Predictors of methotrexate treatment failure in ectopic pregnancy

To determine the possible predictors of methotrexate treatment failure in ectopic pregnancy.

Fifty-eight patients diagnosed with ectopic pregnancy were treated with methotrexate (50 mg/m2). Selected variables in the history of the patients, the signs and symptoms at the time of admission, transvaginal ultrasound findings and serum beta-human chorionic gonadotropin (beta-hCG) levels on day 1 and 3 were evaluated in a logistic regression model to predict treatmentfailure, defined as tubal rupture.

Methotrexate treatment failed in 9 cases (15.5 %). Another 9 cases (15.5%) required a second dose of methotrexate, and no treatment failures were observed in these cases. The presence of subchorionic tubal hematoma in the ectopic gestation (OR = 22.9, CI = 2.7-194.7, p = 0.004), the presence of an embryo (OR = 24, CI = 2.1-269, p = 0.01) and day 1 serum beta-hCG level > or = 3,000 mIU/mL (OR = 27.1, CI = 2.1-342.5, p = 0.01) were the main predictors of treatment failure. Follow-up serum beta-hCG levels > or =3,500 mIU/mL (OR = 42.9, CI = 4.3-421) on day 3 were significant predictors of treatment failure. Follow-up risk score was calculated as > 4 on day 3 by adding day 3 serum beta-hCG level to the admission score. Only 1 treatment failure (2.4%) occurred in 42 patients with an admission score of nil. No treatment failure occurred in 39 patients whose follow-up score was nil. The increase in admission risk (OR = 32.1, CI = 3.8-270, p = 0.001) and follow-up risk (OR = 9.2, CI = 2.4-35.2) were significant predictors of treatment failure.

Transvaginal ultrasound findings are as important as serum beta-hCG level on the first day of methotrexate treatment. In unruptured cases, day 3 serum beta-hCG level is important to reevaluate the decision to continuefollow-up or perform early surgery for increased risk of treatment failure.