E2258. Gastrointestinal Stromal Tumor Radiology: Pathology Correlations and Risk Classifications
  1. Tej Mehta; The Johns Hopkins Hospital; United States Air Force
  2. Renu Berry; The Johns Hopkins Hospital
  3. James Facciola; The Johns Hopkins Hospital
  4. Elliot Fishman; The Johns Hopkins Hospital
  5. Rakhee Gawande; The Johns Hopkins Hospital
To identify contrast enhanced CT (CECT) and histopathologic risk factors for risk stratification of gastrointestinal stromal tumors (GISTs) using the Miettinen risk classification (MRC) and evaluation of recurrent/residual disease.

Materials and Methods:
A retrospectively collected database of all patients with pathologically proven GISTs and available preoperative CECT scans at our institution between 8/14/2003 and 03/19/2020 was analyzed. All CECTs were evaluated by 3 radiologists to characterize size, luminal extent, calcifications, necrosis, ulceration, perforation, enhancement, cystic change, nodal involvement, neovascularity and metastases. Pathology data were reviewed to characterize GIST size, focality, subtype, mitotic rate, necrosis, nodal involvement, lymphatic or vascular invasion and histologic grade. Differences in categorical variables by risk stratification and presence of residual/recurrent disease were assessed using the Pearson ?2 or Fisher exact test. Differences in continuous variables were assessed using independent t-tests, ANOVA or Mann-Whitney U test. Uni/multivariate logistic regression was performed to identify independent predictors of MRC and residual/recurrent disease.

208 GISTS in 201 patients (mean age 62.3 ± 14), 105 male, 96 female were included. The most common tumor site was stomach (51.4%), followed by duodenum (23.1%), jejunum (7.7%), esophagus (5.3%), ileum (4.8%), extra-gastrointestinal sites (4.8%) and colon (2.8%). 20 GISTs were No Risk by the MRC, 22 were Very Low Risk, 36 were Low Risk, 39 were Moderate Risk and 48 were High Risk; 43 had data missing. Presence of necrosis on CECT, metastatic disease and tumor multifocality were associated with higher grade GISTs (p <0.05). Stratified univariate analysis by primary tumor location identified that size greater than 5.1 cm (OR 4.4, 95%CI 1.4-15.2, p=0.01), neovascularity (OR 3.3, 95%CI 1.1-10.3, p=0.04) and presence of metastatic disease (OR 6.6, 95% CI 1.9-25.2, p < 0.01) were significantly associated with residual or recurrent disease for GISTs originating in the stomach. For GISTs arising from the small intestine, only regional nodal involvement (OR 5.8, 95% CI 1.1-35, p = 0.04) and metastatic disease (OR 5.4, 95% CI 1.4-23.1, p = 0.02) were significantly associated with residual or recurrent disease. Insufficient data were available for stratified analysis for GISTs arising from other locations.

Early identification and risk stratification of GISTs is crucial for patient management and prognostication. Primary risk stratification variables include primary tumor location, size, and mitotic rate. CECT assists with risk stratification by identifying benign or aggressive characteristics. In our analysis, presence of necrosis on CECT and on pathologic examination, presence of metastatic disease and tumor multifocality were associated with higher grade GISTs. Additionally, tumor size greater than 5.1 cm, neovascularity, nodal involvement and metastatic disease on initial imaging were associated with residual/recurrent disease, though these factors varied based on primary tumor site.