- Email: [email protected]
Ovarian recovery after laparoscopic enucleation of ovarian cysts: Insights from echographic short-term postsurgical follow-up
Journal of Minimally Invasive Gynecology (2005) 12, 409-414
Original article
Ovarian recovery after laparoscopic enucleation of ovarian cysts: Insights from echographic short-term postsurgical follow-up Massimo Candiani, MD, Maurizio Barbieri, MD, Barbara Bottani, MD, Carlo Bertulessi, MD, Michele Vignali, MD, Benedetta Agnoli, MD, Edgardo Somigliana, MD, and Mauro Busacca, MD From the Department of Obstetrics, Gynecology and Neonatology, Policlinico, L. Mangiagalli Hospital (Drs. Candiani, Barbieri, Bottani, Agnoli, Somigliana); the Università Degli Studi di Milano, (Drs. Candiani, Vignali, Busacca, and Bertulessi); and the Department of Obstetrics and Gynecology, Macedonio Melloni Hospital, University of Milan, (Drs. Vignali, Busacca, and Bertulessi), Milan, Italy. KEYWORDS: Laparoscopy; Ovarian cyst; Ovarian reserve; Ultrasound
Abstract STUDY OBJECTIVE: To evaluate damage to ovarian reserve following laparoscopic cystectomy of benign ovarian cysts. DESIGN: Prospective study (Canadian Task Force classification II-3). SETTING: Tertiary gynecologic endoscopic unit at a university-affiliated hospital. PATIENTS. Thirty-one patients who underwent excision of monolateral (n ⫽ 25) or bilateral (n ⫽ 6) benign ovarian cysts. INTERVENTIONS: Serial transvaginal ultrasound examinations during the first and third postsurgical menstrual cycles. The following ovarian echographic variables were evaluated: antral follicle count, ovarian volume, stromal blood flow, and side of ovulation. Two types of statistical analysis were performed: a paired analysis comparing operated and intact ovaries of the same patient and a prospective analysis comparing ecographic characteristics of the operated gonad at first and second evaluation. MEASUREMENTS AND MAIN RESULTS: Antral follicle count and stromal blood flow were not significantly affected by surgery. While ovarian volume was similar in the operated and in the contralateral intact gonad at the first ultrasound evaluation, the volume of the operated ovary was significantly reduced at the second assessment. The median (interquartile range) of the percentage of this reduction was 33% (18%– 81%). This progressive reduction was confirmed by prospectively analyzing the operated ovaries. An increased probability of ovulation in the intact gonad was observed at both assessments. CONCLUSION: Laparoscopic excision of ovarian cysts is associated with damage to ovarian reserve, at least immediately after surgery. This effect does not appear to be consequent to an injury to ovarian vascularization. © 2005 AAGL. All rights reserved.
Presented at the 33rd annual meeting of the American Association of Gynecologic Laparoscopists, San Francisco, California, November 9 –13, 2004. Corresponding author: Dr. E. Somigliana, Department of Obstetrics, Gynecology and Neonatology (DIOGENE), Policlinico, L. Mangiagalli Hospital, Via Commenda 12, 20122 Milano, Italy. E-mail: [email protected] Submitted November 18, 2004. Accepted for publication March 28, 2005.
1553-4650/$ -see front matter © 2005 AAGL. All rights reserved. doi:10.1016/j.jmig.2005.06.006
410
Journal of Minimally Invasive Gynecology, Vol 12, No 5, September/October 2005
Laparoscopic cystectomy currently is considered the first-line choice for the conservative treatment of benign ovarian cysts and has gained increasing acceptance among gynecologic surgeons.1-4 However, the safety of this technique in terms of ovarian damage to the operated gonad has recently been questioned. In particular, one study showed that in patients previously operated for excision of a unilateral benign ovarian cyst and who successively undergo in vitro fertilization (IVF), the number both of follicles and retrieved oocytes obtained in the operated gonad during ovarian hyperstimulation is markedly reduced when compared with the contralateral intact ovary.5 This observation has been successively confirmed.6-9 A major limitation of these studies is the selection of patients evaluated. Indeed, they all focused on patients who required assisted reproductive technology techniques. Thus, it may be argued that these women are not representative of all patients undergoing laparoscopic excision of benign ovarian cysts. In other words, recruited patients may represent a subgroup of women whose gonads have been more severely damaged. Studies prospectively evaluating ovarian reserve of unselected patients undergoing laparoscopic excision of benign ovarian cysts are scanty. Such studies are difficult to carry out due to difficulties related to the evaluation of ovarian reserve. Indeed, since ovarian function cannot be measured directly, ovarian response to gonadotropin hyperstimulation currently is considered the most appropriate surrogate measurement for ovarian function. However, the induction of ovarian hyperstimulation in an unselected population of surgical patients with the mere purpose to evaluate ovarian function is obviously ethically untenable. Other possibilities include serum hormonal evaluation and ultrasound examination.10,11 The use of serum measurements (i.e., follicle-stimulating hormone [FSH], inhibin B, 17--estradiol, FSH/LH ratio, anti-müllerian hormone) in order to evaluate ovarian damage in patients who have undergone excision of ovarian cysts is of limited value considering that the vast majority of patients undergo monolateral excision of a cyst. The contralateral intact gonad may completely substitute for reduced function of the operated ovary.12 Ultrasound examination appears more appealing in this specific context. Indeed, this approach allows a separate evaluation of the ovaries without exposing the woman to risks or discomfort. Indicators of ovarian reserve most commonly accepted are antral follicle count (AFC), ovarian volume, and ovarian stromal blood flow.11,13,14 To the best of our knowledge, no study has evaluated modification of ultrasound markers of ovarian reserve in an unselected cohort of patients undergoing laparoscopic ovarian cystectomy. In this study, we have prospectively investigated a consecutive series of women who underwent laparoscopic enucleation of benign ovarian cysts. Specifically, patients underwent serial ultrasound examinations during the first three months after surgery focusing on sonographic markers of ovarian reserve.
