MRI markers of adenomyosis severity associated with worse IVF/ICSI outcomes

Adenomyosis is a benign gynaecological condition char - acterised by the infiltration of endometrial tissue and stroma into the myometrium of the uterus, causing disrup- tion in the so-called junctional zone (JZ). 1 The prevalence of adenomyosis is unclear due to lack of consensus in diagnostic method and criteria, with reported prevalence varying widely from 5 to 70%. 2,3 Younger nulliparous women are being more frequently diagnosed with adeno- myosis and it is increasingly being linked to poor obstetric outcomes and infertility. 4–7 A recent meta-analysis showed MRI markers of adenomyosis severity associated with worse IVF/ICSI outcomes Connie Odette Rees1,2,3 , Sehriban Kocyigit1 , Joost Nederend4, Massimo Mischi2, Hubertus A. A. M. van Vliet1,3 and Benedictus Christiaan Schoot1,2,3

Study objective: The aim of this study was to characterise the severity of adenomyosis on MRI in infertile women, and to assess if MRI characteristics of adenomyosis severity are associated with worse IVF/ICSI pregnancy outcomes versus male infertility controls.

This single-centre retrospective study was carried out at Catharina Hospital in Eindhoven, The Netherlands. The MRIs of 124 infertile women undergoing their first, fresh embryo transfer during IVF/ICSI, diagnosed with adenomyosis only (N = 31), or combined adenomyosis and endometriosis (N = 93) were assessed. Measurements of MRI adenomyosis features were performed by two independent investigators. IVF/ICSI outcomes (biochemical pregnancy (BP), ongoing pregnancy (OP) and live birth (LB)) of adenomyosis patients were compared to those of 889 male infertility controls.

Patients with adenomyosis had significantly worse IVF/ICSI outcomes compared to male infertility controls. When assessing individual MRI parameters, adenomyosis patients with a mean junctional zone (JZ) of > 12 mm, a JZ/Myometrium ratio of > 40%, presence of myometrial cysts and presence of endometriosis (specifically deep invasive endometriosis(DIE)) showed statistically significantly worse outcomes compared to patients with milder disease.

The results of this retrospective study suggest that individual MRI markers for severe adenomyosis (mean JZ > 12 mm, myometrial cysts), especially when combined with (severe) endometriosis, may be associated with fewer pregnancies during IVF/ICSI when compared to male infertility controls. Future prospective studies should investigate the prognostic potential of these markers for prediction of IVF/ICSI success.

Adenomyosis, infertility, assisted reproductive technologies, magnetic resonance imaging, pregnancy Date received: 13 January 2023; accepted: 29 July 2023 1 Department of Gynaecology and Obstetrics, Catharina Hospital, Eindhoven, North Brabant, the Netherlands 2 Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, North Brabant, the Netherlands 3 Department of Reproductive Medicine, University Hospital Ghent, Ghent, Belgium 4 Department of Radiology, Catharina Hospital, Eindhoven, North Brabant, the Netherlands Corresponding author: Connie Odette Rees, Department of Gynaecology and Obstetrics, Catharina Hospital, Michelangelolaan 2, Eindhoven, North Brabant 5623EJ, the Netherlands. Email: [email protected] 1195404PEV0010.1177/22840265231195404Journal of Endometriosis and Pelvic Pain DisordersRees et al. research-article2023 Original Research Article Rees et al. 135 detrimental effects of adenomyosis on in vitro fertilisation (IVF) outcomes, with significantly reduced implantation, clinical pregnancy, ongoing pregnancy and live birth in adenomyosis patients. 8 Diagnosis of adenomyosis Conventionally, the diagnosis of adenomyosis was obtained histologically from hysterectomy specimens, and this remains the gold standard. 9 With the advent of improved imaging techniques, the diagnosis can also be made via trans-vaginal ultrasound (TVUS, sensitivity 78%, specific- ity 78%, positive likelihood ratio of 3.5 and a negative like- lihood ratio of 0.28) and magnetic resonance imaging (MRI, sensitivity of 78%, specificity of 88%, a positive likelihood ratio of 6.8 (4.5–10%), and a negative likelihood ratio of 0.25). 10,11 TVUS is arguably less reliable for diag- nosing adenomyosis as it is relatively operator dependent.10 Furthermore, distinguishing adenomyosis from other uter- ine disorders such as leiomyomas or carcinomas can be dif- ficult on TVUS. MRI is therefore often the preferred diagnostic method, specifically in atypical or mild cases of adenomyosis. 9,12 Unfortunately, in contrast to TVUS, which has clear diagnostic criteria (the MUSA criteria13), there are no accepted diagnostic criteria for adenomyosis for MRI. The most widely reported MRI criteria are based on the appearance of the JZ, by looking at the following three fea- tures: (i) a JZ thickness ⩾12 mm; (ii) a ratio of greater than 40% of JZ to myometrium and (iii) a difference greater than 5 mm between the maximum and minimum JZ diameter. There are further reported indirect and direct criteria for adenomyosis (with presence of myometrial cysts seeming most promising) on MRI but their diagnostic and clinical potential remains unclear. 14,15 Junctional zone and infertility Alterations in the JZ have been linked to fertility, as the JZ is influenced by cyclical hormonal changes in accordance with the endometrium. 16,17 The JZ is believed to play an important in role in uterine contractions which are crucial for spermatozoa transport and embryo implantation. 18 A handful studies have specifically investigated whether changes in the JZ could be linked with fertility outcomes. 19,20 Limited studies have investigated the direct link of the type and the severity of adenomyosis to fertility outcomes how- ever, with those that have showing conflicting results. A study by Tamura et al. 21 showed that women with diffuse adenomyosis had worse fertility outcomes. Conversely, a study by Exacoustos et al. 22 showed that focal adenomyosis was more often associated with infertility. By extensively characterising adenomyosis on MRI, the burden of disease could perhaps be definitively correlated with fertility out- comes and thereby inform clinical decision making. Therefore, the aim of this study was to retrospectively quantify and characterise the extent of adenomyosis on MRI in infertile women undergoing IVF/ICSI, and to eval- uate if certain MRI characteristics of adenomyosis severity show worse IVF/ICSI outcomes compared to controls.

