Müllerian anomalies and endometriosis: associations and phenotypic variations

An exhaustive search of the literature was conducted in Cochrane Library, Google Scholar, Ovid Embase, Ovid MEDLINE, PubMed, Scopus, and Web of Science Core Collection databases to find relevant articles published from the inception of each database. Databases were searched using a combination of keywords and controlled vocabulary for Müllerian anomalies and endometriosis, including “Müllerian anomalies”, “uterine malformations”, “female genitourinary malformations”, “congenital malformations of the Müllerian ducts”, “Mayer-Rokitansky-Küster-Hauser syndrome”, “Müllerian aplasia”, “unicornuate uterus”, “bicornuate uterus”, “vaginal atresia”, “cervicovaginal atresia”, “didelphys uterus”, “Herlyn-Werner-Wunderlich syndrome”, “obstructed-hemivagina-and-ipsilateral-renal-anomaly syndrome”, “septate uterus”, “endometriosis”, and “endometrioma”. Studies were included if they reported Müllerian anomalies in combination with endometriosis in humans, and were written in English language. Case series and case-control studies with a minimum number of 10 patients were used to generate tables. Editorials, conference abstracts, reviews, and studies involving animals were excluded. Bilateral Müllerian agenesis is also called Mayer-Rokitansky-Küster-Hauser (MRKH) syndrome, which occurs in 1:4500 women and is characterized by incomplete development of both Müllerian ducts. When isolated, this anomaly involves the Müllerian tubercle and ducts. However, when development of an entire urogenital ridge is affected, unilateral renal agenesis or malformation occurs [ 25 ]. Additional congenital abnormalities are also prevalent in MRKH syndrome (Table  1 ) [ 13 , 25 – 32 ]. These comorbidities involve mostly the renal and skeletal system, and, like the reproductive tract, display variable phenotypes. The most common abnormality was unilateral renal agenesis, which was reported in roughly half of the studies with prevalence ranging from 4 to 66% of cases. Scoliosis was also observed and ranged from 12 to 25%. Other anomalies included horseshoe kidney, dysplastic kidney, ectopic kidney, spina bifida occulta, and bone malformations. Rarer were abnormalities of the cardiovascular system, which included septal defects and duplication of the inferior vena cava [ 27 , 30 , 31 ]. Table 1 Summary of studies on combined uterovaginal agenesis Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Rosenberg HK et al. 1986 [ 25 ] 5 days – 18 yrs 8.3 (1/12) Uterine agenesis (1), rudimentary (1), bicornuate (3), didelphys (6), duplicated (1) Renal agenesis (8), horseshoe kidney (1), crossed fused ectopia (1), mild hydronephrosis (1) Olive DL and Henderson DY 1987 [ 26 ] n/s 15.4 (2/13) Mullerian agenesis/hypoplasia: vaginal (6), cervical (1), fundal (1), tubal (3), combined (9) n/s Uğur M et al. 1997 [ 13 ] n/s 2.4 (1/42) Mullerian agenesis/hypoplasia: vaginal (16), cervical (3), fundal (7), tubal (1), combined (34) n/s Wang Y et al. 2017 [ 27 ] 11–43 8.7 (8/92) Uterine agenesis (20), unilateral rudimentary uterus (9), bilateral rudimentary uterus (63) Renal agenesis (4), isolated pelvic kidney (3), unilateral pelvic kidney (1), unilateral dysplastic kidney (1), unilateral renal agenesis and dysplastic contralateral kidney and ureter (1); scoliosis (23), spina bifida occulta (6), lumbosacral transitional vertebral (6), vertebral deformation (2), bilateral ulnar clubhands (1), hexadactyly (1), rib malformations (1), talipes equinovarus (1); anal atresia (2), inguinal hernia (4), urogenital fistula (4), duplicated inferior vena cava (4), atrial septal defect (4) Asaturova AV et al. 2020 [ 28 ] 20–24 40.4 (7/42) n/s n/s Dabi Y et al. 2020 [ 29 ] 18.9 ± 4.8 (mean) 38.0 (8/21) Bilateral uterine horns (20) Urinary malformation (4), multiple malformations (1) Wang Y et al. 2020 [ 30 ] 9–37 5.5 (11/201) Bilateral rudiments (191), unilateral rudiments (3), uterine agenesis (7) Renal agenesis (15), unilateral dysplastic kidney (4), horseshoe kidney (1), pelvic kidney (1), polycystic kidney (1), hydronephrosis (1); scoliosis (41), spinal bifida occulta (4), hemivertebra (6), rib malformation (1), scapula malformation (1), pelvic bone malformation (1), Klippel Feil syndrome (1), anal atresia (2), atrial septal defects (3), dextrocardia (1) Tian W et al. 2021 [ 31 ] 5–26 67.6 (23/34) Bilateral remnants (34) Renal agenesis (8), scoliosis (4), limb malformations (3), anal atresia (3), cardiac malformations (3) Steinmacher S et al. 2022 [ 32 ] 16–41 2.8 (9/319) Bicornuate (5), gross uterine abnormalities without endometrium (4). Pelvic kidney (2), double renal pelves (2), Abnormal vertebral development (2) n/s = not specified Summary of studies on combined uterovaginal agenesis 18.9 ± 4.8 (mean) n/s = not specified With the exception of Rosenberg et al., none of these studies stratified patients to distinguish endometriosis prevalence in cases of isolated agenesis versus agenesis of an entire urogenital ridge or mesonephric anomalies [ 25 ]. Overall, the association with endometriosis varied considerably among studies, ranging from ~ 2–68% (Table  1 ). In addition to phenotypic heterogeneity, a possible reason for these differences could be the approach in diagnostic evaluation as only few studies investigated the presence of a functional endometrium. When functioning endometrial tissue was confirmed, the prevalence of endometriosis was less variable, ranging from 40 to 67% [ 25 – 27 , 29 – 31 ]. In a study of 92 MRKH cases, 68.4% of patients presented with bilateral rudiments, whereas 9.7% had unilateral rudiments [ 27 ]. However, all unilateral rudimentary uteri showed the presence of endometrium compared to 22% of bilateral rudiments ( p  < 0.001), and endometriosis was significantly more prevalent in unilateral compared to bilateral remnants (22% vs. 4%, p  < 0.01) [ 27 ]. In a separate study, the same group analyzed a total of 385 uterine rudiments from 201 patients with MRKH syndrome [ 30 ]. Using MRI, they classified the remnants into one-, two-, and three-layer differentiation stages (the latter being typical of a