Materials and methods All patients referred to the laparoscopic service of the department of obstetrics and gynecology of the L. Mangiagalli Institute from January through July 2003 were offered the opportunity to enter the study. Inclusion criteria were as follows: 1) age 18 – 40 years; 2) presence of a uni/bilateral ovarian cyst(s); 3) regular menstrual cycles; and 4) agreement to be enrolled in the study. Exclusion criteria were as follows: 1) prior ovarian surgery; 2) surgical necessity to perform adnexectomy; 3) presence of more than one cyst per ovary; 4) diagnosis of malignancy; and 5) oral contraceptive use before surgery. The local ethics committee accepted the protocol of the study, and all patients gave written informed consent. A transvaginal echographic examination was systematically performed before surgery in order to precisely record the dimension of the cyst and rule out a diagnosis of functional ovarian cyst. Operative laparoscopy was performed as previously described.3,15 After initial diagnostic evaluation of the pelvis and abdomen and washing, the ovarian cyst was punctured and opened. The contents were aspirated and visually inspected, and the inner wall of the cyst was checked for possible vegetations. Removal of the was carried out by identifying the cyst wall and removing it from the ovarian cortex by traction with grasping forceps.3 In cases of ultrasonographic diagnosis of a dermoid cyst, every effort was made to excise the entire cyst without spilling its contents. When necessary, hemostasis was achieved with bipolar forceps applied on ovarian parenchyma. No sutures were used for reapproximation of the ovarian edges. All interventions were performed by two skilled surgeons (MBu, MC). All patients were asked to return within day 8 of the first menstrual cycle after surgery to undergo a basal ultrasound evaluation. According to the ecographic situation, patients were given a second and, if necessary, a third appointment to repeat an ultrasound scan in order to determine if ovulation took place and the side of ovulation. The same protocol was repeated after the third postsurgical menstruation. This second evaluation was not performed if pregnancy occurred or if patients were prescribed oral contraceptives for recurrence of symptoms and/or request for hormonal contraception. All ultrasound scans were performed by two of the authors (BB and MBa), who have been experienced in the use of this technique for many years. Operators were blinded regarding the side of previous surgery. All ultrasound examinations were carried out with the same instrument (Aloka, Model SSD-2200, Aloka Co Ltd., Tokyo, Japan) using a 6.5 MHz vaginal probe. Ovarian volume was calculated using the following formula: 4 ⁄ 3 * * (d ⁄ 2)3
(1)
where d was obtained as the mean diameter of the three perpendicular measurements of the ovary. Antral follicle count was defined as the total number of follicles with a
Candiani et al Table 1
Postsurgical ovarian recovery
411
Characteristics of patients enrolled in the study
Characteristic Mean age (yrs) Histology*/No. (%) Endometriosis Dermoid Serous Mucinous Ovarian side/No. (%) Left Right Bilateral Dimension of the cysts Median diameter in cm (range) Median volume in mL (range)
Unilateral cysts (n ⫽ 25)
Bilateral cysts (n ⫽ 6)
Total (N ⫽ 31)
33.5 ⫾ 3.9
32.5 ⫾ 4.6
33.3 ⫾ 4.0
14 5 4 2
(56) (20) (16) (8)
12 (48) 13 (52) NA 4.0 (2.5–9.3) 34 (8–420)
5 1 0 0
(83) (17) (0) (0)
NA NA 6 (100) 4.4 (2.6–7.4) 45 (9–217)
19 6 4 2
(61) (19) (13) (7)
12 (39) 13 (42) 6 (19) 4.0 (2.5–9.3) 34 (8–420)
NA ⫽ not applicable.
diameter less than 10 mm. This variable was counted in each ovary at first ecography of the studied cycle. The stromal blood flow of the ovary was assessed by color Doppler ultrasound. Flow velocity waveforms were obtained from stromal blood vessels away from the ovarian capsule and the utero-ovarian ligament. The “gate” of the Doppler was positioned when the vessel with good color signals was identified on the screen. The peak systolic velocity (PSV) of stromal vessels was calculated electronically when at least three similar, consecutive waveforms of good quality were obtained. Ovulation was defined as the presence of a follicle with a mean diameter of 16 mm or more and successive formation of a corpus luteum. Previous preliminary experiments on 10 healthy volunteers documented intra and interobserver coefficients of variation for all measurements always below 10%. All these women had regular menstrual cycles, and their mean age was 31.1 ⫾ 3.5 years. Quantitative data were expressed as median (interquartile range [IQR]) and compared using nonparametric signed-rank Wilcoxon test for paired data. Qualitative data (side of ovulation) were compared using McNemar test. A p value ⱕ.05 was considered statistically significant. The SPSS (Chicago, IL) statistical package was used for data analysis.
Results During the study period, 59 patients were found to be eligible. Twenty-eight of them refused to participate. Reasons to refuse were inability to adhere to our protocol due to job constraints (n ⫽ 8), living far from our institution (n ⫽ 12), and not interested (n ⫽ 8). Baseline characteristics of patients who refused to enter the study were not significantly different from those of women who agreed to participate.
Baseline characteristics of the 31 patients who agreed to participate are shown in Table 1. Laparoscopic excision was successfully completed in all these women. All recruited
Figure 1 Patients were evaluated after the first menstrual cycle following surgery (first postsurgical ultrasound evaluation). According to the echographic situation, patients were reevaluated a second and, if necessary, a third time in order to determine if ovulation took place and the side of ovulation. The same protocol was repeated after the third postsurgical menstruation (second postsurgical ultrasound evaluation). This second evaluation was not performed in 11 patients. Reasons for not attempting the second evaluation were as follows: six women became pregnant (of whom five had endometriotic ovarian cysts), three required oral contraceptive use for recurrence of symptoms and/or request for hormonal contraception, and two refused to continue the study protocol.
412
Journal of Minimally Invasive Gynecology, Vol 12, No 5, September/October 2005
Table 2 Comparison of the ovarian reserve of the operated and contralateral intact ovaries in patients undergoing excision of a monolateral ovarian cyst Characteristics First postsurgical evaluation (n ⫽ 25) Median ovarian volume in mL (range) Median number of antral follicles (range) Mean peak systolic velocity (cm/s) Side of ovulation (%) Second postsurgical evaluation (n ⫽ 15) Median ovarian volume in mL (range) Median number of antral follicles (range) Mean peak systolic velocity (cm/s) Side of ovulation (%)†
Intact ovary
Operated ovary
p*
9.3 4 11.2 17/25
(6.2–10.9) (1–7) (7.7–16.5) (68)
8.4 4 12.4 8/25
(6.1–16.2) (2–6) (8.0–14.1) (32)
.62 .74 .35 .05
9.8 5.5 13.2 11/11
(6.8–11.5) (2–8.25) (8.0–16.4) (100)
6.6 4 12.2 0/11
(1.9–8.7) (1.0–5.25) (7.4–19.9) (0)
.004 .25 .94 ⬍.001
cm/s ⫽ centimeters per second. *p values were calculated using the nonparametric signed-rank Wilcoxon test for paired data for ovarian volume, number of follicles, and stromal blood flow. Side of ovulation was compared using McNemar test. †Four cases were missing. Four women underwent only the first ecography of the second evaluation but refused to come back for subsequent evaluations.
