• Coronavirus (COVID-19)
• Critical Care Medicine
• Cultural Competency
• Dental Medicine
• Dermatology
• Diabetes and Endocrinology
• Diagnostic Test Interpretation
• Drug Development
• Electronic Health Records
• Emergency Medicine
• End of Life, Hospice, Palliative Care
• Environmental Health
• Equity, Diversity, and Inclusion
• Facial Plastic Surgery
• Gastroenterology and Hepatology
• Genetics and Genomics
• Genomics and Precision Health
• Global Health
• Guide to Statistics and Methods
• Hair Disorders
• Health Care Delivery Models
• Health Care Economics, Insurance, Payment
• Health Care Quality
• Health Care Reform
• Health Care Safety
• Health Care Workforce
• Health Disparities
• Health Inequities
• Health Policy
• Health Systems Science
• History of Medicine
• Hypertension
• Images in Neurology
• Implementation Science
• Infectious Diseases
• Innovations in Health Care Delivery
• JAMA Infographic
• Law and Medicine
• Leading Change
• Less is More
• LGBTQIA Medicine
• Lifestyle Behaviors
• Medical Coding
• Medical Devices and Equipment
• Medical Education
• Medical Education and Training
• Medical Journals and Publishing
• Mobile Health and Telemedicine
• Narrative Medicine
• Neuroscience and Psychiatry
• Notable Notes
• Nutrition, Obesity, Exercise
• Obstetrics and Gynecology
• Occupational Health
• Ophthalmology
• Orthopedics
• Otolaryngology
• Pain Medicine
• Palliative Care
• Pathology and Laboratory Medicine
• Patient Care
• Patient Information
• Performance Improvement
• Performance Measures
• Perioperative Care and Consultation
• Pharmacoeconomics
• Pharmacoepidemiology
• Pharmacogenetics
• Pharmacy and Clinical Pharmacology
• Physical Medicine and Rehabilitation
• Physical Therapy
• Physician Leadership
• Population Health
• Primary Care
• Professional Well-being
• Professionalism
• Psychiatry and Behavioral Health
• Public Health
• Pulmonary Medicine
• Regulatory Agencies
• Reproductive Health
• Research, Methods, Statistics
• Resuscitation
• Rheumatology
• Risk Management
• Scientific Discovery and the Future of Medicine
• Shared Decision Making and Communication
• Sleep Medicine
• Sports Medicine
• Stem Cell Transplantation
• Substance Use and Addiction Medicine
• Surgical Innovation
• Surgical Pearls
• Teachable Moment
• Technology and Finance
• The Art of JAMA
• The Arts and Medicine
• The Rational Clinical Examination
• Tobacco and e-Cigarettes
• Translational Medicine
• Trauma and Injury
• Treatment Adherence
• Ultrasonography
• Users' Guide to the Medical Literature
• Vaccination
• Venous Thromboembolism
• Veterans Health
• Women's Health
• Workflow and Process
• Wound Care, Infection, Healing

Others Also Liked

• Download PDF
• X Facebook More LinkedIn

Shahmoradi Z , Damgacioglu H , Clarke MA, et al. Cervical Cancer Incidence Among US Women, 2001-2019. JAMA. 2022;328(22):2267–2269. doi:10.1001/jama.2022.17806

Manage citations:

© 2024

• Permissions

Cervical Cancer Incidence Among US Women, 2001-2019

• 1 Center for Health Services Research, UTHealth School of Public Health, Houston, Texas
• 2 Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, Maryland
• 3 Department of Pediatrics, Baylor College of Medicine, Houston, Texas
• 4 Department of Public Health Sciences, Medical University of South Carolina, Charleston
• US Preventive Services Task Force USPSTF Recommendation: Screening for Cervical Cancer US Preventive Services Task Force; Susan J. Curry, PhD; Alex H. Krist, MD, MPH; Douglas K. Owens, MD, MS; Michael J. Barry, MD; Aaron B. Caughey, MD, PhD; Karina W. Davidson, PhD, MASc; Chyke A. Doubeni, MD, MPH; John W. Epling Jr, MD, MSEd; Alex R. Kemper, MD, MPH, MS; Martha Kubik, PhD, RN; C. Seth Landefeld, MD; Carol M. Mangione, MD, MSPH; Maureen G. Phipps, MD, MPH; Michael Silverstein, MD, MPH; Melissa A. Simon, MD, MPH; Chien-Wen Tseng, MD, MPH, MSEE; John B. Wong, MD JAMA
• JAMA Patient Page Screening for Cervical Cancer Jill Jin, MD, MPH JAMA
• JAMA Insights Cervical Cancer Screening: More Choices in 2019 George F. Sawaya, MD; Karen Smith-McCune, MD, PhD; Miriam Kuppermann, PhD, MPH JAMA
• JAMA Insights Managing Minimally Abnormal Cervical Cancer Screening Results George F. Sawaya, MD; Robyn Lamar, MD, MPH; Rebecca B. Perkins, MD, MSc JAMA
• JAMA Clinical Guidelines Synopsis Cervical Cancer Screening Guideline for Individuals at Average Risk Julie Chor, MD, MPH; Andrew M. Davis, MD, MPH; Jennifer M. Rusiecki, MD JAMA
• JAMA Patient Page Patient Information: Cervical Cancer Screening Rebecca A. Voelker, MSJ JAMA
• Research Letter Incidence of Cervical Cancer in Puerto Rico, 2001-2017 Ana Patricia Ortiz, PhD, MPH; Karen J. Ortiz-Ortiz, DrPH, MPH; Vivian Colón-López, PhD, MPH; Guillermo Tortolero-Luna, MD, PhD; Carlos R. Torres-Cintrón, MPH; Chi-Fang Wu, PhD; Ashish A. Deshmukh, PhD, MPH JAMA Oncology
• Research Letter Association of Socioeconomic Status With Cervical Cancer Incidence in New York City Stephanie Cham, MD; Alicia Li, BA; J. Alejandro Rauh-Hain, MD, MPH; Ana I. Tergas, MD, MPH; Dawn L. Hershman, MD, MS; Jason D. Wright, MD; Alexander Melamed, MD, MPH JAMA Oncology
• Research Letter Human Papillomavirus Vaccination and Trends in Cervical Cancer Incidence and Mortality in the US Tara Tabibi, BA; Justin M. Barnes, MD, MS; Aneri Shah, BS; Nosayaba Osazuwa-Peters, PhD, BDS, MPH, CHES; Kimberly J. Johnson, PhD, MPH; Derek S. Brown, PhD JAMA Pediatrics

A recent US study reported that previously declining cervical cancer incidence has plateaued between 2012 and 2017. 1 A significant reduction in cervical cancer screening uptake and adherence to guidelines-concordant recommendations has also been reported, particularly among women aged 21 to 29 years. 2 We evaluated calendar trends in cervical cancer incidence by age at diagnosis.

We analyzed the 2001-2019 National Program of Cancer Registries (NPCR) and Surveillance, Epidemiology, and End Results (SEER) data set. This data set includes cancer incidence data from all 50 states and the District of Columbia and covers more than 98% of the US population. Cervical cancer cases were identified based on the International Classification of Diseases for Oncology, Third Edition site codes C53.0 to C53.9 and histology codes 8010 to 8671 and 8940 to 8941. We calculated incidence rates and estimated piecewise-log-linear trends and annual percentage changes (APCs) by 5-year age group. Hysterectomy-corrected incidence trends were assessed by calculating the survey-weighted prevalence of hysterectomy from Behavioral Risk Factor Surveillance System data and then using it to correct the population at risk by removing the proportion of women with hysterectomy from the denominator. For age groups with increasing incidence, we examined trends by race and ethnicity, stage, and histology. Statistical analysis was conducted using SEER*Stat version 8.4.0 and Joinpoint Regression (National Cancer Institute) version 4.9.0. Statistical significance was tested at 2-sided P  < .05. The University of Texas Health Science Center institutional review board deemed the study exempt from review and waived the requirements for informed consent because publicly available data were used.

During 2001-2019, 227 062 cervical cancer cases were reported. Overall, hysterectomy-corrected cervical cancer incidence declined from 12.39/100 000 in 2001 to 9.80/100 000 in 2019 (APC, −1.2% [95% CI, −1.6% to −0.9%]). Reductions in incidence were observed for the youngest (<24 years) and oldest (≥55 years) age groups, and rates were relatively stable in recent years among women aged 35 to 54 years ( Table 1 ). Among women aged 30 to 34 years, after an initial decline from 2001-2012 (incidence, 12.77/100 000 to 10.14/100 000; APC, −2.3% [95% CI, −2.8% to −1.7%]), incidence increased during 2012-2019 (APC, 2.5% [95% CI, 1.4% to 3.6%]), reaching 11.60/100 000 in 2019.

During 2012-2019, for 30- to 34-year-old women, hysterectomy-corrected cervical cancer incidence increased significantly for Hispanic (APC, 3.0% [95% CI, 0.3% to 5.7%]), non-Hispanic White (APC, 2.8% [95% CI, 0.6% to 5.0%]), and other racial and ethnic (APC, 5.0% [95% CI, 2.7% to 7.4%]) groups; the APC for Black women was −0.8% (95% CI, −2.8% to 1.2%) ( Table 2 ). Increases occurred for localized (2.8% [95% CI, 1.3% to 4.3%]) and regional (1.9% [95% CI, 0.7% to 3.1%]) stage cervical cancer as well as for squamous cell carcinoma (2.6% [95% CI, 1.0% to 4.2%]) and adenocarcinoma (3.0% [95% CI, 0.9% to 5.1%]) histology.

Between 2001 and 2019, cervical cancer incidence declined or remained stable among US women except for the 30- to 34-year-old age group, in whom incidence increased 2.5% per year after 2012. The observed increase in incidence among 30- to 34-year-old women could be real as a result of a true increase in cervical cancer incidence or due to increased early detection with a stable disease occurrence. If the increase is real, it could be a result of missed screening opportunities at earlier ages, as suggested by the increase in squamous cell carcinoma and localized disease. It may also stem from a decrease in screening at younger ages. In 2012, the US Preventive Services Task Force recommended an increase in screening interval in 21- to 65-year-old women with cytology every 3 years or in 30- to 65-year-old women with a combination of cytology and human papillomavirus testing every 5 years. 3 Beginning in 2013, declines in screening participation among 21- to 29-year-old women were observed. 2 However, the recommendation for co-screening with human papillomavirus testing and cytology may have led to increased detection of early-stage cancers; if so, the increased incidence would be expected to decline in the future. 4

Study limitations include the unavailability of information regarding risk factors and screening and that hysterectomy data were self-reported and subject to misclassification; however, self-report has been considered a valid approach with accuracy comparable with medical records. 5 Future studies are needed to assess factors that underlie the increase in cervical cancer incidence among 30- to 34-year-old women.

Accepted for Publication: September 11, 2022.

Published Online: November 21, 2022. doi:10.1001/jama.2022.17806

Corresponding Author: Ashish A. Deshmukh, PhD, MPH, Department of Public Health Sciences, Hollings Cancer Center, Medical University of South Carolina (MUSC), 86 Jonathan Lucas St, Charleston, SC 29424 ( [email protected] ).

Author Contributions: Drs Damgacioglu and Deshmukh had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Shahmoradi, Damgacioglu, Sonawane, Deshmukh.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Shahmoradi, Deshmukh.

Critical revision of the manuscript for important intellectual content: Damgacioglu, Clarke, Wentzensen, Richards Montealegre, Sonawane, Deshmukh.

Statistical analysis: Shahmoradi, Damgacioglu, Deshmukh.

Obtained funding: Richards Montealegre, Sonawane, Deshmukh.

Administrative, technical, or material support: Clarke, Wentzensen, Deshmukh.

Supervision: Sonawane, Deshmukh.

Conflict of Interest Disclosures: Dr Sonawane reported receiving personal fees from Value Analytics Labs outside the submitted work. No other disclosures were reported.

Funding/Support: Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health (award numbers R01CA256660, R01CA232888, and P30CA138313) and the National Institute on Minority Health and Health Disparities (award number K01MD016440).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

• Register for email alerts with links to free full-text articles
• Access PDFs of free articles
• Manage your interests
• Save searches and receive search alerts

Clinical Trials

Cervical cancer.

