Ovarian cancer is the most lethal gynecologic malignancy and is usually diagnosed after the cancer has spread within the peritoneal cavity.1 Two phase III trials have found that women with advanced ovarian cancer randomly assigned to neoadjuvant chemotherapy followed by interval debulking surgery have survival equivalent to those assigned to primary debulking surgery and adjuvant chemotherapy.2,3 Although primary debulking surgery was the traditional treatment for advanced ovarian cancer, utilization of neoadjuvant chemotherapy has increased substantially in the United States.4 Recent guidelines from the American Society for Clinical Oncology and the Society for Gynecologic Oncology recommend neoadjuvant chemotherapy and interval surgery for women with high perioperative risk or low likelihood of achieving optimal cytoreduction at primary surgery.5
Laparoscopic interval debulking surgery is an appealing alternative to laparotomy among elderly and infirmed patients who are frequently selected for neoadjuvant chemotherapy; however, little is known about the safety and effectiveness of this approach.6–9 Two small, uncontrolled studies investigated the feasibility of laparoscopic interval debulking surgery and showed promising perioperative and short-term survival outcomes.7,8 A small, historically controlled study, which included 10 women who underwent laparoscopic interval debulking and 11 women in the control group, found a nonsignificant decrease in recurrence-free survival and increased cancer-specific mortality.9
To investigate the safety and effectiveness of laparoscopic interval debulking surgery, this study compared 3-year survival, length of hospitalization, perioperative mortality, risk of readmission, and residual disease between patients who underwent laparoscopic and laparotomic interval debulking surgery in a national cancer registry.
MATERIALS AND METHODS
We conducted a retrospective cohort study using data from the National Cancer Database. The National Cancer Database is a joint program of the American College of Surgeons' Commission on Cancer and the American Cancer Society. This program aggregates data from more than 1,500 hospital tumor registries and includes 70% of all incident cancer diagnoses in the United States.10 The National Cancer Database collects data on demographics, tumor characteristics, cancer-directed therapies, hospital characteristics, and overall survival. Because this investigation used an existing deidentified database, the study protocol was granted exempt status by the Partners Human Research Committee under federal regulation 45 CFR 46.101(b).
Figure 1 shows the study cohort flow diagram. We identified all patients diagnosed with malignant stage IIIC and IV epithelial ovarian cancer from 2010 to 2012 in the 2013 National Cancer Database public use data file. Because pathologic stage is sometimes not known until after initiation of treatment, the stage was defined as American Joint Committee on Cancer 7th-edition clinical stage IIIC or IV when available or as American Joint Committee on Cancer or International Federation of Gynecology and Obstetrics pathologic stage IIIC or IV when clinical stage was not recorded.11,12 We defined epithelial histology using International Classification of Diseases for Oncology, Third Edition codes corresponding to serous, mucinous, clear cell, endometrioid, or other adenocarcinoma of the ovary (Table 1).
We excluded all patients without microscopically confirmed cancer, who had pre-existing cancer diagnoses, who did not undergo definitive surgery for ovarian cancer, or who did not receive chemotherapy. To define a cohort of patients who underwent neoadjuvant chemotherapy followed by interval debulking surgery, we excluded all patients who had surgery before initiating chemotherapy and all those in whom the sequence of surgery and chemotherapy was unknown. Finally, we excluded all patients in whom the surgical approach (laparoscopy or laparotomy) was unknown.
The primary analysis was based on the intended surgical approach. All patients whose surgery was initiated with a minimally invasive modality (traditional or robot-assisted laparoscopy) were categorized as having undergone a planned laparoscopic procedure, irrespective of conversion to laparotomy. Surgical approach was abstracted from medical records by the cancer registrar.
The primary outcome was overall survival defined as months from cancer diagnosis to death or the date of last contact as reported by the cancer registry. Overall survival was examined as probability of survival 3 years after diagnosis (3-year survival) and the hazard of death from any cause during follow-up. Secondary outcomes of interest included median survival, frequency of death within 90 days of surgery, residual disease status after surgery, frequency of unplanned readmission within 30 days of surgical discharge, and length of hospital admission after surgery.
We categorized patient age at the time of diagnosis as less than 40, 40–49, 50–59, 60–69, 70–79, and 80 years or older. Race-ethnicity was defined as Hispanic, non-Hispanic white, non-Hispanic black, non-Hispanic Asian, or other or unknown. We classified tumor grade using a three-point scale. A Charlson Deyo score of one or greater defined the presence of a comorbidity.13,14
We categorized the location of the treating facility by U.S. Census Region (Northeast, South, Midwest, and West).15 Rural–urban status was determined by using patients' county of residence and classified as metropolitan, adjacent to a metropolitan area, or nonmetropolitan by using the U.S. Department of Agriculture's Economic Research Service rural–urban continuum codes.16 The median household income in the participant's ZIP code was used as a proxy for participant income and categorized according to quartiles across all U.S. ZIP codes as estimated by the 2012 American Community Survey. We categorized insurance status as uninsured, privately insured, insured by Medicare, or having another type of government insurance (military or Medicaid). The treating facility was classified as a community cancer program, an academic–research program, an integrated network cancer program, or other. Surgical procedure was categorized as excision of gynecologic organs only (adnexal surgery with or without hysterectomy) or surgery that involved additional cytoreductive maneuvers (including omentectomy or other debulking).
