BACKGROUND. The authors conducted this study to update their experienc
e with patients who have ovarian serous borderline tumors with noninva
sive peritoneal implants, with the objectives of gaining additional in
sight into the biologic behavior of these tumors and understanding bet
ter the effects of postoperative treatment. METHODS. Seventy-three pat
ients who had ovarian serous borderline tumors with noninvasive perito
neal implants were identified in a retrospective review. Major end poi
nts selected for analysis were surgicopathologic response, time to rel
apse, type of relapse, progression free survival, and overall survival
. Univariate and multivariate regression analyses were also performed.
RESULTS. The median follow-up time was 10.3 years. Of 20 patients wit
h macroscopic residual disease at completion of initial surgery who su
bsequently underwent second-look surgery, 3 (15%) had a response to ch
emotherapy. Twenty-two of 73 patients (30%) either developed progressi
ve disease or had a relapse. The median time from the date of diagnosi
s to relapse was 7.1 years. Tissue was available from 20 of the 22 pat
ients who had a relapse; 14 (70%) had invasive low grade serous carcin
omas, and 6 (30%) had recurrent borderline tumors. Age was the only fa
ctor studied that had a significant influence on survival (P = 0.03).
In both univariate and multivariate proportional hazards models, age a
nd residual disease were found to be of borderline significance in pre
dicting cancer specific survival. CONCLUSIONS. Approximately 30% of pa
tients who have ovarian serous borderline tumors with noninvasive peri
toneal implants will develop progressive or recurrent tumors, most com
monly serous carcinomas. The presence of macroscopic residual disease
appears to be a predictor of disease free survival. In this study, how
ever, the authors were unable to elucidate the role of postoperative t
herapy or determine criteria for selecting patients for such therapy.
Cancer 1998;83:2157-63, (C) 1998 American Cancer Society.