Inter-laboratory variability in cytomegalovirus DNA quantification: implications for standardization and clinical monitoring

Abstract

Cytomegalovirus (CMV) infection monitoring is a key element in the management of immunocompromised patients. CMV DNA quantification in plasma or whole blood is the best indicator for clinicians to adjust immunosuppressive or antiviral therapies. Despite the availability of internationally standardized material, the commutability of CMV quantification results across laboratories remains inadequate. To assess inter-laboratory variability in CMV DNA quantification, we conducted a blinded study in seven independent laboratories. Each participant received a panel of 92 specimens for CMV quantification using their routinely used standard platform. While quantifications were highly correlated and reproducible, large discrepancies were observed with differences up to 1.45 log₁₀ IU/mL between techniques for identical specimens. However, quantification scattering was lower for the World Health Organization (WHO) international standard or a commercially tested control (interquartile range = 0.129) than for clinical specimens (0.469; P = 0.0142). Blind quantification of the WHO or the commercial standard indicated that all techniques, except for fully integrated platforms, did not align well with the expected values, and most platforms tended to quantify specimens and standards differently. Recalibration of all platforms against the same standard improved the spread of results, but differences of up to 1.19 log₁₀ IU/mL remained for the same specimens. Achieving commutability in CMV quantification remains an elusive goal. Efforts should focus on improving both the assay calibrators and the run controls, which currently do not appear to simulate the unique characteristics of circulating CMV in patients. Until this is resolved, each transplanted patient should be consistently monitored by the same laboratory on the same platform.IMPORTANCEOur conclusions support previous work on this topic describing the diversity of circulating cytomegalovirus (CMV) DNA forms and the difficulties in standardizing CMV viral load (VL) measurement. The inter-assay reproducibility of CMV VL measurement is primarily influenced by the extraction procedure and the amplicon size generated by the technique. Viral standards generated from cell culture supernatant do not reflect circulating CMV forms from patient samples. We highlight the need to develop a new international standard that better reflects the circulating forms of CMV and demonstrate the risk of tracking a patient's CMV VL in different laboratories.

Keywords: calibration; extraction; fragmentation; immunosuppression; viral load monitoring.

Fig 1

Correlation between the nominal and measured values of the WHO standard when quantified by different systems. Closed squares: Alinity; Closed triangles: Ingenius; Closed circles: Ingenius-Altona; Open triangles: Nuclisens-Rgene; Stars: Qiagen-Artus; Open circles: Cobas; and Open squares: Roche-Altona.

Fig 2

Distribution of VL obtained on clinical specimens. The x-axis represents the mean VL of each clinical specimen as measured by each system. The y-axis represents the overall mean VL obtained from all the systems combined. Closed squares: Alinity; Closed triangles: Ingenius; Closed circles: Ingenius-Altona; Open triangles: Nuclisens-Rgene; Stars: Qiagen-Artus; Open circles: Cobas; and Open squares: Roche-Altona.

Fig 3

Box plot (median, IQR) of the difference between duplicate VL quantification for each system.

Fig 4

Difference to the mean VL for each system according to the type of samples: Clinical specimens (Sple) or WHO/EDX standards (STD). * indicates statistically significant differences.

Fig 5

Difference in standard quantification by each system. Each symbol represents one standard concentration. Closed squares: Alinity; Closed triangles: Ingenius; Closed circles: Ingenius-Altona; Open triangles: Nuclisens-Rgene; Stars: Qiagen-Artus; Open circles: Cobas; and Open squares: Roche-Altona.

Fig 6

Scattering of clinical specimen VL measurements before and after recalibration with the WHO and EDX standards.