Key PointsQuestion
Is there a difference in survival between patients who receive primary cytoreductive surgery compared with neoadjuvant chemotherapy for advanced-stage epithelial ovarian cancer?
Findings
In this cohort study 22 962 patients, 19 836 (86.4%) received primary cytoreductive surgery and 3126 (13.6%) underwent neoadjuvant chemotherapy. Among propensity-score matched groups, the median overall survival was significantly longer in the primary cytoreductive surgery group than in the neoadjuvant chemotherapy group.
Meaning
Primary cytoreductive surgery was associated with improved survival compared with neoadjuvant chemotherapy in otherwise healthy women 70 years or younger with advanced-stage epithelial ovarian cancer.
Importance
Uncertainty remains about the relative benefits of primary cytoreductive surgery (PCS) vs neoadjuvant chemotherapy (NACT) for advanced-stage epithelial ovarian cancer (EOC).
Objective
To compare overall survival of PCS vs NACT in a large national population of women with advanced-stage EOC.
Design, Setting, and Participants
Retrospective cohort study of women with stage IIIC and IV EOC diagnosed between 2003 and 2011 treated at hospitals across the United States reporting to the National Cancer Data Base. We focused on patients 70 years or younger with a Charlson comorbidity index of 0 who were likely candidates for either treatment.
Exposures
Initial treatment approach of PCS vs NACT, examined using an intent-to-treat analysis.
Main Outcomes and Measures
Overall survival, defined as months from cancer diagnosis to death or date of the last contact. We used propensity score matching to compare similar women who underwent PCS and NACT. The association of treatment approach with overall survival was assessed using the Kaplan-Meier method and the log-rank test. We assessed whether the findings were influenced by differences in the prevalence of an unobserved confounder, such as limited performance status (Eastern Cooperative Oncology Group 1-2), preoperative disease burden, and BRCA status.
Results
Among 22 962 patients (mean [SD] age, 56.12 [9.38] years), 19 836 (86.4%) received PCS and 3126 (13.6%) underwent NACT. We matched 2935 patients treated with NACT with similar patients who received PCS. The median follow-up was 56.5 (95% CI, 54.5-59.2) months in the PCS group and 56.3 (95% CI, 54.5-59.8) months in the NACT group in the propensity-matched cohort. Among propensity score–matched groups, the median overall survival was 37.3 (95% CI, 35.2-38.7) months in the PCS group and 32.1 (95% CI, 30.8-34.1) months in the NACT group (P < .001). However, if the NACT group had a higher proportion of women with performance statuses of 1 to 2 compared with those who underwent PCS (60% vs 50%), the association of PCS and improved survival would not be statistically significant.
Conclusions and Relevance
Primary cytoreductive surgery was associated with improved survival compared with NACT in otherwise healthy women with advanced-stage epithelial ovarian cancer aged 70 years or younger. The lower survival in women who received NACT could be explained by a higher prevalence of limited performance status in women undergoing NACT.
Epithelial ovarian cancer (EOC) is associated with the highest case-fatality ratio of all gynecologic cancers. Primary cytoreductive surgery (PCS) followed by chemotherapy consisting of carboplatin and paclitaxel is the traditional treatment regimen for women with advanced EOC.1 Neoadjuvant chemotherapy with interval cytoreductive surgery (NACT) and postoperative chemotherapy is an alternative approach,2,3 often used to treat patients with large tumor burdens, multiple comorbidities, or stage IV disease.4,5
The role of NACT in the treatment of EOC remains controversial. Several retrospective studies have suggested that PCS with optimal cytoreduction is more effective than NACT for women with advanced-stage EOC, with improved progression-free and overall survival.6-10 However, 2 phase 3 randomized clinical trials have demonstrated that NACT is noninferior to PCS for women with stages IIIC and IV EOC with high disease burdens.2,3 Both trials have been criticized because a high number of patients who received PCS had greater than 1 cm of residual disease after surgery, the survival durations were shorter than expected, and a substantial proportion of women received single-agent carboplatin instead of the traditional platinum-based doublet.11,12 Exploratory analyses of the trial of Vergote el al2 revealed that patients with stage IIIC disease and less than 4.5 cm of disease had superior survival when treated with PCS, and patients with stage IV disease had better survival when treated with NACT.13
Despite the evidence supporting the noninferiority of NACT, this approach is not accepted by some gynecologic oncologists in the United States.11,12 The purpose of this study was to compare the overall survival of women with advanced-stage EOC who underwent PCS with those who received NACT in a large US population. We restricted our study sample to women who were 70 years or younger and without medical comorbidities to maximize the likelihood that patients would be eligible for either treatment and reduce bias related to the preferential use of NACT in elderly and frail women.