Displaying 27 studies

The purpose of this study is to answer whether plasma circulating tumor DNA (ctDNA) obtained by serial analysis before, during, and following surgery, radiotherapy, chemotherapy, and/or immunotherapy for cervical cancer will allow for risk stratification, individualized treatment decision making, monitoring of treatment response, and early detection of residual or recurrent disease in patients presenting with human papillomavirus (HPV) mediated cervical cancer.

This clinical trial is studying biomarkers in diagnosing cervical lesions in patients with abnormal cervical cells. Studying biomarkers in abnormal cervical cells may improve the ability to find cervical lesions and plan effective treatment.

The purpose of this study is to develop a pan-gynecologic cancer detection test using gynecologic (unique endometrial, cervical, and ovarian cancer) cancer-specific methylated DNA markers and high-risk human papilloma virus (HR-HPV) detected in vaginal fluid and/or plasma.

The purpose of this study is to see how well radiation therapy and cisplatin with triapine work in combination compared to the standard radiation therapy and cisplatin alone to treat patients with newly diagnosed stage IB2, II, or IIIB-IVA cervical cancer or stage II-IVA vaginal cancer. Radiation therapy uses high energy protons to kill tumor cells and shrink tumors. Drugs used in chemotherapy, such as cisplatin, work in different ways to stop the growth of tumor cells, either by killing the cells, by stopping them from dividing or by stopping them from spreading. Triapine may stop the growth of tumor cells ...

The purpose of this study is to validate a patient self-sampling vaginal collection kit and laboratory testing for the detection of HPV (human papillomavirus) infection. Its secondary aim is to assess the patient perspective on acceptability and feasibility of a self-sampling approach to cervical cancer screening through a survey of study enrollees. The third aim will assess the stability testing component to analyze concordance of multiple self-collected Evalyn brush HPV results within the same study subject.

The purposes of this study are to assess awareness and acceptability of HPV testing for cervical cancer screening and patient self-collection of vaginal swabs among women aged 30 to 65 years old, to assess awareness and acceptability of HPV testing for cervical cancer screening and patient self-collection of vaginal swabs among primary care clinicians, and to identify insights from regional and national thought leaders in the fields of primary care and cervical cancer screening on strategies to improve uptake of primary HPV screening and opinions on patient self-collection by conducting key informant interviews. 

This clinical trial studies lymphedema after surgery in patients with endometrial cancer, cervical cancer, or vulvar cancer. Collecting information over time about how often lymphedema occurs in patients undergoing surgery and lymphadenectomy for endometrial cancer, cervical cancer, and vulvar cancer may help doctors learn more about the disease and plan the best treatment.

This research trial studies comprehensive patient questionnaires in predicting complications in older patients with gynecologic cancer undergoing surgery. Comprehensive patient questionnaires completed before surgery may help identify complications, such as the need for assistance in taking medications, decreased mobility, decreased social activity, and falls, and may improve outcomes for older patients with gynecologic cancer.

This study is a process evaluation of a digital storytelling (DST) intervention. Because the DST intervention has not yet been developed, it is appropriate to use qualitative methods to assess the process.

This study will recruit gifted storytellers for a digital storytelling intervention on breast, cervical, and colorectal cancer. A digital storytelling intervention will be developed to improve breast, cervical, and colorectal cancer screening rates among Hispanic, Spanish speaking individuals. Additionally, to conduct a qualitative assessment of the digital storytelling workshop.

Part A: (ADXS11-001 + MEDI4736 Combination Therapy) will determine the safety and tolerability of the combination and to identify a RP2D.

Part B: Phase 2 design which will randomize subjects 1:1 to either MEDI4736 alone or MEDI4736+ADXS11-001 in subjects who have failed at least 1 prior systemic treatment for their recurrent/persistent or metastatic cervical cancer.

Effective treatments for recurrent gynecological cancer are lacking, and there is a need for novel therapeutic options. One of the barriers to improving outcomes in this subgroup of patients is the paucity of tumor models that can mimic patient characteristics to study novel therapies. Patient-derived xenograft (PDX) models are considerate the most representative pre-clinical model of human cancer, recapitulating the key characteristics of the original patient tumor. Other preclincal models to test drug effcicacy includes ex vivo 3D culture and 2D culture systems. In this study, we make and test preclinical models of gynecological cancers (ovarian, fallopian tube, peritoneal, uterine, ...

The purpose of this study is to quantify the extent of sexual dysfunction in our patients to explore if there is a need for a prospective intervention, with the ultimate goal of improving the quality of life of our patients, with respect to sexual function.

This randomized pilot phase II trial studies how well serum bovine immunoglobulin protein isolate works in improving quality of life and post-operative recovery in patients with gynecological cancer after undergoing surgery. Serum bovine immunoglobulin supplementation may improve the quality of life of patients with gynecological cancer.

This study proposes to develop and maintain a biorepository of blood samples collected from patients receiving definitive chemoradiotherapy for locally advanced rectal cancer, locally advanced pancreatic cancer, non-small cell lung cancer, or cervical cancer. The ultimate goal of this biorepository will be to provide the resource to initiate an exploration of ctDNA as a potential liquid biopsy for GI and Thoracic malignancy detection and surveillance.

A Phase 1a/1b Open-Label Study to Evaluate the Safety, Tolerability, Pharmacokinetics, and Pharmacodynamics of PY314 as a Single Agent and In Combination with Pembrolizumab in Subjects with Advanced Solid Tumors

The purpose of this study is to determine the recommended Phase 2 dose (RP2D) of TBio-6517 when administered by direct injection into tumor(s) alone and when combined with pembrolizumab in patients with solid tumors (RIVAL-01).

This phase I trial studies the side effects and best dose of paclitaxel albumin-stabilized nanoparticle formulation and bevacizumab in treating patients with stage IV melanoma, gynecological cancers, or other malignancies that cannot be removed by surgery. Drugs used in chemotherapy, such as paclitaxel albumin-stabilized nanoparticle formulation, work in different ways to stop the growth of tumor cells, either by killing the cells or by stopping them from dividing. Monoclonal antibodies, such as bevacizumab, can block tumor growth in different ways. Some block the ability of tumor cells to grow and spread. Others find tumor cells and help kill them or carry tumor-killing ...

The purpose of this trial is to evaluate the safety of GEN1046 in patients with malignant solid tumors.

The purpose of the study is to assess the safety, tolerability and effectiveness of experimental medication BMS-986016 administered alone and in combination with nivolumab in patients with solid tumors that have spread and/or cannot be removed by surgery. The following tumor types are included in this study: Non-Small Cell Lung Cancer (NSCLC), gastric cancer, hepatocellular carcinoma, renal cell carcinoma, bladder cancer, squamous cell carcinoma of the head and neck, and melanoma, that have NOT previously been treated with immunotherapy. NSCLC and melanoma that HAVE previously been treated with immunotherapy.

The purpose of this study is to sequence patient germline and tumor samples, and nominate top neoantigen candidates using an in-house developed bioinformatics pipeline, and to validate the neoantigen candidates by laboratory assays using patient peripheral blood immune cells or serum.

The purpose of this study is to test the safety and tolerability of HFB200301 in patients with advanced cancers. There are two parts in this study. During the escalation part, groups of participants will receive increasing doses until a safe and tolerable dose of HFB200301 is determined. During the expansion part, participants will take the dose of study drug that was determined from the escalation part of the study and will be assigned to a group based on the type of cancer the participants have.

The purpose of this study is to determine the prevalence of genetic mutations in cancer patients from various ethnic populations seeking care at Mayo Clinic cancer clinics.

The purpose of this multicenter prospective observational case-control study is to train and validate Adela’s cfMeDIP-seq based methylome profiling platform to detect and differentiate multiple cancer subtypes. In addition, this study includes longitudinal follow-up for a subset of participants to train and validate the methylome profiling platform to detect minimal residual disease and recurrence.

The purpose of this study is to evaluate the safety, tolerability, and preliminary anti-tumor activity of SAR444881 alone and in combination with pembrolizumab or with cetuximab. The study will enroll advanced cancer patients with unresectable or metastatic disease who are refractory to or are not candidates for standard approved therapy and will be comprised of two parts - an initial "3 + 3" dose escalation phase (Part 1) with Sub-Parts 1A (monotherapy SAR444881), 1B (SAR444881 in combination with pembrolizumab) and 1C (SAR444881 in combination with cetuximab) followed by a dose optimization/expansion phase (Part 2), including Sub-Part 2A (Dose Optimization) with Cohorts A1 (SAR444881 in ...

The purpose of this study is to evaluate the challenges, behavioral patterns, and preferences of minority patient participation in clinical trials. Also, to develop and validate a personalized clinical trial educational platform to boost participation among underserved cancer patients.

GRAIL is using deep sequencing of circulating cell-free nucleic acids (cfNAs) to develop assays to detect cancer early in blood. The purpose of this study is to collect biological samples from donors with a new diagnosis of cancer (blood and tumor tissue) and from donors who do not have a diagnosis of cancer (blood) in order to characterize the population heterogeneity in cancer and non-cancer subjects and to develop models for distinguishing cancer from non-cancer.

Falls are common and catastrophic in cancer patients. Cancer patients are vulnerable to falls due to muscle loss. In prescribing exercise in a data driven manner to cancer patients, our hypothesis is this "prescription" for exercise will eventually be demonstrated to reduce the occurrence of injurious falls.

Mayo Clinic Footer

• Request Appointment
• About Mayo Clinic
• About This Site

Legal Conditions and Terms

• Terms and Conditions
• Privacy Policy
• Notice of Privacy Practices
• Notice of Nondiscrimination
• Manage Cookies

Advertising

Mayo Clinic is a nonprofit organization and proceeds from Web advertising help support our mission. Mayo Clinic does not endorse any of the third party products and services advertised.

• Advertising and sponsorship policy
• Advertising and sponsorship opportunities

Reprint Permissions

A single copy of these materials may be reprinted for noncommercial personal use only. "Mayo," "Mayo Clinic," "MayoClinic.org," "Mayo Clinic Healthy Living," and the triple-shield Mayo Clinic logo are trademarks of Mayo Foundation for Medical Education and Research.

Large Study Confirms that HPV Vaccine Prevents Cervical Cancer

October 14, 2020 , by NCI Staff

A new study confirms that widespread use of the HPV vaccine reduces the incidence of cervical cancer, particularly for women who are vaccinated when they are younger.

In what many global health leaders are calling a milestone study, researchers in Sweden have confirmed that widespread use of the human papillomavirus (HPV) vaccine dramatically reduces the number of women who will develop cervical cancer.

In the study of nearly 1.7 million women, the vaccine’s efficacy was particularly pronounced among girls vaccinated before age 17, among whom there was a nearly 90% reduction in cervical cancer incidence during the 11-year study period (2006 through 2017) compared with the incidence in women who had not been vaccinated.

“This is a vaccine against cancer, which can save lives,” said the study’s leader, Jiayao Lei, Ph.D., of the Karolinska Institute in Stockholm.

On Twitter, Noel Brewer, Ph.D., who studies cancer prevention and HPV vaccines at the University of North Carolina, called the study results “incredibly powerful.” The findings were published September 30 in the New England Journal of Medicine .

Studies and clinical trials to date have consistently shown that HPV vaccines are extremely effective at reducing infections with the types of the virus that can lead to cancer, as well as cervical precancers. But because of the long time between infection and cancer, it had yet to be shown that HPV vaccination prevents cervical cancers.

“Because HPV vaccination prevents persistent HPV infection and cervical precancer, the precursors to cervical cancer, we were confident we would eventually observe that HPV vaccination prevents cervical cancer. We also knew it would take time to observe that,” said Aimée R. Kreimer, Ph.D., of NCI’s Division of Cancer Epidemiology and Genetics , who studies HPV vaccines and cancer prevention.

“Cervical cancer can be a devastating diagnosis,” said Abbey Berenson, M.D., Ph.D., who specializes in women’s health at the University of Texas Medical Branch. The study results, Dr. Berenson continued, “send an incredibly important message” about the impact widespread use of the HPV vaccine can have.

The Missing Piece of Evidence

Large clinical trials of HPV vaccines—which enrolled thousands of participants and followed them over time—assessed their ability to prevent cervical infections with cancer-causing types of HPV and the development of precancerous lesions in the cervix that can result from those infections.

The clinical trials did not measure whether the vaccine prevents cervical cancer because precancerous lesions in the cervix found during a clinical trial would be treated, preventing their progression to cancer, Dr. Kreimer explained.

