Improving cervical cancer survival – A multi-faceted strategy to sustain progress for this global problem

Abstract

Cervical cancer is associated with profound socioeconomic and racial disparities in incidence, mortality, morbidity and years of life lost. The last standard-of-care (SOC) treatment innovation for locally advanced cervical cancer occurred in 1999, when cisplatin chemotherapy was added to pelvic radiation therapy (CRT). CRT is associated with a 30–50% failure rate, and there is currently no cure for recurrent or metastatic disease. The enormity of the worldwide clinical problem of cervical cancer morbidity and mortality, as well as the egregiously unchanged mortality rate over the last several decades, are recognized by the National Institutes of Health as an urgent priority. This is reflected within the Office of Research on Women’s Health effort to Advance NIH Research on the Health of Women, as highlighted in a recent symposium¹. In the current review we will address the state of the science and opportunities to improve cervical cancer survival with an emphasis on improving access, utilizing technology in innovative and widely implementable ways, and improving our understanding of cervical cancer biology.

I. Cervical cancer incidence and mortality – trends, disparities, and impact

Worldwide, a total of 604,127 new cases and 341,831 deaths are estimated to have occurred in 2020². Geographically, although >50% of global cervical cancer cases and deaths are reported from the continent of Asia, the age-standardized rate of incidence and mortality is highest in African countries (Figure 1)². The most recent projections for the burden of cervical cancer in the United States includes 14,100 new cases and 4,280 deaths in 2022³. Although the incidence of cervical cancer has trended down over the last 50 years in the US, the mortality rate remained unchanged. This is in contrast to most other cancers, in which a decrease in cancer incidence is accompanied by improved cancer survival over time, largely reflective of improved therapies³.

Figure 1. Global impact of cervical cancer in 2020.

Global incidence and mortality for cervical cancer reproduced with permission from Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: A Cancer Journal for Clinicians 71, 209–249 (2021).

A measurable drop in incidence and mortality of invasive cervical cancer was seen between 1970–2000, largely attributable to screening for pre-neoplastic changes by Papanicolaou (Pap) test, detection of human papilloma virus (HPV) and surrogates, and early intervention efforts⁴. HPV infection first effects young women of child-bearing age, resulting in a peak in cervical cancer incidence in the 4^th decade, affecting one in 359 women from birth to 49 years³. Many women are diagnosed in the prime years of social and economic productivity. Thus, the loss of life and disability associated with cervical cancer is measured by a high number of years of life lost (YLL) and disability adjusted life years (DALY = YLL and years lived with disability (YLD)). In a global study of all cancers, Soerjomataram and colleagues reported that cervical cancer was responsible for the highest age-adjusted DALY for women worldwide⁵. This suggests an enormous global impact of cervical cancer incidence and mortality beyond the absolute numbers.

The disproportionate impact of cervical cancer on Black women, and members of other racial and ethnic minority populations in the US and worldwide is well established, driven by myriad of inequities including systemic racism, differential access to screening and treatment, and other social determinants of health. Indeed, cervical cancer is among the top five widest Black-White mortality gaps among all cancer diagnoses⁶. Similarly, the worldwide burden of cervical cancer disproportionately affects women in low socioeconomic groups, with complex challenges at every step of the cervical cancer continuum. The extent of these disparities and steps needed to systematically address them have been eloquently laid out by experts^6,7, and are discussed further below.

The burden of cervical cancer morbidity and mortality is highest in low- and middle-income countries (LMICs), where it remains a leading cause of death among women². Women living with HIV are at a high risk of developing invasive cervical cancer even in the era of anti-retroviral therapies. Cervical cancer is a leading cause of death for women in sub-Saharan Africa that bears the largest burden of HIV⁸. While the majority of patients in LMICs present with locally advanced disease requiring CRT, resources for treatment are often limited in the same settings leading to poor overall outcomes. A recent analysis suggested that less than half of the countries in Africa have access to external beam radiation therapy (EBRT) and brachytherapy, both of which are essential for curative treatment of cervical cancer⁹. In addition to limited infrastructure, there are often significant challenges in access to timely evidence-concordant care for patients with cervical cancer such as gaps in care coordination, delays in receipt of diagnostic results, lack of education for patients in regard to treatment navigation, lack of knowledge for providers to deliver evidence concordant care, limited data particularly on treatment tolerability, outcomes and quality of life of patients in LMICs^10,11.

