This study shows that a fall in serum hCG between days 0–4 of treatment represents an 85% likelihood of treatment success with no further intervention, medical or surgical, for single-dose methotrexate treatment of ectopic pregnancy. We analysed the early serum hCG trends of 206 women treated with single-dose methotrexate for their ectopic pregnancies in a range of treatment centres, and these are numbers far greater than previously reported (n=30, n=45 and n=129)
A large number of women were excluded from our analysis due to inaccurate timing of (or missed) serum hCG level measurements, in contrast to the requirements of the single-dose methotrexate treatment protocol. Specifically, the analyses performed in this study required serum hCG levels to be available for days 0 or 1, day 4 and day 7 after methotrexate treatment. While this raises the possibility of a selection bias, it is not immediately clear how this may have affected our results. Certainly attendance at such numerous and specific time points for serum sampling requires significant patient (and physician) compliance, but non-compliance is unlikely to have any bearing on treatment outcome, which was the primary outcome assessed in this study. Indeed, the results obtained are consistent with those of previous studies in other populations
[7–9]. This study therefore, strengthens the validity of the prognostic value of a falling serum hCG between days 0–4 after single-dose methotrexate treatment for ectopic pregnancy.
Treatment success for the purposes of this study was defined more strictly than in the current protocol and clinical practice, in that it did not allow for any additional doses of methotrexate. This was for the purposes of rigorously testing the prognostic value of a falling serum hCG between days 0–4 for an actual single (rather than a variable) dose methotrexate treatment course. Allowing for additional doses of methotrexate is likely to only improve the positive predictive value of this measure.
The application of a quantified ≥20% fall in serum hCG between days 0–4 of medical treatment improves the positive predictive value from 85% to 94%. This is likely explained by the fact that the greater the fall in serum hCG between days 0–4 of treatment, the more likely the patient is to experience treatment success. There is a clear trade-off, however, between increasing accuracy of prediction with such a cut-off (specificity) and clinical applicability of this measure to a greater number of women (sensitivity). The sensitivity of a falling serum hCG between days 0–4 fell substantially from 64% to 40% when a ≥20% cut-off was applied, so that the test could only provide meaningful prognostic information to 29% (59/206) of women. In contrast, prognostic information was available for 53% (110/206) of women if any fall in serum hCG between days 0–4 was used as the cut-off for a measure of treatment efficacy, with no difference in prognostic accuracy.
It is possible that women with a falling serum hCG between days 0–4 after single-dose methotrexate for ectopic pregnancies may have had already failing pregnancies and did not require treatment. This was not the case, however, as inspection of the pre-treatment serum hCG trends for women with early falling serum hCG levels after methotrexate indicated that the majority (62%) in fact had rising serum hCG levels prior to treatment. Interestingly, regardless of a rising or falling pre-treatment hCG trend, a falling serum hCG level between days 0–4 predicted single-dose methotrexate treatment success with equal measure in this sub-cohort of women (85% and 84%, respectively).
This study examined ectopic pregnancies with pre-treatment serum hCG levels of ≤3000 IU/L, and the results may not apply to medically treated ectopic pregnancies with pre-treatment serum hCG levels <3000 IU/L. Furthermore, given that a fall in serum hCG between days 0–4 is not 100% predictive of treatment success, it is still prudent to continue monitoring hCG levels until normalisation.