1426Correlation of Cytomegalovirus Viral Load between a Laboratory-Developed Test and a WHO-Calibrated Commercial Assay in Transplant Recipients
Year: 2014
Type: conference-abstract
Abstract: Background. Quantitative CMV nucleic acid testing (QNAT) is the standard laboratory method for the diagnosis and treatment guidance of CMV infection in transplant recipients. Until recently, QNAT using laboratory developed tests (LDT) have not been standardized and results are not directly comparable. Recently, the first WHO International Standard for CMV QNAT (WHO-QNAT) has been released, and clinical laboratories are now calibrating their QNAT to this standard. Our objective is to correlate the viral load measurements between an FDA-approved WHO-QNAT assay and a LDT used by three different clinical laboratory sites. Methods. Plasma samples were quantified for CMV viral load by LDT and WHO-QNAT at three different clinical laboratories of a single institution during the summer of 2013. Spearman's correlation coefficients (ρ) and Lin's concordance correlation coefficients (CCC) were used to compare the viral load levels obtained with WHO-QNAT and LDT. LDT results were transformed into binary result categories based on a pre-defined threshold of 5,000 copies/mL (trigger for antiviral therapy), and WHO-QNAT levels were evaluated using logistic regression and ROC analysis, from which optimal cut-points and sensitivity/specificity (sens/spec) values were obtained. Results. A total of 431 clinical plasma samples were tested by both LDT and WHO-QNAT assays: 123 samples at Lab A, 202 at Lab B, and 106 at Lab C. Combining all data, the optimal viral load equivalent of the pre-defined threshold of 5,000 copies/mL was 1,200 IU/mL (based on WHO-QNAT), with sens/spec of 82.9%/90.2%. For individual sites, the pre-defined threshold was equivalent to 1,810 IU/mL at Lab A (sens/spec 100%/92%), 1,200 IU/mL at Lab B (sens/spec 82%/88%), and 326 IU/mL at Lab C (sens/spec 90%/82%). For Lab A values, the correlation between assays was modest (ρ= 0.777) and concordance was poor (CCC = 0.304). Similar results were observed for Lab B (ρ= 0.793, c = 0.330) and Lab C (ρ= 0.758, CCC = 0.248). Conclusion. Despite using a similar LDT, viral load equivalence to the FDA-approved assay varies among clinical laboratories, possibly due to differences in patient population and other aspects of viral load testing. Understanding factors that contribute to these variations will optimize the ongoing efforts of standardizing testing for CMV viral load. Disclosures. All authors: No reported disclosures.
Source: Open Forum Infectious Diseases
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $4,053
OpenAlex ID: W2301378768