Patient characteristics were compared by using the t test and χ2 test for continuous and categorical variables. To explore independent predictors of laparoscopic interval debulking surgery, we developed a multivariable logistic regression model. Covariates in the logistic regression model were selected a priori and included age group, race, facility type, urban status, median ZIP code income, year of diagnosis, insurance type, presence of comorbid conditions, and stage.
We used the Kaplan-Meier method to plot survival functions for women who underwent interval surgery by laparotomy and laparoscopy and compared these using the log rank test. We estimated the probability of survival 3 years after diagnosis for each group, associated 95% confidence intervals (CIs), and compared these with the Z test for independent proportions. We also estimated the hazard ratio (HR) for overall mortality after planned laparoscopic interval debulking surgery with a univariable Cox proportional hazards model with planned laparotomy serving as the referent group. To adjust for potential confounding we also calculated the HR for overall mortality after planned laparoscopic staging using a Cox proportional hazard model adjusting for covariates that were selected a priori and included age group, race–ethnicity, region, type of facility, urban status, income, presence of comorbidities, year of diagnosis, and cancer stage. Missing data were coded as indicator variables. Differences in frequency of readmission, 90-day mortality, and residual disease status were compared with the χ2 test, and the Wilcoxon rank sum test was used to compare length of postoperative hospitalizations.
We conducted a sensitivity analysis in which patients who underwent conversion from laparoscopy to laparotomy were excluded to assess whether the main results persisted in this subset of patients. A post hoc power calculation was based on the proportional hazard model and used the approach described by Schoenfeld and Richter17 and implemented by Dupont and Plummer.18
We identified, 3,071 women with stage IIIC and IV epithelial ovarian cancer who received neoadjuvant chemotherapy followed by interval debulking surgery between 2010 and 2012 and who met all inclusion criteria (Fig. 1). Debulking surgery was initiated laparoscopically in 450 women (15%), and robotic-assisted laparoscopy comprised 27% of these cases (122 patients). Debulking surgery was completed laparoscopically in 378 women (84% of cases initiated laparoscopically), and conversion to laparotomy was more frequent among those who underwent traditional laparoscopy than among those who underwent robotic-assisted laparoscopic surgery (20% compared with 5%, P<.001).
In univariable analysis (Table 2) we found that laparoscopic interval surgery was associated with region (P=.003) and was used more frequently in the Northeastern and Southern (18% and 15%, respectively) than in the Western and Midwestern regions of the United States (11% and 13%, respectively). Patients living in or adjacent to a metropolitan area were nearly twice as likely to have had laparoscopic interval surgery as those in nonmetropolitan counties (8% compared with 15%, P=.04). We observed a significant increase in the frequency of laparoscopic interval debulking from 2010 to 2012 (11–16%, P<.001). Patients with one or more comorbidities underwent laparoscopic interval surgery more frequently than those without comorbidities (17% compared with 14%, P=.05). Laparoscopic surgery was performed more frequently at community and integrated network cancer programs (15% and 16%, respectively) than academic centers (13%, P=.04) and more frequently among women with stage IV than among those with stage IIIC cancer (16% compared with 13%, P=.02).
To explore which factors were independently associated with receipt of laparoscopic interval debulking surgery, we conducted a multivariable analysis (Table 3). After controlling for demographic, clinical, and hospital-level variables, region, facility type, stage, and year of diagnosis were independent predictors of laparoscopic interval debulking surgery.
Table 4 tabulates surgical characteristics and outcomes among women who underwent planned interval laparoscopic surgery and laparotomy. There were no significant differences in frequency of readmission (5.3% compared with 3.7%) or death within 90 days of surgery (2.8% compared with 2.9%). Compared with women who underwent laparotomy, those who had laparoscopic procedures had significantly shorter postoperative hospitalizations (median 4 compared with 5 days, P<.001). Women who underwent laparoscopic interval debulking surgery had procedures that were limited to excision of gynecologic structures only (no omentectomy or additional cytoreductive maneuvers) more frequently than women who had laparotomy (16.7% compared with 10.1%, P<.001). Residual disease status was available in 1,595 patients (60.9%) who underwent planned laparotomy and in 277 patients (61.6%) who underwent planned laparoscopic interval surgery. Among those with known residual disease status, we observed no difference in the frequency of complete gross resection, optimal cytoreduction (residual disease, 1 cm or less), and suboptimal debulking (residual disease greater than 1 cm), by surgical approach (P=.29).