We identified women who received a diagnosis of advanced-stage (stage IIIC or IV) EOC during 2003 through 2011 from the National Cancer Data Base (NCDB). The NCDB is a nationwide oncology cancer registry that includes information on approximately 70% of all incident cancers diagnosed in the United States and serves as a surveillance mechanism for more than 1430 hospitals participating in the American College of Surgeons Commission on Cancer.14 Because all data are deidentified, this study was deemed exempt by the Massachusetts General Hospital Institutional Review Board.
To better compare NACT with PCS among patients who were likely candidates for either treatment, we studied women who were 70 years or younger with a Charlson comorbidity index of 0. The Charlson Comorbidity Index includes comorbidities likely to affect survival among hospitalized patients with cancer; a score of 0 means that none of the Charlson comorbidities were present.15,16 From 54 382 patients with incident stage IIIC or IV ovarian cancer diagnosed between 2003 and 2011, we excluded women whose International Classification of Diseases for Oncology, 3rd Edition, histologic code did not correspond to a category of EOC (N = 2828), women with other cancers (N = 6321), those without microscopic confirmation (N = 239), and those with unknown follow-up time (N = 4). We next excluded women with Charlson scores greater than 0 and/or who were older than 70 years (N = 17 921), patients who received a diagnosis of ovarian cancer at one institution but were treated at another (N = 550), those missing data on the timing of postoperative chemotherapy, women treated with palliative intent, and women with unknown surgical procedures (N = 2758). Finally, to identify primary (rather than salvage) therapy, we restricted the cohort to patients in whom chemotherapy or surgery was initiated within 3 months of diagnosis, resulting in a final study population of 22 962 women (Figure).
The exposure of interest was initial treatment approach of PCS vs NACT. We performed an intent-to-treat analysis, in which patients who underwent initial PCS were included even if they never received chemotherapy, and those who had initial NACT were included even if they never received surgery. Receipt of treatment, including chemotherapy, in the NCDB has been documented to be reliable.17,18 The primary outcome of interest was overall survival, defined as months from cancer diagnosis to death or date of the last contact.
Control variables included age at diagnosis, race/ethnicity, year of diagnosis, and geographic location of the treating facility. We documented median household income of the ZIP code of residence as estimated by the American Community Survey, categorized in quartiles (<$38 000, $38 000-$47 999, $48 000-$62 999, and ≥$63 000). Insurance status was categorized as uninsured, private insurance, Medicare, or another type of government insurance. The treating facility was categorized according to the Commission on Cancer Accreditation program as a community cancer program, comprehensive community cancer program, academic/research program, or other.
We compared categorical variables using χ2 tests and continuous variables using independent sample t tests. Propensity score matching was undertaken to create a cohort in which patients who underwent PCS and NACT were balanced on covariates that might confound the effect of treatment approach on survival.19 We fit logistic regression models to estimate the probability of receiving NACT; independent variables included age, year of diagnosis, race/ethnicity, treating facility type, insurance status, income, geographic region, rural vs urban status, stage, histologic subtype, and grade. We then matched each woman who received NACT with a woman who underwent PCS who had the same propensity to undergo NACT using a caliper of 0.2.20 We compared standardized differences of covariates in the propensity-matched cohort (eFigure 1 in the Supplement).
We compared overall survival between women who underwent PCS vs NACT in the propensity score–matched cohort using the Kaplan-Meier method and the log-rank test. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals for the association of treatment with survival. We repeated analyses stratified by stage and period of diagnosis in the propensity-matched groups; we also tested for the interaction of treatment approach by stage and treatment approach by period of diagnosis in separate models.
