Multi-Center Clinical Validation of a Mass Spectrometry Immunoassay for the Diagnosis and Monitoring of Multiple Myeloma and Associated Disorders

Abstract

Introduction: Mass spectrometry (MS) technology holds great promise for the investigation of monoclonal proteins (M proteins) in peripheral blood. We present results from a multi-center clinical validation study using quantitative immunoprecipitation MS (QIP-MS). QIP-MS combines isotype-specific immunopurification with matrix-assisted laser-desorption ionization MS, and offers automated, sensitive detection, isotyping and quantification of M proteins. It reports the mass/charge ratio (m/z) of the involved light chain, which serves as molecular fingerprint for monitoring the M protein. Methods: The study included 460 diagnosed monoclonal gammopathy (MG) patients (160 multiple myeloma (MM), 112 smoldering MM (SMM), 120 monoclonal gammopathy of undetermined significance (MGUS), 47 Waldenström's Macroglobulinemia (WM), and 21 AL amyloidosis), and 170 disease controls for assessing diagnostic sensitivity and specificity. Sixty-four MM patients with 439 follow-up samples and 10 WM patients with 91 follow-up samples were included to evaluate the ability of QIP-MS to detect M protein changes related to treatment. Median follow up was 19 and 15 months, respectively. Serum samples were retrospectively analyzed at three sites. QIP-MS was carried out using the automated EXENT® solution (in development, The Binding Site, part of Thermo Fisher Scientific). The assay's diagnostic sensitivity and specificity were calculated based on categorizing results as positive or negative. A positive result was defined in baseline samples as the presence of an M protein which was either an intact immunoglobulin ≥0.200 g/L or a light chain only. Results by QIP-MS were also compared to serum protein electrophoresis (SPE) and immunofixation electrophoresis (IFE). QIP-MS response categories were defined based on M protein changes per international guidelines criteria, and compared to response categories assigned by the treating physician. Complete response (CR) was defined as absence of the M-peak that was observed at baseline, using isotype and m/z value of ±4 as criterion for identity. Results: The overall diagnostic sensitivity of QIP-MS in this study was 95.0%: 93.3% for MGUS; 100.0% for SMM; 94.4% for MM; 100.0% for WM; and 71.4% for AL amyloidosis. The diagnostic specificity of the assay was 68.2%. QIP-MS identified an M protein in more MG patients compared to SPE: 437 (95.0%) vs 398 (86.5%). The positivity rate by QIP-MS vs SPE was 93.3% vs 84.2% in MGUS; 100% vs 92.9% in SMM; 94.3% vs 85.0% in MM; 100% vs 100% in WM and 71.4% vs 47.6% in AL amyloidosis. Method comparison demonstrated a Passing-Bablok slope of 0.8 to 1.2 between QIP-MS and SPE for the quantification of M proteins for each disease group and for each isotype, except for monoclonal IgM and in WM (slope of 1.58). In SPE-positive MG patients, the overall concordance between QIP-MS and IFE for M protein isotype was 97%. The overall concordance rate between QIP-MS response categories and standard response assignment was 55% for MM and 56% for WM: 48% for progressive disease (PD); 63% for stable disease (SD); 46% for minimal response (MR); 71% for partial response (PR); 66% for very good partial response (VGPR); and 25% for CR in MM patients. Among 73 responses categorized as CR, QIP-MS produced a positive result for the original clone in 55 (75.3%) cases. Concordance rates in WM patients were 71% for PD%; 30% for SD; 38% for MR; and 74% for PR; no VGPR or CR were reported by either method. Conclusions: In this study, QIP-MS demonstrated the potential for same or superior diagnostic sensitivity compared to SPE, high concordance with IFE for the M protein isotype and good quantitative agreement with SPE measurements of the M proteins. It reported higher IgM values compared to SPE, likely due to reliance on turbidimetric immunoglobulin measurement for quantitations. Diagnostic specificity was impacted by the identification of minor M proteins, not detectable by SPE, and whose clinical significance requires further investigation. QIP-MS demonstrated “moderate” to “fair” agreement for response assignment in MM and WM, respectively, mostly due to the detection of residual M proteins in a significant proportion of patients in CR, in line with its enhanced analytical sensitivity. These data support the use of QIP-MS as an aid in the diagnosis and monitoring of MGs.