Antral follicle count as a predictor of ovarian responsiveness in women with endometriomas or with a history of surgery for endometriomas Laura Benaglia, M.D.,a Giorgio Candotti, M.D.,a Andrea Busnelli, M.D.,a,b Alessio Paffoni, M.Sc.,a Paolo Vercellini, M.D.,a,b and Edgardo Somigliana, M.D., Ph.D.a,b a degli Department of Obstetrics and Gynecology, Fondazione Ca' Granda, Ospedale Maggiore Policlinico; and b Universita Studi di Milano, Milan, Italy
Objective: To evaluate the accuracy of antral follicular count (AFC) in predicting ovarian responsiveness in ovaries with endometriomas or with a past history of surgical excision of endometriomas. Design: Retrospective review. Setting: Academic hospital. Patient(s): Eighty-three women for a total of 166 gonads. Intervention(s): None. Main Outcome Measure(s): Total number of developing follicles. Result(s): The ovaries were characterized as four groups: [1] unoperated gonads without endometriomas (n ¼ 42, control group), [2] unoperated gonads with endometriomas (n ¼ 46), [3] operated gonads without endometriomas (n ¼ 55), and [4] operated gonads with endometriomas (n ¼ 23). The analyses subsequently considered all ovaries with endometriomas (groups 2 þ 4, n ¼ 69) and all operated ovaries (groups 3 þ 4, n ¼ 78). The capacity of AFC to predict low response (%2 follicles) or hyperresponsiveness (R7 follicles) was evaluated using receiver operating characteristic curves. We used a linear regression model to calculate the adjusted B coefficients. The adjusted B coefficients in unaffected ovaries, in all ovaries with endometriomas, and in all operated ovaries were 0.55 (95% confidence interval [CI], 0.07–1.03), 0.76 (95% CI, 0.54–0.98), and 0.51 (95% CI, 0.26–0.76), respectively. The area under the curve (AUC) for the prediction of low response was 0.83 (95% CI, 0.68–0.99), 0.83 (95% CI, 0.73–0.93), and 0.74 (95% CI, 0.63–0.85), respectively. The AUC for the prediction of hyperresponse was 0.84 (95% CI, 0.70–0.97), 0.74 (95% CI, 0.63–0.85), and 0.77 (0.60–0.94), respectively. Conclusion(s): The accuracy of AFC for predicting ovarian response is similar in unaffected Use your smartphone ovaries, ovaries with endometriomas and ovaries with a history of surgery for endometriomas. to scan this QR code (Fertil SterilÒ 2015;103:1544–50. Ó2015 by American Society for Reproductive Medicine.) and connect to the Key Words: AFC, controlled ovarian hyperstimulation, endometrioma Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/benaglial-afc-ovarian-endometriomas/
T
he impact of ovarian endometriomas and their surgical removal on the ovarian reserve is still a debated issue. There is consistent biological evidence suggesting a detrimental effect of these cysts on the ovarian reserve (1–3). Potential
mechanisms of damage include the diffusion of toxic substances from the endometrioma to the ovarian tissue (2) and the burn-out effect of damage consequent to an enhanced recruitment of primordial follicles and consequent accelerated exhaustion of the ovarian
Received November 25, 2014; revised March 9, 2015; accepted March 10, 2015; published online April 29, 2015. L.B. has nothing to disclose. G.C. has nothing to disclose. A.B. has nothing to disclose. A.P. has nothing to disclose. P.V. has nothing to disclose. E.S. has nothing to disclose. Reprint requests: Laura Benaglia, M.D., Infertility Unit, Fondazione Ca' Granda, Ospedale Maggiore Policlinico, Via M. Fanti 6, 20122 Milan, Italy (E-mail: [email protected]). Fertility and Sterility® Vol. 103, No. 6, June 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.03.013 1544
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
reserve (3). This latter mechanism is strongly supported by two recent contributions that documented a lower number of primordial follicles and an increased proportion of activated and atretic follicles in ovaries with endometriomas (1, 3). In contrast, available clinical data have failed to document a relevant impact. In women with unilateral endometriomas, ovarian responsiveness to hyperstimulation does not differ between the two gonads, and ovarian reserve appears to be only modestly affected in women with bilateral cysts (4–6). VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility® Laparoscopic stripping has been shown to damage the ovarian reserve by most but not all investigators (7–9). Moreover, recent preliminary evidence suggests that alternative surgical techniques such as laser vaporization, partial cystectomy, or surgical suture may cause less or no damage (10–16). Not surprisingly, this controversial scenario has had a clinical impact, and physicians frequently to have deal with a high degree of uncertainty. The decision to operate depends on several clinical factors and sonographic findings (17–20). A neglected but potentially relevant issue is the measurement of the ovarian reserve in ovaries with endometriomas or with a history of surgery for endometriomas. A clear knowledge of the remnant ovarian reserve of an affected or previously operated ovary may influence therapeutic decisions. Unfortunately, the evaluation of ovarian reserve in women with current or past ovarian endometriomas is challenging. In most cases, these cysts are unilateral, thus hampering the validity of hormone assessments such as serum follicle-stimulating hormone (FSH) or inhibin because the contralateral gonad may properly compensate for the reduced function of the affected one. The validity of antim€ ullerian hormone (AMH) is also debatable because the relative contribution of the affected and intact ovaries cannot be definitely discriminated. The accuracy of serum AMH is also questionable in women with bilateral cysts because it cannot be excluded that endometriomas may affect ovarian reserve differently in the same patient. At present, sonographic assessment of antral follicle count (AFC) is the only means of obtaining independent data on the ovarian reserve of an ovary. A huge amount of literature has validated the use of AFC as a surrogate measurement of ovarian reserve (21–24). However, to the best of our knowledge, even though AFC in ovaries with current or past endometriomas has been used repeatedly in research studies (10–13, 15, 16, 25–29), this tool has never been validated. Of note, based on the currently accepted recommendations, AFC should not be assessed in women with ovarian endometriosis or previous ovarian surgery (30). Our study focused on women undergoing in vitro fertilization (IVF) to validate the use of AFC in ovaries with current or past endometriomas. Ovarian responsiveness to hyperstimulation is actually considered the best noninvasive surrogate measurement of ovarian reserve (31). The correlation of AFC and ovarian responsiveness in ovaries with current or past endometriomas may thus provide valuable information on the accuracy of AFC in this context.
MATERIALS AND METHODS Our retrospective study at the Infertility Unit of the Department of Obstetrics and Gynecology of the Fondazione Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy, reviewed the clinical charts of women who underwent IVF cycles between January 2011 and March 2013. The inclusion criteria were as follows: [1] previous surgery for ovarian endometriomas and/or sonographic diagnosis of endometriomas at the time of the cycle, [2] availability of a pre-IVF assessment of AFC between days 2 and 4 of a spontaneous menstrual cycle, VOL. 103 NO. 6 / JUNE 2015
and [3] age %42 years. The exclusion criteria included [1] past history of ovarian surgery for nonendometrial lesions, [2] diagnosis of nonendometrial cysts at the time of IVF cycle, and [3] pre-IVF hormone treatment (progestins, estroprogestins, or gonadotropin-releasing hormone analogs). The women could be enrolled for only one cycle. Women who are referred to our unit are routinely asked to provide informed consent for their clinical data to be used for research purposes, and those who had refused consent were excluded from our study. The local institutional review board approved the study protocol. Our analyses focused on the relationship between AFC and the total number of developing follicles per ovary. Specifically, four different groups of gonads were initially considered: [1] unoperated ovaries without endometriomas (control group), [2] unoperated ovaries with endometriomas, [3] operated ovaries without endometriomas, and [4] operated ovaries with endometriomas. The analyses subsequently considered all ovaries with endometriomas together (groups 2 þ 4), all ovaries with endometriomas with a mean diameter >20 mm (subgroup of the latter), and all operated ovaries (groups 3 þ 4). During the study period, AFC was assessed based on the current available recommendations (30). Briefly, all identifiable antral follicles 2–10 mm in diameter were recorded. The diagnosis of endometrial cysts was performed by transvaginal ultrasound and had to be documented on at least two occasions and at least two menstrual cycles apart. More specifically, ovarian endometrioma was defined as a roundshaped cystic mass with a minimum diameter of 10 mm, with thick walls, regular margins, homogeneous low echogenic fluid content with scattered internal echoes, and without papillary proliferations (32). Women with atypical lesions were excluded, such as cysts whose sonographic appearance was compatible with but not distinctly identifiable as endometriosis. The diameter of the endometriomas was calculated as the mean of three perpendicular diameters. All sonographic evaluations were performed by four physicians who have had a long-term specialization in reproductive medicine, all of whom attended an internal meeting in 2010 to discuss the new recommended modality to assess AFC. In our unit, women with a history of surgery for ovarian endometriosis are routinely asked to provide documentation of the intervention; only those with a histologically confirmed diagnosis of endometriomas were included in the study. During the IVF cycle, the patients selected for IVF were monitored and managed according to standardized clinical protocols, as reported elsewhere (33). Briefly, the patients underwent a transvaginal ultrasound in the early proliferative phase of the cycle preceding the initiation of ovarian hyperstimulation. The AFC assessment was systematically recorded at this time and again on day 3 of the following menstrual cycle if the woman received a flare-up protocol or a protocol with gonadotropin-releasing hormone antagonists (if two AFC assessments were available, the latter one was used for the analysis). The regimen used and the dose of gonadotropins were determined on an individual basis according to the age, day-3 serum FSH value, serum AMH value, and AFC. During 1545
ORIGINAL ARTICLE: ENDOMETRIOSIS the stimulation, women underwent serial transvaginal ultrasounds and serum hormonal assessments when indicated. When leading follicles with a mean diameter >18 mm were visualized, human chorionic gonadotropin (hCG) was administered subcutaneously. All follicles with a mean diameter R11 mm were routinely recorded at each sonographic assessment and at the time of hCG administration to trigger ovulation. Cycles could be cancelled because of low response or hyperresponsiveness after a consultation with the patient. Cancellation for a poor response was considered when the leading follicle had a diameter >18 mm and the total number of follicles was 20. For cancelled cycles, our analysis used the number of follicles with a mean diameter R11 mm at the time of cancellation. Oocyte retrieval was performed transvaginally 36 hours after the hCG injection. Embryo transfer was performed 48 to 72 hours after oocyte collection or, in properly selected patients, at the blastocyst stage. Clinical pregnancy was defined as the ultrasonographic demonstration of an intrauterine gestational sac 4 weeks after embryo transfer. The data analysis was performed with SPSS version 18.0 (IBM). The analysis focused on the ovaries. The total number of developing follicles with a mean diameter R11 mm per ovary was the primary outcome. Data are reported as mean standard deviation (SD), median (interquartile range IQR), or number and percentage, as appropriate. The association between AFC and the total number of developing follicles was initially tested with the Pearson R2 coefficient of correlation. A linear regression model was then used to estimate the B coefficient of regression (95% confidence interval) and to adjust for confounders. We included in the multivariate model age, body mass index, and the mean daily dose of gonadotropin administered. The capacity of AFC to predict low (%2 follicles) or hyperresponsiveness (R7 follicles) was evaluated using receiver operating characteristic (ROC) curves, and the data are presented as the area under the curve (AUC) and 95% confidence interval (CI). P%.05 was considered statistically significant. The required sample size was calculated on the AUC for a low response. Based on preliminary data from our center in unselected women, the expected AUC (SD) in the control group of unaffected ovaries was 0.80 (0.15). Setting type I and II errors at 0.05 and 0.20 and stating as clinically relevant a reduction of the AUC of 0.10 in the three affected groups, we estimated a need for about 36 ovaries per study group. Based on the characteristics of our population, we estimated that this number could be reached with at least 80 women.
