Introduction

A causal relationship between adenomyosis and infertility has been repeatedly suggested [1, 2]. With women delaying their first pregnancy until their late 30s or early 40s, the diagnosis of adenomyosis in infertile patients is becoming increasingly frequent. The presence of adenomyosis may impair fertility by affecting the utero-tubal transport and by altering endometrial function and receptivity [2]. Moreover, several anomalies found in the junctional zone in patients with adenomyosis have been associated with poor reproductive performance, mainly due to perturbed uterine peristalsis. These patients also present with an altered decidualization and abnormally high levels of free radicals in the uterine milieu [1].

Accordingly, impaired fertility outcomes after ART have been reported in two systematic reviews [3, 4]. In one of them, the presence of adenomyosis was associated with a 28% reduction in the likelihood of clinical pregnancy in infertile women undergoing IVF/ICSI [3]. Moreover, the risk of miscarriage among these patients was more than doubled compared to controls. These results are in line with those of a more recent meta-analysis [4], in which adenomyosis resulted to be associated with a 41% decrease in live birth rate after IVF (OR 0.59, 95% CI 0.42–0.82), as well as a more than doubled risk of miscarriage (OR 2.2, 95% CI 1.53–3.15). However, reviews and meta-analyses on this subject are potentially biased by the intrinsic limitations of published studies. Younes and Tulandi [4] considered that studies were heterogeneous, with differences in women’s age, duration of infertility, type of downregulation protocol used, number and quality of transferred embryos, and number of IVF cycles performed. Moreover, clinical outcomes, modality of infertility diagnosis, and the rate of concomitant endometriosis differed among studies, and some authors made no adjustment for confounding factors. Following the publication of Younes and Tulandi’s review, which highlighted the importance of overcoming such limitations, further studies evaluating the reproductive outcome of women with adenomyosis undergoing either ART [5,6,7,8, 9••, 10, 11, 12••, 13,14,15,16] or conservative surgical treatment [17,18,19,20] have been published. Moreover, standardized and reproducible ultrasonographic and MRI diagnostic criteria for adenomyosis have been established [21, 22].

Given the availability of new and recent studies evaluating this issue, we sought to carry out a narrative review investigating the reproductive outcome of infertile women with adenomyosis who undergo ART or fertility-sparing surgical treatment.

Materials and Methods

An electronic database search on PubMed was performed with the objective of identifying all studies written in English and published between 2011 and 2021. We used the following combinations of terms such as “adenomyosis” AND “infertility”, “adenomyosis” AND “IVF”, and “adenomyosis” AND “conservative surgery” and found 442 articles. Three reviewers (AD, FF, and DD) independently reviewed titles, abstracts, and full article texts to identify eligible studies. Discrepancies were resolved by discussion. A total of 408 studies were excluded for the following reasons: 388 were deemed as not pertinent, while the remaining 20 lacked a control group, did not include proper outcomes, or were meta-analyses. The flow diagram of the literature search results is shown in Fig. 1.

Fig. 1
figure 1

Flow diagram of study identification and selection

Full size image

Thirty-four studies were considered eligible for the present review. Among these, 23 evaluated the outcome of ART in infertile women with adenomyosis [5,6,7,8, 9••, 10, 11, 12••, 13,14,15,16, 23,24,25,26,27,28,29,30,31,32,33], and 11 evaluated the reproductive outcome of women with adenomyosis who had previously undergone conservative surgical treatment [17,18,19,20, 34,35,36,37,38,39,40]. Adenomyosis was diagnosed before conception in all studies. Diagnosis was performed by the means of transvaginal ultrasonography (TVUS) alone in 24 studies; 2D TVUS was used in 22 studies [5, 6, 8, 9••, 12••, 13, 16, 17, 19, 23,24,25,26,27,28,29,30,31, 36,37,38, 40], while 3D TVUS was used in two [10, 32]. In the remaining 10 studies, adenomyosis was diagnosed by means of TVUS and/or MRI [7, 11, 14, 15, 18, 20, 33,34,35, 39].

ART Studies

Five studies were prospective [9••, 10, 26, 29, 32], while the remaining 18 were retrospective [5,6,7,8, 11, 12••, 13,14,15,16, 23,24,25, 27, 28, 30, 31, 38]. A total of 22 studies evaluated autologous IVF/ICSI cycles, whereas the outcome of oocyte donation (OD) cycles was assessed in one study [24].

Women with adenomyosis were compared with women without adenomyosis in 12 articles [9••, 10, 13, 15, 16, 23, 24, 26, 27, 29, 30, 32]. In one study, patients with adenomyosis were compared with both women without adenomyosis and women with co-existent adenomyosis and endometriosis [24]. In Sharma et al.’s study, women with adenomyosis only and those with both adenomyosis and endometriosis were separately evaluated, and controls included women with endometriosis without adenomyosis as well as women with tubal infertility [16].

