Reply: Maternal and neonatal outcomes in women diagnosed with PCOS by different definitions Free

Sir,

We thank Dr Palomba and Dr La Sala for their comments on our study (Kollmann et al., 2015). To the best of our knowledge, our population is one of the largest cohorts of pregnant women with polycystic ovary syndrome (PCOS) and we therefore believe that our findings are representative allowing valid conclusions. However, we acknowledge that our study had certain limitations related to its retrospective design, which increases the likelihood of bias and omits several parameters (e.g. mode of conception). Our three groups [NIH (National Institute of Health), ESHRE/ASRM (European Society of Human Reproduction and Embryology/American Society of Reproductive Medicine) and AE-PCOS (Androgen Excess and PCOS Society)] are mutually exclusive and one difference between the groups is the ovulatory status. However, our objective was to compare the participants that would be added by the different criteria and not to compare an ovulatory and oligoanovulatory group. We agree with the notion that the ‘ESHRE phenotype’ represents a subgroup of patients with milder metabolic morbidity, supported by our finding of lower incidence in insulin resistance in this very group. However, the aim of the study was to evaluate the impact on perinatal complications, which ultimately did not differ within the PCOS subgroups. It must therefore be stressed that by using the current criteria for PCOS, all affected pregnant women should receive careful surveillance of their pregnancy.

Dr Palomba and Dr La Sala further reemphasize that women with PCOS should be characterized according to all diagnostic and non-diagnostic features in order to define and grossly quantify their obstetric risk and we agree that this is an important issue. However, it may be that current characterization is insufficient to differentiate high risk from low risk. The role of subtle ovarian morphology in this matter is not yet defined. Improvements in the resolution of ultrasound, achieved through both software and hardware developments, has increased our ability of detecting smaller follicles (Martins et al., 2014). Using a follicle number per ovary (FNPO) ≥12, suggested by ASRM/ESHRE consensus as a diagnostic criterion more than 10 years ago (ESRHE/ASRM, 2004) is now below the average FNPO reported by recent studies for some populations (Dewailly et al., 2014). Considering that, it is now suggested to use a FNPO ≥25 to identify women that are likely to have hyperandrogenemia (Dewailly et al., 2014); and by using that, we may have less problems related to different phenotypes; i.e. we would have much less persons classified as having the ‘ESHRE phenotype’. However, as noted by Dr Palomba and Dr La Sala, there is wide variability in phenotypes and pregnancy complications which may hinder the estimation of individual risks.

In conclusion, we believe that our study on this large population adds to current knowledge and will increase the awareness of clinicians in taking care for pregnant women with PCOS.

References