Recent imaging studies in multiple sclerosis have focused on the association between cognitive dysfunction and grey-matter tissue loss, most notably of the cerebral cortex and thalamus. For example, at the American Academy of Neurology annual meeting, an analysis of the SPRINT-MS trial proposed that lower thalamic volume at baseline may be a useful predictor of physical, cognitive and visual disability in progressive MS (Nicholson et al. AAN 2022;S12.007). Thalamic volume at the time of relapse may also be predictive of cognitive recovery as assessed by the Symbol Digit Modalities Test (SDMT) (Weinstock et al. AAN 2022;P14.010). As such, thalamic volume (but not T2 lesion volume in this study) may serve as a marker of cognitive reserve.
Of emerging interest is cerebellar atrophy and its impact on cognition. A prospective study of 164 RRMS and PMS patients reported that atrophy of the cerebellum and surrounding structures was associated with lower cognitive scores (Bonacchi et al. AAN 2022;S14.006; printed in J Neurol 2022, epublished March 1, 2022).
However, determining the anatomic correlate of individual cognitive impairments and the relationship between cognitive and physical disability may be challenging. For example, a recent analysis of the EXPAND dataset of siponimod in SPMS reported no association between MRI measures of brain atrophy or lesion burden and EDSS or SDMT worsening (Koch et al. Mult Scler 2022;28:561-572). However, brain volume loss at 48 weeks was associated with significant worsening on the 9-Hole Peg Test (9HPT).
Conflicting findings may be due, in part, to the imaging technique used and the cognitive domain being evaluated. A recent study reported that cortical thinning in specific areas was associated with different domains of cognitive dysfunction (Davion et al. Rev Neurol (Paris) 2022;178:326-336). Cortical thinning in the left frontal superior and right precentral gyri was associated with impaired visuo-spatial memory; thinning in the left Broca’s area and the right superior temporal gyrus was associated with impaired verbal memory. Worsening on the 9HPT was associated with cerebellar damage in a retrospective analysis of the CombiRx trial, which may suggest that poorer test scores were the result of cerebellar symptoms rather than impaired cognition (Petracca et al. Eur J Neurol 2022;29:515-521). In contrast, deterioration in language skills may indicate damage to short association/U-fibres (Buyukturkoglu et al. AAN 2022;P18.010).
Impaired information processing speed appears to be a better indicator of disability progression. A recent prospective study of 85 MS patients reported that worse performance in information processing speed at baseline was a significant independent predictor of higher EDSS scores at follow-up (Hechenberger et al. Mult Scler Relat Disord 2022;57:103353). Cognitive assessment was with the Brief Repeatable Battery of Neuropsychological Tests (BRB-N), which comprises the selective reminding, 10/36 spatial recall, SDMT, PASAT and word list generation tests.
In clinical practice, periodic SDMT assessments may be employed to identify patients at risk of EDSS progression. This was demonstrated in a post-hoc analysis of the EXPAND extension, which stratified patients in quartiles according to baseline EDSS score (Penner et al. AAN 2022;P5.006). Patients with the worst baseline SDMT score had a 1.8-fold higher risk of progression to EDSS >7 versus those with the lowest baseline SDMT score; the predictive value was lower (HR 1.12) in the siponimod-treated group, which may be an indication of fewer disabling events with therapy.
A recent suggestion has been to combine SDMT with other measures of progression to provide a better indication of an individual patient’s treatment response. In a post-hoc analysis of EXPAND, there were significant improvements in EDSS and SDMT, and less impact on T25FW and 9HPT for the study group as a whole (Bovis et al. AAN 2022;P16.008). However, the response to individual tests varied widely and a majority of siponimod-treated patients had a significant response on a least one of the four assessments. This suggests that a broader range of investigations may be needed to detect the impact of treatments on physical and cognitive progression.