J Clin Oncol 26: by American Society of Clinical Oncology INTRODUCTION

VOLUME 26 NUMBER 1 JANUARY 1 2008 JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T Tumor Residual After Surgical Cytoreduction in Prediction of Clinical Outcome in Stage IV Epithelial Ovarian Cancer: A Gynecologic Oncology Group Study William E. Winter III, G. Larry Maxwell, Chunqiao Tian, Michael J. Sundborg, G. Scott Rose, Peter G. Rose, Stephen C. Rubin, Franco Muggia, and William P. McGuire From the Department of Obstetrics and Gynecology, Gynecologic Oncology, Brooke Army Medical Center, Ft Sam Houston, TX; Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, Walter Reed Army Medical Center, Washington, DC; Gynecologic Oncology Group Statistical and Data Center, Roswell Park Cancer Institute, Buffalo; Division of Medical Oncology, New York University Medical Center, New York, NY; Department of Obstetrics and Gynecology, MetroHealth Medical Center, Cleveland, OH; Department of Gynecologic Oncology, University of Pennsylvania Cancer Center, Philadelphia, PA; and Harry and Jeanette Weinberg Cancer Institute, Franklin Square Hospital Center, Baltimore, MD. Submitted June 27, 2007; accepted September 25, 2007; published online ahead of print at www.jco.org on November 19, 2007. Supported by National Cancer Institute grants to the Gynecologic Oncology Group Administrative Office (Grant No. CA 27469) and the Gynecologic Oncology Group Statistical Office (Grant No. CA 37517). Authors disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: G. Larry Maxwell, MD, Division of Gynecologic Oncology, Walter Reed Army Medical Center, Rm 6743, Bldg 2, 6900 Georgia Ave, Washington, DC 20015; e-mail: george.maxwell@na.amedd.army.mil. 2008 by American Society of Clinical Oncology 0732-183X/08/2601-83/$20.00 DOI: 10.1200/JCO.2007.13.1953 A B S T R A C T Purpose To identify factors predictive of poor prognosis in a similarly treated population of women with stage IV epithelial ovarian cancer (EOC). Patients and Methods A retrospective review of 360 patients with International Federation of Gynecology and Obstetrics stage IV EOC who underwent primary surgery followed by six cycles of intravenous platinum/ paclitaxel was performed. A proportional hazards model was used to assess the association of potential prognostic factors with progression-free survival (PFS) and overall survival (OS). Results The median PFS and OS for this group of stage IV ovarian cancer patients was 12 and 29 months, respectively. Multivariate regression analysis revealed that histology, malignant pleural effusion, intraparenchymal liver metastasis, and residual tumor size were significant prognostic variables. Whereas patients with microscopic residual disease had the best outcome, patients with 0.1 to 1.0 cm residual disease and patients with 1.1 to 5.0 cm residual disease had similar PFS and OS. Patients with a residual size more than 5 cm had a diminished PFS and OS when compared with all other groups. Median OS for microscopic, 0.1 to 5.0 cm, and more than 5.0 cm residual disease was 64, 30, and 19 months, respectively. Conclusion Patients with more than 5 cm residual disease have the shortest PFS and OS, whereas patients with 0.1 to 1.0 and 1.1 to 5.0 cm have similar outcome. These findings suggest that ultraradical cytoreductive procedures might be targeted for selected patients in whom microscopic residual disease is achievable. Patients with less than 5.0 cm of disease initially and significant disease and/or comorbidities precluding microscopic cytoreduction may be considered for alternative therapeutic options other than primary cytoreduction. J Clin Oncol 26:83-89. 2008 by American Society of Clinical Oncology INTRODUCTION Epithelial ovarian cancer (EOC) is the leading cause of death among gynecologic malignancies in the United States. 1 Approximately 15% of EOC is diagnosed as International Federation of Gynecology and Obstetrics stage IV disease. Overall, median survival for patients with stage IV disease is approximately 15 to 23 months, 2-5 with an estimated 5-year survival of 20%. 