functional uterus). The authors found that endometriosis was present in 39.1% of patients with three-layer differentiation compared to 1.6% of the one- and two-layer differentiation class, consistent with a relationship between the degree of endometrial functionality and endometriosis. Agenesis or atresia of the lower reproductive tract including cervix and vagina result from anomalies of the Müllerian tubercle [ 24 , 33 ]. Typically, these cases are associated with a normal uterine morphology, but fusion or resorption defects may occur, resulting in upper tract anomalies such as didelphys or bicornuate uterus. In addition, segmentary atresia of one of the Müllerian ducts lead to unicornuate uterus presentations. Finally, as discussed above, developmental anomalies of the mesonephric duct are responsible for the co-occurrence of renal anomalies (Table  2 ) [ 34 – 45 ]. Unilateral renal agenesis was again the most reported comorbidity, which occurred in up to 37% of cases. Other anomalies were less frequent and included pelvic kidney, fused kidneys as well as scoliosis. Aside from Zhang et al., and Xu et al., who reported lower rates, endometriosis was generally highly associated with cervico-vaginal agenesis, ranging from 30 to 95% of cases (Table  2 ). Table 2 Summary of studies on cervico-vaginal agenesis or atresia Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Uğur M et al. 1997 [ 13 ] n/s 68.4 (13/19) Mullerian agenesis/hypoplasia: vaginal (16), cervical (3), fundal (7), tubal (1), combined (34) n/s Deffarges JV et al. 2001 [ 34 ] 12–18 44.4 (8/18) Unicornuate (2), hypoplastic (1), bilateral tubal agenesis (1) Renal agenesis (1), ureteral duplicity (1), pelvic kidney (1) Fedele L et al. 2008 [ 35 ] 12–17 58.3 (7/12) Unicornuate (4) Renal agenesis (2), Fallot tetralogy (1) Kriplani A et al. 2012 [ 36 ] 12–19 85.7 (12/14) Unicornuate (6) Renal agenesis (5) Kisku S et al. 2014 [ 37 ] 11–34 70.0 (14/20) Didelphys (2) Renal agenesis (1), cross renal fused ectopia (1), renal agenesis + rectovestibular fistula + segmental thoracolumbar spine scoliosis (1), rectovestibular fistula + ventricular septal defect + bilateral auditory anomalies (1) Song X et al. 2016 [ 38 ] 10–20 54.2 (52/96) Unicornuate (13), bicornutate (10) Renal agenesis (9), pelvic kidney (1), horseshoe kidney (1), ectopic urethral orifice (1). Scoliosis (2), six finger malformation (1), sacral canal cyst (1) Zhang H et al. 2017 [ 39 ] 10–18 13.3 (2/15) Unicornuate (1), didelphys (4) Renal agenesis (2), Scoliosis (2) Xu S et al. 2019 [ 40 ] 15 days – 28 yrs 8.3 (1/12) n/s Ectopic kidney (1) Kang J et al. 2020 [ 41 ] 12–19 80.0 (8/10) Septate (1), bicorporeal (1) Scoliosis (1) Ding Y et al. 2021 [ 42 ] 16.42 ± 5.78 (mean) 60.5 (23/38) Septate (2), bicorporeal (5), hemiuterus (8) n/s Pan HX et al. 2021 [ 43 ] 11–26 95.6 (22/23) Septate (2), bicorporeal (2) Duplex kidney (1). Lumbar/thoracic scoliosis (6), cervical curvature straight (1) Regan L and Dewhurst J 2022 [ 44 ] 14–29 57.1 (8/14) Bicornuate (5), gross uterine abnormalities without endometrium (4). Pelvic kidney (2), double renal pelves (2), Abnormal vertebral development (2) Candiani M et al. 2023 [ 45 ] 11–15 31.2 (5/16) Unicornuate (5) Renal agenesis (2) n/s = not specified Summary of studies on cervico-vaginal agenesis or atresia Renal agenesis (9), pelvic kidney (1), horseshoe kidney (1), ectopic urethral orifice (1). Scoliosis (2), six finger malformation (1), sacral canal cyst (1) Duplex kidney (1). Lumbar/thoracic scoliosis (6), cervical curvature straight (1) n/s = not specified Retrograde menstruation is generally believed to be the main pathogenic mechanism for endometriosis in obstructive anomalies including cervico-vaginal agenesis and MRKH syndrome, and usually found in the form of endometriomas [ 14 , 19 , 26 , 41 , 46 ]. Song et al. reported that in a cohort of 52 patients with cervical atresia and endometriosis, the prevalence of endometriomas was 76% [ 38 ]. The authors argued that decreased peritoneal fluid movement due to the presence of the sigmoid colon on the left side could lead to left-sided cysts [ 38 ]. However, other studies reported absence of endometriosis in 32–40% of cases when a functioning endometrium was demonstrated, arguing that other factors were likely involved [ 26 , 31 ]. The existence of other mechanisms is particularly relevant when absence of uterine tissue is suspected. In these instances, however, it is critical to achieve a definite diagnosis of the pathology. Acién reported a case of endometriosis in a woman with MRKH syndrome with complete agenesis of vagina and uterus and proposed the case as proof against the retrograde menstruation theory [ 18 ]. Interestingly, the authors later found a functional rudimentary horn in the same woman, confirming that retrograde menstruation was involved [ 47 ]. Although MRI is the primary non-invasive diagnostic choice for Müllerian anomalies, it does not demonstrate the presence of endometrial activity within Müllerian tissues [ 48 ]. Therefore, surgical/histological evidence must be attained prior to excluding retrograde menstruation. Importantly, reports of endometriosis in cases of confirmed uterine agenesis or following surgical correction of obstruction present evidence of coelomic metaplasia or other mechanisms [ 20 , 21 ]. In a case series of congenital vaginal obstruction or agenesis, endometriosis was observed between 6 months to 5 years following surgical repair of the anomaly [ 22 ]. The authors suggested that although it was possible that prior endometriotic implants from retrograde menstruation may have remained active, coelomic metaplasia, lymphatic/vascular metastasis, immunologic deficiency, or genetic factors were the likely pathogenic mechanisms. Transverse or oblique vaginal septum is the result of segmentary atresia and failed canalization of the developing vagina [ 33 ]. The resulting partial or complete outflow obstruction is associated with a high rate of early endometriosis, seemingly supporting the pathogenic mechanism of retrograde menstruation. However, only case reports or small case series are present in the