Diagnostic implications of enumerating and reporting beta fraction(s) for the detection of beta-migrating monoclonal immunoglobulins in serum protein electrophoresis

Abstract

BackgroundModern serum protein electrophoresis (SPE) resolves serum proteins in 6 major fractions, splitting beta (B) into beta1 (B1) and beta2 (B2). Beta-migrating monoclonal immunoglobulins (B-MC) commonly integrate into normal-looking peak shape(s) but may increase the fraction value, forming the basis for reflex testing. The objectives of this study were (1) to ascertain the value of beta fraction(s) reporting, and (2) to compare the diagnostic performance between different beta-flagging approaches, particularly ↑B versus ↑B1 and/or ↑B2. MethodsWe retrospectively studied 22,900 consecutive SPEs, and identified 3974 paired SPE and immunofixation electrophoresis (IFE) results obtained from the Sebia Capillarys™ 2 and Hydrasys™ electrophoresis systems respectively. ↑B, ↑B1 and ↑B2 were defined as fraction concentrations >11, 6 and 5 g/L respectively, and their respective diagnostic metrics calculated using IFE as the reference standard. Results32 beta-gamma bridges were B-MC negative and thus excluded. Of 3942 cases, 142, 18, 285 and 300 had ↑B, ↑B1, ↑B2 and ↑B1 and/or ↑B2 respectively, while their corresponding sensitivities for B-MC were 0.38, 0.09, 0.45 and 0.54 respectively. Comparing ↑B and ↑B1 and/or ↑B2, ↑B had significantly lower sensitivity but higher specificity (0.98 Vs 0.95) and positive predictive value (0.47 Vs 0.31). All 4 beta fraction increases had odds ratios ranged from 14 to 118. ConclusionBeta increases were associated with increased odds for B-MC, hence providing value and justification for their reporting and reflex testing. ↑B1 and/or ↑B2 detected more B-MC than ↑B, supporting separate reporting of B1 and B2.