2,5 Previous studies of prognostic factors in stage IV EOC have focused on surgical cytoreduction and the association between residual disease size and clinical outcome. 2,3,5-8 Unfortunately, interpretation is difficult because of nonuniform study design and analysis, and the retrospective nature of the studies. Consequently, controversies abound on the optimal management of stage IV EOC, specifically in the surgical management of these patients. Many of the studies have used less than 1 cm or less than 2 cm in defining optimal residual, but others have used from less than 0.5 cm to less than 3.0 cm. Many of the previous studies also have included patients receiving adjuvant chemotherapy other than platinum and paclitaxel-based regimens that are considered to be a contemporary standard. Finally, many of the studies have pooled both stage III and IV patients in the analysis. These two stages represent two different groups of patients warranting more individualized treatment decisions. 9 During the last 13 years, the Gynecologic Oncology Group (GOG) has conducted four 83

Winter et al randomized phase III trials that included patients with stage IV EOC treated with intravenous chemotherapy consisting of a platinum-based agent and paclitaxel after primary surgical cytoreduction. We used this large and similarly treated group of patients with stage IV EOC to perform a retrospective analysis of the pooled data aimed at identifying independent prognostic factors and evaluating the clinical impact of residual disease size, in particular, after surgical cytoreduction. PATIENTS AND METHODS The current study was a retrospective review of data from patients with stage IV EOC treated with platinum and paclitaxel combination chemotherapy on one of four prospective randomized clinical trials conducted by the GOG (Appendix, online only): protocols 111, 10 132, 11 152, 12 and 162. 13 Patients enrolled onto the experimental arm of GOG 111 and the control arms of GOG 132, 152, and 162 were treated with primary surgical cytoreduction and six cycles of a 24-hour infusion of intravenous paclitaxel 135 mg/m 2, followed by intravenous cisplatin 75 mg/m 2. We also included the experimental arms of GOG 152 and 162 because each of these trials demonstrated statistically equivalent clinical outcomes between the control and experimental arms and both contained a platinum and paclitaxel-based regimen. Table 1 lists the treatment regimens from each protocol that were included in this study. Details regarding eligibility criteria, treatment, and outcome for each particular study have been published previously. Patients provided written informed consent consistent with all federal, state, and local requirements before receiving protocol therapy. The primary end points for all four studies were progression-free survival (PFS) and overall survival (OS). PFS was calculated from the date of study enrollment to the date of disease recurrence (confirmed on physical, serologic, or radiologic examination), death, or most recent follow-up visit. OS was calculated from the date of study enrollment to the date of death or last contact. Laboratory and clinical data were collected at specified intervals per protocol. In addition to GOG surgical reporting forms, all operative, radiology, and pathology reports were reviewed to abstract specific data regarding surgical procedures, stage IV disease site(s), largest diameter of intraoperative disease and postoperative residual disease (intra- and extraperitoneal), and site of largest diameter of postoperative residual disease. Of note, visible gross tumor residual size less than 1.0 cm was recorded in 0.1-cm increments based on operative reports and GOG surgical reporting forms. All hospital and treatment summaries were reviewed to collect data on length of hospital stay and postoperative complications. Complications included length of stay more than 15 days; unplanned intensive care unit admission; transfusion of more Table 1. GOG Treatment Protocols and Study Populations GOG Protocol No. Patient Eligibility Treatment Regimen 111 Suboptimal ( 1 cm residual) stage III/IV EOC 132 Suboptimal ( 1 cm residual) stage III/IV EOC 152 Suboptimal ( 1 cm residual) stage III/IV EOC 162 