literature, making it difficult to assess prevalence rates [ 18 , 19 , 49 , 50 ]. While none of seven women with transverse vaginal septum showed endometriosis in a report from Kapczuk et al., Deligeoroglou et al. presented a case series of seven adolescents with endometriosis and transverse or oblique vaginal septum [ 19 , 49 ]. In the latter cohort, one patient had a bicornuate uterus, two had didelphys uterus, and congenital abnormalities of the urinary tract were present in six patients [ 49 ]. Unicornuate uterus results from lack or underdevelopment of one of the Müllerian ducts, and is estimated to represent about 10% of all uterine anomalies [ 46 ]. If the anomaly is isolated, both kidneys are present, whereas unicornuate uterus is associates with unilateral renal agenesis in cases of agenesis or hypoplasia of an entire urogenital ridge [ 33 ]. In the studies included, unicornuate uterus was often reported with renal anomalies, the most common being unilateral kidney agenesis (Table  3 ) [ 13 , 26 , 51 – 66 ]. Similar to MRKH syndrome, other malformations included horseshoe and ectopic kidney, but compared to MRKH, skeletal anomalies were less common. Table 3 Summary of studies on unicornuate uterus Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Buttram VC and Gibbons WE 1979 [ 51 ] n/s 15.8 (3/19) Non-communicating cavitated rudimentary horn (1), communicating cavitated rudimentary horn (6), non cavitated rudimentary horn (6), no horn (6) Renal agenesis (1), crossed renal ectopy (1), moderate hydronephrosis (1) d Heinonen PK 1983 [ 52 ] 16–79 20.0 (4/20) Non-communicating cavitated rudimentary horn (7), non-communicating non-cavitated rudimentary horn (7), no horn (2), cavity in non-communicating rudimentary horn not determined (4) Renal agenesis (5), pelvic kidney (3), horseshoe kidney (1) b Heinonen PK 1983 [ 53 ] n/s 5.0 (1/20) n/s n/s Fedele L et al. 1987 [ 54 ] n/s 14.3 (3/21) Communicating rudimentary horn (1), non-communicating rudimentary horn (4), non-cavitated rudimentary horn (7), no horn (7) Renal agenesis (3), renal ptosis (2), double renal pelvis (1), horseshoe kidney (1) e Olive DL and Henderson DY 1987 [ 26 ] n/s 30.0 (3/10) Communicating rudimentary horn (1), non-communicating rudimentary horn (1), non-cavitated rudimentary horn (4), no horn (4) n/s Fedele L et al. 1990 [ 55 ] 14–33 60.0 (6/10) Non-communicating rudimentary horn (10) n/s Donderwinkel PFJ et al. 1992 [ 56 ] 18–39 17.8 (8/45) n/s Renal agenesis (5), ectopic kidney (4), hypoplastic pyelum (2) a Fedele L et al. 1992 [ 57 ] n/s 55.0 (11/20) Non-communicating cavitated rudimentary horn (11), non cavitated rudimentary horn (9) n/s Moutos DM et al. 1992 [ 58 ] n/s 48.3 (14/29) n/s Renal agenesis (5), malrotated kidney (1), renal ptosis (1) f Liu MM 1994 [ 59 ] 16–34 31.8 (7/22) Non-communicating cavitated rudimentary horn (1), communicating cavitated rudimentary horn (1) n/s Fedele L et al. 1995 [ 60 ] n/s 48.8 (20/41) Communicating rudimentary horn (1), non-communicating rudimentary horn (8), non-cavitated rudimentary horn (19), no horn (13) n/s Fedele L et al. 1996 [ 61 ] 14–40 46.9 (23/49) Communicating rudimentary horn (3), non-communicating rudimentary horn (14), non-cavitated rudimentary horn (19), no horn (13) Renal agenesis (5), ectopic kidney (3), renal agenesis and contralateral ectopic kidney (1), double renal pelvis (2), horseshoe kidney (2), unilateral medullary sponge kidney (1) Heinonen PK 1997 [ 62 ] 15–79 21.4 (9/42) Non-communicating cavitated rudimentary horn (15), non-communicating non-cavitated rudimentary horn (15), no horn (7) Renal agenesis (8), pelvic kidney (3), horseshoe kidney (1), renal malrotation (1); scoliosis (1), arthrogryposis (1), hypoplasia of radii (1), hemivertebra (1); auditory defects (2) c Uğur M et al. 1997 [ 13 ] n/s 23.1 (3/13) Communicating rudimentary horn (4), non-communicating rudimentary horn (3), non-cavitated rudimentary horn (4), no horn (2) n/s Fedele L et al. 2005 [ 63 ] 14–26 50.0 (5/10) Cavitated non-communicating rudimentary horn (10) Renal agenesis (3) Acién P and Acién M 2010 [ 64 ] n/s 13.5 (5/37) n/s Renal agenesis (10); skeletal anomalies (5) Piriyev E and Römer T 2020 [ 65 ] n/s 62.5 (10/16) Rudimentary horn (13), no horn (3) n/s Tellum T et al. 2023 [ 66 ] 34.0 ± 9.5 (mean) 17.5 (57/326) Communicating functional horn (4), non-communicating non-functional horn (126), non-communicating functional horn (88), no horn (108) Renal agenesis (39), pelvic kidney (11), other urinary tract malformation (18), genetic syndrome (9), cloacal anomaly (6) n/s = non specified a Total number of malformations out of 35 intravenous pyelograms or ultrasound b Total number of malformations out of 15 intravenous pyelograms performed c Total number of malformations out of 34 intravenous pyelograms/renography/ultrasound d Total number of malformations out of 13 intravenous pyelograms performed e Total number of malformations out of 16 intravenous urographies f Total number of malformations out of 12 intravenous pyelograms Summary of studies on unicornuate uterus n/s = non specified a Total number of malformations out of 35 intravenous pyelograms or ultrasound b Total number of malformations out of 15 intravenous pyelograms performed c Total number of malformations out of 34 intravenous pyelograms/renography/ultrasound d Total number of malformations out of 13 intravenous pyelograms performed e Total number of malformations out of 16 intravenous urographies f Total number of malformations out of 12 intravenous pyelograms Unicornuate uterus can present with or without a rudimentary horn, which may or may not be communicating with the contralateral horn. A non-communicating rudimentary horn was seen in the majority of patients, and these had higher risk of developing endometriosis and pelvic inflammatory disease [ 52 ]. Overall, the prevalence of endometriosis ranged from 5 to 62% (Table  3 ). As noted previously, this variability may be a reflection of lack of stratification according to clinical presentations and associated comorbidities. In general, lower prevalence was usually reported in studies that included higher numbers