[Use of biochemical markers to predict the outcome of pregnancies conceived by in vitro fertilization].

BACKGROUND

Pregnancies obtained after in vitro fertilization and embryo transfer are at increased risk for an adverse outcome compared with women who conceive naturally. Multiple gestations also occur more frequently after in vitro fertilization. Therefore, there is a need for markers that accurately detect the establishment of pregnancy and predict its outcome as early as possible, allowing for modification of monitoring and treatment if required. Ultrasound examination is part of the routine follow-up after in vitro fertilization, but a gestational sac is not reliably visible until 33-37 days after ovulation induction. As a result, there is an ongoing effort to find endocrine markers that can earlier detect the establishment of pregnancy and forecast its outcome.

OBJECTIVE

The authors' aim was to assess the predictive value of the following potential serum markers, measured in the second week after embryo transfer in samples collected prospectively during the past ten years at the Division of Assisted Reproduction of their department: total beta-hCG (theoretical post-embryo transfer day 11 values, calculated from levels in two samples collected with a difference of two days, based on the mathematical model describing its exponential increase in early pregnancy), inhibin A, and CA-125.

METHODS

Data of patients undergoing IVF or intracytoplasmic sperm injection and embryo transfer between 1995 and 2001 were analyzed. Establishment of pregnancy was assessed by measuring total beta-hCG concentrations in two serum samples collected between 8 and 16 days after ET with a difference of two days. Measurement of inhibin A and CA-125 levels was performed in the same samples. Logistic regression analyses were used to study the association of these serum markers and the number of retrieved oocytes and transferred embryos with pregnancy outcome. Receiver-operating characteristic (ROC) curves were constructed to identify optimal cutoff levels for outcomes and to assess overall predictive accuracy.

RESULTS AND CONCLUSIONS

(1) Day 11 total beta-hCG can be used to compare hCG levels in samples from different sampling days and to predict early pregnancy losses and multiple ongoing pregnancies with high sensitivity and specificity. (2) Inhibin A concentrations are more accurate than day 11 hCG levels for predicting preclinical abortion after IVF but they have no advantage in forecasting ongoing or multiple ongoing pregnancies. (3) Prognostic accuracy of CA-125 measurements for the prediction of pregnancy as well as its outcome is inferior to that achieved with inhibin A.