2373. Multiparametric MRI Characterization for Diagnosing Acute Cholecystitis
Authors * Denotes Presenting Author
  1. Gabriel Duhancioglu *; University of Arizona
  2. Conner Reynolds; University of Arizona
  3. Kanika Gupta; University of Arizona
  4. Hina Arif; University of Arizona
To evaluate the use of multi-parametric magnetic resonance imaging (mpMRI) to confidently diagnose acute cholecystitis (AC) and accurately differentiate it from its common mimics, such as reactive gallbladder wall inflammation (GBWI) due to acute hepatitis or volume overload.

Materials and Methods:
For this study, 40 consecutive patients presenting to our department with a suspected clinical diagnosis of AC received abdominal imaging using standard clinical mpMRI with T2 weighted, gradient echo T1 pre/postcontrast, MRCP, and DWI sequences. The radiologist, who was blinded to etiology and pathological diagnosis, assessed the following imaging features to reach the diagnosis of AC or reactive GBWI: increased enhancement of liver parenchyma in region of GB fossa, GBW thickness, fluid within GBW on MRCP, GBW splitting, increased mural enhancement, wall irregularity, gallstones or sludge, periportal edema, pericholecystic T2 signal, and restricted diffusion. These features were then compared with patients’ pathological studies and clinical outcome within 6-months of mpMRI examination. Sensitivity, specificity, PPV, and NPV were calculated. Correlation (spearman rho), logistic regression models, and area under the curve (AUC) analysis of receiver operating characteristic (ROC) were used for statistical analyses (SASv9.4, Cary, NC).

All patients diagnosed with AC on mpMRI underwent cholecystectomy which confirmed the diagnosis by pathology. All patients who showed reactive GBWI underwent medical management without surgery for AC. AC was significantly correlated (p < 0.0219) with increased arterial phase enhancement of GB fossa, absence of fluid in wall (MRCP), GB wall splitting, GB wall irregularity, presence of sludge or gallstones, increased pericholecystic T2 signal and restricted diffusion. Sensitivity, specificity, positive and negative predictive value were 100% when all of the above factors were considered. Multivariable logistic regression employing forward selection identified increased pericholecystic T2 signal (p = 0.004) and splitting of wall (p < 0.0001) as significant predictors and achieved an AUC±SE of 0.9886 ± 0.0099.

mpMRI shows extremely high accuracy in diagnosing and differentiating AC from common mimics, such as reactive GBWI.