of patients with non-cavitated rudimentary horn or no horn [ 51 , 54 , 66 ]. Conversely, higher rates of endometriosis were seen in studies with larger proportions of communicating and non-communicating cavitated rudimentary horn [ 55 , 63 , 65 ]. After stratifying for the type of unicornuate uterus presentation, Piriyev and Römer histologically detected endometriosis in 87.5% of patients when endometrium was found in an obstructed rudimentary horn, compared to 25% when no active endometrium was present, and 66.7% in patients without a rudimentary horn [ 65 ]. Fedele et al. found that endometriosis was more prevalent in cases where the rudimentary horn was cavitated and non-communicating (57.2%) compared to communicating (33%), non-cavitated (42.1%), or when no horn was present (46.2%) [ 61 ]. In a retrospective cohort study of 326 women with unicornuate uterus compared to 326 controls, Tellum et al. found that the rate of endometriosis was significantly higher in the unicornuate uterus group (17.5% vs. 10.7%, p  = 0.018) [ 66 ]. Not surprisingly, the risk of developing endometriosis was higher in women with a functional rudimentary horn compared to those with unicornuate uterus but no functional horn [ 66 ]. Infertility is often a presenting complaint in unicornuate uterus cases, and several studies included infertile women to evaluate if endometriosis was a possible factor [ 52 , 57 , 60 , 66 ]. Fedele et al. reported endometriosis in 48.8% of 41 women with unicornuate uterus, the majority of whom were infertile patients [ 60 ]. Although the authors acknowledged that endometriosis could contribute to infertility, they found similar reproductive performances between women with and without endometriosis [ 60 ]. In a separate study that excluded patients with obstructive abnormalities, Fedele et al. compared infertile subjects with non-obstructive Müllerian abnormalities to infertile controls and found that the prevalence of endometriosis was significantly greater in unicornuate uterus compared to other Müllerian anomalies (55% vs. 28%, p  < 0.05) [ 57 ]. Obstructed hemivagina and ipsilateral renal anomaly (OHVIRA) syndrome, previously called Herlyn-Werner-Wunderlich (HWW) syndrome, is a complex anomaly characterized by uterine duplicity (didelphys, bicornuate, septate), obstructed hemivagina, and renal dysplasia ipsilateral to the hemivagina [ 67 , 68 ]. This Müllerian disorder is induced by an associated Wolffian anomaly. The ureteral buds sprout from the opening of the Wolffian ducts in the urogenital sinus. Therefore, agenesis or developmental defects involving one of the mesonephric ducts would lead to absence of the ureteral bud, with consequent renal agenesis and blind ipsilateral hemivagina [ 69 ]. Like other Müllerian anomalies, significant variations of the uterine anatomy have been reported (Table  4 ) [ 19 , 70 – 87 ]. Candiani et al. diagnosed didelphys uterus in 58% of patients, double uterus in 31%, and bicornuate in 11% [ 71 ]. In a cohort of 87 patients, the uterus was didelphys in 77% of cases, bicornuate in 11.5%, and septate bicollis in 11.5% [ 72 ]. Obstructed hemivagina was present in 95.4% of patients, whereas the remaining 4.5% had unilateral cervical atresia [ 72 ]. Similar findings were reported in a cohort of 51 patients, showing didelphys uterus in 78.4% of cases, complete septate in 17.6%, and bicornuate in 3.9% [ 84 ]. The renal anomaly usually presented as renal agenesis in postpubertal patients and multicystic dysplastic kidney in prepubertal patients [ 70 , 73 , 75 , 77 , 78 , 82 , 88 , 89 ]. In patients with ages ranging 11–42 years, Fedele et al. reported ipsilateral renal agenesis in 95.4% of cases and ipsilateral hypoplastic kidney in 4.6%, although the authors did not provide a breakdown of ages [ 72 ]. Overall, laterality of the renal anomaly was reported in 564 patients, and showed that renal agenesis was right-sided in 341 cases (60.5%) and left-sided in 223 cases (39.5%). Additional associated malformations were rare, and included skeletal malformations, cervical aplasia, and unilateral vaginal agenesis [ 19 , 78 , 82 ]. Table 4 Summary of studies on OHVIRA syndrome Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Stassart JP et al. 1992 [ 70 ] 6–26 46.7 (7/15) n/s Renal agenesis (15) Candiani GB et al. 1997 [ 71 ] 9–36 27.8 (10/36) Didelphys (21), double (11), bicornuate (4) Renal agenesis (36) Fedele L et al. 2013 [ 72 ] 11–42 13.8 (12/87) Didelphys (67), septate bicollis (10), bicornuate (10) Renal agenesis (83), dysplastic hypoplastic kidney (4) Tong J et al. 2013 [ 73 ] 10–50 17.1 (12/70) n/s Renal agenesis (70) Wang S et al. 2013 [ 74 ] 11–33 5.8 (3/52) n/s Renal agenesis (50), ectopic kidney (1), renal malformation (1) Sabdia S et al. 2014 [ 75 ] n/s 60.0 (6/10) Cervical aplasia (2) Renal agenesis (10) Tong J et al. 2014 [ 76 ] 13–37 19.1 (18/94) n/s Renal agenesis (94) Zhu L et al. 2015 [ 77 ] 12.86 ± 1.84 (mean, group 1) 21.68 ± 7.43 (mean, group 2) 20.2 (16/79) Complete obstruction (24, group 1), incomplete obstruction (55, group 2) Renal agenesis (79) Kapczuk K et al. 2018 [ 78 ] 11–17 5.5 (1/18) Didelphys (16), septate (2) Renal agenesis (18). Scoliosis (1), scoliosis + lumbarization S1 + butterfly vertebra S2 (1) Troncon JK et al. 2018 [ 79 ] 11–29 14.3 (2/14) Didelphys (14) Renal agenesis (14) Dabi Y et al. 2020 [ 80 ] 16.6 ± 4.14 (mean) 16.4 (13/79) Didelphys (79) Renal aplasia (73), renal dysgenesis (2), renal duplicity (2) Gungor Ugurlucan F et al. 2020 [ 81 ] 10–48 3.4 (1/29) n/s Renal agenesis (28), multicystic dysplastic kidney (1) Kim YN et al. 2021 [ 82 ] 0–32 60.0 (6/10 who underwent laparoscopy in a cohort of 65 patients) Didelphys (65) Renal agenesis (34), multicystic dysplastic kidney (31) Yi S et al. 2021 [ 83 ] 11–34 5.9 (1/17) Complete septate uterus (17) Renal agenesis (17) Candiani GB et al. 2022 [ 84 ] 12–36 17.6 (9/51) Didelphys (40), septate (9), bicornuate (2) Renal agenesis (49), dysplastic kidney (2) Zarfati A et al. 2022 [ 85 ] 0.5–15.7 3.6 (1/28) Didelphys (28) Renal agenesis (23), multicystic dysplastic kidney (5) Zhang J et al. 2020 [ 86 ] 11–28 15.8 (3/19) Cervical