Comorbidities and Their Influence on Outcomes and Infectious Complications in Autoimmune Encephalitis: A Multicenter Cohort Study

Abstract

Background and objectives: Comorbidities greatly influence the course of many diseases. However, systematic data on comorbidities in patients with autoimmune encephalitis (AE) are scarce. We aimed to characterize comorbidities in patients with common AE variants and assess their influence on outcome and occurrence of infectious complications.

Methods: This multicenter, retrospective cohort study analyzed adult patients with definite anti-N-methyl-d-aspartate receptor (NMDAR), anti-leucine-rich glioma-inactivated-1 (LGI1), anti-contactin-associated protein-like-2 (CASPR2), and anti-immunoglobulin-like cell adhesion molecule-5 (IgLON5) AE registered by the GErman NEtwork for REsearch on AuToimmune Encephalitis between June 2004 and July 2023. Preexisting conditions (PECs), secondary diagnoses, and infectious complications documented during hospitalization were analyzed. Outcome was evaluated using a modified Rankin Scale (mRS), with unfavorable outcome defined as mRS >2 after a minimum of 12 months of follow-up.

Results: Among 308 patients with AE (144 NMDAR-AE, 98 LGI1-AE, 47 CASPR2-AE, and 19 IgLON5-AE), nearly half had cardiovascular and metabolic/endocrine, one-third neurologic, and one-fifth psychiatric comorbidities. Accompanying autoimmunity was observed in 12.7%. Univariable analysis showed that the presence of ≥3 PECs (OR 2.80, 95% CI 1.57-4.92), especially cardiovascular (OR 1.93, 95% CI 1.09-3.30) and psychiatric PECs (OR 3.84, 95% CI 1.96-7.31), was associated with unfavorable outcome. Multivariable regression analysis confirmed psychiatric PECs as independent risk factors (OR 4.55, 95% CI 1.99-10.60). During hospitalization, 13.6% of patients developed severe infections, although these were not associated with unfavorable outcome (OR 1.94, 95% CI 0.97-3.89). AE disease severity (OR 5.41, 95% CI 1.38-27.67) and intensive care unit admission emerged as the only independent predictors of severe infections (OR 20.76, 95% CI 7.02-75.10).

Discussion: As premorbid psychiatric conditions are main factors associated with unfavorable outcomes, these patients would highly benefit from integrated interdisciplinary treatment centers, or at least heightened awareness of these factors. Concomitant autoimmunity affecting other organs is frequent and should be sought. The risk of severe infections during the acute phase of AE is moderate and, given their lack of effect on outcome, should not justify withholding appropriate immunotherapy, even in elderly patients with comorbidities. Future prognostic models should incorporate comorbidities, particularly psychiatric ones, to enhance risk assessment and guide personalized care strategies.

Figure 1. Patient Selection

Screening of 490 patients with antibody positivity to NMDAR, LGI1, CASPR2, or IgLON5 from the GENERATE database, resulting in 308 patients remaining after applying the inclusion criteria. Definitive antibody-positive AE was determined according to the consensus criteria. AE = autoimmune encephalitis; CASPR2 = contactin-associated protein-like-2; GENERATE = GErman NEtwork for REsearch on AuToimmune Encephalitis; IgLON5 = immunoglobulin-like cell adhesion molecule-5; LGI1 = leucine-rich glioma-inactivated-1; NMDAR = N-methyl-D-aspartate receptor.

Figure 2. Prevalence of Comorbidities in 4 Autoimmune Encephalitis Subtypes, Grouped by Preexisting Conditions and Secondary Diagnoses

(A) Preexisting conditions (PECs) and (B) secondary diagnoses (SDs) categorized by affected organ systems. (C) Most common PECs and (D) most common SDs in the overall cohort. “Comorbidities” serves as an umbrella term encompassing both PECs, defined as previously diagnosed diseases that were documented in the patient's medical history before the current medical event, and SDs, defined as additional diagnoses made during hospitalization alongside the primary condition. The y-axis (A and B) or x-axis (C and D) represent prevalence (%), calculated within each subtype. Note that the stacked bars show contributions from different subtypes and may exceed 100% in total. Blue bars represent the NMDAR-AE cohort (n = 144), red bars the LGI1-AE cohort (n = 98), green bars the CASPR2-AE cohort (n = 47), and purple bars the IgLON5-AE cohort (n = 19). Diagnoses were identified directly from the GENERATE database, as well as from discharge summaries and inpatient documentation. AE = autoimmune encephalitis; CASPR2 = contactin-associated protein-like-2; GENERATE = GErman NEtwork for REsearch on AuToimmune Encephalitis; IgLON5 = immunoglobulin-like cell adhesion molecule-5; LGI1 = leucine-rich glioma-inactivated-1; NMDAR = N-methyl-D-aspartate receptor; PEC = preexisting condition; SD = secondary diagnosis.