Stage III ( 1cmRP; 1cm IP residual) and stage IV EOC/PSPC and cisplatin 75 mg/m 2 for 6 cycles and cisplatin 75 mg/m 2 for 6 cycles and cisplatin 75 mg/m 2 for 6 cycles interval debulking 24- or 96-hour infusion, cisplatin 75 mg/m 2 for 6 cycles No. of Stage IV Patients 59 54 23 224 Abbreviations: GOG, Gynecologic Oncology Group; EOC, epithelial ovarian cancer; PSPC, papillary serous carcinoma; RP, retroperitoneal sites; IP, intraperitoneal sites. Includes 4 patients with PSPC. than 4 units of packed red cells; estimated blood loss of more than 2,000 ml; GI anastomotic leaks; urinary tract injury; neural injury; vascular injury; and unplanned return to the operating room, readmission, or death within 30 days of primary surgery. Clear-cell carcinomas were assigned grade 3. Baseline performance status (PS) before initiating chemotherapy was defined according to GOG criteria (0 normal activity; 1 symptomatic, fully ambulatory; 2 symptomatic, in bed less than 50% of the time). The Cox proportional hazards model was used to identify the independent prognostic factors as well as to estimate their effects on PFS and OS adjusted for covariates. Kaplan-Meier survival curves based on residual disease status and other clinical factors were calculated and compared using the Table 2. Patient Demographic and Clinical Characteristics (N 360) Characteristic No. of Patients % Age, years Median 58.8 Range 24.4-85.8 50 97 26.9 50-59 89 24.7 60-69 106 29.4 70 68 18.9 Race White 317 88.1 Black 28 7.8 Other 15 4.2 GOG performance status 0 97 26.9 1 203 56.4 2 60 16.7 Tumor grade 1 24 6.7 2 112 31.1 3 224 62.2 Serous 268 74.4 Endometrioid 28 7.8 Mucinous 7 1.9 Clear cell 12 3.3 Adenocarcinoma unspecified 9 2.5 Mixed epithelial 22 6.1 Undifferentiated 9 2.5 Other 5 1.4 Residual size, cm Median 3.0 Range 0.0-40.0 0.0 (no visible) 29 8.1 0.1-1.0 78 21.7 1.1.1-2.0 50 13.9 2.1-3.0 40 11.1 3.1-4.0 30 8.3 4.1-5.0 44 12.2 5.1-6.0 30 8.3 6.0 59 16.4 Distant 45 12.5 Parenchymal liver 64 17.8 Pleural effusion 172 47.8 Subcutaneous 32 8.9 Others 3 0.8 Multiple sites 44 12.2 Abbreviation: GOG, Gynecologic Oncology Group. 84 JOURNAL OF CLINICAL ONCOLOGY

Stage IV Ovarian Cancer log-rank test. All statistical tests were two tailed with a significance level set at 5%. All statistical analyses were performed on Statistical Analysis Software, version 9.1 (SAS Institute, Cary, NC). RESULTS Data from the 360 patients with stage IV invasive EOC who underwent primary surgical cytoreduction followed by paclitaxel/platinum chemotherapy while participating in one of four GOG clinical trials (Table 1) were analyzed for the present study. Table 2 lists the demographic and clinical characteristics of the study population. The median age of patients was 59 years (range, 24 to 86 years) and the majority were white (88%), with a pretreatment GOG performance status of 0 to 1 (83%). Clinically, 74% of them were serous cell type and 62% were classified as tumor grade 3. The median size of largest residual disease (intra- and extraperitoneal) was 3.0 cm; only 8% of patients had microscopic (no visible) disease and almost one fourth had more than 5 cm of residual disease. Malignant pleural effusion was the most common site of stage IV disease (48%). Univariate Analysis Three hundred four recurrences and 320 deaths were identified within a median follow-up period of 28 months. The median PFS was Table 3. Median PFS and OS by Patient Characteristics No. of PFS OS Characteristic Patients Median (months) P Median (months) P Age group, years 50 97 12.9.61 32.2.68 50-59 89 12.3 26.6 60-69 106 12.7 28.5 70 68 9.8 25.6 Race White 317 12.6.08 28.1.84 Black 28 12.5 31.4 Other 15 5.7 32.2 GOG performance status 0 97 12.9.57 31.2.18 1 203 12.4 28.1 2 60 12.3 22.1 Serous 268 13.0.02 29.2.001 Endometrioid 28 10.7 28.2 Clear cell/mucinous 19 4.9 9.8 Mixed epithelial 22 13.8 34.9 Others 23 11.2 25.0 Tumor grade 1 24 17.5.22 28.5.44 2 112 12.3 31.2 3 224 12.2 26.6 Tumor residual, cm Incremental disease residual 0.0 (no visible) 29 20.1.003 64.1.002 0.1-1.0 78 13.0 28.7 1.1-2.0 50 12.9 25.1 2.1-3.0 40 13.8 33.6 3.1-4.0 30 13.2 36.7 4.1-5.0 44 12.2 26.9 5.1-6.0 30 7.0 16.2 6.0 59 8.9 22.6 Categorical disease residual 0.0 (no visible) 29 20.1.0007 64.1.0006 0.1-1.0 78 13.0 28.7 1.1-5.0 164 12.9 30.7 5.0 89 8.5 19.5 Distant 45 13.4.16 38.7.005 Parenchymal liver 64 10.6 20.7 Pleural effusion 172 13.0 28.5 Subcutaneous/other 35 11.7 37.9 Multiple sites 44 9.5 24.1 Abbreviations: PFS, progression-free survival; OS, overall survival; GOG, Gynecologic Oncology Group. www.jco.org 85

