Rituximab Treatment and Long-term Outcome of Patients With Autoimmune Encephalitis: Real-world Evidence From the GENERATE Registry

Abstract

Background and objectives: To determine the real-world use of rituximab in autoimmune encephalitis (AE) and to correlate rituximab treatment with the long-term outcome.

Methods: Patients with NMDA receptor (NMDAR)-AE, leucine-rich glioma-inactivated-1 (LGI1)- AE, contactin-associated protein-like-2 (CASPR2)-AE, or glutamic acid decarboxylase 65 (GAD65) disease from the GErman Network for Research on AuToimmune Encephalitis who had received at least 1 rituximab dose and a control cohort of non-rituximab-treated patients were analyzed retrospectively.

Results: Of the 358 patients, 163 (46%) received rituximab (NMDAR-AE: 57%, CASPR2-AE: 44%, LGI1-AE: 43%, and GAD65 disease: 37%). Rituximab treatment was initiated significantly earlier in NMDAR- and LGI1-AE (median: 54 and 155 days from disease onset) compared with CASPR2-AE or GAD65 disease (median: 632 and 1,209 days). Modified Rankin Scale (mRS) scores improved significantly in patients with NMDAR-AE, both with and without rituximab treatment. Although being more severely affected at baseline, rituximab-treated patients with NMDAR-AE more frequently reached independent living (mRS score ≤2) (94% vs 88%). In LGI1-AE, rituximab-treated and nontreated patients improved, whereas in CASPR2-AE, only rituximab-treated patients improved significantly. No improvement was observed in patients with GAD65 disease. A significant reduction of the relapse rate was observed in rituximab-treated patients (5% vs 13%). Detection of NMDAR antibodies was significantly associated with mRS score improvement. A favorable outcome was also observed with early treatment initiation.

Discussion: We provide real-world data on immunosuppressive treatments with a focus on rituximab treatment for patients with AE in Germany. We suggest that early and short-term rituximab therapy might be an effective and safe treatment option in most patients with NMDAR-, LGI1-, and CASPR2-AE.

Class of evidence: This study provides Class IV evidence that rituximab is an effective treatment for some types of AE.

Figure 1. Study Population Profile

Patient numbers in the different study subpopulations are depicted. *Patients excluded because of insufficient data on rituximab dosing (n = 2), concomitant diagnosis of MS (n = 2), retraction of consent for the GENERATE registry (n = 1), or not fulfilling the ab criteria for inclusion (n = 9). **Patients excluded because of insufficient data on immunosuppressive treatment (n = 1) or not fulfilling the ab criteria for inclusion (n = 31). CA, cerebellar ataxia; CASPR2 = contactin-associated protein-like-2; Enc. = encephalitis; GAD65 = glutamic acid decarboxylase 65; GENERATE = GErman Network for Research on AuToimmune Encephalitis; LGI1 = leucine-rich glioma-inactivated-1; NMDAR = NMDA receptor; SPS = stiff-person syndrome.

Figure 2. Venn/Euler Diagrams Showing Applied Mono- and Combination First-Line and Second-Line Immunotherapies

The numbers of patients treated with the respective prior first-line immunotherapies (A–H) and second-line immunotherapies (I–P) in the rituximab cohort (A–D) and in the control cohort (E–H) are depicted for the different ab subgroups (A, E, I, and M: NMDAR-AE; B, F, J, and N: GAD65 disease; C, G, K, and O: LGI1-AE; and D, H, L, and P: CASPR2-AE). Other second-line therapies included bortezomib (n = 6 in patients with NMDAR-AE treated with rituximab), daratumumab (n = 1 in patients with NMDAR-AE treated with rituximab), tacrolimus (n = 1 in patients with GAD65 disease treated with rituximab and n = 1 in patients with GAD65 disease not treated with rituximab), and basiliximab (n = 1 in patients with GAD65 disease treated with rituximab). Areas of Venn diagrams are proportional to the case numbers relative to the respective subgroup. (A–H) Proportions of combination first-line therapy relative to none/monotherapy were compared using the Fisher exact test. ***p < 0.001, **p < 0.01, and *p < 0.05. (I–P) Proportions of treatment with cyclophosphamide or other therapies relative to steroid-sparing therapies and no treatment were compared using the Fisher exact test. ***p < 0.001, **p < 0.01, and *p < 0.05. AZA = azathioprine; CASPR2 = contactin-associated protein-like-2; cyc = cyclophosphamide; GAD65 = glutamic acid decarboxylase 65; IVIG = IV immunoglobulin; MMF = mycophenolate mofetil; MTX = methotrexate; NMDAR = NMDA receptor; LGI1 = leucine-rich glioma-inactivated-1; PLEX = plasma exchange.

Figure 3. Rituximab Regimens Used in Patients With AE and the Outcome According to Subtypes of AE

(A–F) In different subgroups (NMDAR-AE, GAD65 disease, LGI1-AE, and CASPR2-AE), the duration in days from disease onset to initiation of rituximab treatment (A), the duration in days from initiation of first-line therapy to initiation of rituximab treatment (B), the number of rituximab infusions (C), the total cumulative rituximab dose (D), the duration in days from the first to the last rituximab infusion (E), and the number of patients receiving induction therapy (rituximab treatment <6 months) or induction + maintenance therapy (rituximab treatment ≥6 months) (F) are depicted. Bars indicate the median. Normality testing was performed using the D'Agostino-Pearson omnibus test. Continuous variables were compared using the Kruskal-Wallis test followed by the Dunn multiple comparisons test, and ordinal variables were compared using the Fisher exact test. ****p<0.0001 ***p < 0.001, **p < 0.01, and *p < 0.05. (G) mRS scores in the different ab subgroups were compared in the rituximab cohort and in the control cohort. The distribution of mRS scores is depicted at 4 time points: I, maximal mRS at symptom onset; II, mRS at initiation of rituximab treatment (from −2 months to +4 months from rituximab onset); III, mRS 4–12 months after initiation of rituximab treatment; IV, mRS at last follow-up with at least >12 months after rituximab treatment. The line represents the change in mRS scores dividing favorable mRS scores (0–2) and nonfavorable mRS scores (≥3). The ordinal χ² test was applied to compare the distribution of mRS scores. ****p<0.0001 ***p < 0.001, **p < 0.01, and *p < 0.05. CASPR2, contactin-associated protein-like-2; GAD65 = glutamic acid decarboxylase 65; mRS = modified Rankin Scale; NMDAR = NMDA receptor; LGI1 = leucine-rich glioma-inactivated-1.