LAUSR

The analysis of serum for the presence of free light chains has become an important adjunct to testing by serum protein electrophoresis (SPE) and immunofixation electrophoresis (IFE) [1]. The first commercial free light chain assays using polyclonal reagents was released in 2001 by Binding Site Inc. (FreeliteTM). Subsequently, in 2011, Siemens released monoclonal-based free light chain assays. The interest in free light chain testing has increased with a more than doubling of the subscription rates to the College of American Proficiency Testing Survey from 2010. We validated the Diazyme free light chain assay, a latex particle enhanced immunoturbidimetric assay that uses polyclonal antibodies and was adapted to the Siemens Advia 1800 analyzer. As this assay was not FDA approved on this analyzer, we considered it as a “lab developed test.” We conducted a clinical concordance study compared to the Freelite assay as the predicate. A total of 222 samples were obtained from 175 different patients submitted for routine SPE and IFE (agarose gel-based from Sebia), and free light chain analysis as part of a routine workup for multiple myeloma. There were 131 patients with no myeloma (including those with monoclonal gammopathy of unknown significance, MGUS) and 42 patients exhibiting a monoclonal band, including 4 with light chains disease, and 2 with amyloidosis. The majority of the myeloma patients had IgG-kappa (25 patients) with at least one of each of the other 5 major subtypes (no IgD or IgE myelomas). This study was conducted as part of a quality assurance and validation study for potential implementation of this assay in our hospital laboratory. Samples were permanently deidentified once the required medical information was extracted. As this was not a research study, institutional review board approval was deemed to be unnecessary. The recommended reference range for the Diazyme assay is 2.37–20.73mg/L for free kappa (κ) and 4.23–27.69mg/L for free lambda (λ), and 3.30–19.40 and 5.71–26.30mg/L for the Binding Site free κ and free λ assays, respectively. The reference range for κ /λ for the Diazyme assays is 0.22–1.74, and 0.26–1.65 for the Binding Site Assay. The precision of the Diazyme assay for 27 days for two levels of controls (nominal values 1.50 and 3.21mg/L for κ and 2.49 and 4.93mg/L for λ) was 6.4% and 3.2% for κ free light chains (level 1 and 2, respectively) and 3.2% and 1.6% for free λ light chains, respectively. The correlation between the two assays for free κ and λ was produced a linear correlation equation of y = 1.03 × −18.9mg/L, r = 0.95 and y = 1.06 × +4.1mg/L r = 0.97, respectively (the log–log scale representation is shown in Fig. 1A and B, Medcalc Inc.). Five points were removed from the λ correlation because they were below the Diazyme assay's limit of quantitation of 4.23mg/L. The regression for the kappa to lambda ratio was y = 0.32 × +1.90, r = 0.97 (Fig. 1C). The few significant outliers that were observed (e.g., points A–E, Fig. 1A–C) were all outside the reference range for these analytes (none were removed from the regression equations). A review of the cases showed that none of these would have resulted in a change the diagnosis of myeloma, which at our institution, is largely made on the basis of the clinical presentation, the immunofixation electrophoresis finding, conducted on all myeloma and MGUS patients, and bone marrow biopsies where appropriate. Serum free light chains are usually used to monitor disease progress and remission. Based on manufacturer's claims, the Diazyme assay had a wider analytical measurement range for free κ and free λ (150 and 200 mg/L, respectively than the Binding Site assay (56.2 and 74.8mg/L respectively) resulting in fewer numbers of samples that required dilution after the initial test (9.5% and 1.4% for Diazyme, respectively versus 34% and 18% for Binding Site, respectively). Using the recommended reference intervals for the two kits, the degree of concordance was 99% for free κ, with both discordant results being near the cutoff concentration (i.e., within 10% of the cutoff limit to account for assay imprecision). The concordance for