The role of contraception in the development of postmolar gestational trophoblastic tumor

From January 1974 to June 1988, 299 evaluable patients were referred to the John I. Brewer Trophoblastic Disease Center of Northwestern University Cancer Center for treatment and/or follow-up of a hydatidiform mole (N = 162) or postmolar gestational trophoblastic tumor (N = 137). The type of contraception and other prognostic factors before and after evacuation were correlated with the development of gestational trophoblastic tumor using both univariate and multivariate analysis. There was no relationship between pre-hydatidiform mole contraception and the development of gestational trophoblastic tumor. Oral contraceptives (OCs) were used by 139 patients (46%), barrier methods by 141 patients (47%), intrauterine devices (IUDs) by two patients (1%), and no contraception by 17 patients (6%). The risk of developing gestational trophoblastic tumor was compared between patients using versus not using: OCs--33 versus 57% (P less than .001), barrier methods--53 versus 40% (P = .30), IUD--100 versus 46% (P = .21), and any contraceptive method--43 versus 88% (P less than .001). The dose of estrogens could be determined in 75 patients taking OCs; 14 of 49 (29%) of the patients taking less than 50 micrograms versus nine of 26 (35%) taking 50 micrograms or more developed gestational trophoblastic tumor (P = .78). Stepwise logistic regression analysis demonstrated that the type of contraceptive used was the most important prognostic factor in gestational trophoblastic tumor development (P less than .0001), followed by the occurrence of theca-lutein cysts (P less than .0001), Asian maternal race (P = .02), lesser time from the last menstrual period (P = .005), and greater maternal age (P = .04).(ABSTRACT TRUNCATED AT 250 WORDS)