Comparative Study of Paraneoplastic and Nonparaneoplastic Autoimmune Encephalitis With GABABR Antibodies

Background and Objectives While patients with paraneoplastic autoimmune encephalitis (AE) with gamma-aminobutyric-acid B receptor antibodies (GABABR-AE) have poor functional outcomes and high mortality, the prognosis of nonparaneoplastic cases has not been well studied. Methods Patients with GABABR-AE from the French and the Dutch Paraneoplastic Neurologic Syndromes Reference Centers databases were retrospectively included and their data collected; the neurologic outcomes of paraneoplastic and nonparaneoplastic cases were compared. Immunoglobulin G (IgG) isotyping and human leukocyte antigen (HLA) genotyping were performed in patients with available samples. Results A total of 111 patients (44/111 [40%] women) were enrolled, including 84 of 111 (76%) paraneoplastic and 18 of 111 (16%) nonparaneoplastic cases (cancer status was undetermined for 9 patients). Patients presented with seizures (88/111 [79%]), cognitive impairment (54/111 [49%]), and/or behavioral disorders (34/111 [31%]), and 54 of 111 (50%) were admitted in intensive care unit (ICU). Nonparaneoplastic patients were significantly younger (median age 54 years [range 19–88] vs 67 years [range 50–85] for paraneoplastic cases, p < 0.001) and showed a different demographic distribution. Nonparaneoplastic patients more often had CSF pleocytosis (17/17 [100%] vs 58/78 [74%], p = 0.02), were almost never associated with KTCD16-abs (1/16 [6%] vs 61/70 [87%], p < 0.001), and were more frequently treated with second-line immunotherapy (11/18 [61%] vs 18/82 [22%], p = 0.003). However, no difference of IgG subclass or HLA association was observed, although sample size was small (10 and 26 patients, respectively). After treatment, neurologic outcome was favorable (mRS ≤2) for 13 of 16 (81%) nonparaneoplastic and 37 of 84 (48%) paraneoplastic cases (p = 0.03), while 3 of 18 (17%) and 42 of 83 (51%) patients had died at last follow-up (p = 0.008), respectively. Neurologic outcome no longer differed after adjustment for confounding factors but seemed to be negatively associated with increased age and ICU admission. A better survival was associated with nonparaneoplastic cases, a younger age, and the use of immunosuppressive drugs. Discussion Nonparaneoplastic GABABR-AE involved younger patients without associated KCTD16-abs and carried better neurologic and vital prognoses than paraneoplastic GABABR-AE, which might be due to a more intensive treatment strategy. A better understanding of immunologic mechanisms underlying both forms is needed.


Introduction
2][3][4] The clinical feature usually consists of a limbic encephalitis (LE), with antiseizure medication-resistant epileptic seizures, confusion, and anterograde amnesia, but it can also present with rapidly progressive dementia without prominent seizures. 2,3pproximately 50%-60% of patients have an underlying malignancy, most commonly small cell lung cancer (SCLC) in elderly smoker men, and generally neurologic symptoms precede and ultimately lead to the tumor diagnosis. 4,5Despite increasing knowledge on some pathophysiologic features of GABA B R-AE, 6,7 data regarding long-term outcomes are still scarce. 3,4][10] Nevertheless, whether patients with SCLC and nonparaneoplastic cases differ in clinical onset and long-term prognosis has not been studied in detail yet. 3,11,12In this study, we aimed to compare the clinical presentation, immunogenetic characteristics, and neurologic outcome of paraneoplastic and nonparaneoplastic GABA B R-AE identified in the French and the Dutch Paraneoplastic Neurologic Syndromes Reference Centers between 2011 and 2022.

