Impact of Endometriosis on the Progression of Inflammatory Bowel Diseases: A Multicenter Retrospective Study

Inflammatory bowel diseases (IBDs), encompassing Crohn's disease (CD) and ulcerative colitis (UC), are chronic and progressive intestinal disorders that affect quality of life and healthcare costs ( 1 , 2 ). Their pathophysiology is complex and involves microbial dysbiosis, genetic and environmental factors, and immune system dysregulation ( 3 ). IBDs were characterized by variable degrees of disease heterogeneity, including differences in disease location, progression, and therapeutic response ( 4 ). Endometriosis is also a chronic and heterogeneous inflammatory disease characterized by ectopic endometrial tissue located outside the uterus, typically in the abdomen and pelvis ( 5 ). Endometriosis affects 5%–15% of reproductive-age women and 5%–50% of infertile women ( 6 , 7 ). Its pathophysiology is complex and includes hormonal, immunological, and inflammatory factors. Retrograde menstruation, leading to the deposition of endometrial cells in the peritoneal cavity, may play a role. These ectopic endometrial cells proliferate under the effect of estrogen and promote immune system dysregulation ( 6 , 8 ). Dysfunctions of the innate and adaptive immune systems in the peritoneal environment include resistance of endometrial cells to phagocytosis and apoptosis, abnormalities in B and T lymphocytes, and elevated levels of proinflammatory cytokines such as tumor necrosis factor (TNF)-alpha ( 8 – 10 ). Digestive involvement in endometriosis represents the most common extrapelvic location, with an incidence of 4%–37% among women with endometriosis ( 11 ). Endometriosis and IBD are both immune-mediated inflammatory disorders, share similar symptoms (chronic abdominal pain, changes in bowel movements, rectal bleeding), and predominantly affect young patients, thereby impacting the gynecological and overall quality of life of affected women ( 12 ). Differential diagnosis between these 2 conditions can be difficult, and they can also coexist ( 11 – 15 ). Both significantly affect patients' quality of life and result in substantial cost ( 5 , 16 ). Few studies have evaluated the interaction between endometriosis and IBD ( 6 , 17 ). Our objective was to characterize the IBD phenotype in patients with coexisting endometriosis and to assess the impact of endometriosis on the course of IBD.

We conducted a retrospective, multicenter, exposed-unexposed cohort study across 18 Groupe d'Etudes Thérapeutiques des Affections Inflammatoires du tube Digestif centers. All female patients older than 18 years with a confirmed diagnosis of IBD were eligible for inclusion. We identified patients with IBD with a concomitant well-defined diagnosis of symptomatic endometriosis between June 2022 and March 2023. The diagnosis of endometriosis was based on clinical findings confirmed by gynecologist, laparoscopic visualization of lesions, or pelvic MRI. These patients were then matched to 1 or 2 patients diagnosed with IBD only. The matching was performed according to the following criteria: type of IBD (CD or UC), age at IBD diagnosis (±5 years), the year of IBD diagnosis (±4 years), and date of last visit (±2 years). Matched patients were identified from the Lille University Hospital IBD database (Integralis) and were extracted between April and July 2023. The patterns of care were consistent across participating centers and followed the latest recommendations for the management of IBD. This study was conducted in accordance with local ethical standards and legal requirements under the reference methodology MR-004 (registration number 2229845 v0). Patients were enrolled using an online registry. Investigators from each center were asked to complete an electronic Case Report Form. The recorded data included: age, date of last consultation, smoking status, date and age at IBD diagnosis, type of IBD, family history of IBD, presence of extraintestinal manifestations, history of IBD-related complications (hospitalization, stricture, fistula, intra-abdominal abscess), presence of perianal fistulizing disease, history of IBD-related surgery, and current and previous IBD medical therapies. CD and UC have been described according to the Montreal classification. For patients with endometriosis, the collected data included: age