Comparison of Molecular Tests for Detection and Quantification of Cell-Associated Cytomegalovirus DNA
Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, Georgia,1 Department of Clinical Pathology, Cleveland Clinic Foundation, Cleveland, Ohio,2 Department of Immunology/Microbiology, Statistics and Data Analysis Center, Harvard School of Public Health,3 Division of Infectious Diseases, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts,4 Department of Virology, Eykman Winkler Institute for Clinical Microbiology, Utrecht University Hospital, Utrecht, The Netherlands,5 Department of Pediatric Infectious Diseases, University of California at San Diego, San Diego, California,6 Department of Immunology/Microbiology, Rush-Presbyterian-St. Luke's Medical Center, Chicago, Illinois7f#rh, 百拇医药
Received 28 February 2003/ Returned for modification 25 April 2003/ Accepted 28 May 2003f#rh, 百拇医药
ABSTRACTf#rh, 百拇医药
A cell-based standard was developed to compare the COBAS Amplicor CMV Monitor test, the Hybrid Capture System CMV DNA test, and the NucliSens CMV test. The standard was prepared by infecting human foreskin fibroblasts (HFFs) with cytomegalovirus (CMV) strain AD169 at low multiplicity of infection (0.03) and harvesting the cells at 6 h postinfection. Buffy coat cells were added to produce concentrations of from 0 to 105 HFFs per 106 total cells. Five laboratories performed the Amplicor PCR test and two laboratories performed the NucliSens and Hybrid Capture tests. The Amplicor PCR test was 1.5 to 2.0 log10 more sensitive than the Hybrid Capture test. The specificities of the Amplicor PCR and Hybrid Capture tests were 100 and 93.8%, respectively. The linear range of the Amplicor PCR and Hybrid Capture tests were 2 to 4.48 log10 and 3.48 to at least 5.0 log10 CMV target cells, respectively. The standard deviations of the Amplicor PCR and Hybrid Capture tests varied throughout their linear range, and for both tests the variability was greater for lower concentrations of input CMV DNA. These data allow the direct comparison of viral load values between the Amplicor and Hybrid Capture tests. The analytical sensitivity of the NucliSens test could not be determined by using the 6-h standard, because the low multiplicity of infection and short culture time did not allow for adequate transcription of pp67 late mRNA measured in the test. Extending the incubation time of the standard to 24 h increased the analytical sensitivity of the NucliSens test to 3.0 log10 target cells.
INTRODUCTIONp9., 百拇医药
Cytomegalovirus (CMV) is an important pathogen in immunosuppressed patients, including persons with AIDS and organ transplant recipients. The clinical utility of quantitative CMV DNA tests in managing these patient groups has been well documented (1, 4, 6, 12-14). In transplant recipients, CMV load is used to diagnose active CMV disease, screen patients for the use of preemptive therapy, and monitor the response to antiviral therapy (1, 4, 6, 11). In patients infected with human immunodeficiency virus type 1, the risk of developing CMV disease has been reported to be directly related to the quantity of CMV DNA in plasma (13, 14).p9., 百拇医药
There are several tests available for detecting and quantifying CMV DNA from clinical samples. The COBAS Amplicor CMV Monitor test (Roche Diagnostics, Indianapolis, Ind.) is a DNA PCR test that amplifies a 365-bp region of the CMV polymerase gene. Using a plasma-based standard, the linear range of the test is reported to be between 400 and 50,000 copies/ml of plasma (3). The test has been designed for plasma but can be modified for use with leukocytes or whole blood specimens. The Hybrid Capture System CMV DNA test (Digene Corp., Gaithersburg, Md.) is a signal amplification method that uses an RNA probe which targets 17% of the CMV genome. The target is detected by using antibodies that specifically bind RNA-DNA hybrids. The test is designed for whole blood specimens with a linear range of 1,400 to 560,000 copies/ml (17). The availability of commercial tests has been helpful in standardizing test performance, although neither the Amplicor PCR nor the quantitative Hybrid Capture tests have been cleared by the U.S. Food and Drug Administration (FDA; a qualitative version of the Hybrid Capture test has been cleared by the FDA). The NucliSens CMV test (bioMérieux, Durham, N.C. [FDA cleared]) is a qualitative RNA amplification test that detects pp67 mRNA, which is transcribed late in the viral replicative cycle. Since the NucliSens test is isothermal, it can detect pp67 RNA without interference of CMV DNA. Patients with active CMV disease express high levels of pp67 (7), which are not expressed during latent infection. In addition to these commercially available tests, "laboratory-developed" DNA PCR tests are widely used in clinical laboratories. The performance of laboratory-developed tests varies due to differences in specimen type, nucleic acid extraction method, primers, quantitation standard, and detection method (16).
There have been a few studies comparing the Amplicor PCR test to laboratory-developed PCR tests. Comparison of a LightCycler PCR test to the Amplicor PCR test showed that the Amplicor PCR test yielded viral load results that were 1 log10 higher than those of the LightCycler test (9). Comparison of the Amplicor PCR test with a different laboratory-developed test showed the viral load values to be 0.5 to 0.7 log10 higher with the laboratory-developed test than with the Amplicor PCR test (3). These results underscore the variation in viral load measurements that are obtained by using different quantitative tests. To address this problem, we developed a standard reagent to compare the performance of the Amplicor PCR test, the Hybrid Capture test, and the NucliSens CMV test. We chose a cell-based standard for two reasons: first, CMV is highly cell associated (5, 10) and, second, an infected-cell standard is the only type that would allow the comparison of all three tests.m, http://www.100md.com
MATERIALS AND METHODS9, 百拇医药
Cell-based standard. Human foreskin fibroblasts (HFFs) were inoculated with the CMV laboratory strain AD169 at a multiplicity of infection (MOI) of 0.03. After a 4-h adsorption period, the inoculum was removed. The monolayer was washed withphosphate-buffered saline, and fresh medium was added for an additional 2-h incubation. The 6-h total incubation time was designed to allow the virus to enter the cells but was short enough to avoid intracellular replication of CMV. At this low MOI only 3 per 100 target cells would be expected to be infected with CMV. At 6 h postinfection the HFFs were harvested by trypsinization, washed, and counted. The HFFs were diluted with uninfected buffy coat cells to concentrations of 0 to 105 HFF per 106 total cells. Then, 200-µl aliquots containing 106 total cells were stored at -70°C until they were tested in the Amplicor test. The aliquots (200 µl, 106 total cells) for the Hybrid Capture test were centrifuged (14,000 x g), the supernatant was removed, and the pellets were stored at -70°C until testing was performed. For the evaluation of the NucliSens test a second 24-h standard was designed with a 4-h adsorption period, followed by a 20-h incubation period. Buffy coat cells were added as described above for the other tests. For the NucliSens test, 200-µl aliquots of the standard were added to 0.9 ml of NucliSens lysis buffer; the tubes were then rocked for 30 min before storage at -70°C.
