Investigating Cerebellar Atrophy by MRI as a Biomarker in Stiff Person Syndrome Spectrum Disorder and Anti-GAD65-associated Cerebellar Ataxia
Jonathan Krett1, Daniela Riveros1, Herbert Chen1, Hanyeh Afshar1, Doris Lin2, Scott Newsome1
1Neurology, 2Radiology and Radiological Science, Johns Hopkins University School of Medicine
Objective:
To determine whether cerebellar atrophy by MRI is associated with stiff person syndrome spectrum disorder (SPSD) phenotype.
Background:
Cerebellar and brainstem dysfunction are associated with greater disability in stiff person syndrome spectrum disorder (SPSD). Biomarkers, including MRI findings, that herald a more aggressive disease course may facilitate earlier identification of high-risk patients.
Design/Methods:
We identified patients in our stiff person syndrome (SPS) center database (1997-2024) with ≥1 brain MRI available. We included patients with SPSD: classic SPS, partial SPS, SPS-plus, and progressive encephalomyelitis with rigidity and myoclonus (PERM). Anti-GAD65-associated pure cerebellar ataxia (pCA) patients were also included. The first available and up to one follow-up brain MRI were reviewed by a neuroradiologist blinded to patient diagnosis/phenotype for qualitative determination of disproportionate cerebellar atrophy. Classic SPS/partial SPS (“classic SPS”) and SPS-plus/PERM/p (“non-classic SPSD”) were grouped for statistical comparison.
Results:
Of 315 screened patients, 162 were excluded (insufficient imaging, 23 non-SPSD/pCA phenotype), resulting in a cohort n=153 (91 classic SPS, 62 non-classic SPSD as defined above).The non-classic SPSD group included 44 (71%) with SPS-plus, 8 (13%) pCA, 10 (16%) PERM. Classic SPS and non-classic SPSD groups did not differ in age, sex, or disease duration. initial MRI was present in 49/62 (79%) non-classic SPSD vs. 20/91 (22%, p<0.001) with classic SPS phenotypes. ExcludingpCA, cerebellar atrophy was present in 42/54 (77.8%) of non-classic SPSD cases. atrophy progression on follow-up MRI occurred in /33 (33%) with non-classic SPSD vs. 1 of 35 (3%, p<0.001) classic SPS at median MRI interval 29 (IQR 13-70) months and 31 (IQR 16-65; p=0.75) months, respectively.
Conclusions:
Cerebellar atrophy is more common in SPSD patients known to have a poorer prognosis related to cerebellar/brainstem dysfunction compared to those with classic SPS. Prospective volumetric studies are needed to confirm this finding and its potential implications.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.