Figure 3. Prevalence of Autoimmune Comorbidities Across 4 Subtypes of Autoimmune Encephalitis

NMDAR (n = 144, blue), LGI (n = 98, red), CASPR2 (n = 47, green), IgLON5 (n = 19, purple), and combined data across all subtypes (n = 308, orange). Autoimmune diseases are listed on the y-axis, the x-axis represents the prevalence as a percentage (%). Each bar indicates the frequency of each autoimmune comorbidity within each encephalitis subtype, highlighting subtype-specific comorbidity patterns among patients. Diagnoses were identified directly from the GENERATE database, as well as from discharge summaries and inpatient documentation. AE = autoimmune encephalitis; CASPR2 = contactin-associated protein-like-2; GENERATE = GErman NEtwork for REsearch on AuToimmune Encephalitis; IgLON5 = immunoglobulin-like cell adhesion molecule-5; LGI1 = leucine-rich glioma-inactivated-1; NNMDAR = N-methyl-D-aspartate receptor.

Figure 4. Risk Factors Associated With Unfavorable Functional Outcome in Patients With Autoimmune Encephalitis (Univariable Analysis)

Factors associated with unfavorable functional outcomes in patients with antibody-positive autoimmune encephalitis (n = 308): NMDAR (n = 144), LGI1 (n = 98), CASPR2 (n = 47), and IgLON5 (n = 19). (A) Associations of clinical and demographic factors with unfavorable functional outcome. (B) Associations of preexisting conditions, categorized by organ system, with unfavorable functional outcome. Each diamond represents the odds ratio, with horizontal lines indicating the 95% CIs. Odds ratios >1 indicate an association with an unfavorable functional outcome, defined as an mRS >2. Outcome was measured after a follow-up period of at least 12 months. Univariable analysis was performed using the Fisher exact test. Diagnoses and infectious episodes were identified directly from the GENERATE database, as well as from discharge summaries and inpatient documentation. The severity of infectious complications was retrospectively assessed using the CTCAE (v5.0) for infections. Note that the variable “≥3 immunotherapies” refers to the use during the first hospitalization (disease peak) and that the variable “infectious complication CTCAE grade ≥ III” refers only to the occurrence during the first hospitalization (disease peak). CASEmax = Clinical Assessment Scale in Autoimmune Encephalitis at disease peak; CASPR2 = contactin-associated protein-like-2; CTCAE = Common Terminology Criteria for Adverse Events; GENERATE = GErman NEtwork for REsearch on AuToimmune Encephalitis; ICU = intensive care unit; IgLON5 = immunoglobulin-like cell adhesion molecule-5; LGI1 = leucine-rich glioma-inactivated-1; mRS = modified Rankin Scale; mRSmax = modified Rankin Scale at disease peak; NMDAR = N-methyl-D-aspartate receptor; PEC = preexisting condition; SD = secondary diagnosis.

Figure 5. Factors Associated With Severe Infectious Complications in Hospitalized Patients With Autoimmune Encephalitis (Univariable Analysis)

Factors associated with severe infectious complications in patients with antibody-positive autoimmune encephalitis (n = 308): NMDAR (n = 144), LGI1 (n = 98), CASPR2 (n = 47), and IgLON5 (n = 19). (A) Associations of clinical, demographic, and therapeutic factors with severe infectious complications. (B) Associations of preexisting conditions, categorized by organ system, with severe infectious complications. Each diamond represents the odds ratio, with horizontal lines indicating the 95% CIs. Odds ratios >1 indicate an association with the occurrence of severe infectious complications, defined as CTCAE grade ≥ III. Univariable analysis was performed using the Fisher exact test. Diagnoses and infectious episodes were identified directly from the GENERATE database, as well as from discharge summaries and inpatient documentation. The severity of infectious complications was retrospectively assessed using the CTCAE (v5.0) for infections. Note that only infectious episodes during the first hospitalization (disease peak) were assessed and that the variable “≥3 immunotherapies” refers to use during the first hospitalization (disease peak). CASE_max = Clinical Assessment Scale in Autoimmune Encephalitis at disease peak; CASPR2 = contactin-associated protein-like-2; CTCAE = Common Terminology Criteria for Adverse Events; GENERATE = GErman NEtwork for REsearch on AuToimmune Encephalitis; ICU = intensive care unit; IgLON5 = immunoglobulin-like cell adhesion molecule-5; LGI1 = leucine-rich glioma-inactivated-1; mRS = modified Rankin Scale; mRS_max = modified Rankin Scale at disease peak; NMDAR = N-methyl-D-aspartate receptor; PEC = preexisting condition.