Study Design and Patient Selection
We retrospectively included patients diagnosed with GABA B R-AE from January 2011 to June 2022 at 2 European Reference Network sites (ERN-RITA) for Paraneoplastic Neurologic Syndromes (PNS): the French Reference Center in Lyon (France) and the Dutch Reference Center in Rotterdam (The Netherlands).GABA B R antibodies (GABA B R-abs) were identified in CSF and/ or serum by immunohistochemistry/immunohistofluorescence using commercial and/or in-house cell-based assay (CBA), using either HEK293 or CHO cells expressing GABA B 1a and GABA B 2 subunits as previously described. 2,3In addition, antibodies against potassium channel tetramerization domain containing 16 (KCTD16-abs) were identified in CSF and/or serum with inhouse CBA, as described elsewhere. 3This observational retrospective multicentric cohort study is reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. 13ta Collection Data on clinical presentation, treatment, and results of ancillary investigations (EEG, MRI, CSF findings, presence of coexisting antibodies including KCTD16-abs in serum and/or CSF, and oncological investigations) were retrospectively collected by independent physicians in each reference center (F.L., J.K., M.C.-H., M.d.B., and J.d.V.) from medical reports obtained at the time of diagnosis and by request to patients' treating physicians at the time of the study.Diagnostic criteria by Graus et al. were applied to confirm the diagnosis of AE. 14,15 Primary and Secondary Outcomes We compared the initial characteristics, treatment, neurologic outcome, and survival of paraneoplastic and nonparaneoplastic cases, only in patients for whom the presence of a cancer was confirmed or considered to have been sufficiently explored.Duration of follow-up was defined from symptom onset to the last medical visit or death.A favorable neurologic outcome was considered as a modified Rankin scale (mRS) ≤ 2 after immunotherapy and/or oncological treatment, whereas neurologic improvement was defined as an mRS improvement of ≥1 point.An episode of relapse was defined as a new onset or worsening of encephalitic symptoms after an initial improvement or stabilization of at least 2 months.

Immunoglobulin G Isotypes and Human Leukocyte Antigen Genotyping
In a subset of French patients for whom samples were available, immunoglobulin G (IgG) isotyping and human leukocyte antigen (HLA) genotyping were performed.IgG isotyping was performed using mouse anti-human antibodies that specifically recognize IgG1 or IgG4, as previously described. 16Genotypes at four-digit resolution for HLA class I (loci A, B, and C) and class II (loci DPA1, DPB1, DQA1, DQB1, DRB1, DRB3, DRB4, and DRB5) were imputed from available genome-wide association data (GWAS); allele carrier frequencies were then compared between patients (entire cohort and according to cancer status) and controls (239 healthy participants provided by the Stanford Center for Sleep Sciences and Medicine) using logistic regression controlled by principal component analysis; multiple comparisons were corrected by the Bonferroni method, and corrected p < 0.05 were considered statistically significant. 17LA analysis was performed with R software.

Statistical Analysis
Quantitative variables are reported as mean (SD) or median (range or interquartile range), while qualitative variables are reported as number (percentage).Paraneoplastic and nonparaneoplastic cases were compared using the Fisher exact test and the Student t-test.Neurologic outcome and survival were compared among the 2 groups using logistical regression and Kaplan-Meier curves, respectively, and were secondary adjusted for confounding variables (age, sex, immunosuppressive drugs [including second-line immunotherapy or chemotherapy], and ICU admission) using multivariate logistical regression and Cox model, respectively.Statistical significance was established as p < 0.05.Statistical analyses were performed using jamovi 18 and R version 4.0.3(R Foundation for Statistical Computing, Vienna, Austria).

Standard Protocol Approvals, Registrations, and Patient Consents
Approval from the ethics committee of the Hospices Civils de Lyon (IRB00013204) was obtained.Informed consent was not required, but information about the study was transmitted to the patients or their physician.

Data Availability
Patient-related data will be shared on reasonable request from any qualified investigator, maintaining anonymization of the individual patients.

General Cohort
We retrospectively identified 114 patients with GABA B Rabs in the French and Dutch databases and finally included 111 patients in the analysis (Figure 1); among them, a total of 54 patients (22 French and 32 Dutch) have previously been reported. 2,3Three patients were excluded because GABA B R-abs were doubtful and suggested by only 1 CBA test in the serum.Immunofluorescences were negative, and the CSF was negative for 2 patients and unavailable for the last one.We finally considered this CBA testing in the serum as false positive results; the clinical pictures of these 3 patients were clearly explained by the presence of another autoantibody (Hu-abs in 2 patients and VGCC-abs in 1).The main clinical and paraclinical characteristics of the 111 included patients are presented in Table 1.Median age at onset was 66 years (range 19-88), and 44 of 111 (40%) were women.Seizures were the first symptom at disease onset in 88 of 111 (79%) patients; 55 of 88 (63%) had generalized tonic-clonic seizures, 11 of 88 (13%) had focal

