Maintenance therapy in ovarian cancer refers to a cohort of women achieving response to initial adjuvant chemotherapy who then go on to additional therapy in the hopes of extending time to recurrence or inducing a lasting remission. The concept is not new and retains its scientific and clinical relevance when one understands the natural history of this disease.
It has been long known that patients achieving clinical remission following surgery and chemotherapy (defined by normalization of CA-125, negative physical exam, and frequently, normal imaging) have a 60% to 70% chance of developing recurrent disease in the next 2 years.1 This poor predictive performance of our clinical parameters spawned increasing interest in performing operative assessment: the so-called second-look procedure.
Guidelines for an appropriate intraoperative assessment when doing this procedure were crafted to reduce the chance of a sampling error and dictated—in the absence of macroscopic disease—the resection of 20 to 30 random biopsies from high-risk areas of the abdomen. Despite identifying a cohort of women without macroscopically or microscopically positive disease (“pathologic” complete responders in approximately 50% of second-look procedures), the recurrence risk at 2 years was still 40%.2 This high false-negative rate both relegated routine second-look procedure to a research-only option and confirmed the need for some type of therapy to improve these sobering odds.
Randomized Trials
To date, no less than a dozen and a half randomized phase III trials have been conducted in women with ovarian cancer to address this risk.3 They have evaluated not only multiple different types of agents, including chemotherapy, immunotherapy, antiangiogenesis therapy, and anti–growth factor therapy, but also different delivery strategies, including intraperitoneal, intravenous, and radiation. Most have not demonstrated a prolongation in any survival endpoint.
The AGO-OVAR 16 trial, reported by du Bois and colleagues and reviewed in this issue of The ASCO Post, showed improvement in progression-free survival with pazopanib (Votrient) maintenance therapy.4 Including this trial, four antiangiogenesis studies (bevacizumab [Avastin],5,6 nintedanib,7 pazopanib4) and one chemotherapy study (paclitaxel8) have improved progression-free survival without improving overall survival.
The current trial, in addition to the trial of 1 year of paclitaxel, randomly assigned patients largely in clinical remission after induction therapy. This contrasts with the two bevacizumab trials and the nintedanib trial, in which randomization was performed before adjuvant chemotherapy was administered. Thus, AGO-OVAR 16 included the most favorable subgroup of patients, since they had reached eligibility by having a complete or near-complete clinical response to front-line therapy. Nevertheless, the result was similar to that observed in the other “positive” trials.
Study Context
Before assigning a value to the treatment outcome, it is important to consider the context in which this study was conducted. While the population odds for progression following a complete clinical remission are high, they are not absolute; some patients (approximately 25%–30%) will be cured following surgery and adjuvant chemotherapy. Individually, we are unable to prospectively identify who these women are; for them, maintenance therapy only produces toxicity and likely diminution of quality of life.
This same uncertainty can also influence investigators and patients to see treatment-related adverse events as more problematic than under conditions of treatment for existing disease. This can disproportionately drive treatment discontinuation, which occurred in the current trial in one-third of patients receiving pazopanib. Although the most common reason for discontinuation in this cohort was hypertension, it is a rate substantially higher than the rate seen in renal cancer patients treated with single-agent pazopanib.9
In addition, one could be critical of the initiating dose, as substantial modifications were required after cycle 1 (particularly in the East Asian cohort). These observations highlight that the tolerance “yardstick” is higher in the maintenance setting relative to other treatment settings.
Progression-Free Survival Issues
Another context to consider when reviewing these results is understanding the “value” of increasing progression-free survival. It is clear that one cannot make an asymptomatic patient more asymptomatic by administering therapy. Most women undergoing consideration of maintenance therapy are asymptomatic from a cancer standpoint. This fact implores investigators to carefully evaluate the impact of a prospective therapy on daily functionality.
Since it is likely that therapy will differentially affect those getting drug vs placebo, the magnitude and directionality are extremely important in understanding how an adverse event profile is tolerated. And since no measure of quality of life or patient-reported outcomes was reported in this initial analysis of AGO-OVAR 16, we are left to consider the survival data in a vacuum.
Perhaps even more revealing of the primary treatment effect is how these patients do into the next line of therapy. One effect of prolonging progression-free survival is a (desired) delay in the next line of therapy for those who ultimately experience a recurrence. We know from the current trial that more patients on the placebo arm (61% vs 50%) were undergoing subsequent anticancer therapy at the time of report. However, without a measure of their quality of life, it is unknown whether this prolongation in time to initiation of subsequent therapy among the pazopanib patients actually sustained their quality of life relative to those who had started anticancer therapy in the placebo arm.