patients underwent the first postsurgical evaluation. Twenty of them (65%) were also evaluated at the second assessment. The precise trial profile diagram is shown in Figure 1. Histologic diagnoses of the 20 women evaluated at second assessment were as follows: endometriomas (12 women), dermoid cyst (4), serous cyst (2), and mucinous cyst (2). Results from ultrasound evaluations in women who underwent monolateral excision of an ovarian cyst are reported in Table 2. Paired comparisons of the operated and the contralateral intact gonad are reported. Whereas ovarian volume was similar in the operated and in the contralateral intact gonad at first ultrasound evaluation, a statistically significant reduction in basal volume was observed at the second assessment. The median (IQR) of the percentage of reduction was 33% (18%– 81%). Conversely, total number of follicles and stromal blood flow results were similar at both evaluations. An increased probability of ovulation in the intact gonad was observed at both assessments (Table 2). Ovulation was identified in all patients who underwent bilateral excision of ovarian cysts (six and five women at first and second examination, respectively). The progressive reduction in basal volume of the ovaries soon after surgery was confirmed by prospectively analyzing the operated ovaries. Specifically, data on 25 ovaries from 15 patients who underwent unilateral excision of ovarian cysts and from 5 patients who underwent bilateral excision were available. Median (IQR) basal volume of the operated ovaries dropped from 8.4 (4.1–18.3) to 6.1 (2.1– 8.8) mL (p ⫽ .004, Figure 2). The median (IQR) of the percentage of reduction was 45% (5%– 60%). No differences could be demonstrated regarding the total number of follicles (from five [3– 6] to four [3–5], p ⫽ .34) and the mean PSV (from 12.4 [8.7–14.2] to 12.0 [8.2–18.2] cm/s, p ⫽ .59). Of note, we have repeated the same analysis for the 15 available unoperated ovaries: no difference was observed for both variables (data not shown).
Discussion In this study, we documented that ovarian volume after surgical excision of an ovarian cyst progressively decreases over the first three months after surgery. Conversely, AFC and ovarian stromal blood flow do not appear to be influenced by surgery.
postsurgical ultrasound postsurgical ultrasound Figure 2 Ovarian volume of the operated ovary at first menstrual cycle after surgery (first postsurgical ultrasound evaluation) and after the third postsurgical menstruation (second postsurgical ultrasound evaluation) in 25 operated gonads. A statistically significant reduction in ovarian volume over time was documented. * p ⫽ .004
Candiani et al
Postsurgical ovarian recovery
Ovarian volume has been reported to be a reliable indicator of ovarian reserve by several authors.16 Thus, based on the results observed in our study, it may be inferred that excision of an ovarian cyst is associated with significant damage to ovarian reserve. Specifically, the median (IQR) of the percentage of reduction of the ovarian volume was estimated to be 33% (18%– 81%). Of note, this reduction took place progressively over time. Indeed, the volume of the operated and contralateral intact gonad was similar on first postsurgical follow-up whereas this difference became significant at ultrasound assessment performed two months later. In line with this finding, a study recently documented that ovarian volume of ovaries previously operated on for endometriomas is reduced when compared with contralateral intact gonads. The mean time between surgery and ultrasound evaluation was 15 ⫾ 10 months.17 Of note, the same authors failed to observe this reduction when dermoid cysts were considered.18 Recent findings support that AFC should also be considered a reliable indicator of ovarian reserve.19 In our study, we failed to observe a reduction in AFC despite the marked reduction documented for ovarian volume. Similarly, stromal blood flow, another recognized indicator of ovarian reserve, did not differ in the operated and in the contralateral intact gonad. Using three-dimensional Doppler technology to quantify ovarian stromal blood flow, researchers recently reported similar results in women selected for IVF who previously underwent laparoscopic excision of endometriomas.20 Discrepancies among the results observed using different sonographic tests for ovarian reserve are difficult to explain. It is interesting to note that a similar pattern of modification of ultrasound markers of ovarian reserve has been reported recently among women who received chemotherapy for cancer treatment during their childhood.21 Exposure to antineoplastic agents is known to severely damage ovarian reserve. On the other hand, the observed pattern of changes on sonographic evaluations in our study may help clarify the possible mechanisms that cause gonad injury. In this context, potential deleterious insults are the amount of ovarian tissue that may be removed inadvertently during cystectomy and the damage that may be inflicted on the ovarian stroma and vascularization by both surgery-related local inflammation and electrosurgical coagulation during hemostasis. Results from the present study document that vascularization is not compromised, thus suggesting that blood support to the ovary is not hampered by surgery. Our results support a role for local inflammation. The reduction of ovarian volume observed over the first three months after surgery may be consequent to this type of process. However, it cannot be excluded that this reduction simply may be the consequence of an initial ovarian edema later disappearing that may initially mask the inadvertent removal of healthy ovarian tissue. Finally, the possibility that the damage may at least in part be caused by the cyst per se, so that some degree of
413 injury might be already present at the time of surgery, cannot be ruled out. A statistically significant difference in the rate of ovulation also was observed in this study. Ovulation took place in the intact gonad in 17 out of 25 patients (68%) and in all patients (11 out of 11, 100%) at first and second assessment, respectively. This result is in line with a previous study that reported a decreased rate of spontaneous ovulation in gonads of infertile women previously operated on for ovarian cysts.6 It has to be noted, however, that the side of ovulation is not currently considered a marker of ovarian function. In our study, we simply aimed to evaluate whether ovulation took place, and the side of ovulation was not a primary endpoint. Our study was not designed to evaluate whether the rate of ovulation is reduced in operated ovaries. Indeed, if the alternative-ovulation-side theory is true, assessment at first and third postsurgical cycle would invariably lead to observed ovulation of the same side. Finally, it has to be noted that ovulation at first postsurgical assessment was always documented in women undergoing bilateral excision of ovarian cysts. Thus, results from our study suggest that mechanisms of ovulation are presumably transiently impaired in the operated gonad at first cycle after surgery but do not provide data to support a reduced rate of ovulation in these ovaries. Overall, we believe that the reduced rate of ovulation in the operated ovary should be interpreted with caution. Larger studies recruiting unselected women and specifically designed to address this issue are required to clarify whether the rate of ovulation is reduced in operated gonads. As previously mentioned, most studies on the topic of ovarian reserve after surgery are provided by infertility centers and are consequently limited by the selection of patients.5-9 On the other hand, unselected studies are difficult to perform. Despite the short follow-up period in our study, the second postsurgical assessment was performed in only 65% of patients (20 of 31). Reasons patients dropped out may have influenced the results observed. Indeed, it cannot be excluded that the damage might have been less severe in women who achieved a pregnancy (six women), or it might have been more severe in women who did not choose to further adhere to the study protocol after the first ultrasound evaluation (two women) or who required oral contraceptives after surgery (three women). Patient selection is a crucial topic in this area. Further insights on damage associated with laparoscopic excision of ovarian cysts should be drawn from large series with longer follow-up of unselected women. An important aspect that has to be underlined is that the current study deals with damage to ovarian reserve, not with fertility potential. Even if confirmed, documentation of surgery-mediated damage to ovarian reserve cannot be used to argue against surgical treatments. Indeed, surgery currently is considered an effective therapy
414
Journal of Minimally Invasive Gynecology, Vol 12, No 5, September/October 2005
in infertile women with endometriosis at advanced stages. The probability to conceive spontaneously after surgical removal of endometriotic ovarian cysts has been reported to vary between 24% and 67%.22 Of note, although a reduction in ovarian volume is an established indicator of ovarian reserve in patients who have not had surgery, its value in women undergoing ovarian surgery has not been demonstrated. In other words, it might be speculated that the reduction in ovarian volume in this context may represent a quantitative but not a qualitative injury to ovarian reserve. This hypothesis warrants confirmation.
