Antibody Prevalence and Immunotherapy Response in Chinese Patients with Epilepsy and Encephalopathy Scores for Patients with Different Neuronal Surface Antibodies
Introduction
In recent decades, the discovery of specific neurological autoantibodies has significantly advanced our understanding of the relationship between immune mechanisms and autoimmune encephalopathy or epilepsy. The International League Against Epilepsy (ILAE) officially recognized immune etiology as one of the six major etiological groups of epilepsy in 2017. Studies have shown that over 10% of epilepsy cases may have an underlying autoimmune origin, with a significant proportion of patients with epilepsy of unknown etiology testing positive for neuronal antibodies. Early diagnosis and timely immunotherapy are critical for improving clinical outcomes in these patients. To aid in this process, scoring systems such as the Antibody Prevalence in Epilepsy (APE) score, its modified version (APE2), and the Antibody Prevalence in Chinese Patients with Epilepsy and Encephalopathy (APE2-CHN) score have been developed. These tools aim to predict the likelihood of autoimmune encephalopathy or epilepsy and guide immunotherapy decisions. This study evaluates the application of APE2 and APE2-CHN scores in Chinese patients with epilepsy and encephalopathy, focusing on their ability to predict the presence of neuronal surface antibodies and assess immunotherapy responses.
Methods
The study enrolled 1,365 patients with epileptic seizures as the primary feature at Xuanwu Hospital, Capital Medical University, from June 2016 to June 2020. Patients with metabolic abnormalities or structural brain lesions explaining their seizures were excluded. Among them, 915 patients with epilepsy of unknown etiology underwent testing for autoimmune antibodies in serum and/or cerebrospinal fluid (CSF). Of these, 191 patients tested positive for neuronal surface antibodies, including N-methyl-D-aspartate receptor (NMDAR) antibodies, leucine-rich glioma-inactivated protein 1 (LGI1) antibodies, contactin-associated protein 2 (CASPR2) antibodies, AMPA receptor (AMPA2R) antibodies, and gamma-aminobutyric acid B receptor (GABAR-B) antibodies. Patients with non-specific neuronal antibodies, such as Hu, Yo, and glutamic acid decarboxylase 65 (GAD65) antibodies, were excluded. All patients were assessed using APE2, APE2-CHN, Response to Immunotherapy with Epilepsy and Encephalopathy (RITE2), and RITE2-CHN scores based on clinical, imaging, and laboratory data. Treatment efficacy was evaluated using the modified Rankin score (mRS), with responders defined as those showing a ≥1-point improvement in mRS or a >50% reduction in seizure frequency after six months of immunotherapy.
Results
Of the 915 patients, 191 (20.87%) tested positive for neuronal surface antibodies. The most common antibodies were NMDAR (115 patients), followed by LGI1 (47 patients), CASPR2 (8 patients), AMPA2R (4 patients), and GABAR-B (11 patients). Six patients had multiple antibodies. The sensitivity and specificity of APE2 ≥4 in predicting neuronal surface antibodies were 74.35% and 81.77%, respectively, while APE2-CHN ≥4 showed sensitivity and specificity of 75.92% and 84.53%, respectively. Eight patients with APE2 scores <4 had APE2-CHN scores ≥5, all of whom exhibited memory decline as a prominent symptom. APE2-CHN scores demonstrated higher sensitivity for predicting NMDAR antibodies but lower sensitivity for LGI1 antibodies. Among the 191 antibody-positive patients, 187 (97.91%) received immunotherapy, with 142 (74.35%) showing clinical improvement. Patients with LGI1 antibodies and RITE2-CHN scores ≥8 responded particularly well to immunotherapy.
Discussion
The study highlights the utility of APE2 and APE2-CHN scores in predicting the presence of neuronal surface antibodies in Chinese patients with epilepsy and encephalopathy. Both scores showed similar predictive accuracy, but APE2-CHN was more effective in identifying patients with memory decline, speech disorders, and decreased consciousness. The APE2-CHN score, which incorporates additional clinical variables, proved particularly valuable for patients with complex symptoms. However, the predictive value of these scores varied depending on the type of antibody. APE2 and APE2-CHN scores were most sensitive for NMDAR antibodies but less effective for CASPR2 antibodies. This discrepancy may be due to the distinct clinical presentations associated with different antibodies. For example, anti-CASPR2 encephalitis often involves peripheral nerve damage, which may not be fully captured by the scoring systems.
In terms of immunotherapy response, RITE2 and RITE2-CHN scores were useful predictors, particularly for patients with LGI1 antibodies. However, high APE2 or APE2-CHN scores were associated with poorer immunotherapy outcomes, likely reflecting the severity of clinical symptoms in these patients. The study also found that patients with abnormal mental behavior, dystonia, and decreased consciousness were less likely to respond to treatment, underscoring the importance of early intervention in autoimmune encephalopathy.
Limitations
The study has several limitations. First, it focused exclusively on patients with specific neuronal surface antibodies, excluding those with non-specific antibodies such as Hu, Yo, and GAD65. Second, the number of patients with certain antibodies, such as AMPA2R, was small, which may have affected the statistical power of the analysis. Finally, the study was conducted at a single center, which may limit the generalizability of the findings.
Conclusion
APE2-CHN and RITE2-CHN scores are valuable tools for predicting the presence of neuronal surface antibodies and assessing immunotherapy responses in Chinese patients with epilepsy and encephalopathy. APE2-CHN is particularly useful for patients with cognitive and speech impairments, while RITE2-CHN is effective in predicting treatment outcomes for anti-LGI1 encephalitis. However, high APE2 or APE2-CHN scores may indicate a poor response to immunotherapy, emphasizing the need for early diagnosis and intervention in autoimmune encephalopathy. Future studies should aim to validate these findings in larger, multicenter cohorts and explore the utility of these scoring systems in patients with non-specific neuronal antibodies.
doi.org/10.1097/CM9.0000000000001701
Was this helpful?
0 / 0