at menarche, family history of endometriosis, date of diagnosis, endometriosis-related symptoms, and mode of diagnosis (clinical, radiological, or surgical). Endometriosis lesions were classified according to 3 phenotypes ( 5 ): superficial peritoneal lesions, ovarian lesions (endometrioma), and deep infiltrating endometriosis (involving penetration into adjacent organs). Data on hormonal treatments (oral contraceptive pill including both combined or progestin-only, gonadotropin-releasing hormone) and surgical treatments related to endometriosis, were also collected. The impact of endometriosis on the IBD progression was assessed using a composite severity criterion, which included colectomy or colonic stricture for patients with UC, and bowel damage (stenosis, fistula or intra-abdominal abscess) or intestinal surgery for patients with CD. No specific weighting was assigned to each parameter, and the composite criterion was analyzed as a binary outcome, indicating the occurrence of at least one of these events. This composite severity criterion was previously used in a previous study ( 18 ). For the primary end point, all IBD severity events (as defined by the composite criterion) were considered in the statistical analyses. Secondary end points included 2 subgroup analyses: one evaluating patient whose endometriosis was diagnosed before IBD, and another evaluating patients who underwent surgical treatment of endometriosis. Qualitative variables were presented as percentages, whereas quantitative variables were expressed as mean (±SD) or median (±interquartile range) in case of non-Gaussian distribution. Normal distribution was assessed graphically using histograms and the Shapiro-Wilk normality test. Baseline characteristics of patients with IBD were compared between the exposed group (patients with endometriosis) and the unexposed group (patients without endometriosis) using the χ 2 test or the Fisher exact test. The composite severity criterion was described in both groups by the Kaplan-Meier method, taking the date of IBD diagnosis as the baseline and the 20-year follow-up as the end point. The effect of exposure (presence of endometriosis) was assessed using a marginal Cox model (model with a corrected variance-covariance matrix, known as the robust or sandwich estimator) to account for the matching between the 2 groups. The proportional hazards assumption was first evaluated using Schönefeld residuals. Hazard ratios (HRs) with 95% confidence interval (CI) derived from the Cox model were reported as the measure of effect size for the exposed vs unexposed groups. A sensitivity analysis was performed excluding patients (and the matching patients) whose endometriosis diagnosis occurred after their IBD diagnosis. Subgroup analyses according to the surgical status of endometriosis during follow-up were performed by including the corresponding interaction terms in the Cox model. Statistical analysis was conducted using SAS software (version 9.4; SAS Institute, Cary, NC). A 2-sided P value <0.05 was considered statistically significant.

Overall, 207 patients with IBD and endometriosis (CD n = 149; UC n = 58), and 409 matched patients with IBD (CD n = 296; UC n = 113) were included. All patients matched to 2 patients, except for 5 patients who were matched with 1 patient. Patient characteristics at inclusion are summarized in Table 1 . At IBD diagnosis, the median age was 23 years (18–29). The median disease duration was 10 years (5.75–17). Active smoking at inclusion was reported in 19% (39/205) of patients with endometriosis vs 24.5% (76/310) of patients without endometriosis ( P = 0.13). Patients with IBD and endometriosis had more extraintestinal manifestations, particularly articular manifestations, compared with patients without endometriosis (30.4% vs 10.5%; P < 0.001). A family history of IBD was more common in patients without endometriosis (28.9% vs 17.1%; P = 0.003). For patients with UC, a significant difference was observed in disease extent ( P = 0.026), with a higher proportion of proctitis in the endometriosis group (29.3% vs 12.6%) and more left-sided colitis in the nonexposed group (37.8% vs 27.6%). No significant differences were observed among patients with CD regarding disease location, upper gastrointestinal involvement, perianal disease, or disease behavior. Characteristics