Study design and setting This single-centre retrospective case-control study was conducted at the Catharina Hospital in Eindhoven, the Netherlands, a regional referral centre for fertility and endometriosis treatment. Patients were included between the years of 2007 and 2020. This study was ethically approved by the local institutional review board and the regional Medical Ethical Committee with study number nWMO-2020.005/W20.045, in March 2020. Eligibility criteria IVF/ICSI patients between the ages of 18 and 42 years, undergoing their first, fresh embryo transfer in our centre between 2008 and 2020 were eligible. Study population IVF/ICSI patients that received an MRI at our hospital (according to local MRI protocol, see Supplemental Appendix 1) on suspicion of adenomyosis and/or endome- triosis were chosen as our study group. In order to confirm the initial diagnosis of adenomyosis and/or endometriosis of the MRI’s that were conducted, a reassessment was made of these MRI’s by pelvic radiologists and a study investigator (CR). The diagnosis of adenomyosis on MRI was made based on one of three criteria: (I) JZ thickness ⩾12 mm on T2 either focally or diffusely, (II) the presence of high signal intensity foci (HSI) in the myometrium on T1 and/or T2 concordant with an adenomyotic cyst, (III) JZ/myometrium ratio of >40% on T2. The diagnosis for endometriosis was based on the presence of one of the fol- lowing criteria; (I) hyperintense (multiple) ovarian cysts on T1 and hypointense intensity on T2, (II) endometriosis plaques and (III) deep infiltrating endometriosis. Patient medical files were then assessed in order to identify which of these patients had undergone IVF/ICSI treatment and received their first fresh embryo transfer (ET) in our fertil- ity department. Patients were included regardless of the timing of the MRI in relation to IVF/ICSI treatment. Control population The control group included women undergoing IVF/ICSI treatment due to a male factor only, with normal uteri on TVUS, or MRI where available. We chose only to include controls on the basis of normal uterus on imaging and only a male factor to minimise the chance of including undiag- nosed adenomyosis patients into the control group. 136 Journal of Endometriosis and Pelvic Pain Disorders 15(3-4) Exclusion criteria Patients who did not undergo ET, or only underwent fro- zen ET were excluded. Patients who explicitly objected to the usage of their medical data for research purposes were also excluded. IVF/ICSI treatment protocol Included patients had to meet the local eligibility require- ments of IVF/ICSI treatment protocol (see Supplemental Appendix 2). Patients first received pituitary downregulation with a recombinant GnRH agonist (Decapeptyl®, Ferring Gmbh, Germany), followed by ovarian stimulation. For ovarian stimulation either recombinant follicle stimulating hormone (Gonal-F®, Merck B.V . the Netherlands) or human menopausal gonadotrophin (Menopur®, Ferring B.V . the Netherlands, Fostimon®, Goodlife Fertility B.V . the Netherlands) was used. Fertilisation of the oocytes occurred the same day as oocyte retrieval, either by IVF or ICSI depending on the patients’ medical indication. Three days after oocyte retrieval, ET (single or double) was carried out after administration of a human gonadotrophin (HCG, Pregnyl®, Merck Sharp & Dohme, Canada) boost. According to the local and alpha scoring criteria (see Supplemental Appendix 3) selection of the best quality embryos for trans- fer was carried out. Luteal support was maintained with intravaginal progesterone (Uterogestan®, Besins Healthcare, the Netherlands) and was initiated after ET. Pelvic magnetic resonance imaging The standard MRI protocol for pelvic examinations at this hospital included the following sequences; T2-weighted turbo spin echo (T2-TSE) sequences in the sagittal, axial and coronal planes, as well as T1-weighted turbo spin echo (T1-TSE) sequences in the axial plane. All scans were car- ried out with either a 1.5T or 3T MRI system (Phillips, Ingenia, the