II. Screening and Prevention

Since beginning to see the decrease in cervical cancer incidence and mortality attributed to early screening^12,13, best practices have focused on optimizing criteria for screening initiation and exit, screening intervals, and the use of reflex, concomitant (e.g., contesting), or primary high-risk human papillomavirus (hrHPV) testing. In 2004, cotesting became the preferred strategy in women 30 years or older after multi-national cross-sectional analyses demonstrated superior sensitivity to detect high-grade neoplasia compared to cytology or hrHPV testing alone^14–16. More recently, primary HPV testing has demonstrated enhanced sensitivity and affords more protection from progression to high-grade neoplasia compared to cytology alone^17,18. Despite the demonstrated efficacy of primary hrHPV screening, implementation remains limited.

Implementation and uptake of the HPV prophylactic vaccine is a priority. The 9-valent vaccine is recommended for all individuals aged 11–12 years, though it can be given to those aged 9–45, ¹⁹. A randomized double-blinded study involving 14,215 women demonstrated excellent safety and efficacy and is proposed to prevent up to 90% of cervical cancers – a 20% improvement on the quadrivalent vaccine²⁰. Despite the effectiveness of large-scale screening and HPV vaccination, cervical cancer disparities remain. Black patients and other racial and ethnic minorities experience higher cervical cancer incidence and mortality and lower HPV vaccination completion rates than their White counterparts^21,22. Efforts to improve geographic access to cancer prevention, patient education, community trust in health providers, and awareness of implicit biases are necessary to achieve equity in cervical cancer outcomes.

III. Surgery for early stage disease

Surgery remains a mainstay of treatment for patients with early-stage cervical cancer²³. The specific recommended surgical procedure is driven by cancer stage and patients’ fertility desires^24,25. For example, cervical conization may be both diagnostic and therapeutic for patients with suspected microinvasive disease (defined as stage IA1 without lymphovascular space invasion (LVSI)); it characterizes the degree of stromal invasion, is curative if margins are negative, and may improve survival outcomes in patients who have completed childbearing and ultimately undergo a radical hysterectomy (RH)^23,26–28. If conization margins are positive, extrafascial hysterectomy or modified RH with pelvic lymph node (LN) dissection are appropriate salvage options²⁴. For patients with suspected 1A2, IB1, IB2, and IIA1 disease, RH with bilateral pelvic LN dissection (+/− sentinel LN mapping) is preferred. While an older prospective trial demonstrated similar 15-year overall survival (OS) outcomes for 125 stage IB1 and IIA patients randomized to Piver-Rutledge class III or class I (extrafascial) RH (90% vs. 74%, p = 0.11)²⁹, more recent retrospective studies have confirmed improved survival for stage IB1 patients undergoing RH compared to simple hysterectomy^30,31.

Over the last several years, the surgical approach to RH has shifted away from minimally invasive techniques due to worse survival outcomes associated with robotic and laparoscopic approaches compared to open cases. The Laparoscopic Approach to Cervical Cancer Trial (an international, non-inferiority, randomized control trial), along with other observational studies, demonstrated superior survival and similar quality-of-life for early-stage patients undergoing laparotomy compared to robotic or traditional laparoscopy^32,33. This trial involved 631 patients with stage 1A1 with LVSI, IA2, or IB1 disease, but was underpowered to detect survival differences specifically among patients with tumors < 2 cm. Trials comparing different surgical techniques among these lower-risk patients are ongoing^34,35.

