Relationship Between Morphological Remodeling and Angiographic Types of Spontaneous Isolated Superior Mesenteric Artery Dissection After Conservative Management: Determinant Affecting Serial Radiologic Courses

OBJECTIVE: To monitor the radiological courses of symptomatic spontaneous isolated superior mesenteric artery dissection (SISMAD) after conservation, clarify the relationship between its morphological change and initial imaging classification, and identify these factors that affect dissection remodeling. METHODS: Eighty-nine conservative patients with SISMAD who underwent periodic follow-up of computed tomography angiography (CTA) were enrolled. Initial morphologic classification, imaging features and dissection remodeling were analyzed retrospectively. Logistic regression was used to identify predictors for remodeling. Receiver operating characteristics were performed for cutoff threshold. RESULTS: Zerbib classification was adapted and initial CT appearance divided eighty-nine patients into: type I (15.7%), patent false lumen (FL) with both entry and re-entry; type II (37.1%), “cul-de-sac” shaped FL without re-entry; type III (27.0%), thrombosed FL with ulcer-like-projection (ULP); type IV (18%), intramural hematoma; type V (0%), dissecting aneurysm; and type VI (2.2%), total or partial occlusion of superior mesenteric artery (SMA). Follow-up CTA revealed complete remodeling (33.7%), partial remodeling (16.9%), no change (25.8%), type change (13.5%) and dissection progression (10.1%). There was no dissection-related mortality. Type I (92.9%) sustained patent FL and no angiographic change. Type II showed partial remodeling (42.4%), no change (27.3%) and dissection progression (27.3%), and the length of FL enhancement positively predicted dissection progression with the cutoff of 40.3 millimeters. Type III achieved complete remodeling (58.3%) or evolved into type II (41.7%), and the distance between SMA orifice and ULP negatively predicted type change with the cutoff of 23.5 millimeters. Type IV (87.5%) achieved complete remodeling due to hematoma absorption. One patient underwent stent placement for the evolution of ULP into an enlarged blind-ending FL 2 months after conservation. CONCLUSION: After conservation, patent FL with a distal re-entry is no morphological change, FL thrombosis tends to be resolved, and the “cul-de-sac” shaped FL without re-entry is partially shortened, no change or progressively dilated. FL enhancement length ≥ 40.3 millimeters is a predictor for the blinding-end FL enlargement. Thrombosed FL with ULP evolves into a patent “cul-de-sac” shaped FL when the distance between SMA orifice to ULP is less than 23.5 millimeters. A careful follow-up is necessary for the lesions with demonstrated predictors.