The impact of endometriosis, endometrioma and ovarian cystectomy on assisted reproductive technology

Abstract

Purpose

To assess outcomes in assisted reproductive technology (ART) in infertile women with endometriosis with respect to their concomitant endometrioma status and surgical history in our department.

Methods

This is a retrospective case control study which analyzes informational data obtained at a university hospital. The study drew from a patient pool of 332 cases (877 cIVF/ICSI cycles) that took place in our department from 2006 to 2008. Sixty‐one cases (97 cycles) had major indications for cIVF/ICSI with endometriosis. We classified groups from these 61 cases as follows: an unoperated endometrioma group (A) with 31 cycles, a postoperative endometrioma group (B) with 51 cycles, and a no endometrioma group (C) with 15 cycles. We analyzed and compared these three groups and also included a non‐endometriosis tubal infertility group (D) with 27 cycles.

Results

In the control group (D), serum FSH levels and the cancellation rates were significantly lower than those of other groups, and the number of developing follicles was higher. E2 levels before oocyte aspiration in the postoperative endometrioma group (B) was lower. Implantation, pregnancy, delivery and miscarriage rates were not significantly different among the four groups.

Conclusion

The results suggest that endometriosis causes a decrease in endocrinologic ovarian function whether or not an endometrioma is also present. As for E2 level before oocyte aspiration, our results suggest that ovarian reserves might be reduced by endometrioma excision, but this is difficult to evaluate. In the endometriosis groups, cancellation rates were significantly higher, although when embryos were transferred the pregnancy rates were not significantly different when compared with the non‐endometriosis group. As for infertile women with endometriomas, our results suggest that preexisting ovarian reserve is reduced by the presence of endometriosis, and ovarian reserve might also be reduced by excision of endometriomas.

Introduction

Endometriosis is diagnosed as the presence of endometrial‐like tissue that is organized in both glandular epithelium and stroma outside the uterus. Endometriosis is an estrogen‐dependent benign gynecological disorder that markedly decreases the quality of life in women of reproductive age due, in part, to the accompanying dysmenorrhea, chronic pelvic pain and dyspareuria. Concern about treatment for infertility in women with endometriosis is a common experience in the clinic.

There have been many reports on the association of endometriosis and sterility, and on the mechanism that causes the associated infertility. Some causes of infertility are ovarian dysfunction including premature ovarian failure and luteinized unruptured follicle, tubal infertility with associated pick up failure in the fimbriae of uterine tubes with pelvic adhesion, and the influence of peritoneal fluid [1, 2, 3, 4, 5]. Methods of treatment for infertility in women with endometriosis are medication, surgery, and assisted reproductive technology (ART). Medicinal treatments include gonadotropin‐releasing hormone (GnRH) agonist, danazol and progestin, but a meta‐analysis of five randomized controlled trials by Hughes et al. was not able to find an evident effect on pregnancy rates due to these medical treatments [6]. As for surgical treatment, there have been randomized controlled trials which have compared ablation of lesions and intraperitoneal irrigation to non‐treatment of infertile women with minimal and mild endometriosis. Opinion is divided as to whether surgery increases the pregnancy ratio or does not change it [7, 8, 9]. Many reports have found that ART is among the most effective treatments for patients with endometriosis‐associated infertility. However, infertile women with endometriosis who undergo conventional in vitro fertilization (cIVF) or intracytoplasmic sperm injection (ICSI) often respond with significantly decreased rates of fertilization, implantation rates, and number of oocytes retrieved, resulting in low pregnancy rates [10]. According to data on transfer of embryos derived from donor oocytes, implantation, fertilization and miscarriage are not affected by uterine conditions of stages III–IV endometriosis patients [11]. In fact, it was suggested that even severe endometriosis, while having an effect on the eggs, does not have an impact on the uterine environment.

In cases of infertility with endometriosis, and especially with endometrioma, it is controversial whether to first perform ART or first perform an operation. We assessed infertile women with endometriosis on whom we performed cIVF/ICSI in our department.

Materials and methods

Patient selection

Out of 332 cases corresponding to 877 cIVF/ICSI cycles in 2006–2008 in our department, 107 cases (32.2%) (300 cycles) were diagnosed with either endometriosis by laparoscopic surgery or with ovarian endometrioma identified by ultrasound or magnetic resonance imaging. We excluded cases exceeding four cIVF/ICSI cycles based on the tendency of the data to statistically significantly decrease the pregnancy rate after the third cycle attempt [12, 13, 14], and the cycles of GnRH agonist, buserelin acetate nasal spray used, including short or long protocols, as controlled ovarian hyper‐stimulation. Sixty‐one cases corresponding to 97 cycles had a major indication for ART with endometriosis. We classified these cases as follows: an unoperated endometrioma group (A) with 18 cases corresponding to 31 cycles (19 cycles of which were bilateral endometriomas); a postoperative endometrioma group (B) with 36 cases corresponding to 51 cycles (25 cycles bilateral endometriomas); and a no endometrioma group (C) (who were recognized with endometriosis by laparoscopy that had been carried out for other purposes) with 7 cases corresponding to 15 cycles. We compared cIVF/ICSI outcomes of these three groups and also included a fourth group (D) of 21 cases corresponding to 27 cycles as a control group with tubal infertility but without endometriosis (Fig. 1).