RESULTS Eighty-three women fulfilled our selection criteria. Twentyeight (34%) had no cysts at the time of recruitment. Fourteen women (17%) were diagnosed bilateral endometriomas, and 22 (26%) and 19 (23%) women had a diagnosis of unilateral right and left endometriomas, respectively. Eleven women were diagnosed more than one endometrioma per ovary. The mean SD diameter of the cysts was 20 10 mm. The baseline characteristics of these patients are shown in 1546
TABLE 1 Baseline clinical characteristics of the studied women (n [ 83). Characteristic
Data 35.4 3.5 2.9 1.4 8 (10) 21.2 3.1 7.8 3.3 1.7 1.7
Age (y) Duration of infertility (y) Previous pregnancies BMI (kg/m2) Day-3 serum FSH (IU/L) AMH (ng/mL) No. of interventions for endometriosis None 1 R2 rASRM classificationa III IV CA-125 (IU/mL) Concomitant male factor
25 (30) 49 (59) 9 (11) 27 (47) 31 (53) 41 28 40 (48)
Note: Date is presented as mean SD or number (%). AMH ¼ antim€ ullerian hormone; BMI ¼ body mass index; FSH ¼ follicle-stimulating hormone; rASRM ¼ revised American Society of Reproductive Medicine guidelines. a Data refer to operated patients. Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
Table 1. The characteristics of the IVF cycles are summarized in Supplemental Table 1 (available online). Seven cycles (8%) were cancelled, all of them for poor response. The studied ovaries (n ¼ 166) were divided into four groups (see Supplemental Fig. 1): [1] nonoperated gonads without endometriomas (controls) (n ¼ 42), [2] nonoperated gonads with endometriomas (n ¼ 46), [3] previously operated gonads without endometriomas (n ¼ 55), and [4] previously operated gonads with endometriomas (n ¼ 23). The main characteristics of ovarian stimulation for the four groups are summarized in Supplemental Table 2 (available online). The crude correlations between AFC and number of developing follicles in the four groups are shown in Figure 1. Statistically significant correlations emerged in all analyses. R2 values varied between 0.19 and 0.52. The results from the univariate and multivariate regression analyses are summarized in Table 2. Overall, all tested correlations were statistically significant with the exception of the adjusted correlation in the group of previously operated gonads without recurrences (B coefficient ¼ 0.34, P¼ .07). The results from ROC curves for the prediction of poor response (%2 follicles) and hyperresponse (R7 follicles) are presented in Table 3. The data were insufficient to draw a reliable estimate for the prediction of hyperresponsiveness in the group of previously operated gonads with recurrences. Statistically significant AUCs emerged for all groups for the prediction of a poor response and for all but one for the prediction of hyperresponsiveness (the exception being the group of previously operated gonads without recurrences). All the analyses were repeated by grouping all ovaries with endometriomas (groups 2 and 3, n ¼ 69), all ovaries with endometriomas with a mean diameter >20 mm (n ¼ 29, 22 from group 2 and 7 from group 3), and all operated ovaries (groups 3 and 4, n ¼ 78). Correlations between AFC and the number of developing follicles were statistically VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility®
FIGURE 1
Correlations between antral follicle count (AFC) and the total number of follicles at the time of human chorionic gonadotropin (hCG) administration. Data are presented separately for (A) nonoperated gonads without endometriomas (n ¼ 42), (B) nonoperated gonads with endometriomas (n ¼ 46), (C) previously operated gonads without endometriomas (n ¼ 55), and (D) previously operated gonads with endometriomas (n ¼ 23). The Pearson R2 coefficients of correlation were 0.31, 0.52, 0.19 and 0.48, respectively. All were statistically significant (P%.001). The curve lines represent the 95% confidence interval of the estimated linear regression (straight line). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
significant in all three groups (P< .001 for all). The corresponding R2 values were 0.52, 0.28, and 0.60, respectively. The results from the univariate and multivariate regression analyses and ROC curves for the prediction of poor response
or hyperresponsiveness are shown in Tables 2 and 3. All results were statistically significant. We also compared the ratio between developed follicles and AFC among the studied groups. However, we failed to
TABLE 2 Correlation between antral follicle count and ovarian responsiveness according to ovarian condition. Crude analyses Study group Nonoperated gonads Without endometriomas With endometriomas Previously operated gonads Without endometriomas With endometriomas Gonads with endometriomas (operated or nonoperated) Total >20 mm Operated gonads (with or without endometriomas)
Adjusted analyses
No. of ovaries
AFC
No. of follicles
B coefficient (95% CI)
P value
42 46
6.0 3.4 4.7 3.9
5.9 4.5 5.1 4.6
0.74 (0.39, 1.09) 0.86 (0.60, 1.11)
< .001 < .001
0.55 (0.07, 1.03) 0.77 (0.47, 1.07)
55 23
2.4 2.5 3.4 2.9
2.7 3.1 3.7 3.1
0.55 (0.24, 0.86) 0.74 (0.39, 1.08)
.001 < .001
0.34 (0.03, 0.71) 0.78 (0.37, 1.19)
69 29 78
4.3 3.6 3.9 4.3 2.7 2.6
4.6 4.2 4.1 4.9 3.0 3.1
0.84 (0.64, 1.03) 0.88 (0.60, 1.17) 0.63 (0.40, 0.86)
< .001 < .001 < .001
0.76 (0.54, 0.98) 0.79 (0.47, 1.12) 0.51 (0.26, 0.76)
B coefficient (95% CI)
P value .025 < .001 .07 .001 < .001 < .001 < .001
Note: The B coefficients were calculated using regression analyses. Adjustments were performed using a model that included age, body mass index, and daily dose of gonadotropins. AFC ¼ antral follicle count; CI ¼ confidence interval. Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
VOL. 103 NO. 6 / JUNE 2015
1547
ORIGINAL ARTICLE: ENDOMETRIOSIS
TABLE 3 Receiver operating characteristic curve analyses on the predictivity of AFC according to ovarian condition.
Study group Nonoperated gonads Without endometriomas With endometriomas Previously operated gonads Without endometriomas With endometriomas Gonads with endometriomas (operated or nonoperated) Total >20 mm Operated gonads (with or without endometriomas)
Poor response (£2 follicles)
Hyperresponse (‡7 follicles)
No. of ovaries
No.
AUC (95% CI)
P value
No.