The Swedish study, however, looked back in time at a huge population of women. And for their study, the Swedish researchers had two factors in their favor: the individual-level data in the country’s nationwide public health registry and the fact that, beginning in 2007, the country has conducted a series of nationwide HPV vaccination programs.

The Swedish study isn’t the first large population-based study of HPV vaccines. In Australia, for example, researchers have shown that the country’s universal HPV vaccination program, launched in 2007, led to massive declines in infections with the HPV types covered by the vaccine , while also protecting against HPV infections in unvaccinated people, a phenomenon known as herd immunity .

It’s logical to conclude that if the vaccine slashes infections with cancer-causing HPV types and the development of advanced precancerous cervical lesions in women, then far fewer diagnoses of invasive cervical cancers should follow in the ensuing years, Dr. Berenson said.

Nevertheless, no studies had gone on long enough to deliver that logic to its anticipated result.

For HPV Vaccination: The Younger, the Better

The Swedish study is the largest to compare cervical cancer diagnoses among women who did and did not receive an HPV vaccine. In Sweden, the only HPV vaccine available during the time period studied was one that protects against four HPV types: HPV 6, HPV 11, HPV 16, and HPV 18. Infections with types 16 and 18 are responsible for approximately 70% of cervical cancers, and types 6 and 11 cause 90% of genital warts.

All of the females followed in the study were between the ages of 10 and 30. Approximately 528,000 of them had received at least one dose of the vaccine between 2006 and 2017, and the remaining 1.14 million had not been vaccinated. More than 80% of those vaccinated received the vaccine before they were 17 years old.

Overall, 19 of the vaccinated women were diagnosed with cervical cancer during the study period, compared with 538 of the unvaccinated women. After adjusting for different factors that can influence cervical cancer risk, those numbers translated into a 63% reduced risk of being diagnosed with cervical cancer among females who had been vaccinated compared with those who hadn’t.

The nearly 90% reduction in cervical cancer among women who were vaccinated at a younger age makes sense, said Dr. Kreimer.

Many women who received the vaccine after age 17 would be more likely to have HPV infections at the time of vaccination, and the vaccine only works to prevent infections, not stop existing infections. “Thus, the older girls were more likely to have had infections at the time of vaccination that were, therefore, not preventable and could progress to cancer.”

According to the study’s senior researcher, Pär Sparén, Ph.D., also of Karolinska, the findings affirm the need for broader use of the HPV vaccine among women in low- and middle-income countries, in which cervical cancer is often one of the leading causes of death.

“The evidence … highlights the importance of continuing to introduce HPV vaccination programs and maintaining a high [vaccine] coverage, preferably for girls at young age, to maximize the benefits,” Dr. Sparén said.

A Boost for Vaccinations?

The Swedish study does have some limitations. For example, it couldn’t account for factors such as the extent to which women followed in the study were screened for cervical cancer, the study team reported. The researchers also couldn’t capture the number of vaccine doses each person in the vaccination group received.  

“But that’s not a big limitation for this type of study. It’s not a dose question,” Dr. Kreimer said. For this study, she continued, “They were saying, ‘We established a vaccine program in a population, and this is how [well] it worked.’”

Although rates of HPV vaccination have increased among adolescents and teens in the United States , they still are lower than public health officials would like. Dr. Berenson said she’s hopeful that the findings from the Swedish study can provide a boost.

“[The findings] provide a very good discussion point around age of vaccination,” she said. And that’s needed, she added, because parents sometimes are reluctant to have their daughters receive the HPV vaccine at the recommended age, which is 11-12 years old.

“They often tell us they want to wait until she’s older—until she’s 18—saying, ‘She can make the decision for herself,’” Dr. Berenson said. “This study gives [pediatricians] good evidence to say, ‘We understand why you may feel that way, but you are missing the opportunity of much higher efficacy if she gets vaccinated at a younger age.’”

It’s taken some time, but the findings from the Swedish study complete the story of the HPV vaccine, Dr. Kreimer said. “This provides the final key piece of evidence in the pathway from [HPV] infection to cancer," she said, "and HPV vaccination protects against all of it.”

Featured Posts

March 27, 2024, by Edward Winstead

March 21, 2024, by Elia Ben-Ari

March 5, 2024, by Carmen Phillips

• Biology of Cancer
• Cancer Risk
• Childhood Cancer
• Clinical Trial Results
• Disparities
• FDA Approvals
• Global Health
• Leadership & Expert Views
• Screening & Early Detection
• Survivorship & Supportive Care
• February (6)
• January (6)
• December (7)
• November (6)
• October (7)
• September (7)
• February (7)
• November (7)
• October (5)
• September (6)
• November (4)
• September (9)
• February (5)
• October (8)
• January (7)
• December (6)
• September (8)
• February (9)
• December (9)
• November (9)
• October (9)
• September (11)
• February (11)
• January (10)

• Open access
• Published: 18 May 2024

Correlates of intention to screen for cervical cancer among adult women in Kyotera District, Central Uganda: a community based cross-sectional study

• Richard Kabanda 1 ,
• Arthur Kiconco 2 , 3 ,
• Anguzu Ronald 1 ,
• Kirsten M. M. Beyer 1 &
• Steven A. John 4  

BMC Women's Health volume  24 , Article number:  296 ( 2024 ) Cite this article

Metrics details

Introduction

Cervical cancer continues to pose a major public health challenge in low-income countries. Cervical cancer screening programs enable early detection and effectively reduce the incidence of cervical cancer as well as late-stage diagnosis and mortality. However, screening uptake remains suboptimal in Uganda. This study assessed correlates of intention to screen for cervical cancer among women in the Kyotera district of Central Uganda.

We analyzed cross-sectional data collected to determine the effectiveness of community audio towers (CATs) as a modality of health communication to support cervical cancer prevention. Women ( n  = 430) aged 21–60 years without a prior history of cervical cancer screening were surveyed about demographics, sources of health information and cervical cancer screening intentions in 2020. We used generalized linear modelling with modified Poisson regression and backwards variable elimination to identify adjusted prevalence ratios and 95% confidence intervals (CI) to determine factors associated with intention to screen for cervical cancer.

Half (50.2%) of the participants had intentions to screen for cervical cancer within twelve months and 26.5% had moderate knowledge about cervical cancer. Nearly half (46.0%) considered themselves at risk of cervical cancer. Compared to residents who primarily received their health information from social media and radio, participants who received health information primarily from CATs (aPR:0.64, 95% CI:0.52–0.80, p  < 0.001) and TV (aPR:0.52, 95% CI:0.34–0.82, p  = 0.005) had a lower prevalence of intention to screen for cervical cancer. The prevalence of intentions to screen for cervical cancer in twelve months was higher among those resided in town councils (aPR:1.44, 95% CI:1.12–1.86, p  = 0.004) compared to rural areas, and higher among those who considered themselves to be at risk of cervical cancer (aPR:1.74, 95% CI:1.28–2.36, p  < 0.001) compared to those who did not.

Conclusions

We found suboptimal prevalence of intentions to screen for cervical cancer among women in central Uganda. Additional research and implementation projects are needed to increase cervical cancer screening. Targeting risk perceptions and behavioral approaches to increase intentions could be effective in future intervention work. Based on urban-rural differences, additional work is needed to support equitable sharing of information to support cancer prevention messaging; CATs and TV may best help reach those with lower intentions to screen based on our research.

Peer Review reports

Cervical cancer continues to pose a major global public health burden, with over 340,000 deaths annually [ 1 ]; projections estimate this number increasing to 400,000 annual deaths by 2030 [ 2 ]. Cervical cancer is the fourth most common cancer among women globally [ 3 ], with an estimated 604,127 new cases of cervical cancer in 2020 [ 1 ] and an anticipated increase to 700,000 by 2030 [ 4 ]. Cervical cancer is among the common human papillomavirus (HPV)-related diseases, with nearly all cases of cervical cancer attributable to HPV infection; specifically, HPV types 16 and 18 are known to cause 70% of cervical cancers and precancerous cervical lesions [ 5 , 6 , 7 ].

There are significant socioeconomic disparities in cervical cancer incidence rates, with national rates increasing as the Human Development Index (HDI) decreases; the poor, especially in low- and middle-income countries (LMICs), shoulder the largest disease burden [ 1 ]. The highest cervical cancer incidence occurs in Africa, followed by Latin America, Asia, and Melanesia. Within sub-Saharan Africa, the 2020 age-adjusted incidence rate for cervical cancer was highest in eastern Africa, estimated at 40 cases per 100,000 women-years [ 1 ]. The 2023 age-adjusted incidence and mortality rates for Uganda were 56.2 and 41.4 respectively [ 8 ]. In the same year, the annual estimates indicated that 6,959 women were diagnosed with cervical cancer and 4,607 died from the disease—making it the first most frequent cancer among women in Uganda [ 8 ].

Despite disparities in cervical cancer incidence rates, resources for prevention, diagnosis and treatment are limited in most LMICs [ 9 , 10 ]. Although preventable and curable if identified at an early stage, cervical cancer remains a top cancer killer of women in low-resource settings [ 11 ]. The HIV/AIDS epidemic is also believed to exasperate high rates of cervical cancer incidence and mortality, as the risk of development, progression, and recurrence of HPV-induced cervical precursor lesions and cervical cancer are higher among women living with HIV (WLHIV) [ 12 , 13 , 14 , 15 , 16 ]. Despite reductions in HIV new infections in Uganda, the HIV prevalence remains high at 7.2% among women compared to 4.3% among men [ 17 ].

The World Health Organization (WHO) Global Cervical Cancer Elimination Initiative (GCCEI) aims to reduce incidence below a threshold of 4 cases per 100,000 women-years in every country [ 2 ]. Cervical cancer is the number one cause of cancer-related deaths among women in Uganda [ 18 ], and the WHO estimates approximately 3,915 Ugandan women were diagnosed with cervical cancer and 2,160 died from the disease in 2014 [ 19 ]. In Uganda, cervical cancer screening guidelines recommend visual inspection of the cervix with acetic acid (VIA) annually for women living with HIV and every 3 years for those HIV-negative [ 20 ].

Cervical cancer screening programs enable the detection of cervical lesions before they become cancerous, which can effectively reduce the incidence of cervical cancer by 75–90% [ 21 , 22 ]. Screening also results in earlier detection of cancer, improving prognosis among those diagnosed and treated. As such, population-based cervical cancer screening programs are effective in reducing cervical cancer mortality [ 23 , 24 ]. Despite these statistics, only a small percentage (estimated at 19%) of women have been screened for cervical cancer in LMICs, compared to 63% in high-income countries [ 25 ]. In Uganda, it is estimated that the percentage of women who had ever screened for cervical cancer ranged from 9 to 10% and only about 7.5% had screened in the last 5 years in 2023 [ 8 ].

Further, researchers have previously attributed low cervical cancer screening uptake to a number of key factors, including limited resources required for successful screening programs [ 25 , 26 ], cervical cancer knowledge gaps [ 27 , 28 , 29 , 30 ], fear of positive diagnosis [ 31 ], and lower risk perception and negative attitudes [ 32 ]. The SARS-CoV-2 (i.e., COVID-19) pandemic is also believed to have led to delays in diagnosis and treatment due closures of health facilities, disruptions in access due to loss of insurance as people were laid off from work, and fear of COVID-19 exposure by those eligible for screening and care [ 33 ]. Most cervical cancer prevention programs aimed at increasing screening uptake usually focus on modifiable contextual factors such as knowledge, women’s intentions, and service availability, among others. However, few studies have assessed correlates of intention to screen for cervical cancer. As such, we assessed correlates of intention to screen for cervical cancer among adult women in Kyotera District, Central Uganda.

Study design

This was a cross-sectional analytical study based on secondary analysis of data collected at baseline for a study to determine the efficacy of community audio towers (CATs) as a health communication channel used in the prevention of cervical cancer in rural communities in Uganda [ 34 ]. The primary study was carried out between March and June 2020. It compared the use of CATs to disseminate messages on cervical cancer versus other health communication channels and cervical cancer screening among women aged 21 to 60 years. This analysis focused on data collected at baseline, prior to the use of the CATs for dissemination of cervical cancer-prevention messaging.

Study setting

This study was carried out in Kyotera district, located in the south-central region, southwest of Kampala Capital City in Uganda. Kyotera District headquarters are approximately 182 km from Kampala and forty-seven kilometers from Masaka City. Kyotera District was created from Rakai District in the year 2015 by an Act of Parliament but started operating as an independent district and local government on July 1, 2017, with two counties of Kakuuto and Kyotera. The district is primarily rural and borders with Kalangala, Masaka, Rakai, and Lwengo districts in Uganda and the Missenyi district in the south, which is in the Kagera region of Republic of Tanzania.