Winter et al 12.3 months (95% CI, 11.3 to 13.2 months). Forty-eight percent of patients had disease progression within the first year and only 19% remained progression free by the end of the second year. Univariate analysis, which included age, race, GOG PS, histology, tumor grade, residual disease size, and stage IV sites (Table 3), revealed that residual disease, multiple stage IV sites, and mucinous/clear cell histology were associated with poor PFS and OS (P.05). A significant association between liver metastases and decreased OS was also identified (P.05). Other variables, including performance status, tumor grade, age, and race, did not appear to have a significant association with clinical outcomes. The effects of residual disease were assessed in 1-cm increments from microscopic disease to more than 6 cm. Gross residual disease was found to be associated with decreased PFS and OS. The correlation between residual disease size and median PFS and OS seemed to segregate into three categories: microscopic, 0.1 to 5 cm, and more than 5 cm of residual disease. Multivariate Regression Analysis All variables considered as potential prognostic factors (ie, age, race, GOG PS, tumor grade, histology, residual tumor size, and stage IV site) were included in a Cox proportional hazards multivariate regression model for preassessment. Age, race, GOG PS, and tumor grade were excluded secondary to lack of independent association with clinical outcomes. The final model included three independent variables: histology, stage IV site, and residual disease size. These results are listed in Table 4. When patients with microscopic disease were used as the referent, the adjusted hazard ratio (HR) for disease progression in patients with 0.1 to 5.0 and more than 5.0 cm was 2.09 (95% CI, 1.36 to 3.21; P.0008) and 2.95 (95% CI, 1.85 to 4.69; P.0001), respectively, and the relative risk between more than 5.0 cm and 0.1 to 5.0 cm residual disease size was 1.41 (95% CI, 1.08 to 1.84; P.01). The overall survival results were similar to those shown for PFS. The adjusted HR for death in patients with 0.1 to 5.0 and more than 5.0 cm was 1.87 (95% CI, 1.18 to 2.96; P.008) and 2.73 (95% CI, 1.66 to 4.49; P.0001), respectively, and the group with more than 5.0 cm also showed significantly increased risk for death compared with the group with 0.1 to 5.0 cm (HR 1.46; 95% CI, 1.11 to 1.92; P.006). Analysis of the data was also performed using both 1 and 2 cm in defining optimal cutoff residual tumor size. Our results confirmed data reported in previous investigations using these thresholds for optimal disease in their analysis (2 to 3; 5; 7 to 8 cm). Though these analyses using 1- and 2-cm cutoffs were suggestive of this relationship, the association between tumor residual size and clinical outcome was best illustrated by the three-tier classification as delineated in the univariate analysis. Furthermore, the prognosis for those patients with residual size of 0.1 to 1.0 cm and 1.0 to 5.0 cm was comparable (HR 1.08; 95% CI, 0.80 to 1.45; P.61 for PFS v HR 0.95; 95% CI, 0.70 to 1.29; P.73 for OS using residual size of 0.1 to 1.0 cm as the referent). Patients with residual tumor more than 5.0 cm had an increased risk of disease progression (HR 1.56; 95% CI, 1.20 to 2.03, P.0008) and death (HR 1.57; 95% CI, 1.20 to 2.05; P.001) when compared with all other patients. Of note, there was no difference in outcomes if the suboptimal disease was found at intra- and/or extraperitoneal (or intraparenchymal) sites (data not shown). Only those with multiple stage IV disease sites had a borderline increased risk of disease progression (HR 1.54; 95% CI, 0.99 to 2.40; Prognostic Factor Table 4. Multivariate Analysis of Prognostic Factors PFS HR 95% CI P HR 95% CI