AUC (95% CI)
P value
42 46
10 15
0.83 (0.68–0.99) 0.83 (0.71–0.95)
.002 < .001
16 15
0.84 (0.70–0.97) 0.73 (0.58–0.88)
< .001 .011
55 23
33 10
0.71 (0.57–0.85) 0.81 (0.63–0.99)
.009 .013
8 4
0.69 (0.47–0.91) NA
.10 NA
69 29 78
25 13 43
0.83 (0.73–0.93) 0.82 (0.66–0.98) 0.74 (0.63–0.85)
< .001 .004 < .001
19 6 12
0.80 (0.68–0.91) 0.79 (0.56–1.00) 0.77 (0.60–0.94)
< .001 .033 .003
Note: AFC ¼ antral follicle count; AUC ¼ area under the curve; CI ¼ confidence interval; NA ¼ not applicable (the low number of cases with hyperresponse hampered a reliable evaluation of the AUC). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
document statistically significant differences (data not shown).
DISCUSSION Based on our findings, AFC is a reliable means by which to estimate ovarian reserve in gonads with endometriomas or with a past history of surgical excision of endometriomas. In particular, the presence of an endometrioma does not affect the diagnostic performance of AFC. We found that B coefficients and the AUC were indeed very similar in gonads with endometriomas and in unaffected ovaries. On the other hand, AFC appears less accurate in previously operated gonads (albeit still significantly associated with ovarian responsiveness), in particular in the absence of a recurrence. The study was not sufficiently powered for definite intergroup comparisons, so further evidence on this point is required. However, if confirmed, the observation of a lower accuracy of AFC in operated gonads would be intriguing. It may be speculated that the surgery-related damage rather than surgery per se may reduce the predictive power of AFC. Indeed, the performance appears to be lower in operated gonads without recurrences compared with operated gonads with recurrences, and there is evidence that endometrioma recurrence is more likely in less injured gonads. In operated gonads, ovarian responsiveness to hyperstimulation is actually higher in those with a recurrent cysts than in those without recurrences (34); this may be a consequence of a higher chance of ovulation (and thus endometrioma formation) in less injured ovaries (35, 36). Combining this information in a precise interpretative hypothesis is, however, difficult and beyond the scope of the present study. Our conclusion on the reliability of AFC in affected and previously operated gonads is also supported by some external findings. Indeed, even if to the best of our knowledge this is the first study validating AFC in ovaries with current or past endometriomas, some indirect evidence in the literature already supported our conclusion. In particular, several 1548
studies designed to investigate the damage consequent to the surgical excision of ovarian endometriomas showed similar modifications in AFC and serum AMH (10, 11, 15, 16, 25–28). However, it has to be recognized that some unexplained inconsistencies emerged in a minority of these studies (10, 25, 29) and in a recent meta-analysis (29). Estimating the ovarian reserve of an ovary with an endometrioma or of a previously operated gonad may be clinical useful. For instance, in women of child-bearing age, the finding of a low AFC in an ovary with an endometrioma may favor surgery (the ovarian reserve being already compromised), whereas a high AFC may favor expectant management or egg banking before surgery. In this regard, it has to be pointed out that some investigators actually suggest ovarian hyperstimulation and oocyte freezing before surgery in selected cases (37, 38), and it has recently been shown that the quality of the oocytes obtained from ovaries with endometriomas is not affected (4, 6). Of further relevance here is the possibility of drawing on the overall judgment of the woman and the potential effects of surgery, considering both ovaries together. Again, as an example, the accurate evaluation of AFC may be of utmost relevance in a woman previously operated on one ovary who has a recurrence requiring surgery on the contralateral side. If the AFC of the operated ovary is adequate, the woman can face surgery without the risk of a compromised ovarian reserve after the surgery because at least one ovary would have an appropriate ovarian reserve. Conversely, presurgical storage of oocytes may be considered if the AFC of the contralateral gonad is low because if the ovary to be operated on is damaged, the overall ovarian reserve of the woman could be impaired after surgery. Some limitations of our study should be recognized. First, we used the number of developing follicles as a marker of ovarian reserve, and this outcome may be questioned. Data from ovarian biopsies would have been more appropriate but were not available with our study design. Ovarian responsiveness to hyperstimulation is actually the most reliable VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility® surrogate means of measuring ovarian reserve (31). Of note, we deemed it more appropriate to refer to the number of follicles rather than the number of oocytes retrieved because this outcome is not influenced by technical difficulties with oocyte retrieval, a difficulty that is expected to be more frequent in women with endometriosis. Moreover, cancelled cycles would have to be excluded, which may introduce a bias. Finally, AFC is intended to reflect the amount of ovarian reserve, not the quality of the folliculogenesis (and thus the oocytes retrieved). The population studied may be another point of concern. In particular, the dimension of the endometriomas was generally small (the mean diameter was 20 mm). In this regard, it has to be pointed out that the subgroup analysis focusing on larger cysts (including only the 29 gonads with a mean diameter >20 mm) showed extremely similar results. A third limitation would be related to the retrospective study design, but we do not consider this a limitation of considerable relevance. Our unit systematically assesses AFC before to an IVF cycle because this information plays a critical role in our therapeutic decisions. All physicians in our unit have long-term experience with ultrasound scanning, and they all assess AFC based on the currently available recommendations (30). Moreover, women who did not receive an AFC assessment in their early follicular phase were excluded from the study. In fact, we consider the use of data recorded outside an experimental context to be an advantage of our study because it better reflects normal clinical practice. Moreover, the baseline characteristics among studied patients may differ. We failed to observe statistically significant differences in the ovarian stimulation variables (see Supplemental Table 2), but we cannot definitely rule out subtle differences among the groups, a limitation that would be difficult to overcome even in prospective studies. In any case, we performed regression analyses that included age and gonadotropin doses in our model with the specific aim of protecting our findings from confounding variables. Finally, it must be recognized that our sample size was insufficient to draw definite conclusions about the relative accuracy of AFC among the different study groups. In conclusion, AFC is a reliable means by which to assess ovarian reserve in ovaries with endometriomas or with a history of surgery for endometriomas. In these situations, AFC can be used to collect important information for guiding physicians in their therapeutic decisions.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
REFERENCES 1.
2.
3.
Kitajima M, Defr ere S, Dolmans MM, Colette S, Squifflet J, Van Langendonckt A, et al. Endometriomas as a possible cause of reduced ovarian reserve in women with endometriosis. Fertil Steril 2011;96: 685–91. P, et al. Sanchez AM, Papaleo E, Corti L, Santambrogio P, Levi S, Vigano Iron availability is increased in individual human ovarian follicles in close proximity to an endometrioma compared with distal ones. Hum Reprod 2014;29:577–83. Kitajima M, Dolmans MM, Donnez O, Masuzaki H, Soares M, Donnez J. Enhanced follicular recruitment and atresia in cortex derived from ovaries with endometriomas. Fertil Steril 2014;101:1031–7.
VOL. 103 NO. 6 / JUNE 2015
20. 21.
22. 23.
