S188 Idiopathic Myointimal Hyperplasia of Mesenteric Veins: A Systematic Review and Individual Patient Regression Analysis

Abstract

Introduction: Idiopathic myointimal hyperplasia of the mesenteric veins (IMHMV) is an uncommon cause of colonic ischemia for which surgical treatment is recommended. We describe the clinical, radiologic and endoscopic findings in IMHMV patients to provide clinicians with a framework for the identification of this rare disease. Methods: We performed a systematic review of seven databases for cases of IMHMV and identified additional cases from Yale New Haven Hospital records. To identify features specifically associated with colonic ischemia due to IMHMV, we performed multivariate logistic regression analysis incorporating data from a large, well-characterized multicenter cohort of biopsy-proven ischemic colitis (n=923). Results: A total of 123 patients with IMHMV were identified from 58 publications and 3 unpublished cases at Yale New Haven Hospital (80% male, mean age 53, 56% Caucasian). Symptoms at presentation were most commonly abdominal pain (87%) and diarrhea (67%). The most affected areas were the sigmoid colon (89%) and rectum (66%). Median time from symptom onset to diagnosis was 4 months. Complications, including obstruction and perforation due to diagnostic delay occurred in 33% of patients. The most common radiologic feature was thickening of the affected segment of colon (97%). Anatomic vascular abnormalities including non-opacification of the inferior mesenteric vein were observed in 38% of patients. 97% of patients ultimately underwent curative segmental colectomy. Mean follow-up time was 23.6 months. Compared to biopsy-proven non-IMHMV colonic ischemia, IMHMV was significantly associated with younger age (p=0.010), male sex (p=0.002), rectal involvement on imaging (p=0.027) and on endoscopic evaluation (p=0.004), mucosal ulcerations on endoscopy (p=0.049) and absence of rectal bleeding on presentation (p< 0.001) (Table). Conclusion: IMHMV is a rare, underreported cause of colonic ischemia that predominantly involves the left colon. Using the largest IMHMV cohort to date and comparing to a multicenter cohort of patients with ischemic colitis, we identify clinical, endoscopic, and radiologic characteristics of this entity. Our findings suggest younger age, rectal involvement, and absence of rectal bleeding as clinical features to help identify select patients presenting with colonic ischemia as having higher likelihood of IMHMV and therefore consideration of upfront surgical management. Table 1. - Clinical, endoscopic, and radiologic findings associated with IMHMV associated colonic ischemia Variable # patients IMHMV # patients Non-IMHMV P value Demographics Age 85 55.5 (17.0) 922 69.5 (13.2) < 0.001 Male sex 85 78.8% (67) 918 28.7% (265) < 0.001 Clinical Presentation Abdominal pain 85 84.7% (72) 918 82.6% (758) 0.764 Rectal bleeding 85 57.6% (49) 913 76.2% (696) < 0.001 Pain without bleeding 85 35.3% (30) 911 22.7% (207) 0.011 Bleeding without pain 85 8.2% (7) 911 18.2% (166) 0.017 Non-bloody diarrhea 85 24.7% (21) 911 11.0% (100) < 0.001 Clinical Comorbidities Hypertension 61 21.3% (13) 921 75.1% (692) < 0.001 Diabetes mellitus 61 13.1% (8) 921 30.0% (276) 0.005 Stroke 61 4.9% (3) 920 10.8% (99) 0.193 Hypercoagulability 61 6.6% (4) 768 1.0% (8) 0.008 Multivariate logistic regression Variable aOR 95% CI P value Age, per year 0.941 0.919 – 0.964 0.010 Sex (Male) 10.10 4.71 – 21.67 0.002 Rectal bleeding on presentation 0.006 0.001 – 0.025 < 0.001 Non-bloody diarrhea on presentation 1.46 0.60 – 3.55 0.673 Diabetes mellitus 0.569 0.249 – 1.350 0.494 Rectal involvement on imaging 6.71 2.84 – 15.85 0.027 Right-sided involvement on endoscopy 0.620 0.213 – 1.806 0.655 Rectal involvement on endoscopy 12.61 5.22 – 30.51 0.004 Mucosal ulcerations on endoscopy 4.06 1.99 – 8.25 0.049 Baseline characteristics and regression analysis of IMHMV and non-IMHMV patients included in the final multivariate model. Number of patients included in the analysis are indicated. Continuous variables presented as mean and standard deviation. Categorical variables are presented as proportion and number of individuals. aOR = adjusted odds ratio. 95% CI = 95% confidence interval.