dysplasia (1) Renal agenesis (19) Zhang Y et al. 2022 [ 87 ] 13.13–23.34 (mean) 13.0 (6/46) Didelphys (40), septate (6) Renal agenesis (46) Kapczuk K et al. 2023 [ 19 ] 11–17 43.5 (10/23) Didelphys (23) Renal agenesis (23); scoliosis (1) n/s = non specified Summary of studies on OHVIRA syndrome 12.86 ± 1.84 (mean, group 1) 21.68 ± 7.43 (mean, group 2) Renal agenesis (18). Scoliosis (1), scoliosis + lumbarization S1 + butterfly vertebra S2 (1) n/s = non specified The prevalence of endometriosis ranged from 3 to 60% (Table  4 ). However, the documentation of endometriosis was typically limited to patients who had undergone reparative laparoscopy or laparotomy and thus did not include all patients in the study cohorts, therefore underestimation of disease frequencies was possible [ 70 , 71 , 82 , 83 , 85 ]. Tong et al. and Zhu et al. both noted that patients with complete obstruction of the hemivagina were more likely to have endometriosis than those with incomplete obstruction [ 73 , 77 ]. They hypothesized that complete obstruction led to a higher volume of retrograde flow compared to incomplete obstruction where blood could partly drain through the patent vaginal canal, supporting other reports [ 70 , 78 , 86 , 88 , 89 ]. These studies also pointed to a delayed diagnosis of both OHVIRA and endometriosis as contributing to more advanced endometriosis due to a longer duration of retrograde flow. Cohort studies of didelphys uterus were few, and showed endometriosis in 16.7–40% of cases (Table  5 ) [ 13 , 26 , 57 ]. None of these studies, however, mentioned additional variations of reproductive tract anatomy or the presence of associated congenital malformations. Therefore, it is not clear if all these cases were isolated Müllerian anomalies or some included mesonephric involvement. Table 5 Summary of studies on didelphys uterus Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Uğur M et al. 1997 [ 13 ] n/s 16.7 (2/12) n/s n/s Olive DL and Henderson DY 1987 [ 26 ] n/s 40.0 (4/10) n/s n/s Fedele L et al. 1992 [ 57 ] n/s 35.3 (6/17) n/s n/s n/s = non specified Summary of studies on didelphys uterus n/s = non specified Heinonen retrospectively analyzed a cohort of 49 patients with didelphys uterus, including 8 with OHVIRA, and found endometriosis in 7 among the 45 women who underwent laparoscopy (16%) [ 90 ]. However, the author did not specify how many were didelphys or OHVIRA cases. Similarly, in a cohort of patients with didelphys uterus or OHVIRA, Moutos et al. reported a combined endometriosis prevalence of 8.7% (2/23) without providing further details [ 58 ]. Although retrograde menstruation was mentioned in some obstructive case reports, several authors proposed other mechanisms [ 90 – 94 ]. Heinonen reported that 9/49 (18%) cases had an obstructed hemivagina, but only 1 out of 9 cases (11%) had endometriosis, providing an argument against retrograde menstruation [ 90 ]. Hori et al. presented a case of ectopic endometriosis of the alimentary tract in a patient with didelphic uterus [ 94 ]. The patient did not show any ovarian disease, and the authors suggested that epithelial metaplasia or metastatic transplantation were likely responsible for the disease. Similar to didelphys, bicornuate uterus results from the incomplete fusion of the Müllerian ducts during organogenesis. Case series of bicornuate uterus were also few, and showed a wide variation in endometriosis prevalence, ranging between 0% and 72.2% (Table  6 ) [ 13 , 17 , 53 , 57 , 65 ]. Most studies described the type of separation between the two uterine cavities (i.e. complete vs. partial), but only one reported the presence of additional malformations, which involved the renal system [ 17 ]. Table 6 Summary of studies on bicornuate uterus Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Heinonen PK 1983 [ 53 ] n/s 1.4 (1/69) Complete (8), partial (61) n/s Fedele L et al. 1992 [ 57 ] n/s 31.2 (10/32) Complete (26), partial (6) n/s Creatsas G et al. 1994 [ 17 ] 11–19 0.0 (0/11) Obstructed hemivagina (4), obstructed cervical os (3), rudimentary horn (4) Renal agenesis (3), fused pelvic kidney (1) Uğur M et al. 1997 [ 13 ] n/s 17.2 (5/29) Complete (12), partial (17) n/s Piriyev E and Römer T 2020 [ 65 ] n/s 72.2 (13/18) n/s n/s n/s = non specified Summary of studies on bicornuate uterus n/s = non specified In a study where bicornuate and didelphic uterus were combined, Acién and Acién reported an endometriosis prevalence of 9.7% (12/123) [ 64 ]. Unilateral kidney agenesis was the most commonly reported anomaly associated with these conditions. The authors noted that subjects with bicornuate or didelphic uterus and renal agenesis were more likely to have endometriosis than those without renal anomalies (20.0% [9/45] vs. 3.8% [3/78], respectively) [ 64 ]. A retrospective analysis of 800 women with Müllerian anomalies in the restricted gene pool population of LaCrete (Canada) showed that bicornuate and unicornuate uterus were the most common (21.7% for both) uterine anomalies [ 95 ]. In this selected population, reproductive tract anomalies in general occurred 5.5 times more frequently than in the general population suggesting the existence of a genetic component [ 95 ]. The author also found some five cases of endometriosis and several associated renal malformations but did not report any statistics or clarify in which class of Müllerian anomalies they occurred. Due to the limited number of studies, a causal or direct relationship between bicornuate uterus and endometriosis remains unclear [ 96 , 97 ]. Few case reports described endometriosis in patients with hindered menstrual flow [ 98 , 99 ]. However, several authors reported the type of endometriosis as scar endometriosis developing after past surgery [ 100 – 102 ]. This type of endometriosis usually occurs from the iatrogenic seeding of endometrial tissue following surgical procedures including exploratory laparoscopy, hysterectomy, or Cesarean section. In these cases, the concomitant Müllerian anomaly is unlikely to be directly associated with endometriosis. Septate uterus results from the incomplete resorption of the