Kyotera District was part of Rakai where the first case of HIV/AIDS in Uganda was discovered at the Uganda-Tanzania border of Mutukula [ 35 ]. The district is known for its high HIV prevalence, currently standing at 11.1% [ 36 ]. There is a known link between HIV/AIDS and cancers, including cancer of the cervix, which shares similar risk factors. Although there are no disaggregated data showing the district prevalence of cervical cancer, the prevalence of cancer of the cervix is likely high in Kyotera.

Inclusion and exclusion criteria

The study population consisted of women aged 21–60 years living in Kyotera district. To be eligible, participants were required to: (1) be aged 21–60 years; (2) have lived in Kyotera for at least 3 months; and (3) have direct access to information as narrow-casted from CATs. Participants were excluded if (1) they had previously screened for cervical cancer in the past three years or one year for those LHIV; and (2) intended to relocate in the proceeding 16 weeks at the time of the survey. All participants consented to participate in this study.

Sample size and sampling

Sample size.

The sample size for the parent study was 480 participants. Each cluster (village) had sixty participants, and eight clusters were included. Fifty (50) of the participants had screened in the previous three years and were excluded from this sample; thus, the final sample for analysis was 430 participants.

Sampling and recruitment procedures

The initial process of sampling was based on the composition of Kyotera district in terms of counties. Kyotera has two counties, and each of these forms a health subdistrict. Recruitment of participants from clustered villages was done by systematic sampling from a list of households registered by community health workers to have the targeted age group. Where households had more than one eligible participant, only one was sampled and the lottery method was used to select one.

Study variables

Dependent variable.

The dependent variable for this study was intention to screen for cervical cancer. This was measured using three questions: (i) If never screened for cervical cancer, would you like to be screened? With responses: ‘Yes,’ ‘No’ or ‘I do not know.’ ‘No’ and ‘I do not know’ were merged as No; (ii) If yes above, when do you intend to have the screening done? With responses: in three months, six months, one year, not sure and never; (iii) Where would you like to go for the screening? With responses: nearby government hospital, private health facility, regional referral hospital, or any other. A previous study in an area closer to the study area measured intention using two questions of whether one intended to go for screening and when [ 32 ], but we added a third question of where they intended to go for the screening. Those who responded ‘yes’ in the first question, intention to go in either three or six months or one year and indicated where they intended to go for screening were considered to have intentions to screen.

Independent variables

We measured cervical cancer knowledge using a 20-item scale consisting of four constructs: risk factors (six items); signs and symptoms (eight items); eligibility for screening (5 items); and routine cervical cancer screening recommendations (one item). A previous study in Eastern Uganda considered all women who scored above the average for 20-point possible answers to be more knowledgeable, while those who scored below the average were considered to have less knowledge [ 37 ]. For this study, we considered those who scored in the 75th percentile to be knowledgeable and those whose scores fell below the 75th percentile to be less knowledgeable. The other independent variables considered were age (categorized into 20–29, 30–39, 40–49 and 50–59); marital status (single, cohabiting/married, commercial Sex Worker/divorced/widowed); work status (employed, student/not-working); regular income (yes and no); highest level of education (A-level+, O-level, PLE, none); residence (rural area, town board, town council); family cancer history (don’t know, no, and yes); common source of health information (social media/FM radio, TV, CATs, health worker, and any other); perceived self-risk for cervical cancer (yes, no, I do not know); fear of getting diagnosed with cancer (yes, no, I do not know, refused to answer); and fear of the cervical cancer screening procedure (yes, no, I do not know, refused to answer).

Statistical analysis

Descriptive statistics were reported using frequency distributions of the participant characteristics at individual level. Bivariate and multivariable analyses were conducted using generalized linear modelling with modified Poisson regression; prevalence ratios (PR) instead of odds ratios were used because of the high prevalence of intention to screen for cervical cancer [ 38 , 39 ]. We built our final analytic model using backwards elimination, where only variables with a p -value ≤ 0.2 were considered for the adjustment stage to determine the factors independently associated with intention to screen for cervical cancer. Collinearity between independent variables was assessed using pairwise correlation analysis. Data were analysed using Stata/SE 17, and statistical significance was considered at p  < 0.05; 95% confidence intervals are reported.

Baseline characteristics of the participants

Half of the participants had intentions to be screened for cervical cancer. In addition, half the participants were aged 20–29 years, and nearly three-quarters (73%) were married or cohabiting, as shown in Table  1 below. There was almost an equal distribution in residence for rural, town board and town council. Three-quarters (76%) of the participants reported no family history of cervical cancer, and 46% considered themselves at risk of cervical cancer. Half (52.5%) of the participants had full-time jobs, 62% had a regular source of income, and only 12.8% had more than A-level (high school equivalent) education.

CATs were mentioned as the main source of health information for nearly half (49.8%); more than half (64%) also feared being diagnosed with cervical cancer, while 40.9% feared the screening procedure. Approximately one-quarter (26.5%) had moderate knowledge about cervical cancer.

Correlates of intention to screen for cervical cancer among adult women

At the bivariate level, knowledge of cervical cancer, residence, common source of health information and perceived risk of getting cervical cancer were associated with intention to screen for cervical cancer. After adjusting for potential confounders, only participants’ residence, common source of health information and perceived risk of getting cervical cancer were independently associated with intention to screen for cervical cancer. Participants who resided in the town council were 44% more likely to have intentions to screen for cervical cancer compared to those who lived in the rural areas (see Table  2 ).

Compared to participants who mentioned FM radio or social media as their main source of health information, those who mentioned television were 48% less likely to have intentions to screen for cervical cancer, while those who mentioned CATs/health workers were 36% less likely to have intentions to screen for cervical cancer. Participants who perceived themselves to be at risk of cervical cancer were 74% more likely to have intentions to screen compared to those who did not. The full results are presented in Table  2 .

This study assessed the correlates of intention to screen for cervical cancer in Kyotera district, Central Uganda. We found 50.2% of the participants had intentions of being screened for cervical cancer. This prevalence was slightly lower than the 63% reported in a neighbouring district of Masaka in 2013 [ 32 ] and the 61% reported in rural Indonesia in 2016 [ 40 ]. However, it is higher than the prevalence of 45.3% in Ethiopia in 2017 [ 41 ]. Given the length of time since these prior estimates, we classify the prevalence of intention to screen in our study as suboptimal; we would anticipate increasing rates of screening over time and that there is a gap in cervical cancer related health promotion. As such, additional efforts are needed to promote cervical cancer screening in this area.

Participants who resided in the town council were 44% more likely to have intentions to screen for cervical cancer compared to those who lived in rural areas. This could be attributed to geographic proximity to health services located in urban areas, as well as the potential differences in income between urban and rural areas. Women who lived in urban and semiurban areas in Eastern Uganda were four times and two times more likely to have high knowledge about cervical cancer than their rural counterparts, respectively [ 37 ]. Our findings and others indicate there is a disparity in intentions to screen, which likely translates to differences in screening uptake. More equitable approaches to service delivery are warranted, including increased funding to support health education and cervical cancer screening promotion.

CATs were mentioned as the main source of health information for 49.8% of participants, compared to only 8.4% reported health workers as their main source of health information. The proportion that reported health workers was lower than the 15.1% reported from health facilities in Eastern Uganda [ 37 ]. Another study conducted in a neighbouring district had found that women who had discussions on cervical cancer with health care providers reported more intentions to screen for cervical cancer [ 32 ]. Even among women in Thailand, having received a recommendation from health care providers was associated with decisions to attend cervical cancer screening [ 42 ]. As such, integrating cervical cancer screening into health workers education packages and disseminating information via CATs may be effective health communication delivery mechanisms, where available. This may be feasible in Kyotera since a previous study found that the majority of the health workers believed CATs were accessible and easier to communicate on health issues; however, fewer than 20% used them [ 43 ].

We found only 46% of the sample considered themselves at risk of cervical cancer compared to 76.0% who perceived themselves to be at risk of cervical cancer in another study in Eastern Uganda, as reported in 2017 [ 44 ]. Risk perceptions were identified to be particularly important since those who perceived themselves to be at risk of cervical cancer were 74% more likely to have intentions to screen compared to those who did not; these findings align with prior reports in a neighbouring district [ 32 ]. Relatedly, a family history of cervical cancer was not associated with higher intentions of screening in this study, but it was reported to be associated among women in rural areas of Indonesia [ 40 ]; further research is needed to identify potential differences between family history and the impact on cancer screening. Multiple approaches for conducting effective Health education should be strengthened including use of print and interpersonal communication, as this could to help increase risk perception, increase intentions for screening, and ultimately aid in increasing uptake of cervical cancer screening.

Although being knowledgeable about cervical cancer was not associated with intentions to screen for cervical cancer after adjustment, it is important to note that only 26.5% of the participants had moderate knowledge (> 75th percentile) about cervical cancer. Increasing knowledge about cervical cancer is a critical area for further intervention given its importance in decision-making; yet it is likely this factor significant at the bivariate level was no longer significant after adjustment because of potential correlation with other social determinants of health (e.g., rurality). Knowledge has been reported to be associated with intention to undergo Pap smear testing in rural areas of Indonesia [ 40 ]. Therefore, improving knowledge about cervical cancer literacy could improve screening uptake.

Over half of the participants feared being diagnosed with cervical cancer, while 40.9% feared the screening procedure. Although these fears were not associated with intentions to screen for cervical cancer after adjustment, they could remain potential barriers to screening. These findings are inconsistent with previous studies conducting in neighbouring districts, Thailand, and Ethiopia [ 32 , 42 , 45 ]. Additional qualitative research could help identify nuance in these reports and is recommended; decreasing barriers to screening and managing a diagnosis are important to support patients in cancer prevention.

Strengths and limitations

We applied approaches to maximize the validity of the findings of this study. First, we assessed the outcome variable with more than a single question to only consider those who indicated the intention as well as when and where they would go for screening as those with intention to minimize social desirability bias. The district-wide sampling and the rich distribution of participants by age are other strengths of this study and are key to representativeness and thus generalizability of the study findings across the district and similar contexts. Despite these, some limitations are acknowledged. First, there could have been some people who still indicated intentions without actual intentions. In addition, the inherent limitations of cross-sectional study design including recall and difficulties with self-reporting on other variables other than intention cannot miss acknowledgement.

In this study, we found only half of adult women sampled in the Kyotera district, Central Uganda, had intentions for cervical cancer screening, and only 46% considered themselves at risk of cervical cancer. Urban residence, risk perception, and CATs as a source of health information were associated with higher intentions to screen for cervical cancer. The urban-rural difference calls for equity in cervical cancer health education and service delivery. In addition to other communication channels, targeting health information sharing via CATS and interactive TV educational messages may help reach those with lower intentions to screen.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Singh D, et al. Global estimates of incidence and mortality of cervical cancer in 2020: a baseline analysis of the WHO Global Cervical Cancer Elimination Initiative. Lancet Global Health. 2023;11(2):e197–206.

Article   CAS   PubMed   Google Scholar  

World Health Organization. Accelerating the elimination of cervical cancer as a global public health problem. World Health Organization. Regional Office for South-East Asia; 2019.

WCRFI. Cervical cancer statistics. 2023 2022 [cited 2023 3/6/2023]; https://www.wcrf.org/cancer-trends/cervical-cancer-statistics/ .

Arbyn M, et al. Estimates of incidence and mortality of cervical cancer in 2018: a worldwide analysis. Lancet Global Health. 2020;8(2):e191–203.

Article   PubMed   Google Scholar  

WHO. Human papilloma virus (HPV) and cervical Cancer: fact sheet. World Health Organization; 2016.

Markowitz LE, et al. Human papillomavirus vaccination: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep. 2014;63(RR–05):1–30.

PubMed   Google Scholar  

de Sanjose S, et al. Human papillomavirus genotype attribution in invasive cervical cancer: a retrospective cross-sectional worldwide study. Lancet Oncol. 2010;11(11):1048–56.

ICO/IARC, Uganda Human Papillomavirus and related cancers, fact sheet 2023 (2023-03-10) . 2023, the Catalan Institute of Oncology (ICO) and the International Agency for Research on Cancer (IARC): Barcelona, Spain.