Netherlands). In order to minimise the effects of bowel motions/spasms and uterine peristalsis on image interpretation, all patients were administered an antispas- modic agent (1 mL of 20 mg/mL Buscopan®, Sanofi, Paris, France) intravenously or intramuscularly. The slice thick- ness used was generally 3 mm, with slight variations rang- ing from 3 to 5 mm. Minor changes existed in the protocol throughout the years, but had no significant impact on the diagnostic quality of the MRIs. In case of patients receiv- ing multiple MRI’s, the MRI performed closest to IVF/ ICSI treatment was chosen for measurements. Full details can be found in Supplemental Appendix 1. MRI measurements The MRI features assessed with a brief definition, unit, calculation and stratification can be found in Table 1, and an illustration of measurements taken shown in Figure 1. Measurements were performed independently by two study investigators (COR and SK) using Sectra IDS7 ver - sion 21.1 (Linköping, Sweden). The investigators’ meas- urements were subsequently compared and measurements were considered equal when there was a difference ⩽1 mm. A pelvic radiologist was consulted when doubts presented about the performed measurements. For all measurements, the junctional zone was defined as a low signal intensity region between the high signal intensity region of the endometrium and the intermediate signal intensity region of the outer myometrium on T2. In case of an ill-defined junctional zone which inhibited accurate measurement, the MRI was labelled as having ‘poor JZ definition’. Focal adenomyosis was defined as focal widening of the JZ, or an ill-defined low signal intensity region of the uterine wall. In order to avoid mistaking uterine contractions for focal lesions, the precise location of the low signal inten- sity region was assessed in three directions (using a locali- sation cursor). When this low intensity region was not seen in other directions, it was categorised as a uterine contrac- tion. Diffuse adenomyosis was defined in case of a diffuse thickening of the JZ of ⩾12 mm showing a low T2 signal intensity with indistinct margins (Figure 1(c)). Outcomes The primary outcomes of this study included: (I) biochem- ical pregnancy (BP, a positive HCG test on day 16 post- ET), (II) ongoing pregnancy (OP, presence of a foetal cardiac activity on ultrasound 11 weeks after ET), (III) live birth (LB, delivery of a viable foetus >24 weeks of gesta- tional age). Secondary outcomes included the MRI charac- teristics of adenomyosis as shown in Table 1. Data sources and management Patient characteristics, radiology reports and MRI data were retrieved from the electronic hospital patient records programme HIX (Chipsoft 6.1, Amsterdam, The Netherlands). Data concerning the IVF and ICSI cycles was retrieved from the Dutch national fertility database (Landelijk Specialistisch Fertiliteits Dossier, LFSD, Stichting Automatisering Fertiliteit (SAF), Utrecht, the Netherlands)). All patient data and MRI measurements were recorded in a secure electronic database (Research Manager version 5.53 (Cloud9 software, Deventer, the Netherlands)) and were later exported to IBM SPSS statis- tics (version 27) for data analysis. Statistical analysis The Shapiro-Wilk test was applied to assess normal distri- bution of data. Normally distributed data were presented Rees et al. 137 as mean ± standard deviation (SD) and in case of non- normally distributed data as median (interquartile range). Between-group differences were assessed using the inde- pendent T-test or Mann-Whitney U test for the continuous variables, and for categorical variables a Chi-squared test or Fisher’s exact test was performed (with Bonferroni cor- rection). Subsequently, a multivariate logistic regression analysis for IVF/ICSI outcomes was carried out correct- ing for age at IVF, IVF or ICSI treatment, number of transferred embryos and embryo quality, leading to adjusted odds ratios (aOR). with 95% confidence interval (95% CI). Overall, a p -value of < 0.05 was considered significant.