P Serous 1.0 1.0 Endometrioid 1.44 0.95 to 2.18.09 1.48 0.98 to 2.25.06 Clear cell/mucinous 2.34 1.43 to 3.85.0008 2.34 1.40 to 3.92.001 Mixed epithelial 0.76 0.47 to 1.22.26 0.77 0.47 to 1.25.28 Other 0.97 0.61 to 1.53.88 1.07 0.67 to 1.70.78 Distant 1.0 1.0 Parenchymal liver 1.22 0.81 to 1.86.34 1.68 1.07 to 2.63.02 Pleural effusion 1.18 0.83 to 1.69.35 1.73 1.17 to 2.54.006 Subcutaneous/other 1.10 0.68 to 1.80.69 1.30 0.78 to 2.15.31 Multiple sites 1.54 0.99 to 2.40.05 2.24 1.40 to 3.59.0008 Tumor residual, cm Optimal: microscopic residual 0.0 (no visible) 1.0 1.0 0.1-1.0 1.99 1.24 to 3.18.004 1.93 1.17 to 3.20.01 1.0-5.0 2.15 1.38 to 3.34.0007 1.83 1.14 to 2.94.01 5.0 2.96 1.86 to 4.71.0001 2.72 1.65 to 4.47.0001 Optimal: 1.0 cm residual 1.0 1.0 1.0 1.0 1.49 1.16 to 1.92.002 1.30 1.00 to 1.59.05 Optimal: 2.0 cm 2.0 1.0 1.0 2.0 1.27 1.01 to 1.61.04 1.17 0.92 to 1.49.21 Abbreviations: PFS, progression-free survival; OS, overall survival. OS 86 JOURNAL OF CLINICAL ONCOLOGY

Stage IV Ovarian Cancer P.05). Patients with parenchymal liver metastases, pleural effusions, or multiple stage IV disease sites experienced decreased survival (HR 1.68; 95% CI, 1.07 to 2.63; P.02 v HR 1.73; 95% CI, 1.17 to 2.54; P.006 v HR 2.24; 95% CI, 1.40 to 3.59; P.0008, respectively) relative to patients with distant site disease (eg, supraclavicular/axillary lymphadenopathy, parenchymal lung metastases, mediastinal adenopathy, and distal vaginal/perineal metastases). Mucinous/clear cell tumors were predictive of increased disease progression (HR 2.34; 95% CI, 1.43 to 3.85; P.0008) and poor survival (HR 2.34; 95% CI, 1.40 to 3.92; P.001). PFS and OS based on residual disease size are also illustrated in Figure 1. The overall median PFS and OS for stage IV patients were 12 and 28 months, respectively. The estimated 3-year PFS and OS were 13% and 38%, respectively. The median PFS (Fig 1A) for patients with microscopic residual disease (20 months) was superior to that in patients with any amount of gross residual disease. Although the median PFS was 13 months for patients with 0.1 to 1.0 and 1.1 to 5.0 cm of residual disease, patients with more than 5.0 cm of residual disease had a shorter median PFS compared with all other groups (8.5 months). Similar results were seen with OS (Fig 1B), where the median OS for patients with microscopic residual disease (64 months) was superior to that in patients with any amount of gross residual disease. Patients with 0.1 to 1.0 and 1.1 to 5.0 cm of residual had statistically similar median OS: 29 and 31 months, respectively. Patients with more than 5.0 cm of residual disease had a shorter median OS compared with all other groups (19 months). These associations were consistent over the time period of the protocols included in this study. When evaluating the association of clinicopathologic factors with residual disease status, there was no difference between the residual disease groups and demographic, clinical, and pathologic factors (Table 5). Our data also revealed that variations in the size of residual tumor (ie, microscopic, 0.1 to 5.0, and 5 cm residual disease) were not associated with an increased risk of perioperative complications (ie, length of stay 15 days, estimated blood loss 2,000 ml, transfusion 4 units, urinary tract infection, neural injury, vascular injury, unplanned admission to the intensive care unit, unplanned readmission 30 days, unplanned return to OR 30 days, or death 30 days). In addition, the stage IV site did not seem to have significant association with residual disease group distributions. DISCUSSION A PFS Probability B OS Probability 1.0 0.8 0.6 0.4 0.2 0 1.0 0.8 0.6 0.4 0.2 0 6 12 18 24 30 36 42 48 54 60 Time on Study (months) Time on Study (months) 0 cm 0.1-1.0 cm 1.1-5.0 cm > 5.0 cm 0 cm 0.1-1.0 cm 1.1-5.0 cm > 5.0 cm 6 12 18 24 30 36 42 48 54 60 Fig 1. Kaplan-Meier estimate of (A) progression-free survival (PFS) and (B) overall survival (OS) by residual tumor size. Although several series have evaluated prognostic factors in stage IV EOC, 2-8,14 to our knowledge, this study represents the largest investigation of stage IV EOC patients reported to date. Only women