Benaglia L, Bermejo A, Somigliana E, Faulisi S, Ragni G, Fedele L, et al. In vitro fertilization outcome in women with unoperated bilateral endometriomas. Fertil Steril 2013;99:1714–9. Somigliana E, Marchese MA, Frattaruolo MP, Berlanda N, Fedele L, Vercellini P. Serum anti-mullerian hormone in reproductive aged women with benign ovarian cysts. Eur J Obstet Gynecol Reprod Biol 2014;180: 142–7. Filippi F, Benaglia L, Paffoni A, Restelli L, Vercellini P, Somigliana E, et al. Ovarian endometriomas and oocyte quality: insights from in vitro fertilization cycles. Fertil Steril 2014;101:988–93. Somigliana E, Benaglia L, Vigano' P, Candiani M, Vercellini P, Fedele L. Surgical measures for endometriosis-related infertility: a plea for research. Placenta 2011;32(Suppl 3):S238–42. Raffi F, Metwally M, Amer S. The impact of excision of ovarian endometrioma on ovarian reserve: a systematic review and meta-analysis. J Clin Endocrinol Metab 2012;97:3146–54. P, Vercellini P, Fedele L. SurgiSomigliana E, Berlanda N, Benaglia L, Vigano cal excision of endometriomas and ovarian reserve: a systematic review on serum antim€ ullerian hormone level modifications. Fertil Steril 2012;98: 1531–8. Tsolakidis D, Pados G, Vavilis D, Athanatos D, Tsalikis T, Giannakou A, et al. The impact on ovarian reserve after laparoscopic ovarian cystectomy versus three-stage management in patients with endometriomas: a prospective randomized study. Fertil Steril 2010;94:71–7. Donnez J, Lousse JC, Jadoul P, Donnez O, Squifflet J. Laparoscopic management of endometriomas using a combined technique of excisional (cystectomy) and ablative surgery. Fertil Steril 2010;94:28–32. Var T, Batioglu S, Tonguc E, Kahyaoglu I. The effect of laparoscopic ovarian cystectomy versus coagulation in bilateral endometriomas on ovarian reserve as determined by antral follicle count and ovarian volume: a prospective randomized study. Fertil Steril 2011;95:2247–50. Coric M, Barisic D, Pavicic D, Karadza M, Banovic M. Electrocoagulation versus suture after laparoscopic stripping of ovarian endometriomas assessed by antral follicle count: preliminary results of randomized clinical trial. Arch Gynecol Obstet 2011;283:373–8. Ferrero S, Venturini PL, Gillott DJ, Remorgida V, Leone Roberti Maggiore U. Hemostasis by bipolar coagulation versus suture after surgical stripping of bilateral ovarian endometriomas: a randomized controlled trial. J Minim Invasive Gynecol 2012;19:722–30. Yu HT, Huang HY, Soong YK, Lee CL, Chao A, Wang CJ. Laparoscopic ovarian cystectomy of endometriomas: surgeons' experience may affect ovarian reserve and live-born rate in infertile patients with in vitro fertilization-intracytoplasmic sperm injection. Eur J Obstet Gynecol Reprod Biol 2010;152:172–5. Zaitoun MM, Zaitoun MM, El Behery MM. Comparing long term impact on ovarian reserve between laparoscopic ovarian cystectomy and open laparotomy for ovarian endometrioma. J Ovarian Res 2013;6:76. Garcia-Velasco JA, Somigliana E. Management of endometriomas in women requiring IVF: to touch or not to touch. Hum Reprod 2009;24: 496–501. P, Abbiati A, Barbara G, Crosignani PG. Vercellini P, Somigliana E, Vigano Surgery for endometriosis-associated infertility: a pragmatic approach. Hum Reprod 2009;24:254–69. P, Fedele L. The Vercellini P, Crosignani PG, Abbiati A, Somigliana E, Vigano effect of surgery for symptomatic endometriosis: the other side of the story. Hum Reprod Update 2009;15:177–8. P, Somigliana E, Fedele L. Endometriosis: pathogenesis Vercellini P, Vigano and treatment. Nat Rev Endocrinol 2014;10:261–75. Hansen KR, Hodnett GM, Knowlton N, Craig LB. Correlation of ovarian reserve tests with histologically determined primordial follicle number. Fertil Steril 2011;95:170–5. Nelson SM. Biomarkers of ovarian response: current and future applications. Fertil Steril 2013;99:963–9. Broer SL, van Disseldorp J, Broeze KA, Dolleman M, Opmeer BC, Bossuyt P, et al. Added value of ovarian reserve testing on patient characteristics in the prediction of ovarian response and ongoing pregnancy: an individual patient data approach. Hum Reprod Update 2013;19:26–36.
1549
ORIGINAL ARTICLE: ENDOMETRIOSIS 24.
25.
26.
27.
28.
29.
30.
La Marca A, Sunkara SK. Individualization of controlled ovarian stimulation in IVF using ovarian reserve markers: from theory to practice. Hum Reprod Update 2014;20:124–40. Biacchiardi CP, Piane LD, Camanni M, Deltetto F, Delpiano EM, Marchino GL, et al. Laparoscopic stripping of endometriomas negatively affects ovarian follicular reserve even if performed by experienced surgeons. Reprod Biomed Online 2011;23:740–6. Ercan CM, Duru NK, Karasahin KE, Coksuer H, Dede M, Baser I. Ultrasonographic evaluation and anti-mullerian hormone levels after laparoscopic stripping of unilateral endometriomas. Eur J Obstet Gynecol Reprod Biol 2011;158:280–4. Urman B, Alper E, Yakin K, Oktem O, Aksoy S, Alatas C, et al. Removal of unilateral endometriomas is associated with immediate and sustained reduction in ovarian reserve. Reprod Biomed Online 2013;27:212–6. Jang WK, Lim SY, Park JC, Lee KR, Lee A, Rhee JH. Surgical impact on serum anti-M€ ullerian hormone in women with benign ovarian cyst: A prospective study. Obstet Gynecol Sci 2014;57:121–7. Muzii L, Di Tucci C, Di Feliciantonio M, Marchetti C, Perniola G, Panici PB. The effect of surgery for endometrioma on ovarian reserve evaluated by antral follicle count: a systematic review and meta-analysis. Hum Reprod 2014;29:2190–8. Broekmans FJ, de Ziegler D, Howles CM, Gougeon A, Trew G, Olivennes F. The antral follicle count: practical recommendations for better standardization. Fertil Steril 2010;94:1044–51.
1550
31.
32.
33. 34.
35.
36.
37.
38.
Broekmans FJ, Kwee J, Hendriks DJ, Mol BW, Lambalk CB. A systematic review of tests predicting ovarian reserve and IVF outcome. Hum Reprod Update 2006;12:685–718. Savelli L. Transvaginal sonography for the assessment of ovarian and pelvic endometriosis: how deep is our understanding? Ultrasound Obstet Gynecol 2009;33:497–501. Busnelli A, Somigliana E, Benaglia L, Leonardi M, Ragni G, Fedele L. In vitro fertilization outcomes in treated hypothyroidism. Thyroid 2013;23:1319–25. Somigliana E, Benaglia L, Vercellini P, Paffoni A, Ragni G, Fedele L. Recurrent endometrioma and ovarian reserve: biological connection or surgical paradox? Am J Obstet Gynecol 2011;204:529.e1–5. Vercellini P, Somigliana E, Vigano P, Abbiati A, Barbara G, Fedele L. ‘‘Blood on the tracks’’ from corpora lutea to endometriomas. BJOG 2009;116:366–71. P, De Matteis S, Barbara G, Fedele L. PostVercellini P, Somigliana E, Vigano operative endometriosis recurrence: a plea for prevention based on pathogenetic, epidemiological and clinical evidence. Reprod Biomed Online 2010;21:259–65. Elizur SE, Chian RC, Holzer HE, Gidoni Y, Tulandi T, Tan SL. Cryopreservation of oocytes in a young woman with severe and symptomatic endometriosis: a new indication for fertility preservation. Fertil Steril 2009;91:293.e1–3. Garcia-Velasco JA, Domingo J, Cobo A, Martínez M, Carmona L, Pellicer A. Five years' experience using oocyte vitrification to preserve fertility for medical and nonmedical indications. Fertil Steril 2013;99:1994–9.
VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility®
SUPPLEMENTAL FIGURE 1
Diagram of the study population. The analyses focused on the ovaries. The arrows show the contribution of the different categories of women (based on history of surgery for endometrioma and presence of endometrioma) to the four study groups: unoperated gonads without endometriomas, unoperated gonads with endometriomas, operated gonads without endometriomas, operated gonads with endometriomas. (N ¼ number of women; n ¼ number of gonads.) *Four women had homolateral recurrences and nine had contralateral recurrences. Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
VOL. 103 NO. 6 / JUNE 2015
1550.e1
ORIGINAL ARTICLE: ENDOMETRIOSIS
SUPPLEMENTAL TABLE 1 IVF cycle characteristics in recruited women (n [ 83). Characteristic
Mean ± SD, median (IQR), or n (%)
Stimulation protocol Long protocol GnRH antagonist Short protocol Cancelled cycle Total dose of administered FSH (IU) Duration of stimulation (d) No. of oocyte retrieveda No. of oocytes usedb Technique usedb IVF ICSI Fertilization rateb No. of embryos transferredc 1 2 3 Clinical pregnancy rate Implantation rate
37 (45%) 29 (35%) 17 (20%) 7 (8%) 3,118 1,492 9.7 3.0 7.1 5.6 4.7 2.9 19 (28%) 49 (72%) 75% (50–100%) 24 (43%) 30 (54%) 2 (3%) 21 (25%) 25 (28%)
Note: FSH ¼ follicle-stimulating hormone; GnRH ¼ gonadotropin-releasing hormone; ICSI ¼ intracytoplasmic sperm injection; IQR ¼ interquartile range; SD ¼ standard deviation. a Data refers to patients underwent oocyte retrieval (n ¼ 76). b Data refers to patients who retrieved R1 suitable oocyte (n ¼ 68). c Data refers to patients who obtained R1 viable embryo (n ¼ 56). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
1550.e2
VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility®
SUPPLEMENTAL TABLE 2 Main stimulation characteristics in the four study groups. Nonoperated gonads Characteristic No. of gonads Stimulation protocol Long protocol GnRH antagonist Short protocol Cancelled cycle Total dose of administered FSH (IU) Duration of stimulation (d) Total no. of follicles
Previously operated gonads
Without endometriomas
With endometriomas
Without endometriomas
With endometriomas
42
46
55
23
22 (52%) 12 (29%) 8 (19%) 1 (2%) 3,059 1,547
25 (54%) 14 (31%) 7 (15%) 5 (11%) 2,905 1,502
21 (38%) 20 (36%) 14 (26%) 8 (14%) 3,327 1,372
6 (26%) 12 (52%) 5 (22%) 0 (0%) 3,171 1,637
.07 .59
10.3 2.6 5.9 4.5
9.7 2.6 5.1 4.6
9.3 3.7 2.6 3.1
9.8 2.4 3.6 3.1
.47 < .001a
P value – .24
Note: FSH ¼ follicle-stimulating hormone; GnRH ¼ gonadotropin-releasing hormone. a At post hoc analysis, a statistically significant difference emerged for the comparison between previously operated gonads without recurrences and nonoperated gonads without endometriomas (P< .001) and between previously operated gonads without recurrences and nonoperated gonads with endometriomas (P¼ .01). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
VOL. 103 NO. 6 / JUNE 2015
1550.e3
Objective: To evaluate the accuracy of antral follicular count (AFC) in predicting ovarian responsiveness in ovaries with endometriomas or with a past history of surgical excision of endometriomas. Design: Retrospective review. Setting: Academic hospital. Patient(s): Eighty-three women for a total of 166 gonads. Intervention(s): None. Main Outcome Measure(s): Total number of developing follicles. Result(s): The ovaries were characterized as four groups: [1] unoperated gonads without endometriomas (n ¼ 42, control group), [2] unoperated gonads with endometriomas (n ¼ 46), [3] operated gonads without endometriomas (n ¼ 55), and [4] operated gonads with endometriomas (n ¼ 23). The analyses subsequently considered all ovaries with endometriomas (groups 2 þ 4, n ¼ 69) and all operated ovaries (groups 3 þ 4, n ¼ 78). The capacity of AFC to predict low response (%2 follicles) or hyperresponsiveness (R7 follicles) was evaluated using receiver operating characteristic curves. We used a linear regression model to calculate the adjusted B coefficients. The adjusted B coefficients in unaffected ovaries, in all ovaries with endometriomas, and in all operated ovaries were 0.55 (95% confidence interval [CI], 0.07–1.03), 0.76 (95% CI, 0.54–0.98), and 0.51 (95% CI, 0.26–0.76), respectively. The area under the curve (AUC) for the prediction of low response was 0.83 (95% CI, 0.68–0.99), 0.83 (95% CI, 0.73–0.93), and 0.74 (95% CI, 0.63–0.85), respectively. The AUC for the prediction of hyperresponse was 0.84 (95% CI, 0.70–0.97), 0.74 (95% CI, 0.63–0.85), and 0.77 (0.60–0.94), respectively. Conclusion(s): The accuracy of AFC for predicting ovarian response is similar in unaffected Use your smartphone ovaries, ovaries with endometriomas and ovaries with a history of surgery for endometriomas. to scan this QR code (Fertil SterilÒ 2015;103:1544–50. Ó2015 by American Society for Reproductive Medicine.) and connect to the Key Words: AFC, controlled ovarian hyperstimulation, endometrioma Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/benaglial-afc-ovarian-endometriomas/