In four studies, women with adenomyosis who were pre-treated with GnRH agonists [6, 8, 28, 31] were compared with those not pre-treated, while in one study, pre-treatment with 52-mg 20-mg/d levonorgestrel-releasing IUD (LNG-IUD) was compared with no pre-treatment [14]. In Niu et al.’s study, women chose either treatment with GnRH agonists + HRT or a pure HRT protocol before frozen-thawed embryo transfer (FET) depending on their own preference [28]. Conversely, Park et al. compared the outcome of fresh embryo transfer (ET) cycles with or without pre-treatment with GnRH agonists with FET cycles with GnRH agonist pre-treatment [31]. In three further studies, all women were pre-treated with GnRH agonists [5, 12••, 16]. Three other studies reported data from subgroups of women undergoing pre-treatment with GnRH agonists [7, 15, 33]. Two studies compared ART outcomes in women with adenomyosis receiving a long vs ultra-long GnRH agonist protocol [9••, 11].

One study compared patients with adenomyosis and CA125 < 35 IU/ml with those with adenomyosis and CA125 > 35 IU/mL [33]. Li et al. compared women with adenomyosis who had a uterine volume < 99 cc with those with adenomyosis and a uterine volume > 99 cc [5]. Another study compared different values of myometrial thickness: < 2.00 cm, 2.00–2.49 cm, and ≥ 2.50 cm [25]. Iwasawa et al. compared women with different location and extension of adenomyosis within the myometrium, distinguishing between advanced (full thickness), extrinsic (located on the serosal side), and intrinsic (located on the endometrial side) adenomyosis [7]. Lastly, three studies compared outcomes in women with focal vs diffuse adenomyosis [11, 29, 31].

One study compared patients with adenomyosis with those with endometriosis [12••]. Only 11 studies reported the presence of endometriosis in enrolled women [5, 7, 8, 11, 12••, 15, 16, 23, 24, 30, 32].

Surgical Treatment Studies

Eleven studies evaluated the reproductive outcome of women with adenomyosis who had previously undergone conservative surgical treatment [17,18,19,20, 34,35,36,37,38,39,40]. Six studies were retrospective [17,18,19, 34, 38, 39], and five were prospective [20, 35,36,37, 40].

Ten studies were surgical series, without a control group of women not undergoing surgery [17,18,19,20, 35,36,37,38,39,40]. In three of these ten studies, surgical treatment was followed by either 3 months [19] or 6 months of GnRH agonist treatment [36, 38]. The remaining study compared surgical treatment with a 6-month medical treatment with GnRH agonists [34].

Four studies only included naturally conceived pregnancies [20, 34, 36, 37], while seven studies included both pregnancies from natural conception and IVF [17,18,19, 35, 38,39,40].

Five studies included both women with focal and diffuse adenomyosis [17,18,19, 37, 40], three studies included women with adenomyoma only [34, 36, 39], while three further studies included women with diffuse adenomyosis only [20, 35, 38]. In one study, all enrolled women had both adenomyosis and endometriosis [17].

Surgery was performed at laparotomy in seven studies [20, 34,35,36,37,38, 40] and at laparoscopy in two studies [17, 39], while in two studies, focal adenomyosis was treated at laparoscopy and diffuse adenomyosis at laparotomy [18, 19]. Surgical complications were reported in four studies. Two groups of authors described radical adenomyomectomy using a triple-flap technique [20, 35], and three performed a radical excision of adenomyosis leaving a 1-cm margin of tissue above the endometrium and a 1-cm margin of tissue below the serosal surface [18, 20, 35].

The weight of the removed lesions was specified in seven studies [18,19,20, 35, 36, 38, 39], and the size of the removed lesions was specified in three studies [19, 34, 37]. Eight studies evaluated the mode of delivery [18,19,20, 34, 35, 38,39,40]. Peripartum complications were reported in six studies [17, 18, 34, 36, 39, 40].

Outcomes

Implantation Rate

In ART studies, implantation rate was defined as the number of gestational sacs detected by the means of TVUS divided by the number of transferred embryos [5, 7, 9••, 11, 12••, 14, 23, 26, 29, 30, 33].

Clinical Pregnancy Rate

Clinical pregnancy rate was defined as clinical pregnancies/cycles in all ART studies. In the studies analyzing surgical treatment, clinical pregnancy rate was defined as clinical pregnancies/patients. Clinical pregnancy was defined as the detection of an intrauterine gestational sac by the means of TVUS [5,6,7,8, 10, 11, 29, 30] or as the ultrasonographic evidence of an embryo with a heartbeat either 4 weeks [16, 25], 5 weeks [12••], 6 weeks [15, 18, 26, 32, 33], 7 weeks [23, 28], 8 weeks [9••, 14, 27], or 12 weeks [17, 39] after embryo transfer.

