Whole-brain cortical morphometry and structural covariance network disruption in subacute sclerosing panencephalitis

Subacute sclerosing panencephalitis (SSPE) is a progressive, fatal encephalitis caused by persistent measles virus infection. Quantitative cortical imaging in SSPE remains poorly characterised, and the topology of the cortical structural covariance network (SCN) has not been examined.

To characterise whole-brain cortical structural alterations in SSPE using surface-based morphometry (SBM) and voxel-based morphometry (VBM), evaluate their clinical correlates, and describe the topology of the cortical SCN.

Twenty-eight patients with SSPE and 48 age- and sex-matched healthy controls underwent 3.0 T volumetric T1-weighted MRI. SBM was performed in FreeSurfer using the Desikan-Killiany (DK) atlas, adjusted for age and sex; VBM used the FSL-VBM pipeline with threshold-free cluster enhancement, adjusted for age, sex and total intracranial volume. Clinical associations used Pearson correlation with false discovery rate (FDR) correction and Spearman sensitivity analyses; severity was quantified with the Neurological Disability Index. Group-level SCNs were constructed from age- and sex-adjusted regional thickness correlations; graph metrics were computed across 16 network densities and summarised as the area under the metric-versus-density curve, so that findings do not depend on a single arbitrary threshold. Group differences were assessed by subject-label permutation testing with bootstrap confidence intervals.

VBM identified widespread bilateral grey matter volume reduction extending from cortex into deep grey matter and cerebellum, with additional right-lateralised clusters in the planum temporale and frontal pole. Vertex-wise SBM demonstrated 35 bilateral cortical thickness clusters and 21 cortical volume clusters; in the covariate-adjusted regional analysis, 38 of 68 DK regions showed significant thinning after FDR correction, and 18 of 68 regions showed significant volume reduction, most prominently the bilateral frontal pole and posterior cortices. Cortical thickness and volume correlated negatively with disease duration across most regions, but not with clinical severity. SCN metrics differed consistently in direction across all densities - longer characteristic path length, higher modularity and local efficiency, and loss of degree assortativity in patients - but no metric difference survived subject-label permutation testing after FDR correction (smallest adjusted p = 0.24); SCN findings are therefore reported as descriptive.

SSPE produces widespread, duration-dependent cortical thinning and grey matter volume loss involving association, limbic and primary cortices. Group-level SCN topology showed a consistent directional pattern of network reorganisation that did not reach corrected statistical significance at this sample size. Prospective, longitudinal and multi-centre studies are required before any clinical application of these measures can be considered.