T
he impact of ovarian endometriomas and their surgical removal on the ovarian reserve is still a debated issue. There is consistent biological evidence suggesting a detrimental effect of these cysts on the ovarian reserve (1–3). Potential
mechanisms of damage include the diffusion of toxic substances from the endometrioma to the ovarian tissue (2) and the burn-out effect of damage consequent to an enhanced recruitment of primordial follicles and consequent accelerated exhaustion of the ovarian
Received November 25, 2014; revised March 9, 2015; accepted March 10, 2015; published online April 29, 2015. L.B. has nothing to disclose. G.C. has nothing to disclose. A.B. has nothing to disclose. A.P. has nothing to disclose. P.V. has nothing to disclose. E.S. has nothing to disclose. Reprint requests: Laura Benaglia, M.D., Infertility Unit, Fondazione Ca' Granda, Ospedale Maggiore Policlinico, Via M. Fanti 6, 20122 Milan, Italy (E-mail: [email protected]). Fertility and Sterility® Vol. 103, No. 6, June 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.03.013 1544
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
reserve (3). This latter mechanism is strongly supported by two recent contributions that documented a lower number of primordial follicles and an increased proportion of activated and atretic follicles in ovaries with endometriomas (1, 3). In contrast, available clinical data have failed to document a relevant impact. In women with unilateral endometriomas, ovarian responsiveness to hyperstimulation does not differ between the two gonads, and ovarian reserve appears to be only modestly affected in women with bilateral cysts (4–6). VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility® Laparoscopic stripping has been shown to damage the ovarian reserve by most but not all investigators (7–9). Moreover, recent preliminary evidence suggests that alternative surgical techniques such as laser vaporization, partial cystectomy, or surgical suture may cause less or no damage (10–16). Not surprisingly, this controversial scenario has had a clinical impact, and physicians frequently to have deal with a high degree of uncertainty. The decision to operate depends on several clinical factors and sonographic findings (17–20). A neglected but potentially relevant issue is the measurement of the ovarian reserve in ovaries with endometriomas or with a history of surgery for endometriomas. A clear knowledge of the remnant ovarian reserve of an affected or previously operated ovary may influence therapeutic decisions. Unfortunately, the evaluation of ovarian reserve in women with current or past ovarian endometriomas is challenging. In most cases, these cysts are unilateral, thus hampering the validity of hormone assessments such as serum follicle-stimulating hormone (FSH) or inhibin because the contralateral gonad may properly compensate for the reduced function of the affected one. The validity of antim€ ullerian hormone (AMH) is also debatable because the relative contribution of the affected and intact ovaries cannot be definitely discriminated. The accuracy of serum AMH is also questionable in women with bilateral cysts because it cannot be excluded that endometriomas may affect ovarian reserve differently in the same patient. At present, sonographic assessment of antral follicle count (AFC) is the only means of obtaining independent data on the ovarian reserve of an ovary. A huge amount of literature has validated the use of AFC as a surrogate measurement of ovarian reserve (21–24). However, to the best of our knowledge, even though AFC in ovaries with current or past endometriomas has been used repeatedly in research studies (10–13, 15, 16, 25–29), this tool has never been validated. Of note, based on the currently accepted recommendations, AFC should not be assessed in women with ovarian endometriosis or previous ovarian surgery (30). Our study focused on women undergoing in vitro fertilization (IVF) to validate the use of AFC in ovaries with current or past endometriomas. Ovarian responsiveness to hyperstimulation is actually considered the best noninvasive surrogate measurement of ovarian reserve (31). The correlation of AFC and ovarian responsiveness in ovaries with current or past endometriomas may thus provide valuable information on the accuracy of AFC in this context.
MATERIALS AND METHODS Our retrospective study at the Infertility Unit of the Department of Obstetrics and Gynecology of the Fondazione Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy, reviewed the clinical charts of women who underwent IVF cycles between January 2011 and March 2013. The inclusion criteria were as follows: [1] previous surgery for ovarian endometriomas and/or sonographic diagnosis of endometriomas at the time of the cycle, [2] availability of a pre-IVF assessment of AFC between days 2 and 4 of a spontaneous menstrual cycle, VOL. 103 NO. 6 / JUNE 2015
and [3] age %42 years. The exclusion criteria included [1] past history of ovarian surgery for nonendometrial lesions, [2] diagnosis of nonendometrial cysts at the time of IVF cycle, and [3] pre-IVF hormone treatment (progestins, estroprogestins, or gonadotropin-releasing hormone analogs). The women could be enrolled for only one cycle. Women who are referred to our unit are routinely asked to provide informed consent for their clinical data to be used for research purposes, and those who had refused consent were excluded from our study. The local institutional review board approved the study protocol. Our analyses focused on the relationship between AFC and the total number of developing follicles per ovary. Specifically, four different groups of gonads were initially considered: [1] unoperated ovaries without endometriomas (control group), [2] unoperated ovaries with endometriomas, [3] operated ovaries without endometriomas, and [4] operated ovaries with endometriomas. The analyses subsequently considered all ovaries with endometriomas together (groups 2 þ 4), all ovaries with endometriomas with a mean diameter >20 mm (subgroup of the latter), and all operated ovaries (groups 3 þ 4). During the study period, AFC was assessed based on the current available recommendations (30). Briefly, all identifiable antral follicles 2–10 mm in diameter were recorded. The diagnosis of endometrial cysts was performed by transvaginal ultrasound and had to be documented on at least two occasions and at least two menstrual cycles apart. More specifically, ovarian endometrioma was defined as a roundshaped cystic mass with a minimum diameter of 10 mm, with thick walls, regular margins, homogeneous low echogenic fluid content with scattered internal echoes, and without papillary proliferations (32). Women with atypical lesions were excluded, such as cysts whose sonographic appearance was compatible with but not distinctly identifiable as endometriosis. The diameter of the endometriomas was calculated as the mean of three perpendicular diameters. All sonographic evaluations were performed by four physicians who have had a long-term specialization in reproductive medicine, all of whom attended an internal meeting in 2010 to discuss the new recommended modality to assess AFC. In our unit, women with a history of surgery for ovarian endometriosis are routinely asked to provide documentation of the intervention; only those with a histologically confirmed diagnosis of endometriomas were included in the study. During the IVF cycle, the patients selected for IVF were monitored and managed according to standardized clinical protocols, as reported elsewhere (33). Briefly, the patients underwent a transvaginal ultrasound in the early proliferative phase of the cycle preceding the initiation of ovarian hyperstimulation. The AFC assessment was systematically recorded at this time and again on day 3 of the following menstrual cycle if the woman received a flare-up protocol or a protocol with gonadotropin-releasing hormone antagonists (if two AFC assessments were available, the latter one was used for the analysis). The regimen used and the dose of gonadotropins were determined on an individual basis according to the age, day-3 serum FSH value, serum AMH value, and AFC. During 1545
ORIGINAL ARTICLE: ENDOMETRIOSIS the stimulation, women underwent serial transvaginal ultrasounds and serum hormonal assessments when indicated. When leading follicles with a mean diameter >18 mm were visualized, human chorionic gonadotropin (hCG) was administered subcutaneously. All follicles with a mean diameter R11 mm were routinely recorded at each sonographic assessment and at the time of hCG administration to trigger ovulation. Cycles could be cancelled because of low response or hyperresponsiveness after a consultation with the patient. Cancellation for a poor response was considered when the leading follicle had a diameter >18 mm and the total number of follicles was 20. For cancelled cycles, our analysis used the number of follicles with a mean diameter R11 mm at the time of cancellation. Oocyte retrieval was performed transvaginally 36 hours after the hCG injection. Embryo transfer was performed 48 to 72 hours after oocyte collection or, in properly selected patients, at the blastocyst stage. Clinical pregnancy was defined as the ultrasonographic demonstration of an intrauterine gestational sac 4 weeks after embryo transfer. The data analysis was performed with SPSS version 18.0 (IBM). The analysis focused on the ovaries. The total number of developing follicles with a mean diameter R11 mm per ovary was the primary outcome. Data are reported as mean standard deviation (SD), median (interquartile range IQR), or number and percentage, as appropriate. The association between AFC and the total number of developing follicles was initially tested with the Pearson R2 coefficient of correlation. A linear regression model was then used to estimate the B coefficient of regression (95% confidence interval) and to adjust for confounders. We included in the multivariate model age, body mass index, and the mean daily dose of gonadotropin administered. The capacity of AFC to predict low (%2 follicles) or hyperresponsiveness (R7 follicles) was evaluated using receiver operating characteristic (ROC) curves, and the data are presented as the area under the curve (AUC) and 95% confidence interval (CI). P%.05 was considered statistically significant. The required sample size was calculated on the AUC for a low response. Based on preliminary data from our center in unselected women, the expected AUC (SD) in the control group of unaffected ovaries was 0.80 (0.15). Setting type I and II errors at 0.05 and 0.20 and stating as clinically relevant a reduction of the AUC of 0.10 in the three affected groups, we estimated a need for about 36 ovaries per study group. Based on the characteristics of our population, we estimated that this number could be reached with at least 80 women.
