Imprecision of tumour biomarker measurements on Roche Modular E170 platform fulfills desirable goals derived from biological variation

Background: Monitoring of test imprecision is one of the most important quality indicators in clinical laboratories. Imprecision goals should be derived from biological variation. The aim of this study was to evaluate the imprecision of eight tumour biomarker assays routinely measured on the Modular E170 system. Methods: Method coefficient of variations (CVs) were obtained by an appropriate Internal Quality Control programme based on the measurement every working day of a fresh-frozen human serum pool with biomarkers concentrations around the clinical cut-offs. We evaluated data collected along the whole year 2008 (n range: 21-461); monthly CVs and their cumulative means were calculated and compared with corresponding goals. Results: Biomarkers concentration means and average yearly CVs (desirable goals in parentheses) were as follows: α-fetoprotein, 9.6 μg/L, 3.9% (6.0%); CA125, 41.2 U/L, 2.8% (12.4%); CA15.3, 32.7 U/L, 3.1% (3.1%); CA19.9, 35.1 U/L, 2.8% (8.0%); CEA, 7.7 μg/L, 4.3% (6.4%); prostate-specific antigen (PSA), 4.1 μg/L, 4.3% (9.1%); CYFRA 21.1, 2.4 μg/L, 5.7% (11.3%); and ferritin, 64.5 μg/L, 4.0% (7.1%). Conclusions: Our study shows that in routine laboratory practice and over a clinically and analytically relevant time-span, the imprecision of the tumour biomarker measurements on the Roche Modular E170 fulfills desirable goals. For four assays (CA125, CA19.9, PSA and CYFRA 21.1) the optimum CV can even be achieved.