Ferlay J et al. GLOBOCAN 2012: Cancer incidence and mortality worldwide: IARC CancerBase no. 11. Lyon, France: International Agency for Research on Cancer . 2013.

Boyle P, Levin B. World cancer report 2008. IARC Press, International Agency for Research on Cancer; 2008.

Denny L, et al. Cervical cancer. Disease Control priorities, (volume 3): Cancer. World Bank; 2015. H. Gelband, Jha, P., Sankaranarayanan, R. and Horton, S., Editor.

Tate DR, Anderson RJ. Recrudescence of cervical dysplasia among women who are infected with the human immunodeficiency virus: a case-control analysis. Am J Obstet Gynecol. 2002;186(5):880–2.

Ghebre RG, et al. Cervical cancer control in HIV-infected women: past, present and future. Gynecologic Oncol Rep. 2017;21:101–8.

Article   Google Scholar  

Ginindza TG, Sartorius B. Projected cervical cancer incidence in Swaziland using three methods and local survey estimates. BMC Cancer. 2018;18:1–10.

Chibwesha C, et al. Estimating the burden of cervical disease among HIV-infected women accessing screening services in South Africa: a model-based analysis. South Afr Med J. 2018;108(3):235–9.

Article   CAS   Google Scholar  

Simonds HM, et al. Completion of and early response to chemoradiation among human immunodeficiency virus (HIV)-positive and HIV‐negative patients with locally advanced cervical carcinoma in South Africa. Cancer. 2012;118(11):2971–9.

MOH. The Uganda Population-based HIV Impact Assessment (UPHIA 2020–2021). Uganda: Kamapala, Uganda: Ministry of Health (MOH); 2022.

Google Scholar  

Nakisige C, Schwartz M, Ndira AO. Cervical cancer screening and treatment in Uganda. Gynecologic Oncol Rep. 2017;20:37–40.

Bruni L et al. ICO Information Centre on HPV and Cancer (HPV Information Centre) Human papillomavirus and related diseases in India. Summary Report, 2015: pp. 12–23.

Nakisige C, Schwartz M, Ndira AO. Cervical cancer screening and treatment in Uganda. Gynecol Oncol Rep. 2017;20:37–40.

Article   PubMed   PubMed Central   Google Scholar  

Saslow D, et al. American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J Clin. 2002;52(6):342–62.

May DS, et al. The national breast and cervical Cancer Early Detection Program: report on the first 4 years of mammography provided to medically underserved women. AJR Am J Roentgenol. 1998;170(1):97–104.

Peirson L, et al. Screening for cervical cancer: a systematic review and meta-analysis. Syst Reviews. 2013;2:35–35.

Landy R, et al. Impact of cervical screening on cervical cancer mortality: estimation using stage-specific results from a nested case–control study. Br J Cancer. 2016;115(9):1140–6.

Gakidou E, Nordhagen S, Obermeyer Z. Coverage of cervical cancer screening in 57 countries: low average levels and large inequalities. PLoS Med. 2008;5(6):e132.

Holowaty P, et al. Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst. 1999;91(3):252–8.

Aweke YH, Ayanto SY, Ersado TL. Knowledge, attitude and practice for cervical cancer prevention and control among women of childbearing age in Hossana Town, Hadiya Zone, Southern Ethiopia: community-based cross-sectional study. PLoS ONE. 2017;12(7):e0181415.

Nwankwo KC, et al. Knowledge attitudes and practices of cervical cancer screening among urban and rural Nigerian women: a call for education and mass screening. Eur J Cancer Care. 2011;20(3):362–7.

Sudenga SL, et al. Knowledge, attitudes, practices, and perceived risk of cervical cancer among Kenyan women: brief report. Int J Gynecol Cancer. 2013;23(5):895–9.

Ndejjo R, et al. Uptake of Cervical Cancer Screening and Associated Factors among women in rural Uganda: A Cross Sectional Study. PLoS ONE. 2016;11(2):e0149696.

Mupepi SC, Sampselle CM, Johnson TR. Knowledge, attitudes, and demographic factors influencing cervical cancer screening behavior of Zimbabwean women. J Womens Health (Larchmt). 2011;20(6):943–52.

Twinomujuni C, Nuwaha F, Babirye JN. Understanding the low level of Cervical Cancer Screening in Masaka Uganda using the ASE Model: A Community-based survey. PLoS ONE. 2015;10(6):e0128498.

Siegel RL et al. Cancer statistics, 2023 CA: A Cancer Journal for Clinicians, 2023. 73(1): pp. 17–48.

Kabanda R. Availability, use and efficacy of community audio towers in health communication for prevention of cervical cancer in the communities of Kyotera District, Uganda [Unpublished doctoral dissertation] . 2021, Uganda Martyrs University.

Serwadda D, et al. Slim disease: a new disease in Uganda and its association with HTLV-III infection. Lancet. 1985;2(8460):849–52.

UAC, Factsheet - facts on HIV and AIDS in Uganda 2021 (based on data ending 31st December 2020) . 2021, Uganda AIDS Commission Secretariat: Kampala, Uganda.

Mukama T, et al. Women’s knowledge and attitudes towards cervical cancer prevention: a cross sectional study in Eastern Uganda. BMC Womens Health. 2017;17(1):9–9.

Thompson ML, Myers JE, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done? Occup Environ Med. 1998;55(4):272.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Tamhane AR, et al. Prevalence odds ratio versus prevalence ratio: choice comes with consequences. Stat Med. 2016;35(30):5730–5.

Sumarmi S, et al. Factors associated with the intention to undergo pap smear testing in the rural areas of Indonesia: a health belief model. Reproductive Health. 2021;18(1):138.

Getahun T, Kaba M, T Derseh B. Intention to screen for cervical Cancer in Debre Berhan Town, Amhara Regional State, Ethiopia: application of theory of Planned Behavior. J Cancer Epidemiol. 2020;2020:3024578.

Budkaew J, Chumworathayi B. Factors associated with decisions to attend cervical cancer screening among women aged 30–60 years in Chatapadung Contracting Medical Unit, Thailand. Asian Pac J Cancer Prev. 2014;15(12):4903–7.

Kabanda R, et al. Utilisation of Community Audio Towers in Health Education for Prevention of Cervical Cancer by Health Workers in Kyotera District, Uganda: a cross-sectional study. Risk Manag Healthc Policy. 2021;14:3667–73.

Mukama T, et al. Women’s knowledge and attitudes towards cervical cancer prevention: a cross sectional study in Eastern Uganda. BMC Womens Health. 2017;17(1):1–8.

Getahun T, Kaba M, Derseh BT. Intention to Screen for Cervical Cancer in Debre Berhan Town, Amhara Regional State, Ethiopia: Application of Theory of Planned Behavior Journal of cancer epidemiology, 2020. 2020: pp. 3024578–3024578.

Download references

This study received no funding from any agency be it public, commercial, or not-for-profit.

Author information

Authors and affiliations.

Institute for Health and Equity, Medical College of Wisconsin, Milwaukee, WI, 53226, USA

Richard Kabanda, Anguzu Ronald & Kirsten M. M. Beyer

Uganda Martyrs University, Nkozi, Uganda

Arthur Kiconco

Ministry of Health, Kampala, Uganda

Department of Psychiatry and Behavioral Medicine, Medical College of Wisconsin, Milwaukee, WI, 53226, USA

Steven A. John

You can also search for this author in PubMed   Google Scholar

Contributions

RK conceptualized the study and collected data; AK contributed to study conceptualization, analyzed data and drafted the manuscript; AR, KMMB, SAJ, RK and AK all reviewed the manuscript.

Corresponding author

Correspondence to Richard Kabanda .

Ethics declarations

Ethics approval and consent to participate.

This study was approved by the Institutional Review Board (IRB) of Mildmay Uganda (#REC 0801–2020) and the Uganda National Council of Science and Technology (UNCST) (SS 5233). Administrative clearance was also sought from the Ministry of Health and Kyotera District Local Government. Participants provided written informed consent before being interviewed. Confidentiality and privacy were maintained throughout the research process.

Consent for publication

Not applicable.

Competing interests

The authors declare 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.

Kabanda, R., Kiconco, A., Ronald, A. et al. Correlates of intention to screen for cervical cancer among adult women in Kyotera District, Central Uganda: a community based cross-sectional study. BMC Women's Health 24 , 296 (2024). https://doi.org/10.1186/s12905-024-03129-5

Download citation

Received : 05 September 2023

Accepted : 03 May 2024

Published : 18 May 2024

DOI : https://doi.org/10.1186/s12905-024-03129-5

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

• Cervical cancer screening
• Health information source

BMC Women's Health

ISSN: 1472-6874

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Study to Understand Cervical Cancer Early Endpoints and Determinants (SUCCEED)

Researchers now know that human papillomavirus (HPV) infection is the necessary but not sufficient cause of cervical cancer. Cervical pathogenesis evolves as follows: normal cervical tissue, to oncogenic HPV infection, to precancer and then to invasive cancer. The majority of women with oncogenic HPV infections will not develop cancer, and most HPV infections, even those with associated cellular changes, regress in 1-2 years, probably eradicated or controlled by cellular immune response. Moreover, while invasive cancer and precancer are histologically well-defined, the histological classification of low-grade lesions, now better defined as HPV infection, is very heterogeneous and poorly reproducible. Identifying women at highest risk for cancer prior to neoplastic progression is therefore a challenge. At present, researchers are unable to predict with any accuracy which HPV infections will progress and which are among the majority that regress. It is therefore of etiologic interest and of public health benefit to develop a method for identifying the HPV-infected women at risk for progressing to precancer and invasion.

In order to develop an accurate and reproducible division of precursor lesions (HPV infection and precancer), researchers must gain knowledge about the molecular distinctions at each progressive disease state. The goal of SUCCEED is to comprehensively assess biomarkers of risk for progressive cervical neoplasia, and thus develop biomarkers that can distinguish those at highest risk of cervical cancer from those with benign infection.

Investigators in DCEG have implemented a cross-sectional study to develop a comprehensive list of potential biomarkers by examining biospecimens and cervical tissues from over 2,500 women with HPV infection, precancer, and cancer. Several HPV-based markers, including HPV genotyping and detection of HPV integration have been evaluated in SUCCEED. A major focus is the evaluation of gene expression profiles to gain an accurate and comprehensive in vivo picture of cervical carcinogenesis. Investigators will validate the most promising identified candidate biomarkers in a prospective design by assessing their predictive values for key outcomes related to progression (HPV persistence, diagnosis of precancer) or non-progression (HPV clearance).

View publications related to SUCCEED .

For more information, contact Nicolas Wentzensen .

Clinical Epidemiology Unit Research Areas 

• Open access
• Published: 15 May 2024

Efficacy and safety of different chemotherapy regimens concurrent with radiotherapy in the treatment of locally advanced cervical cancer

• Yaping Wu 1 ,
• Peng Jiang 1 ,
• Zhiying Chen 1 ,
• Bin Dong 1 &
• Yongchun Zhang 1  

BMC Cancer volume  24 , Article number:  589 ( 2024 ) Cite this article

Metrics details

Evaluate the efficacy and safety of different chemotherapy regimens concurrent with radiotherapy in treating locally advanced cervical cancer (LACC).

Retrospective data was collected from LACC patients who were treated at our institution. These patients were categorized into three groups: the single-agent cisplatin (DDP) chemoradiotherapy group, the paclitaxel plus cisplatin (TP) chemoradiotherapy group, and the nanoparticle albumin-bound (nab-) paclitaxel combined with cisplatin (nPP) chemoradiotherapy group. The primary endpoints were overall survival (OS) and progression-free survival (PFS) and the secondary endpoints were objective response rate (ORR) and incidence of adverse events (AEs).

A total of 124 patients were enrolled (32 in the DDP group, 41 in the TP group, and 51 in the nPP group). There were differences in OS ( P  = 0.041, HR 0.527, 95% CI 0.314–0.884) and PFS ( P  = 0.003, HR 0.517, 95% CI 0.343–0.779) between the three groups. Notably, the 2-year OS rate was significantly higher in the nPP group compared to the DDP group (92.2% vs. 85.4%, P  = 0.012). The 2-year PFS rates showed a marked increase in the TP group (78.0% vs. 59.4%, P  = 0.048) and the nPP group (88.2% vs. 59.4%, P  = 0.001) relative to the DPP group, with multiple comparisons indicating that the 2-year PFS rate was significantly superior in the nPP group versus the DDP group (88.2% vs. 59.4%, P  = 0.001). Moreover, the ORR was also significantly higher in the nPP group than in the DDP group ( P  = 0.013); and no statistically significant differences were found in the incidence of AEs among the groups ( P  > 0.05).

