Purpose or Case Report: This retrospective pilot study was performed to assess the utility of limited protocol MRI for appendicitis in excluding ovarian torsion by assessing ovarian symmetry with regard to size and appearance comparing torsed cases to controls.
Methods & Materials: Retrospective review of limited appendicitis MRI cases was IRB approved. 699 female patients were imaged between 12/2013 and 7/2019. Imaging findings and impressions were collected from the medical record.
7 cases of surgically proven torsion were identified.Ten age-matched, pubertal status-matched, and MRI field-strength-matched controls for each torsion case were selected for this pilot study.
Ovarian dimensions, T2 signal intensity sampled to exclude follicles or cysts, and diffusion coefficient of the most restricted stroma excluding hemorrhage were tabulated.
Differences in size and appearance between normal and torsed ovaries were analysed utilizing ratios of bigger to smaller ovaries for ovarian volume, more intense to less intense ovaries for T2 signal, and more restricted to less restricted ovaries for diffusion coefficients.
Median values, quartiles, ranges, and odds ratios were calculated for each ratio variable and Mann-Whitney tests were performed.
Results: Of 70 selected controls, 66 had adequate ovarian visualization. Of the 66, 2 control patients and one torsion case could not be evaluated for diffusion coefficient.Control patients had symmetric appearing ovaries in most cases.The torsed ovary was always larger than the normal ovary. Using a 3:1 volume ratio, we observed 7 true positives, 61 true negatives, and 5 false positives. There were no false negatives. The 3:1 volume ratio in exclusion of torsion was 100% sensitive (95% CI 59% - 100%) and 92.4% specific (95% CI 83.2% to 97.5%) with a p value of 0.134. The median T2 signal ratio was larger in torsion cases than in controls: 2.0, range 1.3 - 5.5, p value 0.025, although the torsed ovary could be brighter or darker than the normal ovary. The median diffusion coefficient ratio was larger in torsion cases than controls: 3.3, range 1.7 - 7.0, p value 0.008, although the torsed ovary could have a higher or lower diffusion coefficient than the normal ovary.
Conclusions: The limited MRI appendicitis protocol can exclude ovarian torsion. An ovarian volume ratio greater than 3:1 has a sensitivity for torsion of 100%. Asymmetric T2 signal and diffusion restriction can suggest the presence of torsion, but further research is needed with a larger cohort of cases.