Nucleic acid testing. Viral load testing was done by using the COBAS Amplicor CMV Monitor Test (Roche Diagnostics) and the Hybrid Capture System CMV DNA test (version 2.0; Digene Corp.). For the Amplicor test, 200-µl aliquots of the cell suspensions were added to lysis buffer and tested according to the manufacturer's protocol. For the Hybrid Capture test, the pellets from the 200-µl aliquots were resuspended in lysis buffer and tested according to the manufacturer's protocol. For the NucliSens test, lysis buffer tubes containing the 200-µl samples were tested according to the manufacturer's protocol. The results of the NucliSens test were reported as positive or negative.g!x, 百拇医药
Five laboratories performed the Amplicor test, with each laboratory testing eight replicates of 10 concentrations of the standard containing a range of target cells from 3 to 105. The Hybrid Capture test was done in two laboratories, each of which tested two different panels of the standard. The panel consisted of eight replicates of seven concentrations of the standard, with a range of 100 to 105 target cells. NucliSens testing was performed in two laboratories, with each laboratory testing a panel that consisted of eight replicates of the same target cell concentrations used for the Amplicor test.
Statistical analysis. The results of the quantitative tests were log10 transformed for analysis. The linear range was examined by comparison to a 45° line. Crude within-laboratory standard deviations were tools for detecting departure from the linear range. One of the Amplicor panels contained four missing results; these observations were excluded from the analysis. Methods accommodating censored values were used (8) when we calculated the mean copy number for the Hybrid Capture test at an input concentration of 1,000 target cells because some replicates had viral load levels below the limit of detection. For the NucliSens test, one specimen produced an invalid test result and was positive on repeat testing, this specimen was analyzed as a positive result. Another specimen had two consecutive invalid results as determined by using the NucliSens test; this specimen was excluded from analysis.]'8du.{, http://www.100md.com
RESULTS]'8du.{, http://www.100md.com
The Amplicor test was able to detect 50% of replicates with 1.48 log10 (i.e., 30) CMV target cells and reached a sensitivity of 100% when the input concentration was 2.0 log10 (i.e., 100) target cells (Table 1). For the Hybrid Capture test, at least 50% of the replicates were detected when the input CMV concentration was 3.0 log10 (i.e., 1,000) target cells, 97% of replicates were detected at 3.48 log10 (i.e., 3,000) target cells, and the sensitivity was 100% at an input concentration of 4.0 log10 (i.e., 10,000) target cells. The specificity of the Amplicor test was 100%, whereas the specificity of the Hybrid Capture test was 93.8% (Table 1).
fig.ommitted-4, 百拇医药
TABLE 1. Results of Amplicor and Hybrid Capture testing-4, 百拇医药
The means and standard deviations of the viral load values from all laboratories are shown in Table 2 by input CMV target cell concentration. One replicate with an input concentration of 5.0 log10 CMV target cells was found to contain no virus when tested in the Amplicor PCR test. When this outlier was included in the calculation of the mean and standard deviation (a value of 2.3 log10 [i.e., 200] was used), the mean viral load was 4.496 log10 copies/106 cells with a standard deviation of 0.37 log10 copies/106 cells. When this value was excluded from the calculation, the mean was 4.55 log10 copies/106, and the standard deviation was 0.11 log10 copies/106 cells. We felt that the outlier with no viral load detected was due to a technical error either in preparing the standard or in performing the viral load testing. Further discussions of the test performance refer to the values obtained with the outlier excluded, since the underlying results of the study are unchanged by the exclusion.
fig.ommitted4;, 百拇医药
TABLE 2. Mean results of the Amplicor and Hybrid Capture testing4;, 百拇医药
The linear range of the Amplicor test was 2.0 to 4.48 log10 (i.e., 100 to 30,000) CMV target cells, whereas the Hybrid Capture test was linear, with results from 3.48 to at least 5.0 log10 (i.e., 3,000 to at least 100,000) CMV target cells (Fig. 1). We did not test a high enough concentration of target cells to determine the upper limit of the linear range of the Hybrid Capture test. For concentrations of CMV target DNA that were within the linear range of both tests (3.48, 4.0, and 4.48 log10), the viral load values were within twofold.4;, 百拇医药
fig.ommitted4;, 百拇医药
FIG. 1. Plot of the mean log10 CMV DNA copy number versus the input concentration of the quantitation standard for the Amplicor PCR test (A) and Hybrid Capture test (B). The open circle in panel A represents the mean with the outlier excluded.4;, 百拇医药
The standard deviations of the tests varied throughout their linear ranges. The standard deviation of the Amplicor PCR test ranged from 0.10 to 0.37 log10 CMV copies/106 cells. When the outlier at 5.0 log10 target cells was excluded, the standard deviation ranged from 0.10 to 0.28 log10 CMV copies/106 cells (Fig. 2). The standard deviation was higher at lower input concentrations. Similarly, for the Hybrid Capture tests, more variability was observed for the lower concentrations of CMV DNA compared to the higher concentration, with standard deviations ranging from 0.17 to 0.65 log10 CMV copies/106 cells (Fig. 2). A standard deviation of 0.15 log10 is used by the Virology Quality Assessment Laboratory program to ensure that a laboratory could maintain the precision required to have 90% power to detect a fivefold (0.69 log10) difference in copy number between two samples in the same batch (18). Based on these data, overall changes in viral load measured in the Amplicor and Hybrid Capture tests need to meet or exceed fivefold (0.69 log10 copies) to be interpreted as clinically significant, particularly at the lower end of these tests.