Miscarriage Rate

Miscarriage was defined as the loss of a clinical pregnancy before 12 weeks [8, 14, 23, 26], 20 weeks [10, 16, 40], 22 weeks [6], 24 weeks [30], or 28 weeks [9••, 38] or regardless of gestational age [7, 10, 12••, 24, 25, 32, 34,35,36, 39]. Thalluri and Tremellen also included ectopic pregnancies [27]. One study defined miscarriage as the loss of a clinical pregnancy before 24 weeks, distinguishing between early miscarriage when it occurred before 12 weeks and late miscarriage when it occurred between 12 and 24 weeks [5]; similarly, Lan et al. defined early miscarriage as that occurring before 12 weeks and late miscarriage as that occurring before 24 weeks [11]. One study defined miscarriage as the ultrasonographic detection of an intrauterine gestational sac in the absence of fetal heartbeat [15].

Live Birth Rate

Live birth rate was defined as the delivery of an alive fetus/clinical pregnancy following 20 weeks [23], 22 weeks [6], 24 weeks [5], and 26 weeks of gestation [16] or regardless of gestational age [7, 8, 9••, 10, 11, 12••, 15, 24, 25, 27, 29,30,31,32, 34,35,36]. Four studies defined an ongoing pregnancy as a clinical pregnancy continuing past 12 weeks of gestation [14, 26, 28, 37]. One study evaluated delivery rate [30].

Among the studies analyzing ART results, only one reported mode of delivery and obstetric outcomes [7]. Conversely, of the 11 surgical studies, eight reported mode of delivery [18,19,20, 34, 35, 38,39,40], three reported abnormal placentation [17, 34, 39], and one reported uterine rupture during pregnancy [40].

Results

ART Studies

A total of 3791 women with adenomyosis and 6841 without adenomyosis were included. The findings of studies including women who conceived by means of ART are reported in Table 1. Mean (± SD) age was 36.3 ± 1.49 in women with adenomyosis and 33.5 ± 1 in those without adenomyosis.

Table 1 Reproductive outcome in women with and without adenomyosis who underwent ART
Full size table

Implantation Rate

Thirteen out of 23 studies evaluated this outcome. Overall, mean (± SD) implantation rate was 30.0% ± 7.5% in women with adenomyosis and 31.6% ± 12.1% in women without adenomyosis.

Among studies comparing women with adenomyosis and patients without adenomyosis, two out of six described a reduced implantation rate in the former group [9••, 26]. In particular, Salim et al. observed a significantly lower implantation rate in adenomyosis vs controls (18.8% vs 29.4%) [26]; similarly, Hou et al. found a significantly lower implantation rate when comparing patients with adenomyosis undergoing ultra-long or long GnRH agonist protocols versus controls (43.5% and 36.9% vs 49.5%) [9••]. No difference in implantation rate was observed between adenomyosis and controls by Costello et al. (28.3% vs 31.6%) [23], Zhang et al. (28.4% vs 31.7%) [13], and Benaglia et al. (32% vs 21%) [29]. Martinez-Conejero et al. found no significant difference in implantation rate when comparing women with adenomyosis both with controls (29.6% vs 30.8%) and with women with endometriosis plus adenomyosis (29.6% vs 33.3%) [24].

Among women with adenomyosis, a significantly higher implantation rate was observed by Niu et al. in those pre-treated as compared to those not pre-treated with GnRH agonists (32.5% vs 16.1%) [28]. Liang et al. reported a significantly higher implantation rate among women with adenomyosis pre-treated with LNG-IUD, compared to those not pre-treated (32.1% vs 22.1%) [14].

Among women with adenomyosis undergoing an ultra-long vs long protocol with GnRH agonists, Lan et al. observed no difference in implantation rate between the two groups (36.2% vs 30.2%) [11].

Huang et al. found no significant difference in implantation rates in women with adenomyosis when comparing those with levels of CA125 < 35 IU/ml and those with levels > 35 IU/mL (28.4% vs 22.9%) [33]. In Youm et al.’s study, a significantly lower implantation rate was observed in women with a myometrial thickness > 2.49 cm, compared to those whose myometrial thickness was < 2.0 cm and 2.0–2.49 cm, respectively (12.3% vs 22.8% and 12.3% vs 21.9%). Among women whose myometrial thickness ranged between 2.0 and 2.49 cm, the implantation rate was significantly higher when sonographic criteria for adenomyosis were absent rather than present (27.2% vs 12.4%) [25]. Analyzing women with adenomyosis pre-treated with GnRH agonists, Li et al. found no difference in implantation rates between those with a uterine volume < 98.81 cc and those with a uterine volume > 98.81 cc (43.3% vs 40%) [5]. Benaglia et al. found no difference in implantation rates when comparing focal and diffuse adenomyosis (32% vs 32%) [29].

Comparing women with adenomyosis to those with endometriosis, both pre-treated with GnRH agonists, Zhang et al. found a significantly lower implantation rate in the former group (31.9% vs 46.7%) [12••].

Clinical Pregnancy Rate

All 23 ART studies evaluated this outcome. Overall, mean (± SD) clinical pregnancy rate/cycle was 39.4% ± 12.8% in women with adenomyosis and 49.8% ± 16.5% in women without adenomyosis.