midline uterine septum after embryonic fusion of the Müllerian ducts. Complete septate uterus is associated with a septum extending to the cervix and sometimes, with a longitudinal vaginal septum, while partial uterine septum extends no further than the internal cervical os [ 9 ]. In the selected studies, partial septum was more represented than complete (941 cases vs. 482, respectively) (Table  7 ) [ 103 – 116 ]. However, very few studies described the presence of additional malformations. Comorbid congenital anomalies included renal agenesis, atrophic kidney, and ureter duplication [ 51 , 64 , 107 ]. Endometriosis rates were variable, ranging from 0 to 77.1%, and results were often conflicting. By comparing groups with and without septate uterus, Fedele et al., Uğur et al., and Demir et al. found no association between uterine septum and endometriosis [ 13 , 57 , 111 ]. Similarly, a retrospective Finnish study reported only a 3.3% rate of endometriosis in patients with complete septate uterus [ 107 ]. In contrast, a retrospective study by Nawroth et al. found endometriosis to be significantly more common in patients with partial or complete septate uterus compared to women with normal uterine cavities (25.8% vs. 15.2% endometriosis, respectively) [ 108 ]. Piriyev and Romer noted that in a cohort of 179 women with septate uterus, 92.6% of those with a complete septum had endometriosis versus 74.3% of patients with a partial septum [ 65 ]. Table 7 Summary of studies on septate uterus Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Buttram VC et al. 1974 [ 103 ] 20–33 32.1 (9/28) Complete (6), partial (22) n/s Buttram VC and Gibbons WE 1979 [ 51 ] n/s 0.0 (0/67) Complete (14), partial (53) Duplication of collecting system (3), atrophic kidney (1) Heinonen PK 1983 [ 53 ] n/s 0.0 (0/72) Complete (27), partial (45) n/s Olive DL and Henderson DY 1987 [ 26 ] n/s 63.6 (7/11) Complete (2), partial (9) n/s Fedele L et al. 1992 [ 57 ] n/s 26.9 (31/115) Complete (24), partial (91) n/s Fedele L et al. 1993 [ 104 ] 18–39 35.5 (11/31) n/s n/s Haberal A et al. 1996 [ 105 ] 22–37 16.9 (10/59) Complete (17), partial (42) n/s Uğur M et al. 1997 [ 13 ] n/s 16.1 (10/62) Complete (3), partial (59) n/s Grimbizis G et al. 1998 [ 106 ] 22–45 26.1 (12/46) Complete (10), partial (36) n/s Heinonen PK 2006 [ 107 ] 14–46 3.3 (2/61) Complete (61) Renal agenesis (5), double ureter (6), obstructed hemivagina + renal agenesis (4) Nawroth F et al. 2006 [ 108 ] 29.4 ± 4.7 years (mean) 25.8 (31/120) Complete (24), partial (96) n/s Darwish AM et al. 2009 [ 109 ] 19–25 18.7 (6/32) Complete (32) No renal anomalies Wang JH et al. 2009 [ 110 ] 22–39 12.0 (3/25) Complete (25) n/s Acién P and Acién M 2010 [ 64 ] n/s 5.8 (3/52) Complete (26), partial (26) Renal agenesis (1) Demir B et al. 2010 [ 111 ] 26.9 ± 5.5 (mean) 9.8 (9/92) Complete (39), partial (53) n/s Gergolet M et al. 2010 [ 112 ] 29.62 ± 4.5 (mean) 20.1 (36/179) n/s n/s Galal AF et al. 2016 [ 113 ] 20–34 10.0 (5/50) b Hypoplastic uterus (2) n/s LaMonica R et al. 2016 [ 114 ] n/s a 37.0 b n/s n/s Sukur YE et al. 2018 [ 115 ] n/s 13.0 (7/54) Complete (54) n/s Piriyev E and Römer T 2020 [ 65 ] n/s 77.1 (138/179) Complete (27), partial (152) n/s Bean E et al. 2022 [ 23 ] 16–51 33.3 (4/12) n/s n/s Chang Y et al. 2023 [ 116 ] 20–40 8.0 (28/348) Complete (91), partial (257) n/s n/s = non specified a women of reproductive age b prevalence of septate uterus in endometriosis patients Summary of studies on septate uterus n/s = non specified a women of reproductive age b prevalence of septate uterus in endometriosis patients Other studies used a different approach and investigated the rate of septate uterus among women with and without endometriosis. A prospective study from LaMonica et al. found that uterine septum was significantly more common in patients with histologically-confirmed endometriosis compared to those without (37% versus 27%, respectively) [ 114 ]. Consistent with these findings, a retrospective analysis by Matalliotakis et al. compared a group of 425 infertile patients with endometriosis to a control group of 200 women with tubal or male factor infertility and no endometriosis [ 117 ]. The endometriosis group contained 7 patients (1.6%) with septate uterus, while only 1 septate uterus (0.5%) was identified in the control group [ 117 ]. In light of these observations, the association between septate uterus and endometriosis remains unclear. One contributing factor to the contradicting results may be that patients in the control populations typically had a significantly higher incidence of primary infertility compared to the septate uterus groups. Hence, it is possible that the prevalence of endometriosis may be overrepresented in the control populations. Another potential source of inconsistency is the evolving definition of septate uterus. For example, current ASRM guidelines require a septum length > 1 cm, whereas the previous 2016 ASRM classification defined fundal indentation > 1.5 cm as septate, and the American Fertility Society guidelines of 1988 did not provide measurable criteria [ 5 , 118 ]. In contrast, the European Society of Human Reproduction and Embryology/ European Society for Gynaecological Endoscopy (ESHRE/ESGE) define complete septate uterus as a fundal indentation exceeding 50% of the uterine wall thickness and septa extending to the internal cervical os [ 119 , 120 ]. These differences are likely to have contributed to inconsistent inclusion criteria across studies. There is currently limited data available about endometriosis in relation to arcuate uterus, which is defined by the ASRM as internal fundal indentation < 1 cm [ 9 ]. Uğur et al. found endometriosis in 3 of 9 women (33.3%) with arcuate uterus, whereas other case series found 3 and 2 patients with endometriosis in groups of 14 and 12 arcuate uterus respectively (Table  8 ) [ 13 , 23 , 57 ]. Perhaps more informative is a larger study by Piriyev and Romer, which reported endometriosis in 48 of 55 patients (77.6%) [ 65 ]. Table 8 Summary of studies on arcuate uterus Reference Age (range, yrs) Endometriosis prevalence (%) Reproductive tract variations ( n ) Additional malformations ( n ) Heinonen PK 1983 [ 53 ] n/s n/s 0.0 (0/23) n/s Fedele L et al. 1992 [ 57 ] n/s n/s 21.4 (3/14) n/s Piriyev E and Römer T 2020 [ 65 ] n/s n/s 77.6 (45/58) n/s Bean E et al. 2022 [ 23 ] 16–51 16–51 16.7 (2/12) n/s n/s = non specified Summary of studies on arcuate uterus n/s = non specified Robert’s uterus is a partially obstructive variant of septate uterus characterized by an asymmetric septum dividing the uterine cavity into a non-communicating hemicavity and a patent contralateral cavity. The external contour of the uterus and the urinary system are typically normal. Data on endometriosis in Robert’s uterus are limited and none of the few case series available included more than five patients [ 121 – 123 ].