Figure 1.

Patient criteria flowchart. Out of 107 cases, 300 cycles diagnosed as endometriosis, we excluded cases exceeding four cIVF/ICSI cycles and controlled ovarian hyper‐stimulation cycles including short or long protocols. Sixty‐one cases corresponding to 97 cycles had a major indication for ART with endometriosis. We classified these as an unoperated endometrioma group A with 31 cycles, a postoperative endometrioma group B with 51 cycles, and a no endometrioma group C with 15 cycles. We compared these three groups and a fourth 27 cycle control group of non‐endometriosis tubal infertility (group D)

Background

Patient backgrounds at the time of laparoscopic surgery are presented in Table 1. We compared endometriosis groups A–C at laparoscopic surgery. In the unoperated endometrioma group (A), 5 cases, corresponding to 12 cycles out of 14 cases, had not become pregnant by cIVF/ICSI procedures and underwent laparoscopic surgery. Ages at surgery (mean ± SD) of the A–C groups were, respectively, 38.0 ± 2.8, 35.3 ± 4.2 and 38.3 ± 2.4 years old. There was no significant difference (N.S.) in age among groups A–C. Revised‐American Society for Reproductive Medicine (R‐ASRM) scores (median) for groups A–C were 72, 44, and 40 pts, respectively. Most of the endometrioma cases were severe (stages III and IV) endometriosis, therefore there was a significant difference in the endometrioma group with respect to the stage of endometriosis.

Table 1.

Background at laparoscopic surgery

	Endometriosis			p value
	Unoperated endometrioma group A a 12 cycles	Postoperative endometrioma group B 51 cycles	No Endometrioma group C 15 cycles
Ages at surgery [years old, mean ± SD(range)]	38.1 ± 2.6 (33–41)	35.5 ± 4.0 (29–41)	38.3 ± 2.4 (34–41)	N.S.
R‐ASRM scores [pts; median (range)]	72 (36–132)	44 (14–144)	40 (2–44)	0.025
Severe endometriosis (stage III and IV) [cycle (%)]	12/12 (100)	48/51 (94.1)	9/15 (60.0)	<0.01

We compared endometriosis groups A–C at laparoscopic surgery. There was no significant difference in ages at surgery. Cases in the endometrioma groups A and B had significantly higher R‐ASRM scores than did group C. Almost all cases in the endometrioma groups (A, B) were severe endometriosis (stages III, IV)

^aAn unoperated endometrioma group A: 5 cases, corresponding to 12 cycles out of 14 cases, had not become pregnant by conventional IVF/ICSI and underwent laparoscopic surgery

Backgrounds at cIVF/ICSI are shown Table 2. Ages at cIVF/ICSI (mean ± SD) of the 4 groups (A–D) were, respectively, 38.4 ± 3.0, 36.5 ± 3.4, 39.3 ± 2.1 and 36.2 ± 3.1 years old (N.S.). There were no significant differences in controlled ovarian stimulation, the number of cIVF/ICSI cycles and fertilization procedures, as presented in Table 2.

Table 2.

Background at conventional IVF/ICSI

	Endometriosis			Non‐endometriosis
	Unoperated endometrioma group A 31 cycles	Postoperative endometrioma group B 51 cycles	No endometrioma group C 15 cycles	Tubal infertility group D 27 cycles	p value
Ages at conventional IVF/ICSI [years old, mean ± SD (range)]	38.4 ± 3.0 (33–42)	36.5 ± 3.4 (30–42)	39.3 ± 2.1 (32–41)	36.2 ± 3.1 (27–40)	N.S.
Controlled ovarian stimulation [cycle (%)]
Spontaneous cycle	3 (9.7)	12 (23.5)	1 (6.7)	3 (11.1)	N.S.
Clomiphene citrate cycle	23 (74.2)	25 (49.0)	8 (53.3)	15 (55.6)
Clomiphene citrate—hMG/rFSH cycle a	5 (16.1)	14 (27.5)	6 (40.0)	9 (33.3)
Number of conventional IVF/ICSI cycles for a patient [cycle (%)]
First cycle	19 (61.3)	26 (51.0)	9 (60.0)	18 (66.7)	N.S.
Second cycle	8 (25.8)	16 (31.4)	4 (26.7)	7 (25.9)
Third cycle	4 (12.9)	9 (17.6)	2 (13.3)	2 (7.4)
Fertilization procedure [cycle (%)]
Conventional IVF	11 (35.5)	15 (29.4)	6 (40.0)	17 (63.0)	N.S.
ICSI	13 (41.9)	23 (45.1)	5 (33.3)	9 (33.3)
Empty follicle	7 (22.6)	13 (25.5)	4 (26.7)	1 (3.7)	N.S.

