HCG supplementation of controlled ovarian stimulation cycles

Abstract

Sir, We read with great interest the paper by Thuesen et al. (2012) in Human Reproduction on the effects of controlled ovarian stimulation (COS) with HCG supplementation on progesterone (P4) pretriggering values and ART outcome. According to Thuesen et al., HCG administration during COS with recombinant FSH increased P4 values. This observation is in contrast to a previously suggested reduction due to androgen conversion. Moreover, in the Thuesen study, pregnancy rates were similar in the different groups, irrespective of pre-triggering P4 levels. Thus, this is another study that seemingly refutes the prognostic value of pre-triggering P4 levels. Conversely, a number of studies suggest a clear negative prognostic effect of P4 pre-triggering increase (Bosh et al., 2010; Xu et al., 2012). Thus, there is an intriguing lack of agreement between authors on the prognostic role of P4 value. Some authors have suggested that the lack of agreement on this relevant issue could be due to P4 cut-off levels chosen to discriminate for good or bad prognosis (Bosh et al., 2010). We have recently made observations of the discrepancy between the prognostic value of P4 pre-triggering levels in two different clinical settings using different immuno-metric methods for P4 determination. Interestingly, the sensitivity was completely different between the two methods employed: 0.2 versus 0.08 ng/ml, respectively. The prognostic value was really good with the method with low sensitivity–high specificity but totally absent with the other. Even more interestingly samples from the same patients tested with both methods showed that 4/6 values ,1.4 ng/ml with one method were higher with the other. So even with appropriate cut-off levels, patients could be wrongly assigned to high or low P4 value group due to methodological bias (unpublished data). Unfortunately, this observation remains as yet unpublished as (according to the review team) ‘this article adds little to the literature’. On the contrary, reading the paper by Thuesen et al. from one of the most important centers of ART in north Europe using a method for P4 evaluation with 0.03 ng/ml of sensitivity, we are not fully convinced that the pivotal importance of the specificity of the method for premature luteinization evaluation is a largely shared knowledge between experts. In other words, due to the need to discriminate small increments of P4 levels, we remain convinced that the choice of extremely sensitive methods could be misleading due to their lack of specificity. The choice of the method is irrelevant when dosing high P4 levels for the diagnosis of ovulation but absolutely critical when looking for a small P4 increase.