Diagnostic approach for multiple sclerosis: optimizing algorithms for intrathecal synthesis of immunoglobulins.

Background: The kappa-free light chain (κFLC) index has shown its value in detecting the intrathecal synthesis of immunoglobulins. We aimed to evaluate the diagnostic performance of the κFLC index for multiple sclerosis (MS) and compare different algorithms proposed in the literature to optimize its use for our population.

Methods: Based on the results of the oligoclonal bands (OCBs) and κFLC index of 255 patients with suspected MS different optimization strategies were evaluated, for which the optimal κFLC index cut-off thresholds were calculated.

Results: The best diagnostic performance was achieved by using a reflexive algorithm, in which OCBs are only performed according to the κFLC index result. With a single cut-off (κFLC index = 7.9), an accuracy of 92.2% was obtained (sensitivity = 92.4%, specificity = 92%) with an OCB performance rate of 58.1%. When applying 2 cut-offs (κFLC index = 4.2 and 13), the accuracy was the same (92.2%, sensitivity = 89.6%, specificity = 94%), but the OCB performance rate dropped to 29.4%.

Conclusion: The 2-step strategy proposed with κFLC determination followed by OCB analysis in the borderline cases appears to be the most suitable solution, further optimized by adjusting the decision thresholds to 4.2 < κFLC index < 13, resulting in high accuracy and the most saving of OCBs.