Data-driven disease subgrouping in ALS: a multicenter cerebral functional connectivity study.

backgroundIn a clinically-heterogeneous disease such as ALS, it is crucial to identify early disease changes that impair real-world functioning. The lack of consensus across clinical approaches, coupled with the subjectiveness of their evaluation, impedes our understanding of disease processes underlying early and advanced disease. This study presents neuroimaging as a potential supplementary approach that provides objectivity to the identification and evaluation of disease stage-specific ALS subgroups.

methodsCerebral functional connectivity and its association with clinical function was evaluated in 174 ALS patients and 165 healthy controls enrolled in the Canadian ALS Neuroimaging Consortium (CALSNIC). Participants were subgrouped using two approaches: (1) a data-driven hierarchical clustering of cerebral activation and 2) contemporary clinical criteria. The data-driven approach utilized data from resting-state functional magnetic resonance imaging. The clinical approach utilized three clinical subgrouping methods - two derived from trial enrollment criteria for the drugs Riluzole and Edaravone, and the third on the median disease progression rate of the patient sample.

resultsEach subgrouping approach identified two patient subgroups with different symptom durations, disease progression rates, and cognitive/motor/lung functions - albeit with differences across approaches. The data-driven approach identified greater spatial extents of cerebral connectivity alterations compared to the clinical approaches.

conclusionObservations of clinical and cerebral connectivity differences were specific to the stratification approach. Given the ability of the data-driven approach to identify alterations in both clinical and cerebral function corresponding to disease stage, this approach presents a potential biomarker for patient stratification, clinical trial enrichment, disease and therapeutic monitoring.