Conclusions

In LACC treatment, the two cisplatin-based doublet chemotherapy regimens are associated with better outcomes, with the nab-paclitaxel plus cisplatin regimen showing better efficacy than the paclitaxel plus cisplatin regimen. Furthermore, the AEs associated with these regimens were deemed tolerable. These findings could provide a reference for the clinical treatment of LACC. However, further prospective studies are needed to verify it.

Peer Review reports

Cervical cancer is the fourth most frequently diagnosed cancer and the fourth leading cause of cancer death in women, representing a major global health challenge. According to statistics, there are about 600,000 new cases of cervical cancer worldwide every year, and nearly 90% of them occur in low-income and middle-income countries [ 1 , 2 ]. Although early-stage cervical cancer is often curable, 40-50% of patients are diagnosed at the locally advanced stage [International Federation of Obstetricians and Gynecologists (FIGO2009/2018), IB2-IVA stage/IB3-IVA stage] [ 3 , 4 ]. Cisplatin-based concurrent chemoradiotherapy combined with brachytherapy is the standard treatment for LACC. After completion of CCRT, the 5-year OS of patients is about 65–70%, and nearly 40% have recurrence or metastasis [ 5 , 6 ]. Reducing distant metastasis and improving the long-term survival of LACC patients is still a clinical challenge. Many studies have been conducted at home and abroad to improve the efficacy of concurrent chemotherapy regimens in treating LACC, but which regimen is better is not conclusive. In previous studies, paclitaxel combined with platinum has shown potent activity and is a commonly used combination chemotherapy regimen; however, the increased incidence of adverse reactions reduced patient compliance [ 7 , 8 , 9 ]. Nab-paclitaxel is an albumin-bound, solvent-free form of paclitaxel in nanoparticles, which is now widely used for the treatment of breast cancer, locally advanced or metastatic non-small cell lung cancer, and ovarian cancer [ 10 ]. Moreover, the efficacy in drug-resistant, metastatic, and recurrent cervical cancer has also been demonstrated [ 11 ]. However, there are few reports on its use in treating LACC. In this study, nab-paclitaxel was used in CCRT for LACC and investigated the efficacy and safety of different chemotherapy regimens concurrent radiotherapy for LACC.

Patient characteristics

The clinical data of 124 patients with LACC who underwent CCRT and intracavitary brachytherapy (ICBT) in our hospital from March 2018 to January 2021 were retrospectively analyzed. The patients were divided into three groups according to different treatment protocols: 32 patients in the DDP group, 41 in the TP group and 51 in the nPP group.

The inclusion criteria were as follows: (I) Diagnosed with cervical cancer through histopathological examination, including cervical squamous cell carcinoma and cervical adenocarcinoma; (II) stage IB3–IVA; (III) no previous surgical treatment; (IV) no previous history of radiotherapy; (V) no bone marrow suppression, and liver and kidney functions were generally normal.

The exclusion criteria were: (I) Presence of other cancer types; (II) pregnant or lactating women; (III) incomplete data on clinical treatment; (IV) severe complications or severe infection in important organs such as heart and lung; (V) change of chemotherapy regimen during the course of treatment.

Gynecological examination, magnetic resonance imaging (MRI), positron emission computed tomography (PET-CT), computed tomography (CT), or Ultrasound were used to evaluate the local tumor, lymph node status, and tumor metastasis. The maximum diameter of a tumor is measured based on CT or MRI scan results prior to treatment. There was no statistical difference in baseline data such as age, mass size, pathological type, clinical stage, and initial hemoglobin among the three groups ( P  > 0.05). (Table  1 )

Following the careful exclusion of contraindications to chemoradiotherapy, patients proceeded with radiation therapy localization and treatment planning, concurrently initiating a cycle of systemic chemotherapy. In the DDP group, patients were administered a single dose of 75 mg/m² of single-agent cisplatin, repeated on a three-week interval. For the TP group, patients received a combination of paclitaxel and cisplatin, with a single dose consisting of 135 mg/m² of paclitaxel and 75 mg/m² of cisplatin, also repeated at three-week intervals. As for the nPP group, patients were treated with a combination of nab-paclitaxel and cisplatin, receiving a single dose of 200 mg/m² of nab-paclitaxel and 75 mg/m² of cisplatin, again on a three-weekly repetition schedule. Throughout the radiotherapy phase, chemotherapy was synchronously administered once every three weeks, approximating to two complete cycles of concurrent chemotherapy. Within one week following radiotherapy, an additional 0 to 1 cycle of chemotherapy was given as needed, culminating in an overall total of 3 to 4 chemotherapy cycles for each patient.

Radiotherapy (RT) included external beam radiation therapy (EBRT) and intracavitary brachytherapy. In EBRT, Image-guided Intensity-Modulated Radiation Therapy was utilized. The patient was positioned in the supine posture secured with thermoplastic immobilization molds. Enhanced CT scans were performed at a slice thickness of 5 millimeters, with the superior boundary at the level of the upper margin of the tenth thoracic vertebra and the inferior boundary approximately 10 centimeters below the ischial tuberosity. The clinical target volume (CTV) included the primary cervical lesion area and the lymph node area, encompassing the entire cervix, parametria, uterine corpus, partial or complete vagina, as well as the draining lymph nodes of the obturator, internal iliac, external iliac, common iliac, and presacral regions. Depending on the specific case, the inguinal and para-aortic lymph node drainage areas may or may not be included. The planning target volume (PTV) was defined as an expansion of 3–5 millimeters from the CTV to establish the planning treatment volume. The dose of CTV was 45.0 ~ 50 Gy/25-27fraction (f), for patients with positive parametrial involvement, a simultaneous integrated boost up to 58–62 Gy/25-27f was administered locally. Once the primary tumor shrinks to less than 3 cm in diameter, typically after about 15 fractions of EBRT or upon completion of EBRT, three-dimensional image-guided high-dose-rate brachytherapy was conducted under CT guidance, with a dose of 24–26 Gy/4f. The equivalent dose of 2 Gy (EQD2) of ERBT combined with ICBT was 82–88 Gy. The total treatment course spanned approximately 7 weeks. The treatment flow diagram of the three groups is shown in Fig.  1 .

Chemotherapy dose was maintained or reduced after giving symptomatic treatment to patients who developed hepatic dysfunction, severe gastrointestinal reactions, or bone marrow suppression after treatment. If the patient could not tolerate the adverse effects of the treatment, the chemotherapy was stopped.

Treatment flow diagram. CCRT, concurrent chemoradiotherapy; EBRT, external beam radiation therapy; CTV, clinical target volume; CTV-n, gross target volume of lymph nodes; ICBT, intracavitary brachytherapy

Evaluation of tumor response

Short-term efficacy: Efficacy was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST 1.1). The short-term outcomes were classified into complete response (CR), partial response (PR), stable disease (SD), progressive disease (PD), and objective response rate (ORR) = (CR + PR)/total × 100%.

Long-term efficacy: Patients were subjected to routine medical reviews and telephonic check-ups every 3–6 months post-treatment. All subjects received a minimum of two years’ follow-up, or until their demise, whichever occurred earlier. Overall Survival (OS) was defined as the duration from the initiation of treatment until death due to any cause or until the designated endpoint of follow-up. Progressive Free Survival (PFS) was determined as the interval between the start of treatment and the first instance of disease progression, death from any cause, or the end of the follow-up period.

AEs: AEs were assessed according to the Common Terminology Criteria for Adverse Events version 5.0 (CTCAE v5.0), mainly including bone marrow suppression, gastrointestinal reactions, allergic reaction, peripheral neurotoxicity, radiation enteritis, radiation cystitis and hepatic impairment and renal impairment. with a focus on key manifestations such as bone marrow suppression, gastrointestinal adverse reactions, allergic responses, peripheral neuropathy, radiation-induced enteritis, radiation cystitis, as well as hepatic and renal dysfunctions.

Statistical analysis

SPSS 26.0 (IBM, USA) was used for the statistical analysis. Measured data were expressed as mean ± standard deviation. One-way analysis of variance (ANOVA) was used for comparisons between groups. The quantitative index was converted into frequency and percentage. The chi-square test was used for comparisons among the three groups. OS and PFS were estimated using the Kaplan–Meier method, and the Log-rank test compared differences in survival curves. P  < 0.05 (two-tailed) was considered significant. Using Bonferroni adjustment method (test level a = 0.05/3 = 0.017, P <0.017) made multiple comparisons between the three groups.

All patients completed EBRT and ICRT, and two patients in the TP group discontinued chemotherapy after one cycle due to adverse reactions. The differences in the number of chemotherapy cycles and EBRT, ICRT, and EQD2 doses were not statistically significant among the three groups. (Table  2 )

Short-term efficacy

In the DDP group, 17 cases (53.1%) achieved CR, 6 cases (18.8%) achieved PR, and ORR was 71.9% (23/32). In the TP group, 25 cases (61.0%) achieved CR, 10 cases (24.4%) achieved PR, and ORR was 85.4% (35/41). In the nPP group, 37 cases (72.5%) achieved CR, 10 cases (19.6%) achieved PR, and ORR was 92.2% (47/51). There was a difference in ORR between the three groups (71.9% vs. 85.4% vs. 92.2%, P  = 0.044). Multiple comparisons revealed that the ORR was significantly higher in the nPP group than in the DDP group (71.9% vs. 92.2%, P  = 0.013). In contrast, no statistically significant difference was seen in the TP group compared with the DDP group (85.4% vs. 71.9%, P  = 0.157). (Table  3 )

Long-term efficacy

Survival follow-up was until January 2023, with a median follow-up of 30 months (10–49 months). The 1- and 2-year OS rates were 93.8% vs. 97.6% vs. 98.0% and 78.1% vs. 85.4% vs. 92.2% in the DDP, TP, and nPP groups, respectively, with statistically significant differences ( P  = 0.041, log-rank test). 1- and 2-year PFS rates were 75.0% vs. 87.8% vs. 92.2% and 59.4% vs. 78.0% vs. 88.2% in the DDP, TP, and nPP groups, respectively, with statistically significant differences ( P  = 0.003, log-rank test). (Fig.  2 )

The 2-year OS rate in the nPP group was significantly higher than in the DDP group (92.2% vs. 78.1%, P  = 0.012, HR = 0.525, 95%CI = 0.307–0.899). There was no significant difference in the 2-year OS rate between the TP group and the DDP group (85.4% vs. 78.1%, P  = 0.237). The 2-year PFS rates were higher in both the TP groups (78.0% vs. 59.4%, P  = 0.048) and the nPP groups (88.2% vs. 59.4%, P  = 0.001) than in the DDP group, but multiple comparisons suggested that the 2-year PFS rate was higher in the nPP group compared with the DDP group (88.2% vs. 59.4%, P  = 0.001, HR = 0.525, 95% CI = 0.348–0.791). (Table  4 )

Kaplan-Meier analysis of the three groups, and percentages of 1-year and 2-year PFS and OS for patients in the three group. PFS, progression-free survival; OS, overall survival

AEs were tolerable in all three groups, and the most common AEs were myelosuppression and gastrointestinal reactions. There was no significant difference in AEs between the groups ( P  > 0.05). (Table  5 )

National Comprehensive Cancer Network (NCCN) clinical practice guidelines for cervical cancer have recommended cisplatin-based CCRT as standard treatment for LACC [ 12 ] after five large-sample randomized controlled trials conducted by Gynecologic Oncology Group (GOG), Radiation Therapy Oncology Group (RTOG), and Southwest Oncology Group (SWOG) in the United States reported that concurrent radiotherapy could improve survival in cervical cancer [ 13 ]. However, many patients have residual tumors after treatment, leading to tumor progression or death.