Among studies comparing patients with adenomyosis and controls, six out of 12 described a reduced pregnancy rate in the former group [9••, 15, 16, 26, 27, 32]. In particular, a significantly lower pregnancy rate in women with adenomyosis vs controls was observed by Salim et al. (22.2% vs 47.2%) [26], Thalluri and Tremellen (23.6% vs 44.6%) [27], Mavrelos et al. (29.2% vs 42.6%) [32], and Stanekova et al. (55.9% vs 84.6%) [15]. A significantly lower pregnancy rate compared to controls was observed by Hou et al. in women with adenomyosis undergoing an ultra-long (63.8% vs 68.4%) or a long GnRH agonist protocol (50.5% vs 68.4%) [9••]. Sharma et al. reported a significantly lower pregnancy rate in women with both adenomyosis and endometriosis compared to patients with tubal infertility (controls versus adenomyosis plus endometriosis: 34.5% vs 22.7%), as in both the adenomyosis group and the adenomyosis plus endometriosis group, compared to the endometriosis only group (23.4% vs 36.6% and 22.7% vs 36.6%). However, they did not observe a lower pregnancy rate in the adenomyosis group compared to the tubal infertility group (23.4% vs 34.5%) [16]. Conversely, no difference in pregnancy rate was observed between women with adenomyosis and controls in studies by Costello et al. (35.1% vs 31.1%) [23], Benaglia et al. (43% vs 29%) [29], Yan et al. (36.4% vs 45.5%) [30], Neal et al. (80% vs 75%) [10], and Zhang et al. (42.2% vs 42.8%) [13]. Martinez-Conejero et al. found no significant difference in pregnancy rates when comparing patients with adenomyosis versus controls (40% vs 44.4%) and versus women with endometriosis plus adenomyosis (40% vs 44.2%) [24].

Among women with adenomyosis, Niu et al. observed a significantly higher clinical pregnancy rate in those pre-treated with GnRH agonists as compared to those not pre-treated (51.3% vs 24.8%) [28], whereas no difference was observed between the two groups by Chen et al. (30.8% vs 42.6%) [8] and Li et al. (40.6% vs 42.5%) [6]. Park et al. observed no difference when comparing clinical pregnancy rate in not pre-treated fresh ET (25.2%) with both pre-treated fresh ET (30.5%) and pre-treated frozen ET (39.5%) [31]. Liang et al. found a significantly higher pregnancy rate among women with adenomyosis who were pre-treated with LNG-IUD as compared to not pre-treated women (44% vs 33.5%) [14].

Among patients with adenomyosis undergoing an ultra-long vs long protocol with GnRH agonists, a significantly higher clinical pregnancy rate for the former group was observed by Hou et al. (63.8% and 50.5%) [9••], whereas no difference was observed by Lan et al. (52.6% vs 43%) [11]. However, in Lan et al.’s study, the ultra-long protocol was associated with a significantly higher rate of clinical pregnancy in women with diffuse adenomyosis (55.5% vs 37.9%) [11].

When comparing levels of CA125 < 35 IU/ml with levels > 35 IU/mL in women with adenomyosis, Huang et al. found no significant difference in pregnancy rate (35.7% vs 31.8%). Also, no significant difference was found when comparing women who were pre-treated with GnRH agonists with those not pre-treated (32.2% vs 32.3%) [33]. In Youm et al.’s study, a significantly lower pregnancy rate was observed in women with a myometrial thickness > 2.49 cm compared to women with myometrial thickness < 2.0 cm and 2–2.49 cm, respectively (31.5% vs 56.4% and 31.5% vs 53.1%). In women with a myometrial thickness ranging between 2.0 and 2.49 cm, the clinical pregnancy rate was significantly higher when sonographic criteria for adenomyosis were absent rather than present (63.5% vs 34.5%) [25]. When analyzing women with adenomyosis pre-treated with GnRH agonists, Li et al. described a comparable clinical pregnancy rate between those with a uterine volume < 98.81 cc and those with a uterine volume > 98.81 cc (51.8% vs 52%) [5]. Iwasawa et al. observed comparable pregnancy rates between women with advanced, extrinsic, and intrinsic adenomyosis (25% vs 33.3% vs 22.2%, respectively) [7]. Comparable pregnancy rates between women with diffuse and focal adenomyosis were observed by Benaglia et al. (46% vs 40%) [29] and Park et al. (43.5% vs 35%) [31]. In the latter study, women were pre-treated with GnRH agonists.

Comparing patients with adenomyosis to those with endometriosis, both pre-treated with GnRH agonists, Zhang et al. found a significantly lower clinical pregnancy rate in the former group (47.1% vs 64.4%) [12••].

Miscarriage Rate

Twenty-two out of 23 studies evaluated this outcome [5,6,7,8, 9••, 10, 11, 12••, 13,14,15,16, 23,24,25,26,27, 29,30,31,32,33]. Overall, the mean (± SD) miscarriage rate was 26.3% ± 16.9% in women with adenomyosis and 15.3% ± 10.6% in women without adenomyosis.