In sensitivity analyses, we repeated all analyses after excluding women who received chemotherapy without interval cytoreductive surgery from the NACT group because of a concern that these women may have been too sick to undergo surgery. Additionally, we restricted the analysis to high-volume centers, defined as hospitals that treated at least 20 ovarian cancer cases per year,21,22 and women with high-grade serous cancer. For these analyses, we refit the propensity score models.
Although we adjusted for numerous observed characteristics using propensity score analyses, propensity scores cannot adjust for unmeasured confounders. Therefore, we tested the sensitivity of our associations between treatment group and survival to several potential unmeasured confounders.23,24 To do this, we assumed that there were unmeasured factors, that is, preoperative disease burden, performance status, and BRCA status, that were associated with both the treatment received and survival. We then updated our estimates of the differences in survival after adjusting for these additional unobserved variables using specific assumptions about differences in the prevalence of these characteristics in women receiving PCS vs NACT and the confounder’s association with survival. Based on the Gynecologic Oncology Group (GOG)-182 study estimate of the effects of preoperative disease burden on overall survival, we estimated that women with high preoperative disease burden have an HR of death of 1.33 compared with women with a moderate disease burden.25 We instituted a range of assumptions about the prevalence of high disease burden, which is likely to have been more prevalent among women undergoing NACT. Similarly, the impact of performance status of 1 or 2, compared with a performance status of 0, in women with EOC is associated with an HR of time to death of 1.71.26 We also assumed that performance status of 1 or 2 was more prevalent among women undergoing NACT. Finally, the impact of BRCA status was also assessed. BRCA mutation is associated with a better prognosis in women with EOC; in a large series, the HR of time to death was 0.72 in women with a BRCA mutation compared with those without the mutation.27 We assumed that BRCA mutations were more prevalent among the PCS group, assuming this group may be more likely to pursue aggressive treatment of their cancer with PCS.
Statistical tests were 2 sided and differences were considered statistically significant at P < .05. All statistical analyses were performed using R 3.0.3 (R Foundation for Statistical Computing, Vienna, Austria).
Among 22 962 included patients, 19 836 (86.4%) underwent PCS and 3126 (13.6%) patients received NACT. Of those who underwent NACT, 813 (26.0%) never received interval cytoreductive surgery, while 2925 (14.7%) of women treated with PCS did not receive chemotherapy within 6 months of surgery. Table 1 summarizes selected patient characteristics and eTable 1 in the Supplement summarizes additional demographic and clinical characteristics of the study population. Women who had NACT were significantly older and less likely to receive a diagnosis of stage IIIC disease.
In a multivariable model, factors associated with use of NACT included older age at diagnosis, more recent diagnosis, serous histologic subtype, and stage IV vs IIIC disease (eTable 2 in the Supplement). After stratification by stage, associations were similar. Propensity score matching between the PCS and NACT groups yielded 2935 matched pairs (93.9% of 3126 patients) in the NACT group (Table 1 and eTable 1 in the Supplement). After propensity score matching, all characteristics were balanced in the 2 groups (eFigure 1 in the Supplement).
The median follow-up was 56.5 (95% CI, 54.5-59.2) months in the PCS group and 56.3 (95% CI, 54.5-59.8) months in the NACT group in the propensity-matched cohort (eFigure 2 in the Supplement shows Kaplan-Meier survival curves for the propensity-matched groups for the entire population). The median overall survival in the propensity-matched cohort was 37.3 (95% CI, 35.2-38.7) months for the PCS group and 32.1 (95% CI, 30.8-34.1) months in the NACT group (P < .001). Within the cohort, 30- and 60-day survival rates were 98.8% and 97.4% in the NACT group, respectively, and 97.4% and 94.7% in the PCS group. In addition, 5- and 10-year survival rates were 25.3% and 12.2% in the NACT group, and 32.8% and 14.2% in women who underwent PCS.
Overall, women in the NACT group had an 18% increased hazard of death from all causes compared with PCS (HR, 1.18; 95% CI, 1.11-1.26) (Table 2). When stratified by period of diagnosis, the association of treatment approach with survival was similar for each period (P = .73 for interaction of treatment approach by time period; data not shown) (Table 2). Results differed slightly after stratification by stage; in women with stage IIIC disease, the propensity-matched HR in women who received NACT compared with PCS was 1.24 (95% CI, 1.11-1.37); and in stage IV the propensity-matched HR was 1.13 (95% CI, 1.04-1.23); however, these differences were not statistically significant (P = .38 for interaction).