Patient inclusion and characteristics One hundred twenty-four women with MRI-diagnosed adenomyosis were included (see Figure 2). Thirty-one women had only adenomyosis and 93 women had both adenomyosis and endometriosis. Eight hundred eighty- nine patients undergoing IVF/ICSI treatment due to male factor only were included in the control group. Table 2 shows demographic and IVF/ICSI treatment characteris- tics of both groups. Controls more often underwent ICSI treatment (80.1 vs 25.0%, p < 0.001) and also had more viable oocytes (8.0 vs 7.0, p = 0.015) and embryos (5.0 vs 4.0, p < 0.001) ver- sus adenomyosis patients. Embryo quality was similar between groups (p = 0.112), but did differ (p < 0.001) when looking only at the second transferred embryo where applicable. MRI characteristics A summary of all MRI characteristics can be seen in Table 3. Figure 3 shows several illustrative examples of adenomyosis. Thirty-six women received an MRI in advance of their first IVF/ICSI cycle, of which twelve women achieved pregnancy, and 24 did not. Seventy- eight women underwent MRI after fertility treatment, of Table 1. MRI characteristics of adenomyosis: objective adenomyosis MRI features, based on Rees et al. 14 MRI feature Definition Unit Stratification Average Junctional Zone thickness (AJZ) Mean of JZ measurement at 6 points of the uterus: Fundus, Mid-corpus, Isthmus, measuring the anterior and posterior wall at each point in the mid-sagittal plane. mm >7 mm, >10 mm, >12 mm Maximal Junctional Zone Thickness (JZMax) Maximal diameter of JZ from those measured at the 6 points as described above, with location mm >7 mm, >10 mm, >12 mm Minimal Junctional Zone Thickness (JZMin) Minimal diameter of JZ from those measured at the 6 points as described above mm Junctional Zone Differential (JZDiff) As a measure of JZ irregularity Difference between maximal and minimal JZ mm >5 mm Junctional Zone Asymmetry (JZAsymm) Difference between anterior and posterior JZ (based on measurement at 6 points previously described) mm Junctional Zone to Myometrium Ratio (JZ/Myo Ratio) Ratio of Junctional zone to full myometrium thickness (measured at 6 points previously described) % >40% Uterine volume Uterine length × width × height × 0.523. 23 The length of measured in the sagittal plane from the outer ostium of the cervix until the fundus. Width was measured in the axial plane, and height in the transverse plane at the mid-corpus. mm 3 Average Uterine Wall Thickness Uterine wall thickness measured from endometrium to myometrium, at 6 points previously described mm Adenomyotic foci volume Volume of adenomyotic foci in 3 orientations Calculated using the formula of a sphere: 4 3 .π.r3 mm3 60 mm3 (Number of) HSI adenomyotic foci (Myometrial Cysts) Visible high signal intensity (HSI) myometrial foci (compared to normal myometrium) on T1 or T2- weighted imaging a >5 Adenomyosis Signal intensity ratio (SIR) Signal intensity ratio of adenomyotic tissue compared to that of the rectus muscle on T2 imaging (as measured using Region of Interest (ROI) circles). aOn the total MRI scan, not per image slice. Duplicate counting foci in various slices was avoided by tracking lesions across slices using the localisa- tion cursor. 138 Journal of Endometriosis and Pelvic Pain Disorders 15(3-4) Figure 1. Measurements on MRI of Adenomyosis characteristics. All MRI images shown were performed on a T2-TSE. Figure a, b and c are shown in a sagittal plane, d in the transverse plane. (a) The JZ thickness (yellow line) and the myometrium thickness (red line) measured at the level of fundus (F), mid-corpus (M) and isthmus (I) in the anterior (A) and posterior (P) wall of the uterus. (b) The measurements of JZ thickness and myometrium thickness. (c) A uterus with diffuse adenomyosis with a JZ ⩾ 12 mm. (d) Uterus volume determination using the width in the transverse plane (78.8 mm). JZ is showing high signal intensity foci’s determined with dashed yellow circles. Volume of focal adenomyosis shown in the red circle. Rees et al. 139 which nineteen women became pregnant and 59 did not. There was no significant difference in terms of pregnancy observed between the group with MRI prior to fertility treatment and the group that had it afterwards (p = 0.83, Supplemental Table S4). Most women (83/124) had an MRI within 5 years of fertility treatment. No significant difference in terms of pregnancy rate was found when comparing these women to those received an MRI out- side this time-frame (p = 0.91, Supplemental Table S4). IVF/ICSI Outcomes A complete overview of IVF/ICSI outcomes for the adeno- myosis versus controls is seen in Supplemental Table S6. Overall, patients with MRI-diagnosed adenomyosis showed significantly fewer biochemical and ongoing preg- nancies and live births versus controls in crude analysis (25 vs 36.3%, p = 0.013, 15.6 vs 29.4%, p = 0.001, and 14.0 vs 26.8%, p = 0.009, respectively). A sub-analysis of Table 2. Patient characteristics of IVF/ICSI patients with MRI-diagnosed adenomyosis versus male infertility controls. Adenomyosis (N = 124) Control Group (N = 889) p-Value BMI in kg/m2 (Median, IQR) 23.95 (21.30–28.33) 23.60 (21.44–27.04) 0.48 Infertility time in months (Median, IQR) 35.00 (23.50–53.50) 27.00 (20.00–41.00) 0.11 Age during 1st IVF cycle (years, Median, IQR) 33.00 (30.00–35.50) 31.00 (28.00–35.00) 0.042 Year of IVF Treatment (Median, (IQR)) 2011 (2008–2016) 2011 (2009–2014) 0.54 Cycle length (days, Median, IQR) 28.00 (28.00–30.00) 28.00 (28.00–30.00) 0.43 Primary subfertility 79 (63.7%) 616 (69.4%) 0.92 Secondary subfertility 36 (29.0%) 272 (30.6%) Indication for fertility treatment Male factor 20 (16.3%) 889 (100%) <0.001 Female factor§ 13 (10.6%) Combined 26 (21.1%) Endometriosis 37 (30.1%) Idiopathic 27 (22.0%) Type of treatment (N (%)) IVF 93 (75.0) 177 (19.9) <0.001 ICSI 31 (25.0) 712 (80.1) Number of viable oocytes (Median, IQR) 7.0 (4.0–10.0) 8.0 (5.0–12.0) 0.015 Number of viable embryos (Median, IQR) 4.0 (2.0–6.0) 5.0 (3.0–8.0) <0.001 Fertilisation rate (%, Median, IQR) 60.0 (45.0–75.0) 57.1 (40.0–72.0) 0.21 Number of embryos transferred (N (%)) Single 83 (68.0) 570 (64.2) 0.42 Double 39 (32.0) 318 (35.8) Embryo quality of first transferred embryo (N (%)) a Super 28 (23.5) 298 (33.7) 0.11 Good 19 (16.0) 138 (15.6) Fair 48 (40.3)* 314 (50.6) Moderate 22 (18.5) 133 (14.9) Poor 2 (1.7) 33 (3.7) Embryo quality of second transferred embryo (N (%))a Super 2 (5.7)* 0 (0.0) <0.001 Good 8 (22.9) 63 (20.1) Fair 19 (54.3) 217 (69.1) Moderate 5 (14.3)* 0 (0.0) Poor 1 (2.9)* 34 (10.8) *An asterisk denotes statistically significant difference versus control. aSee Supplemental Appendix 3, Table S3 for details on embryo quality criteria. §Female factor infertility included indications such as: ovulation disorders, tubal factor and cervical issues as IVF/ICSI indications. Figure 2. Flowchart of patient selection and inclusion. 140 Journal of Endometriosis and Pelvic Pain Disorders 15(3-4) patients with only fair-to-super quality embryos was also carried out (see Supplementary File, Table S5), with com- parable results (p = 0.010, p = 0.002 and p = 0.014, respec- tively in crude analysis). Subsequently, IVF/ICSI outcomes for patients with markers for adenomyosis or endometriosis severity were compared to controls (see Supplemental Table S6 for full results). Table 4 presents a sub-analysis of adenomyosis MRI markers and reports the live birth rate (LBR) and adjusted odds ratio’s for LB versus male infertility controls cor - rected for age at IVF, embryo quality, number of trans- ferred embryos, and year of treatment. Results for BP and OP are shown in Supplemental Table S7. Adenomyosis patients overall showed statistically significantly fewer LB compared to controls (aOR 0.560 (95% CI 0.318–0.988, p = 0.045). The sub-analysis for patients with only fair-to- super quality embryos showed comparable results (aOR 0.457 (95% CI 0.239–0.877, p = 0.018) Furthermore, the MRI markers of myometrial cysts, JZ-Diff >5 mm, JZ/ Myometrium ratio >40% and added presence of endome- triosis remained statistically significantly associated with fewer LB versus controls.

Previous studies have suggested that adenomyosis nega- tively affects reproductive outcomes, however, a lack of consensus in diagnostic criteria on MRI makes the relation- ship between disease severity and IVF/ICSI outcomes unclear. Hence, we investigated known MRI markers of adenomyosis severity in relation to IVF/ICSI outcomes. Our study showed a wide range of adenomyosis characteristics in infertile women, reflecting the varied nature of the dis- ease, and highlighting the challenges in its diagnosis and clinical presentation. Results showed that within adenomy- osis (and endometriosis) patients, patients with certain MRI markers (namely concomitant endometriosis, myometrial cysts, JZ Diff >5 mm and/or a JZ/Myometrium ratio >40%) exhibited significantly worse IVF/ICSI versus male infertil- ity controls (p < 0.05). These findings were confirmed when correcting for confounders in multivariate analysis. Despite the fact that there are limited comparable stud- ies, it can be said that our results are consistent with the current literature. In a study by Meylaerts et al., 19 a thick- ened AJZ and JZmax on MRI were associated with infer - tility. A similar study by Maubon et al. 20 examined the influence of JZ thickness in infertile women on implanta- tion rates during IVF, and showed that a thickened JZ was a negative predictor for embryo implantation. This study also investigated JZ cut-offs and showed a implantation failure rate of 95.8% for patients with an AJZ> 7 mm and a JZmax >10 mm compared to patients with