Linear estimates of rates of disease progression as predictors of survival in patients with ALS entering clinical trials

Maximal voluntary isometric grip and foot dorsiflexion (FD) strength and forced vital capacity (FVC) were obtained in 62 patients with ALS at or close to enrollment into two clinical trials. The agents tested did not slow disease progression. Isometric strength data were standardized, and the worse side was taken. FVC was expressed as a percentage of the predicted value (FVC%). We derived linear estimates of rates of disease progression based on the isometric myometry and FVC measures and on disease duration. Forty one patients were known to have died or to have undergone tracheostomy for ventilatory support. Probability of tracheostomy-free survival was calculated using the Kaplan-Meier method. The measured values, the linear estimates for rates of decline of these values, gender, age at onset, bulbar vs. spinal onset, height and weight were tested as risk factors within the Cox proportional hazards model, using regression techniques. When tested individually, estimates of rates of decline based on all three measures (FD, grip and FVC%) were the only statistically significant risk factors (P<0.005). Multivariate analysis resulted in a 3-variable model (chi-square=75.3, P<0.00001) in which estimated rates of decline of FD strength and of FVC%, and bulbar onset were independently significant (P<0.0001, P<0.0007 and P<0.05, respectively). We conclude that linear estimates of the rate of disease progression till enrollment into a clinical trial may be better predictors of patient survival than demographic data or discrete biologic measures. (C) 1998 Elsevier Science B.V. All rights reserved.