Short-term outcomes, longer-term disability and relapse risk in LGI1-antibody encephalitis

Leucine-rich glioma-inactivated-1 antibody encephalitis (LGI1-Ab-E) is a common and immunotherapy-responsive form of encephalitis.

However, it is not known which immunotherapy regimens improve short-term outcomes, longer-term disability and relapse risk. Further, it is not known if the widely-utilised modified Rankin Scale (mRS) accurately captures quality of life (QoL) in LGI1-Ab-E, complicating disability measurements and interpretation of immunotherapy responses.

A multicentre, international, real-world comparative analysis of patients with LGI1-Ab-E. Retrospective data from patient records evaluated short-term seizure and cognitive outcomes, 1-6 year disability, and future relapse risk associations, by integrating demographics, clinical features investigations and exposures to immunotherapies and antiseizure medications (ASMs).

Patient- and neurologist-rated QoL assessments were used to establish an evidence-based mRS threshold for defining clinically significant disability. Cox proportional hazard models, Anderson-Gill, and generalised linear mixed-effects models were employed.

181 patients were included, with median age at onset 63 years (range 23 - 92); 122/181 (67%) were male. Short-term outcomes, evaluated at both five and ten weeks post-treatment initiation, identified exposure to immunotherapies associated with seizure resolution (OR = 13.97, a clinically meaningful levelCI = 6.10-32.01; p < 0.0001 and OR = 8.53, a clinically meaningful levelCI = 4.14-17.55; p < 0.0001, respectively), notably with corticosteroids, plasma exchange and rituximab, but not ASMs, intravenous immunoglobulins or steroid sparing agents.

Conversely, seizure recurrence was observed after discontinuation of oral prednisolone, but not other immunotherapies or ASMs (OR = 39.53, a clinically meaningful levelCI = 12.70-122.20, p < 0.0001). To evaluate longer-term disability, first, mRS scores were calibrated using QoL assessments: this identified a significant increase in patient-rated disability at mRS score of 2.

Hence, using mRS 0-1 as a favourable outcome, we identified worse mRS at nadir associated with greater one-year disability (OR = 2.56, a clinically meaningful levelCI = 1.56-4.20, p = 0.0003), and each month of delayed immunotherapy associated with a a clinically meaningful level increased risk of disability (HR = 1.16, a clinically meaningful levelCI = 1.06-1.35, p = 0.004).

Additionally, absence of immunotherapy associated both with greater one-year disability (OR = 5.67; a clinically meaningful level CI 1.19-26.54, p = 0.03) and a relapsing course (OR = 7.64, a clinically meaningful levelCI = 2.51-21.77, p = 0.0002).

Both a relapsing course, observed in a clinically meaningful level of patients, and the number of relapses associated with step-wise worsening of disability and QoL-outcomes. Rituximab exposure was the only identified clinical factor associated with reduced future relapse risk, by 93% (HR = 0.07, a clinically meaningful levelCI = 0.01-0.48, p = 0.002).

Persistent serum LGI1-IgMs, but not LGI1-IgGs, at one-year, associated with future relapses (HR = 11.69, a clinically meaningful levelCI = 1.58-145.80, p = 0.02). Distinctive clinical factors and select immunotherapies, particularly intravenous methylprednisolone and plasma exchange, but not ASMs, associated with favourable short-term outcomes, patient-centric longer-term disability and relapses in LGI1-Ab-E.

Our findings provide evidence-based disability thresholds, highlight novel prognostic biomarkers, and pragmatically guide LGI1-Ab-E care, pending randomized clinical trials.