of IBD according to the diagnosis of endometriosis P values were calculated using marginal binary logistic regression models for binary outcomes or using mixed ordinal logistic regression for ordinal outcomes. A P value < 0.05 is considered statistically significant. Endometriosis (−), patients without endometriosis; Endometriosis (+), patients with endometriosis; IBD, inflammatory bowel disease; NA, not applicable; TNF, tumor necrosis factor. Clinical characteristics of endometriosis are summarized in Table 2 . The median age at menarche was 13 years (11.9–14). Approximately, 10% of patients had a familial history of endometriosis. In most cases, IBD was diagnosed before endometriosis (n = 151/200, 75.5%). Among the 207 patients with endometriosis, diagnosis was mainly based on pelvic MRI (73.9%, 153/207) or laparoscopic visualization (23.2%, 48/207). Endometriosis was classified as superficial peritoneal in 17.0% (33/194), ovarian in 32.0% (62/194), and deep infiltrating in 70.6% (137/194) of patients. Intestinal involvement was observed in 23.7% (46/194) of patients. The main symptoms at diagnosis were gynecological (dysmenorrhea: 72.2%, dyspareunia: 36.7%), and 23% of patients reported painful defecation during menstruation. Regarding endometriosis-related treatments, 65.2% (128/197) of patients received hormonal therapies and 47.3% (98/207) underwent surgical intervention, including bowel resection in 20 patients. Clinical details of endometriosis CD, Crohn's disease; GnRH, gonadotropin-releasing hormone; IBD, inflammatory bowel disease; IQR, interquartile range; UC, ulcerative colitis. Surgery was significantly less frequent in patients with CD and endometriosis compared with those with CD only (32.2% vs 43.9%; P = 0.011). The time to first surgery was longer in the endometriosis group (3.5 years [0–10] vs 3 years [1–7]). No differences were observed in the prevalence of perianal manifestations or in the need for perianal surgery between the 2 groups (Table 3 ). In the survival analysis, patients with CD and endometriosis had a less severe disease course compared those without endometriosis (HR 0.68, 95% CI 0.50–0.92, P = 0.011) (Figure 1 a). The probability of survival free from a complicated course of CD was estimated at 85%, 78%, and 72% at 1, 3, and 5 years, respectively, for patients with endometriosis, compared with 82%, 70%, and 62% for those without endometriosis (Supplementary Table 1, Supplementary Digital Content 1, http://links.lww.com/CTG/B434 ). Among patients whose endometriosis was diagnosed before CD (n = 34), the result remained unchanged but was not statistically significant (HR = 0.69, 95% CI 0.33–1.41, P = 0.31). Similarly, in the subgroup of patients who underwent surgical treatment of endometriosis during follow-up (n = 72), a trend toward a less severe course of CD was observed (HR 0.65, 95% CI 0.42–1, P = 0.05) (Table 4 ). Comparison of CD characteristics according to the diagnosis of endometriosis P values were calculated using marginal binary logistic regression models for binary outcomes or using mixed ordinal logistic regression for ordinal outcomes. A P value < 0.05 is considered statistically significant. CD, Crohn's disease; Endometriosis (−), patients without endometriosis; Endometriosis (+), patients with endometriosis; NA, not applicable; TNF, tumor necrosis factor. Survival analysis of IBD progression according to the diagnosis of endometriosis. ( a ) Survival analysis of CD progression according to the diagnosis of endometriosis. ( b ) Survival analysis of UC progression according to the diagnosis of endometriosis. CD, Crohn's disease; HR, hazard ratio; IBD, inflammatory bowel disease; UC, ulcerative colitis. Impact of endometriosis on IBD course P values were calculated using marginal binary logistic regression models for binary outcomes or using mixed ordinal logistic regression for ordinal outcomes. A P value < 0.05 is considered statistically significant. Composite severity criterion: colonic stricture or colectomy for patients with UC, surgery or bowel damage (stenosis, fistula or intra-abdominal abscess) for patients with CD. CD, Crohn's disease; Endometriosis (−), patients without endometriosis; Endometriosis (+), patients with endometriosis; IBD, inflammatory bowel disease; n/N, number of patients experienced the composite criterion/number total