Performance Characteristics of the COBAS AMPLICOR Hepatitis C Virus MONITOR Test, Version 2.0

Clinical Chemistry / EVALUATION OF QUANTITATIVE HEPATITIS C VIRUS ASSAY Performance Characteristics of the COBAS AMPLICOR Hepatitis C Virus MONITOR Test, Version 2.0 Maria Erali, MS, 1 Edward R. Ashwood, MD, 1,2 and David R. Hillyard, MD 1,2 Key Words: Hepatitis C virus; HCV RNA quantitation; COBAS AMPLICOR HCV MONITOR test, version 2.0; QUANTIPLEX HCV bdna test Abstract We evaluated the performance characteristics of the COBAS AMPLICOR Hepatitis C Virus (HCV) MONITOR Test, version 2.0. Dilution studies using patient specimens demonstrated a lower limit of detection of 1,000 copies per milliliter. The assay was linear from 1,000 to 1 million HCV RNA copies per milliliter. Within-run precision and between-run precision were acceptable (approximately 0.100 and 0.14 SD for log 10 [copies per milliliter]). A comparison of this version of the test (y), with the manual AMPLICOR HCV MONITOR Test, version 1.0 (x), yielded the following Deming regression equation: y = 1.004(± 0.04)x + 0.654(± 0.22); S y/x D = 0.6; n = 92; r 2 = 0.846; r = 0.920. Further comparison of the COBAS version 2.0 assay (x) with the QUANTIPLEX HCV bdna Test (y) yielded the following Deming regression equation: y = 0.94 (±0.10)x + 0.47 (±0.717); S y/x D = 0.194; n = 26; r 2 = 0.600; r = 0.774. Version 2.0 detected the spectrum of HCV genotypes better than version 1.0. The present study was undertaken to evaluate the performance characteristics of the second-generation AMPLICOR HCV MONITOR Test (Roche Diagnostics, Indianapolis, IN) as it is configured for the COBAS AMPLICOR Analyzer (Roche Diagnostics). Hepatitis C virus (HCV) is a positive-strand RNA virus that is a major cause of viral hepatitis worldwide. The virus has been classified into 6 major genotypes and a large number of subtypes based on the variability of the nucleic acid sequence. 1,2 Since the advent of serologic screening for HCV in 1989, the rate of new infection has declined from 250,000 per year to an estimated 0,000 per year. The total number of persons infected with HCV in the United States is almost 4 million, 2 and as many as 10,000 deaths a year can be attributed to HCV infection. Tests that determine the level of HCV RNA in serum are useful for managing HCV infection. Qualitative and quantitative assays are available for detecting HCV RNA. Quantitative serum HCV RNA results may be useful as prognostic indicators of the severity of disease and response to therapy. They also may be used for monitoring response to therapy and relapse after the end of treatment. There are 2 major commercially available, research use only, methods for determining HCV RNA levels that are based on different techniques: (1) target amplification by reverse transcriptase polymerase chain reaction (RT-PCR) (Roche AMPLICOR) and (2) signal amplification by branched DNA (bdna) (Chiron QUANTIPLEX, Chiron, Emeryville, CA). Others have reported differences between these assays in sensitivity, reproducibility, agreement, and ability to detect genotypes. -7 With the availability of improved treatment regimens, it is important to have reliable tests for measuring HCV RNA levels to provide meaningful information about 180 Am J Clin Pathol 2000;114:180-187 American Society of Clinical Pathologists