Among studies comparing women with adenomyosis with those without adenomyosis, six out of 12 described an increased miscarriage rate in the former group [9••, 13, 15, 16, 24, 26]. In particular, when comparing adenomyosis vs controls, Salim et al. observed a significantly higher miscarriage rate (50% vs 2.8%) [26], Zhang et al. reported a significantly higher rate of early miscarriage (13.3% vs 5.6%) and a comparable rate of late miscarriage (5.0% vs 2.2%) [13], while Stanekova et al. observed a significantly higher rate of early miscarriage, including miscarriage at a biochemical stage (53% vs 19.7%) [15]. A significant increase in miscarriage rate was observed by Martinez-Conejero et al. when comparing women with adenomyosis both versus controls (13.1% vs 7.2%) and versus women with endometriosis plus adenomyosis (13.1% vs 6.1%) [24]. Sharma et al. reported a significantly higher miscarriage rate in both the adenomyosis group and the adenomyosis plus endometriosis group compared to the tubal infertility group (40% vs 13% and 35% vs 13%); a higher miscarriage rate was also observed in the adenomyosis group and in the adenomyosis plus endometriosis group, compared to the endometriosis group (40% vs 14.6% and 35% vs 14.6%, respectively) [16].

A significantly higher miscarriage rate when comparing patients with adenomyosis undergoing an ultra-long or long GnRH agonist protocol and controls was observed by Hou et al. (17.4% and 25.5% vs 10.4%) [9••]. No difference in miscarriage rates between adenomyosis and controls was observed by Costello et al. (15.4% vs 27.1%) [23], Thalluri and Tremellen (25% vs 10.3%) [27], Benaglia et al. (19% vs 36%) [29], Yan et al. (50% vs 36.9%) [30], Mavrelos et al. (4.8% vs 16.3%) [32], and Neal et al. (10.5% vs 7.7%) [10].

Among women with adenomyosis, a significantly lower miscarriage rate was observed in those pre-treated as compared to those not pre-treated with GnRH agonists by Stanekova et al. (82.4% vs 35.7%) [15], whereas no difference was observed between the two groups by Chen et al. (31.3% vs 11.6%) [8], Li et al. (41.5% vs 44.2%) [6], and Huang et al. (26.3% vs 33.3%) [33]. Park et al. observed no difference comparing miscarriage rates in not pre-treated fresh ET (6.1%) vs both pre-treated fresh ET (9.5%) and pre-treated frozen ET (13.9%) [31]. Liang et al. observed a comparable miscarriage rate between patients with adenomyosis who were pre-treated with LNG-IUD and those not pre-treated (3.4% vs 9.3%) [14].

Among women with adenomyosis undergoing an ultra-long vs long protocol with GnRH agonists, a significantly lower early miscarriage rate for the former group was observed by Lan et al. (12% vs 26.5%). In the same study, the late miscarriage rate was not significantly different between the two groups (8.0% vs 2.9%) nor was it significantly different between diffuse and focal adenomyosis [11].

Analyzing women with adenomyosis, Huang et al. found no significant difference in miscarriage rates when comparing women with levels of CA125 < 35 IU/ml and those with CA125 > 35 IU/mL (20% vs 35.7%) [33]. In Youm et al.’s study, a significantly higher miscarriage rate was observed in women with a myometrial thickness > 2.49 cm, compared to women with a myometrial thickness < 2.0 cm and 2.0–2.49 cm, respectively (52.2% vs 12.9% and 52.2% vs 20.9%). Among women with a myometrial thickness ranging between 2.0 and 2.49 cm, the miscarriage rate was significantly lower when sonographic criteria for adenomyosis were absent rather than present (12.1% vs 50%) [25].

When analyzing women pre-treated with GnRH agonists, Li et al. observed a higher rate of both miscarriage and early miscarriage (< 12 weeks) in those with a bulky uterus > 98.81 cc, compared to those with a uterine volume < 98.81 cc (51.2% vs 16.3% and 38.5% vs 13.9%, respectively) [5]. Iwasawa et al. observed comparable miscarriage rates between women with advanced, extrinsic, and intrinsic adenomyosis (64% vs 33.3% vs 50%, respectively) [7]. In Park et al.’s study, among pre-treated women, miscarriage rate did not differ between those with diffuse and focal adenomyosis (14.3% vs 15% and 7.1% vs 13%, respectively) [31].

Comparing women with adenomyosis to those with endometriosis, both pre-treated with GnRH agonists, Zhang et al. found a significantly higher miscarriage rate in the former group (33.3% vs 13.4%) [12••].