Conclusion The present study supports the opinion that laparoscopic excision of ovarian cysts is associated with damage to ovarian reserve, at least immediately after surgery. Moreover, our results tend to rule out a role for an injury of ovarian vascularization in determining this effect. Several aspects remain to be clarified. Does the damage continue to progress after the first 3 months? What is the rate of gonads severely damaged? Is the reduction of ovarian reserve a consequence of surgery or is the damage already present before surgery? If the former, are there less-harmful alternative therapeutical approaches?
References 1. Sutton CJ, Ewen SP, Jacobs SA, et al. Laser laparoscopic surgery in the treatment of ovarian endometriomas. J Am Assoc Gynecol Laparosc. 1997;4:319-323. 2. Canis M, Rabischong B, Houlle C, et al. Laparoscopic management of adnexal masses: a gold standard? Curr Opin Obstet Gynecol. 2002; 14:423-428. 3. Chapron C, Vercellini P, Barakat H, et al. Management of ovarian endometriomas. Hum Reprod Update. 2002;8:591-597. 4. Chapron C, Fauconnier A, Goffinet F, et al. Laparoscopic surgery is not inherently dangerous for patients presenting with benign gynaecologic pathology: results of a meta-analysis. Hum Reprod. 2002;17: 1334-1342. 5. Nargund G, Cheng WC, Parsons J. The impact of ovarian cystectomy on ovarian response to stimulation during in-vitro fertilization cycles. Hum Reprod. 1996;11:81-83.
6. Loh FH, Tan AT, Kumar J, et al. Ovarian response after laparoscopic ovarian cystectomy for endometriotic cysts in 132 monitored cycles. Fertil Steril. 1999;72:316-321. 7. Ho HY, Lee RK, Hwu YM, et al. Poor response of ovaries with endometrioma previously treated with cystectomy to controlled ovarian hyperstimulation. J Assist Reprod Genet. 2002;19:507-511. 8. Somigliana E, Ragni G, Benedetti F, et al. Does laparoscopic excision of endometriotic ovarian cysts significantly affect ovarian reserve? Insights from IVF cycles. Hum Reprod. 2003;18:2450-2453. 9. Somigliana E, Ragni G, Benedetti F, et al. Does laparoscopic excision of ovarian cysts significantly affect ovarian reserve? Insights from in vitro fertilization cycles. Presented at the 59th Annual Meeting of the American Society for Reproductive Medicine, San Antonio, Texas, October 11–15, 2003. 10. Bukman A, Heineman M: Ovarian reserve testing and the use of prognostic models in patients with subfertility. Hum Reprod Update. 2001;7:581-590. 11. Tarlatzis B, Zepiridis L, Grimbizis G, et al. Clinical management of low ovarian response to stimulation for IVF: a systematic review. Hum Reprod Update. 2003;9:61-76. 12. Lass A. The fertility potential of women with a single ovary. Hum Reprod Update. 1999;5:546-550. 13. Chan C, Ng E, Li C, et al. Impaired ovarian blood flow and reduced antral follicle count following laparoscopic salpingectomy for ectopic pregnancy. Hum Reprod. 2003;18:2175-2180. 14. Ng E, Yeung W, Fong D, et al. Effects of age on hormonal and ultrasound markers of ovarian reserve in Chinese women with proven fertility. Hum Reprod. 2003;18:2169-2174. 15. Muzii L, Bianchi A, Croce C, et al. Laparoscopic excision of ovarian cysts: is the stripping technique a tissue-sparing procedure? Fertil Steril. 2002;77:609-614. 16. Lass A, Brinsden P. The role of ovarian volume in reproductive medicine. Hum Reprod Update. 1999;5:256-266. 17. Exacoustos C, Zupi E, Amadio A, et al. Laparoscopic removal of endometriomas: sonographic evaluation of residual functioning ovarian tissue. Am J Obstet Gynecol. 2004;191:68-72. 18. Zupi E, Exacoustos C, Szabolcs B, et al. Laparoscopic approach to dermoid cysts: combined surgical technique and ultrasonographic evaluation of residual functioning ovarian tissue. J Am Assoc Gynecol Laparosc. 2003;10:154-158. 19. Bancsi L, Broekmans F, Eijkemans M, et al. Predictors of poor ovarian response in in vitro fertilization: a prospective study comparing basal markers of ovarian reserve. Fertil Steril. 2002;77:328-336. 20. Wu MH, Tsai SJ, Hsiao KY, Chang FM. Three-dimensional Doppler imaging of ovarian stromal blood flow in women with endometriosis undergoing in vitro fertilization. Ultrasound Obstet Gynecol. 2003;21: 480-485. 21. Bath L, Wallace W, Shaw M, et al. Depletion of ovarian reserve in young women after treatment for cancer in childhood: detection by anti-Mullerian hormone, inhibin B and ovarian ultrasound. Hum Reprod. 2003;18:2368-2374. 22. Jones KD, Sutton CJG. Pregnancy rates following ablative laparoscopic surgery for endometriomas. Hum Reprod. 2002;17:782-785.
