Clinical Use of κ Free Light Chains Index as a Screening Test for Multiple Sclerosis

Luisa Agnello, PhD; Bruna Lo Sasso, PhD; Giuseppe Salemi, MD; Patrizia Altavilla, BS; Emanuela Maria Pappalardo, BS; Rosalia Caldarella, BS; Francesco Meli, BS; Concetta Scazzone, BS; Giulia Bivona, MD; Marcello Ciaccio, MD, PhD

Disclosures

Lab Med. 2020;51(4):402-407. 

In This Article

Results

A total of 56 patients were enrolled. MS was diagnosed in 39 patients: 34 with relapsing remitting MS, 4 with primary progressive MS, and 1 with clinically isolated syndrome. The remained 17 patients were diagnosed as having neurological diseases (NDs), including cognitive decline (n = 8), amyotrophic lateral sclerosis (ALS; n = 4), brain tumor metastasis (n = 1), Wernicke encephalopathy (n = 1), autoimmune encephalitis (n = 2), and intracranial hypotension (n = 1). Basic characteristics of the study population are described in Table 1.

CSF κFLC levels and κFLCi were significantly higher in MS than in ND (P <.001 and P <.001, respectively; Figure 1). In contrast, serum κFLC levels were significantly lower in patients with MS than patients patients with ND (P = .01). A total of 94% of patients with MS and 17% of those with ND tested positive for OCBs (P <.001). A significant difference in κFLCi was detected between patients with OCB negativity and OCB positivity (2.1 [1.6–4.8] vs 32.4 [13.7–62.7]; P <.001]).

Figure 1.

Distribution of cerebrospinal fluid (CSF) kappa free light chain κFLC (A) and kappa free light chain index (κFLCi; B) in patients with multiple sclerosis (MS) and other neurological diseases (NDs).

The diagnostic performance of κFLCi for MS diagnosis was evaluated by performing ROC-curve analysis. The best cutoff for κFLCi was 4.3, with sensitivity of 94.9% and specificity of 76.5%. The area under the curve (AUC) was 0.849 (95% confidence interval [CI], 0.604–0.847; P <.001; Figure 2). To further evaluate the clinical use of κFLCi as a screening test for MS, we chose to test the cutoff of 2.9, which was associated with higher sensitivity (97.4%) in our study, although it has lower specificity (64.7%). To evaluate κFLC analysis as an initial test to identify individuals eligible for OCB detection, we performed a posthoc analysis, testing the algorithm presented in Figure 3. Among all patients, 45 had a κFLCi of 2.9 or greater, with 40 testing OCB positive and 5 testing OCB negative. Overall, 92% of patients with κFLCi of 2.9 or greater and testing OCB positive were diagnosed as having MS. Only 3 patients with ND had κFLCi of 2.9 or greater and had tested OCB positive were diagnosed as having ALS, autoimmune encephalitis, and intracranial hypotension. All patients with a κFLCi of less than 2.9 were diagnosed as having ND, except for 1 patient who received the diagnosis of MS. We were intrigued that this patient also tested OCB negative, which suggests that no intratechal synthesis of FLC was detectable in the CSF. Using this approach, we would have avoided OCB analysis in 20% (11/56) of the patients suspected of having MS.

Figure 2.

Sensitivity, specificity, positive likehood ratio, and negative likehood ratio of kappa free light chain index (κFLCi) for the diagnosis of multiple sclerosis (MS).

Figure 3.

Algorithm for multiple sclerosis (MS) diagnosis based on kappa free light chain (κFLC) index and oligoclonal band (OCB) detection.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....