In sensitivity analyses, we assessed the association of treatment group with all-cause survival after excluding patients who only received chemotherapy. Among women who received either PCS or NACT followed by interval cytoreductive surgery, in the propensity score–adjusted analysis, the median overall survival was 40.1 (95% CI, 38.2-41.7) months in the women in the PCS group and 32.4 (95% CI, 30.9-34.2) months in the NACT group (P < .001) (adjusted HR, 1.29; 95% CI, 1.21-1.38) (eFigure 3 in the Supplement shows Kaplan-Meier survival curves for the propensity-matched groups for this population). Among women who received treatment at high-volume centers, in the propensity score–adjusted analysis, the median overall survival was 37.7 (95% CI, 36.1-39.9) months in the women in PCS group and 32.0 (95% CI, 30.7-34.0) months in the NACT group (P < .001) (adjusted HR, 1.20; 95% CI, 1.13-1.28). When patients with high-grade serous cancer were analyzed separately, in the propensity score–adjusted analysis, the median overall survival was 44.0 (95% CI, 41.6-46.6) months in the women in PCS group and 35.6 (95% CI, 33.9-38.0) months in the NACT group (P < .001) (adjusted HR, 1.28; 95% CI, 1.16-1.41).
In analyses assessing the sensitivity of our findings to potential unmeasured confounders, we tested whether differences in the burden of disease, performance status, or BRCA status25-27 could explain our findings. Our findings were robust to rather large differences in the prevalence of high disease burden and BRCA status (eTable 3 in the Supplement). For example, given prior evidence that a high preoperative disease burden is associated with higher mortality (HR = 1.33) compared with moderate disease burden, if 95% of women receiving NACT had high disease burden compared with 60% of women in the PCS group, the association of PCS with better survival would still be statistically significant. The findings were less robust for potential moderate-sized differences in performance status. For example, given prior evidence that a limited performance status (Eastern Cooperative Oncology Group [ECOG] 1-2) is associated with higher mortality (HR = 1.71) compared with a performance status of 0, if 60% of women receiving NACT had an ECOG performance status of 1 to 2 compared with 50% in the PCS group, then the association of PCS and improved survival would no longer be statistically significant.
In this large national study examining primary treatment for otherwise healthy women 70 years or younger with advanced-stage EOC, PCS was associated with improved survival of approximately 5 months compared with NACT. Primary cytoreductive surgery was associated with improved survival even when we restricted our analyses to patients who received NACT without interval cytoreductive surgery, patients with high-grade serous cancers, or those treated at high-volume centers. The findings were robust to large differences in 2 potential unobserved confounders, high preoperative disease burden and BRCA status, but not large differences in the prevalence of limited performance status.
Two clinical trials showed noninferior survival in women randomized to NACT, compared with PCS. However, these trials have substantial shortcomings. The trial of Vergote et al2 is limited by the short overall survival and low proportion of patients who were optimally cytoreduced to less than 1 cm of residual disease (41.6%). The median overall survival of patients randomized to PCS was only 29 months vs 30 months in the NACT group. Similarly, in the Chemotherapy or Upfront Surgery (CHORUS) trial3 the median overall survival of patients randomized to PCS was 23 months vs 24 months in the NACT group, and only 41% of the patients in the PCS arm were optimally cytoreduced to less than 1 cm of residual disease.
Prior retrospective studies and a meta-analysis have demonstrated improved overall survival in women who received PCS.6-10,28 However, observational studies do not provide the same level of evidence as randomized clinical trials. In observational studies, women who receive NACT are usually older, have more comorbid conditions, have higher preoperative tumor burdens, and likely more limited performance status. These factors can result in large differences in the distribution of prognostic factors between women receiving NACT and PCS.
In the present investigation, we restricted the population to healthy women 70 years or younger to identify women who were likely candidates for either treatment. Additionally, we used propensity score matching to adjust for many variables that have traditionally been associated with treatment approach. We also tested the robustness of our estimates to potential unobserved confounders. These analyses suggest that our finding of lower survival with NACT was robust to differences in the prevalence of high preoperative disease burden and BRCA mutation status. On the other hand, the lower survival in women who received NACT could be explained by a higher prevalence of limited performance status in women undergoing NACT if the prevalence of ECOG performance status of 1 to 2 is higher in the NACT vs PCS group (ie, 60% vs 50%).
