Study Detects Zero Documented Lambert-Eaton Cases in Small Cell Lung Cancer

A retrospective cross-sectional study conducted across a major United States academic health system identified zero documented diagnoses of Lambert-Eaton myasthenic syndrome among 436 patients treated for small cell lung cancer. Published in Frontiers in Neurology by researchers at Georgetown University School of Medicine, the investigation evaluated patient records from May 24, 2019, to May 2025 across MedStar Health, a ten-hospital academic healthcare network. Within this clinical cohort diagnosed during the study period, the observed frequency of documented Lambert-Eaton myasthenic syndrome was 0 out of 436 patients, yielding an estimated 95 percent confidence interval of 0 to 0.84 percent.

Lambert-Eaton myasthenic syndrome is recognized as the paraneoplastic neurological syndrome most frequently associated with small cell lung cancer. However, historical documentation rates display substantial divergence across differing study designs and healthcare settings. Four previously published series in which unselected cohorts of patients with small cell lung cancer underwent dedicated neurological examinations reported Lambert-Eaton myasthenic syndrome prevalences ranging between 1.3 percent and 3.8 percent. In contrast, prior epidemiological analyses relying on administrative healthcare claims data have reported diagnostic frequencies below 0.2 percent, highlighting unresolved questions regarding how frequently the disorder is identified in routine clinical environments.

To evaluate patients treated in ordinary clinical care, the investigators implemented a multi-pronged screening strategy designed to overcome specific medical coding limitations. Because the International Classification of Diseases, Tenth Revision (ICD-10) contains no specific diagnostic code for small cell lung cancer, the research team established two distinct screening pathways. The first pathway, designated Arm 1A, utilized structured ICD-10 diagnosis codes combined with chemotherapy-based query filters. The second pathway, Arm 1B, employed a large language model to identify and classify electronic oncology clinical notes containing the phrase “small cell,” with every positive classification subsequently verified against the full note text. Through these combined queries, 436 patients with small cell lung cancer were identified, including 425 confirmed by note text classification and 11 identified solely through structured coding queries.

In parallel, the researchers identified patients diagnosed with Lambert-Eaton myasthenic syndrome across the health system without date restrictions, querying electronic health records for relevant diagnostic codes, medication orders for amifampridine, and targeted text searches. Each candidate record then underwent manual chart review for clinical adjudication. Across the health system records, the team identified nine patients meeting established diagnostic criteria for Lambert-Eaton myasthenic syndrome, categorized as two definite, two probable, and five possible cases. Among these nine patients, seven had an associated malignancy, including three with non-small cell lung cancer, one with small cell lung cancer, one with lymphoma, one with astrocytoma, and one with breast cancer, and four patients had died. An additional patient with small cell lung cancer and possible Lambert-Eaton myasthenic syndrome was detected solely through the unrestricted query and had died prior to the study period.

The study also assessed how frequently patients with small cell lung cancer received documented neurological evaluations during their initial oncologic care. Among the 425 patients identified through oncology note text classification, only 18 individuals had an index oncology note that recorded a formal neurology encounter or an electrodiagnostic study. This observed rate indicates that documented neurological consultations and diagnostic electrophysiology testing were infrequently noted in routine oncology documentation for this patient population.

Although this finding is consistent with clinical underrecognition, the researchers noted that the retrospective study design cannot definitively separate underrecognition from true disease absence, care received at outside facilities, or incomplete electronic documentation. The authors emphasized that clinicians should consider Lambert-Eaton myasthenic syndrome when patients with small cell lung cancer present with unexplained proximal weakness, hyporeflexia, autonomic symptoms, or bulbar complaints, noting also that the lack of histology-specific coding for small cell lung cancer alongside a discrete code for Lambert-Eaton myasthenic syndrome presents an ongoing barrier to large-scale electronic health record research.

The research was conducted by Chelsea Chang, Laura Hidalgo, and corresponding author Nicholas S. Streicher from the Georgetown University School of Medicine in Washington, D.C., with the findings appearing in the Neuroepidemiology section of Frontiers in Neurology.

Subject of Research: Documented frequency of Lambert-Eaton myasthenic syndrome in small cell lung cancer

Article Title: Documented frequency of Lambert-Eaton myasthenic syndrome among patients with small cell lung cancer in a large United States health system

Article References: Chang, C. W. M., Hidalgo, L., & Streicher, N. S. (2026). Documented frequency of Lambert-Eaton myasthenic syndrome among patients with small cell lung cancer in a large United States health system. Frontiers in Neurology, 17, Article 1873372. https://doi.org/10.3389/fneur.2026.1873372

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1873372

Keywords: Lambert-Eaton myasthenic syndrome, small cell lung cancer, paraneoplastic syndrome, MedStar Health, retrospective cross-sectional study, electronic health records, underrecognition, large language models, ICD-10 coding, amifampridine, electrodiagnostic study, neurology encounter, proximal weakness, hyporeflexia, autonomic symptoms, bulbar complaints, neuroepidemiology