Repetitive nerve stimulation as a diagnostic aid for distinguishing cervical spondylotic amyotrophy from amyotrophic lateral sclerosis.

PURPOSE: To identify and compare the features of compound muscle action potential (CMAP) decrements in repetitive nerve stimulation (RNS) in patients with cervical spondylotic amyotrophy (CSA) and in patients with amyotrophic lateral sclerosis (ALS).

METHODS

The cohort consisted of 43 CSA (distal-type to proximal-type ratio: 27-16) and 35 ALS patients. Five muscles, including abductor pollicis brevis (APB), abductor digiti minimi (ADM), biceps brachii (BB), middle deltoid (Del), and upper trapezius (Trap), were tested by 3-Hz RNS. Decrements greater than cutoff values (APB > 5.8%; ADM > 4.8%; BB > 5.2%; Del > 6%; Trap > 5.1%) determined using receiver operating characteristic (ROC) curves were defined as abnormal, and the conventional criterion (≥10%) was also considered.

RESULTS

A significant CMAP decrement (>cutoff values) was recorded from at least one tested muscle in 91.4% of ALS patients, and was most common in the proximal muscle, a finding that differed significantly from CSA patients (32.6%, P < 0.05). The application of cutoff values greatly improved the sensitivity of RNS over the conventional criterion (≥10%) for the detection of ALS (P < 0.05). The specificity of this technique remained higher when performing RNS in the proximal muscles, especially in the upper trapezius (AUC = 0.864, sensitivity = 0.643, and specificity = 1.000). The decrement percentages were significantly greater in the proximal muscles of ALS patients than in those of the CSA patients (P < 0.05). In addition, illness duration was not correlated with decrement percentage in either patient group, and no difference in the frequency of decrement among different ALS diagnostic categories was observed (P > 0.05).

CONCLUSIONS

The application of RNS, especially in proximal muscles, may provide a simple accurate and noninvasive supplementary test for distinguishing CSA from ALS, even in the early stage of these diseases. A combination of RNS, needle EMG, clinical features and cervical magnetic resonance imaging may yield sufficient diagnostic information to differentiate CSA and ALS.