Cancer Associations
A total of 71 of 111 (64%) patients were diagnosed with cancer; 64 had SCLC (90%), the remaining ones presenting with lepidic pulmonary adenocarcinoma, poorly differentiated lung adenocarcinoma, small cell neuroendocrine prostatic cancer, neuroendocrine rectal cancer, neuroendocrine medullary thyroid carcinoma, pancreatic adenocarcinoma, and malignant thymoma (one each, 1%); all of them were considered paraneoplastic cases.Furthermore, 13 of 111 patients (12%) had a diagnosis of probable lung cancer because a lung and/or mediastinal mass suspicious of malignancy was identified on a thoracic CT scan and/or PET scan.These patients also presented asthenia and loss of weight but did not reach a final tissuebased diagnosis; 9 patients died before a definitive diagnosis could be made, while 4 patients were still alive at last follow-up (only initial data without follow-up for 1 patient; 5 months, 7 months, and 8 months of follow-up for the 3 others).Among them, 1 patient had a biopsy failure and 3 patients had unconclusive biopsies.These 13 patients were also classified as paraneoplastic cases.
Conversely, 18 of 111 patients (16%) were diagnosed as nonparaneoplastic GABA B R-AE because there was no evidence of an underlying cancer after thorough screening, which in all but one case included a PET scan and/or a full-body CT scan, and a median (interquartile range  No treatment 36/107 (34)   Abbreviations: IQR = interquartile range, SCLC = small cell lung cancer.
[IQR]) follow-up of 22 (7-37) months; the remaining patient died, and no evidence of cancer was found during autopsy.Nine patients (9/111, 8%) could not be classified because of insufficient tumor screening or missing data.

Treatment and Outcomes
First-line immunotherapy (IV immunoglobulins, steroids, and/or plasma exchange) was initiated with a median (IQR) of 27 (16-43)
IgG subclasses 1 and 4 for GABA B R-abs were analyzed in 10 patients, and all samples were positive for IgG1 and negative for IgG4 (7 nonparaneoplastic and 3 paraneoplastic; 8 in CSF and 2 in serum).