RESULTS Eighty-three women fulfilled our selection criteria. Twentyeight (34%) had no cysts at the time of recruitment. Fourteen women (17%) were diagnosed bilateral endometriomas, and 22 (26%) and 19 (23%) women had a diagnosis of unilateral right and left endometriomas, respectively. Eleven women were diagnosed more than one endometrioma per ovary. The mean SD diameter of the cysts was 20 10 mm. The baseline characteristics of these patients are shown in 1546
TABLE 1 Baseline clinical characteristics of the studied women (n [ 83). Characteristic
Data 35.4 3.5 2.9 1.4 8 (10) 21.2 3.1 7.8 3.3 1.7 1.7
Age (y) Duration of infertility (y) Previous pregnancies BMI (kg/m2) Day-3 serum FSH (IU/L) AMH (ng/mL) No. of interventions for endometriosis None 1 R2 rASRM classificationa III IV CA-125 (IU/mL) Concomitant male factor
25 (30) 49 (59) 9 (11) 27 (47) 31 (53) 41 28 40 (48)
Note: Date is presented as mean SD or number (%). AMH ¼ antim€ ullerian hormone; BMI ¼ body mass index; FSH ¼ follicle-stimulating hormone; rASRM ¼ revised American Society of Reproductive Medicine guidelines. a Data refer to operated patients. Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
Table 1. The characteristics of the IVF cycles are summarized in Supplemental Table 1 (available online). Seven cycles (8%) were cancelled, all of them for poor response. The studied ovaries (n ¼ 166) were divided into four groups (see Supplemental Fig. 1): [1] nonoperated gonads without endometriomas (controls) (n ¼ 42), [2] nonoperated gonads with endometriomas (n ¼ 46), [3] previously operated gonads without endometriomas (n ¼ 55), and [4] previously operated gonads with endometriomas (n ¼ 23). The main characteristics of ovarian stimulation for the four groups are summarized in Supplemental Table 2 (available online). The crude correlations between AFC and number of developing follicles in the four groups are shown in Figure 1. Statistically significant correlations emerged in all analyses. R2 values varied between 0.19 and 0.52. The results from the univariate and multivariate regression analyses are summarized in Table 2. Overall, all tested correlations were statistically significant with the exception of the adjusted correlation in the group of previously operated gonads without recurrences (B coefficient ¼ 0.34, P¼ .07). The results from ROC curves for the prediction of poor response (%2 follicles) and hyperresponse (R7 follicles) are presented in Table 3. The data were insufficient to draw a reliable estimate for the prediction of hyperresponsiveness in the group of previously operated gonads with recurrences. Statistically significant AUCs emerged for all groups for the prediction of a poor response and for all but one for the prediction of hyperresponsiveness (the exception being the group of previously operated gonads without recurrences). All the analyses were repeated by grouping all ovaries with endometriomas (groups 2 and 3, n ¼ 69), all ovaries with endometriomas with a mean diameter >20 mm (n ¼ 29, 22 from group 2 and 7 from group 3), and all operated ovaries (groups 3 and 4, n ¼ 78). Correlations between AFC and the number of developing follicles were statistically VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility®
FIGURE 1
Correlations between antral follicle count (AFC) and the total number of follicles at the time of human chorionic gonadotropin (hCG) administration. Data are presented separately for (A) nonoperated gonads without endometriomas (n ¼ 42), (B) nonoperated gonads with endometriomas (n ¼ 46), (C) previously operated gonads without endometriomas (n ¼ 55), and (D) previously operated gonads with endometriomas (n ¼ 23). The Pearson R2 coefficients of correlation were 0.31, 0.52, 0.19 and 0.48, respectively. All were statistically significant (P%.001). The curve lines represent the 95% confidence interval of the estimated linear regression (straight line). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
significant in all three groups (P< .001 for all). The corresponding R2 values were 0.52, 0.28, and 0.60, respectively. The results from the univariate and multivariate regression analyses and ROC curves for the prediction of poor response
or hyperresponsiveness are shown in Tables 2 and 3. All results were statistically significant. We also compared the ratio between developed follicles and AFC among the studied groups. However, we failed to
TABLE 2 Correlation between antral follicle count and ovarian responsiveness according to ovarian condition. Crude analyses Study group Nonoperated gonads Without endometriomas With endometriomas Previously operated gonads Without endometriomas With endometriomas Gonads with endometriomas (operated or nonoperated) Total >20 mm Operated gonads (with or without endometriomas)
Adjusted analyses
No. of ovaries
AFC
No. of follicles
B coefficient (95% CI)
P value
42 46
6.0 3.4 4.7 3.9
5.9 4.5 5.1 4.6
0.74 (0.39, 1.09) 0.86 (0.60, 1.11)
< .001 < .001
0.55 (0.07, 1.03) 0.77 (0.47, 1.07)
55 23
2.4 2.5 3.4 2.9
2.7 3.1 3.7 3.1
0.55 (0.24, 0.86) 0.74 (0.39, 1.08)
.001 < .001
0.34 (0.03, 0.71) 0.78 (0.37, 1.19)
69 29 78
4.3 3.6 3.9 4.3 2.7 2.6
4.6 4.2 4.1 4.9 3.0 3.1
0.84 (0.64, 1.03) 0.88 (0.60, 1.17) 0.63 (0.40, 0.86)
< .001 < .001 < .001
0.76 (0.54, 0.98) 0.79 (0.47, 1.12) 0.51 (0.26, 0.76)
B coefficient (95% CI)
P value .025 < .001 .07 .001 < .001 < .001 < .001
Note: The B coefficients were calculated using regression analyses. Adjustments were performed using a model that included age, body mass index, and daily dose of gonadotropins. AFC ¼ antral follicle count; CI ¼ confidence interval. Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
VOL. 103 NO. 6 / JUNE 2015
1547
ORIGINAL ARTICLE: ENDOMETRIOSIS
TABLE 3 Receiver operating characteristic curve analyses on the predictivity of AFC according to ovarian condition.
Study group Nonoperated gonads Without endometriomas With endometriomas Previously operated gonads Without endometriomas With endometriomas Gonads with endometriomas (operated or nonoperated) Total >20 mm Operated gonads (with or without endometriomas)
Poor response (£2 follicles)
Hyperresponse (‡7 follicles)
No. of ovaries
No.
AUC (95% CI)
P value
No.
AUC (95% CI)
P value
42 46
10 15
0.83 (0.68–0.99) 0.83 (0.71–0.95)
.002 < .001
16 15
0.84 (0.70–0.97) 0.73 (0.58–0.88)
< .001 .011
55 23
33 10
0.71 (0.57–0.85) 0.81 (0.63–0.99)
.009 .013
8 4
0.69 (0.47–0.91) NA
.10 NA
69 29 78
25 13 43
0.83 (0.73–0.93) 0.82 (0.66–0.98) 0.74 (0.63–0.85)
< .001 .004 < .001
19 6 12
0.80 (0.68–0.91) 0.79 (0.56–1.00) 0.77 (0.60–0.94)
< .001 .033 .003
Note: AFC ¼ antral follicle count; AUC ¼ area under the curve; CI ¼ confidence interval; NA ¼ not applicable (the low number of cases with hyperresponse hampered a reliable evaluation of the AUC). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
document statistically significant differences (data not shown).
