Nomogram for survival after primary surgery for bulky stage IIIC ovarian carcinoma

Abstract

Objective.

Nomograms have been developed for numerous malignancies to predict a specific individual's probability of long-term survival based on known prognostic factors. To date, only one prediction model has been reported for patients with epithelial ovarian carcinoma (EOC). The objective of this study was to develop a more accurate survival nomogram for patients with bulky stage IIIC EOC.

Patients and methods.

Nomogram predictor variables included age, tumor grade, histologic type, preoperative platelet count, ascites, and residual disease after primary cytoreduction. Disease-specific survival was estimated by the Kaplan–Meier method. Cox proportional hazards regression was used for multivariate analysis, which was the basis for the nomogram. The concordance index was used as an accuracy measure with bootstrapping to correct for optimistic bias.

Results.

A total of 424 evaluable patients with bulky stage IIIC EOC underwent primary surgery at our institution during the study period of 1/89 to 12/03. All patients received postoperative platinum-based systemic chemotherapy. EOC-specific survival at 5 years was 51%. Using the six predictor variables, a nomogram was constructed and internally validated using bootstrapping. It was shown to have excellent calibration with a bootstrap corrected concordance index of 0.67, which was more accurate in predicting survival at this stage than the previously published model (concordance index = 0.53).

Conclusion.

Utilizing six readily accessible predictor variables, our nomogram more accurately predicted 5-year disease-specific survival for bulky stage IIIC EOC than the previously published model. This tool may be useful for patient counseling, determination of clinical trial eligibility, and postoperative management.

Introduction

The American Cancer Society estimates that there will be 22,430 newly diagnosed cases of ovarian cancer and 15,280 deaths due to the disease in 2007, making ovarian carcinoma the fifth leading cause of cancer deaths in American women [1]. The majority of these newly diagnosed cases will be designated as stage IIIC disease as per the staging system of the International Federation of Gynecology and Obstetrics (FIGO) due to bulky (> 2 cm) upper abdominal disease [2]. The standard management of these patients includes cytoreductive surgery followed by platinum-based chemotherapy [3].

Numerous factors, such as age, race, tumor grade, histology, FIGO stage, preoperative serum CA125 levels and platelet counts, the presence or absence of ascites, residual disease after surgery, and various molecular markers, have been reported to be important prognostic factors for women with ovarian cancer [4], [5], [6], [7]. Prognostic nomograms attempt to combine proven prognostic factors to quantify risk as precisely as possible in order to explicitly predict outcomes. Nomograms have been developed as predictive tools for outcomes in malignancies such as prostate cancer, sarcoma, and gastric carcinoma [8], [9], [10]. To date, however, only one study has attempted to develop a nomogram for survival in ovarian cancer [11]. This previous model has not gained widespread acceptance or utilization.

The objective of this study was to develop a nomogram to predict the probability of a 5-year survival after primary cytoreductive surgery, specifically for bulky stage IIIC ovarian carcinoma, and to compare its accuracy with the previously published model.