Background at IVF/ICSI is shown. There was no significant difference in ages at cIVF/ICSI. Controlled ovarian stimulation, number of cIVF/ICSI cycles and fertilization procedures show no significant difference

^aWe administered hMG or rFSH until 450 IU in total as the clomiphene citrate—hMG/rFSH cycle

Conventional IVF/ICSI procedure

When cIVF/ICSI was performed, we had patients come to our hospital at menstrual day 3 to undergo transvaginal ultrasonography. We examined serum LH, FSH and E2 levels to assess ovarian reserve. Clomiphene citrate cycles and clomiphene citrate—hMG or recombinant FSH (rFSH) cycles were used for controlled ovarian stimulation in our department. Briefly, we administered 50 mg clomiphene citrate (Clomid^®, Shionogi, Japan) twice daily for 5 days starting at day 3 of the menstruation cycle as the clomiphene citrate cycle, and additionally injected 150 IU hMG (HMG Fujiseiyaku^®, Fuji Pharma, Japan) or rFSH (Follistim^®, Schering‐Plough, Japan) from day 5 on alternate days until the patient received 450 IU in total as the clomiphene citrate—hMG/rFSH cycle. Controlled ovarian stimulation approaches were chosen depending on the ovarian reserve status and previous ART data. Approaches employed were: spontaneous cycle, clomiphene citrate cycle, clomiphene citrate—hMG/rFSH cycle and short or long protocols. In all cycles, we (1) administered 5,000 IU HCG (Pregnyl^®, Schering‐Plough, Japan) by intramuscular injection 2 days before oocyte aspiration, (2) aspirated follicles transvaginally, and (3) performed cIVF or ICSI procedures depending on spermatic findings and previous fertilization rates. Cleavage embryos at 2–3 days after fertilization were recognized and transferred to the patients’ uteruses.

Laparoscopic procedure

We performed surgery for all endometriosis cases with laparoscopy. In some cases, the GnRH agonist, Goserelin acetate (Zoladex^® 1.8 mg depot Kissei Pharmaceutical, Japan), was injected every 4 weeks for 3–6 months as a preoperative medication. The surgical procedure for endometriosis used in our department has previously been described elsewhere. Briefly, we routinely performed the 4‐puncture method under general anesthesia by endotracheal intubation in the lithotomy position [15, 16]. In cases with endometriomas, we excised ovarian cyst walls—the so‐called “striping method”—and sewed purse‐string sutures with Vicryl^® 2‐0 (Ethicon, Japan) in consideration of potential damage to the ovarian reserve. In cases with peritoneal lesions and adhesion, we performed excision, ablation and adhesiotomy by whatever means were appropriate and possible for the given case. We confirmed bilateral tubal patency with 20 mg Indigocarmine (Daiichi Sankyo, Japan) and subsequently irrigated intraperitoneally. Depending on the intraperitoneal circumstances, fibrin glue was coated to prevent re‐adhesion.

Statistical analysis

Comparisons between the two groups were made using Student's t test, the Mann–Whitney U test, or Fischer's exact test, as appropriate, and comparisons among the 3 or 4 groups were made using the Kruskal–Wallis test or Fischer's exact test, with p < 0.05 as the level of statistical significance.

Results

Outcomes of cIVF/ICSI are presented in Table 3. Serum FSH levels at menstrual day 3 (mean ± SD) of the 4 groups (A–D) were, respectively, 10.5 ± 4.8, 11.2 ± 5.8, 10.9 ± 5.1, and 7.7 ± 2.0 mIU/ml. FSH levels in control group D were significantly lower than the levels in the other groups (p < 0.01). Serum E2 levels before oocyte aspiration were 864 ± 624, 661 ± 631, 778 ± 300, and 861 ± 419 pg/ml. E2 levels in the postoperative endometrioma group (B) were significantly lower (p < 0.01). The number of developing follicles in the control group (D) was higher (p = 0.034), but the number of retrieved, fertilized eggs and high grade embryo rates (Grade I or II as per the Veeck scale) showed no significant differences. Cancellation rate of embryo transfer (including cases of empty follicle, degenerating egg and unfertilized egg) were, respectively, 35.5, 47.1, 40.0 and 14.8%. Endometriosis groups (A, B and C) had significantly higher cancellation rates (p = 0.036). Implantation, pregnancy, delivery and miscarriage rates were not significantly different.

