Cervical cancer continues to be a global burden for women, with > 500,000 cases and 275,000 deaths reported annually. Resource-rich countries have seen a dramatic reduction in the prevalence of invasive cervical cancer due to widely accessed screening programs (and further reductions are expected with the increased use of the human papillomavirus [HPV] vaccine). However, these gains have not been realized in resource-poor countries, in which barriers in the infrastructure and implementation of screening programs and the access to treatment remain significant. Despite these findings, much progress is still required to improve the outcomes for women diagnosed with invasive cervical cancer. These include better approaches to reduce the risks of local and distant recurrences, less toxic treatments without sacrificing treatment efficacy, and better therapies for women with recurrent and metastatic disease.
Unfortunately, conducting clinical trials in patients with cervical cancer is becoming increasingly challenging. Multiple factors contribute to this, including the reduced number of women eligible for clinical trials, a lack of accessibility to these trials for those who are candidates (particularly among women with limited resources, in whom cervical cancer continues to be a major health dilemma), and logistical issues in the conduct of international collaborations. In addition, funding agencies are also focusing on biologically based hypotheses to advance knowledge of disease in addition to clinical outcome.
The Cervical Cancer Task Force in collaboration with the Gynecologic Oncology Group (GOG) convened a state of the science (SOTS) meeting, bringing together investigators and content experts with an interest in cervical cancer clinical trials. This article summarizes the presentations given during the meeting, which we anticipate will provide progress toward the conduct of the next generation of clinical trials in this disease.
Primary Therapy: Where Are We Now?
As a result of screening, most cervical cancers can be identified early and cured with surgery. However, the lack of routine population-based screening in some parts of the United States and in resource-poor countries results in the majority of unscreened patients presenting with locally advanced disease. Historically, and in many parts of the world, patients typically are treated with surgery for limited disease, with radiotherapy (RT) (when available) used predominantly for those with locally advanced disease, either as adjuvant therapy or in lieu of surgery for those who are not surgical candidates. Chemotherapy is reserved solely for the treatment of patients with distant metastatic or recurrent disease.
Approximately 30% to 40%, 60%, and 80% to 100% of patients with International Federation of Gynecologic Oncology (FIGO) stage II, III, and IV disease, respectively, were not cured with standard RT regimens alone. Limitations in the tolerance of normal tissue to increased RT doses meant that RT doses could not be escalated without increased toxicity. Given these issues, the concurrent administration of chemotherapy during RT (chemoradiation) has been extensively evaluated as an additional treatment modality.
The GOG has extensively studied the role of chemoradiation in patients with cervical cancer using multiple agents including hydroxyurea, cisplatin, and 5-fluorouracil. Five randomized trials (Table 1) were subsequently completed with a consistent benefit in survival outcomes noted with chemoradiation compared with RT alone in 4 of them,[4-10] although 1 relatively smaller Canadian trial did not demonstrate this survival advantage. These trials had various eligibility criteria but represented a variety of important clinical scenarios including women with bulky stage IB (tumors measuring > 4 cm) to stage IVA disease and those with stage IA2 and IB disease with high-risk factors (positive lymph nodes, parametrial extension, and/or positive surgical margins) after radical hysterectomy. These results prompted a clinical alert by the National Cancer Institute in 1999 that recommended the addition of concurrent chemotherapy to RT when RT is used as a primary treatment for women with cervical cancer. In a 2010 Cochrane meta-analysis of 13 randomized trials that compared chemoradiation with primary RT, chemoradiation resulted in a 20% reduction in the risk of death (hazards ratio [HR], 0.69; 95% confidence interval [95% CI], 0.61-0.77) and a 34% reduction in the risk of disease recurrence (HR, 0.66; 95% CI, 0.59-0.73). There was also evidence that compared with primary RT alone, chemoradiation further reduced the odds of a local disease recurrence (odds ratio [OR], 0.59; 95% CI, 0.50-0.69) and the development of distant disease (OR, 0.81; 95% CI, 0.65-1.01), although the latter did not achieve statistical significance.
|How will success be defined?|
|• What is the threshold for significance in defining improvements in progression-free or overall survival? Rates of long-term survival?|
|• How does cost factor into the designation of trial endpoints? Does this vary within or between communities? Countries?|
|• In resource-limited areas, what endpoints are preferable?|
|What knowledge is needed?|
|• What advances will be most readily transported to patient care?|
|• Do we need genetic/genomic targets?|
|• Where will predictive biomarkers be most useful? What kinds of biomarkers?|
|• What is the role of imaging in the management of cervical cancer? How do we take into account the variation in resources at the community, regional, national, and international levels?|
|How can we improve trial designs to be more practical and hence more fruitful?|
What is the role for neoadjuvant chemotherapy?