Clinical Chemistry / ORIGINAL ARTICLE prognosis and response to therapy. The quantitative AMPLICOR HCV MONITOR Test, version 2.0 (Roche Diagnostics) is the second-generation format for the RT-PCR AMPLICOR HCV MONITOR Test, version 1.0. The version 2.0 chemistry has been modified by the manufacturer in an attempt to improve the efficiency of nucleic acid amplification and is expected to provide enhanced detection of genotypes. In addition, the version 2.0 is available on the COBAS AMPLICOR Analyzer, which allows automated amplification, detection, and reporting for HCV RNA quantitation. We studied the sensitivity and reproducibility of the AMPLICOR HCV MONITOR Test, version 2.0, and compared the version 2.0 assay with the version 1.0. An analysis of the ability of version 2.0 to detect genotypes was made, and results from the AMPLICOR HCV MONITOR Test, version 2.0 were compared with the QUANTIPLEX HCV RNA 2.0 bdna Test. Materials and Methods Clinical Specimens Correlation and precision studies were done using EDTA-anticoagulated plasma or serum samples submitted to ARUP Laboratories, Salt Lake City, UT, for HCV quantitation. Whole blood was drawn in serum separation tubes (SST, Becton Dickinson, Franklin Lakes, NJ) or tubes containing EDTA as anticoagulant and processed within 0 minutes. As described in the AMPLICOR HCV MONITOR package insert, serum or plasma was separated from cells by centrifugation at 1,500g for 20 minutes at room temperature, divided into aliquots, and frozen for storage at 20 C or lower. HCV Quantitation by the AMPLICOR HCV MONITOR Test, Version 1.0 HCV RNA was quantitated using the AMPLICOR HCV MONITOR Test, version 1.0 (Roche Diagnostics) according to the manufacturer s instructions. Viral lysis was done by adding 100 µl of serum or plasma to a lysis solution containing guanidine thiocyanate and the Roche-supplied Quantitation Standard (QS) required for determining RNA levels. The QS is an in vitro transcribed RNA molecule with HCV primer binding sites and a unique probe binding site. The QS is added at a known concentration to each specimen to provide a reference for quantification of HCV RNA. Following lysis, nucleic acid extraction was accomplished by isopropanol precipitation and ethanol wash. The nucleic acid was resuspended in the Roche-supplied specimen diluent containing tris(hydroxymethyl)aminomethane hydrochloride, synthetic poly ra RNA, EDTA, and sodium azide. Extracted samples were added to a PCR reagent mixture containing recombinant Thermus thermophilus (rtth) DNA polymerase in a buffer with the appropriate concentrations of manganese, primers, and nucleotide triphosphates to allow RT-PCR of a 244-base-pair product located in the 5 untranslated region of the HCV genome. AmpErase (uracil N-glycosylase) was incorporated by Roche into the PCR reagent mixture with deoxyuridine triphosphate (dutp) to ensure selective amplification of target nucleic acid. Thermal cycling was done in a GeneAmp PCR System 9600 (Perkin-Elmer, Norwalk, CT). Detection was accomplished using an enzyme-based colorimetric microwell assay format that uses solid-phase probes complementary to the HCV and QS amplicons. Absorbance measurements were done on an El X 800 plate reader (Bio-Tek Instruments, Winooski, VT). Viral RNA levels were calculated as HCV RNA copies per milliliter according to a formula that includes the absorbance values for the QS and HCV amplicons. HCV Quantitation by the COBAS AMPLICOR HCV MONITOR Test, Version 2.0 HCV RNA was quantitated using the COBAS AMPLICOR HCV MONITOR Test, version 2.0 according to the manufacturer s directions. Viral lysis was performed as in the version 1.0 manual assay. Following lysis, nucleic acid extraction was accomplished by isopropanol precipitation and ethanol wash. The nucleic acid was resuspended in 1 ml of specimen diluent. Fifty microliters of the extracted samples were added to 50 µl of the version 2.0 PCR reagent mixture. This reagent is modified in the version 2.0 assay to contain dimethyl sulfoxide in the buffer containing rtth DNA polymerase, manganese, primers, nucleotide triphosphates (including dutp), and AmpErase. The primers allowed RT-PCR of a 244-base-pair product located in the 5 untranslated region of the HCV genome. Thermal cycling and detection were done on the COBAS AMPLICOR Analyzer with no further technician intervention. Detection was accomplished on the COBAS AMPLICOR Analyzer in a colorimetric format using suspensions of magnetic particles coated with probes