receiving adjuvant chemotherapeutic regimens that consisted of platinum with paclitaxel were included in this analysis to provide conclusions that would be applicable to patient care in a contemporary practice setting. The strength of this study lies in the large size and similarity of the population. Adherence to strict protocol regimens of primary surgical assessment and cytoreduction of disease burden and postoperative therapy provides enhanced statistical power to detect independent variables predictive of poor outcome and a precise estimation of the magnitude of the effect while controlling for confounding influences. Variation in the method of defining optimal residual disease also has led to the variable degree of association between residual disease size and clinical outcome. Investigators using 1 cm as the defining size for optimal residual disease have reported median survival of 38 months, compared with 10 months for those with suboptimal residual disease size (P.05). 3 Curtin et al 2 used 2 cm as a threshold for defining optimal residual disease and also observed an improved median survival in stage IV patients with less than 2 cm residual disease (40 months) compared with those with 2 cm residual disease (18 months; P.01). Using multivariate analysis, we evaluated the data using either 1 or 2 cm as a threshold in defining optimal disease. The risk of recurrence was almost 50% greater among patients with more than 1 cm residual disease compared with patients with less than 1 cm residual disease, and overall survival was also significantly lower in suboptimally debulked patients defined using this threshold (HR 1.30; 95% CI, 1.00 to 1.59; P.05). Analysis of the data using less than 2 cm as a defining measure of optimal residual disease revealed that patients with suboptimal residual disease size had a decreased PFS but there was no significant association with decreased OS. Although our initial analysis had confirmed the historic data, we subsequently chose to evaluate the data without any preconceived plans for dichotomization at a previously reported threshold www.jco.org 87

Winter et al Characteristic Table 5. Patient Characteristics by Residual Disease Residual Disease (cm) 0.0 (no visible) 0.1-5.0 5.0 No. % No. % No. % Age group, years 50 12 41.4 68 28.1 17 19.1.36 50-59 5 17.2 61 25.2 23 25.8 60-69 8 25.6 68 28.1 30 33.7 70 4 13.8 45 18.6 19 21.4 Race White 27 93.1 214 88.4 76 85.4.35 Black 0 0.0 18 7.4 10 11.2 Other 2 6.9 10 4.1 3 3.4 GOG performance status 0 11 37.9 66 27.3 20 22.5.10 1 18 62.1 134 55.4 51 57.3 2 0 0.0 42 17.4 18 22.2 Serous 18 62.1 190 78.5 60 67.4.20 Endometrioid 3 10.3 17 7.0 8 9.0 Clear cell/mucinous 4 13.8 8 3.3 7 7.9 Mixed epithelial 2 6.9 12 5.0 8 9.0 Other 2 6.9 15 6.2 6 6.7 Tumor grade 1 3 10.3 13 5.4 8 9.0.71 2 9 31.0 77 31.8 26 29.2 3 17 58.6 152 62.8 55 61.8 Distant 4 13.8 35 14.5 6 6.7.16 Parenchymal liver 4 13.8 37 15.3 23 25.8 Pleural effusion 17 58.6 116 47.9 39 43.8 Subcutaneous/other 3 10.3 21 8.7 11 12.4 Multiple sites 1 3.5 33 13.6 10 11.2 Abbreviation: GOG, Gynecologic Oncology Group. P of optimal residual. Median PFS and OS were noted to cluster into three groups based on size of residual disease: microscopic (0 cm), 0.1 to 5.0 cm, and more than 5 cm. Hazard proportions analysis confirmed that both PFS and OS were significantly worse in patients with either 0.1 to 5 cm or more than 5 cm compared with patients with microscopic disease (Table 4). In the illustration of survival data, we chose to show survival for microscopic disease, 0.1 to 1.0 cm, 1.1 to 5.0 cm, and more than 5 cm in an effort to further evaluate the outcome among patients with 0.1 to 1.0 cm and 1.1 to 5 cm residual disease. Neither disease progression nor survival was different between patients with 0.1 to 1.0 cm and those with 1.1 to 5.0 cm of residual disease. Patients with more than 5 cm of residual disease had both worse PFS and OS compared with any other group (Fig 1). Our data revealed that variations in the degree of size of residual tumor were not associated with an increased risk of complications, suggesting that even with potential selection bias, there does not seem to