fig.ommitted#(rjr, http://www.100md.com
FIG. 2. Plot of the standard deviations of the log10 CMV DNA copy number versus the input concentration of the quantitation standard for the Amplicor test (A) and Hybrid Capture test (B). The value indicated by the asterisk represents the standard deviation with the outlier excluded.#(rjr, http://www.100md.com
Qualitative results obtained with the NucliSens test are shown in Table 3. At input concentrations of between 3.0 and 3.5 log10, 50% of the replicates could be detected, with the analytical sensitivity reaching 100% at 4.5 log10 CMV target cells. This low level of sensitivity is probably due to the manner in which the cell standard was designed. A low MOI (0.03) and short time of culture (6 h) were used in an effort to have each infected HFF contain ~ 1 copy of CMV DNA, although at this low MOI not every cell is infected. However, the short period of infection would predict that the late mRNA (pp67) measured in the NucliSens test was only transcribed at very low levels. To further evaluate this, we repeated the infection incubating for 6 and 24 h, with the assumption that the longer culture time would allow for more expression of pp67 mRNA. Replicates of 100 and 1,000 input target cells from the 6- and 24-h standards were tested in the NucliSens test in two laboratories (Table 4). A greater number of replicates were positive at both input concentrations (100 and 1,000 cells) with the 24-h standard than with the 6-h standard. Based on these findings we concluded that the analytical sensitivity of the NucliSens test could not be accurately determined with the 6-h standard used to assess the Amplicor and Hybrid Capture tests.
fig.ommittedz$, 百拇医药
TABLE 3. Results of NucliSens testingz$, 百拇医药
fig.ommittedz$, 百拇医药
TABLE 4. Comparison of 6- and 24-h standards with the NucliSens assayz$, 百拇医药
DISCUSSIONz$, 百拇医药
Quantitative CMV DNA tests are being used with increasing frequency in the management of CMV disease in human immunodeficiency virus type 1-infected patients and in transplant recipients. Although two commercial quantitative tests are available (the Amplicor Monitor CMV and Hybrid Capture tests), many laboratories prefer to use laboratory-developed tests. The use of these different tests complicates the interpretation of studies because viral load levels among the different tests are not always comparable. To address this problem, our study compares the sensitivity, specificity, linear range, and variation over a linear range of the Amplicor PCR and Hybrid Capture tests based on a cell-based standard. To our knowledge, this is the first reported evaluation of these tests using the same standard specimen type. Our results show that the Amplicor PCR test is 1.5 to 2.0 log10 more sensitive than the Hybrid Capture test. Whether the increased analytical sensitivity seen for the Amplicor PCR test versus the Hybrid Capture test is clinically significant requires further investigation. The linear range of the Amplicor test was limited to 2.5 log10 (2.0 to 4.5 log10), whereas the full extent of the linear range of the Hybrid Capture test could not be evaluated because we did not test input a DNA concentration greater than 5.0 log10 target cells. Based on our data, the upper linear range exceeds that of the Amplicor test by at least 0.5 log10. For concentrations of CMV target DNA that were within the linear range of both tests, the viral load values were remarkably similar, i.e., within twofold (0.12 log10). These data were very useful when the results of both tests were compared by using whole blood specimens. However, the Amplicor PCR test as formulated by the manufacturer and as performed in most laboratories uses plasma specimens. The data presented here do not provide information on how values compare when the viral load is measured in plasma specimens with the Amplicor PCR test and in whole-blood specimens with the Hybrid Capture test. The viral load has been shown to be higher in polymorphonuclear leukocytes than in plasma (2, 5, 10). A recent study compared viral loads obtained by using plasma specimens in the Amplicor PCR test and whole-blood specimens in the Hybrid Capture test (15). At the lower limit of detection of the Amplicor PCR test (2.60 log10 copies/ml, 400 copies/ml), the corresponding Hybrid Capture measurements were 1.2 to 1.3 log10 higher in numerical value than the Amplicor PCR value (3.86 log10 or 7,244 genomes/ml). When the Amplicor PCR viral load was 5.0 log10 copies/ml of plasma, the corresponding measurement by the Hybrid Capture test was 5.23 log10 or 169,000 genomes/ml of whole blood.
The variation in viral load values was greater overall for the Hybrid Capture test than that seen with the Amplicor PCR test. For both tests, the standard deviation was greater at lower concentrations of input CMV DNA. Based on our data, changes in viral load observed with the Amplicor PCR test will need to meet or exceed fivefold (0.69 log10) to represent a biologically significant change in viral replication. The 5-fold difference will apply at viral load values of greater than ~ 3.5 log10 (~ 3,200 copies/ml), whereas a >5-fold difference will apply for values of <3.5 log10. For the Hybrid capture test, changes in viral load at the lower end of the test (<3.6 log10, ~ 4,000 copies) may need to approach 10-fold to be considered a biologically significant change in viral replication; for values greater than 4.0 log10, a fivefold change in viral load will likely be clinically relevant.j, http://www.100md.com
The analytical sensitivity of the qualitative NucliSens test could not be determined by using our cell based standard. It appears the low MOI and short culture time (6 h) did not allow for adequate transcription of pp67, the late mRNA measured in the test. This interpretation is supported by the results of our comparison of standards made by culturing cells for either 6 or 24 h because a greater number of replicates of standards containing 100 or 1,000 cells were detected after 24 h of infection compared to 6 h. Developing a standard to compare the quantitative CMV DNA tests and the NucliSens test will be challenging because the biologic basis of the tests differs. We would expect even greater amounts of late mRNA transcripts if a longer incubation period were used, i.e., 96 h. Under these conditions, it is likely that the sensitivity of the NucliSens test would increase above that of the 24-h standard.
Of the 176 samples tested in the NucliSens assay, 2 (1.1%) gave an invalid result, which is in close agreement to the invalid rate reported in a previous study (<0.5%) (4). There were no invalid results with the Hybrid Capture or Amplicor assays.y}u!, 百拇医药
Both the Amplicor PCR and Hybrid Capture tests are technically straightforward to perform. The Amplicor test is more sensitive and specific than the Hybrid Capture test, but the Hybrid Capture test remains linear at higher concentrations of CMV target cells. In spite of differences between test performance, the values obtained with both tests are within twofold with the cell-based standard. For both tests, it is important to avoid overinterpretation of small changes in viral load, particularly near the limits of detection of the tests. However, studies have shown that both the Amplicor and the Hybrid Capture tests are useful in monitoring viral load changes in response to therapy (1, 4, 12).y}u!, 百拇医药
In the present study, we developed the first cell-based standard for CMV quantitation, and we used this standard for direct comparison of the quantitative Amplicor PCR and Hybrid Capture tests. In extending the study to include the qualitative NucliSens test, we were able to demonstrate that the biological properties of this test based on late mRNA transcription differ from the two quantitative DNA tests. It is very important to consider this last point in assessing clinical studies that compare these different tests.
