It is not clear to what extent cerebral degeneration impacts on survival in patients with amyotrophic lateral sclerosis (ALS). Previous research indicates that regional brain concentrations of neurochemicals such as N-acetylaspartate (NAA) might serve as biomarkers for the degree of cerebral involvement in ALS.OBJECTIVE
To assess whether abnormal ratios of cerebral neurochemicals, quantified by magnetic resonance spectroscopic imaging (MRSI), correlate with survival in patients with ALS.DESIGN AND INTERVENTION
This controlled trial, conducted in Canada, recruited 63 patients with upper motor neuron dysfunction who met the El Escorial criteria for probable or definite ALS. Forty-nine patients had limb-onset symptoms and 14 patients had bulbar-onset symptoms; the median age was 61.0 years, and 44 patients were male. Between July 1996 and May 2002, all patients underwent MRSI scanning to determine the levels of creatine (Cr), choline (Cho) and NAA in the primary motor cortex. At the time of the scan, symptom duration was <5 years. Patients were then followed up until March 2004. The control group comprised 18 age-matched and gender-matched people without neurological disorders.OUTCOME MEASURES
The main outcome variables were the metabolic ratios of cortex NAA:Cho, NAA:Cr and Cho:Cr, and the correlations between these metabolic ratios and survival.RESULTS
Compared with healthy controls, patients with ALS had a 13% reduction in metabolic NAA:Cho ratio (P <0.001), a 5% reduction in NAA:Cr ratio (P = 0.01) and an 8% increase in Cho:Cr ratio (P = 0.01), with median values of 2.11 (interquartile range 1.87-2.30), 2.34 (2.17-2.47) and 1.15 (1.06-1.23), respectively. During the study period, 36 patients with ALS died; the median survival after MRSI in the ALS group was 24 months. The NAA: Cho ratio had the best sensitivity-specificity profile (sensitivity 67%, specificity 83%) and was therefore used for the subsequent analyses. Patients with NAA:Cho <2.11 survived for a median of 19.4 months, and their risk of death was twice that of patients with NAA:Cho >2.11 (hazard ratio 2.05, 95% CI 1.12-4.03; P = 0.02). Multivariate Cox proportional hazards analysis identified a decreased NAA:Cho ratio, advanced age and a shorter symptom duration as independent risk factors for mortality, with hazard ratios of 0.24 (95% CI 0.08-0.72; P = 0.01), 1.03 (95% CI 1.00-1.06; P = 0.04) and 0.96 (95% CI 0.93-0.99; P = 0.01), respectively. Gender and site of the initial symptoms did not have any impact on survival. In a subgroup of 32 patients, survival analysis was performed to test the effects of ALS Functional Rating Scale score and forced vital capacity (both determined within 2 weeks of the MRSI scan) in addition to those of NAA:Cho. In this subanalysis, only reduced NAA:Cho remained an independent predictor of survival (hazard ratio 0.14, 95% CI 0.03-0.68, P = 0.02).CONCLUSION
In patients with ALS, neurochemical evidence of impaired neuronal integrity in the motor cortex predicts a reduced survival.