Paraneoplastic or not? Sirtuin 2 in anti-N-methyl-D-aspartate receptor encephalitis

Background and purpose: N-methyl- d- aspartate receptor (NMDAR) and leucine-rich glioma-inactivated protein 1 (LGI1) encephalitis are important types of autoimmune encephalitis (AE) with significant morbidity. In this study, we used a proteomic approach in search of novel clinically relevant biomarkers in these types of encephalitides. Methods: Swedish and Czech tertiary neuroimmunology centers collaborated in this retrospective exploratory study. Fifty-eight cerebrospinal fluid (CSF) samples of 28 patients with AE (14 definite NMDAR, 14 with definite LGI1 encephalitis) and 30 controls were included. CSF samples were analyzed using proximity extension assay technology (Olink Target 96 Inflammation panel). For each CSF sample, 92 proteins were measured. Clinical variables were retrospectively collected, and correlations with protein levels were statistically analyzed. Results: Patients and controls differed significantly in the following 18 biomarkers: TNFRSF9, TNFRSF12, TNFRSF14, TNF beta, TNF alpha, IL7, IL10, IL12B, IFN gamma, CD5, CD6, CASP8, MMP1, CXCL8, CXCL10, CXCL11, IL20RA, and sirtuin 2 (SIRT2). In LGI1 encephalitis, no clinically useful association was found between biomarkers and clinical variables. In the NMDAR encephalitis group, SIRT2, TNF beta, and CD5 were significantly associated with ovarian teratoma. For SIRT2, this was true even for the first patients' CSF sample (SIRT2 without vs. with tumor, mean +/- SD = 2.2 +/- 0.29 vs. 2.88 +/- 0.48; p = 0.007, 95% confidence interval = -1.15 to -0.22; r statistic in point-biserial correlation (rpb) = 0.66, p = 0.011). SIRT2 was positively correlated with age (rpb = 0.39, p = 0.018) and total hospital days (r = 0.55, p = <0.001). Conclusions: SIRT2 should be investigated as a biomarker of paraneoplastic etiology in NMDAR encephalitis.