Kappa Index Versus CSF Oligoclonal Bands in Predicting Multiple Sclerosis and Infectious/Inflammatory CNS Disorders

Abstract

Background: Cerebrospinal fluid (CSF) kappa free light chains (KFLC) are gaining increasing interest as markers of intrathecal immunoglobulin synthesis. The main aim of this study was to assess the diagnostic accuracy (AUC) of the kappa index (CSF/serum KFLC divided by the CSF/serum albumin ratio) compared to CSF oligoclonal IgG bands (OCB) in predicting Multiple Sclerosis (MS) or a central nervous system infectious/inflammatory disorder (CNSID).

Methods: We enrolled patients who underwent a diagnostic spinal tap throughout two years. KFLC levels were determined using a Freelite assay (Binding Site) and the turbidimetric Optilite analyzer.

Results: Of 540 included patients, 223 had a CNSID, and 84 had MS. The kappa index was more sensitive (0.89 versus 0.85) and less specific (0.84 versus 0.89), with the same AUC (0.87) as OCB for MS diagnosis (optimal cut-off: 6.2). Adding patients with a single CSF IgG band to the OCB-positive group slightly increased the AUC (0.88). Likewise, the kappa index (cut-off: 3.9) was more sensitive (0.67 versus 0.50) and less specific (0.81 versus 0.97), with the same AUC (0.74) as OCB, for a CNSID diagnosis.

Conclusion: The kappa index and CSF OCB have comparable diagnostic accuracies for a MS or CNSID diagnosis and supply the clinician with useful, complementary information.

Keywords: cerebrospinal fluid; free light chains; kappa index; multiple sclerosis; oligoclonal bands.

Figure 1

Box-plots showing median kappa index values and interquartile ranges (IQR) amongst different diagnostic subgroups (A), and amongst different sub-groups on a logarithmic scale (B). Upper whiskers show the largest observation ≤ (third quartile + 1.5 × IQR); lower whiskers show the smallest value ≥ (lower quartile + 1.5 × IQR); black dots show outliers. MS: Multiple Sclerosis; CIS: Clinically Isolated Syndrome, DEM: acquired demyelinating central nervous system (CNS) disorder; INF: CNS infectious disorder; EA: autoimmune/paraneoplastic encephalitis; EPI: epilepsy; PNS: peripheral nervous system disorder; NEO: CNS neoplasm; VASC: vascular disorder; DEG: degenerative neurological disorder; MISC: miscellaneous diagnoses. CNSID: central nervous system inflammatory/infectious disorders; NID: non-inflammatory disorders.

Figure 2

KFLC Reibergram in MS patients. KFLC Reibergram showing MS patients’ CSF/serum KFLC quotients (Q kappa) in relation to their CSF/serum albumin quotients (Qalb). KFLC intrathecal synthesis can be assumed when Q kappa is above the upper line, depicting the hyperbolic border line Q kappa (lim). The dashed line shows the Q kappa mean, and the lower line, the lower limit of the reference range (Q kappa low). The graph was created using the free software available at available at: Albaum. Available online: www.albaum.it (accessed on 21 October 2020).

Figure 3

KFLC Reibergrams of CNSID and NID patients. CNSID patients′ data (A) and NID patients′ data (B) on KFLC Reibergrams showing the CSF/serum KFLC quotients (Q kappa) in relation to the CSF/serum albumin quotients (Qalb). KFLC intrathecal synthesis can be assumed when Q kappa is above the upper line, depicting the hyperbolic border line Q kappa (lim). The dashed line shows the Q kappa mean, and the lower line, the lower limit of the reference range (Q kappa low). Graphs were created using the free software available at Albaum. Available online: www.albaum.it (accessed on 21 October 2020).