Remarkably, nearly 15% of women never initiated chemotherapy after PCS while approximately 26% of the women treated with NACT did not undergo surgery. Similar to our findings, a study of women with stage III and IV EOC diagnosed between 1995 and 2008 using Surveillance, Epidemiology, and End Results–Medicare data showed that 15% of women underwent surgery alone.29 Similarly, Wright et al30 noted that 19% of older patients with stage II to IV ovarian cancer who underwent primary surgery never received postoperative chemotherapy and 66% of women treated with initial chemotherapy never underwent surgery. In comparison, in the study by Vergote et al,2 12% of the patients in the NACT arm did not receive surgery. The decision not to perform surgery in a substantial number of patients after NACT may be related to the existence of platinum-resistant disease, or other unobserved clinical factors (eg, limited performance status). Although our inclusion of patients treated with chemotherapy only in the NACT cohort could bias survival estimates in the primary intent-to-treat analysis, when patients receiving chemotherapy alone were removed, NACT remained associated with lower survival compared with PCS.
In this study, PCS was associated with improved survival in women with both stage IIIC and IV disease. Although the estimate of the effect of treatment approach on survival was larger in patients with stage IIIC disease, the interaction term was not statistically significant, suggesting that there was no significant difference in our findings by stage. Similarly, when we performed an analysis stratified by period of diagnosis (because the types of patients undergoing NACT may have changed after the publication of the study by Vergote et al),2,31,32 we found no differences.
One of the strengths of this study is that it includes information on nearly 70% of the women with a diagnosis of ovarian cancer in the United States. Nevertheless, it is important to note several limitations. We lacked information about the extent of postoperative residual disease and specific chemotherapy regimens and doses. The differences in survival between treatment approaches could be attributed to other unmeasured factors, such as more aggressive tumor characteristics in women who received NACT. In addition, we had no information on how treatment decisions were made; these are often based on the likelihood of cure and perceptions of how specific interventions will affect patients’ quality of life. The results of the present study are limited to otherwise healthy women 70 years or younger and should not be extrapolated to other populations. Finally, if most surgeons in the United States considered PCS superior to NACT during the study period, then the patients who underwent NACT, even though relatively young and healthy, may have been even more different from those in the PCS group than they appeared.
We found that PCS was associated with improved survival compared with NACT in otherwise healthy women 70 years or younger with advanced-stage EOC. This difference was robust to large differences in 2 potential unobserved confounders, high disease burden and BRCA status; however, the lower survival in women who received NACT could be explained by differences in another unobserved confounder, limited performance status, if the rates are higher in patients who underwent NACT vs PCS. Such information will be important to collect for future observational studies. Future research should focus on which patients benefit most from PCS or NACT to tailor the treatment of women with advanced-stage EOC.
Accepted for Publication: August 18, 2016.
Corresponding Author: J. Alejandro Rauh-Hain, MD, Division of Gynecologic Oncology, Vincent Obstetrics and Gynecology, Massachusetts General Hospital, 55 Fruit St, Yawkey 9 E, Boston, MA 02114 (jrauh-hain@partners.org).
Published Online: November 17, 2016. doi:10.1001/jamaoncol.2016.4411
Author Contributions: Dr Rauh-Hain had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Rauh-Hain, Melamed, Wright, Gockley, Schorge, del Carmen, Keating.
Acquisition, analysis, or interpretation of data: Rauh-Hain, Melamed, Clemmer, Keating.
Drafting of the manuscript: Rauh-Hain, Wright, Gockley, del Carmen.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Rauh-Hain, Melamed, Clemmer, del Carmen, Keating.
Administrative, technical, or material support: Rauh-Hain, Gockley, Schorge.
Study supervision: Schorge, del Carmen.
Conflict of Interest Disclosures: None reported.
Funding/Support: This work was supported by the Deborah Kelly Center for Outcomes Research, Massachusetts General Hospital. This work is supported by grant R25CA092203 from the National Cancer Institute at the National Institutes of Health. Dr Wright is supported by grant K07 CA166210, and Dr Keating, by grant K24CA181510 from the National Cancer Institute.
