Differentiating perforated from non-perforated appendicitis on contrast-enhanced magnetic resonance imaging.

BACKGROUND: The role of magnetic resonance imaging (MRI) in pediatric appendicitis is increasing; MRI findings predictive of appendiceal perforation have not been specifically evaluated.

OBJECTIVE: To assess the performance of MRI in differentiating perforated from non-perforated appendicitis.

MATERIALS AND METHODS: A retrospective review of pediatric patients undergoing contrast-enhanced MRI and subsequent appendectomy was performed, with surgicopathological confirmation of perforation. Appendiceal diameter and the following 10 MRI findings were assessed: appendiceal restricted diffusion, wall defect, appendicolith, periappendiceal free fluid, remote free fluid, restricted diffusion within free fluid, abscess, peritoneal enhancement, ileocecal wall thickening and ileus. Two-sample t-test and chi-square tests were used to analyze continuous and discrete data, respectively. Sensitivity and specificity for individual MRI findings were calculated and optimal thresholds for measures of accuracy were selected.

RESULTS: Seventy-seven patients (mean age: 12.2 years) with appendicitis were included, of whom 22 had perforation. The perforated group had a larger mean appendiceal diameter and mean number of MRI findings than the non-perforated group (12.3 mm vs. 8.6 mm; 5.0 vs. 2.0, respectively). Abscess, wall defect and restricted diffusion within free fluid had the greatest specificity for perforation (1.00, 1.00 and 0.96, respectively) but low sensitivity (0.36, 0.25 and 0.32, respectively). The receiver operator characteristic curve for total number of MRI findings had an area under the curve of 0.92, with an optimal threshold of 3.5. A threshold of any 4 findings had the best ability to accurately discriminate between perforated and non-perforated cases, with a sensitivity of 82% and specificity of 85%.

CONCLUSION: Contrast-enhanced MRI can differentiate perforated from non-perforated appendicitis. The presence of multiple findings increases diagnostic accuracy, with a threshold of any four findings optimally discriminating between perforated and non-perforated cases. These results may help guide management decisions as MRI assumes a greater role in the work-up of pediatric appendicitis.