Abstract
BACKGROUND: The relapse rate in myasthenia gravis (MG) is high, and promising therapies have emerged; however, identifying potential predictive factors for relapse remains a challenge. This study aimed to explore the association between circulating natural killer (NK) cell levels and the risk of recurrence in MG. METHODS: This retrospective cohort study included 265 patients with MG whose data were included in the Neurology Department of the Affiliated Hospital of Guizhou Medical University database between March 2015 and March 2022. Data from electronic medical records were collected, which included the patients' circulating NK cell count (exposure variable) and demographic/clinical characteristics (covariates). The primary outcome was the 1-year MG recurrence rate. RESULTS: The study revealed a non-linear relationship between peripheral NK cell count and MG recurrence, with an inflection point at 5.38. Below this threshold, the risk of recurrence was low with higher NK cell counts (relative risk (RR): 0.23, 95% confidence interval (CI) [0.11-0.490]); above this threshold, no significant association was observed (RR: 1.43, 95% CI [0.62-3.34]). Furthermore, the NK cell proportion showed no significant linear or non-linear association with MG recurrence risk (RR: 0.84, 95% CI [0.57-1.2]). CONCLUSION: This study provides epidemiological evidence of a potential association between peripheral NK cell count and MG recurrence risk, suggesting an immunoregulatory protective effect within a specific NK cell count range. These findings may inform more personalized MG treatment strategies, warranting further validation in larger and more diverse cohorts.