specific for the HCV and QS amplicons. Absorbance measurements were done by the COBAS AMPLICOR Analyzer. Viral RNA levels were calculated by the instrument and reported as HCV RNA copies per milliliter. HCV Quantitation by the QUANTIPLEX HCV RNA 2.0 bdna Test HCV RNA quantitation of plasma samples by the QUANTIPLEX HCV RNA 2.0 bdna Test was performed by an outside reference laboratory. The bdna method used signal amplification based on the hybridization of probes specific to sequences in the 5 untranslated region of the American Society of Clinical Pathologists Am J Clin Pathol 2000;114:180-187 181

Erali et al / EVALUATION OF QUANTITATIVE HEPATITIS C VIRUS ASSAY HCV genome. Additional probes and synthetic bdna molecules were used to attach multiple enzyme labeled probes that were used in a chemiluminescent reaction to detect HCV RNA in the sample. Quantitation of HCV RNA in the QUANTIPLEX HCV RNA 2.0 bdna Test was accomplished by comparing unknowns to an external calibration curve. Results are reported as HCV RNA equivalents (Eq) per milliliter. Dilution Samples Clinical specimens that were quantitated for HCV RNA levels in the COBAS AMPLICOR HCV MONITOR Test, version 2.0, were diluted serially in HCV seronegative normal human plasma. Dilutions were prepared to provide values throughout the expected range of the assay, 1,000 to 1 million HCV RNA copies per milliliter. Duplicate and triplicate determinations were made on the diluted samples in the COBAS AMPLICOR HCV MONITOR Test, version 2.0. The 1,000 copies per milliliter limit of detection corresponds to 100 extracted copies of nucleic acid (100 µl 1,000 copies per milliliter) and 5 copies introduced to the PCR (50 µl/1,000 µl 100 copies). HCV Genotype Genotype determination was done using RT-PCR amplification followed by nucleic acid sequence analysis. Nucleic acid was extracted from samples by guanidine thiocyanate lysis and isopropanol precipitation with an ethanol wash. RT- PCR was performed in PCR reagent mixture containing rtth polymerase, manganese, primers, nucleotide triphosphates (including dutp), AmpErase, and dimethyl sulfoxide. The primers were identical to the primers used in the AMPLICOR HCV MONITOR tests and allowed RT-PCR of a 244-basepair product located in the 5 untranslated region of the HCV genome. Amplified product was purified using a Qiagen QIAquick PCR purification column (Qiagen, Valencia, CA) and cycle sequenced using Big Dye Terminator chemistry (PE-Applied Biosystems, Foster City, CA). Product from the cycle sequencing was purified by precipitation with an ethanol salt solution and underwent polyacrylamide gel electrophoresis on the ABI Prism 77 DNA Sequencer (PE- Applied Biosystems) for determination of nucleotide sequence. Genotype identification was made by reference to a sequence analysis database program created with HCV genotype sequence information found in the National Center for Biotechnology Information GenBank database. Data Analysis Before analysis, all HCV RNA results were converted to log 10 (copies per milliliter). Reproducibility was evaluated by determining the SD and the percentage of the coefficient of variation for within-run and between-run replicates. Linear regression analysis was used to analyze observed and expected values in the dilution study. Deming regression analysis was used for comparison of methods. Results Linearity Study Serial dilutions of 11 samples were prepared, and the HCV RNA levels were measured in the COBAS AMPLICOR HCV MONITOR Test, version 2.0. The individual dilutions were tested in triplicate for all except 4 dilutions. One dilution was run in duplicate owing to space restrictions on the instrument. Calculations were done using duplicate measurements on other samples, which in each case had 1 replicate that did not quantify. The mean observed values for the 4 samples run in duplicate were.2, 2.9,.6, and.0 log 10 (copies per milliliter). Expected