Special Feature
Primary Ovarian Surgical Cytoreduction: When? How Much? Why?
By Robert L. Coleman, MD, Professor, University of Texas; M.D. Anderson Cancer Center, Houston, TX, is Associate Editor for OB/GYN Clinical Alert.
Dr. Coleman reports no financial relationship to this field of study.
Nearly every reference source discussing the primary management of advanced ovarian cancer will highlight the role of surgical cytoreduction as an important first step in the care of this disease. Citing control of symptoms, improved chemotherapy exposure, reduction of potentially chemoresistant tumor cell volumes, and favorable immunological effects, advocates of the approach further support their position by highlighting improvement in nearly every outcome parameter that has been addressed when primary surgery (PS) is performed and performed well. These include higher rates of tumor response, more frequent negative second look surgery, and longer progression-free (PFS) and overall survival (OS). They will further highlight that alternative approaches have not yielded better outcomes, and in many reports, have reported worse outcomes. These approaches include a secondary attempt at cytoreduction following an initial maximal attempt, "interval cytoreduction" (IC), and a pre-emptive approach where chemotherapy following biopsy to establish the diagnosis is administered ahead of a maximal surgical, "neoadjuvant chemotherapy" (NACT). Since ovarian cancer is a chemosensitive disease, post-chemotherapy disease volumes are likely to be significantly reduced providing both a better opportunity to achieve a more complete surgical resection of remaining disease and do so with reduced morbidity. Advocates of the neoadjuvant approach further highlight that induction chemotherapy helps identify those women unlikely to benefit from surgical exploration, as it is unlikely that a woman will gain any increase in her clinical situation with disease not responsive to chemotherapy. Since the single most important independent predictor of overall survival in chemo-naïve patients is the volume of postoperative residual disease, advocates of each approach bring the increased likelihood of complete surgical resection, that is, removal of all visible tumor, to the debate.
Globally, there is wide variation in the proportion of patients considered for each of these approaches, but the latter, NACT, has been more uniformly practiced in patients with poor performance status, or in those with extensive unresectable intra-abdominal or extra-abdominal parenchymal metastatic disease. As a result, comparative data between series of patients categorized by surgical approach are heavily biased and difficult, if not impossible, to interpret. Several meta-analyses have been done in this respect, reaching alternative conclusions. However, in no case has the NACT approach been found to be superior to the primary surgical approach. Nevertheless, in the absence of a randomized phase III trial comparing PS to NACT, the debate was largely framed by inferences from predominately retrospective and a few prospective non-randomized clinical trials, with no clear consensus as to the appropriateness of the approach.
Interval Cytoreduction
Before discussing PS and NACT and the study by Vergote and colleagues that addressed this question,1 a review of the intermediate approach, IC, is warranted. In this approach, patients in whom a maximal attempt at initial surgery was performed and was unsuccessful in removing disease bulk to less than 1 cm are treated with chemotherapy for (usually) 3 cycles and then returned for a second surgery to remove any remaining disease. Following the second surgery, patients are usually returned to chemotherapy for 3-6 additional cycles (see Figure 1, below). There have been two randomized phase III studies that have been reported addressing this approach. In the first, 425 women with suboptimally cytoreduced (> 1 cm tumor residuum) stage IIIC and IV disease were treated with cyclophosphamide and cisplatin for 3 cycles and, if not progressive (319; 75%), randomized to a second surgical attempt vs continued therapy. In the second, 550 women were registered and treated with paclitaxel and cisplatin followed by randomization (425; 77%) to a secondary surgery or continued chemotherapy. Figure 2 (below) highlights the features and outcomes of these two trials. Since the GOG trial enrolled a substantially higher proportion of patients who underwent a maximal initial cytoreductive effort, and in light of its use of contemporary chemotherapy, the GOG trial is considered by most as the definitive trial of the IC approach. As a result, it is widely thought that an unsuccessful maximal attempt at primary surgery by a qualified gynecologic oncologist is reflective of a disease biology that is also characteristic of poor chemosensitivity. As a result, this approach is rarely practiced in the United States.