IV. Chemoradiation for locally advanced cervical cancer

Several technological advances establish the current standard of care for locally advanced cervical cancer (LACC) – concurrent cisplatin-based CRT. Staging, prognostication, and treatment delivery have all been refined by incorporation of 3-dimensional (3D) imaging including (18)F-fluorodeoxyglucose (FDG) positron emission tomography (PET), computed tomography (CT), and magnetic resonance imaging (MRI). EBRT volumes adjusted if LN metastases are identified on FDG-PET, CT and/or MRI may translate to improved disease free survival³⁶. 3D-image-guided intracavitary brachytherapy (3D-IGBT) is associated with a 10% overall survival improvement compared to 2-dimensional imaging ³⁷, and is largely accepted as the standard of care where resources are available. The trans-European GEC-ESTRO group recently reported results of MRI-guided adaptive brachytherapy in locally advanced cervical cancer (EMBRACE-I), resulting in excellent local control and morbidity compared to historical controls³⁸. Advanced imaging has also been incorporated in the assessment of response to definitive chemoradiation^39,40, and early detection of recurrence⁴¹, and may provide useful insight into the biologic heterogeneity that is a hallmark of this disease^42,43.

Radiotherapy delivery techniques have also advanced, contributing not only to optimized local control and disease outcomes³⁸, but also reduction in therapy-related toxicity and detriment to quality of life with highly conformal techniques like intensity modulated radiation therapy (IMRT)⁴⁴. Further optimization of current radiation treatment is likely with volume-based and adaptive approaches, and will maximally exploit the extremely high radiation doses achievable with brachytherapy. Prospective studies of hypofractionated EBRT are critically needed. Additionally, novel radiation delivery techniques may provide opportunities to deliver radiosensitizers, immunotherapies and other potent anti-cancer agents more directly to the tumor, improving anticancer effect and minimizing systemic toxicities^45,46. An ongoing challenge is to ensure that trained physicians, advanced imaging and radiation treatment equipment are accessible throughout the world, and particularly in regions with high rates of cervical cancer.

V. Advanced disease and supportive care with broad implementation

Distant metastatic spread of cervical cancer is associated with poor long term survival. Some efforts, including comprehensive chemoradiation for patients with supraclavicular or LN-limited metastatic disease, suggests that an aggressive approach for select patients is warranted⁴⁷. Improved survival was demonstrated with stereotactic body radiation therapy (SBRT) to oligometastases in patients with other cancers ^48,49. Early retrospective data suggests similar efficacy and low toxicity profile in patients with oligometastatic cervical cancer⁵⁰. Systemic therapy with singlet cytotoxic chemotherapies are largely ineffective, showing only about a 10% response rate with significant toxicities. Targeted therapies combined with platinum doublet chemotherapy have shown the most promise to date. Targeting of the vascular endothelial growth factor receptor (VEGFR), and immune checkpoint therapy have provided incremental, though meaningful, improvements in overall survival for patients with recurrent and metastatic cervical cancer^51,52. Similarly, the drug-antibody conjugate tisotimab vedotin, which targets tumor expression of tissue factor (TF) to deliver intra-tumoral mono-methyl auristatin E, demonstrated promising results in early clinical trials⁵³. It is increasingly clear that real strides will be made with improved understanding of cervical cancer biology.

CRT, although curative, can negatively impact sexual and bladder functioning, psychosocial wellbeing, and overall quality of life. In LMICs, long term quality of life and challenges in survivorship care are largely understudied. In the last decade, most of the work in survivorship care has been done in high-income countries (HICs) and focuses on cancers more prevalent in HICs (e.g., breast, colorectal). Given the disproportionately high burden of cervical cancer in LMICs, there is a pressing need to develop interventions that can best address and improve survivorship care and outcomes specifically in LMICs^58–60. While there may be some interventions that can be adapted from HICs, without empirical understanding of the contextual barriers in LMICs, previous interventions may be inappropriate or potentially ineffective in these settings. The field of implementation science has turned its focus toward ensuring not only effectiveness but also equity of implementation⁶¹. At the global level, inequities in cervical cancer outcomes are stark, highlighting the crucial need for implementation studies that would help disseminate standard of care interventions, including those that rely upon advanced imaging, radiation, and novel biologics. In 2020, the World Health Organization (WHO) adopted a global strategy to eliminate cervical cancer that sets measurable targets for HPV vaccination (90%), cervical cancer screening (70%), and treatment for women with pre-cancer and invasive cancer (90%). This unified global effort has translated into national efforts within countries to achieve these 90–70-90 targets. Furthermore, institutes such as the NCI and IAEA have helped to further this goal by enhancing funding for research and implementation projects across the globle, with substantial focus on cervical cancer in low- and middle-income countries^62,63.

