Performance evaluation of CXCL10 ELISA “cosmic” kit to measure CXCL10 in cerebrospinal fluid of patients with HTLV-1-associated myelopathy

Abstract

Objectives

This study aimed to validate the clinical utility of cerebrospinal fluid (CSF) CXCL10 measurements in HTLV-1-associated myelopathy (HAM) using a CXCL10 ELISA “Cosmic” kit, a more widely applicable method than cytometric bead array (CBA).

Methods

CSF CXCL10 levels were measured in 165 samples from 111 patients with HAM and 18 controls using a CXCL10 ELISA “Cosmic” kit. We assessed the following: (1) CSF CXCL10 concentrations by HAM activity level (high, moderate, and low) versus controls; (2) correlation with CBA; (3) cutoff values, sensitivity, and specificity for differentiating among HAM activity levels; (4) changes in HAM activity after steroid therapy; and (5) relationship between HAM activity and prognosis in patients undergoing steroid therapy. A correlation coefficient of ≥0.9 with CBA was the primary endpoint.

Results

The median CSF CXCL10 levels in the high, moderate, low, and control groups were 4016.0, 841.0, 112.8, and 102.5 pg/mL, respectively. The ELISA findings were highly correlated with the CBA findings (r = 0.99). Cutoff values were set at 2500 pg/mL (sensitivity, 93.3%; specificity, 100%) to distinguish between high and moderate activity and 180 pg/mL (sensitivity, 81.8%; specificity, 100%) for low to moderate activity comparable to CBA. The new cutoffs enabled the detection of HAM activity changes and prediction of motor disability progression under steroid therapy.

Conclusion

CXCL10 ELISA “Cosmic” kit findings were strongly correlated with CBA findings, meeting the primary endpoint and demonstrating comparable sensitivity and specificity for distinguishing HAM activity. This product shows a promising ability to determine the therapeutic strategy.