Discussion
In this study, the characteristics and prognosis of patients with paraneoplastic and nonparaneoplastic GABA B R-AE were investigated.We first showed that nonparaneoplastic forms of GABA B R-AE can affect younger patients and seem to have better neurologic outcome after immunotherapy.We then confirmed KCTD16-abs as a strong biomarker of an underlying SCLC, as previously described 3 ; KCTD16-abs should therefore be requested in all cases of GABA B R-AE without a known tumor, and their presence should warrant thorough and repeated tumor screening.
Although there was no difference between groups regarding the disease severity, nonparaneoplastic cases were more likely to receive second-line immunotherapy; this may reflect a tendency to use more aggressive treatments in younger patients with no oncological comorbidity or to avoid secondline immunotherapy in patients with cancer receiving chemotherapy, because of the risk of drug interactions leading to increased toxicity.Adjusted analyses on confounding factors suggested that a younger age and a more intensive treatment strategy (i.e., immunosuppressive drugs as second-line immunotherapy or chemotherapy) were associated with a better survival, arguing toward the use of immunosuppressive treatments in both paraneoplastic and nonparaneoplastic cases of GABA B R-AE. ICU admission was also found to be associated with a worse neurologic outcome, which may be explained by the occurrence of ICU complications, known to be strongly associated with worse long-term functional outcomes of patients with an AE, more than the individual AE subtype. 10 interest, we found 2 different demographic profiles in the nonparaneoplastic group: young nonsmoking women and older smoking patients, the latter were similar to typical paraneoplastic GABA B R-AE patients.][21] However, the absence of KCTD16-abs in these patients argues against this hypothesis.Among the nonparaneoplastic group, 5 of 18 patients had finally a duration of follow-up of only 3-5 months that might be insufficient to clearly exclude a paraneoplastic origin because many PNS may precede tumor diagnosis of several months to years. 15Furthermore, the lack of an obvious environmental trigger in nonparaneoplastic GABA B R-AE could suggest the existence of a genetic predisposition.However, we did not find any HLA association, although the number of patients included in the HLA genotyping was very low.Notably, all patients (paraneoplastic and nonparaneoplastic) had GABA B R-abs of IgG1 subtype.8][29][30][31][32] Nevertheless, nonparaneoplastic GABA B R-AE cases are scarcer in European descent populations [1][2][3]5,11,30,33 than in Asian ones, 8,9,[27][28][29]31,32,34,35 which we also observed in our own series because all paraneoplastic cases with available DNA were of European descent and, conversely, almost 40% of nonparaneoplastic patients were not of European descent. This diference might reflect distinct HLA associations that we were unable to detect possibly due to the small sample size or might be related to other genetic loci or even environmental triggers (e.g., less smoking in Asian populations 31 ); further immunogenetic research might be particularly revealing in non-European descent populations.In this study, patients classified as nonparaneoplastic underwent thorough and repeated tumor screening during an extended period of follow-up, except for 1 elderly patient who died early without treatment and for whom KCTD16-abs were positive but without any proof of cancer at autopsy (admittedly, a microscopic tumor could have been missed). Lower rates of paranelastic cases in previous studies could also reflect either methodological issues such as insufficiently extensive tumor screening or too short follow-ups.
The rate of relapse in our study (14%, similar between paraneoplastic and nonparaneoplastic cases) was also lower than previously described (19%-33%) in 3 other Asian series of patients with GABA B R-AE. 9,31,36 However, only 3 of 166 patients (2%) in these studies were treated with second-line immunotherapy compared with ours (27%), which could explain the more frequent relapsing episodes and support the early use of second-line immunotherapies to prevent relapses; this has been recently suggested in other subtypes of AE. 37 A long delay (more than 28 days) between disease and first-line immunotherapy was also hypothesized to raise the risk of relapse; however, the nonparaneoplastic group in our study evolved more favorably than the paraneoplastic group, even if on average they were treated later.The mortality rate found in our study (49%) was consistent with previously published studies from Northern countries (40-57%) 2,3,11 ; the lower mortality found in some studies from Southern countries (23-42%) 8,9,31 may be due to the greater frequency of nonparaneoplastic forms in Asian descent population.
Our study has some limitations, most of them being inherent to the retrospective methodology.Even if low in our study and not included in the analysis, missing data can still skew the results.To obtain the most homogeneous and comparable cohort, we deliberately excluded 3 patients with GABA B R-abs in serum because of discrepancies of immunologic testing, atypical clinical courses, and the co-occurrence of other neuronal antibodies that was believed to be more explanatory for the clinical presentation (Figure 1); this, however, may have led to omit a yet unknown neurologic phenotype of GABA B R-AE.Among the paraneoplastic group, 13 patients were included with a probable diagnosis of lung cancer, without reaching a tissue-based diagnosis of tumor; even with highly indicative clinical course and radiologic features, an erroneous diagnosis of cancer is still possible.Finally, the presence of cancer in the paraneoplastic group is also a major bias to compare the outcome and mortality because the cancer itself inevitably affects mortality of the paraneoplastic group.Finally, despite being helpful and easy to use, the choice of mRS as one of the outcome criteria might probably be insufficiently sensitive to detect clinically meaningful differences in outcome, especially in patients with prominent cognitive involvement.
This work also has some strengths, most importantly, all included cases were diagnosed or confirmed in a European Reference Network site for PNS, which uses in-house, highly reliable, diagnostic techniques. 38This European bicentric study is to our knowledge the largest Northern country cohort of GABA B R-AE.
To conclude, our study shows differences in demographic profile between paraneoplastic and some nonparaneoplastic GABA B R-AE cases, and suggests a different disease course in neurologic outcome between these 2 groups.The younger age as well as the more frequent use of second-line immunotherapy and the less common ICU admission of nonparaneoplastic patients may partly explain the latter result.We also confirmed the strong value of KCTD16-abs to identify paraneoplastic GABA B R-AE.In the light of these observations, disease mechanisms and immunogenetic features of paraneoplastic and nonparaneoplastic forms of GABA B R-AE should be further investigated to better understand these differences.
This study was performed within the framework of the LABEX CORTEX of the Université Claude Bernard Lyon 1, within the program "Investissements d'Avenir" (ANR-11-LABX-0042) operated by the ANR.SMC, VPS, and EM were supported by NIH (NIH-5U01NS120885-02).

Figure 3
Figure 3 Bar Diagram Showing the Association of Paraneoplastic Cases With KCTD16-abs

Table 2
Characteristics of Tumor Diagnosis and Oncological Treatments

Table 3
Detailed Characteristics of Nonparaneoplastic Patients

Table 3
Detailed Characteristics of Nonparaneoplastic Patients (continued)

Table 3
Detailed Characteristics of Nonparaneoplastic Patients (continued)

Table 4
Favorable Neurologic Outcome (mRS ≤2) Comparison Between Paraneoplastic and Nonparaneoplastic Cases of GABA B R-AE: Logistical Regression Model Adjusted for Confounding Factors