Understanding how Müllerian anomalies arise and clinically present – including associated malformations and diseases – is critical to facilitate proper diagnosis and the development of effective management plans. Early detection could reduce the risk of endometriosis and facilitate the management of ancillary disorders of the renal or skeletal systems or complications related to subfertility and pregnancy. The first step in working up patients with cyclical pain and amenorrhea or dysmenorrhea for the diagnosis of Müllerian abnormalities is usually physical exam combined with pelvic ultrasonography [ 124 ]. Ultrasound is widely available and can be useful in detecting obstructive presentations, but is not the most reliable diagnostic method [ 124 ]. MRI is preferred as it shows up to 95% correlation with laparoscopic findings and is a valuable tool for the evaluation of associated complications [ 125 ]. However, although MRI is an efficient method for the diagnosis of Müllerian anomalies, it is not accurate to assess the presence of endometriosis [ 126 ]. Therefore, judicious use of laparoscopy may be necessary to reveal or manage other complications including endometriosis and ovarian tumors, in addition to guiding surgical reconstruction of upper genital tract structures [ 127 , 128 ]. The risk of endometriosis for specific classes of Müllerian anomalies is still a matter of debate, and studies analyzing endometriosis in Müllerian anomalies showed a wide variability in prevalence. There are several possible explanations for the observed discrepancies. First, both Müllerian anomalies and endometriosis are complex, multifactorial conditions with variable phenotypic presentations that pose classification challenges for clinical analyses [ 9 , 10 ]. When attempting to classify Müllerian anomalies, it would be important to consider the embryological events regulating the development of the genitourinary system to better correlate genital abnormalities with comorbid pathologies. For example, the involvement of Wolffian duct or mesonephric anomalies represent phenotypic classes that may have specific associations with endometriosis. However, lack of stratification of phenotypes was a significant limitation of the literature reviewed here. In addition, most papers did not report several details of endometriosis, including staging, type, and location of lesions. Similarly, past family history of genital malformations or endometriosis was rarely recorded or reported, further complicating assessment of causal or direct relationship. Second, the etiology of Müllerian anomalies and endometriosis remain largely unknown. Familiar clustering showed that genetic factors play a significant role [ 2 , 129 ]. However, the genetics of both disorders are characterized by heterogeneity, incomplete penetrance, and variable expressivity [ 130 – 132 ]. These challenges were rarely addressed among studies of concomitant Müllerian anomalies and endometriosis, and genetic analyses were rarely performed [ 133 , 134 ]. Karyotyping analyses were almost exclusively limited to MRKH cases, with two studies identifying a 10:17 translocation and a ring 18 chromosome [ 25 , 51 ]. For other anomalies, only two studies reported chromosomal investigations. Acién et al. identified an increased length of heterochromatin on the long arm of chromosome 9 (46,XX9qh+) in a woman with didelphys uterus, but the variant was considered “without clinical consequences” [ 135 ]. Fenn performed karyotyping in a patient with bicornuate uterus and found a mutation at chromosome 17q [ 98 ]. Although the author did not provide specific information on the chromosomal abnormality, they proposed it as possible cause for the susceptibility of endometriosis [ 98 ]. Notably, molecular genetic testing was not performed in any of the literature addressing concomitant Müllerian anomalies and endometriosis. Several genetic loci, including copy number and single nucleotide variants have been identified using microarray, genome sequencing, and GWAS in separate studies of MRKH syndrome or endometriosis [ 130 – 132 , 136 ]. Interestingly, some of the genetic variants associated with endometriosis are also candidate variants for Müllerian anomalies, including WNT4, HOXA10, and HNF1B [ 130 , 137 – 139 ]. Future genomic analyses in patients with both conditions may play a significant role in improving our understanding of the etiology and molecular mechanisms involved. Finally, the complex cellular events responsible for endometriosis and their relationship with Müllerian anomalies remain unclear. In general, retrograde menstruation is widely accepted in obstructive anomalies, and it has also been proposed in cases where menstrual flow may be impaired but not completely obstructed, including septate and unicornuate uterus [ 23 , 65 , 105 , 110 – 113 ]. However, when endometriosis ensues in cases of Müllerian agenesis, other mechanisms are more likely, including coelomic metaplasia, Müllerian remnants, and genetic predisposition [ 13 , 18 , 20 – 23 ]. A recent meta-analysis tried to provide some clarity on these pathogenic drivers [ 140 ]. A significantly higher prevalence of endometriosis was found in obstructive versus non-obstructive anomalies, seemingly supporting the retrograde menstruation theory. However, the authors cautioned that other mechanisms may be involved