values for each dilution series were established by using the observed result for that series that was greater than.0 and less than 6.0 HCV RNA log 10 (copies per milliliter), usually the result closest to 4.5 log 10 (copies per milliliter). The observed results used to establish the expected values were not used in the final analysis. The dilutions that had measurable HCV RNA (n = 5) were evaluated in a plot of the mean of the triplicate or duplicate observed values vs the expected values Figure 1. These data show that the COBAS AMPLICOR HCV MONITOR Test, version 2.0, is almost linear over the range of.0 to 6.0 HCV RNA log 10 (copies per milliliter). The linearity becomes compromised at measurements made above 6.0 and below.0 HCV RNA log 10 (copies per milliliter) as shown by the leveling off of the data in these ranges. Within-Run Reproducibility Within-run reproducibility was evaluated by using the triplicate determinations made on 56 dilutions evaluated in the linearity study. The SDs of the log 10 (copies per milliliter) were calculated and plotted against the average log 10 (copies per milliliter) for each sample. The trend in data was analyzed using a second-order polynomial curve fit Figure 2. A SD for log 10 (copies per milliliter) of 0.150 or less is considered acceptable within-run reproducibility. The within-run reproducibility for samples with higher HCV RNA levels was well within this specification; however, as the level of HCV RNA approached the.0 log 10 (copies per milliliter) detection limit of the assay, the SD trended upward. Between-Run Reproducibility Between-run reproducibility was evaluated by using the high and low positive controls included with the COBAS 182 Am J Clin Pathol 2000;114:180-187 American Society of Clinical Pathologists

Clinical Chemistry / ORIGINAL ARTICLE 7 Observed 6 5 4 2 2 4 5 6 7 Standard Deviation 0.5 0.4 0. 0.2 0.1 0.0 2.5.5 4 4.5 5 5.5 6 6.5 Expected Figure 1 Observed (y) vs expected (x) hepatitis C virus (HCV) RNA log 10 (copies per milliliter) for a dilution series measured in the COBAS AMPLICOR HCV MONITOR Test, version 2.0 (Roche Diagnostics, Indianapolis, IN). The equation for the linear regression line is y = 0.85 (± 0.016)x + 0.594 (± 0.080); S y/x = 0.148; n = 5; r 2 = 0.982; r = 0.991. The diagonal dashed line is the line of unity. The upper and lower horizontal dashed lines are 1 million and 1,000 copies per milliliter, respectively. Figure 2 Within-run precision profile. The SD of triplicate determinations vs the mean hepatitis C virus (HCV) RNA log 10 (copies per milliliter). The solid trend line was generated by using a second-order polynomial curve fit. The horizontal dashed line is an SD of 0.150. The vertical dashed line is the limit of detection of the assay,.0 log 10 (copies per milliliter). AMPLICOR HCV MONITOR, version 2.0 kits. Reproducibility was measured on 1 lot of reagents over 27 days with different operators and different instruments. The total number of determinations made was 70 for the high control and 72 for the low control. The mean values for the high and low controls were 150,000 HCV RNA copies per milliliter (5.1 log 10 [copies per milliliter]) and 8,000 HCV RNA copies per milliliter (.9 log 10 [copies per milliliter]), respectively. The percentage of the coefficient of variation of the copies per milliliter for the high control was 66%, and the SD of the log 10 (copies per milliliter) was 0.19. For the low control, the percentage of the coefficient of variation of the copies per milliliter was %, and the SD of the log 10 (copies per milliliter) was 0.15. A Levey-Jennings plot of the data is presented in Figure. Comparison With Manual AMPLICOR HCV MONITOR Test, Version 1.0 A total of 146 samples were measured in singlet in the manual version 1.0 and in COBAS version 2.0 AMPLICOR HCV MONITOR tests. Of the samples, 92 had levels of HCV RNA that quantified in both assays. A Deming