ACKNOWLEDGMENTS*td(, http://www.100md.com
This work was supported in part by NIH contract NO1AI35172 to J.B., Roche Diagnostics Systems, Digene Corp., and Organon Teknika (now bioMérieux).*td(, http://www.100md.com
We thank Jessica Ingersoll and Alicia Green of Emory University Hospital, Sal Scianna at Rush-Presbyterian-St. Luke's Medical Center, and Colleen Starkey and Debbie Kohn at the Cleveland Clinic for technical assistance.*td(, http://www.100md.com
REFERENCES*td(, http://www.100md.com
Aitken, C., W. Barrett-Muir, C. T. Millar, K., J. Thomas, F. Sheridan, D. Jeffries, M. Yaqoob, and J. Breuer. 1999. Use of molecular tests in diagnosis and monitoring of cytomegalovirus disease following renal transplantation. J. Clin. Microbiol. 37:2804-2807.*td(, http://www.100md.com
Boivin, G., J. Handfield, E. Toma, G. Murray, R. Lalonde, and M. G. Bergeron. 1998. Comparative evaluation of the cytomegalovirus DNA load in polymorphonuclear leukocytes and plasma of human immunodeficiency virus-infected subjects. J. Infect. Dis. 177:355-360.
Caliendo, A. M., R. Schuurman, B. Yen-Lieberman, S. A. Spector, J. Andersen, R. Manjiry, C. Crumpacker, N. S. Lurain, A. Erice, et al. 2001. Comparison of quantitative and qualitative PCR tests for cytomegalovirus DNA in plasma. J. Clin. Microbiol. 39:1334-1338.k%os[, 百拇医药
Caliendo, A. M., K. St. George, J. Allega, A. C. Bullotta, L. Gilbane, and C. R. Rinaldo. 2002. Distinguishing cytomegalovirus (CMV) infection and disease with CMV nucleic acid tests. J. Clin. Microbiol. 40:1581-1586.k%os[, 百拇医药
Gerna, G., M. Zavattoni, F. Baldanti, A. Sarasini, L. Chezzi, P. Grossi, and M. Grazia Revello. 1998. Human cytomegalovirus (HCMV) leukodnaemia correlates more closely with clinical symptoms than antigenemia and viremia in heart and heart-lung transplant recipients with primary HCMV infection. Transplantation 65:1378-1385.k%os[, 百拇医药
Humar, A., D. Gregson, A. M. Caliendo, A. McGeer, G. Malkan, M. Krajden, P. Corey, P. Greig, S. Walmsley, G. Levy, and T. Mazzulli. 1999. Clinical utility of quantitative cytomegalovirus viral load determination for predicting cytomegalovirus disease in liver transplant recipients. Transplantation 68:1305-1311.
Middeldorf, J., P. Sillekens, and J. Lunenberg. 2000. Diagnosis of active HCMV infection: the mRNA approach. Organ Tissues 2:99-107.';j*c, http://www.100md.com
Persson, T., and H. Rootzen. 1977. Simple and highly efficient estimators for a type 1 censored normal sample. Biometrika 64:123-128.';j*c, http://www.100md.com
Razonable, R. R., R. A. Brown, M. J. Espy, A. Rivero, W. Kremers, J. A. Wilson, C. M. Groettum, T. F. Smith, and C. V. Paya. 2001. Comparative quantitation of cytomegalovirus (CMV) DNA in solid organ transplant recipients with CMV infection by using two high-throughput automated systems. J. Clin. Microbiol. 39:4472-4476.';j*c, http://www.100md.com
Razonable, R. R., R. A. Brown, J. A. Wilson, C. M. Groettum, W. Kremers, M. J. Espy, T. F. Smith, and C. V. Paya. 2002. The clinical use of various blood compartments for cytomegalovirus (CMV) DNA quantitation in transplant recipients with CMV disease. Transplantation 73:968-973.';j*c, http://www.100md.com
Sia, I. G., J. A. Wilson, M. J. Espy, C. V. Paya, and T. F. Smith. 2000. Evaluation of the COBAS AMPLICOR CMV monitor test for detection of viral DNA in specimens taken from patients after liver transplantation. J. Clin. Microbiol. 38:600-606.
Sia, I. G., J. A. Wilson, C. M. Groettum, M. J. Espy, T. F. Smith, and C. V. Paya. 2000. Cytomegalovirus (CMV) DNA load predicts relapsing CMV infection after solid organ transplantation. J. Infect. Dis. 181:717-720.v|, http://www.100md.com
Spector, S. A., K. Hsia, M. Crager, M. Pilcher, S. Cabral, and M. J. Stempien. 1999. Cytomegalovirus (CMV) DNA load is an independent predictor of CMV disease and survival in advanced AIDS. J. Virol. 73:7027-7030.v|, http://www.100md.com
Spector, S. A., R. Wong, K. Hsia, M. Pilcher, and M. J. Stempien. 1998. Plasma cytomegalovirus (CMV) DNA load predicts CMV disease and survival in AIDS patients. J. Clin. Investig. 101:497-502.v|, http://www.100md.com
Tong, C. Y. W., L. E. Cuevas, H. Williams, and A. Bakran. 2000. Comparison of two commercial methods for measurement of cytomegalovirus load in blood samples after renal transplantation. J. Clin. Microbiol. 38:1209-1213.v|, http://www.100md.com
von Muller, L., W. Hampl, J. Hinz, H. Meisel, A. Reip, E. Engelmann, R. Heilbronn, B. Gartner, O. Kramer, H. Einsele, H. Hebart, T. Ljubicic, J. Loffler, and T. Mertens. 2002. High variability between results of different in-laboratory tests for cytomegalovirus (CMV) monitoring and a standardized quantitative plasma CMV PCR test. J. Clin. Microbiol. 40:2285-2287.v|, http://www.100md.com
Wattanamano, P., J. L. Clayton, J. J. Kopicko, P. Kissinger, S. Elliot, C. Jarrott, S. Rangan, and M. A. Beilke. 2000. Comparison of three tests for cytomegalovirus detection in AIDS patients at risk for retinitis. J. Clin. Microbiol. 38:727-732.v|, http://www.100md.com
Yen-Lieberman, B., D. Brambilla, B. Jackson, J. Bremer, R. Coombs, M. Cronin, S. Herman, D. Katzenstein, S. Leung, H. J. Lin, P. Palumbo, S. Rasheed, J. Todd, M. Vahey, and P. Reichelderfer. 1996. Evaluation of a quality assurance program for quantification of human immunodeficiency virus type 1 RNA in plasma and AIDS Clinical Trials Group virology laboratories. J. Clin. Microbiol. 34:2695-2701.(Angela M. Caliendo Belinda Yen-Lieberman Jovana Baptista Janet Andersen Clyde Crumpacker Rob Schuurm)
Received 28 February 2003/ Returned for modification 25 April 2003/ Accepted 28 May 2003f#rh, 百拇医药
ABSTRACTf#rh, 百拇医药
