Objective: This study aimed to reveal the diagnostic utility of Gold Coast (GC) criteria in Japanese patients with amyotrophic lateral sclerosis (ALS) by comparing the sensitivity/specificity with revised El Escorial (R-EE) and Awaji criteria, because its utility has not been studied in Asian ALS. Methods: Consecutive 639 patients (529 with ALS and 110 with ALS mimics), who were suspected of ALS and referred to three Japanese ALS centers, were enrolled. Diagnostic accuracy and characteristics of false positive and negative in GC criteria were compared with those of the Awaji and R-EE criteria. Patients were categorized as definite, probable or possible ALS according to each criterion. Results: The sensitivity of GC criteria (96.8%, 95% confidence interval [CI]: 95.3-98.3%) was higher than that of Awaji (89.6%, 95% CI: 87.0-92.2%) and R-EEC (89.2, 95% CI: 86.6-91.8%) criteria (both, p < 0.001). The specificity was also higher with GC criteria (77.3%, 95% CI: 69.5-85.1%) than Awaji (65.5%, 95% CI: 56.6-74.4%) and R-EEC (66.4, 95% CI: 57.6-75.2%) criteria (both, p < 0.01). Using GC criteria, patients with cervical spondylosis and Parkinson's syndrome tended to be diagnosed with ALS (i.e. "false positive"). Additionally, ALS patients diagnosed only by GC criteria less frequently had upper motor neuron (UMN) signs, compared with the other two criteria. Conclusion: Gold Coast criteria improve diagnostic accuracy for ALS in an Asian population, especially in patients with subtle UMN signs.
Keywords: Amyotrophic lateral sclerosis; Awaji criteria; El Escorial criteria; Gold Coast criteria.