Evaluation of the two different real time polymerase chain reaction methods used for bk virus (bkv) quantification and bkv genotype assignment

Abstract

BK virus (BKV) viral load quantification has a distinct role in the clinical control of BKV nephropathy and organ rejection among renal transplant recipients. In this study, it was aimed to compare BKV DNA measurement values performed with two different real-time polymerase chain reaction (PCR) methods and to determine BKV genotypes in renal transplant recipients. Totally, 150 clinical samples tested previously in two different laboratories (Lab-1 and Lab-2) from adult and pediatric renal transplantation patients were included in the study. Fifty plasma samples of 50 different patients from Lab-1, 50 plasma and 50 urine samples of 58 different patients from Lab-2 were included in the study. Viral nucleic acid extraction was performed with automatized systems in Lab-1 and Lab-2 (EZ1, Qiagen, Germany and MagNA Pure 96, Roche Diagnostics, Germany; respectively;). Real-time PCR procedure was carried out in Lab-1 with an amplification mixture of primer, probe sequences targeting VP-1 gene region using RotorGene (Qiagen, Germany) and in Lab-2 with an amplification mixture of primer, probe sequences targeting VP-2 gene region using ABI Prism 7500 (Applied Biosystems, USA). BKV genotyping was performed with multiplex PCR using primer, probe sequences for BKV genotypes I-IV. In both of the laboratories, 82 (54.6%) of the samples were found as positive, 37(24.6%) samples were found as negative and a moderate agreement was found between qualitative results of two real-time PCR methods (k= 0.56, p< 0.001). Median viral load values were 4.1 x 10(4) copies/ml (321-6 x 10(9)) in Lab-1 and 3.3 x 10(5) copies/ml (224-8.3 x 10(10)) in Lab-2 for positive samples. According to the lineer regression analysis of quantitative results, moderate (R-2 = 0.52, p< 0.001) and high (R-2 = 0.88, p< 0.001) correlation was found for plasma (n= 52) and urine (n= 30) samples, respectively. Bland-Altman analysis yielded a mean difference of -0.58 log(10) for all samples. For plasma samples mean difference was -0.29 log(10), while it was -1.1 log(10 )for urine samples. In all samples, Lab-1 measurements were lower than Lab-2 measurements. A mean difference of -1.1 log(10) indicated that the measurement values of Lab-2 were more higher than Lab-1 measurments with an average of 1.1 log(10). Supporting this result, 71.9% of the samples had a measurement difference more than 0.5 log 10 and 29.2% of the samples had a measurement difference more than 1 log(10). Only 28.1% of the samples were measured within clinically acceptable log difference range (less than 0.5 log(10)).BKV genotyping was performed only for 74 different patient samples with sufficient copy numbers and genotype I (81.7%), IV (15.5%), II (1.4%), I+IV (1.4%) were detected. When the results were compared; 66.6% (n= 12) of the genotype IV samples had more than 1 log(10) and 83.3% of them had more than 0.5 log(10) viral load measurement difference. Correlation and linear regression analyzes were insufficient for the comparison of the results of the two different tests. It will be appropriate for each center to monitor patients with the same test until the international BKV standard developed by the World Health Organization is optimized. The clinical correlation of the tests is limited to the currently used test. The result of incorrect BKV quantification affects the clinical decision. Measurements less than the actual value will lead to the development of BKV nephropathy, and higher measurements will lead to unnecessary allograft biopsy and unnecessary reduction of immunosuppression.