Live Birth Rate

This outcome was evaluated in 20 out of 23 studies [5,6,7,8, 9••, 10, 11, 12••, 13,14,15,16, 23,24,25,26, 28,29,30, 33]. Overall, the mean (± SD) live birth rate was 29.9% ± 13.7% in women with adenomyosis and 42.6% ± 18.2% in women without adenomyosis. Seven out of 10 studies comparing women with adenomyosis and women without adenomyosis described a reduced live birth rate in the former group [9••, 13, 15, 16, 24, 26, 30]. In particular, a significantly lower live birth rate in the adenomyosis vs control group was observed by Yan et al. (24.8% vs 33.3%) [30], Stanekova et al. (47% vs 80%) [15], and Zhang et al. (22.8% vs 33.3%) [13]. A significantly lower live birth rate in adenomyosis vs controls was observed by Salim et al. (11.1% vs 45.9%) [26]. A significantly lower live birth rate was observed by Hou et al. in women with adenomyosis undergoing an ultra-long (52.4% vs 58.5%) or a long GnRH agonist protocol compared to controls (37.6% vs 58.5%) [9••]. Sharma et al. reported a significantly lower live birth rate in both the adenomyosis group and in the adenomyosis plus endometriosis group, compared to the tubal infertility group (12.5% vs 27.5% and 11.4% vs 27.5%, respectively). Similar results were reported in both the adenomyosis group and the adenomyosis plus endometriosis group, compared to the endometriosis group (12.5% vs 26.5% and 11.4% vs 26.5%, respectively) [16]. A significant decrease in live birth rate was observed by Martinez-Conejero et al. when comparing women with adenomyosis with controls (26.8% vs 37.1%) and with women with endometriosis plus adenomyosis (26.8% vs 38%) [24]. No difference in live birth rate was observed between adenomyosis and controls by Costello et al. (29.7% vs 26.1%) [23], Benaglia et al. (35% vs 18%) [29], and Neal et al. (69.5% vs 66.5%) [10].

A significantly higher live birth rate was observed in women with adenomyosis who were pre-treated with GnRH agonists as compared to those not pre-treated by Niu et al. (48.9% vs 21.4%) [28], whereas no difference was observed by Li et al. (23.7% vs 23.7%) [6]. Interestingly, Chen et al. reported a higher live birth rate in women with adenomyosis not pre-treated with GnRH agonists as compared to those pre-treated (37.7% vs 21.2%) [8]. Liang et al. found a significantly higher live birth rate among women with adenomyosis pre-treated with LNG-IUD compared to not pre-treated women (41.8% vs 29.5%) [14].

Lan et al. observed no difference in live birth rate when comparing women with adenomyosis undergoing an ultra-long vs long protocol with GnRH agonists (41.6% vs 30.4%). However, the ultra-long protocol was associated with a significantly higher rate of live birth in women with diffuse adenomyosis (43.4% vs 25.9%), whereas no difference was observed in women with focal adenomyosis [11].

Huang et al. found no significant differences in live birth rate between women with serum levels of CA125 < 35 IU/ml and those with levels > 35 IU/mL (28.6% vs 20.5%). Similarly, no differences were observed when comparing women pre-treated with GnRH agonists with those not pre-treated [33]. Youm et al. observed a significantly lower live birth rate in women with myometrial thickness > 2.49 cm compared to women with a myometrial thickness < 2.0 cm and 2.0–24.9 cm, respectively (15.1% vs 46.9% and 15.1% vs 40.7%). Among women with a myometrial thickness ranging between 2.0 and 2.49 cm, live birth rate was significantly higher when sonographic criteria for adenomyosis were absent rather than present (53.8% vs 17.2%) [25].

Analyzing women with adenomyosis pre-treated with GnRH agonists, Li et al. described a higher live birth rate in those with a uterine volume < 98.81 cc compared to those with a uterine volume > 98.81 cc (43.4 vs 25.3%) [5]. Iwasawa et al. observed a comparable live birth rates between women with advanced, extrinsic, and intrinsic adenomyosis (9% vs 22.2% vs 11.1%, respectively) [7]. Benaglia et al. observed comparable live birth rates between women with diffuse versus focal adenomyosis (33% vs 36%) [29].

Comparing women with adenomyosis to those with endometriosis, both pre-treated with GnRH agonists, Zhang et al. found a significantly lower live birth rate in the former group (31.4% vs 54.8%) [12••].

Mode of Delivery and Abnormal Placentation

In a series of 16 women with adenomyosis conceiving by ART, the rate of cesarean section was 62.5%, the rate of preterm delivery was 18.8%, and the rate placenta previa was 6.2% [7].

Surgical Treatment Studies

The reproductive outcome of a total of 961 women who previously underwent surgical treatment of adenomyosis was evaluated in 11 studies [17,18,19,20, 34,35,36,37,38,39,40]. The findings of patients who underwent conservative surgery are reported in Table 2.

Table 2 Reproductive outcome of women with and without adenomyosis who underwent conservative surgical treatment
Full size table

The mean weight of the removed lesions was 133.3 g ± 194.3 g [18,19,20, 35, 36, 38, 39].

