Antibody prevalence and immunotherapy response in Chinese patients with epilepsy and encephalopathy scores for patients with different neuronal surface antibodies

Epilepsy affects over 50 million people worldwide, and in recent years, researchers have uncovered a critical link between the immune system and certain forms of the condition: autoimmune epilepsy, where the body’s own antibodies attack neuronal surface proteins. For Chinese patients with epilepsy or encephalopathy of unknown cause, two scoring systems—Antibody Prevalence in Chinese Patients with Epilepsy and Encephalopathy (APE2-CHN) and Response to Immunotherapy in Chinese Patients with Epilepsy and Encephalopathy (RITE2-CHN)—have emerged as tools to predict antibody presence and treatment outcomes. A 2021 study from Xuanwu Hospital, Capital Medical University (Beijing, China) explores how these scores perform for different neuronal surface antibodies, offering actionable insights for clinicians caring for this patient population.

Study Background and Methods

The research team analyzed data from 1365 patients with epileptic seizures as their primary symptom, seen at Xuanwu Hospital between June 2016 and June 2020. They excluded patients with metabolic abnormalities or structural brain lesions (which could explain seizures) and focused on 915 patients with epilepsy of unknown etiology whose serum or cerebrospinal fluid (CSF) was tested for autoimmune antibodies.

Antibodies targeting key neuronal proteins—including N-methyl-D-aspartate receptor (NMDAR), leucine-rich glioma-inactivated protein 1 (LGI1), contactin-associated protein 2 (CASPR2), and others—were measured using a cell-based assay (EUROIMMUN), a gold-standard method for autoimmune encephalitis testing.

All patients were scored using four systems:

APE2: A global score predicting antibody presence (cutoff ≥4).
APE2-CHN: A modified version for Chinese patients, adding cognitive decline, speech impairment, and decreased consciousness (cutoff ≥5).
RITE2: Predicts immunotherapy response (cutoff ≥6).
RITE2-CHN: Modified for Chinese patients (cutoff ≥8).

Treatment response was defined as:

A ≥1 improvement in the modified Rankin Score (mRS) (a measure of disability) at follow-up, or
A >50% reduction in seizure frequency 6 months after immunotherapy (for patients with seizures alone).

Key Findings

Of the 915 patients, 191 (20.9%) tested positive for neuronal surface antibodies:

115 for NMDAR (most common),
47 for LGI1,
8 for CASPR2,
4 for AMPA receptor (AMPAR),
11 for GABAB receptor (GABAB-R),
6 with multiple antibodies.

APE2-CHN Outperforms APE2 for Antibody Prediction

APE2 (cutoff ≥4) correctly identified 74.4% of antibody-positive patients (sensitivity) and ruled out 81.8% of negative cases (specificity). APE2-CHN (cutoff ≥5) improved this to 75.9% sensitivity and 84.5% specificity—a small but meaningful gain for Chinese patients.

Notably, 8 patients had an APE2 score <4 but an APE2-CHN score ≥5—all of these patients had memory decline as their most prominent symptom, a feature APE2-CHN explicitly includes. This shows the score’s ability to catch cases missed by the global APE2 system.

APE2-CHN Works Best for Anti-NMDAR Encephalitis

When broken down by antibody type, APE2-CHN’s performance varied:

Anti-NMDAR: 87.8% sensitivity (best among all antibodies). These patients often have diverse symptoms (psychiatric changes, seizures, movement disorders) that align with APE2-CHN’s criteria.
Anti-LGI1: 53.2% sensitivity (lowest). Anti-LGI1 encephalitis typically causes faciobrachial dystonic seizures (FBDS)—brief, frequent arm/face movements—which APE2-CHN does not prioritize.
Anti-CASPR2: 50% sensitivity. These patients may have Morvan syndrome (peripheral nerve damage, sleep issues), which falls outside the score’s focus on brain-related symptoms.

Most Patients Benefit from Immunotherapy—But Response Depends on Antibody Type

Nearly all (187/191, 97.9%) antibody-positive patients received immunotherapy:

126 used one first-line treatment (glucocorticoids, intravenous immunoglobulin [IVIG], or plasma exchange),
52 used two first-line treatments,
9 used three,
18 needed second-line immunosuppressants (e.g., rituximab).

A total of 142 (74.4%) patients benefited from treatment. Response rates varied by antibody:

Anti-LGI1: 75.4% responded well (highest rate),
Anti-NMDAR: 57.8% responded,
Anti-CASPR2: 87.5% responded (small sample size).

The RITE2-CHN score (cutoff ≥8) was most useful for predicting response in anti-LGI1 encephalitis—85.7% of patients with a score ≥8 improved. However, high APE2-CHN scores (≥5) often correlated with poor response: patients with severe symptoms (e.g., consciousness decline, movement disorders) had higher scores but were harder to treat.

What This Means for Clinicians

APE2-CHN is more relevant for Chinese patients: Its inclusion of cognitive decline, speech issues, and consciousness changes—common in local autoimmune encephalitis cases—makes it a better first-line tool than APE2.
Prioritize APE2-CHN for memory/speech symptoms: The score is especially good at catching anti-NMDAR cases where memory loss or speech problems are key.
RITE2-CHN guides anti-LGI1 treatment: For patients with LGI1 antibodies, a RITE2-CHN score ≥8 signals a high chance of responding to immunotherapy.
High scores can mean worse outcomes: Severe symptoms (e.g., unconsciousness) drive up APE2-CHN scores but also make treatment harder—clinicians should not delay immunotherapy for these patients.

Limitations

Small sample sizes: Only 4 patients had AMPAR antibodies, limiting statistical power for rare cases.
Excluded non-specific antibodies: Patients with non-neuronal surface antibodies (e.g., GAD65) were excluded, so results do not apply to these cases.

Conclusions

For Chinese clinicians, APE2-CHN is a stronger tool than APE2 for identifying patients with anti-NMDAR antibodies—especially those with memory or speech issues. RITE2-CHN helps gauge immunotherapy response for anti-LGI1 encephalitis, the second-most common autoimmune form. However, neither score is perfect:

They struggle with antibodies like CASPR2 that cause peripheral symptoms.
High APE2-CHN scores can indicate severe disease that’s harder to treat.

Ultimately, these scores are best used alongside clinical judgment—helping clinicians prioritize antibody testing and immunotherapy for patients most likely to benefit.

Jia Y, Wang HF, Zhang MY, Wang YP. Antibody prevalence and immunotherapy response in Chinese patients with epilepsy and encephalopathy scores for patients with different neuronal surface antibodies. Chin Med J 2021;134:2985–2991. doi.org/10.1097/CM9.0000000000001701

Was this helpful?

0 / 0