Early (Days 1–4) post-treatment serum hCG level changes predict single-dose methotrexate treatment success in tubal ectopic pregnancy

Abstract

Abstract STUDY QUESTION What is the capacity of the change between Day 1 and Day 4 post-treatment serum human chorionic gonadotropin (hCG) levels for predicting single-dose methotrexate treatment success in tubal ectopic pregnancy? SUMMARY ANSWER Any fall in Days 1–4 serum hCG signified an 85% (95% CI 76.8–90.6) likelihood of treatment success for women with tubal ectopic pregnancy (initial hCG of ≥1000 and ≤5000 IU/l) managed with single-dose methotrexate. WHAT IS KNOWN ALREADY For those with tubal ectopic pregnancy managed by single-dose methotrexate, current guidelines advocate intervention if Days 4–7 hCG fails to fall by >15%. The trajectory of hCG over Days 1–4 has been proposed as an early indicator that predicts treatment success, allowing early reassurance for women. However, almost all prior studies of Days 1–4 hCG changes have been retrospective. STUDY DESIGN, SIZE, DURATION This was a prospective cohort study of women with tubal ectopic pregnancy (pre-treatment hCG of ≥1000 and ≤5000 IU/l) managed with single-dose methotrexate. The data were derived from a UK multicentre randomized controlled trial of methotrexate and gefitinib versus methotrexate and placebo for treatment of tubal ectopic pregnancy (GEM3). For this analysis, we include data from both treatment arms. PARTICIPANTS/MATERIALS, SETTING, METHODS Participants were categorized according to single-dose methotrexate treatment success or failure. Treatment success for this analysis was defined as complete and uneventful resolution of tubal ectopic pregnancy to serum hCG <30 IU/l following single-dose methotrexate treatment without additional treatment. Patient characteristics of the treatment success and failure groups were compared. Changes in Days 1–4, 1–7, and 4–7 serum hCG were evaluated as predictors of treatment success through receiver operating characteristic curve analysis. Test performance characteristics were calculated for percentage change ranges and thresholds including optimal classification thresholds. MAIN RESULTS AND THE ROLE OF CHANCE A total of 322 women with tubal ectopic pregnancy were treated with single-dose methotrexate. The overall single-dose methotrexate treatment success rate was 59% (n = 189/322). For any fall in serum hCG on Days 1–4, likelihood ratios were >3, while for any fall of serum hCG >20% on Days 1–7, likelihood ratios reached 5. Any rise of serum hCG on Days 1–7 and 4–7 strongly reduced the chance of success. Any fall in Days 1–4 hCG predicted single-dose methotrexate treatment success with a sensitivity of 58% and specificity 84%, resulting in positive and negative predictive values of 85% and 57%, respectively. Any rise in Days 1–4 serum hCG <18% was identified as an optimal test threshold that predicted treatment success with 79% sensitivity and 74% specificity, resulting in 82% positive predictive value and 69% negative predictive value. LIMITATIONS, REASONS FOR CAUTION Our findings may be limited by intervention bias resulting from existing guidelines which influences evaluation of hCG changes reliant on Day 7 serum hCG levels. WIDER IMPLICATIONS OF THE FINDINGS Examining a large prospective cohort, we show the value of Days 1–4 serum hCG changes in predicting single-dose methotrexate treatment success in tubal ectopic pregnancy. We recommend that clinicians provide early reassurance to women who have a fall or only a modest (<18%) rise in Days 1–4 serum hCG levels, that their treatment will likely be effective. STUDY FUNDING/COMPETING INTEREST(S) This project was supported by funding from the Efficacy and Mechanism Evaluation programme, a Medical Research Council and National Institute for Health Research partnership (grant reference number 14/150/03). A.W.H. has received honoraria for consultancy for Ferring, Roche, Nordic Pharma and AbbVie. W.C.D. has received honoraria from Merck and Guerbet and research funding from Galvani Biosciences. L.H.R.W. has received research funding from Roche Diagnostics. B.W.M. is supported by a NHMRC Investigator grant (GNT1176437). B.W.M. also reports consultancy for ObsEva and Merck and travel support from Merck. The other authors declare no competing interests. TRIAL REGISTRATION NUMBER This study is a secondary analysis of the GEM3 trial (ISRCTN Registry ISRCTN67795930).