of patients; NA, not applicable; TNF, tumor necrosis factor; UC, ulcerative colitis. The need for immunosuppressants was significantly lower in patients with CD and endometriosis compared with those with CD only (65.1% vs 76.3%, P = 0.008) (Table 3 ). Similarly, the use of biologics was less frequent in the endometriosis group (72.5% vs 88.9%; P < 0.001). This difference was consistent when analyzed separately for anti-TNF agents (71.8% vs 87.2%; P < 0.001) and other biologics (20.8% vs 36.1%; P < 0.001). The rate of colectomy was not statistically different between the 2 groups (19.0% vs 11.5%, P = 0.20). For 1 patient, the colectomy was performed for both UC and endometriosis with intestinal involvement. The time to colectomy was longer in the endometriosis group (6 years [3–7.25] vs 3 years [1–9]). No patients with UC developed colonic stenosis during follow-up (Table 5 ). In the survival analysis, patients with UC and endometriosis did not have a significant different disease course compared with those without endometriosis (HR 1.73, 95% CI 0.74–4.00, P = 0.20) (Figure 1 b). The probability of survival free from a complicated course of UC was estimated at 97%, 97%, and 90% at 1, 3, and 5 years, respectively, for patients with endometriosis, compared with 97%, 95%, and 93% for those without endometriosis (Supplementary Table 1, Supplementary Digital Content 1, http://links.lww.com/CTG/B434 ). When restricting the analysis to patients whose endometriosis was diagnosed before UC (n = 15), results remained unchanged (HR 2.08, 95% CI 0.29–14.87, P = 0.47). Similarly, in the subgroup of patients who underwent surgical treatment of endometriosis during follow-up (n = 26), no significant difference was observed between the 2 groups (HR 2.05, 95% CI 0.46–9.1, P = 0.20) (Table 4 ). Comparison of UC characteristics according to the diagnosis of endometriosis P values were calculated using marginal binary logistic regression models for binary outcomes or using mixed ordinal logistic regression for ordinal outcomes. A P value < 0.05 is considered statistically significant. Endometriosis (−), patients without endometriosis; Endometriosis (+), patients with endometriosis; NA, not applicable; TNF, tumor necrosis factor; UC, ulcerative colitis. Among patients with UC, the need for immunosuppressants was significantly lower in the endometriosis group (46.6% vs 66.4%; P = 0.005) (Table 5 ). Similarly, the use of biologics was less frequent in the endometriosis group (56.9% vs 75.2%; P < 0.009). This difference was consistent when analyzed separately for anti-TNF agents (53.4% vs 70.8%; P = 0.014) and other biologics (29.3% vs 46.0%; P = 0.037). In addition, the use of 5-aminosalicylates was significantly higher in the endometriosis group (94.8% vs 83.2%; P = 0.033).

This study demonstrated that patients with both endometriosis and CD have a less severe disease progression (i.e., surgery and bowel damage) compared with patients with CD without endometriosis (HR 0.68, 95% CI 0.50–0.92, P = 0.011). By contrast, no significant difference was observed in the disease course of patients with UC with vs without endometriosis (HR 1.73, 95% CI 0.74–4.00). A case-control study conducted by Lee et al involving 51 patients with both IBD and endometriosis (26 CD, 23 UC) and 102 matched controls reported that the coexistence of IBD with endometriosis did not affect disease phenotype, the use of immunosuppressants or anti-TNF agents, or the rate of IBD-related surgery ( 17 ). In this study, authors did not find differences in the behavior (inflammatory, stricturing, or fistulizing) and the location of CD. However, in our larger multicenter cohort, we found that patients with CD and endometriosis were significantly less frequently treated with immunosuppressants and biologics compared with those without endometriosis. Regarding UC, our study found no significant difference in colectomy rates according to endometriosis status (11.5% [13/113] vs 19.0% [11/58]; P = 0.20). However, colectomy occurred later in the disease course among patients with coexisting endometriosis. It is noteworthy that 1 patient underwent colectomy for both UC and endometriosis with intestinal involvement. This result should be interpreted cautiously given the small number of patients and limited number of events (11/58). Furthermore, no patients in