A cell-based standard was developed to compare the COBAS Amplicor CMV Monitor test, the Hybrid Capture System CMV DNA test, and the NucliSens CMV test. The standard was prepared by infecting human foreskin fibroblasts (HFFs) with cytomegalovirus (CMV) strain AD169 at low multiplicity of infection (0.03) and harvesting the cells at 6 h postinfection. Buffy coat cells were added to produce concentrations of from 0 to 105 HFFs per 106 total cells. Five laboratories performed the Amplicor PCR test and two laboratories performed the NucliSens and Hybrid Capture tests. The Amplicor PCR test was 1.5 to 2.0 log10 more sensitive than the Hybrid Capture test. The specificities of the Amplicor PCR and Hybrid Capture tests were 100 and 93.8%, respectively. The linear range of the Amplicor PCR and Hybrid Capture tests were 2 to 4.48 log10 and 3.48 to at least 5.0 log10 CMV target cells, respectively. The standard deviations of the Amplicor PCR and Hybrid Capture tests varied throughout their linear range, and for both tests the variability was greater for lower concentrations of input CMV DNA. These data allow the direct comparison of viral load values between the Amplicor and Hybrid Capture tests. The analytical sensitivity of the NucliSens test could not be determined by using the 6-h standard, because the low multiplicity of infection and short culture time did not allow for adequate transcription of pp67 late mRNA measured in the test. Extending the incubation time of the standard to 24 h increased the analytical sensitivity of the NucliSens test to 3.0 log10 target cells.
INTRODUCTIONp9., 百拇医药
Cytomegalovirus (CMV) is an important pathogen in immunosuppressed patients, including persons with AIDS and organ transplant recipients. The clinical utility of quantitative CMV DNA tests in managing these patient groups has been well documented (1, 4, 6, 12-14). In transplant recipients, CMV load is used to diagnose active CMV disease, screen patients for the use of preemptive therapy, and monitor the response to antiviral therapy (1, 4, 6, 11). In patients infected with human immunodeficiency virus type 1, the risk of developing CMV disease has been reported to be directly related to the quantity of CMV DNA in plasma (13, 14).p9., 百拇医药
There are several tests available for detecting and quantifying CMV DNA from clinical samples. The COBAS Amplicor CMV Monitor test (Roche Diagnostics, Indianapolis, Ind.) is a DNA PCR test that amplifies a 365-bp region of the CMV polymerase gene. Using a plasma-based standard, the linear range of the test is reported to be between 400 and 50,000 copies/ml of plasma (3). The test has been designed for plasma but can be modified for use with leukocytes or whole blood specimens. The Hybrid Capture System CMV DNA test (Digene Corp., Gaithersburg, Md.) is a signal amplification method that uses an RNA probe which targets 17% of the CMV genome. The target is detected by using antibodies that specifically bind RNA-DNA hybrids. The test is designed for whole blood specimens with a linear range of 1,400 to 560,000 copies/ml (17). The availability of commercial tests has been helpful in standardizing test performance, although neither the Amplicor PCR nor the quantitative Hybrid Capture tests have been cleared by the U.S. Food and Drug Administration (FDA; a qualitative version of the Hybrid Capture test has been cleared by the FDA). The NucliSens CMV test (bioMérieux, Durham, N.C. [FDA cleared]) is a qualitative RNA amplification test that detects pp67 mRNA, which is transcribed late in the viral replicative cycle. Since the NucliSens test is isothermal, it can detect pp67 RNA without interference of CMV DNA. Patients with active CMV disease express high levels of pp67 (7), which are not expressed during latent infection. In addition to these commercially available tests, "laboratory-developed" DNA PCR tests are widely used in clinical laboratories. The performance of laboratory-developed tests varies due to differences in specimen type, nucleic acid extraction method, primers, quantitation standard, and detection method (16).
There have been a few studies comparing the Amplicor PCR test to laboratory-developed PCR tests. Comparison of a LightCycler PCR test to the Amplicor PCR test showed that the Amplicor PCR test yielded viral load results that were 1 log10 higher than those of the LightCycler test (9). Comparison of the Amplicor PCR test with a different laboratory-developed test showed the viral load values to be 0.5 to 0.7 log10 higher with the laboratory-developed test than with the Amplicor PCR test (3). These results underscore the variation in viral load measurements that are obtained by using different quantitative tests. To address this problem, we developed a standard reagent to compare the performance of the Amplicor PCR test, the Hybrid Capture test, and the NucliSens CMV test. We chose a cell-based standard for two reasons: first, CMV is highly cell associated (5, 10) and, second, an infected-cell standard is the only type that would allow the comparison of all three tests.m, http://www.100md.com
MATERIALS AND METHODS9, 百拇医药
Cell-based standard. Human foreskin fibroblasts (HFFs) were inoculated with the CMV laboratory strain AD169 at a multiplicity of infection (MOI) of 0.03. After a 4-h adsorption period, the inoculum was removed. The monolayer was washed withphosphate-buffered saline, and fresh medium was added for an additional 2-h incubation. The 6-h total incubation time was designed to allow the virus to enter the cells but was short enough to avoid intracellular replication of CMV. At this low MOI only 3 per 100 target cells would be expected to be infected with CMV. At 6 h postinfection the HFFs were harvested by trypsinization, washed, and counted. The HFFs were diluted with uninfected buffy coat cells to concentrations of 0 to 105 HFF per 106 total cells. Then, 200-µl aliquots containing 106 total cells were stored at -70°C until they were tested in the Amplicor test. The aliquots (200 µl, 106 total cells) for the Hybrid Capture test were centrifuged (14,000 x g), the supernatant was removed, and the pellets were stored at -70°C until testing was performed. For the evaluation of the NucliSens test a second 24-h standard was designed with a 4-h adsorption period, followed by a 20-h incubation period. Buffy coat cells were added as described above for the other tests. For the NucliSens test, 200-µl aliquots of the standard were added to 0.9 ml of NucliSens lysis buffer; the tubes were then rocked for 30 min before storage at -70°C.
