Monoclonal Paraprotein May Interfere with the Roche Direct HDL-C Plus Assay

We report here the observation of a falsely decreased HDL-cholesterol (HDL-C) result obtained with the Roche HDL-C Plus assay (polyethylene glycol-modified enzymes, sulfated α-cyclodextrin, and dextran sulfate; cat no. 1930648) in an asymptomatic 78-year-old woman (1). The total cholesterol of the patient was 3.2 mmol/L, the triglycerides were 1.3 mmol/L, and the HDL-C was suppressed (<0) on a Roche Modular analyzer. The same analysis was repeated on the Johnson & Johnson Vitros (with magnetic separation of particles; cat. no. 1042523) and the Beckman LX-20 (with a polyanion-polymer/detergent reagent; cat. no. 650207) analyzers. The HDL-C concentrations according to these methods were 1.2 and 1.0 mmol/L, respectively. On the Roche analyzer, the absorbance of the blank was higher than the final absorbance result, giving a negative value by subtraction. Lipoprotein electrophoresis revealed a normal α band. We therefore suspected the presence of a paraprotein that would precipitate with the first reagent, possibly related to dextran. Serum total proteins and albumin were 101 and 40 g/L, respectively. Serum protein electrophoresis revealed a monoclonal band in the γ region that was unknown to the treating physician. This monoclonal band was characterized by immunofixation as an IgM-κ band that was estimated at 30 g /L. We selected other sera from patients known to have a monoclonal band and assayed these sera for HDL-C with two different assays: the Roche HDL-C Plus assay and a precipitation method using heparin-manganese (Table 1). HDL-C results of patients with known monoclonal paraproteins. HDL-C results of patients with known monoclonal paraproteins. The interference was identified by visual inspection of the absorbance graphs and confirmed with the precipitation method. We identified other cases of falsely decreased results in some samples, independent of the concentration or type of paraprotein involved (Table 1). The exact incidence of this bias was not known, but it was possibly not uncommon because we found several cases in a small series of patients. Such interference does not always lead to suppression of a HDL-C result because it could simply lower the result. This decrease would go unnoticed with the current Roche application. In our view, some modifications could be introduced to screen for this phenomenon. This interference by monoclonal bands with the HDL-C direct assay is not reported in the 1997 edition of Effects of Disease on Clinical Laboratory Tests by Friedman and Young (2) or in the manufacturer’s package insert, except for the following statement: “In rare cases, increased immunoglobulin concentrations can lead to falsely increased HDL-cholesterol results”. There is no mention of falsely decreased results in patients with paraproteins. Because the prevalence of paraproteins in the population is significant, it is conceivable that some patients with benign paraproteins may have a falsely decreased HDL-C result that could alter their cardiovascular risk estimates and treatment by physicians.