the UC cohort developed colonic stenosis. Therefore, the severity of UC was assessed solely based on the need for colectomy, a stringent yet incomplete criterion given current conservative strategies facilitated by advancements in medical treatment alternatives ( 19 – 21 ). Moreover, patients with UC and endometriosis were less frequently treated with systemic corticosteroid therapy (60.3% vs 72.6%; P = 0.10), immunosuppressants (46.6% vs 66.4%; P = 0.005), and biologics (56.9% vs 75.2%; P = 0.009), whereas more often receiving 5-aminosalicylates (94.8% vs 83.2%; P = 0.033), compared with patients without endometriosis. These findings suggest a potentially milder disease course in patients with UC with endometriosis, although this remains to be confirmed in larger prospective studies. Lee et al did not find any difference in disease extent for patients with UC with or without endometriosis ( 17 ). On the contrary, our results showed a significant difference in the location of UC ( P = 0.026), with more proctitis in the endometriosis group (29.3% vs 12.6%) and more left-sided colitis in the group without endometriosis (37.8% vs 27.6%). In a Danish cohort study, the authors reported that women with endometriosis have 1.5-fold increased risk of developing IBD compared with women from the general population. This risk is higher in women with surgically confirmed endometriosis ( 8 ). In our cohort, 47.3% (n = 98/207) of patients underwent surgery for endometriosis. The decision to perform surgical intervention instead of medical therapy for endometriosis depends on various factors, including patient's expectations, symptoms, disease location, disease severity, previous treatment failure, and reproductive considerations such as infertility or pregnancy desire ( 22 ). In our study, surgical treatment of endometriosis did not influence the progression of CD (HR 0.65, 95% CI 0.42–1.00, P = 0.05). By contrast, Lee et al observed, in a small subgroup of patients with CD with surgically confirmed endometriosis, a higher prevalence of stricturing disease compared with controls (n = 5, 33.3% vs n = 2, 4.3%; OR 11.8, 95% CI 2.03–69.0) ( 17 ). However, in our larger sample, we did not confirm this finding, as the proportion of stricturing disease was similar between patients with CD surgically treated for endometriosis and their controls (40% [14/35] vs 37.1% [26/70]). For patients with UC, surgery for endometriosis during the follow-up (n = 26/58 [45.6%]) likewise had no impact on disease progression (HR 2.05, 95% CI 0.46–9.10, P = 0.20). The incidence of IBD and endometriosis commonly impact women of reproductive age. Several studies have suggested that sex hormones influence the activity of IBD ( 23 ) and that oral contraceptives use may increase the risk of developing IBD ( 24 ). Endometriosis is also a hormone-dependent condition, with hormonal therapy serving as the mainstay of treatment, which could affect the risk of developing IBD. In our study, 49 patients had pre-existing endometriosis at the time of IBD diagnosis. When examining the temporal sequence between the diagnoses of the 2 diseases, we found no significant difference in IBD progression among patients whose endometriosis preceded IBD (for CD: HR 0.69, 95% CI 0.33–1.41, P = 0.31; for UC: HR 2.08, 95% CI 0.29–14.87, P = 0.47). However, our study did not establish whether hormonal treatments used for endometriosis could potentially influence the course of IBD. As in IBD, inflammation is a key major mechanism in the pathophysiology of endometriosis, promoting the implantation and growth of ectopic endometrial tissue ( 7 ). TNF alpha, a major proinflammatory cytokine, promotes the proliferation of endometrial cells, and its levels are increased in the peritoneal fluid of women with endometriosis, correlating with disease extent ( 25 ). The clinical overlap between endometriosis and IBD, such as pelvic pain and bowel movement disorders, may contribute to diagnostic uncertainty and delays ( 12 ). In our study, the mean time between IBD diagnosis and the subsequent diagnosis of endometriosis was 9 years. Importantly, our analyses consider all IBD-related events regardless the timing of the endometriosis diagnosis. Although surgery remains the gold standard for the diagnosis of endometriosis, imaging techniques based on ultrasound and MRI have improved