be an increased level of mortality associated with cytoreduction to microscopic disease. Inconsistent intraoperative descriptions of disease size and distribution before and after cytoreduction prevented us from more precisely assessing the proportion of tumor removed as a result of surgical cytoreduction for all of the patients. However, only 19, 18, and five patients had 50%, 25%, and 1% of the tumor burden removed, respectively, based on GOG surgical reporting forms. Thus, almost 90% of patients underwent removal of more than 50% of their tumor burden, suggesting that significant proportions of disease volume were removed as a result of cytoreductive efforts. Although this study did not evaluate prospectively the association of surgical cytoreduction with survival, the data still provide provocative evidence to suggest that the traditional surgical approach of primary cytoreduction to less than 1 cm in stage IV disease should be re-evaluated in an era of platinum and paclitaxel-based chemotherapy. For instance, patients with more than 5 cm of disease may benefit from surgery that is tailored according to whether the patient can undergo cytoreduction to microscopic disease or 0.1 to 5 cm residual disease. Ultraradical procedures might be justified in cases in which the patient can be completely debulked to microscopic disease. More tempered surgical cytoreduction aimed at reducing the size of the largest residual implant to less than 5 cm may be considered for those individuals in whom resection to microscopic disease cannot be achieved safely, given that the patient s outcome seems to be similar irrespective of whether they undergo resection to either 0.1 to 1.0 or 1.1 to 5.0 cm. Patients presenting with 0.1 to 5 cm disease size might be considered for a diagnostic surgical procedure, such as laparoscopy, hand-assisted laparoscopy, or minilaparotomy, in an attempt to best assess resectability. Patients with disease that is considered safely resectable to microscopic disease may then undergo laparotomy 88 JOURNAL OF CLINICAL ONCOLOGY

Stage IV Ovarian Cancer and complete surgical cytoreduction. Median OS of patients with microscopic residual disease in the current study (64 months) compares favorably with stage III patients with microscopic residual disease (72 months) from our previous study. 15 Possible alternatives (eg, neoadjuvant therapy, interval debulking, and so on) would be reserved for those in whom microscopic cytoreduction is not deemed possible because their outcome appears uninfluenced by the extent of cytoreduction that falls short of microscopic residual disease. Currently, preoperative identification of patients most likely to achieve cytoreduction to microscopic disease is limited on the basis of current radiologic imaging and laboratory testing. 9,16,17 Furthermore, some patients with stage IV disease are diagnosed only postoperatively by imaging or malignant pleural cytology. Therefore, some initial form of surgery would be recommended in all stage IV ovarian cancer patients who are medically fit to undergo surgery. Whether surgical cytoreductive effort, tumor biology, or a combination of both is more predictive of tumor residual remains to be determined. Nonetheless, residual tumor size has a significant association with PFS and OS among patients with stage IV EOC. Previous studies demonstrating improved outcome in patients with small postoperative residual tumor size 6-9 have prompted many surgeons to adopt aggressive surgical practice patterns in the cytoreduction of metastatic ovarian cancer. However, there has not been a prospective randomized trial comparing surgical cytoreductive efforts among advancedstage EOC patients performed to date. Until more definitive phase III data are made available, our current findings may assist the gynecologic oncologist in his/her approach at cytoreduction in EOC patients with stage IV disease. AUTHORS DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a U are those for which no compensation was received; those relationships marked with a C were compensated. For a detailed description of the disclosure categories, or for more information about