Original article
Ovarian recovery after laparoscopic enucleation of ovarian cysts: Insights from echographic short-term postsurgical follow-up Massimo Candiani, MD, Maurizio Barbieri, MD, Barbara Bottani, MD, Carlo Bertulessi, MD, Michele Vignali, MD, Benedetta Agnoli, MD, Edgardo Somigliana, MD, and Mauro Busacca, MD From the Department of Obstetrics, Gynecology and Neonatology, Policlinico, L. Mangiagalli Hospital (Drs. Candiani, Barbieri, Bottani, Agnoli, Somigliana); the Università Degli Studi di Milano, (Drs. Candiani, Vignali, Busacca, and Bertulessi); and the Department of Obstetrics and Gynecology, Macedonio Melloni Hospital, University of Milan, (Drs. Vignali, Busacca, and Bertulessi), Milan, Italy. KEYWORDS: Laparoscopy; Ovarian cyst; Ovarian reserve; Ultrasound
Abstract STUDY OBJECTIVE: To evaluate damage to ovarian reserve following laparoscopic cystectomy of benign ovarian cysts. DESIGN: Prospective study (Canadian Task Force classification II-3). SETTING: Tertiary gynecologic endoscopic unit at a university-affiliated hospital. PATIENTS. Thirty-one patients who underwent excision of monolateral (n ⫽ 25) or bilateral (n ⫽ 6) benign ovarian cysts. INTERVENTIONS: Serial transvaginal ultrasound examinations during the first and third postsurgical menstrual cycles. The following ovarian echographic variables were evaluated: antral follicle count, ovarian volume, stromal blood flow, and side of ovulation. Two types of statistical analysis were performed: a paired analysis comparing operated and intact ovaries of the same patient and a prospective analysis comparing ecographic characteristics of the operated gonad at first and second evaluation. MEASUREMENTS AND MAIN RESULTS: Antral follicle count and stromal blood flow were not significantly affected by surgery. While ovarian volume was similar in the operated and in the contralateral intact gonad at the first ultrasound evaluation, the volume of the operated ovary was significantly reduced at the second assessment. The median (interquartile range) of the percentage of this reduction was 33% (18%– 81%). This progressive reduction was confirmed by prospectively analyzing the operated ovaries. An increased probability of ovulation in the intact gonad was observed at both assessments. CONCLUSION: Laparoscopic excision of ovarian cysts is associated with damage to ovarian reserve, at least immediately after surgery. This effect does not appear to be consequent to an injury to ovarian vascularization. © 2005 AAGL. All rights reserved.
Presented at the 33rd annual meeting of the American Association of Gynecologic Laparoscopists, San Francisco, California, November 9 –13, 2004. Corresponding author: Dr. E. Somigliana, Department of Obstetrics, Gynecology and Neonatology (DIOGENE), Policlinico, L. Mangiagalli Hospital, Via Commenda 12, 20122 Milano, Italy. E-mail: [email protected] Submitted November 18, 2004. Accepted for publication March 28, 2005.
1553-4650/$ -see front matter © 2005 AAGL. All rights reserved. doi:10.1016/j.jmig.2005.06.006
410
Journal of Minimally Invasive Gynecology, Vol 12, No 5, September/October 2005
Laparoscopic cystectomy currently is considered the first-line choice for the conservative treatment of benign ovarian cysts and has gained increasing acceptance among gynecologic surgeons.1-4 However, the safety of this technique in terms of ovarian damage to the operated gonad has recently been questioned. In particular, one study showed that in patients previously operated for excision of a unilateral benign ovarian cyst and who successively undergo in vitro fertilization (IVF), the number both of follicles and retrieved oocytes obtained in the operated gonad during ovarian hyperstimulation is markedly reduced when compared with the contralateral intact ovary.5 This observation has been successively confirmed.6-9 A major limitation of these studies is the selection of patients evaluated. Indeed, they all focused on patients who required assisted reproductive technology techniques. Thus, it may be argued that these women are not representative of all patients undergoing laparoscopic excision of benign ovarian cysts. In other words, recruited patients may represent a subgroup of women whose gonads have been more severely damaged. Studies prospectively evaluating ovarian reserve of unselected patients undergoing laparoscopic excision of benign ovarian cysts are scanty. Such studies are difficult to carry out due to difficulties related to the evaluation of ovarian reserve. Indeed, since ovarian function cannot be measured directly, ovarian response to gonadotropin hyperstimulation currently is considered the most appropriate surrogate measurement for ovarian function. However, the induction of ovarian hyperstimulation in an unselected population of surgical patients with the mere purpose to evaluate ovarian function is obviously ethically untenable. Other possibilities include serum hormonal evaluation and ultrasound examination.10,11 The use of serum measurements (i.e., follicle-stimulating hormone [FSH], inhibin B, 17--estradiol, FSH/LH ratio, anti-müllerian hormone) in order to evaluate ovarian damage in patients who have undergone excision of ovarian cysts is of limited value considering that the vast majority of patients undergo monolateral excision of a cyst. The contralateral intact gonad may completely substitute for reduced function of the operated ovary.12 Ultrasound examination appears more appealing in this specific context. Indeed, this approach allows a separate evaluation of the ovaries without exposing the woman to risks or discomfort. Indicators of ovarian reserve most commonly accepted are antral follicle count (AFC), ovarian volume, and ovarian stromal blood flow.11,13,14 To the best of our knowledge, no study has evaluated modification of ultrasound markers of ovarian reserve in an unselected cohort of patients undergoing laparoscopic ovarian cystectomy. In this study, we have prospectively investigated a consecutive series of women who underwent laparoscopic enucleation of benign ovarian cysts. Specifically, patients underwent serial ultrasound examinations during the first three months after surgery focusing on sonographic markers of ovarian reserve.
