Persistent cognitive deficits in anti-LGI1 encephalitis are linked to a reorganization of structural brain networks

Importance : Despite immunotherapy, most patients with anti-leucine-rich, glioma-inactivated 1 encephalitis (LGI1-E) develop long-term cognitive deficits that persist for years after peak illness. However, the structural brain changes that underlie these deficits remain poorly understood. Objective : To study the relationship between cognitive outcomes and white matter (WM) networks in LGI1-E. Design & Setting : Cross-sectional study. German university center. Participants : 25 patients with LGI1-E (19/25 male [76%], mean age: 63 ± 12 years) and 25 age- and sex-matched healthy controls (HC), recruited between January 2013 and April 2019. Main Outcomes and Measures : Clinical assessments including the modified Rankin Scale (mRS) and Clinical Assessment Scale in Autoimmune Encephalitis (CASE); comprehensive cognitive testing; WM tractography using diffusion-weighted MRI. Results : All patients had received first-line immunotherapy, and two-thirds underwent second-line immunotherapy. Patients showed a significant reduction in mRS scores from peak illness to post-acute follow-up (z = -3.8, p < 0.001, n = 20), with 85% presenting "good" functional outcomes (post-acute mRS ≤ 2), paralleled by a significant reduction in CASE scores (z = -3.5, p < 0.001, n = 20). Despite this overall improvement, however, cognitive symptoms were highly prevalent at peak illness (95% of patients affected) and strongly persisted into the post-acute disease stage (85% affected). Neuroimaging at post-acute follow-up (median: 12 months from onset) revealed that LGI1-E is characterized by (i) significantly reduced whole-brain structural connectivity (t = -2.16, p = 0.036, d = -0.61), (ii) a cortico-subcortical hypoconnectivity cluster that strongly affects the hippocampus but also severely impacts extra-limbic brain systems, (iii) systematic limbic and extra-limbic decreases in node degree — a graph-theoretical measure of overall connectedness, and (iv) a "topological reorganization" of structural brain networks, marked by a bidirectional shift in the relative importance of individual brain regions in the network. Importantly, the extent of this network reorganization was significantly related to persistent cognitive deficits in the domains of verbal memory (r = -0.57, p = 0.007, n = 21), attention (r = -0.47, p = 0.030, n = 21), and executive functions (r = -0.60, p = 0.010, n = 17). Conclusion and Relevance : This study characterizes LGI1-E as a network disease that affects both limbic and extra-limbic brain systems and shows that a reorganization of WM networks is linked to multi-domain cognitive deficits in the post-acute disease stage — despite immunotherapy and good overall recovery. These findings highlight the need for extended treatment strategies to improve long-term cognitive outcomes and propose a sensitive new neuroimaging marker to include in prospective clinical trials. ### Competing Interest Statement The authors have declared no competing interest.