a smaller JZ (p < 0.0001). In our study on the other hand, a higher JZ cut-off of 12 mm was significantly associated with worse IVF/ICSI outcomes. The majority of women in our popu- lation already had a relatively thickened junctional zone due to the presence of adenomyosis, so the threshold prof- fered in the aforementioned study may well not be applica- ble to our population. Table 3. MRI characteristics of IVF/ICSI patients with MRI- diagnosed adenomyosis. Adenomyosis patients (n = 124) Age at MRI (years, mean, SD) 34.96 (5.46) MRI conducted prior to IVF/ICSI treatment (N (%)) 46 (37.1) MRI within 5 years of fertility treatment (N (%)) 83 (66.9) Adenomyosis type Focal 58 (47.9) Diffuse 31 (25.6) Cystic 5 (4.1) Focal and cystic 17 (14.0) Diffuse and cystic 10 (8.3) Missing 3 (2.4%) Average JZ (mm, mean, SD) 8.68 mm (3.57) >7 mm 89 (71.8) >10 mm 41 (33.1) >12 mm 20 (16.1) >15 mm 9 (7.3) Maximal JZ (mm, mean, SD) 17.05 mm (8.70) >7 mm 115 (92.7) >10 mm 104 (83.4) >12 mm 94 (77.4) >15 mm 68 (54.8) JZ differential (mm, mean, SD) 13.36 mm (8.68) >5 mm 110 (88.7) Average JZ/myometrium ratio (mean, SD) 0.46 (0.14) >40% 92 (74.2) JZ asymmetry (mm, mean, SD) 0.07 mm (0.81) >2 mm 22 (17.7) Presence of high signal intensity foci (myometrial cysts) 60 (48.8) >5 HSI foci 24 (19.4) T1-high signal HSI Foci 24 (19.4) Uterine length, in sagittal direction (mm, median, IQR) 78.90 (18.50) Maximal focal lesion (mm, median, IQR) 23.30 (13.0) Presence of endometriosis (N (%)) 93 (75.0) Presence of plaques 86 (69.4) Endometriomas 62 (50.0) Deep invasive endometriosis (DIE) 26 (21.0) Presence of fibroids (N (%)) Yes 5 (4.2) Rees et al. 141 A handful of recent studies have investigated individ- ual adenomyosis MRI characteristics and IVF/ICSI out- comes.24–26 A study by Iwasawa et al.25 found that patients with the extrinsic adenomyosis subtype had better fertility outcomes compared to other adenomyosis subtypes. Our study did not find a clear difference in IVF/ICSI outcomes between adenomyosis subtypes however. A possible explanation for this lies in the diverse categorisations of adenomyosis that exist, making consensus of certain MRI markers difficult. 9 Our finding that the added presence of (deep invasive) endometriosis affects IVF/ICSI preg- nancy outcomes has been described before. 22,27,28 One recent study also found that women with combined aden- omyosis (irrespective of subtype) and endometriosis on MRI had fewer live births compared to endometriosis alone. 26 Bourdon et al. 26 additionally reported a signifi- cantly lower live birth rate in women with adenomyosis and endometriosis exhibiting myometrial cysts. Our study showed a similar relationship between live birth rate in relation to myometrial cysts (aOR 0.420, p = 0.049). Overall therefore, our results support that adenomyosis in combination with (extensive) endometriosis could be seen as a more severe form of disease, and that these patients may form a specific subgroup potentially needing specific treatment protocols. Strength and limitations Our study has several strengths. First, the fact that the measurements were performed by two independent study investigators, reduces information bias. We also re- assessed all included MRIs during the study instead of relying on radiology report, accounting for differences in adenomyosis diagnosis over time and thereby increasing the internal consistency of our data. Correcting our results for relevant IVF/ICSI confounders also increases the reli- ability of our findings. This study admittedly has limitations. First, despite our sample size being comparable to previous studies, the number of absolute pregnancies achieved in our study group is low (only 14% live births in patients with adeno- myosis), reducing the power of the results. This is also reflected in the larger confidence intervals in the multivari- ate logistic regression analysis. It is possible that our study population represents a group of women with more severe disease (due to their infertility and indication for MRI in Figure 3. Examples of adenomyotic uteri in the study population: (a) uterus with focal adenomyosis in the posterior wall, (b) uterus with cystic adenomyosis in the anterior wall and a diffusely widened JZ, (c) enlarged uterus with a myoma in the anterior wall and diffuse adenomyosis with hyperintense foci in the JZ, and an ovarian endometrioma, and (d) uterus with various myomas as well as a diffusely enlarged JZ with scattered hyperintense foci (myometrial cysts). 