A-210 Comparison of Two Lipoprotein(a) Methods: Validation, Sample Stability, and Patient Characteristics

Abstract

Abstract Background Lipoprotein(a) [Lp(a)] is now thought of as an independent risk factor for coronary heart disease, stroke, and calcific aortic valve stenosis, and drugs specifically targeting Lp(a) are showing promise in clinical trials. Our laboratory recently validated two commercially available methods and used them to characterize a randomly selected group of samples from an outpatient clinic in the United States. Since the assays have different stated sample stabilities, this was also compared. Methods Both the Roche Lp(a) and the Randox Lp(a) immunoturbidimetric methods were run on a Roche cobas c501 platform across several months. In the U.S., the Roche assay is calibrated in mg/dL, while Randox provides both mg/dL and nmol/L units. Assays were tested for reproducibility, intermediate precision and accuracy. Stability of samples and AMR were established. Additionally, 78 residual samples were randomly selected to assess the percentage with high Lp(a) in our patient population. Results Method validation showed acceptable within-run and between-run precision as both assays met or exceeded manufacturer’s specifications. The stated AMR for the Roche assay is 6.0 – 80.0 mg/dL; linearity was verified through 90.0 mg/dL using the Randox calibrators. For the Randox assay, linearity was verified for the manufacturer’s AMR of 5.2–206.0 nmol/L using estimated values assigned to the Roche calibrators. Both assays allow for x3 dilutions to extend the range. Method comparison of the Randox results (nmol/L) was performed with samples sent to Quest. Deming regression results were acceptable (r = 0.995, slope 1.2, y-intercept—2.3). The stated sample stability from Roche is 8 h at room temperature and 48 h at 2–8oC; however, our results suggest 7 h at room temperature and up to 10 days at 2–8oC. Our study also extended the number of freeze-thaws to two. The Randox package insert gives no information on freeze-thaws or room temperature stability, but states stability at 4oC for 14 days. In our assessment, some samples were not stable past 10 days at 2–8oC. We found samples stable for 2 days at room temperature and up to one month at −20oC. Three freeze-thaws was acceptable. Our patient population included 49 Caucasians, 19 Blacks, 2 Asians, 3 Hispanics, and 5 with race unknown. The median Lp(a) for Blacks was 151.1 nmol/L (65.1 mg/dL) while for Caucasians it was 17.9 nmol/L (14.5 mg/dL) for samples within range. The extended range with dilutions allowed all high values to be quantified; however, 12 were below the AMR for the Roche method and 8 below the AMR for Randox method. Based on the Roche assay, 20 (25.6%) subjects had Lp(a) > 50 mg/dL and would be considered to have increased cardiovascular risk. By the Randox assay, the same 20 had an Lp(a) > 100 nmol/L, while 19 had Lp(a) > 125 nmol/L. Conclusions Both assays met performance goals and showed no significant difference in % subjects with elevated risk for cardiovascular events using typical cut-offs. Our data suggests greater stability of samples at room temperature and additional stability to freeze-thaws when using the Randox assay.