Nucleic acid testing. Viral load testing was done by using the COBAS Amplicor CMV Monitor Test (Roche Diagnostics) and the Hybrid Capture System CMV DNA test (version 2.0; Digene Corp.). For the Amplicor test, 200-µl aliquots of the cell suspensions were added to lysis buffer and tested according to the manufacturer's protocol. For the Hybrid Capture test, the pellets from the 200-µl aliquots were resuspended in lysis buffer and tested according to the manufacturer's protocol. For the NucliSens test, lysis buffer tubes containing the 200-µl samples were tested according to the manufacturer's protocol. The results of the NucliSens test were reported as positive or negative.g!x, 百拇医药
Five laboratories performed the Amplicor test, with each laboratory testing eight replicates of 10 concentrations of the standard containing a range of target cells from 3 to 105. The Hybrid Capture test was done in two laboratories, each of which tested two different panels of the standard. The panel consisted of eight replicates of seven concentrations of the standard, with a range of 100 to 105 target cells. NucliSens testing was performed in two laboratories, with each laboratory testing a panel that consisted of eight replicates of the same target cell concentrations used for the Amplicor test.
Statistical analysis. The results of the quantitative tests were log10 transformed for analysis. The linear range was examined by comparison to a 45° line. Crude within-laboratory standard deviations were tools for detecting departure from the linear range. One of the Amplicor panels contained four missing results; these observations were excluded from the analysis. Methods accommodating censored values were used (8) when we calculated the mean copy number for the Hybrid Capture test at an input concentration of 1,000 target cells because some replicates had viral load levels below the limit of detection. For the NucliSens test, one specimen produced an invalid test result and was positive on repeat testing, this specimen was analyzed as a positive result. Another specimen had two consecutive invalid results as determined by using the NucliSens test; this specimen was excluded from analysis.]'8du.{, http://www.100md.com
RESULTS]'8du.{, http://www.100md.com
The Amplicor test was able to detect 50% of replicates with 1.48 log10 (i.e., 30) CMV target cells and reached a sensitivity of 100% when the input concentration was 2.0 log10 (i.e., 100) target cells (Table 1). For the Hybrid Capture test, at least 50% of the replicates were detected when the input CMV concentration was 3.0 log10 (i.e., 1,000) target cells, 97% of replicates were detected at 3.48 log10 (i.e., 3,000) target cells, and the sensitivity was 100% at an input concentration of 4.0 log10 (i.e., 10,000) target cells. The specificity of the Amplicor test was 100%, whereas the specificity of the Hybrid Capture test was 93.8% (Table 1).
fig.ommitted-4, 百拇医药
TABLE 1. Results of Amplicor and Hybrid Capture testing-4, 百拇医药
The means and standard deviations of the viral load values from all laboratories are shown in Table 2 by input CMV target cell concentration. One replicate with an input concentration of 5.0 log10 CMV target cells was found to contain no virus when tested in the Amplicor PCR test. When this outlier was included in the calculation of the mean and standard deviation (a value of 2.3 log10 [i.e., 200] was used), the mean viral load was 4.496 log10 copies/106 cells with a standard deviation of 0.37 log10 copies/106 cells. When this value was excluded from the calculation, the mean was 4.55 log10 copies/106, and the standard deviation was 0.11 log10 copies/106 cells. We felt that the outlier with no viral load detected was due to a technical error either in preparing the standard or in performing the viral load testing. Further discussions of the test performance refer to the values obtained with the outlier excluded, since the underlying results of the study are unchanged by the exclusion.
fig.ommitted4;, 百拇医药
TABLE 2. Mean results of the Amplicor and Hybrid Capture testing4;, 百拇医药
The linear range of the Amplicor test was 2.0 to 4.48 log10 (i.e., 100 to 30,000) CMV target cells, whereas the Hybrid Capture test was linear, with results from 3.48 to at least 5.0 log10 (i.e., 3,000 to at least 100,000) CMV target cells (Fig. 1). We did not test a high enough concentration of target cells to determine the upper limit of the linear range of the Hybrid Capture test. For concentrations of CMV target DNA that were within the linear range of both tests (3.48, 4.0, and 4.48 log10), the viral load values were within twofold.4;, 百拇医药
fig.ommitted4;, 百拇医药
FIG. 1. Plot of the mean log10 CMV DNA copy number versus the input concentration of the quantitation standard for the Amplicor PCR test (A) and Hybrid Capture test (B). The open circle in panel A represents the mean with the outlier excluded.4;, 百拇医药
The standard deviations of the tests varied throughout their linear ranges. The standard deviation of the Amplicor PCR test ranged from 0.10 to 0.37 log10 CMV copies/106 cells. When the outlier at 5.0 log10 target cells was excluded, the standard deviation ranged from 0.10 to 0.28 log10 CMV copies/106 cells (Fig. 2). The standard deviation was higher at lower input concentrations. Similarly, for the Hybrid Capture tests, more variability was observed for the lower concentrations of CMV DNA compared to the higher concentration, with standard deviations ranging from 0.17 to 0.65 log10 CMV copies/106 cells (Fig. 2). A standard deviation of 0.15 log10 is used by the Virology Quality Assessment Laboratory program to ensure that a laboratory could maintain the precision required to have 90% power to detect a fivefold (0.69 log10) difference in copy number between two samples in the same batch (18). Based on these data, overall changes in viral load measured in the Amplicor and Hybrid Capture tests need to meet or exceed fivefold (0.69 log10 copies) to be interpreted as clinically significant, particularly at the lower end of these tests.
fig.ommitted#(rjr, http://www.100md.com
FIG. 2. Plot of the standard deviations of the log10 CMV DNA copy number versus the input concentration of the quantitation standard for the Amplicor test (A) and Hybrid Capture test (B). The value indicated by the asterisk represents the standard deviation with the outlier excluded.#(rjr, http://www.100md.com
Qualitative results obtained with the NucliSens test are shown in Table 3. At input concentrations of between 3.0 and 3.5 log10, 50% of the replicates could be detected, with the analytical sensitivity reaching 100% at 4.5 log10 CMV target cells. This low level of sensitivity is probably due to the manner in which the cell standard was designed. A low MOI (0.03) and short time of culture (6 h) were used in an effort to have each infected HFF contain ~ 1 copy of CMV DNA, although at this low MOI not every cell is infected. However, the short period of infection would predict that the late mRNA (pp67) measured in the NucliSens test was only transcribed at very low levels. To further evaluate this, we repeated the infection incubating for 6 and 24 h, with the assumption that the longer culture time would allow for more expression of pp67 mRNA. Replicates of 100 and 1,000 input target cells from the 6- and 24-h standards were tested in the NucliSens test in two laboratories (Table 4). A greater number of replicates were positive at both input concentrations (100 and 1,000 cells) with the 24-h standard than with the 6-h standard. Based on these findings we concluded that the analytical sensitivity of the NucliSens test could not be accurately determined with the 6-h standard used to assess the Amplicor and Hybrid Capture tests.