Role of the Funder/Sponsor: Support from the Deborah Kelly Center for Outcomes Research assisted in the design and conduct of the study; collection, management, analysis, and interpretation of the data but had no role in preparation, review, and approval of the manuscript; and decision to submit the manuscript for publication. Support from the National Cancer Institute assisted in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, and approval of the manuscript but had no role in the decision to submit the manuscript for publication.
2.Vergote
I, Tropé
CG, Amant
F,
et al; European Organization for Research and Treatment of Cancer-Gynaecological Cancer Group; NCIC Clinical Trials Group. Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer.
N Engl J Med. 2010;363(10):943-953.
PubMedGoogle ScholarCrossref 3.Kehoe
S, Hook
J, Nankivell
M,
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(9990):249-257.
PubMedGoogle ScholarCrossref 4.Morgan
RJ
Jr, Alvarez
RD, Armstrong
DK,
et al; National Comprehensive Cancer Network. Ovarian cancer: clinical practice guidelines in oncology.
J Natl Compr Canc Netw. 2006;4(9):912-939.
PubMedGoogle Scholar 5.Worley
MJ
Jr, Guseh
SH, Rauh-Hain
JA,
et al. What is the optimal treatment for obese patients with advanced ovarian carcinoma?
Am J Obstet Gynecol. 2014;211(3):231.e1-231.e9.
PubMedGoogle ScholarCrossref 6.Bristow
RE, Tomacruz
RS, Armstrong
DK, Trimble
EL, Montz
FJ. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: a meta-analysis.
J Clin Oncol. 2002;20(5):1248-1259.
PubMedGoogle ScholarCrossref 7.Eisenhauer
EL, Abu-Rustum
NR, Sonoda
Y,
et al. The addition of extensive upper abdominal surgery to achieve optimal cytoreduction improves survival in patients with stages IIIC-IV epithelial ovarian cancer.
Gynecol Oncol. 2006;103(3):1083-1090.
PubMedGoogle ScholarCrossref 8.Hoskins
WJ, Bundy
BN, Thigpen
JT, Omura
GA. The influence of cytoreductive surgery on recurrence-free interval and survival in small-volume stage III epithelial ovarian cancer: a Gynecologic Oncology Group study.
Gynecol Oncol. 1992;47(2):159-166.
PubMedGoogle ScholarCrossref 9.Aletti
GD, Dowdy
SC, Gostout
BS,
et al. Aggressive surgical effort and improved survival in advanced-stage ovarian cancer.
Obstet Gynecol. 2006;107(1):77-85.
PubMedGoogle ScholarCrossref 10.Chi
DS, Eisenhauer
EL, Zivanovic
O,
et al. Improved progression-free and overall survival in advanced ovarian cancer as a result of a change in surgical paradigm.
Gynecol Oncol. 2009;114(1):26-31.
PubMedGoogle ScholarCrossref 11.Schorge
JO, Clark
RM, Lee
SI, Penson
RT. Primary debulking surgery for advanced ovarian cancer: are you a believer or a dissenter?
Gynecol Oncol. 2014;135(3):595-605.
PubMedGoogle ScholarCrossref 13.van Meurs
HS, Tajik
P, Hof
MH,
et al. Which patients benefit most from primary surgery or neoadjuvant chemotherapy in stage IIIC or IV ovarian cancer? an exploratory analysis of the European Organisation for Research and Treatment of Cancer 55971 randomised trial.
Eur J Cancer. 2013;49(15):3191-3201.
PubMedGoogle ScholarCrossref 14.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(8):488-490.
PubMedGoogle ScholarCrossref 15.Klabunde
CN, Potosky
AL, Legler
JM, Warren
JL. Development of a comorbidity index using physician claims data.
J Clin Epidemiol. 2000;53(12):1258-1267.
PubMedGoogle ScholarCrossref 16.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(5):373-383.
PubMedGoogle ScholarCrossref 17.Winchester
DP, Stewart
AK, Phillips
JL, Ward
EE. The National Cancer Data Base: past, present, and future.
Ann Surg Oncol. 2010;17(1):4-7.