VI. Cervical cancer biology – opportunities and challenges

In addition to improving global access to current standard therapies and emerging biologic therapies for cervical cancer, it is clear that improving overall survival will require a better understanding of the diverse tumor biology underlying response to therapy, recurrence and metastasis.

Multiple studies have demonstrated that HPV status and HPV genotype are prognostic biomarkers in cervical cancer patients^64–66. Patients with no detectable HPV in their tumors have poor survival outcomes compared with patients with HPV+ tumors, and several studies have shown that patients with HPV α−7 group or HPV genotypes other than HPV 16 have worse outcomes than patients with HPV α−9 or HPV 16 (HPV16+) tumors^67–70. A recent study of a large cohort of tumor specimens from Ugandan patients with cervical carcinomas, many of whom were HIV-positive, suggested HPV-clade-specific differences in tumor epigenomic features, gene expression and pathway dysregulation⁷¹. The heterogeneity of HPV genotypes in cervical cancer, as well as differences in viral copy number, integration status and associated viral gene transcription, are possible explanations for why HPV status alone is not always a binary biomarker for CRT sensitivity in cervix, as has been demonstrated for HPV related head and neck cancer. Genotype diversity and copy number may also impact the sensitivity of circulating HPV DNA (ctDNA) to detect disease status^72–75. Further study is needed to understand how HPV specific biomarkers can be used to in the clinical management of cervical cancer.

In addition to HPV ctDNA, circulating serum levels of squamous cell carcinoma antigen (SCCA), first described in this Journal in 1977 as tumor antigen-4 (TA-4), is prognostic for cervical cancer outcomes^76–78. SCCA is measureable in approximately two-thirds of patients with cervical squamous- and adenocarcinomas, and can serve as an early indicator of response to therapy, or of recurrence⁷⁹. Additionally, SCCA, also known as protease inhibitors SERPINB3 and SERPINB4 – serves a molecular role in protecting tumor cells against radiation-induced cell death, suggesting that this prognostic biomarker may also be a viable therapeutic target⁸⁰.

A systematic review of biomarkers for cervical cancer treated with CRT identified 82 molecular markers studied in retrospective cohorts⁸¹. SCCA and cyclooxygenase-2 (COX-2) were associated with worse overall and disease free survival after CRT. Overactive ribonucleotide reductase (RNR) serves as an IHC biomarker of poor metabolic response to cisplatinum-CRT⁸². Triapene, a therapeutic inhibitor of RNR inhibits DNA repair in response to CRT, is being assessed in an ongoing Phase III randomized clinical trial for LACC through the NRG Oncology group, after promising results from a small randomized trial⁸³. Continued translation of biomarkers to new personalized therapies is urgently needed for cervical cancer.