fig.ommittedz$, 百拇医药
TABLE 3. Results of NucliSens testingz$, 百拇医药
fig.ommittedz$, 百拇医药
TABLE 4. Comparison of 6- and 24-h standards with the NucliSens assayz$, 百拇医药
DISCUSSIONz$, 百拇医药
Quantitative CMV DNA tests are being used with increasing frequency in the management of CMV disease in human immunodeficiency virus type 1-infected patients and in transplant recipients. Although two commercial quantitative tests are available (the Amplicor Monitor CMV and Hybrid Capture tests), many laboratories prefer to use laboratory-developed tests. The use of these different tests complicates the interpretation of studies because viral load levels among the different tests are not always comparable. To address this problem, our study compares the sensitivity, specificity, linear range, and variation over a linear range of the Amplicor PCR and Hybrid Capture tests based on a cell-based standard. To our knowledge, this is the first reported evaluation of these tests using the same standard specimen type. Our results show that the Amplicor PCR test is 1.5 to 2.0 log10 more sensitive than the Hybrid Capture test. Whether the increased analytical sensitivity seen for the Amplicor PCR test versus the Hybrid Capture test is clinically significant requires further investigation. The linear range of the Amplicor test was limited to 2.5 log10 (2.0 to 4.5 log10), whereas the full extent of the linear range of the Hybrid Capture test could not be evaluated because we did not test input a DNA concentration greater than 5.0 log10 target cells. Based on our data, the upper linear range exceeds that of the Amplicor test by at least 0.5 log10. For concentrations of CMV target DNA that were within the linear range of both tests, the viral load values were remarkably similar, i.e., within twofold (0.12 log10). These data were very useful when the results of both tests were compared by using whole blood specimens. However, the Amplicor PCR test as formulated by the manufacturer and as performed in most laboratories uses plasma specimens. The data presented here do not provide information on how values compare when the viral load is measured in plasma specimens with the Amplicor PCR test and in whole-blood specimens with the Hybrid Capture test. The viral load has been shown to be higher in polymorphonuclear leukocytes than in plasma (2, 5, 10). A recent study compared viral loads obtained by using plasma specimens in the Amplicor PCR test and whole-blood specimens in the Hybrid Capture test (15). At the lower limit of detection of the Amplicor PCR test (2.60 log10 copies/ml, 400 copies/ml), the corresponding Hybrid Capture measurements were 1.2 to 1.3 log10 higher in numerical value than the Amplicor PCR value (3.86 log10 or 7,244 genomes/ml). When the Amplicor PCR viral load was 5.0 log10 copies/ml of plasma, the corresponding measurement by the Hybrid Capture test was 5.23 log10 or 169,000 genomes/ml of whole blood.
The variation in viral load values was greater overall for the Hybrid Capture test than that seen with the Amplicor PCR test. For both tests, the standard deviation was greater at lower concentrations of input CMV DNA. Based on our data, changes in viral load observed with the Amplicor PCR test will need to meet or exceed fivefold (0.69 log10) to represent a biologically significant change in viral replication. The 5-fold difference will apply at viral load values of greater than ~ 3.5 log10 (~ 3,200 copies/ml), whereas a >5-fold difference will apply for values of <3.5 log10. For the Hybrid capture test, changes in viral load at the lower end of the test (<3.6 log10, ~ 4,000 copies) may need to approach 10-fold to be considered a biologically significant change in viral replication; for values greater than 4.0 log10, a fivefold change in viral load will likely be clinically relevant.j, http://www.100md.com
The analytical sensitivity of the qualitative NucliSens test could not be determined by using our cell based standard. It appears the low MOI and short culture time (6 h) did not allow for adequate transcription of pp67, the late mRNA measured in the test. This interpretation is supported by the results of our comparison of standards made by culturing cells for either 6 or 24 h because a greater number of replicates of standards containing 100 or 1,000 cells were detected after 24 h of infection compared to 6 h. Developing a standard to compare the quantitative CMV DNA tests and the NucliSens test will be challenging because the biologic basis of the tests differs. We would expect even greater amounts of late mRNA transcripts if a longer incubation period were used, i.e., 96 h. Under these conditions, it is likely that the sensitivity of the NucliSens test would increase above that of the 24-h standard.
Of the 176 samples tested in the NucliSens assay, 2 (1.1%) gave an invalid result, which is in close agreement to the invalid rate reported in a previous study (<0.5%) (4). There were no invalid results with the Hybrid Capture or Amplicor assays.y}u!, 百拇医药
Both the Amplicor PCR and Hybrid Capture tests are technically straightforward to perform. The Amplicor test is more sensitive and specific than the Hybrid Capture test, but the Hybrid Capture test remains linear at higher concentrations of CMV target cells. In spite of differences between test performance, the values obtained with both tests are within twofold with the cell-based standard. For both tests, it is important to avoid overinterpretation of small changes in viral load, particularly near the limits of detection of the tests. However, studies have shown that both the Amplicor and the Hybrid Capture tests are useful in monitoring viral load changes in response to therapy (1, 4, 12).y}u!, 百拇医药
In the present study, we developed the first cell-based standard for CMV quantitation, and we used this standard for direct comparison of the quantitative Amplicor PCR and Hybrid Capture tests. In extending the study to include the qualitative NucliSens test, we were able to demonstrate that the biological properties of this test based on late mRNA transcription differ from the two quantitative DNA tests. It is very important to consider this last point in assessing clinical studies that compare these different tests.
