Evaluation of L-index interference limits on Roche cobas c502 and c702 immunoturbidimetric assays using endogenously lipemic specimens and intralipid spiking

Abstract

ObjectivesSpecimen lipemia is a primary concern with turbidimetric and nephelometric assays due the potential interference caused by light scattering or absorption. The purpose of this study was to evaluate lipemic interference thresholds across seven FDA-cleared assays using patient specimens with varying degrees of endogenous lipemia pre- and post-ultracentrifugation (UC) and with Intralipid spiking. MethodsUsing an IRB-approved protocol, residual human serum specimens (n=42; L-indices, 1-1769; H-index ≤85; I-index ≤2) were obtained. Baseline and post-UC testing was conducted across assays on cobas c502 and c702 instruments (Roche Diagnostics; Indianapolis, IN). Serum indices and triglyceride (TRIG) concentrations were also measured pre- and post-UC. Intralipid spiking studies with human AB serum were also conducted. Lipoprotein subfraction analysis (Lipoprint; Quantimetrix; Redondo Beach, CA) was performed on three additional patient specimens with elevated TRIG post-UC to determine which TRIG-containing lipoprotein fraction(s) remain. ResultsSeveral assays showed L-index thresholds derived from endogenously lipemic specimens that were below previously defined limits from the package inserts (PIs) [new (prior)]: AAT 400 (500); CERU 100 (200); HAPTO 450 (600); TRSF 250 (500). L-index limits derived from Intralipid spiking were generally higher than those listed in PIs. UC did not adversely impact results in non-lipemic or lipemic specimens. UC was effective at clearing lipemic interference, although persistence of residual VLDL was often observed. ConclusionsThis study provides an analysis of L-index thresholds for seven immunoturbidimetric assays. Due to the variety of human lipoproteins, limits defined using endogenously lipemic patient specimens may be different from those derived from spiking studies using Intralipid.