Endometrioma excision and ovarian reserve; do assessments by antral follicle count and anti-Mullerian hormone yield contradictory results?

Sir, We read the systematic review regarding the effect of endometrioma excision on ovarian reserve as assessed by antral follicle count by Muzii et al. with great interest (Muzii et al., 2014). Muzii et al. suggest that sample stability issues limit the value of anti-Mullerian hormone (AMH) assay in evaluating ovarian reserve and therefore that conclusions of two previous meta-analyses, both of which reported a detrimental effect of endometrioma excision on ovarian reserve as assessed by AMH, might be erroneous (Raffi et al., 2012; Somigliana et al., 2012; Muzii et al., 2014). It is true that handling and storage conditions of the sample at room temperature or −20°C before analysis can lead to systematic variations in AMH readings (Rustamov et al., 2014). However, as all samples would be handled similarly in a particular center it is unlikely that a systematic error would be introduced in pre- and post-operative comparisons. It could be argued that preoperatively collected samples would systematically yield higher readings than post-operative samples, if the former were kept in storage at room temperature or −20°C for a longer period before actual testing. This could augment the difference between preoperative and post-operative readings. However, only one of the eight studies included in the meta-analysis by Raffi et al., reported storing samples at −20°C (Biacchiardi et al., 2011), while samples were either analyzed directly without cryostorage (Tsolakidis et al., 2010; Hwu et al., 2011) or were stored at −70 or −80°C before actual measurements in the remaining five studies (Ercan et al. 2010, 2011; Hirokawa et al., 2011; Kitajima et al., 2011; Lee et al., 2011). AMH levels are expected to remain stable at temperatures of −70 or −80°C (Rustamov et al., 2014). When the meta-analysis of Raffi et al. is repeated with exclusion of the study in which samples were stored at −20°C the conclusion remains essentially unchanged (Biacchiardi et al., 2011; Raffi et al., 2012). Serum AMH levels still appear to be significantly decreased by an average of 0.9 ng/ml following endometrioma excision (Fig. 1). Weighted mean difference in serum AMH after laparoscopic endometrioma excision. Data represent re-analysis of studies included in Raffi et al. (2012), with the exclusion of serum samples stored at −20°C until analysis. (Adapted from Raffi et al., 2012). Muzii et al. claim that antral follicle count (AFC) can be a better marker of ovarian reserve (Muzii et al., 2014). We disagree with this view for several reasons. Firstly, when compared with AFC, AMH, despite its limitations, was found to have similar correlation with primordial follicle count and similar capacity for predicting ovarian response to stimulation (Hansen et al., 2011; Broer et al., 2013a,b). Secondly, inter-cycle and intra-individual variation of AMH was found to be significantly less than that of AFC in several studies (van Disseldorp et al., 2010). Last but not least, reliability of AFC is questionable in the presence of an endometrioma as visualization of antral follicles may be obscured by a cyst occupying a substantial portion of the ovary. This could lead to underestimation of AFC preoperatively and obscure any decrease in the post-operative period. AMH can be expected to outperform AFC in this regard. Muzii et al. suggest that AFC enables individual evaluation of the ovary, harboring the endometrioma that was subsequently removed, and therefore that AFC could better reflect the effect of surgery on ovarian reserve than AMH, which is a systemic marker indicating the combined reserve of both ovaries (Muzii et al., 2014). However, they have neglected this point in their meta-analysis, including studies that report the total AFC for both ovaries, even for patients who had a unilateral endometrioma (Tsolakidis et al., 2010; Celik et al., 2012; Uncu et al., 2013; Urman et al., 2013; Alborzi et al., 2014). When the meta-analysis is limited to studies that compared AFC in the same ovary before and after surgery, the results are somewhat changed, demonstrating a strong trend toward decreased AFC following surgery (Fig. 2) (Biacchiardi et al., 2011; Ercan et al., 2011; Var et al., 2011; Zaitoun et al., 2013). Antral follicle counts in the same ovary before and after excision of endometrioma. Data represent re-analysis of studies included in Muzii et al. (2014), with the exclusion of studies involving non-operated gonad in antral follicle count. (Adapted from Muzii et al., 2014). In conclusion, we think it is not justified to suggest (i) AFC is a better marker of ovarian reserve in the presence of an endometrioma or (ii) AMH and AFC yield contradicting results for comparisons of ovarian reserve before and after endometrioma excision. In our opinion both tests suggest a decrease in ovarian reserve following endometrioma excision. It would still be prudent to warn patients, especially those suffering from infertility, about the detrimental effect of endometrioma excision on ovarian reserve, which is clearly demonstrated in studies comparing oocyte yield from operated and intact gonads in the IVF setting (Somigliana et al., 2011).