Encoded within the HPV genome are 2 oncogenes that target host tumor suppressors resulting in increased tumor cell sensitivity to genotoxic stress, including chemotherapy and radiation. HPV E6 targets p53, and HPV E7 targets pRb and p130. Through E6 and E7, HPV induces cell immortalization resulting in a hyperproliferative state that facilitates viral replication. Maintenance of this hyperproliferative state by E6 and E7, together with binding numerous additional proteins that are essential for the execution of cell cycle checkpoints, DNA repair and apoptosis, render HPV positive cells sensitive to therapies that induce or prolong DNA damage. Indeed, preclinical evidence is accumulating that HPV positive cervical cancers are susceptible to biological therapies that target the DNA Damage Response (DDR) pathway, (i.e., ATM, ATR, DNA-PK and PARP inhibitors)^84–88. Sensitivity to this class of drugs, however, is not uniform amongst HPV positive cervical tumors, even within the same HPV genotype. Work is ongoing to determine how additional biological differences, such as HPV alternative transcript expression, may influence chemoradiation and DDR inhibitor sensitivity⁸⁹. Recent work has also highlighted the potential for other aspects of the tumor microbiome, in addition to HPV, such as bacterial and fungal species, to influence tumor responses to chemoradiation^90–93. Much of this developing work highlights the connection between the microbiome and the tumor immune microenvironment, which may support the potential for radiation to successfully induce systemic anti-tumor immune responses.

The majority of cervical cancers are caused by persistent infection with HPV, and this has generated enthusiasm for immunotherapy; however clinical trials using immune checkpoint blockade (ICB) as monotherapy in cervical cancer show limited efficacy^94,95. Most recently, a press release from the CALLA trial, a randomized phase III study of the addition of durvalumab during and after chemoradiation in cervical cancer, indicate no benefit of the addition of durvalumab⁹⁶, and we await further analyses with the official publication of these results. The publication of the KEYNOTE 826 trial of cisplatin plus placebo versus pembrolizumab (anti-PD-1) with or without bevacizumab (anti-VEGF) in persistent, recurrent or metastatic cervical cancer demonstrated a 2-month improvement in progression free survival, suggesting that the combination of ICB + chemotherapy may hold promise in metastatic/recurrent cervical cancer⁵². Interestingly, in this trial tumor PD-L1 expression was not a reliable predictor of response to anti-PD-1 therapy. Currently, no reliable pretreatment biomarkers exist to guide treatment choices or predict response to SOC CRT +/−ICB in cervical cancer. Identifying and validating candidate predictive biomarkers in cervical cancer is an unmet need that can be used in the future to select more effective personalized treatment.

Recently, a major development in the field of radiation oncology has been to use RT in combination with other agents as a means to stimulate host systemic anti-tumor immune responses. The success of this strategy has been supported by results from multiple preclinical and clinical studies in other cancer types, although the precise timing and combination of RT dose and fractionation to achieve these effects is still debated. Several ongoing or recently completed clinical trials are testing RT + ICB and CRT + ICB with variable sequencing in cervical cancer, including the CALLA trial with preliminary results indicating no benefit to durvalumab + CRT over CRT alone. Although many of the trials are still in process, a few have included preliminary analysis of tumor and blood collected from patients during treatment⁹⁷. Together with recently published longitudinal data from our group⁹⁸, the results of these preliminary translational analyses demonstrate that SOC CRT is associated with 1) changes in tumor cell and HPV associated gene expression that can support tumor cell survival post RT; 2) decreased numbers of CD4+ and CD8+ T cells within the cervix TME; 3) increased expression of activation markers ICOS and PD-1 on circulating CD4 T cells; 4) increases in circulating levels of memory T cells; 5) increases in circulating levels of protumorigenic cytokines TGF-β and GM-CSF that may contribute to myeloid shift in the bone marrow and increased “M2 like” TAMs in to the cervix TME. Overall these observations support a complex picture of immune stimulating and suppressive properties of SOC CRT in cervical cancer that are driven in part by CRT induced changes in tumor cells, TAMs and T cells. The ideal combination may include RT plus immunotherapy without concurrent cytotoxic chemotherapy. Furthermore, brachytherapy may be more efficient at stimulating host anti-tumor immune responses than external beam, particularly when regional LN are included in the RT target volume. Well designed Phase I clinical trials with corresponding translational studies together with functional and mechanistic preclinical work are needed to unlock the potential of immunotherapy in cervical cancer.