ACKNOWLEDGMENTS*td(, http://www.100md.com
This work was supported in part by NIH contract NO1AI35172 to J.B., Roche Diagnostics Systems, Digene Corp., and Organon Teknika (now bioMérieux).*td(, http://www.100md.com
We thank Jessica Ingersoll and Alicia Green of Emory University Hospital, Sal Scianna at Rush-Presbyterian-St. Luke's Medical Center, and Colleen Starkey and Debbie Kohn at the Cleveland Clinic for technical assistance.*td(, http://www.100md.com
REFERENCES*td(, http://www.100md.com
Aitken, C., W. Barrett-Muir, C. T. Millar, K., J. Thomas, F. Sheridan, D. Jeffries, M. Yaqoob, and J. Breuer. 1999. Use of molecular tests in diagnosis and monitoring of cytomegalovirus disease following renal transplantation. J. Clin. Microbiol. 37:2804-2807.*td(, http://www.100md.com
Boivin, G., J. Handfield, E. Toma, G. Murray, R. Lalonde, and M. G. Bergeron. 1998. Comparative evaluation of the cytomegalovirus DNA load in polymorphonuclear leukocytes and plasma of human immunodeficiency virus-infected subjects. J. Infect. Dis. 177:355-360.
Caliendo, A. M., R. Schuurman, B. Yen-Lieberman, S. A. Spector, J. Andersen, R. Manjiry, C. Crumpacker, N. S. Lurain, A. Erice, et al. 2001. Comparison of quantitative and qualitative PCR tests for cytomegalovirus DNA in plasma. J. Clin. Microbiol. 39:1334-1338.k%os[, 百拇医药
Caliendo, A. M., K. St. George, J. Allega, A. C. Bullotta, L. Gilbane, and C. R. Rinaldo. 2002. Distinguishing cytomegalovirus (CMV) infection and disease with CMV nucleic acid tests. J. Clin. Microbiol. 40:1581-1586.k%os[, 百拇医药
Gerna, G., M. Zavattoni, F. Baldanti, A. Sarasini, L. Chezzi, P. Grossi, and M. Grazia Revello. 1998. Human cytomegalovirus (HCMV) leukodnaemia correlates more closely with clinical symptoms than antigenemia and viremia in heart and heart-lung transplant recipients with primary HCMV infection. Transplantation 65:1378-1385.k%os[, 百拇医药
Humar, A., D. Gregson, A. M. Caliendo, A. McGeer, G. Malkan, M. Krajden, P. Corey, P. Greig, S. Walmsley, G. Levy, and T. Mazzulli. 1999. Clinical utility of quantitative cytomegalovirus viral load determination for predicting cytomegalovirus disease in liver transplant recipients. Transplantation 68:1305-1311.
Middeldorf, J., P. Sillekens, and J. Lunenberg. 2000. Diagnosis of active HCMV infection: the mRNA approach. Organ Tissues 2:99-107.';j*c, http://www.100md.com
Persson, T., and H. Rootzen. 1977. Simple and highly efficient estimators for a type 1 censored normal sample. Biometrika 64:123-128.';j*c, http://www.100md.com
Razonable, R. R., R. A. Brown, M. J. Espy, A. Rivero, W. Kremers, J. A. Wilson, C. M. Groettum, T. F. Smith, and C. V. Paya. 2001. Comparative quantitation of cytomegalovirus (CMV) DNA in solid organ transplant recipients with CMV infection by using two high-throughput automated systems. J. Clin. Microbiol. 39:4472-4476.';j*c, http://www.100md.com
Razonable, R. R., R. A. Brown, J. A. Wilson, C. M. Groettum, W. Kremers, M. J. Espy, T. F. Smith, and C. V. Paya. 2002. The clinical use of various blood compartments for cytomegalovirus (CMV) DNA quantitation in transplant recipients with CMV disease. Transplantation 73:968-973.';j*c, http://www.100md.com
Sia, I. G., J. A. Wilson, M. J. Espy, C. V. Paya, and T. F. Smith. 2000. Evaluation of the COBAS AMPLICOR CMV monitor test for detection of viral DNA in specimens taken from patients after liver transplantation. J. Clin. Microbiol. 38:600-606.
Sia, I. G., J. A. Wilson, C. M. Groettum, M. J. Espy, T. F. Smith, and C. V. Paya. 2000. Cytomegalovirus (CMV) DNA load predicts relapsing CMV infection after solid organ transplantation. J. Infect. Dis. 181:717-720.v|, http://www.100md.com
Spector, S. A., K. Hsia, M. Crager, M. Pilcher, S. Cabral, and M. J. Stempien. 1999. Cytomegalovirus (CMV) DNA load is an independent predictor of CMV disease and survival in advanced AIDS. J. Virol. 73:7027-7030.v|, http://www.100md.com
Spector, S. A., R. Wong, K. Hsia, M. Pilcher, and M. J. Stempien. 1998. Plasma cytomegalovirus (CMV) DNA load predicts CMV disease and survival in AIDS patients. J. Clin. Investig. 101:497-502.v|, http://www.100md.com
Tong, C. Y. W., L. E. Cuevas, H. Williams, and A. Bakran. 2000. Comparison of two commercial methods for measurement of cytomegalovirus load in blood samples after renal transplantation. J. Clin. Microbiol. 38:1209-1213.v|, http://www.100md.com
von Muller, L., W. Hampl, J. Hinz, H. Meisel, A. Reip, E. Engelmann, R. Heilbronn, B. Gartner, O. Kramer, H. Einsele, H. Hebart, T. Ljubicic, J. Loffler, and T. Mertens. 2002. High variability between results of different in-laboratory tests for cytomegalovirus (CMV) monitoring and a standardized quantitative plasma CMV PCR test. J. Clin. Microbiol. 40:2285-2287.v|, http://www.100md.com
Wattanamano, P., J. L. Clayton, J. J. Kopicko, P. Kissinger, S. Elliot, C. Jarrott, S. Rangan, and M. A. Beilke. 2000. Comparison of three tests for cytomegalovirus detection in AIDS patients at risk for retinitis. J. Clin. Microbiol. 38:727-732.v|, http://www.100md.com
Yen-Lieberman, B., D. Brambilla, B. Jackson, J. Bremer, R. Coombs, M. Cronin, S. Herman, D. Katzenstein, S. Leung, H. J. Lin, P. Palumbo, S. Rasheed, J. Todd, M. Vahey, and P. Reichelderfer. 1996. Evaluation of a quality assurance program for quantification of human immunodeficiency virus type 1 RNA in plasma and AIDS Clinical Trials Group virology laboratories. J. Clin. Microbiol. 34:2695-2701.(Angela M. Caliendo Belinda Yen-Lieberman Jovana Baptista Janet Andersen Clyde Crumpacker Rob Schuurm)