It is unknown whether the toxicity experienced in patients treated with pazopanib, but who had not shown disease progression, was as bad as the toxicity in placebo-assigned patients now receiving chemotherapy. In this regard, a positive trend in quality of life would provide some context to the delay in progression.
Study Endpoints
An additional lens through which one should look at this trial is that of endpoints. The primary endpoint for this trial was progression-free survival; the investigators were attempting to increase the median progression-free survival by 47%. They largely achieved this (46% increase). However, more important than a point estimate is the hazard over exposure.
In this study, exposure to pazopanib (24 months of therapy) covered about 62% of the expected progression events (those expected without treatment, or placebo). This implies that prolonging therapy would be unlikely to add benefit to the primary endpoint.
In addition, while the study was initially designed to have 80% power to evaluate a 27% improvement in overall survival, the likelihood of preserving a progression-free survival benefit with an estimated 38-month postprogression survival is more likely less than 20%.10 Several factors contribute to this effect, such as uncontrolled use of additional therapy, therapeutic crossover, intent of additional therapy, subsequent treatment discovery, and surgery.
In addition, confounding effects are stronger the more proximal to diagnosis the intervention is studied. This additionally highlights the need for context when assessing progression-free survival as an endpoint. The Society of Gynecologic Oncology has recently produced a white paper and a guidance document on endpoints in ovarian cancer.11,12
Closing Thoughts
In all, GlaxoSmithKline has determined, despite meeting its primary endpoint, to suspend development of the agent in this setting. The adverse event profile, particularly among East Asian patients (which was associated with a statistical detriment in overall survival), appears to tip the risk-benefit ratio away from declaring the intervention a therapeutic advance. We await the quality-of-life and genomic-profiling data to better understand how the regimen was truly tolerated and whether there are intrinsic manifestations that shed light on the adverse events and efficacy observed. ■
Disclosure: Dr. Coleman reported no potential conflicts of interest.
References
1. Ozols RF, Bundy BN, Greer BE, et al: Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: A Gynecologic Oncology Group study. J Clin Oncol 21:3194-3200, 2003.
2. Greer BE, Bundy BN, Ozols RF, et al: Implications of second-look laparotomy in the context of optimally resected stage III ovarian cancer: A non-randomized comparison using an explanatory analysis: A Gynecologic Oncology Group study. Gynecol Oncol 99:71-79, 2005.
3. Coleman RL, Monk BJ, Sood AK, et al: Latest research and treatment of advanced-stage epithelial ovarian cancer. Nat Rev Clin Oncol 10:211-224, 2013.
4. du Bois A, Floquet A, Kim J-W, et al: Incorporation of pazopanib in maintenance therapy of ovarian cancer. J Clin Oncol 32:3374-3382, 2014.
5. Perren TJ, Swart AM, Pfisterer J, et al: A phase 3 trial of bevacizumab in ovarian cancer. N Engl J Med 365:2484-2496, 2011.
6. Burger RA, Brady MF, Bookman MA, et al: Incorporation of bevacizumab in the primary treatment of ovarian cancer. N Engl J Med 365:2473-2483, 2011.
7. du Bois A, Kristensen G, Ray-Coquard I, et al: AGO-OVAR 12: A randomized placebo-controlled GCIG/ENGOT intergroup phase III trial of standard frontline chemotherapy +/- nintedanib for advanced ovarian cancer. Int J Gynecol Cancer 23(8 suppl):Abstract, 2013.
8. Markman M, Liu PY, Wilczynski S, et al: Phase III randomized trial of 12 versus 3 months of maintenance paclitaxel in patients with advanced ovarian cancer after complete response to platinum and paclitaxel-based chemotherapy: A Southwest Oncology Group and Gynecologic Oncology Group trial. J Clin Oncol 21:2460-2465, 2003.
9. Sternberg CN, Davis ID, Mardiak J, et al: Pazopanib in locally advanced or metastatic renal cell carcinoma: Results of a randomized phase III trial. J Clin Oncol 28:1061-1068, 2010.
10. Broglio KR, Berry DA: Detecting an overall survival benefit that is derived from progression-free survival. J Natl Cancer Inst 101:1642-1649, 2009.
11. Herzog TJ, Alvarez RD, Secord A, et al: SGO guidance document for clinical trial designs in ovarian cancer: A changing paradigm. Gynecol Oncol 135:3-7, 2014.
12. Herzog TJ, Armstrong DK, Brady MF, et al: Ovarian cancer clinical trial endpoints: Society of Gynecologic Oncology white paper. Gynecol Oncol 132:8-17, 2014.
Dr. Coleman is Professor and Deputy Chair, Department of Gynecologic Oncology & Reproductive Medicine, The University of Texas MD Anderson Cancer Center, Houston.