142 Journal of Endometriosis and Pelvic Pain Disorders 15(3-4) the first place), introducing an element of selection bias, and adding to the low pregnancy rate. Additionally, most MRI diagnoses of adenomyosis were made after IVF/ICSI treatment (78/124 patients). One could question whether the adenomyosis was present to a similar extent at the time of fertility treatment. We do not believe this to be a rele- vant issue however, as a sub-analysis based on the timing of the MRI in relation to IVF/ICSI treatment did not affect the results. Moreover, adenomyosis is known to be a disor- der that develops over a lifetime, and can be assumed to be present throughout the reproductive life-phase. 29 Another element of our study to consider when interpreting our findings is the choice of control group. The majority of our control group did not undergo MRI, which means that the presence of adenomyosis in this group cannot be com- pletely excluded, despite normal TVUS findings and lack of clinical adenomyosis symptoms. It is possible therefore that there are some undiagnosed adenomyosis patients in the control group, which may affect the final analysis. Furthermore, patients in our study cohort could have had other indications for infertility treatment in addition to endometriosis/adenomyosis, whereas our control group in theory only had male infertility. This inevitably introduces a further element of bias into our case group. Clinical and future implications Our results support that specific MRI markers of adeno- myosis and endometriosis severity may be associated with worse IVF/ICSI outcomes compared to male infertility controls. In this context, there is arguably value in thor - oughly assessing severity and extent of adenomyosis pre- conceptionally, especially when in combination with endometriosis. Detailed mapping of adenomyosis on MRI may improve clinical counselling and management of adenomyosis and/or endometriosis patients considering IVF/ICSI. Our data shows that the lower IVF/ICSI preg- nancy rates are mainly seen in patients with combined adenomyosis and endometriosis, with adenomyosis alone seemingly not enough to cause convincingly worse fertil- ity outcomes in our study population. If adenomyosis and endometriosis are seen as a spectrum of the same disease, the combined diseases constitute more severe disease, and thus have a greater impact on reproductive ability. We did not include patients with only endometriosis so could not assess its potentially confounding effect here. However, previous work by our group has shown that endometriosis alone has less effect on IVF/ICSI outcomes than combined disease. 27 Larger future studies with a prospective design should confirm these results, and aid in creating more per- Table 4. Live birth rate for adenomyosis severity MRI markers versus controls. Analysed subgroup (N) LBR adenomyosis cases (%) LBR control group (N = 889, %) aOR (95% CI)* p-Value** Adenomyosis patients with myometrial cysts (n = 60) 7 (12.1) 233 (26.8) 0.420 (0.177−0.997) 0.049 Adenomyosis patients without myometrial cysts (n = 63) 10 (16.1) 0.556 (0.255−1.214) 0.14 Adenomyosis patients with mean JZ 12 mm (N = 20) 2 (10.5) 0.374 (0.082−1.700) 0.20 Adenomyosis patients with JZ-diff 5 mm (N = 110) 16 (15.0) 0.519 (0.273−0.990) 0.046 Adenomyosis patients with JZ- myometrium 40% (N = 92) 12 (13.3) 0.453 (0.222−0.921) 0.029 Diffuse adenomyosis (N = 31) 5 (16.7) 0.583 (0.209−1.201) 0.30 Focal adenomyosis (N = 58) 9 (15.5) 0.525 (0.230−1.626) 0.13 Adenomyosis alone (n = 31) 5 (17.2) 0.652 (0.225−1.889) 0.43 Adenomyosis and endometriosis (n = 93) 12 (14.1) § 0.440 (0.219−0.886) 0.021 Adenomyosis without DIE (n = 98) 15 (15.8) 0.542 (0.280−1.050) 0.070 Adenomyosis with DIE (n = 26) 2 (7.7)§ 0.272 (0.061−1.212) 0.088 LBR: live birth rate; aOR: adjusted Odds Ratio; DIE: deep invasive endometriosis. *Multivariate logistic regression adjusted for: age at time of IVF, IVF or ICSI treatment, embryo quality, year of IVF treatment and number of trans- ferred embryos. **p-Value for logistic regression analysis. §Denotes p < 0.05 versus controls in crude analysis. Rees et al. 143 sonalised management and treatments for (infertile) aden- omyosis patients.

This study assessed a number of MRI parameters (added presence of endometriosis (DIE), mean JZ >12 mm, mean JZ/Myometrium ratio of >40% and presence of myome- trial cysts) that could function as markers for IVF/ICSI outcomes and aid in counselling patients prior to starting treatment. We believe further (prospective) research should be encouraged. Mapping out the severity and the extent of adenomyosis and endometriosis in correlation to further clinical (fertility) outcomes could aid in clinical management of (infertile) women with the disease. Declaration of conflicting interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding The authors received no financial support for the research, authorship, and/or publication of this article. ORCID iDs Connie Odette Rees https://orcid.org/0000-0001-7742-3050 Sehriban Kocyigit https://orcid.org/0000-0002-8318-1163 Supplemental material Supplemental material for this article is available online.