Because the curability and local control of patients with locally advanced cervical cancer are related to the size of the tumor at the time of presentation, cytoreducing the tumor before definitive therapy (ie, surgery or RT) has been proposed as another way to treat this disease. Despite the lack of randomized trials demonstrating efficacy, a 2012 meta-analysis of 6 trials demonstrated that both overall survival (HR, 0.77; 95% CI, 0.62-0.96) and progression-free survival (HR, 0.75; 95% CI, 0.45-0.99) were improved with the use of neoadjuvant therapy. However, significant heterogeneity, the small number of trials included in this analysis, and a lack of a uniform treatment regimen evaluated in the trials limit these conclusions. At this time, a trial of neoadjuvant chemotherapy before radical hysterectomy versus concurrent cisplatin-based chemoradiation is being conducted in Europe (European Organization for Research and Treatment of Cancer [EORTC] 55994/NCT00039338).
Is there a role for systemically dosed chemotherapy after chemoradiation?
One trial has raised the potential survival advantages associated with the use of systemic chemotherapy after definitive chemoradiation. However, the results of this trial require confirmation before it is used as an acceptable alternative to definitive chemoradiation. In this trial, Duenas-Gonzalez et al evaluated the administration of chemoradiation with cisplatin versus cisplatin plus gemcitabine. Women assigned to treatment with the combination regimen also received additional systemic therapy after chemoradiation was completed. With a median follow-up of 3 years, women who received cisplatin plus gemcitabine (with RT and then as post-RT treatment) had a 32% reduction in disease progression (HR, 0.68; 95% CI, 0.49-0.95) and a higher rate of progression-free survival at 3 years compared with those treated with cisplatin alone (74% vs 65%, respectively). In addition, cisplatin plus gemcitabine resulted in a 32% survival advantage (HR, 0.68; 95% CI, 0.49-0.95). However, there were methodological problems with follow-up in this study, and a second publication by the authors purported no benefit in various subgroups.
Despite these results, there are reports from other investigators that the combination of cisplatin, gemcitabine, and pelvic irradiation results in unacceptable toxicity. At this time, 2 international phase 3 trials are currently ongoing to further evaluate the role of additional chemotherapy (carboplatin and paclitaxel for 4 cycles) after cisplatin-based chemoradiation (ANZGOG 0902/GOG 0274/NCT01414608 and Radiation Therapy Oncology Group [RTOG] 0724/NCT00980954).
The Approach to Women With Recurrent or Metastatic Disease
The GOG has been committed to clinical trials for women with advanced and recurrent cervical cancer, and a comprehensive review of GOG trials is discussed elsewhere. Unfortunately, none of the various platinum-based combinations was found to improve survival outcomes when compared with the most commonly administered regimen, cisplatin plus paclitaxel. One study performed by the Japanese Gynecologic Oncology Group (JGOG 0505) suggested that using carboplatin instead of cisplatin could dramatically reduce toxicity without adversely impacting survival, particularly in women previously treated with cisplatin-based chemoradiation.
A major step forward in the treatment of patients with recurrent or metastatic cervical cancer was recently achieved with the results of GOG 240, which randomly assigned patients to treatment with 1 of 2 chemotherapy regimens (cisplatin plus paclitaxel vs paclitaxel plus topotecan) and to bevacizumab versus no bevacizumab. There was no difference in outcomes noted between the chemotherapy regimens. However, when compared with chemotherapy alone, the addition of bevacizumab significantly prolonged overall survival (median, 17 months vs 13.3 months; HR, 0.71 [9% CI, 0.54-0.94]), improved the overall response rate (48% vs 36%), and improved progression-free survival (median, 8.2 months vs 5.9 months; HR, 0.67 [95% CI, 0.54-0.82]). These benefits were offset by an increase in toxicities with the addition of bevacizumab, including National Cancer Institute Common Toxicity Criteria for Adverse Events (CTCAE) version 3.0 grade ≥ 2 hypertension (25% vs 2%), gastrointestinal fistula formation (3% vs 0%), neutropenia (35% vs 26%), and thromboembolism (8% vs 1%). Despite these toxicities, the results of GOG 240 represent what to our knowledge is the first time that overall survival was improved over standard therapy, particularly when the drug can be accessed. The question of where to go from here for patients with locally advanced and recurrent or metastatic disease drove the second focus of the symposium.
