Objective: This study aimed to develop a nomogram predicting the probability of relapse in individual patients who have surgery for borderline ovarian tumors (BOTs).
Methods: This retrospective study included 801 patients with BOT diagnosed between 1985 and 2008 at 6 gynecologic cancer centers. We analyzed covariates that were associated with the risk of developing a recurrence by multivariate logistic regression. We identified a parsimonious model by backward stepwise logistic regression. The 5 most significant or clinically important variables associated with an increased risk of recurrence were included in the nomogram. The nomogram was internally validated.
Results: Fifty-one patients developed a recurrence after a median observation period of 57 months. Age at diagnosis, the International Federation of Gynecology and Obstetrics stage, cell type, preoperative serum CA125, and type of surgery (radical vs fertility-sparing) were associated with an increased risk of recurrence and were used in the nomogram. Bootstrap-corrected concordance index was 0.67 and showed good calibration.
Conclusions: Five factors that are commonly available to clinicians treating patients with BOT were used in the development of a nomogram to predict the risk of recurrence. The nomogram will be useful to counsel patients about risk-reduction strategies to minimize the risk of recurrence or to inform patients about a very low risk of recurrence making intensive follow-up unwarranted.
*Queensland Centre for Gynaecological Cancer, Royal Brisbane & Women’s Hospital; †University of Queensland, Brisbane, Australia; ‡Queen Mary Hospital, Hong Kong; §Radboud University Medical Centre, Nijmegen, The Netherlands; ∥Royal Women’s Hospital Melbourne; ¶Royal Prince Alfred Hospital, Sydney; #School of Women’s and Infants’ Health, University of Western Australia, Perth; and **School of Public Health, Queensland University of Technology, Brisbane, Australia.
Address correspondence and reprint requests to Andreas Obermair, MD, FRANZCOG, CGO, School of Medicine, The University of Queensland, Brisbane, Royal Brisbane and Women’s Hospital, 6th Floor, Ned Hanlon Bldg, Butterfield St, Herston, QLD 4029, Australia. E-mail: Obermair@powerup.com.au.
This study was partly funded by the Royal Brisbane and Women’s Hospital Research Foundation, Brisbane, Australia.
The authors declare no conflicts of interest.
Article Category: Cancer therapy
Received July 27, 2012
Accepted October 29, 2012