Specificity of serum and urine protein electrophoresis for the diagnosis of monoclonal gammopathies.

Serum protein electrophoresis (SPE) is the hub of laboratory testing for monoclonal gammopathies and is a component of almost all diagnostic panels (1–3 ). The sensitivity of SPE as a screening test, however, is dependent on how SPE abnormalities are defined (4 ). Discrete, quantifiable electrophoretic bands (M-spikes) are almost always monoclonal proteins. Subtle abnormalities such as fuzzy bands, increased or fractions, or decreased fractions may reflect the presence of a monoclonal protein, or they may have other causes. Because these subtle abnormalities can reflect serious diseases, laboratories often reflex samples with SPE abnormalities to immunofixation electrophoresis (IFE) for confirmation, even if IFE was not ordered. To evaluate the specificity of these reflex triggers, we reviewed IFE results for reflexed protein electrophoresis assays. We performed SPE, urine protein electrophoresis (UPE), and urine IFE with agarose gel reagents on the Helena REP system (Helena Laboratories) and performed serum IFE with Sebia Hydragel reagent sets (Sebia). All SPE assay results that were reflexed between October 1 and November 3, 2008, and all UPE results reflexed between November 1 and November 30, 2009, were categorized. The triggers for reflexing SPE results were abnormal discrete bands (M-spikes), fuzzy or questionable fuzzy bands, hypogammaglobulinemia ( 5.5 g/L), increased fractions (between 16 g/L and 19 g/L; fractions 20 g/L were fractionated as M-spikes), and an increased 2 fraction ( 14 g/L). UPE abnormalities were M-spikes, fuzzy bands, or the patient having a previously identified serum monoclonal protein. Of 5992 sera for which SPE (but not IFE) was ordered, 790 samples (13.2%) were reflexed to serum IFE because of an abnormality that had not previously been characterized. IFE identified 43% of the abnormalities as a monoclonal protein (100% of all M-spikes and 28% of the other SPE abnormalities; see Table 1). Among the 206 fuzzy, restricted bands, 112 monoclonal proteins (54%) were detected. Four of the 112 IFE-positive results were monoclonal free light chains. Among the 263 questionable fuzzy bands, 47 (18%) were monoclonal proteins. Of 85 hypogammaglobulinemic bands, 10 (12%) were positive, and 5 were monoclonal free light chains. Of 2826 urine samples for which UPE (but not urine IFE) was ordered, 135 (4.8%) were reflexed to urine IFE, which identified 56% as monoclonal: 89% of M-spikes, 51% of fuzzy bands, and 44% of urine samples that had a known serum monoclonal protein. IFE confirmed all 169 serum M-spikes in this study as monoclonal proteins (100% specificity), and 23 of the 26 urine M-spikes were positive by IFE (89% specificity). Although fibrinogen and hemoglobin bands are not uncommon in serum samples, most laboratories