Survival information was available in 445 patients (98.9%) who underwent planned laparoscopic interval debulking surgery and for 2,598 patients (99.1%) who had planned laparotomy. Median follow-up was 32.0 months (95% CI 31.2–33.0). We observed 183 deaths among women who underwent planned laparoscopy and 1,045 among those who had laparotomy. Figure 2 illustrates the survival functions for patients by planned surgical approach. We found no significant difference in unadjusted survival between women who underwent planned laparoscopic approach and those who underwent laparotomy (HR 1.12; 95% CI 0.96–1.31; P=.15). Median survival was 33.8 months (95% CI 31.9–40.6) and 37.6 months (95% CI 36.1–39.6) for women who underwent planned laparoscopy and laparotomy, respectively. The probability of surviving 3 years after diagnosis was 47.5% (95% CI 41.4–53.5) for patients who underwent planned interval laparoscopy and 52.6% (95% CI 50.3–55.0) for those who had planned laparotomy (P=.12).
Adjustment for age group, race, region, facility type, urban status, income, presence of comorbidities, year of diagnosis, and stage did not alter the association between planned laparoscopic surgery and overall survival (adjusted HR 1.09; 95% CI 0.93–1.28; P=.26). When we excluded from the analysis women who underwent conversion from laparoscopy to laparotomy, we found no significant difference in unadjusted (HR 1.16; 95% CI 0.98–1.37; P=.08) or adjusted (adjusted HR 1.12; 95% CI 0.94–1.32; P=.20) survival estimates by surgical approach. Post hoc power analysis demonstrated that with 1,231 deaths at a significance level of 0.05, this study had 80% power to detect an 8% difference in 3-year survival between the laparoscopy and laparotomy groups (53% compared with 45%).
In this cohort study, we investigated the use and effectiveness of laparoscopic interval debulking surgery in women who received neoadjuvant chemotherapy for advanced epithelial ovarian cancer. We found that interval surgery was initiated laparoscopically in 15% of patients and completed laparoscopically in 11% of patients. We observed no benefit or detriment in survival associated with receipt of laparoscopic interval debulking surgery. Additionally, we observed shorter postoperative hospitalization after laparoscopic surgery and no difference in risk of unplanned readmission or perioperative death. These findings suggest that in well-selected patients, such as those who have a complete response to neoadjuvant chemotherapy, laparoscopic interval debulking may be a safe and effective alternative to laparotomy.
The feasibility of laparoscopic debulking surgery after neoadjuvant chemotherapy has been demonstrated previously in several small studies.7–9 In a prospective observational study, Gueli Alletti et al7 performed laparoscopic interval debulking in 30 women with clinical response to neoadjuvant chemotherapy, achieving resection of all visible disease in 29 women. All patients were alive with a median follow-up of 10.5 months.7 In a retrospective study of 30 women, Corrado et al8 found that laparoscopic cytoreduction was associated with low rates of intraoperative and postoperative complications (3.3% and 6.6%, respectively). At study completion, 26 of 30 patients were alive without recurrence with a median follow-up of 15 months.8 In a historically controlled study of 10 women who underwent laparoscopic interval debulking and 11 women who had laparotomy, Favero et al9 found a nonsignificant decrease in cancer-specific survival and a nonsignificant reduction in the chemotherapy-free interval among women undergoing laparoscopy. Although these studies suggest that laparoscopic debulking may be feasible, small sample size and the absence of any control group in all but the smallest study preclude meaningful comparison between laparoscopy and laparotomy.7–9
In the current study, we observed no difference in survival between women who underwent laparoscopic interval debulking and those who underwent laparotomy. These results were consistent in the primary intention-to-treat analysis and after excluding cases in which laparoscopic procedures were converted to a laparotomy. Survival between the groups remained equivalent after controlling for a range of potential confounders, including age, facility type, comorbidities, and stage.
Among the 61% of women with known residual disease status, the mode of interval surgery did not alter the frequency of optimal cytoreduction. This finding is consistent with the prior series that reported high rates of complete cytoreduction achieved with laparoscopic interval debulking.7,8 Interestingly, we found that surgeons performing laparoscopic interval debulking surgery were more likely to omit resection of extrapelvic structures, yet this did not affect the extent of residual disease or survival among women undergoing laparoscopic interval surgery. These findings suggest that patients selected for laparoscopic debulking may have a lower burden of disease than those who underwent laparotomy. Alternatively, it is possible that disease that would have been palpable at laparotomy was not visualized during laparoscopy and was, therefore, underreported.