ASCO s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: Franco Muggia, Bristol-Myers Squibb (C); William P. McGuire, Unither (C), GlaxoSmithKline (C), Genetech (C), Novartis (C) Stock Ownership: None Honoraria: Franco Muggia, Ortho Biotech; William McGuire, GlaxoSmithKline, Novartis Research Funding: William P. McGuire, GlaxoSmithKline Expert Testimony: None Other Remuneration: None AUTHOR CONTRIBUTIONS Conception and design: William E. Winter III, G. Larry Maxwell, G. Scott Rose Financial support: G. Scott Rose Administrative support: G. Scott Rose Provision of study materials or patients: William E. Winter III, Peter G. Rose, Stephen C. Rubin, William P. McGuire Collection and assembly of data: William E. Winter III, Michael J. Sundborg Data analysis and interpretation: William E. Winter III, G. Larry Maxwell, Chunqiao Tian, Franco Muggia, William P. McGuire Manuscript writing: William E. Winter III, George Larry Maxwell, Chunqiao Tian, G. Scott Rose, Peter G. Rose, Franco Muggia, William P. McGuire Final approval of manuscript: William E. Winter III, George Larry Maxwell, G. Scott Rose, Peter G. Rose, Stephen C. Rubin, Franco Muggia, William P. McGuire REFERENCES 1. Jemal A, Siegel R, Ward E, et al: Cancer Statistics, 2006. CA Cancer J Clin 56:106-130, 2006 2. Curtin JP, Malik R, Venkatraman ES, et al: Stage IV ovarian cancer: Impact of surgical debulking. Gynecol Oncol 64:9-12, 1997 3. Bristow RE, Montz FJ, Lagasse LD, et al: Survival impact of surgical cytoreduction in stage IV epithelial ovarian cancer. Gynecol Oncol 72:278-287, 1999 4. Repetto L, Chiara S, Pace M, et al: Prognostic factors in stage IV ovarian carcinoma treated with platinum-based regimens. Tumori 76:274-277, 1990 5. Akahira JI, Yoshikawa H, Shimizu Y, et al: Prognostic factors of stage IV epithelial ovarian cancer: A multicenter retrospective study. Gynecol Oncol 81:398-403, 2001 6. Goodman HM, Harlow BL, Sheets EE, et al: The role of cytoreductive surgery in the management of stage IV epithelial ovarian carcinoma. Gynecol Oncol 46:367-371, 1992 7. Liu PC, Benjamin I, Morgan MA, et al: Effect of surgical debulking on survival in stage IV ovarian cancer. Gynecol Oncol 64:4-8, 1997 8. Munkarah AR, Hallum AV, Morris M, et al: Prognostic significance of residual disease in patients with stage IV epithelial ovarian cancer. Gynecol Oncol 64:13-17, 1997 9. Bristow RE, Tomacruz RS, Armstrong DK, et al: Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: A meta-analysis. J Clin Oncol 20:1248-1259, 2002 10. McGuire WP, Hoskins WJ, Brady MF, et al: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 334:1-6, 1996 11. Muggia FM, Braly PS, Brady MF, et al: Phase III randomized study of cisplatin versus paclitaxel versus cisplatin and paclitaxel in patients with suboptimal stage III or IV ovarian cancer: A Gynecologic Oncology Group Study. J Clin Oncol 18:106-115, 2000 12. Rose PG, Nerenstone S, Brady MF, et al: Secondary surgical cytoreduction for advanced ovarian carcinoma. N Engl J Med 351:2489-2497, 2004 13. Spriggs DR, Brady MF, Rubin S, et al: Phase III randomized trial of intravenous cisplatin plus a 24- or 96-hour infusion of paclitaxel in epithelial ovarian cancer: A Gynecologic Oncology Group study. J Clin Oncol 25:4466-4471, 2007 14. Scholz HS, Benedicic C, Haas J, et al: Stage IV ovarian cancer: Prognostic factors and survival beyond 5 years. Anticancer Res 21:3729-3732, 2001 15. Winter WE III, Maxwell GL, Tian C, et al: Prognostic factors for stage III epithelial ovarian cancer: A Gynecologic Oncology Group study. J Clin Oncol 25:3621-3627, 2007 16. Memarzadeh S, Lee SB, Berek JS, et al: CA125 levels are a weak predictor of optimal cytoreductive surgery in patients with advanced epithelial ovarian cancer. Int J Gynecol Cancer 13:120-124, 2003 17. Chi DS, Venkatraman ES, Masson V, et al: The ability of preoperative serum CA-125 to predict optimal primary tumor cytoreduction in stage III epithelial ovarian carcinoma. Gynecol Oncol 77:227-231, 2000 Appendix The Appendix is included in the full-text version of this article, available online at www.jco.org. It is not included in the PDF version (via Adobe Reader ). www.jco.org 89