regression analysis was performed on the data Figure 4. The overall correlation of the 2 assays was good; however, the values obtained in the COBAS version 2.0 assay were on average about 0.654 log 10, or almost 5-fold higher than in the manual version 1.0. This variation might have been due to differences in the ability of the 2 assays to detect the various HCV genotypes. To study the possible effect of genotype on HCV RNA levels between the methods, a number of the more discrepant samples were chosen for genotype determination. Genotype Determination of Discrepant Samples Twenty samples that differed in the COBAS version 2.0 from the manual version 1.0 by at least 0.50 log 10 (copies per milliliter), or a -fold difference in copies per milliliter, were analyzed to determine the HCV genotype. The genotypes for the samples tested are given in Table 1. The genotype distribution in this population was as follows: 1a, 4 of 20 (20%); 1b, of 20 (15%); 2a, 2 of 20 (10%); a, 10 of 20 (50%); and 4a, 1 of 20 (5%). Comparison With QUANTIPLEX HCV RNA 2.0 bdna Test Forty-five samples were measured in both the COBAS HCV MONITOR Test version 2.0 and the QUANTIPLEX HCV 2.0 bdna Test. Twenty-six samples that quantified in both assays were compared by Deming regression analysis Figure 5. American Society of Clinical Pathologists Am J Clin Pathol 2000;114:180-187 18

Erali et al / EVALUATION OF QUANTITATIVE HEPATITIS C VIRUS ASSAY 7 6 5 4 0 6 12 18 24 0 6 42 48 54 60 66 72 Cobas version 2.0 6 5 4 Run Number Figure A Levey-Jennings plot of the high control (closed circle) and low control (open circle) for the COBAS AMPLICOR HCV MONITOR Test, version 2.0 (Roche Diagnostics, Indianapolis, IN). For the high control, n = 70 measured over 27 days by operators on instruments. For the low control, n = 72 measured over 27 days by operators on instruments. The mean log 10 (copies per milliliter) for the high and low controls is 5.1 and.9, respectively. The dashed error bars are 2 SD units. 4 5 Manual version 1.0 6 7 Figure 4 Hepatitis C virus (HCV) RNA log 10 (copies per milliliter) for 92 clinical specimens as determined by the COBAS version 2.0 (y) and the manual version 1.0 (x) AMPLICOR HCV MONITOR tests (Roche Diagnostics, Indianapolis, IN). The equation for the Deming regression line (solid line) is y = 1.004 (±0.04)x + 0.654 (±0.22); S y/x D = 0.6; r 2 = 0.846; r = 0.920. The diagonal dashed line is the line of unity. Table 1 Hepatitis C Virus Genotypes Determined for 20 Samples * HCV RNA (copies/ml) Genotype Manual Version 1.0 COBAS Version 2.0 Fold Difference 1a 5,200 4,000 6 590,000 2,200,000 4 1,600 12,000 7 7,000 67,000 10 1b 800,000 2,600,000 2,700 1,000 5,600 100,000 28 2a 92,000 720,000 8 20,000,200,000 10 a 2,500 52,000 21 60,000,700,000 10 68,000 70,000 11 64,000 470,000 7 94,000 870,000 9 58,000 60,000 6 1,900 240,000 128 5,400 99,000 18 22,000 240,000 11 26,000 180,000 7 4a 170,000 2,400,000 14 * The samples chosen for genotyping quantified greater than -fold higher, or 0.50 log 10 (copies per milliliter), in the COBAS version 2.0 (AMPLICOR HCV MONITOR Test, version 2.0, Roche Diagnostics, Indianapolis, IN) than in the manual version 1.0 (AMPLICOR HCV MONITOR Test, Roche Diagnostics). 184 Am J Clin Pathol 2000;114:180-187 American Society of Clinical Pathologists

Clinical Chemistry / ORIGINAL ARTICLE Quantiplex HCV bdna HCV RNA log 10 (Eq/mL) 7 6 5 4 Discussion 4 5 6 7 Cobas Amplicor HCV Monitor, v2.0 Figure 5 Hepatitis C virus (HCV) RNA log 10 (copies per milliliter) for 26 clinical specimens as determined by the QUANTIPLEX HCV RNA 2.0 bdna Test (y) (Chiron, Emeryville, CA) and the COBAS AMPLICOR HCV MONITOR Test, version 2.0 (x) (Roche Diagnostics, Indianapolis, IN). The equation for the Deming regression line (solid line) is y = 0.94 (±0.10)x + 0.47 (±0.717); S y/x D = 0.194; r 2 = 0.600; r = 0.774. The diagonal dashed line is the line of unity. The vertical dashed line is