Diagnostic accuracy of nerve ultrasonography for the detection of peripheral neuropathy in type 2 diabetes.

Abstract

Nerve conduction studies (NCS) are the current objective measure for diagnosis of peripheral neuropathy in type 2 diabetes but do not assess nerve structure. This study investigated the utility of peripheral nerve ultrasound as a marker of the presence and severity of peripheral neuropathy in type 2 diabetes. A total of 156 patients were recruited, and nerve ultrasound was undertaken on distal tibial and distal median nerves. Neuropathy severity was graded using the modified Toronto Clinical Neuropathy Scale (mTCNS) and Total Neuropathy Score (TNS). Studies were undertaken by a single ultrasonographer blinded to nerve conduction results. A stepwise increase in tibial nerve cross-sectional area (CSA) was noted with increasing TNS grade (p< 0.001) and each mTCNS quartile (p< 0.001). Regression analysis demonstrated a correlation between tibial nerve CSA and neuropathy severity (p< 0.001). Using receiver operator curve analysis, tibial nerve CSA of >12.88 mm yielded a sensitivity of 70.5% and specificity of 85.7% for neuropathy detection. Binary logistic regression revealed that tibial nerve CSA was a predictor of abnormal sural sensory nerve action potential amplitude (odds ratio=1.239, 95% confidence interval [CI]=1.142-1.345) and abnormal neuropathy score (odds ratio=1.537, 95% confidence interval [CI]=1.286-1.838). Tibial nerve ultrasound has good specificity and sensitivity for neuropathy diagnosis in type 2 diabetes. The study demonstrates that tibial nerve CSA correlates with neuropathy severity. Future serial studies using both ultrasound and NCS may be useful in determining whether changes in ultrasound occur prior to development of nerve conduction abnormalities and neuropathic symptoms.