Comparing PS and NACT
Recently, the first randomized phase III comparing the other two approaches, PS and NACT, was published. This highly anticipated clinical trial's results have been in the public domain and under publication consideration for approximately 2 years, which highlights the peer-reviewed scrutiny the manuscript received before its publication in the September 2010 issue of the New England Journal of Medicine. The study was a collaborative effort of two independently funded trial organizations, the EORTC-Gynaecological Cancer Group (EORTC-GCG) and National Cancer Institute of Canada (NCIC) Clinical Trials Group. Eligible patients were those with biopsy-proven stage IIIC and IV disease or those with a fine needle aspirate and additional evidence of metastatic disease including a biomarker profile (CA125:CEA > 25) favoring mullerian origin. The study design was 1:1 randomization of eligible patients to either primary surgical cytoreduction or neoadjuvant chemotherapy, which was defined as 3 cycles of paclitaxel and carboplatin at standard doses and infusion schedule. Planned adjuvant therapy in the PS arm was at least 6 cycles of paclitaxel and carboplatin; in the NACT arm, surgical cytoreduction was performed in the absence of disease progression after 3 cycle of paclitaxel and carboplatin, followed by surgical debulking and at least 3 additional cycles of chemotherapy (see Figure 3). A provision was made early on in the trial that suboptimally cytoreduced PS patients could undergo an interval secondary surgery; however, upon the revelation of the data outlined above, this was subsequently removed. Of 670 randomized patients, 632 were ultimately randomized. The population had extensive disease burden with 62% of patients harboring metastatic lesions of 10 cm or greater. Optimal cytoreduction (defined as postoperative tumor residuum of 1 cm or less) was achieved in 42% of PS patients and 81% of NACT patients. Postoperative adverse events rates (hemorrhage, infection, venous thrombosis) and mortality were significantly higher after PS compared to NACT. The statistical endpoint was non-inferiority; in other words, the trial was powered to address that the NACT approach was not more than 25% inferior to standard PS in overall survival. The hazard ratio for death between the arms was 0.98 (90% confidence interval [CI], 0.84-1.13; P = 0.01 for non-inferiority), rejecting the null hypothesis that NACT was inferior to PS. The same was true for progression (hazard ratio, 1.01; 90% CI, 0.89-1.15). As has been documented previously, resection of all macroscopic disease was the strongest independent predictive factor for overall survival. The authors concluded that while NACT is a viable alternative to PS in patients with bulky stage IIIC-IV disease, complete tumor resection should be the goal of surgery, whether performed initially or after NACT.
The Controversy
The study is a tribute to the cooperative group charter and their ability to address questions that would be otherwise impossible to formally answer. Further, since the trial includes centers in multiple geographic locations and consists of surgeons of various skill sets, the results are ultimately generalizable to those who would fit the eligibility criteria. So why the controversy? The root of the current debate lies not in the conduct of the trial or the comparative nature as designed. There fundamentally is even little controversy in the conclusions; the focus of the debate lies in the observed median outcomes in both arms. That is, a median PFS of 12 months and OS around 30 months are far lower than what would be expected in this population (Stage IIIC-IV) and suggests the surgical effort in the control group is substandard. In fact, GOG protocol 111, a randomized clinical trial of cyclophosphamide and cisplatin vs paclitaxel and cisplatin in exclusively suboptimally cytoreduced ovarian cancer patients (> 1 cm residual disease) had a median PFS of 18 months and an OS of 38 months. Even the mixed population of GOG protocol 182 provided a median PFS of 16 months and OS of 44 months in more than 4300 patients from U.S. and European centers. As a result, the optimal debulking rates by country were evaluated and, indeed, there was variation in optimal rates ranging from around 40% to more than 80%. However, there was strong correlation between the optimal debulking rates within centers for both PS and NACT, which may explain the consistency of the results. In all, no differences could be discerned between participating sites, patients, and outcomes that could explain the absolute values.
An exhaustive subgroup analysis was provided by the study organization as a supplement to the article to address questions about the trial's results that have been raised since initial presentation. However, this has done little to quench the thirst for re-evaluation in centers or organizations with higher rates of optimal resection rates, and potentially among patients with more favorable risk profiles. Recent documentation of the frequency and impact of complete tumor resection may lend credence to this effort. While there is variance among institutions in the rate of PS optimal debulking defined by the 1 cm yardstick, the impact on overall survival is far lower than the impact of zero tumor residual, which varies by an even greater degree globally. In addition, this effect on survival could have been diluted in the current study as zero tumor residual was accomplished more frequently after NACT in centers with low rates of zero tumor residual after PS. In this fashion, the impact of the strongest predictive variable may not have been realized in the PS cohort.
Conclusion
The take home message from this trial and discussion is that the NACT approach, which heretofore had been relegated predominately to our poorest prognosis and sickest patients, indeed has a viable role in the management of ovarian cancer. This is not dissimilar to other solid tumors, such as colorectal and breast cancer, where the approach has also been associated with organ preservation and a prime opportunity to obtain tissue for post-therapeutic biomarker assessment of effect. This valuable asset will ultimately lead to a more "personalized" or directed therapy following surgery, which could greatly enhance treatment effect. However, until data from a NACT approach the expected norms in the United States following PS, the controversy surrounding a potential underappreciated adverse effect on PFS and OS will loom. Unfortunately, it is not clear that any trial of this sort will ever be accomplished in the United States, though a trial is currently being conducted in Europe. Until then, a bias will continue to drive primary treatment approaches in ovarian cancer.
Reference
- Vergote I, et al. Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer. N Engl J Med 2010;363:943-953.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content