Characterizing conventional ankle MRI findings of nerve and muscle changes in diabetic patients. A case-control study

Foot and ankle amputation is a feared complication of diabetic neuropathy and diabetes mellitus (DM) accounts for 80% of all in-hospital amputations. Magnetic resonance neurography is an effective tool in characterizing neuromuscular sequelae of the disease. However, conventional ankle MRI is more commonly performed and has not been studied to assess neuromuscular changes of DM. The objective is to characterize neuromuscular changes of diabetic patients in a case-control study using conventional ankle MRI. Between November 2019 and July 2021, 110 consecutive ankle MRI scans (n = 102 patients) at our county hospital were reviewed and met the inclusion criteria. Patients were divided into two cohorts, diabetic (N = 63) and non-diabetic (N = 39). Demographics, HgbA1c, and reason for MRI study were collected via retrospective chart review. The presence of intramuscular edema-like signal, pattern of the edema, muscle fatty infiltration, and measurements of the cross-sectional area of the posterior, medial, and lateral tibial nerves (PTN, MPN, and LPN) was recorded blinded to the clinical findings by two readers. Muscle edema-like signal was much more likely to be found in DM (odds ratio 19.5, 95% CI 7.0-54.6, p < 0.001). DM also showed increase of 0.87 in the mean grade of muscle fatty infiltration (p < 0.001). There were higher rates of nerve T2 hyperintensity (odds ratio 14.0, 95% CI 3.1-62.7, p < 0.001) and the measured areas of the PTN, MPN, and LPN were also larger in DM compared to their non-diabetic counterparts (PTN: 0.16 cm${}^2$ vs. 0.10 cm${}^2$, p < 0.01; MPN: 0.09 cm${}^2$ vs. 0.05 cm${}^2$, p < 0.01; LPN: 0.07 cm${}^2$ vs. 0.04 cm${}^2$, p < 0.05). Conventional ankle MRIs can be used to detect DM-related neuromuscular changes.