To enhance the therapeutic effect and prognosis of cervical cancer, a variety of novel treatment strategies are actively being explored. The development of molecular targeted therapies has opened up new avenues. The GOG 240 Phase III clinical trial combined bevacizumab with chemotherapy for Stage IVB/recurrent/refractory cervical cancer patients, significantly improving their OS rate [ 14 ]. And based on this study, the Food and Drug Administration (FDA) approved bevacizumab for the first-line treatment of recalcitrant/recurrent/metastatic cervical cancer. In the RTOG 0417 phase II trial [ 15 ], bevacizumab combined with chemoradiotherapy was applied to stage IB to IIIB cervical cancer patients. The 3-year OS and disease-free survival (DFS) were 81.3% and 68.7%, respectively, while the incidence of grade 3 and 4 adverse events was 26.5% and 10.2%, respectively. However, since this treatment did not demonstrate superiority over historical controls with standard cisplatin chemoradiotherapy, further investigation with bevacizumab was not pursued. Bevacizumab increased the incidence of AEs such as hypertension, thrombosis, and gastrointestinal fistula while improving its efficacy [ 16 ], and its use has been greatly limited as drug resistance has developed. Immune checkpoint inhibitors are currently a hot spot of research in various oncology therapies. PD-1/PD-L1 inhibitors have shown remarkable efficacy in treating recurrent/metastatic cervical cancer [ 17 , 18 ], and the FDA has approved pembrolizumab and nivolumab for treating recurrent/metastatic cervical cancer. Regarding immunotherapy for LACC, the CALLA Phase III clinical trial combined durvalumab or placebo with concurrent chemoradiotherapy in treating LACC patients [ 19 ], however, the results indicated that durvalumab combined with CCRT failed to improve the PFS of LACC patients. The KEYNOTE-A18 Phase III trial results suggested that combining immunotherapy with CCRT might achieve synergistic effects in patients with locally advanced cervical cancer. While targeted therapy and immunotherapy remain limited in their application for LACC, cytotoxic chemotherapy remains indispensable in the treatment of LACC [ 20 ].

Petrelli et al. [ 21 ] conducted a meta-analysis of 1500 patients, demonstrating that the combination of CCRT with a cisplatin-based dual agent significantly improved the OS ( P  = 0.0002, OR 0.65, 95%CI 0.51–0.81) and PFS ( P  = 0.006, OR 0.71, 95% CI 0.55–0.91) compared to weekly cisplatin single-agent CCRT. Similarly, another meta-analysis published by Ma et al. [ 22 ] showed that CCRT with a platinum-based doublet significantly improved OS ( P  = 0.01, HR 0.75, 95% CI 0.60–0.94) and PFS ( P  = 0.01, HR 0.78, 95% CI 0.65–0.94) compared with CCRT combined with cisplatin monotherapy. Consistent with past research, the combination of paclitaxel and platinum is often the favored chemotherapy regimen for treating LACC considering its impact on ORR and PFS [ 8 ]. However, the traditional formulation of paclitaxel is associated with reduced patient adherence due to its high frequency of AEs, such as myelosuppression and allergic reactions, thereby impacting overall treatment outcomes [ 23 ], and in this study, two patients in the TP group discontinued concurrent chemotherapy due to AEs of chemotherapy. Nab-paclitaxel, a 130 nano-meter albumin-bound paclitaxel complex, binds to specific receptors on the surface of tumor vascular endothelial cells, facilitating the uptake of paclitaxel into tumor cells via albumin-mediated endocytosis. It increases the concentration of paclitaxel in the tumor stroma, aggregates more anti-tumor drugs to the lesion, and ultimately enhances treatment outcomes. Furthermore, the absence of the requirement for co-solvents and desensitization pretreatment renders nab-paclitaxel more convenient and safer to administer [ 24 ].

Our study used nab-paclitaxel in CCRT to compare the efficacy and safety of chemotherapy regimens of single-agent cisplatin, paclitaxel combined with cisplatin, and nab-paclitaxel combined with cisplatin concurrent radiotherapy for LACC. The results showed differences in OS ( P  = 0.041, HR 0.527, 95%CI 0.314–0.884) and PFS ( P  = 0.003, HR 0.517, 95%CI 0.343–0.779) rates among the three groups. The 2-year OS rate was higher in patients in the nPP group than in the DDP group (92.2% vs. 78.1, P  = 0.012 < 0.017, HR 0.525, 95% CI 0.307–0.899). However, there was no significant difference in the 2-year OS rate between the TP group and the DDP group (85.4 vs. 78.1%, P  = 0.237), which may be related to the short follow-up time. An extended follow-up duration may be able to reveal a statistically discernible disparity in OS rates between the two groups. The 2-year PFS rate was higher in both the TP and nPP groups than in the DDP group (78.0% vs. 59.4%, P  = 0.048; 88.2% vs. 59.4%, P  = 0.001), and multiple comparisons suggested that the 2-year PFS rate was significantly higher in the nPP group compared with the TP group ( P  = 0.001 < 0.017, HR 0.525, 95% CI 0.348–0.791). There were significant differences in ORR among the three groups (71.9% vs. 85.4% vs. 92.2%, P  = 0.044), and multiple comparisons suggested that the ORR in the nPP group was significantly higher than that in the DDP group (92.2% vs. 71.9%, P  = 0.013 < 0.017). However, there was no significant difference in ORR between the TP group and the DDP group (85.4% vs. 71.9%, P  = 0.157); increasing the sample size may be able to observe a statistical difference between the two groups. The cisplatin-alone group in our study seemed to have done much worse than what was reported in the EMBRACE I trial [ 25 ], which may be related to the higher clinical stage of the patients in the groups we enrolled. Increasing the number of patients could further refine our study. The incidence of AEs was higher in the TP and nPP groups than in the DDP group. However, the differences were not statistically significant ( P  > 0.05), indicating that the safety of combining paclitaxel or nab-paclitaxel with single-agent cisplatin was tolerable.

A phase II clinical study published by the GOG in 2012 investigated the efficacy and safety of nab-paclitaxel monotherapy in patients with advanced and recurrent cervical cancer. It showed that nab-paclitaxel has considerable activity and moderate toxicity in treating resistant, metastatic, and recurrent cervical cancer [ 11 ]. Li et al. [ 26 ] employed a combination of nab-paclitaxel and nedaplatin for patients with advanced and recurrent cervical cancer. Their findings showed an ORR of 50.0%, an OS of 16.6 months, a PFS of 9.1 months, and a Grade 3 incidence of thrombocytopenia and anemia at 7.4% and 18.5%, respectively. And no cases of hypersensitivity reactions were reported, suggesting that nab-paclitaxel presents encouraging efficacy and acceptable toxicity profiles. Currently, nab-paclitaxel is approved as a second-line treatment option for patients with recurrent/metastatic cervical cancer. Yu et al. [ 27 ] investigated the effect of neoadjuvant chemotherapy consisting of nab-paclitaxel and platinum (NACT-nPP) in patients with LACC. It showed that 72 (92.3%) patients in the NACT-nPP group and 96 (82.1%) patients in the control group achieved CR ( P  = 0.042). Grade 3 or higher acute hematologic AEs were manageable in the NACT-nPP group (46.2%, 36/78), demonstrating the efficacy and safety of nab-paclitaxel neoadjuvant therapy combined with CCRT for LACC. However, despite these findings, there is still debate about whether neoadjuvant therapy confers tangible benefits to LACC patients and currently, most LACC patients continue to receive primary treatment through CCRT [ 28 ].

This study compared the efficacy and safety of three chemotherapy regimens of single-agent cisplatin, paclitaxel plus cisplatin, and nab-paclitaxel plus cisplatin combined with radiotherapy in the treatment of LACC. The results showed that the two cisplatin-based double-agent chemotherapy regimens were associated with improved outcomes than the single-agent cisplatin regimen, and the AEs were tolerable. Compared with traditional paclitaxel, albumin-bound paclitaxel was associated with improved outcomes in OS, PFS, and ORR of LACC, along with improved treatment compliance of patients. Nonetheless, it’s crucial to acknowledge that this study was a single-center retrospective study, the retrospective nature could introduce potential selection biases. Additionally, the sample size was relatively small, the follow-up period was comparatively short, and at data cutoff, outcomes for many patients remained unknown. Consequently, further validation through larger-scale, prospective studies is required to substantiate these findings.

For LACC, nPP regimen concurrent with radiotherapy appears to offer superior benefits compared to TP regimen and DPP regimen and adverse reactions are tolerable. The findings from this study can provide valuable reference for the clinical treatment of LACC.

Data availability

The datasets generated and/or analysed during the current study are not publicly available due to the experiment is still in progress but are available from the corresponding author on reasonable request.

Abbreviations

Locally advanced cervical cancer

Single-agent cisplatin

Paclitaxel combined with cisplatin

Nanoparticle albumin-bound (nab-) paclitaxel combined with cisplatin

Overall survival

Progression-free survival

Objective response rate

Adverse events

Concurrent chemoradiotherapy

Intracavitary brachytherapy

Magnetic resonance imaging

Positron emission computed tomography

Computed tomography

Radiotherapy

External beam radiation therapy

American Radiation Therapy Oncology Collaborative Group

Gross target volume

Clinical target volume

Intensity-modulated radiation therapy

Equivalent dose of 2 Gy

Organs at risk

Response Evaluation Criteria in Solid Tumors

Complete response

Partial response

Stable disease

Progressive disease

One-way analysis of variance

National Comprehensive Cancer Network

Gynecologic Oncology Group

Radiation Therapy Oncology Group

Southwest Oncology Group

Disease-free survival

Sung H, et al. Global Cancer statistics 2020: GLOBOCAN estimates of incidence and Mortality Worldwide for 36 cancers in 185 countries. Cancer J Clin. 2021;71:209–49. https://doi.org/10.3322/caac.21660 .

Article   CAS   Google Scholar  

Cohen PA, Jhingran A, Oaknin A, Denny L. Cervical cancer. Lancet (London England). 2019;393:169–82. https://doi.org/10.1016/s0140-6736(18)32470-x .

Article   PubMed   Google Scholar  

Gennigens C, De Cuypere M, Hermesse J, Kridelka F, Jerusalem G. Optimal treatment in locally advanced cervical cancer. Expert Rev Anticancer Ther. 2021;21:657–71. https://doi.org/10.1080/14737140.2021.1879646 .

Article   CAS   PubMed   Google Scholar  

Naga Ch P, Gurram L, Chopra S, Mahantshetty U. The management of locally advanced cervical cancer. Curr Opin Oncol. 2018;30:323–9. https://doi.org/10.1097/cco.0000000000000471 .

Green JA, et al. Survival and recurrence after concomitant chemotherapy and radiotherapy for cancer of the uterine cervix: a systematic review and meta-analysis. Lancet (London England). 2001;358:781–6. https://doi.org/10.1016/s0140-6736(01)05965-7 .

da Costa SCS, et al. Neoadjuvant Chemotherapy with Cisplatin and Gemcitabine followed by Chemoradiation Versus Chemoradiation for locally Advanced Cervical Cancer: a randomized phase II trial. J Clin Oncology: Official J Am Soc Clin Oncol. 2019;37:3124–31. https://doi.org/10.1200/jco.19.00674 .

McGuire WP, Blessing JA, Moore D, Lentz SS, Photopulos G. Paclitaxel has moderate activity in squamous cervix cancer. A Gynecologic Oncology Group study. J Clin Oncology: Official J Am Soc Clin Oncol. 1996;14:792–5. https://doi.org/10.1200/jco.1996.14.3.792 .

Della Corte L, et al. Advances in paclitaxel combinations for treating cervical cancer. Expert Opin Pharmacother. 2020;21:663–77. https://doi.org/10.1080/14656566.2020.1724284 .

Minion LE, Tewari KS. Cervical cancer - state of the science: from angiogenesis blockade to checkpoint inhibition. Gynecol Oncol. 2018;148:609–21. https://doi.org/10.1016/j.ygyno.2018.01.009 .

Article   PubMed   PubMed Central   Google Scholar  

10 Kundranda MN, Niu J. Albumin-bound paclitaxel in solid tumors: clinical development and future directions. Drug Des Devel Ther. 2015;9:3767–77. https://doi.org/10.2147/dddt.S88023 .

Article   CAS   PubMed   PubMed Central   Google Scholar  

Alberts DS, et al. Phase II trial of nab-paclitaxel in the treatment of recurrent or persistent advanced cervix cancer: a gynecologic oncology group study. Gynecol Oncol. 2012;127:451–5. https://doi.org/10.1016/j.ygyno.2012.09.008 .

Hsu HC, Li X, Curtin JP, Goldberg JD, Schiff PB. Surveillance epidemiology and end results analysis demonstrates improvement in overall survival for cervical cancer patients treated in the era of concurrent chemoradiotherapy. Front Oncol. 2015;5:81. https://doi.org/10.3389/fonc.2015.00081 .

