pone-0020582-g004: Probability of survival for CS 4, CS 4+ and CS 5.Probability of survival for the 3 different endpoints (CS 4, black line; CS 4+, blue line; CS 5, red line). For all logrank tests, CS 4 was used as the reference when comparing CS 4 vs. CS 4+ and CS 4 vs. CS 5, whereas CS 4+ was used as the reference when comparing CS 4+ vs. CS 5. The rate of reaching endpoint is significantly different (P = 0.021) between CS 4 (clinical score of 4 = functional paralysis of both hindlimbs), CS 4+ [CS 4 plus the earliest of a) weight loss ≥20% vs. body weight immediately prior to a clinical score of 2, b) weight loss ≥20% vs. peak body weight, c) body condition score <2, or d) a righting reflex >20 s (CS 5)], and CS 5 (clinical score of 5 = CS 4 and righting reflex >20 s). Mice reached CS 4 at a rate of 34% faster vs. CS 5 (HR = 1.34; 95% CI 1.10, 1.74; P = 0.006) and 24% faster vs. CS 4+ (HR = 1.24; 95% CI 1.00, 1.59; P = 0.046). Mice reached CS 4+ at a non-significant rate of 9% faster vs. CS 5 (HR = 1.09; 95% CI 0.88, 1.38; P = 0.410).

Mentions: A logrank test showed a significant difference in the rate at which endpoint was reached between CS 4, CS 4+ and CS 5 (P = 0.021; Figure 4). Mice reached CS 4 at a rate 34% faster vs. CS 5 (HR = 1.34; 95% CI 1.10, 1.74; P = 0.006) and 24% faster vs. CS 4+ (HR = 1.24; 95% CI 1.00, 1.59; P = 0.046). Mice reached CS 4+ at a non-significant rate of 9% faster vs. CS 5 (HR = 1.09; 95% CI 0.88, 1.38; P = 0.410).

Solomon JA, Tarnopolsky MA, Hamadeh MJ - PLoS ONE (2011)

Bottom Line: Mice reached CS 4 at 123.9±10.3 d, CS 4+ at 126.6±9.8 d and CS 5 at 127.6±9.8 d, all significantly different from each other (P<0.001).There was a significant positive correlation between CS 4 and CS 5 (r = 0.95, P<0.001), CS 4 and CS 4+ (r = 0.96, P<0.001), and CS 4+ and CS 5 (r = 0.98, P<0.001), with the Bland-Altman plot showing an acceptable bias between all endpoints.Converging ethics and discovery, would adopting CS 4 as endpoint compromise the acquisition of insight about the effects of interventions in animal models of ALS?

Affiliation: School of Kinesiology and Health Science, Faculty of Health, York University, Toronto, Ontario, Canada.

Abstract: There is no consensus among research laboratories around the world on the criteria that define endpoint in studies involving rodent models of amyotrophic lateral sclerosis (ALS). Data from 4 nutrition intervention studies using 162 G93A mice, a model of ALS, were analyzed to determine if differences exist between the following endpoint criteria: CS 4 (functional paralysis of both hindlimbs), CS 4+ (CS 4 in addition to the earliest age of body weight loss, body condition deterioration or righting reflex), and CS 5 (CS 4 plus righting reflex >20 s). The age (d; mean ± SD) at which mice reached endpoint was recorded as the unit of measurement. Mice reached CS 4 at 123.9±10.3 d, CS 4+ at 126.6±9.8 d and CS 5 at 127.6±9.8 d, all significantly different from each other (P<0.001). There was a significant positive correlation between CS 4 and CS 5 (r = 0.95, P<0.001), CS 4 and CS 4+ (r = 0.96, P<0.001), and CS 4+ and CS 5 (r = 0.98, P<0.001), with the Bland-Altman plot showing an acceptable bias between all endpoints. Logrank tests showed that mice reached CS 4 24% and 34% faster than CS 4+ (P = 0.046) and CS 5 (P = 0.006), respectively. Adopting CS 4 as endpoint would spare a mouse an average of 4 days (P<0.001) from further neuromuscular disability and poor quality of life compared to CS 5. Alternatively, CS 5 provides information regarding proprioception and severe motor neuron death, both could be important parameters in establishing the efficacy of specific treatments. Converging ethics and discovery, would adopting CS 4 as endpoint compromise the acquisition of insight about the effects of interventions in animal models of ALS?