PubMedGoogle ScholarCrossref 18.Mallin
K, Palis
BE, Watroba
N,
et al. Completeness of American cancer registry treatment data: implications for quality of care research.
J Am Coll Surg. 2013;216(3):428-437.
PubMedGoogle ScholarCrossref 19.Rosenbaum
PR, Rubin
DB. The central role of the propensity score in observational studies for causal effects.
Biometrika. 1983a;70:41-55.
Google ScholarCrossref 20.Austin
PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples.
Stat Med. 2009;28(25):3083-3107.
PubMedGoogle ScholarCrossref 21.Bristow
RE, Palis
BE, Chi
DS, Cliby
WA. The National Cancer Database report on advanced-stage epithelial ovarian cancer: impact of hospital surgical case volume on overall survival and surgical treatment paradigm.
Gynecol Oncol. 2010;118(3):262-267.
PubMedGoogle ScholarCrossref 22.Bristow
RE, Puri
I, Diaz-Montes
TP, Giuntoli
RL, Armstrong
DK. Analysis of contemporary trends in access to high-volume ovarian cancer surgical care.
Ann Surg Oncol. 2009;16(12):3422-3430.
PubMedGoogle ScholarCrossref 23.Rosenbaum
PR, Rubin
DB. Assessing sensitivity to an unobserved binary covariate in an observational study with binary outcomes.
J R Stat Soc [Ser A]. 1983;45:212-218.
Google Scholar 24.Lin
DY, Psaty
BM, Kronmal
RA. Assessing the sensitivity of regression results to unmeasured confounders in observational studies.
Biometrics. 1998;54(3):948-963.
PubMedGoogle ScholarCrossref 25.Horowitz
NS, Miller
A, Rungruang
B,
et al. Does aggressive surgery improve outcomes? interaction between preoperative disease burden and complex surgery in patients with advanced-stage ovarian cancer: an analysis of GOG 182.
J Clin Oncol. 2015;33(8):937-943.
PubMedGoogle ScholarCrossref 26.Teramukai
S, Ochiai
K, Tada
H, Fukushima
M; Japan Multinational Trial Organization OC01-01. PIEPOC: a new prognostic index for advanced epithelial ovarian cancer—Japan Multinational Trial Organization OC01-01.
J Clin Oncol. 2007;25(22):3302-3306.
PubMedGoogle ScholarCrossref 27.Chetrit
A, Hirsh-Yechezkel
G, Ben-David
Y, Lubin
F, Friedman
E, Sadetzki
S. Effect of BRCA1/2 mutations on long-term survival of patients with invasive ovarian cancer: the national Israeli study of ovarian cancer.
J Clin Oncol. 2008;26(1):20-25.
PubMedGoogle ScholarCrossref 28.Bristow
RE, Eisenhauer
EL, Santillan
A, Chi
DS. Delaying the primary surgical effort for advanced ovarian cancer: a systematic review of neoadjuvant chemotherapy and interval cytoreduction.
Gynecol Oncol. 2007;104(2):480-490.
PubMedGoogle ScholarCrossref 29.Lin
JJ, Egorova
N, Franco
R, Prasad-Hayes
M, Bickell
NA. Ovarian cancer treatment and survival trends among women older than 65 years of age in the United States, 1995-2008.
Obstet Gynecol. 2016;127(1):81-89.
PubMedGoogle ScholarCrossref 30.Wright
JD, Ananth
CV, Tsui
J,
et al. Comparative effectiveness of upfront treatment strategies in elderly women with ovarian cancer.
Cancer. 2014;120(8):1246-1254.
PubMedGoogle ScholarCrossref 31.Dewdney
SB, Rimel
BJ, Reinhart
AJ,
et al. The role of neoadjuvant chemotherapy in the management of patients with advanced stage ovarian cancer: survey results from members of the Society of Gynecologic Oncologists.
Gynecol Oncol. 2010;119(1):18-21.
PubMedGoogle ScholarCrossref 32.Cornelis
S, Van Calster
B, Amant
F, Leunen
K, van der Zee
AG, Vergote
I. Role of neoadjuvant chemotherapy in the management of stage IIIC-IV ovarian cancer: survey results from the members of the European Society of Gynecological Oncology.
Int J Gynecol Cancer. 2012;22(3):407-416.
PubMedGoogle ScholarCrossref