Exploiting the Genomic Diversity of Cervical Cancer
Although not yet fully mature, emerging data have suggested that there are molecular alterations present in cervical cancer that differ by histologic subtype. In work from the Dana-Farber Cancer Institute, data indicate that 60% of patients with cervical cancer carry somatic mutations, the most common of which were PIK3CA (phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha) and KRAS. In addition, E322K substitutions in the MAPK1 (mitogen-activated protein kinase 1)/ERK1 (extracellular signal-regulated kinase 1) gene appear recurrently, and may be a targetable finding. It is interesting to note that PIK3CA mutations, found in more than one-third of cancers independent of histological subtype, were associated with shorter survival (67 months vs 90.3 months; HR, 9.3 [95% CI, 2.8-29.5]). In contrast, KRAS mutations were only present in adenocarcinomas, and a rare missense mutation in the epidermal growth factor receptor (EGFR) gene was only present in squamous cell carcinomas. None of these mutations was associated preferentially with either HPV type 16 or HPV type 18 infection. Finally, up to 7% of patients appear to harbor clinically significant mutations involving HER2,[22, 23] suggesting yet another targetable mutation for HER2-directed therapy.
Novel Therapeutic Approaches
A significant percentage of patients with cervical cancer develop disease recurrence after definitive treatment. Even with the results shown in GOG 240 (discussed earlier), the prognosis for patients with recurrent or metastatic cervical cancer remains poor and represents an urgent unmet need worldwide. Therefore, better alternatives to conventional therapy must be explored for the treatment of cervical cancer, including the administration of novel agents that may improve the therapeutic index of definitive chemoradiation and various immunotherapeutic approaches.
The rationale for immunotherapy in cervical cancer is based on the causative role of HPV infection in this disease. There are data to support the theory that the host immunological status and HPV-induced immune evasion are responsible for persistent HPV infection and the subsequent development of cervical cancer. HPV infection invokes a cellular immune response, and regulatory T cells appear to play a role in local immune suppression in HPV-associated tumors. New agents have been developed to interrupt the mechanisms involved in cancer immune evasion, and promising results have been reported with this approach in other tumor types.
The use of bacterial vectors directed against E7 has been shown to induce tumor regression in animal models, and a phase 2 trial with a live-attenuated Listeria monocytogenes vaccine suggests it may have activity in patients with cervical cancer. As presented at the 2013 American Society of Clinical Oncology Annual meeting, > 100 women in India with recurrent/refractory cervical cancer had been treated and 63% were still alive at 6 months of follow-up, with evidence of both complete and partial responses noted in 12 patients. These promising results have led to a trial in the United States using the same live-attenuated L. monocytogenes cancer vaccine under the auspices of the GOG (GOG 265/NCT01266460).
Regulatory pathways that limit the immune response to cancer are also becoming increasingly well characterized. This includes the upregulation of cytotoxic T lymphocyte-associated molecule-4 (CTLA-4), which is important for the activation of cellular immunity. The CTLA-4 receptor on T lymphocytes is a negative regulator of T-cell activation that outcompetes CD28 for binding to B7 on antigen-presenting cells and therefore works as an immune checkpoint molecule that downregulates pathways of T-cell activation. The monoclonal antibody ipilimumab blocks CTLA-4 to promote antitumor immunity, and thus results in an immune response against the tumor. It is specifically being evaluated in a phase 1 clinical trial in patients with cervical cancer (GOG 9929/NCT01711515).
A second coinhibitory pathway on activated T cells uses the inhibitory receptor programmed cell death 1 (PD-1). When PD-1 binds to its ligand, PD-L1, which is often present on tumor cells, the ability of the activated T cell to produce an effective immune response is downmodulated. Antibodies directed against PD-1 or PD-L1 thus may restore or augment an antitumor immune response and produce tumor responses in patients with advanced melanoma. Monoclonal antibodies targeting both PD-1 and PD-L1 currently are being developed to interrupt this pathway and to augment the antitumor immune response and may have a role in the treatment of cervical cancer.
Finally, the role of adoptive immunotherapy using tumor-infiltrating lymphocytes represents another potentially useful approach for patients with advanced or metastatic disease. In this technology, the tumor is excised and enriched for tumor-infiltrating lymphocytes, which are expanded after selection and reinfused to patients after lymphodepletion. Based on responses noted in tumors such as malignant melanoma, this approach is being evaluated in an ongoing trial at the National Cancer Institute for patients with recurrent or refractory metastatic cervical cancer (NCT01266460).