since only half of patients with obstructive anomalies had coexistent endometriosis, and several confounding factors affected the comparison between the two groups, including age and surgical indications. The same study showed a lack of difference in endometriosis rates between women with non-obstructive anomalies versus no anomalies, an apparent argument against coelomic metaplasia or embryonic remnants. However, the authors pointed out that these mechanisms cannot be ruled out since the control group included women with infertility and pelvic pain, both associated with endometriosis [ 140 ]. This analysis further supports the observed lack of standardization in studies involving Müllerian anomalies and endometriosis. In conclusion, the elucidation of the complex relationship between these two conditions hinges on improved experimental designs allowing better stratification and comparative analyses. Most importantly, as both conditions are heterogeneous, multifactorial, and affected by genetic, epigenetic, and environmental factors, the use of interdisciplinary approaches combining clinical investigations, molecular genetics, and population-based analyses become increasingly critical to understand cellular and molecular determinants, and develop better and more personalized management plans.

Müllerian anomalies are anatomical variations of the female reproductive tract resulting from the incomplete development of the Müllerian ducts [ 1 ]. During embryogenesis, the Müllerian (paramesonephric) ducts arise as an invagination of the coelomic epithelium, and elongate posteriorly parallel to the Wolffian (mesonephric) ducts until they fuse into a Y-shaped structure that forms the uterine primordium [ 2 – 4 ]. Resorption of the septum formed by the fusion of the two Müllerian ducts results in the formation of a single canal that will later develop into the uterine cavity [ 5 ]. At the site where the lower portions of the Müllerian ducts fuse with the medial wall of the Wolffian ducts, the Müllerian tubercle forms [ 6 ]. The subsequent cavitation of the Müllerian tubercle will result in the development of the primitive vagina, whose wall is formed by Wolffian duct cells and internally by Müllerian tubercle cells before proper epithelialization from the urogenital sinus occurs [ 4 , 7 , 8 ]. Disruption of any of these complex processes can lead to abnormalities of Müllerian tract structures. In addition, the development of the genital system is closely related to the urinary system as they both derive from the intermediate mesoderm [ 2 ]. Importantly, the Wolffian ducts are involved in the regulation of Müllerian duct development, while also providing a foundation for the formation of the metanephros (permanent kidneys) [ 4 ]. The metanephros are induced by the ureteral buds, which sprouts from the Wolffian ducts at the urogenital sinus. Therefore, anomalies of Wolffian duct development result in malformations of the urinary system in addition to the reproductive tract. At the molecular level, several of the signaling pathways regulating Müllerian duct development are also involved in the development of other organs including the renal and skeletal systems, leading to additional concomitant congenital malformations [ 2 , 9 ]. As a result, Müllerian anomalies occur with a wide range of clinical presentations, severities, and comorbidities including unilateral kidney agenesis, kidney dysplasia, and scoliosis. Müllerian anomalies are also associated with endometriosis, a disease characterized by the abnormal growth of endometrial-like tissue outside of the uterus. Like Müllerian anomalies, endometriosis has a multifactorial pathogenesis and shows high degree of phenotypic variability and severity [ 10 , 11 ]. Several mechanisms have been proposed in the pathogenesis of endometriosis. One of the most accepted is Sampson’s theory of retrograde menstruation, which suggests that during menstruation, some endometrial tissue moves through the oviducts into the peritoneal cavity causing endometriosis [ 12 ]. An alternative theory is coelomic metaplasia, postulating that because endometrium and peritoneum originate from the same embryonic coelomic epithelium, the peritoneal mesothelium can eventually transform into endometrium causing endometriosis [ 12 ]. Another mechanism involves Müllerian embryonic remnants, and suggests that Müllerian structures persisting from the time of embryonic development can adhere to the peritoneum and subsequently develop into endometriosis [ 12 ]. Each theory alone cannot explain the heterogeneity of endometriosis, sometimes occurring even within the same patients. Therefore, it is possible that multiple mechanisms are involved and it is likely that they are affected by genetic predisposition, immune status, and environmental factors. The relationship between Müllerian anomalies and endometriosis is complex and unclear. High rates of endometriosis have been found in obstructive compared to non-obstructive Müllerian anomalies [ 13 ]. In addition, complete resorption of endometriosis has been reported following surgical removal of uterine rudiments or removal of obstruction [ 14 – 19 ]. These studies seem to support the retrograde menstruation mechanism. However, endometriosis has also been found in non-obstructive Müllerian malformations [ 13 , 18 , 20 – 23 ]. These reports refute retrograde menstruation and suggest the existence of alternative mechanisms including coelomic metaplasia or Müllerian remnants. To better characterize the heterogeneity of Müllerian conditions, their anatomical features, and their association with endometriosis, we performed an extensive search for studies on co-occurring Müllerian anomalies and endometriosis. We classified Müllerian anomalies into three main categories (agenesis, fusion, resorption) reflecting the embryological processes of Müllerian duct development, and discussed the involvement of mesonephric anomalies inducing Müllerian abnormalities [ 24 ]. Where available, we reported prevalence, phenotypic variations of the reproductive tract and other systems, and proposed pathogenic mechanisms.