the diagnosis of endometriosis in recent years, reducing delays and providing timely care. Our study was conducted in tertiary referral centers specializing in endometriosis, with diagnoses made using pelvic MRI in 73.9% of cases (153/207) and laparoscopic visualization in 23.2% of cases (48/207). Our cohort included two-third of patients with deep infiltrating endometriosis and 23.4% with intestinal involvement, which is consistent with the literature ( 26 , 27 ). Consequently, our findings may not be generalizable to patients with more superficial forms of endometriosis. Our study represents the largest multicenter cohort to date assessing the impact of endometriosis on the phenotype and the course of IBD, with a long median follow-up exceeding 10 years. However, several limitations should be acknowledged. First, owing to the retrospective nature of the data collection, there are numerous missing data points, which vary significantly depending on the quality of the computerized medical record. Some information was completed through telephone interviews with patients, introducing a potential memory bias. In addition, it is difficult to determine whether surgeons perform fewer digestive resections in patients with both IBD and endometriosis out of concern for complications, which could explain the delay in surgical management. Second, the composite severity criterion has not been validated and does not include objective markers of disease activity. It could be appropriate to expand these results with clinical, biological, radiological, endoscopic, and histological outcomes reflecting the severity of the disease ( 28 ). Finally, data regarding fertility outcomes were not available. In conclusion, a diagnosis of endometriosis, which can mimic IBD, should therefore be considered by the gastroenterologist in female patients of childbearing age. Although patients with endometriosis are supposed to have a higher risk of developing IBD, the presence of endometriosis does not negatively affect the course of IBD. On the contrary, in our study, patients with IBD exposed to endometriosis had less severe disease progression and less exposure to immunosuppressants and biologics. Surgery for endometriosis does not seem to affect the course of IBD and should not be delayed due to the presence of IBD. However, these results should be viewed in light of the limitations of the study.

Guarantor of the article: Lucas Guillo, MD. Specific author contributions: L.G. and P.W.: participated in the study concept and study design. JL performed the statistical analysis of data. L.G., P.W., and L.M.: participated in the analysis and interpretation of data. L.G., O.E.O., and P.W.: wrote the article. All authors critically revised the manuscript. The manuscript was approved by all authors. Financial support: None to report. Potential competing interests: L.G. declares consulting fees for Abbvie, Amgen, Ferring, Janssen, Everzom. P.W. declares personal fees from Abbvie, Ferring, Amgen, Takeda, Biogen, Janssen. The remaining authors have nothing to disclose. Data availability statement: The data underlying this article are available in the article and in its online supplementary material. Study Highlights WHAT IS KNOWN ✓ Endometriosis and inflammatory bowel diseases (IBDs) are immune-mediated inflammatory diseases that predominantly affect young patients. ✓ Women with endometriosis have a higher risk of developing IBD. WHAT IS NEW HERE ✓ Endometriosis does not seem to negatively influence the course of IBD. ✓ Patients with IBD with endometriosis are less exposed to immunosuppressants and biologics. ✓ Endometriosis and inflammatory bowel diseases (IBDs) are immune-mediated inflammatory diseases that predominantly affect young patients. ✓ Women with endometriosis have a higher risk of developing IBD. ✓ Endometriosis and inflammatory bowel diseases (IBDs) are immune-mediated inflammatory diseases that predominantly affect young patients. ✓ Women with endometriosis have a higher risk of developing IBD. ✓ Endometriosis does not seem to negatively influence the course of IBD. ✓ Patients with IBD with endometriosis are less exposed to immunosuppressants and biologics. ✓ Endometriosis does not seem to negatively influence the course of IBD. ✓ Patients with IBD with endometriosis are less exposed to immunosuppressants and biologics.