Surgical complications included small, self-reabsorbing myometrial hematomas in 5.8% of women [35], Asherman’s syndrome in 3.9% (4/103) [40], and need for postoperative blood transfusion in 10% (4/40) [34] and 1% (1/102) [39], respectively. One study reported an 8.9% rate of conversion from laparoscopy to laparotomy [39]

Clinical Pregnancy Rate

Overall, the pregnancy rate was 38.1% (range 2.3–67.4%). Among women who conceived naturally [20, 34, 36, 37], the pregnancy rate was 27.9% (range 2.3–48.2%). In studies evaluating only women with diffuse adenomyosis [20, 35, 38], the pregnancy rate was 48.3% (range 16.7–66.7%), while in studies evaluating only women with focal adenomyosis [34, 36, 39], the pregnancy rate was 44.5% (range 41.1–48.2%).

Miscarriage Rate

Overall the miscarriage rate was 20.9% (range 0–66.7%). Women who conceived naturally [20, 34, 36, 37] had a miscarriage rate of 22.4% (range 0–50%). When comparing patients with diffuse adenomyosis [20, 35, 38] with those with focal adenomyosis [34, 36, 39], miscarriage rate was 6.25% (range 0–66.7%) in the former group and 21.2% (range 14.8–25%) in the latter.

Live Birth Rate

Live birth rate among all women who had undergone conservative surgery was 63.7% (range 20–87.5%). When analyzing only women who conceived naturally [20, 34, 36, 37], live birth rate was 61.8% (range 50–75%). Similarly, live birth rate was 62.5% (range 33.3–87.5%) in women with diffuse adenomyosis [20, 35, 38] and 68.9% (range 55.5–76.2%) in those with focal adenomyosis [34, 36, 39].

Mode of Delivery and Abnormal Placentation

All eight studies describing the mode of delivery reported a 100% rate of elective cesarean section among women with a previous surgical treatment of adenomyosis [18,19,20, 34, 35, 38,39,40]. Abnormal placentation rate was reported in two studies. In the first study, among 12 women with abnormal placentation (12.4%), seven had placenta previa, four had placenta accreta, and one had placenta increta [17]. In the second study, Kishi et al. reported two (6.2%) cases of placenta accreta that were managed by cesarean hysterectomy without fetal or maternal complications [39]. Two (9.5%) cases of uterine rupture in the third trimester were observed in the series by Saremi et al. [40]; one rupture occurred at 37 weeks of gestation and resulted in a stillbirth, whereas the other rupture occurred at 32 weeks and the neonate survived. In both cases, the uterus was repaired after delivery [40].

The rate of preterm delivery, in the four studies evaluating this outcome, was 6.2% [39], 16.5% [17], 13.4% [36], and 33.3% [18], respectively. Among the eight women who returned to the study center for cesarean delivery, Saremi et al. described severe adhesions in one (12.5%), moderate adhesions in two (25%), and mild adhesions in two (25%) [40].

Discussion

In agreement with previous metanalyses [3, 4], our narrative review supports the notion that, among infertile women undergoing ART, adenomyosis is associated with a poorer reproductive outcome. In this regard, the most informative studies are those comparing reproductive outcome in women affected versus not affected by adenomyosis. In the twelve studies published in the last 10 years, a better reproductive outcome was never observed in women with adenomyosis as compared to controls. Conversely, although data may not be conclusive due to the heterogenicity of study designs and population characteristics, adenomyosis was found to negatively affect all the reproductive outcomes, including implantation, pregnancy, miscarriage, and live birth rates. Interestingly, the proportion of studies showing a significant impairment ranged from a minimum of 33% demonstrating a reduced implantation rate to as much as 70% demonstrating a reduced live birth rate. Therefore, these data seem to support the notion that adenomyosis affects implantation rate less than early and late miscarriage rate and especially less than live birth rate. A tentative explanation is that women with adenomyosis suffer from a defective gestational capacity, rather than from pure subfertility. In other words, adenomyosis may not influence the quality of embryos, but it may have an adverse influence on endometrial receptivity [12••], and it may provide a dysfunctional environment for the pregnancy beyond early implantation events [15].

The hypothetical mechanisms responsible for a reduced receptivity and compliance of the pregnant uterus in women with adenomyosis include infiltration of the endometrial-myometrial junction [28], as well as an impairment of the elastic properties of the uterine wall [41]. Such detrimental anatomical conditions may be improved by the administration of GnRH agonists. Accordingly, the five studies that evaluated this issue showed that pre-treated women had an improved reproductive outcome. The proportion of studies reporting a significantly improved pregnancy rate or a reduced miscarriage rate was 20%, whereas the proportion of studies reporting a significantly improved live birth rate was 50%. Because numbers are small, it is difficult to interpret these findings from a pathogenic point of view. However, the fact that pre-treatment with GnRH agonists resulted in a greater improvement in live birth rate as compared to pregnancy and miscarriage rates may be consistent with the hypothesis that adenomyosis negatively affects more the course of an ongoing pregnancy rather than its earlier stages. Further studies are advisable to possibly define the characteristics of women who are more likely to benefit from treatment with GnRH agonists before ART.