Potential novel targets
Radiation causes damage to DNA, leading to apoptosis and cellular death. Molecular pathways of the cancer cell both recognize the need for DNA repair and slow the cell cycle progression until DNA damage has been repaired. Deoxyribonucleotide triphosphate used in the repair of DNA damage after ionizing RT can be salvaged from deoxyribonucleosides or synthesized de novo by ribonucleoside reduction. Ribonucleotide reductase (RNR) activity increases to supply the required deoxyribonucleotide triphosphates used to repair DNA damage. Therefore, these agents (which to the best of our knowledge have been underexplored to date) that decrease the activity of RNR may have synergistic potential in combination with RT in the clinic.
In cervical cancer cell lines, 17-fold elevations in RNR proteins have been detected. Clinically, elevated RNR levels are associated with an increased risk of an incomplete response to chemoradiation and an increased risk of disease recurrence. Therefore, targets that inhibit RNR may be of potential value in combination with RT. One such agent is 3-aminopyridine-2-carboxaldehyde thiosemicarbazone, which was evaluated in a single-institution phase 2 trial in which 25 patients received it 3 times per week with cisplatin-based chemoradiation. With a median follow-up of 20 months, clinical responses were observed in 24 of 25 patients with suggested metabolic complete responses noted in 23 of these 24 patients as evaluated using a positron emission tomography scan at 3 months. These encouraging results have prompted a larger, multiinstitutional, randomized phase 2 trial evaluating this combination (NCT01835171) and was discussed as a potential innovative primary treatment to be further investigated in patients treated for advanced disease with chemoradiation.
Irradiation is a major cause of DNA damage. Thus inhibition of poly (adenosine diphosphate-ribose) polymerase (PARP) provides another potential target for the treatment of cervical cancer. In cervical cancer cell lines, the combination of topotecan and the PARP inhibitor ABT-888 (also known as veliparib) or irradiation and ABT-888 resulted in the inhibition of DNA repair and increased cytotoxicity. Based on these data, 2 trials currently are being conducted among patients with this disease through the GOG (GOG 127W/NCT01266447 and GOG76HH/NCT01281852).
EGFR expression in patients with cervical cancer activates 2 proliferation pathways: one that involves the RAS, RAF, MEK, and mitogen-activated protein kinase (MAP) and a second that involves the phosphatidylinositide 3-kinase (PI3K)/AKT/mammalian target of rapamycin pathway. Therefore, inhibition of EGFR has been a focus of clinical trials in this disease. Therapeutic targets include blockage of EGFR with small molecule inhibitors (cetuximab, gefitinib, erlotinib, and selumetinib), although the combination of cetuximab with RT was deemed not to be of clinical use.[35, 36] Beyond EGFR inhibitors, metformin may play a role by its inhibition of the mammalian target of rapamycin pathway and currently is being study in other cancers.
A Roadmap to Progress
Like all cancers, there has been definable progress in our understanding and intervention in cancer of the cervix over the last decades. This progress has been in the area of understanding the role and etiology of HPV in carcinogenesis, vaccination for prevention, modifying early detection by the incorporation of oncogenic HPV status to cervical smear cytology, and incremental benefits in treatment. The recognition that HPV infection is required for cervical carcinogenesis was instrumental in leading to the prophylactic vaccination program, with its optimistic early outcomes. Given that vaccination may take a generation or more to fully realize a reduction in cervical cancer incidence and mortality, efforts to improve early detection and therapy should continue. These are especially important to address in the needs of women in the resource-poor areas United States and around the world, where cancer of the cervix is the leading cause of cancer death for women. Getting vaccination to be a global success is needed, but currently unduly optimistic. Early detection via the Papanicolaou test changed the landscape of the disease in the United States and the industrialized world, and point-of-contact testing is improving outcomes in Africa.
Chemoradiation and surgery have increased survival for women with early and advanced nonmetastatic disease, leading to new questions of how to reduce treatment intensity and when to preserve fertility. Advanced and recurrent cervical cancer remain both a problem and a very rare challenge, with approximately 4000 cases diagnosed per year in the United States. The population at risk is too often women who do not receive screening or vaccination, and who present with the late-stage disease due to limited health care or other opportunities. Outreach to those populations is an area for attention and should incorporate the growing knowledge and therapeutic armamentarium.
Costs should be considered in the development of new agents so that our results may be more generalized worldwide. Determining the next critical research questions will require prioritization of mission and resources.