Materials and methods All patients referred to the laparoscopic service of the department of obstetrics and gynecology of the L. Mangiagalli Institute from January through July 2003 were offered the opportunity to enter the study. Inclusion criteria were as follows: 1) age 18 – 40 years; 2) presence of a uni/bilateral ovarian cyst(s); 3) regular menstrual cycles; and 4) agreement to be enrolled in the study. Exclusion criteria were as follows: 1) prior ovarian surgery; 2) surgical necessity to perform adnexectomy; 3) presence of more than one cyst per ovary; 4) diagnosis of malignancy; and 5) oral contraceptive use before surgery. The local ethics committee accepted the protocol of the study, and all patients gave written informed consent. A transvaginal echographic examination was systematically performed before surgery in order to precisely record the dimension of the cyst and rule out a diagnosis of functional ovarian cyst. Operative laparoscopy was performed as previously described.3,15 After initial diagnostic evaluation of the pelvis and abdomen and washing, the ovarian cyst was punctured and opened. The contents were aspirated and visually inspected, and the inner wall of the cyst was checked for possible vegetations. Removal of the was carried out by identifying the cyst wall and removing it from the ovarian cortex by traction with grasping forceps.3 In cases of ultrasonographic diagnosis of a dermoid cyst, every effort was made to excise the entire cyst without spilling its contents. When necessary, hemostasis was achieved with bipolar forceps applied on ovarian parenchyma. No sutures were used for reapproximation of the ovarian edges. All interventions were performed by two skilled surgeons (MBu, MC). All patients were asked to return within day 8 of the first menstrual cycle after surgery to undergo a basal ultrasound evaluation. According to the ecographic situation, patients were given a second and, if necessary, a third appointment to repeat an ultrasound scan in order to determine if ovulation took place and the side of ovulation. The same protocol was repeated after the third postsurgical menstruation. This second evaluation was not performed if pregnancy occurred or if patients were prescribed oral contraceptives for recurrence of symptoms and/or request for hormonal contraception. All ultrasound scans were performed by two of the authors (BB and MBa), who have been experienced in the use of this technique for many years. Operators were blinded regarding the side of previous surgery. All ultrasound examinations were carried out with the same instrument (Aloka, Model SSD-2200, Aloka Co Ltd., Tokyo, Japan) using a 6.5 MHz vaginal probe. Ovarian volume was calculated using the following formula: 4 ⁄ 3 * * (d ⁄ 2)3
(1)
where d was obtained as the mean diameter of the three perpendicular measurements of the ovary. Antral follicle count was defined as the total number of follicles with a
Candiani et al Table 1
Postsurgical ovarian recovery
411
Characteristics of patients enrolled in the study
Characteristic Mean age (yrs) Histology*/No. (%) Endometriosis Dermoid Serous Mucinous Ovarian side/No. (%) Left Right Bilateral Dimension of the cysts Median diameter in cm (range) Median volume in mL (range)
Unilateral cysts (n ⫽ 25)
Bilateral cysts (n ⫽ 6)
Total (N ⫽ 31)
33.5 ⫾ 3.9
32.5 ⫾ 4.6
33.3 ⫾ 4.0
14 5 4 2
(56) (20) (16) (8)
12 (48) 13 (52) NA 4.0 (2.5–9.3) 34 (8–420)
5 1 0 0
(83) (17) (0) (0)
NA NA 6 (100) 4.4 (2.6–7.4) 45 (9–217)
19 6 4 2
(61) (19) (13) (7)
12 (39) 13 (42) 6 (19) 4.0 (2.5–9.3) 34 (8–420)
NA ⫽ not applicable.
diameter less than 10 mm. This variable was counted in each ovary at first ecography of the studied cycle. The stromal blood flow of the ovary was assessed by color Doppler ultrasound. Flow velocity waveforms were obtained from stromal blood vessels away from the ovarian capsule and the utero-ovarian ligament. The “gate” of the Doppler was positioned when the vessel with good color signals was identified on the screen. The peak systolic velocity (PSV) of stromal vessels was calculated electronically when at least three similar, consecutive waveforms of good quality were obtained. Ovulation was defined as the presence of a follicle with a mean diameter of 16 mm or more and successive formation of a corpus luteum. Previous preliminary experiments on 10 healthy volunteers documented intra and interobserver coefficients of variation for all measurements always below 10%. All these women had regular menstrual cycles, and their mean age was 31.1 ⫾ 3.5 years. Quantitative data were expressed as median (interquartile range [IQR]) and compared using nonparametric signed-rank Wilcoxon test for paired data. Qualitative data (side of ovulation) were compared using McNemar test. A p value ⱕ.05 was considered statistically significant. The SPSS (Chicago, IL) statistical package was used for data analysis.
Results During the study period, 59 patients were found to be eligible. Twenty-eight of them refused to participate. Reasons to refuse were inability to adhere to our protocol due to job constraints (n ⫽ 8), living far from our institution (n ⫽ 12), and not interested (n ⫽ 8). Baseline characteristics of patients who refused to enter the study were not significantly different from those of women who agreed to participate.
Baseline characteristics of the 31 patients who agreed to participate are shown in Table 1. Laparoscopic excision was successfully completed in all these women. All recruited
Figure 1 Patients were evaluated after the first menstrual cycle following surgery (first postsurgical ultrasound evaluation). According to the echographic situation, patients were reevaluated a second and, if necessary, a third time in order to determine if ovulation took place and the side of ovulation. The same protocol was repeated after the third postsurgical menstruation (second postsurgical ultrasound evaluation). This second evaluation was not performed in 11 patients. Reasons for not attempting the second evaluation were as follows: six women became pregnant (of whom five had endometriotic ovarian cysts), three required oral contraceptive use for recurrence of symptoms and/or request for hormonal contraception, and two refused to continue the study protocol.
412
Journal of Minimally Invasive Gynecology, Vol 12, No 5, September/October 2005
Table 2 Comparison of the ovarian reserve of the operated and contralateral intact ovaries in patients undergoing excision of a monolateral ovarian cyst Characteristics First postsurgical evaluation (n ⫽ 25) Median ovarian volume in mL (range) Median number of antral follicles (range) Mean peak systolic velocity (cm/s) Side of ovulation (%) Second postsurgical evaluation (n ⫽ 15) Median ovarian volume in mL (range) Median number of antral follicles (range) Mean peak systolic velocity (cm/s) Side of ovulation (%)†
Intact ovary
Operated ovary
p*
9.3 4 11.2 17/25
(6.2–10.9) (1–7) (7.7–16.5) (68)
8.4 4 12.4 8/25
(6.1–16.2) (2–6) (8.0–14.1) (32)
.62 .74 .35 .05
9.8 5.5 13.2 11/11
(6.8–11.5) (2–8.25) (8.0–16.4) (100)
6.6 4 12.2 0/11
(1.9–8.7) (1.0–5.25) (7.4–19.9) (0)
.004 .25 .94 ⬍.001
cm/s ⫽ centimeters per second. *p values were calculated using the nonparametric signed-rank Wilcoxon test for paired data for ovarian volume, number of follicles, and stromal blood flow. Side of ovulation was compared using McNemar test. †Four cases were missing. Four women underwent only the first ecography of the second evaluation but refused to come back for subsequent evaluations.