DISCUSSION Based on our findings, AFC is a reliable means by which to estimate ovarian reserve in gonads with endometriomas or with a past history of surgical excision of endometriomas. In particular, the presence of an endometrioma does not affect the diagnostic performance of AFC. We found that B coefficients and the AUC were indeed very similar in gonads with endometriomas and in unaffected ovaries. On the other hand, AFC appears less accurate in previously operated gonads (albeit still significantly associated with ovarian responsiveness), in particular in the absence of a recurrence. The study was not sufficiently powered for definite intergroup comparisons, so further evidence on this point is required. However, if confirmed, the observation of a lower accuracy of AFC in operated gonads would be intriguing. It may be speculated that the surgery-related damage rather than surgery per se may reduce the predictive power of AFC. Indeed, the performance appears to be lower in operated gonads without recurrences compared with operated gonads with recurrences, and there is evidence that endometrioma recurrence is more likely in less injured gonads. In operated gonads, ovarian responsiveness to hyperstimulation is actually higher in those with a recurrent cysts than in those without recurrences (34); this may be a consequence of a higher chance of ovulation (and thus endometrioma formation) in less injured ovaries (35, 36). Combining this information in a precise interpretative hypothesis is, however, difficult and beyond the scope of the present study. Our conclusion on the reliability of AFC in affected and previously operated gonads is also supported by some external findings. Indeed, even if to the best of our knowledge this is the first study validating AFC in ovaries with current or past endometriomas, some indirect evidence in the literature already supported our conclusion. In particular, several 1548
studies designed to investigate the damage consequent to the surgical excision of ovarian endometriomas showed similar modifications in AFC and serum AMH (10, 11, 15, 16, 25–28). However, it has to be recognized that some unexplained inconsistencies emerged in a minority of these studies (10, 25, 29) and in a recent meta-analysis (29). Estimating the ovarian reserve of an ovary with an endometrioma or of a previously operated gonad may be clinical useful. For instance, in women of child-bearing age, the finding of a low AFC in an ovary with an endometrioma may favor surgery (the ovarian reserve being already compromised), whereas a high AFC may favor expectant management or egg banking before surgery. In this regard, it has to be pointed out that some investigators actually suggest ovarian hyperstimulation and oocyte freezing before surgery in selected cases (37, 38), and it has recently been shown that the quality of the oocytes obtained from ovaries with endometriomas is not affected (4, 6). Of further relevance here is the possibility of drawing on the overall judgment of the woman and the potential effects of surgery, considering both ovaries together. Again, as an example, the accurate evaluation of AFC may be of utmost relevance in a woman previously operated on one ovary who has a recurrence requiring surgery on the contralateral side. If the AFC of the operated ovary is adequate, the woman can face surgery without the risk of a compromised ovarian reserve after the surgery because at least one ovary would have an appropriate ovarian reserve. Conversely, presurgical storage of oocytes may be considered if the AFC of the contralateral gonad is low because if the ovary to be operated on is damaged, the overall ovarian reserve of the woman could be impaired after surgery. Some limitations of our study should be recognized. First, we used the number of developing follicles as a marker of ovarian reserve, and this outcome may be questioned. Data from ovarian biopsies would have been more appropriate but were not available with our study design. Ovarian responsiveness to hyperstimulation is actually the most reliable VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility® surrogate means of measuring ovarian reserve (31). Of note, we deemed it more appropriate to refer to the number of follicles rather than the number of oocytes retrieved because this outcome is not influenced by technical difficulties with oocyte retrieval, a difficulty that is expected to be more frequent in women with endometriosis. Moreover, cancelled cycles would have to be excluded, which may introduce a bias. Finally, AFC is intended to reflect the amount of ovarian reserve, not the quality of the folliculogenesis (and thus the oocytes retrieved). The population studied may be another point of concern. In particular, the dimension of the endometriomas was generally small (the mean diameter was 20 mm). In this regard, it has to be pointed out that the subgroup analysis focusing on larger cysts (including only the 29 gonads with a mean diameter >20 mm) showed extremely similar results. A third limitation would be related to the retrospective study design, but we do not consider this a limitation of considerable relevance. Our unit systematically assesses AFC before to an IVF cycle because this information plays a critical role in our therapeutic decisions. All physicians in our unit have long-term experience with ultrasound scanning, and they all assess AFC based on the currently available recommendations (30). Moreover, women who did not receive an AFC assessment in their early follicular phase were excluded from the study. In fact, we consider the use of data recorded outside an experimental context to be an advantage of our study because it better reflects normal clinical practice. Moreover, the baseline characteristics among studied patients may differ. We failed to observe statistically significant differences in the ovarian stimulation variables (see Supplemental Table 2), but we cannot definitely rule out subtle differences among the groups, a limitation that would be difficult to overcome even in prospective studies. In any case, we performed regression analyses that included age and gonadotropin doses in our model with the specific aim of protecting our findings from confounding variables. Finally, it must be recognized that our sample size was insufficient to draw definite conclusions about the relative accuracy of AFC among the different study groups. In conclusion, AFC is a reliable means by which to assess ovarian reserve in ovaries with endometriomas or with a history of surgery for endometriomas. In these situations, AFC can be used to collect important information for guiding physicians in their therapeutic decisions.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
REFERENCES 1.
2.
3.
Kitajima M, Defr ere S, Dolmans MM, Colette S, Squifflet J, Van Langendonckt A, et al. Endometriomas as a possible cause of reduced ovarian reserve in women with endometriosis. Fertil Steril 2011;96: 685–91. P, et al. Sanchez AM, Papaleo E, Corti L, Santambrogio P, Levi S, Vigano Iron availability is increased in individual human ovarian follicles in close proximity to an endometrioma compared with distal ones. Hum Reprod 2014;29:577–83. Kitajima M, Dolmans MM, Donnez O, Masuzaki H, Soares M, Donnez J. Enhanced follicular recruitment and atresia in cortex derived from ovaries with endometriomas. Fertil Steril 2014;101:1031–7.
VOL. 103 NO. 6 / JUNE 2015
20. 21.
22. 23.
Benaglia L, Bermejo A, Somigliana E, Faulisi S, Ragni G, Fedele L, et al. In vitro fertilization outcome in women with unoperated bilateral endometriomas. Fertil Steril 2013;99:1714–9. Somigliana E, Marchese MA, Frattaruolo MP, Berlanda N, Fedele L, Vercellini P. Serum anti-mullerian hormone in reproductive aged women with benign ovarian cysts. Eur J Obstet Gynecol Reprod Biol 2014;180: 142–7. Filippi F, Benaglia L, Paffoni A, Restelli L, Vercellini P, Somigliana E, et al. Ovarian endometriomas and oocyte quality: insights from in vitro fertilization cycles. Fertil Steril 2014;101:988–93. Somigliana E, Benaglia L, Vigano' P, Candiani M, Vercellini P, Fedele L. Surgical measures for endometriosis-related infertility: a plea for research. Placenta 2011;32(Suppl 3):S238–42. Raffi F, Metwally M, Amer S. The impact of excision of ovarian endometrioma on ovarian reserve: a systematic review and meta-analysis. J Clin Endocrinol Metab 2012;97:3146–54. P, Vercellini P, Fedele L. SurgiSomigliana E, Berlanda N, Benaglia L, Vigano cal excision of endometriomas and ovarian reserve: a systematic review on serum antim€ ullerian hormone level modifications. Fertil Steril 2012;98: 1531–8. Tsolakidis D, Pados G, Vavilis D, Athanatos D, Tsalikis T, Giannakou A, et al. The impact on ovarian reserve after laparoscopic ovarian cystectomy versus three-stage management in patients with endometriomas: a prospective randomized study. Fertil Steril 2010;94:71–7. Donnez J, Lousse JC, Jadoul P, Donnez O, Squifflet J. Laparoscopic management of endometriomas using a combined technique of excisional (cystectomy) and ablative surgery. Fertil Steril 2010;94:28–32. Var T, Batioglu S, Tonguc E, Kahyaoglu I. The effect of laparoscopic ovarian cystectomy versus coagulation in bilateral endometriomas on ovarian reserve as determined by antral follicle count and ovarian volume: a prospective randomized study. Fertil Steril 2011;95:2247–50. Coric M, Barisic D, Pavicic D, Karadza M, Banovic M. Electrocoagulation versus suture after laparoscopic stripping of ovarian endometriomas assessed by antral follicle count: preliminary results of randomized clinical trial. Arch Gynecol Obstet 2011;283:373–8. Ferrero S, Venturini PL, Gillott DJ, Remorgida V, Leone Roberti Maggiore U. Hemostasis by bipolar coagulation versus suture after surgical stripping of bilateral ovarian endometriomas: a randomized controlled trial. J Minim Invasive Gynecol 2012;19:722–30. Yu HT, Huang HY, Soong YK, Lee CL, Chao A, Wang CJ. Laparoscopic ovarian cystectomy of endometriomas: surgeons' experience may affect ovarian reserve and live-born rate in infertile patients with in vitro fertilization-intracytoplasmic sperm injection. Eur J Obstet Gynecol Reprod Biol 2010;152:172–5. Zaitoun MM, Zaitoun MM, El Behery MM. Comparing long term impact on ovarian reserve between laparoscopic ovarian cystectomy and open laparotomy for ovarian endometrioma. J Ovarian Res 2013;6:76. Garcia-Velasco JA, Somigliana E. Management of endometriomas in women requiring IVF: to touch or not to touch. Hum Reprod 2009;24: 496–501. P, Abbiati A, Barbara G, Crosignani PG. Vercellini P, Somigliana E, Vigano Surgery for endometriosis-associated infertility: a pragmatic approach. Hum Reprod 2009;24:254–69. P, Fedele L. The Vercellini P, Crosignani PG, Abbiati A, Somigliana E, Vigano effect of surgery for symptomatic endometriosis: the other side of the story. Hum Reprod Update 2009;15:177–8. P, Somigliana E, Fedele L. Endometriosis: pathogenesis Vercellini P, Vigano and treatment. Nat Rev Endocrinol 2014;10:261–75. Hansen KR, Hodnett GM, Knowlton N, Craig LB. Correlation of ovarian reserve tests with histologically determined primordial follicle number. Fertil Steril 2011;95:170–5. Nelson SM. Biomarkers of ovarian response: current and future applications. Fertil Steril 2013;99:963–9. Broer SL, van Disseldorp J, Broeze KA, Dolleman M, Opmeer BC, Bossuyt P, et al. Added value of ovarian reserve testing on patient characteristics in the prediction of ovarian response and ongoing pregnancy: an individual patient data approach. Hum Reprod Update 2013;19:26–36.