The results of this study should be interpreted in light of important limitations, including the lack of random allocation and our inability to verify registry data. Estimates of the association of laparoscopic interval surgery with survival may be biased from patient selection based on factors that are not measured in the data, including burden of disease or functional status. However, because patients selected for laparoscopy tended to have more advanced disease and more comorbidities, we expect such bias to overestimate harm associated with laparoscopic staging. Moreover, we cannot verify surgeon intention in cases that were initiated laparoscopically but completed via laparotomy; it is possible that surgeons initiated some procedures laparoscopically to assess feasibility of open interval debulking. Reassuringly, our findings were not affected by exclusion of such cases. Residual disease status was missing in 39% of cases, calling into question conclusions related to this variable. However, the concordance of residual disease and survival results argues against bias in the missing data. This study has relatively short follow-up, and it is possible that survival difference may emerge with further observation. Additionally, this study had power to detect an 8% difference in 3-year survival, and smaller differences may be clinically relevant. Furthermore, the study was likely underpowered for small difference in the frequency of rare events, such as readmission and perioperative death. We could not investigate the effect of laparoscopic debulking on recurrence-free or disease-specific survival because these outcomes are not reported in the National Cancer Database. Finally, this investigation does not provide evidence for or against the use of neoadjuvant chemotherapy in advanced ovarian cancer.
1. Cannistra SA. Cancer of the ovary [published erratum appears in N Engl J Med 2005;352:104]. N Engl J Med 2004;351:2519–29.
2. Kehoe S, Hook J, Nankivell M, Jayson GC, Kitchener H, Lopes T, et al. Primary chemotherapy versus primary surgery for newly diagnosed advanced ovarian cancer (CHORUS): an open-label, randomised, controlled, non-inferiority trial. Lancet 2015;386:249–57.
3. Vergote I, Tropé CG, Amant F, Kristensen GB, Ehlen T, Johnson N, et al. Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer. N Engl J Med 2010;363:943–53.
4. Melamed A, Hinchcliff EM, Clemmer JT, Bregar AJ, Uppal S, Bostock I, et al. Trends in the use of neoadjuvant chemotherapy for advanced ovarian cancer in the United States. Gynecol Oncol 2016;143:236–40.
5. Wright AA, Bohlke K, Armstrong DK, Bookman MA, Cliby WA, Coleman RL, et al. Neoadjuvant chemotherapy for newly diagnosed, advanced ovarian cancer: Society of Gynecologic Oncology and American Society of Clinical Oncology Clinical Practice guideline. J Clin Oncol 2016;34:3460–73.
6. Nezhat FR, Pejovic T, Finger TN, Khalil SS. Role of minimally invasive surgery in ovarian cancer. J Minim Invasive Gynecol 2013;20:754–65.
7. Gueli Alletti S, Bottoni C, Fanfani F, Gallotta V, Chiantera V, Costantini B, et al. Minimally invasive interval debulking surgery in ovarian neoplasm (MISSION trial-NCT02324595): a feasibility study. Am J Obstet Gynecol 2016;214:503.e1–6.
8. Corrado G, Mancini E, Cutillo G, Baiocco E, Vici P, Sergi D, et al. Laparoscopic debulking surgery in the management of advanced ovarian cancer after neoadjuvant chemotherapy. Int J Gynecol Cancer 2015;25:1253–7.
9. Favero G, Macerox N, Pfiffer T, Köhler C, da Costa Miranda V, Estevez Diz Mdel P, et al. Oncologic concerns regarding laparoscopic cytoreductive surgery in patients with advanced ovarian cancer submitted to neoadjuvant chemotherapy. Oncology 2015;89:159–66.
10. Raval MV, Bilimoria KY, Stewart AK, Bentrem DJ, Ko CY. Using the NCDB for cancer care improvement: an introduction to available quality assessment tools. J Surg Oncol 2009;99:488–90.
11. Edge SB, Burd DR, Compton CC, Fritz AG, Greene FL CC, Trotti A, editors. AJCC cancer staging manual. 7th ed. New York (NY): Springer; 2009.
12. Prat J; FIGO Committee on Gynecologic Oncology. Abridged republication of FIGO's staging classification for cancer of the ovary, fallopian tube, and peritoneum. Cancer 2015;121:3452–4.
13. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol 2000;53:1258–67.
14. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83.
15. United States Census Bureau. Census Regions and Divisions of the United States. Available at: www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf
. Retrieved September 15, 2016.
16. United States Department of Agriculture Economic Research Service. Rural urban continuum codes. Available at: http://www.ers.usda.gov/data-products/rural-urban-continuum-codes.aspx
. Retrieved September 15, 2016.
17. Schoenfeld DA, Richter JR. Nomograms for calculating the number of patients needed for a clinical trial with survival as an endpoint. Biometrics 1982;38:163–70.
18. Dupont WD, Plummer WD. Power and sample size calculations: a review and computer program. Control Clin Trials 1990;11:116–28.