[Comparison of Cytomegalovirus Polymerase Chain Reaction Test Results Obtained with Fully Automated Systems: Roche...

Viral load monitoring is important in identifying patients at risk of developing cytomegalovirus (CMV) related complications after transplantation and for this purpose, quantitative real-time polymerase chain reaction (Rt-qPCR) tests are most commonly used. The main problem in CMV DNA Rt-qPCR tests that make quantitative measurements is that there are significant differences in measurements performed with different kits in different laboratories. Comparability of viral load measurements between laboratories has increased with the introduction of quantitative PCR tests calibrated with the CMV International Quantitation Standard (IQS) developed by the World Health Organization (WHO). However, quantitative agreement between measurements made with different kits has still not been fully achieved. In this study, it was aimed to investigate the quantitative compatibility between measurements made with Cobas 6800 (Roche Diagnostics, Mannheim, Germany) and NeuMoDx (Qiagen, Ann Arbor, USA) CMV DNA Rt-qPCR tests, which are fully automated new generation systems calibrated with the WHO CMV IQS. The results of 214 plasma samples, which were studied simultaneously with Cobas 6800 CMV Rt-qPCR and NeuMoDx CMV Rt-qPCR tests were analyzed. In the tests, the extraction, amplification and detection stages were carried out fully automatically. CMV DNA was detected in 144 (67.28%) samples in both tests and was not detected in 53 (24.76%) samples. Incompatible results were obtained in a total of 17 (7.94%) samples. Good agreement was found between the qualitative results of both tests (kappa= 0.80, p< 0.001). When the quantitative results (n= 129) obtained in the dynamic measurement range of both tests were examined, the median viral load values measured by Cobas 6800 CMV Rt-qPCR and NeuMoDx CMV Rt-qPCR tests were 513 IU/mL (range= 35-37000) and 741 IU/mL (range= 68-48978), respectively. According to the correlation analysis, a very strong correlation was found between the results of both tests (r= 0.94, p< 0.001). According to Bland-Altman analysis; the average difference between the results of the NeuMoDx CMV Rt-qPCR test and the Cobas 6800 CMV Rt-qPCR test was found to be -0.14 log10 [standard deviation (SD)= 0.23] IU/mL and it was determined that the Cobas 6800 CMV Rt-qPCR test had lower measurements than the NeuMoDx CMV Rt-qPCR test. In 120 of 129 samples (93%) whose results were within the dynamic measurement range of both tests, the measurement difference was within ± 0.5 log10 IU/mL and in 9 (7%), it was detected as more than ± 0.5 log10 (median 0.54 log10 IU/ml; range= 0.51-0.81). No measurement difference of more than ± 1.0 log10 was detected in any sample. In this study, quantitative agreement was found in the measurements made in plasma samples with the fully automated Cobas 6800 CMV Rt-qPCR and NeuMoDx CMV Rt-qPCR tests calibrated with the CMV IQS. To the best of our knowledge, a study comparing viral load measurements made with Cobas 6800 and NeuMoDx fully automated systems in the detection of CMV DNA has not yet been conducted, and this is the first study on this subject.