Table 3.

Outcomes of conventional IVF/ICSI cycles

	Endometriosis			Non‐endometriosis
	Unoperated endometrioma group A 31 cycles	Postoperative Endometrioma group B 51 cycles	No endometrioma group C 15 cycles	Tubal infertility group D 27 cycles	p value
Serum FSH level at menstrual day 3 (mIU/ml; mean ± SD)	10.5 ± 4.8	11.2 ± 5.8	10.9 ± 5.1	7.7 ± 2.0	<0.01
Serum E2 level before oocytes aspiration (pg/ml; mean ± SD)	864 ± 624	661 ± 631	778 ± 300	861 ± 419	<0.01
The number of developing follicles/cycle (mean ± SD)	2.6 ± 2.2	2.6 ± 1.6	2.6 ± 0.9	4.1 ± 2.5	0.034
The number of oocytes retrieved/cycle (mean ± SD)	1.6 ± 2.1	1.2 ± 1.1	1.3 ± 1.0	1.8 ± 1.3	N.S.
The numbers of fertilized eggs/cycle (mean ± SD)	1.6 ± 1.4	1.1 ± 0.9	1.0 ± 0.8	1.5 ± 1.0	N.S.
The high grade embryo rate/cycle (%) a	29/50 (58.0)	29/56 (51.8)	11/15 (73.3)	24/41 (58.5)	N.S.
The cancellation rate of embryo transfer/cycle (%) b	11/31 (35.5)	24/51 (47.1)	6/15 (40.0)	4/27 (14.8)	0.036
Implantation rate/cycle (%)	6/31 (19.4)	8/51 (15.7)	4/15 (26.7)	7/27 (25.9)	N.S.
Pregnancy rate/transfer (%)	6/20 (30.0)	8/27 (29.6)	4/9 (44.4)	7/23 (30.4)	N.S.
Delivery rate/transfer (%)	5/20 (25.0)	6/27 (22.2)	3/9 (33.3)	5/23 (21.7)	N.S.
Miscarriage rate/pregnancy (%)	1/6 (16.7)	2/8 (25.0)	1/4 (25.0)	2/7 (28.6)	N.S.

Serum FSH levels in control group D were significantly lower. Serum E2 levels before oocytes aspiration in group B were lower. The number of developing follicles in control group D was higher, but the number of retrieved, fertilized eggs and the high grade embryo rates showed no significant difference. Endometriosis groups A–C had significantly higher cancellation rates. Implantation, pregnancy, delivery and miscarriage rates showed no significant difference

^aGrade I or II with Veeck scale

^bCancellation rates of embryo transfer including cases of empty follicle, degenerating and unfertilized egg

Discussion

It currently is the dominant opinion that removal of endometriomas before IVF does not improve fertility outcomes [17, 18]. Postsurgical ovarian failure is possible after laparoscopic excision of bilateral endometriomas [19, 20]. Therefore, we have to pay great attention to infertility cases with bilateral endometriomas. However, there is a report that the presence of ovarian endometrioma is associated with a reduced fecundity which includes responsiveness to gonadotropins and the number of oocytes retrieved [21].

In cIVF/ICSI performed in our department, FSH levels at day 3 in endometriosis cases were significantly higher than levels in the control group (D). Such a finding suggests that endometriosis causes a decrease in endocrinologic ovarian function whether or not endometrioma is also present. Various physiologically active substances, including lymphocytes, macrophages and cytokines are increased in the peritoneal fluid of endometriosis patients, resulting in peritoneal inflammatory cascade which may in part be responsible for endometriosis‐associated infertility [22, 23]. As for E2 levels before oocyte aspiration, our results suggest that ovarian reserve might be reduced by excision of endometriomas, but this is difficult to evaluate. The cancellation rates of embryo transfer with endometriosis, despite surgery, were significantly higher than in the control group (D), but if embryos were able to be transferred, the pregnancy rate was not significantly different when compared to the non‐endometriosis group. Matalliotakis et al. [24] reported similar outcomes—i.e. that embryo quality and uterine receptivity remain unaffected despite diminished ovarian reserve. On the basis of these results, in cases of infertility, the patient's age and ovarian reserve, as well as endometriosis status must be taken in to consideration when we have to decide whether to first perform laparoscopy or proceed with ART. In our department, our first choice is usually laparoscopic surgery for sterile women with endometriomas of more than 5 cm in diameter which takes in to consideration the risk of rupture, infection and malignant alteration.

Acknowledgments

We express profound gratitude for discussion with the late professor Hiroyuki Takeuchi, and pay our heartfelt respects to him. And we thank K. Itagaki and M. Machida for technical assistance with ART.