patients underwent the first postsurgical evaluation. Twenty of them (65%) were also evaluated at the second assessment. The precise trial profile diagram is shown in Figure 1. Histologic diagnoses of the 20 women evaluated at second assessment were as follows: endometriomas (12 women), dermoid cyst (4), serous cyst (2), and mucinous cyst (2). Results from ultrasound evaluations in women who underwent monolateral excision of an ovarian cyst are reported in Table 2. Paired comparisons of the operated and the contralateral intact gonad are reported. Whereas ovarian volume was similar in the operated and in the contralateral intact gonad at first ultrasound evaluation, a statistically significant reduction in basal volume was observed at the second assessment. The median (IQR) of the percentage of reduction was 33% (18%– 81%). Conversely, total number of follicles and stromal blood flow results were similar at both evaluations. An increased probability of ovulation in the intact gonad was observed at both assessments (Table 2). Ovulation was identified in all patients who underwent bilateral excision of ovarian cysts (six and five women at first and second examination, respectively). The progressive reduction in basal volume of the ovaries soon after surgery was confirmed by prospectively analyzing the operated ovaries. Specifically, data on 25 ovaries from 15 patients who underwent unilateral excision of ovarian cysts and from 5 patients who underwent bilateral excision were available. Median (IQR) basal volume of the operated ovaries dropped from 8.4 (4.1–18.3) to 6.1 (2.1– 8.8) mL (p ⫽ .004, Figure 2). The median (IQR) of the percentage of reduction was 45% (5%– 60%). No differences could be demonstrated regarding the total number of follicles (from five [3– 6] to four [3–5], p ⫽ .34) and the mean PSV (from 12.4 [8.7–14.2] to 12.0 [8.2–18.2] cm/s, p ⫽ .59). Of note, we have repeated the same analysis for the 15 available unoperated ovaries: no difference was observed for both variables (data not shown).
Discussion In this study, we documented that ovarian volume after surgical excision of an ovarian cyst progressively decreases over the first three months after surgery. Conversely, AFC and ovarian stromal blood flow do not appear to be influenced by surgery.
postsurgical ultrasound postsurgical ultrasound Figure 2 Ovarian volume of the operated ovary at first menstrual cycle after surgery (first postsurgical ultrasound evaluation) and after the third postsurgical menstruation (second postsurgical ultrasound evaluation) in 25 operated gonads. A statistically significant reduction in ovarian volume over time was documented. * p ⫽ .004
Candiani et al
Postsurgical ovarian recovery
Ovarian volume has been reported to be a reliable indicator of ovarian reserve by several authors.16 Thus, based on the results observed in our study, it may be inferred that excision of an ovarian cyst is associated with significant damage to ovarian reserve. Specifically, the median (IQR) of the percentage of reduction of the ovarian volume was estimated to be 33% (18%– 81%). Of note, this reduction took place progressively over time. Indeed, the volume of the operated and contralateral intact gonad was similar on first postsurgical follow-up whereas this difference became significant at ultrasound assessment performed two months later. In line with this finding, a study recently documented that ovarian volume of ovaries previously operated on for endometriomas is reduced when compared with contralateral intact gonads. The mean time between surgery and ultrasound evaluation was 15 ⫾ 10 months.17 Of note, the same authors failed to observe this reduction when dermoid cysts were considered.18 Recent findings support that AFC should also be considered a reliable indicator of ovarian reserve.19 In our study, we failed to observe a reduction in AFC despite the marked reduction documented for ovarian volume. Similarly, stromal blood flow, another recognized indicator of ovarian reserve, did not differ in the operated and in the contralateral intact gonad. Using three-dimensional Doppler technology to quantify ovarian stromal blood flow, researchers recently reported similar results in women selected for IVF who previously underwent laparoscopic excision of endometriomas.20 Discrepancies among the results observed using different sonographic tests for ovarian reserve are difficult to explain. It is interesting to note that a similar pattern of modification of ultrasound markers of ovarian reserve has been reported recently among women who received chemotherapy for cancer treatment during their childhood.21 Exposure to antineoplastic agents is known to severely damage ovarian reserve. On the other hand, the observed pattern of changes on sonographic evaluations in our study may help clarify the possible mechanisms that cause gonad injury. In this context, potential deleterious insults are the amount of ovarian tissue that may be removed inadvertently during cystectomy and the damage that may be inflicted on the ovarian stroma and vascularization by both surgery-related local inflammation and electrosurgical coagulation during hemostasis. Results from the present study document that vascularization is not compromised, thus suggesting that blood support to the ovary is not hampered by surgery. Our results support a role for local inflammation. The reduction of ovarian volume observed over the first three months after surgery may be consequent to this type of process. However, it cannot be excluded that this reduction simply may be the consequence of an initial ovarian edema later disappearing that may initially mask the inadvertent removal of healthy ovarian tissue. Finally, the possibility that the damage may at least in part be caused by the cyst per se, so that some degree of
413 injury might be already present at the time of surgery, cannot be ruled out. A statistically significant difference in the rate of ovulation also was observed in this study. Ovulation took place in the intact gonad in 17 out of 25 patients (68%) and in all patients (11 out of 11, 100%) at first and second assessment, respectively. This result is in line with a previous study that reported a decreased rate of spontaneous ovulation in gonads of infertile women previously operated on for ovarian cysts.6 It has to be noted, however, that the side of ovulation is not currently considered a marker of ovarian function. In our study, we simply aimed to evaluate whether ovulation took place, and the side of ovulation was not a primary endpoint. Our study was not designed to evaluate whether the rate of ovulation is reduced in operated ovaries. Indeed, if the alternative-ovulation-side theory is true, assessment at first and third postsurgical cycle would invariably lead to observed ovulation of the same side. Finally, it has to be noted that ovulation at first postsurgical assessment was always documented in women undergoing bilateral excision of ovarian cysts. Thus, results from our study suggest that mechanisms of ovulation are presumably transiently impaired in the operated gonad at first cycle after surgery but do not provide data to support a reduced rate of ovulation in these ovaries. Overall, we believe that the reduced rate of ovulation in the operated ovary should be interpreted with caution. Larger studies recruiting unselected women and specifically designed to address this issue are required to clarify whether the rate of ovulation is reduced in operated gonads. As previously mentioned, most studies on the topic of ovarian reserve after surgery are provided by infertility centers and are consequently limited by the selection of patients.5-9 On the other hand, unselected studies are difficult to perform. Despite the short follow-up period in our study, the second postsurgical assessment was performed in only 65% of patients (20 of 31). Reasons patients dropped out may have influenced the results observed. Indeed, it cannot be excluded that the damage might have been less severe in women who achieved a pregnancy (six women), or it might have been more severe in women who did not choose to further adhere to the study protocol after the first ultrasound evaluation (two women) or who required oral contraceptives after surgery (three women). Patient selection is a crucial topic in this area. Further insights on damage associated with laparoscopic excision of ovarian cysts should be drawn from large series with longer follow-up of unselected women. An important aspect that has to be underlined is that the current study deals with damage to ovarian reserve, not with fertility potential. Even if confirmed, documentation of surgery-mediated damage to ovarian reserve cannot be used to argue against surgical treatments. Indeed, surgery currently is considered an effective therapy
414
Journal of Minimally Invasive Gynecology, Vol 12, No 5, September/October 2005
in infertile women with endometriosis at advanced stages. The probability to conceive spontaneously after surgical removal of endometriotic ovarian cysts has been reported to vary between 24% and 67%.22 Of note, although a reduction in ovarian volume is an established indicator of ovarian reserve in patients who have not had surgery, its value in women undergoing ovarian surgery has not been demonstrated. In other words, it might be speculated that the reduction in ovarian volume in this context may represent a quantitative but not a qualitative injury to ovarian reserve. This hypothesis warrants confirmation.