1. Campo S, Campo V and Benagiano G. Adenomyosis and infertility. Reprod Biomed Online 2012; 24(1): 35–46. 2. Graziano A, Lo Monte G, Piva I, et al. Diagnostic findings in adenomyosis: a pictorial review on the major concerns. Eur Rev Med Pharmacol Sci 2015; 19: 1146–1154. 3. Aleksandrovych V, Basta P and Gil K. Current facts consti- tuting an understanding of the nature of adenomyosis. Adv Clin Exp Med 2019; 28(6): 839–846. 4. Puente JM, Fabris A, Patel J, et al. Adenomyosis in infer- tile women: prevalence and the role of 3D ultrasound as a marker of severity of the disease. Reprod Biol Endocrinol 2016; 14: 60. 5. Bruun MR, Arendt LH, Forman A, et al. Endometriosis and adenomyosis are associated with increased risk of preterm delivery and a small-for-gestational-age child: a systematic review and meta-analysis. Acta Obstet Gynecol Scand 2018; 97(9): 1073–1090. 6. Horton J, Sterrenburg M, Lane S, et al. Reproductive, obstetric, and perinatal outcomes of women with adeno- myosis and endometriosis: a systematic review and meta- analysis. Hum Reprod Update 2019; 25(5): 592–632. 7. Maheshwari A, Gurunath S, Fatima F, et al. Adenomyosis and subfertility: a systematic review of prevalence, diagnosis, treatment and fertility outcomes. Hum Reprod Update 2012; 25(5): 592–632. 8. Nirgianakis K, Kalaitzopoulos DR, Schwartz ASK, et al. Fertility, pregnancy and neonatal outcomes of patients with adenomyosis: a systematic review and meta-analysis. Reprod Biomed Online 2021; 42: 185–206. 9. Chapron C, Vannuccini S, Santulli P, et al. Diagnosing adenomyosis: an integrated clinical and imaging approach. Hum Reprod Update 2020; 26(3): 392–411. 10. Champaneria R, Abedin P, Daniels J, et al. Ultrasound scan and magnetic resonance imaging for the diagnosis of adeno- myosis: systematic review comparing test accuracy. Acta Obstet Gynecol Scand 2010; 26(3): 392–411. 11. Tellum T, Nygaard S and Lieng M. Noninvasive diagnosis of adenomyosis: a structured review and meta-analysis of diagnostic accuracy in imaging. J Minim Invasive Gynecol 2019; 27(2): 408–418.e3. 12. Vannuccini S and Petraglia F. Recent advances in under- standing and managing adenomyosis. F1000Research 2019; 8: F1000 Faculty Rev-283. 13. Van den Bosch T, de Bruijn AM, de Leeuw RA, et al. Sonographic classification and reporting system for diag- nosing adenomyosis. Ultrasound Obstet Gynecol 2019; 53(5): 576–582. 14. Rees CO, Nederend J, Mischi M, et al. Objective measures of adenomyosis on MRI and their diagnostic accuracy—a systematic review & meta-analysis. Acta Obstet Gynecol Scand 2021; 100(8): 1377–1391. 15. Munro MG. Classification systems for adenomyosis. J Minim Invasive Gynecol 2019; 27(2): 296–308. 16. Agostinho L, Cruz R, Osório F, et al. MRI for adenomyosis: a pictorial review. Insights Imaging 2017; 8: 549–556. 17. Meylaerts LJ, Wijnen L, Grieten M, et al. Junctional zone thickness in young nulliparous women according to men- strual cycle and hormonal contraception use. Reprod Biomed Online 2017; 34(2): 212–220. 18. Brosens I, Derwig I, Brosens J, et al. The enigmatic uterine junctional zone: the missing link between reproductive dis- orders and major obstetrical disorders? Hum Reprod 2010; 25: 569–574. 19. Meylaerts LJ, Wijnen L, Ombelet W, et al. Uterine junc- tional zone thickness in infertile women evaluated by MRI. J Magn Reson Imaging 2017; 45(3): 926–936. 20. Maubon A, Faury A, Kapella M, et al. Uterine junctional zone at magnetic resonance imaging: a predictor of in vitro fertilization implantation failure. J Obstet Gynaecol Res 2010; 36(3): 611–618. 21. Tamura H, Kishi H, Kitade M, et al. Clinical outcomes of infertility treatment for women with adenomyosis in Japan. Reprod Med Biol 2017; 16(3): 276-282. 22. Exacoustos C, Lauriola I, Lazzeri L, et al. Complications during pregnancy and delivery in women with untreated rectovaginal deep infiltrating endometriosis. Fertil Steril 2016; 106(5): 1129–1135.e1. 23. Kitamura Y, Allison SJ, Jha RC, et al. MRI of adenomyosis: changes with uterine artery embolization. Am J Roentgenol 2006; 186(3):855–864. 144 Journal of Endometriosis and Pelvic Pain Disorders 15(3-4) 24. Khandeparker MS, Jalkote S, Panpalia M, et al. High- resolution magnetic resonance imaging in the detection of subtle nuances of uterine adenomyosis in infertility. Glob Reprod Heal 2018; 3: e14. 25. Iwasawa T, Takahashi T, Maeda E, et al. Effects of locali- sation of uterine adenomyosis on outcome of in vitro fertilisation/intracytoplasmic sperm injection fresh and fro- zen-thawed embryo transfer cycles: a multicentre retrospec- tive cohort study. Reprod Biol Endocrinol 2021; 19: 1–11. 26. Bourdon M, Santulli P, Bordonne C, et al. Presence of aden- omyosis at MRI reduces live birth rates in ART cycles for endometriosis. Hum Reprod 2022; 37: 1470–1479. 27. Rees CO, Rupert IAM, Nederend J, et al. Women with com- bined adenomyosis and endometriosis on MRI have worse IVF/ICSI outcomes compared to adenomyosis and endome- triosis alone: a matched retrospective cohort study. Eur J Obstet Gynecol Reprod Biol 2022; 271: 223–234. 28. Coccia ME, Rizzello F, Mariani G, et al. Impact of endome- triosis on IVF-ET cycles in young women: a stage depend- ent interference. Acta Obstet Gynecol Scand 2011; 90: 1232–1238. 29. Benagiano G, Brosens I and Habiba M. Adenomyosis: a life-cycle approach. Reprod Biomed Online 2015; 30: 220–332.