Nagy, V.et al. Radiotherapy versus concurrent 5-day cisplatin and radiotherapy in locally advanced cervical carcinoma. Long-term results of a phase III randomized trial. Strahlentherapie und Onkologie: Organ der Deutschen Rontgengesellschaft … et al] 185, 177–183, doi:10.1007/s00066-009-1893-z (2009).

Monk BJ, et al. Phase III trial of four cisplatin-containing doublet combinations in stage IVB, recurrent, or persistent cervical carcinoma: a Gynecologic Oncology Group study. J Clin Oncology: Official J Am Soc Clin Oncol. 2009;27:4649–55. https://doi.org/10.1200/jco.2009.21.8909 .

15 Schefter T, et al. RTOG 0417: efficacy of bevacizumab in combination with definitive radiation therapy and cisplatin chemotherapy in untreated patients with locally advanced cervical carcinoma. Int J Radiat Oncol Biol Phys. 2014;88:101–5. https://doi.org/10.1016/j.ijrobp.2013.10.022 .

Tewari KS, et al. Improved survival with bevacizumab in advanced cervical cancer. N Engl J Med. 2014;370:734–43. https://doi.org/10.1056/NEJMoa1309748 .

Colombo N, et al. Pembrolizumab for Persistent, recurrent, or metastatic cervical Cancer. N Engl J Med. 2021;385:1856–67. https://doi.org/10.1056/NEJMoa2112435 .

Naumann RW, et al. Safety and Efficacy of Nivolumab Monotherapy in recurrent or metastatic cervical, vaginal, or Vulvar Carcinoma: results from the phase I/II CheckMate 358 trial. J Clin Oncology: Official J Am Soc Clin Oncol. 2019;37:2825–34. https://doi.org/10.1200/jco.19.00739 .

Mayadev J, et al. Efficacy and safety of concurrent and adjuvant durvalumab with chemoradiotherapy versus chemoradiotherapy alone in women with locally advanced cervical cancer: a phase III, randomized, double-blind, multicenter study. Int J Gynecol cancer: Official J Int Gynecol Cancer Soc. 2020;30:1065–70. https://doi.org/10.1136/ijgc-2019-001135 .

Article   Google Scholar  

Lorusso D, et al. Pembrolizumab or placebo with chemoradiotherapy followed by pembrolizumab or placebo for newly diagnosed, high-risk, locally advanced cervical cancer (ENGOT-cx11/GOG-3047/KEYNOTE-A18): a randomised, double-blind, phase 3 clinical trial. Lancet (London England). 2024;403:1341–50. https://doi.org/10.1016/s0140-6736(24)00317-9 .

Petrelli F, De Stefani A, Raspagliesi F, Lorusso D, Barni S. Radiotherapy with concurrent cisplatin-based doublet or weekly cisplatin for cervical cancer: a systematic review and meta-analysis. Gynecol Oncol. 2014;134:166–71. https://doi.org/10.1016/j.ygyno.2014.04.049 .

Ma S, et al. Platinum single-agent vs. platinum-based Doublet agent concurrent chemoradiotherapy for locally advanced cervical cancer: a meta-analysis of randomized controlled trials. Gynecol Oncol. 2019;154:246–52. https://doi.org/10.1016/j.ygyno.2019.04.013 .

Marupudi NI, et al. Paclitaxel: a review of adverse toxicities and novel delivery strategies. Exp Opin Drug Saf. 2007;6:609–21. https://doi.org/10.1517/14740338.6.5.609 .

Yardley DA. Nab-Paclitaxel mechanisms of action and delivery. J Controlled Release: Official J Controlled Release Soc. 2013;170:365–72. https://doi.org/10.1016/j.jconrel.2013.05.041 .

Pötter R, et al. MRI-guided adaptive brachytherapy in locally advanced cervical cancer (EMBRACE-I): a multicentre prospective cohort study. Lancet Oncol. 2021;22:538–47. https://doi.org/10.1016/s1470-2045(20)30753-1 .

Li Y, et al. A phase 2 study of nanoparticle albumin-bound paclitaxel plus nedaplatin for patients with advanced, recurrent, or metastatic cervical carcinoma. Cancer. 2017;123:420–5. https://doi.org/10.1002/cncr.30328 .

Yu XL, Wu MF, Ding L, Yang J, Bai SM. Enhanced efficacy of Neoadjuvant Chemotherapy with Nab-Paclitaxel and Platinum for locally Advanced Cervical Cancer. Cancer Manag Res. 2021;13:9297–304. https://doi.org/10.2147/cmar.S343602 .

Lapresa M, Parma G, Portuesi R, Colombo N. Neoadjuvant chemotherapy in cervical cancer: an update. Expert Rev Anticancer Ther. 2015;15:1171–81. https://doi.org/10.1586/14737140.2015.1079777 .

Download references

Acknowledgements

Not applicable.

Author information

Authors and affiliations.

Department of Radiation Oncology, The Affiliated Hospital of Qingdao University, 16 Jiangsu Road, Qingdao, 266003, China

Yaping Wu, Peng Jiang, Zhiying Chen, Wei Li, Bin Dong & Yongchun Zhang

You can also search for this author in PubMed   Google Scholar

Contributions

PJ, YCZ, WL and BD collected the patient’s data. YPW finished and revised the manuscript. YPW and PJ provided the figures and the tables. YCZ provided final approval for the version to be published. All authors contributed to the article and approved the submitted version.

Corresponding author

Correspondence to Yongchun Zhang .

Ethics declarations

Ethics approval and consent to participate.

The experimental protocol was developed in accordance with the ethical guidelines of the Declaration of Helsinki. All experimental protocols were approved by Medical Ethics Committee of the Affiliated Hospital of Qingdao University. The Medical Ethics Committee of the Affiliated Hospital of Qingdao University waived the need for informed consent. Ethics Approval Number: QYFY WZLL 27744.

Consent for publication

Competing interests.

The authors declare 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.

Wu, Y., Jiang, P., Chen, Z. et al. Efficacy and safety of different chemotherapy regimens concurrent with radiotherapy in the treatment of locally advanced cervical cancer. BMC Cancer 24 , 589 (2024). https://doi.org/10.1186/s12885-024-12358-8

Download citation

Received : 17 November 2023

Accepted : 07 May 2024

Published : 15 May 2024

DOI : https://doi.org/10.1186/s12885-024-12358-8

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

• Nab-paclitaxel
• Locally advanced cervical cancer (LACC)
• Concurrent chemoradiotherapy (CCRT)
• Adverse events (AEs)

ISSN: 1471-2407

• Submission enquiries: [email protected]
• General enquiries: [email protected]

Global main menu

• Queen Mary University of London
• News stories

HPV vaccination is a highly effective way of reducing cervical disease, finds research led Professor of Cancer Epidemiology, Peter Sasieni

The human papillomavirus (HPV) vaccination programme in England has been associated with a substantial reduction in cervical disease across all socioeconomic groups .  

HPV is one of the most common sexually transmitted infections and many countries, including the UK, now offer routine vaccination to girls and boys at age 12-13 to protect them against strains that can cause cancer in later life.   

The n ew study, published today by the British Medical Journal (BMJ) and led by Professor Peter Sasieni from the Wolfson Institute of Population Health, analysed cancer data from NHS England for vaccinated and unvaccinated women aged 20-64. Researchers concluded that the vaccine reduced cervical cancer incidence rates by nearly 90% and pre-cancerous conditions by around 95% in women. The study also found that the vaccine was much more effective when taken up by people in year 8 (aged 12-13) than later years.  

Between 1 January 2006 and 30 June 2020 there were 29,968 diagnoses of cervical cancer and 335,228 of grade 3 precancerous cervical lesions (CIN3) in women aged 20-64 years. In the group of women offered vaccination at age 12-13, rates of cervical cancer and CIN3 in the additional year of follow-up were, respectively, 84% and 94% lower than in the older unvaccinated group. Overall, the researchers estimate that by mid-2020, HPV vaccination had prevented 687 cancers and 23,192 CIN3s.  

While the incidence of cervical cancer was higher in areas of the most deprivation, the study showed that the HPV vaccination had prevented the greatest numbers of cervical cancer cases in women in the most deprived areas of England.  

Professor Peter Sasieni, lead author from Queen Mary University of London, said: “Our research highlights the power of HPV vaccination to benefit people across all social groups. Historically, cervical cancer has had greater health inequalities than almost any other cancer and there was concern that HPV vaccination may not reach those at greatest risk. Instead, this study captures the huge success of the school-based vaccination programme in helping to close these gaps and reach people from even the most deprived communities. In the UK, the elimination of cervical cancer as a public health problem in our lifetime is possible with continued action to improve access to vaccination and screening for all.”  

Cancer Research UK’s senior health information manager, Sophia Lowes, said: “Every year, around 3,300 people receive a cervical cancer diagnosis in the UK. This research shows us that HPV vaccination works, and increased coverage can help to bring about a future virtually free from this disease. But we can't lose momentum. We're calling for targeted action to ensure that as many young people as possible get the lifesaving HPV vaccine. Better reporting on uptake by deprivation and ethnicity, along with more research, will help us understand how to reach those most at risk.  

"We encourage people to take up the HPV vaccine if they are eligible. If you are concerned that you or your child has missed out on the HPV vaccine, you can contact your child’s school nurse, school immunisation service or GP surgery to find out more."  

Related items

For media information, contact:

A .gov website belongs to an official government organization in the United States.

A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

• Risk Factors
• Health Equity
• Cancer Data and Statistics
• Cancer Research
• Resource Library
• Types of Cancer
• HPV and Cancer
• Cancer Research Citation Search Tips
• Initiatives

Older Women May Not Get Needed Cervical Cancer Screenings

What to know.

Some women who are 65 or older should be screened for cervical cancer.

One type of cancer that only women can get is cancer of the cervix, or cervical cancer. Most cervical cancer is caused by human papillomavirus (HPV). The only sure way to find out if you have cervical cancer is to get a screening test (a Pap test, an HPV test, or both tests). If you are a woman who has not had her cervix removed by surgery (a hysterectomy), keep getting tested until you are at least age 65.

However, a study found that some women do not continue to get screened for cervical cancer as they get closer to age 65. Unfortunately, you can still get cervical cancer when you are older than 65. The only way to know it is safe to stop being tested after age 65 is if you have had several tests in a row that didn't find cancer within the previous 10 years, including at least 1 test in the previous 5 years.

• For the Pap test alone, you should have 3 normal tests in a row.
• For the Pap-HPV co-test, you should have 2 normal tests in a row.
• Screening after age 65 may be appropriate for some women at high risk, including women with a history of cervical lesions or cancer, women whose mothers took a hormone called diethylstilbestrol (DES) while pregnant, or women who have a weakened immune system. Women at high risk should talk with their doctors about how often to get screened and when to stop being screened.

How the study worked

The researchers wanted to know how many women between ages 41 and 70 had never had a Pap test or co-test, or hadn't had one in the 5 years before the survey. They looked at answers to questions about cervical cancer testing in the National Health Interview Survey. The questions were asked in 2013 and 2015. The researchers also looked at information from two federal cancer registry programs to see how the chance of getting cervical cancer changes with age among women who have not had a hysterectomy. (Cancer registries are used to keep detailed records about cases of cancer.)

Key findings

• The older women get, the more likely it is that they have never been tested or haven't been tested in the previous 5 years.
• About one woman out of 20 between ages 66 and 70 has never been tested.
• An older woman, until she's in her 80s, who has not had a hysterectomy, is at least as likely to get cervical cancer as a younger woman.
• Cervical cancer incidence rates increased with age and were higher for Black women than White women.

What this means

• If you are a woman between 21 and 65 who hasn't had a hysterectomy and hasn't had a cervical cancer screening test in the past 5 years, ask your doctor about getting one.
• If you are a woman older than 65 who hasn't had a hysterectomy, talk to your doctor about your risk for cervical cancer and whether you still might benefit from testing.
• Health care staff, as part of regular checkups, could review the health records of patients 65 years or older, or ask about past cervical cancer screening tests before deciding it's safe to stop screening for cervical cancer.
• For all women between ages 21 and 65 with no recent history of cervical cancer screening, health care staff can counsel them about the importance of getting a screening test.

Data citation

White MC, Shoemaker ML, Benard VB. Cervical cancer screening and incidence by age: unmet needs near and after the stopping age for screening. Am J Prev Med. 2017;53(3):392–395.

Learn how to lower your cancer risk and what CDC is doing to prevent and control cancer.