Aim: Amyotrophic lateral sclerosis (ALS) is fatal late-onset neurodegenerative disorder of adult life, characterized by a progressive impairment of motor function. We recently developed new software able to recognize the spinal canal and spinal cord (SC) tracer uptake on PET/CT images. Our study aims to investigate whether this method permits to identify abnormalities in SC metabolism in ALS patients.

Materials and Methods: We studied 30 patients with spinal-onset ASL at different clinical stage, submitted to FDG-PET/CT following 1-36 months (median 14) after diagnosis. Obtained data were compared with corresponding findings in age and sex-matched healthy controls selected from a published normalcy database. Image analysis was performed according to a previously validated algorithm able to identify all vertebrae from the image data set and to extract the spinal canal as the non- osseous space within the spine volume. The output of the software was therefore the extraction and the 3D-representation of spinal canal volume that served as a mask to recognize SC using a segmentation algorithm based on Hough transformation. Thereafter, mean standardize uptake value (SUV) of cervical and dorsal SC, normalized to the liver (NSUV), was evaluated in comparison with normal subjects.

Results: No differences were found in SC volume between patients and controls. FDG uptake was slightly, yet significantly, higher in ALS patients in the whole SC (NSUV 0.82±0.28 vs 0.70±0.14 p<0.05) and in cervical segment (NSUV 0.99±0.37 vs 0.85±0.20, p<0.05). During follow-up 13 patients died. A potential prognostic role of SC metabolism was suggested by the observation of a higher SC_NSUV in non-survivors compared with the survivor patients (0.71±0.26 vs 0.55±0.16, p<0.05). Kaplan-Meyer approach confirmed the predictive value of SC-NSUV while multivariate analysis confirmed the additive nature of metabolic information (HR = 24.3, 95% CI 2.2-262.8). By contrast no association with prognosis was observed for age, ALSFRS-R score, time elapsed from diagnosis to PET scanning or presence/absence of riluzole treatment.

Conclusion: Our computational approach might represent a new window to explore SC metabolism from PET/CT images. Whether confirmed in larger prospective studies, the prognostic significance of SC metabolic pattern in ALS patients might suggest a relevant role for SC inflammatory response in ALS progression.