Conclusion The present study supports the opinion that laparoscopic excision of ovarian cysts is associated with damage to ovarian reserve, at least immediately after surgery. Moreover, our results tend to rule out a role for an injury of ovarian vascularization in determining this effect. Several aspects remain to be clarified. Does the damage continue to progress after the first 3 months? What is the rate of gonads severely damaged? Is the reduction of ovarian reserve a consequence of surgery or is the damage already present before surgery? If the former, are there less-harmful alternative therapeutical approaches?
References 1. Sutton CJ, Ewen SP, Jacobs SA, et al. Laser laparoscopic surgery in the treatment of ovarian endometriomas. J Am Assoc Gynecol Laparosc. 1997;4:319-323. 2. Canis M, Rabischong B, Houlle C, et al. Laparoscopic management of adnexal masses: a gold standard? Curr Opin Obstet Gynecol. 2002; 14:423-428. 3. Chapron C, Vercellini P, Barakat H, et al. Management of ovarian endometriomas. Hum Reprod Update. 2002;8:591-597. 4. Chapron C, Fauconnier A, Goffinet F, et al. Laparoscopic surgery is not inherently dangerous for patients presenting with benign gynaecologic pathology: results of a meta-analysis. Hum Reprod. 2002;17: 1334-1342. 5. Nargund G, Cheng WC, Parsons J. The impact of ovarian cystectomy on ovarian response to stimulation during in-vitro fertilization cycles. Hum Reprod. 1996;11:81-83.
6. Loh FH, Tan AT, Kumar J, et al. Ovarian response after laparoscopic ovarian cystectomy for endometriotic cysts in 132 monitored cycles. Fertil Steril. 1999;72:316-321. 7. Ho HY, Lee RK, Hwu YM, et al. Poor response of ovaries with endometrioma previously treated with cystectomy to controlled ovarian hyperstimulation. J Assist Reprod Genet. 2002;19:507-511. 8. Somigliana E, Ragni G, Benedetti F, et al. Does laparoscopic excision of endometriotic ovarian cysts significantly affect ovarian reserve? Insights from IVF cycles. Hum Reprod. 2003;18:2450-2453. 9. Somigliana E, Ragni G, Benedetti F, et al. Does laparoscopic excision of ovarian cysts significantly affect ovarian reserve? Insights from in vitro fertilization cycles. Presented at the 59th Annual Meeting of the American Society for Reproductive Medicine, San Antonio, Texas, October 11–15, 2003. 10. Bukman A, Heineman M: Ovarian reserve testing and the use of prognostic models in patients with subfertility. Hum Reprod Update. 2001;7:581-590. 11. Tarlatzis B, Zepiridis L, Grimbizis G, et al. Clinical management of low ovarian response to stimulation for IVF: a systematic review. Hum Reprod Update. 2003;9:61-76. 12. Lass A. The fertility potential of women with a single ovary. Hum Reprod Update. 1999;5:546-550. 13. Chan C, Ng E, Li C, et al. Impaired ovarian blood flow and reduced antral follicle count following laparoscopic salpingectomy for ectopic pregnancy. Hum Reprod. 2003;18:2175-2180. 14. Ng E, Yeung W, Fong D, et al. Effects of age on hormonal and ultrasound markers of ovarian reserve in Chinese women with proven fertility. Hum Reprod. 2003;18:2169-2174. 15. Muzii L, Bianchi A, Croce C, et al. Laparoscopic excision of ovarian cysts: is the stripping technique a tissue-sparing procedure? Fertil Steril. 2002;77:609-614. 16. Lass A, Brinsden P. The role of ovarian volume in reproductive medicine. Hum Reprod Update. 1999;5:256-266. 17. Exacoustos C, Zupi E, Amadio A, et al. Laparoscopic removal of endometriomas: sonographic evaluation of residual functioning ovarian tissue. Am J Obstet Gynecol. 2004;191:68-72. 18. Zupi E, Exacoustos C, Szabolcs B, et al. Laparoscopic approach to dermoid cysts: combined surgical technique and ultrasonographic evaluation of residual functioning ovarian tissue. J Am Assoc Gynecol Laparosc. 2003;10:154-158. 19. Bancsi L, Broekmans F, Eijkemans M, et al. Predictors of poor ovarian response in in vitro fertilization: a prospective study comparing basal markers of ovarian reserve. Fertil Steril. 2002;77:328-336. 20. Wu MH, Tsai SJ, Hsiao KY, Chang FM. Three-dimensional Doppler imaging of ovarian stromal blood flow in women with endometriosis undergoing in vitro fertilization. Ultrasound Obstet Gynecol. 2003;21: 480-485. 21. Bath L, Wallace W, Shaw M, et al. Depletion of ovarian reserve in young women after treatment for cancer in childhood: detection by anti-Mullerian hormone, inhibin B and ovarian ultrasound. Hum Reprod. 2003;18:2368-2374. 22. Jones KD, Sutton CJG. Pregnancy rates following ablative laparoscopic surgery for endometriomas. Hum Reprod. 2002;17:782-785.