Lesion-Independent Thalamic Degeneration Identifies Intrinsically Vulnerable Nuclei Associated with Cognitive Impairment in Multiple Sclerosis

Abstract: Cognitive impairment in multiple sclerosis (MS) is driven by both focal inflammation and compartmentalized neurodegeneration, yet the relative effect of lesion-independent thalamic atrophy on information processing speed (IPS) remains unclear. This retrospective cohort study included 100 participants with MS. Automatic segmentation techniques quantified lesion load and delineated 26 thalamic regions of interest (ROIs). Linear models compared associations between ROI volumes and Symbol Digit Modalities Test (SDMT) performance in lesion-adjusted and unadjusted models. Twenty-one of 26 ROIs showed significant SDMT associations before lesion adjustment; twelve remained significant after adjustment. Lesion-independent associations were observed in the global thalamus, sensory relay nuclei (ventral posterolateral, medial and lateral geniculate), and associative hubs (pulvinar and mediodorsal-parafascicular complex). These intrinsically vulnerable nuclei exhibited significantly lower lesion-mediated effects (13.4%) than those losing significance after adjustment (34.2%, p < 0.001). Our findings suggest that IPS impairment reflects heterogenous contributions from both primary and secondary degeneration, with nucleus-specific phenotyping potentially informing identification of higher risk individuals.

• The paper identified that after lesion adjustment, twelve thalamic nuclei remained significantly associated with slowed information processing, highlighting intrinsic vulnerability.
• The study employed automated MRI segmentation and ANCOVA to distinguish lesion-dependent from lesion-independent degeneration in a cohort of 100 MS patients.
• Results suggest that targeting intrinsically vulnerable thalamic hubs may refine risk stratification and guide neuroprotective interventions in MS.

Lesion-Independent Thalamic Degeneration and Cognitive Impairment in Multiple Sclerosis

Introduction

This work investigates the relationships between thalamic nuclear atrophy and cognitive impairment—specifically, information processing speed (IPS)—in multiple sclerosis (MS), using fully automated segmentation and rigorous lesion adjustment. The study addresses the relative contributions of lesion-mediated and lesion-independent thalamic degeneration, distinguishing nuclei that remain associated with IPS after controlling for lesion load. Historically, thalamic atrophy in MS has often been attributed to secondary degeneration following white matter lesion formation; however, the anatomical and functional heterogeneity of thalamic nuclei and their distinct vulnerabilities have not been adequately examined, particularly in lesion-independent contexts.

Methodology

A cohort of 100 MS patients underwent high-resolution MRI scanning and cognitive assessment using the Symbol Digit Modalities Test (SDMT), a validated measure of IPS. Automated lesion segmentation (LST-AI) and thalamic nuclei segmentation (HIPS-THOMAS) were employed, yielding volumetric data for 26 thalamic ROIs per subject, normalized to intracranial volume. Associations between ROI volumes and SDMT scores were evaluated using ANCOVA, controlling for age, sex, education, and, in adjusted models, total lesion volume. Multiple comparison correction (Benjamini-Hochberg FDR) and mediation analyses further characterized lesion-dependent and independent effects.

Major Findings

Twenty-one of twenty-six thalamic ROIs exhibited significant associations with SDMT performance prior to lesion adjustment, all indicating that reduced nuclear volume correlated with cognitive slowing (lower SDMT scores). However, after controlling for lesion burden, only twelve ROIs retained significance; these were classified as "intrinsically vulnerable" nuclei, with markedly lower lesion-mediated effects (13.4%) than ROIs losing significance (34.2%, $p < 0.001$). The preserved associations after lesion correction implicate global thalamic volume, somatosensory relays (ventral posterolateral, lateral geniculate, medial geniculate), and associative hubs (pulvinar, mediodorsal-parafascicular complex) as particularly susceptible to primary neurodegenerative processes. The lesion-independent nuclei aligned with circuits critical for rapid sensory encoding and attentional regulation—functions supported by continuous relay and integrative activity.

Numerically, the effect size was strongest in the pulvinar and global thalamus, with lateralization favoring the left hemisphere. The incremental explained variance after adding lesion load as a covariate ($\Delta R^2$) was substantially lower for intrinsically vulnerable nuclei (2.4%) compared to those with lesion-dependent associations (5.5%, $p < 0.01$).

Theoretical Implications

This study provides direct evidence for heterogeneous pathways of thalamic neurodegeneration in MS—primary, lesion-independent atrophy in select nuclei, and secondary, lesion-mediated degeneration in others. Sensory and associative nuclei that maintain significant volume-function relationships after lesion adjustment may possess intrinsic vulnerabilities attributed to high metabolic demand, continuous activity, glial dependence, and microglial reactivity, as well as proximity to CSF interfaces. This nucleus-resolved approach challenges standard interpretations of thalamic atrophy as a uniformly secondary phenomenon and supports models of compartmentalized neurodegeneration influenced by both anatomical connectivity and metabolic stress.

Trans-synaptic propagation of degenerative signals, differences in myelin and oligodendrocyte support, and localized microglial activation likely contribute to the selective atrophy patterns observed. The persistence of left-hemisphere dominance for IPS-related nuclei is consistent with specialization for symbolic processing and attentional control.

Practical Implications

Automated segmentation of thalamic nuclei enables scalable and reproducible quantification of nuclear atrophy, supporting its use as a candidate biomarker for cognitive decline in MS. The identification of lesion-independent nuclear vulnerabilities refines risk stratification, potentially enabling earlier intervention in high-risk individuals and nuclei-targeted clinical trial endpoints. Current disease-modifying therapies (DMTs) do not selectively protect these vulnerable nuclei, underlining an unmet need for therapeutics aiming at neuroprotection or glial modulation.

Limitations and Directions for Future Research

The cross-sectional design limits causal inference regarding temporal dynamics of degeneration and cognitive decline. The cohort had mild disability and was recruited from a single center, limiting generalizability. Cognitive assessment was limited to IPS; future work should incorporate broader neuropsychological batteries. It remains plausible that untapped spatial or nonlinear relationships among lesions, nuclei, and cognitive outcomes exist. Longitudinal, multicenter studies employing multimodal imaging and extended cognitive profiling will be required to validate and expand upon these findings.

Further research could elucidate whether similar lesion-independent vulnerabilities exist in other deep grey matter structures, and to test the efficacy of neuroprotective or anti-inflammatory strategies on selectively vulnerable nuclei.

Conclusion

This study fundamentally separates lesion-dependent and lesion-independent contributions to thalamic atrophy and IPS deficits in MS, demonstrating that approximately half of thalamic nuclei retain strong volume-cognition associations independent of lesion load. These intrinsically vulnerable sensory and associative hubs present novel targets for biomarker development and therapeutic intervention, with implications for understanding neurodegenerative mechanisms underlying cognitive decline in MS. The approach provides a template for future nucleus-resolved studies of grey matter pathology in demyelinating disease (2511.21677).