Over the past 2 decades, significant advances have been made through the incorporation of advanced imaging into pretreatment evaluation and the delivery of SOC CRT. Publication of the immense efforts of EMBRACE-I³⁸ joins previous reports in setting optimal goals to implement SOC imaging globally. FDG-PET has demonstrated utility in identifying patient groups at risk of recurrence primarily through enhanced sensitivity for the identification of involved LN regions^99–104. Early identification of involved LN can avoid unnecessary surgeries and guide radiation therapy field design in ways that can improve patient outcomes, and the results of pretreatment imaging are now incorporated into the 2018 revision of FIGO staging system¹⁰⁵. Additionally, post therapy FDG-PET imaging, performed at 3 months after the completion of chemoradiation, can be used as an early marker of treatment response⁴⁰. Studies are ongoing to validate and refine treatment decision making in the setting of persistent and progressive disease on post therapy FDG-PET. FDG-PET is a functional imaging strategy that depends upon tumor glucose uptake. Cervical tumors with increased FDG uptake on pretreatment PET are resistant to SOC CRT, highlighting the connection between tumor metabolic dependencies and resistance to CRT (Figure 2)^39,106. Preclinical studies have now demonstrated that the addition of inhibitors of glucose, glutamine and redox metabolism, some of which are available as oral therapies, can enhance the efficacy of RT^107,108. Clinical trials are being developed to test this strategy and may involve FDG uptake on pretreatment PET (with or without mutations in the PI3K pathway) as a predictive biomarker.

Figure 2. FDG-PET imaging sheds new light on the biology and classification of cervical cancer.

New approaches have been used to integrate information from functional imaging studies such as FDG-PET into the analysis of gene expression data from cervical tumors. In this example, integrated analysis of FDG uptake on PET measured by standardized uptake value (SUVmax) together with tumor gene expression data from RNASeq identified a subcategory of HPV positive squamous cervical cancers that are resistant to standard of care chemoradiation. Reproduced with permission from Zhang, J. et al. Integrating imaging and RNA-seq improves outcome prediction in cervical cancer. J Clin Invest 131, 139232 (2021).³⁹ Copyright 2021, Journal of Clinical Investigation.

Cervical cancer genomics and the potential of personalized medicine –

Although cervical cancers are unlike many other solid tumors in that TP53 mutations are rare, HPV-mediated downregulation of p53 has similar implications for disrupting the normal cellular roles of p53 in cell cycle regulation, DNA-repair, and regulation of cell death. As such, the response to ionizing radiation and other DNA damaging therapies is complex. Several recent translational studies highlight the modes in which cervical tumor cells evade cell death by upregulating protective measures^80,109. Further study to identify tumor vulnerabilities are critical to improving overall response to CRT, and for integrating targeted drugs into personalize treatment approaches for patients with cervical cancer.

Significant advances have been made in recent years in understanding the heterogeneity of cervical cancer through the use of next generation sequencing^110,111. The combined results of these tumor cataloguing studies have generated a wealth of new information about cervical cancer biology and nominated new targets for therapy (Figure 3)¹⁰¹. Moving forward, it is critical to think about genomics and new targets in the context of a more holistic approach to treatment. Well designed and accurately powered validation studies in uniformly treated patient populations with long term follow up data are needed to prioritize new targets. In addition, we need to prioritize treatments, such as hypofractionation and oral therapies that simplify treatment and improve patient access to care. Collaboration with basic scientists is needed to perform functional genomics and preclinical testing to optimize new combination therapies and minimize toxicity. Working together, we can help ensure the success of new agents in the context clinical trials. It is truly an exciting time with the potential to generate real impact to improve the lives of cervical cancer patients.

Figure. 3. Next generation sequencing identifies new targets for cervical cancer treatment.

Tumor cataloguing studies such as The Cancer Genome Atlas Project have identified new targets for personalized medicine in cervical cancer. Reproduced with permission from Cancer Genome Atlas Research Network et al. Integrated genomic and molecular characterization of cervical cancer. Nature 543, 378–384 (2017).¹⁰¹