1549
ORIGINAL ARTICLE: ENDOMETRIOSIS 24.
25.
26.
27.
28.
29.
30.
La Marca A, Sunkara SK. Individualization of controlled ovarian stimulation in IVF using ovarian reserve markers: from theory to practice. Hum Reprod Update 2014;20:124–40. Biacchiardi CP, Piane LD, Camanni M, Deltetto F, Delpiano EM, Marchino GL, et al. Laparoscopic stripping of endometriomas negatively affects ovarian follicular reserve even if performed by experienced surgeons. Reprod Biomed Online 2011;23:740–6. Ercan CM, Duru NK, Karasahin KE, Coksuer H, Dede M, Baser I. Ultrasonographic evaluation and anti-mullerian hormone levels after laparoscopic stripping of unilateral endometriomas. Eur J Obstet Gynecol Reprod Biol 2011;158:280–4. Urman B, Alper E, Yakin K, Oktem O, Aksoy S, Alatas C, et al. Removal of unilateral endometriomas is associated with immediate and sustained reduction in ovarian reserve. Reprod Biomed Online 2013;27:212–6. Jang WK, Lim SY, Park JC, Lee KR, Lee A, Rhee JH. Surgical impact on serum anti-M€ ullerian hormone in women with benign ovarian cyst: A prospective study. Obstet Gynecol Sci 2014;57:121–7. Muzii L, Di Tucci C, Di Feliciantonio M, Marchetti C, Perniola G, Panici PB. The effect of surgery for endometrioma on ovarian reserve evaluated by antral follicle count: a systematic review and meta-analysis. Hum Reprod 2014;29:2190–8. Broekmans FJ, de Ziegler D, Howles CM, Gougeon A, Trew G, Olivennes F. The antral follicle count: practical recommendations for better standardization. Fertil Steril 2010;94:1044–51.
1550
31.
32.
33. 34.
35.
36.
37.
38.
Broekmans FJ, Kwee J, Hendriks DJ, Mol BW, Lambalk CB. A systematic review of tests predicting ovarian reserve and IVF outcome. Hum Reprod Update 2006;12:685–718. Savelli L. Transvaginal sonography for the assessment of ovarian and pelvic endometriosis: how deep is our understanding? Ultrasound Obstet Gynecol 2009;33:497–501. Busnelli A, Somigliana E, Benaglia L, Leonardi M, Ragni G, Fedele L. In vitro fertilization outcomes in treated hypothyroidism. Thyroid 2013;23:1319–25. Somigliana E, Benaglia L, Vercellini P, Paffoni A, Ragni G, Fedele L. Recurrent endometrioma and ovarian reserve: biological connection or surgical paradox? Am J Obstet Gynecol 2011;204:529.e1–5. Vercellini P, Somigliana E, Vigano P, Abbiati A, Barbara G, Fedele L. ‘‘Blood on the tracks’’ from corpora lutea to endometriomas. BJOG 2009;116:366–71. P, De Matteis S, Barbara G, Fedele L. PostVercellini P, Somigliana E, Vigano operative endometriosis recurrence: a plea for prevention based on pathogenetic, epidemiological and clinical evidence. Reprod Biomed Online 2010;21:259–65. Elizur SE, Chian RC, Holzer HE, Gidoni Y, Tulandi T, Tan SL. Cryopreservation of oocytes in a young woman with severe and symptomatic endometriosis: a new indication for fertility preservation. Fertil Steril 2009;91:293.e1–3. Garcia-Velasco JA, Domingo J, Cobo A, Martínez M, Carmona L, Pellicer A. Five years' experience using oocyte vitrification to preserve fertility for medical and nonmedical indications. Fertil Steril 2013;99:1994–9.
VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility®
SUPPLEMENTAL FIGURE 1
Diagram of the study population. The analyses focused on the ovaries. The arrows show the contribution of the different categories of women (based on history of surgery for endometrioma and presence of endometrioma) to the four study groups: unoperated gonads without endometriomas, unoperated gonads with endometriomas, operated gonads without endometriomas, operated gonads with endometriomas. (N ¼ number of women; n ¼ number of gonads.) *Four women had homolateral recurrences and nine had contralateral recurrences. Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
VOL. 103 NO. 6 / JUNE 2015
1550.e1
ORIGINAL ARTICLE: ENDOMETRIOSIS
SUPPLEMENTAL TABLE 1 IVF cycle characteristics in recruited women (n [ 83). Characteristic
Mean ± SD, median (IQR), or n (%)
Stimulation protocol Long protocol GnRH antagonist Short protocol Cancelled cycle Total dose of administered FSH (IU) Duration of stimulation (d) No. of oocyte retrieveda No. of oocytes usedb Technique usedb IVF ICSI Fertilization rateb No. of embryos transferredc 1 2 3 Clinical pregnancy rate Implantation rate
37 (45%) 29 (35%) 17 (20%) 7 (8%) 3,118 1,492 9.7 3.0 7.1 5.6 4.7 2.9 19 (28%) 49 (72%) 75% (50–100%) 24 (43%) 30 (54%) 2 (3%) 21 (25%) 25 (28%)
Note: FSH ¼ follicle-stimulating hormone; GnRH ¼ gonadotropin-releasing hormone; ICSI ¼ intracytoplasmic sperm injection; IQR ¼ interquartile range; SD ¼ standard deviation. a Data refers to patients underwent oocyte retrieval (n ¼ 76). b Data refers to patients who retrieved R1 suitable oocyte (n ¼ 68). c Data refers to patients who obtained R1 viable embryo (n ¼ 56). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
1550.e2
VOL. 103 NO. 6 / JUNE 2015
Fertility and Sterility®
SUPPLEMENTAL TABLE 2 Main stimulation characteristics in the four study groups. Nonoperated gonads Characteristic No. of gonads Stimulation protocol Long protocol GnRH antagonist Short protocol Cancelled cycle Total dose of administered FSH (IU) Duration of stimulation (d) Total no. of follicles
Previously operated gonads
Without endometriomas
With endometriomas
Without endometriomas
With endometriomas
42
46
55
23
22 (52%) 12 (29%) 8 (19%) 1 (2%) 3,059 1,547
25 (54%) 14 (31%) 7 (15%) 5 (11%) 2,905 1,502
21 (38%) 20 (36%) 14 (26%) 8 (14%) 3,327 1,372
6 (26%) 12 (52%) 5 (22%) 0 (0%) 3,171 1,637
.07 .59
10.3 2.6 5.9 4.5
9.7 2.6 5.1 4.6
9.3 3.7 2.6 3.1
9.8 2.4 3.6 3.1
.47 < .001a
P value – .24
Note: FSH ¼ follicle-stimulating hormone; GnRH ¼ gonadotropin-releasing hormone. a At post hoc analysis, a statistically significant difference emerged for the comparison between previously operated gonads without recurrences and nonoperated gonads without endometriomas (P< .001) and between previously operated gonads without recurrences and nonoperated gonads with endometriomas (P¼ .01). Benaglia. AFC and ovarian endometriomas. Fertil Steril 2015.
VOL. 103 NO. 6 / JUNE 2015
1550.e3