In one study only, pre-treatment of women with adenomyosis was accomplished by means of a LNG-IUD. Implantation, clinical pregnancy, and ongoing pregnancy rates were significantly higher in the LNG-IUD group than in the control group. Nevertheless, detrimental effects on the oocyte and/or embryo quality were suspected, but unconfirmed, in the LNG-IUD group. Further research is needed to define the role of medical pre-treatment with LNG-IUD in women with adenomyosis undergoing ART.

Other possible strategies for improving reproductive outcome of ART in infertile women with adenomyosis include long and ultra-long GnRH agonist stimulation protocols. Two studies compared pregnancy rates achieved with a long versus an ultra-long GnRH protocol: in the first study, pregnancy rate was improved for the ultra-long protocol [9••]; in the second study, no significant differences were observed [11]. Therefore, no conclusions can be drawn on which of the two protocols is associated with an improvement, if any, in reproductive outcome in women with adenomyosis. Interestingly, when analyzing infertile women with adenomyosis undergoing the long GnRH agonist protocol, Chen et al. observed a higher live birth rate in those not receiving pre-treatment with GnRH agonists compared to those receiving pre-treatment [8]. This finding, suggesting that GnRH pre-treatment may reduce the efficacy of a long GnRH stimulation protocol, needs confirmation.

Four studies evaluated the possibility of identifying women with adenomyosis who are at higher risk of an adverse reproductive outcome before beginning ART. In two studies, the authors hypothesized that the negative effect of adenomyosis on reproductive outcomes constitutes a continuum, i.e., it becomes clinically significant when a threshold of severity is reached, and it increases proportionally to the extension of disease within the myometrium [32]. In the first study, Youm et al. demonstrated that women with a myometrial thickness greater than 2.5 cm, as compared to less than 2.5 cm, presented significantly worse implantation, pregnancy, miscarriage, and live birth rates [25]. In the second study, Li et al. observed that women with a uterine volume greater than 100 cc faced significantly increased early and late abortion rates and a reduced live birth rate, whereas implantation and pregnancy rates were not affected [5]. In the remaining two studies, categorization of women based on adenomyosis location (extrinsic or intrinsic or diffuse) and a serum CA 125 value greater or lower than 35 were not predictive of an impaired outcome of ART [7, 33]. Surprisingly, among the three studies comparing women with diffuse adenomyosis and women with focal adenomyosis, no difference in reproductive outcomes was observed between the two groups [11, 29, 31].

Further research should be carried out to establish standardized, easy to implement, widely agreed on, and reliable criteria for the categorization of adenomyosis and the prediction of affected women’s reproductive potential. This seems important to adequately counsel women with adenomyosis and to guide them through a shared decision-making process, in the choice between natural conception, conservative surgery, and assisted reproduction, or a combination of these options.

The analysis including a total of 961 women demonstrates the possibility to achieve and carry out a successful pregnancy after surgical removal of adenomyosis. The overall mean pregnancy rate in these patients is substantially similar to that observed in women not operated on (38.1% vs 39.4%). Therefore, surgical treatment does not seem to impact negatively on pregnancy rate in women with adenomyosis. In this regard, it is important to separately evaluate the reproductive outcome of women who underwent conservative surgery for focal adenomyosis from that of women who underwent surgery for diffuse adenomyosis. In fact, surgical removal of focal adenomyosis is easier and usually allows to restore normal uterine anatomy; on the other hand, surgical removal of diffuse adenomyosis may often be challenging and not radical. Nevertheless, according to available evidence, pregnancy rates after surgical removal of focal (44%) [34, 36, 39] versus diffuse (48%) [20, 35, 38] adenomyosis do not seem to differ and also appear comparable to the pregnancy rate observed in non-operated women. Furthermore, birth rates as high as 69% and 62% were observed among women who had undergone surgery for focal and diffuse adenomyosis, respectively. This data suggests that the surgical removal of adenomyosis may be effective in making the uterus a more functional environment for both implantation and maintenance of pregnancy. However, women operated on appear to face a not negligible risk of potentially severe complications of pregnancy such as placenta previa, placenta accreta, uterine rupture, and significant adhesions. For this reason, in our opinion, it is important that the management of ongoing pregnancies and especially deliveries of women who have undergone conservative surgical treatment of adenomyosis are carried out by experienced surgeons in referral centers with a neonatal intensive care unit and a blood bank.

In conclusion, available evidence shows that adenomyosis negatively affects reproductive outcome. Pre-treatment with GnRH agonists is advantageous in improving reproductive outcome in infertile women with adenomyosis undergoing ART, but the actual magnitude of such improvement needs to be better defined. Conservative surgical treatment of both focal and diffuse adenomyosis has been associated with similar pregnancy rates and improved live birth rates when compared to women who have not undergone surgery, although data is based on a limited number of studies.