Some agreement on target parameters is needed to define these new directions (Table 1). Early cervical cancer may be preventable, and when found is highly curable. Recurrent, advanced, and de novo metastatic cervical cancer are rare, requiring an a priori agreement regarding how success will be defined and whether those definitions of success can be attained within a reasonably powered study that can be completed within a period of time that makes the results of interest to the community. The ability to transport success to areas of the world in which advanced-stage invasive cervical cancer is more common may also frame the question of what steps are to be taken next.
Defining measures of success requires agreement concerning measurable endpoints and parameters. The SOTS meeting identified several elements that may have caused bias in the interpretation of recent studies. These include a shift of the inclusive population to one with better performance status, less and/or different disease burden, and varied treatment backbones. Harmonization of populations for inclusion, definitions, diagnostic tools, and treatment interventions are required. This is an area of attention of the Gynecologic Cancer InterGroup, their Cervical Cancer Research Network, and their partner cooperative groups, which will hopefully bring more institutions and countries into the cervical cancer research arena, thereby allowing outreach to more women at risk and among more diverse populations. It also will provide greater quality-control oversight of the treatment plans for those women, improve their care, and permit greater consistency in treatment paradigms.
Further questions remain regarding how to incorporate functional imaging into clinical trials, both to accurately determine disease extent and as an early and assessable surrogate marker for outcomes. In addition, the issue of how to incorporate molecular markers to predict response and toxicity remains largely unaddressed. Finally, we must not forget to address issues of supportive care, both during and after treatment, to better measure symptom burden and test novel modalities that will hopefully assist our patients live with and/or beyond their disease.
These are areas that may yield advances in our understanding of the biology of and outcomes in cervical cancer. Depending on the trial design, outcomes risk falling into the categories of incremental benefit, low cost/benefit balance, and/or complexity, thereby preventing transportability to the populations for whom greatest benefit may be obtained.
As discussed during this SOTS, we will need to advance our understanding of the biology of cervical cancer beyond HPV infection to the signaling pathways affected by HPV to apply novel targeted drugs in a scientifically rigorous fashion. Genomics data have documented KRAS and PI3KCA mutations, and are now awaiting proof of concept by demonstration that these pathways are both turned on and necessary for the cancer success. Endpoints including showing purposeful activation of MEK/ERK and AKT proliferation, invasion, and survival pathways would provide justification for testing active agents targeting these pathways. This will then lead to the need for novel and creative trial designs to allow for the testing of multiple agents in a cost-effective and time-effective fashion in this relatively rare patient population. Versions of multiarm, multistage statistical designs with stringent outcome expectations to allow smaller numbers of patients to enter clinical trials with adequate power were presented.
Therefore, where and how do we advance the treatment of cervical cancer? Clearly, the issues and needs of the world are diverse, varying from improving early detection to managing advanced disease. The developing world sorely needs resources and trained personnel for screening and prevention programs including prophylactic vaccination programs, and adequate resources and trained personnel to appropriately manage invasive cervical cancer. The type of trials required to optimize the prevention, early detection, diagnosis, and treatment are not typical classic treatment trials, but are those designed to evaluate cost-effectiveness, health policy, and public health in the broadest sense.
Conversely, the issues of the developed world extend beyond the constellation of needs to demand more emphasis on understanding disease and applying novel approaches to improve efficacy and/or reduce toxicity in the treatment of cervical cancer, both first-line therapy and for disease recurrence. One cannot ignore the fiscal and practical realities of embarking on large-scale trials to find small incremental benefits in an ever-shrinking prevalent population. Novel strategies using conventional drugs and RT will be difficult to study. However, we are primed to incorporate novel drugs such as immune modulators and biologics aimed at specific targets using small phase 2 studies testing both prognostic and predictive molecular and/or immune markers in enriched populations. Commonality with other HPV-caused malignancies, such as cancers of the oropharynx and anal canal, suggest there might be underdeveloped routes of collaboration. What we think of as a uniform diagnosis of cervical cancer may in fact be a constellation of diverse malignancies with different molecular markers, driven by different mechanisms. If so, we may need to explore and develop refinements in our ability to identify risk groups by molecular or gene profiles.
Small phase 2 studies of novel design to test both prognostic and predictive molecular or immune markers in enriched populations are the most likely next step. In considering these studies, the need to refine our definition of “clinical benefit” was raised as an important step. Researchers, clinicians, and scientists will have to come together in various forums with approaches and ideas that satisfy most if not all of the above and identify the most scientifically compelling and most public health-focused directions in which to concentrate our efforts. These are what we view as the challenge for the future development of cervical cancer clinical trials.