the upper limit of detection of the COBAS assay, 1 million HCV RNA copies per milliliter. The horizontal dashed line is the lower limit of detection of the QUANTIPLEX assay, 200,000 HCV RNA equivalents per milliliter. Compared with quantitative RNA testing for HIV-1, the indications for quantitative HCV measurement remain less certain, and recommendations for testing are poorly standardized. Nevertheless, quantitative serum HCV RNA levels are used widely to estimate disease prognosis, for monitoring therapy, and as a general indicator of clinical outcome. The association of high HCV levels with poor clinical outcome does not warrant testing for exclusion of patients from therapy, but recent data suggest an important role for HCV viral load testing in determining the appropriate length of treatment. Although the use of serial HCV quantification for therapeutic monitoring has been discussed extensively in the literature, especially for interferon monotherapy, 8,9 consensus is lacking. An alternative approach to serial quantitative monitoring is to test the patient at the end of therapy with a sensitive qualitative HCV assay. As more studies on current therapies are completed and as the next generation of more effective HCV antiviral drugs becomes available, more clear-cut indications for quantitative HCV monitoring will hopefully emerge. In addition to these difficulties, the clinician should understand the performance attributes of the various HCV assays referenced in current clinical studies and how they compare with one another. Our evaluation of the COBAS AMPLICOR HCV MONITOR Test, version 2.0, included an assessment of the reproducibility, sensitivity, and dynamic range of the assay, as well as comparison of the assay with the first-generation AMPLICOR HCV MONITOR Test and the QUANTIPLEX HCV bdna Test. In addition, we performed sequence-based genotyping on samples most discrepant between the version 1.0 and 2.0 AMPLICOR assays as an indication of the ability of the new assay to better detect the various genotypes. The COBAS AMPLICOR HCV MONITOR Test, version 2.0, performed reproducibly and reliably. The within-run precision and between-run precision were acceptable with SDs for log 10 (copies per milliliter) of less than 0.150, or a -fold variation in copies per milliliter. The linearity studies supported a reportable range of 1,000 to 1 million HCV RNA copies per milliliter. The reproducibility of the assay was good across the -log dynamic range; however, greater imprecision was evident at the low end. Above 1 million HCV RNA copies per milliliter, the values generated by the assay were reproducible but not reliable, and dilutions are necessary to obtain accurate results. Comparison of the COBAS version 2.0 assay with the manual version 1.0 assay showed generally improved detection of all samples in the COBAS version 2.0 assay. This improvement was reflected as an average 5-fold increase in HCV RNA levels when measured in the COBAS version 2.0 assay vs the manual version 1.0 assay. Since large changes in viral level are expected for patients with HCV undergoing therapy, a 5-fold difference in values between the 2 assay versions should not mislead clinicians. However, for serial values obtained for 1 patient with the 2 versions, care should be taken in interpreting results when the HCV RNA levels do not decrease dramatically. When the genotypes of samples with large viral load increases in the COBAS version 2.0 assay were determined, it was evident that the ability of the test to detect various genotypes is much improved over the version 1.0 assay. A number of earlier studies suggested that HCV quantitation was higher in infections with genotype 1b and lower in infections with type 2a or 2b. 10,11 It is now evident that the method used to measure the levels of HCV RNA greatly influenced the apparent HCV level. The first-generation Roche assay for HCV, version 1.0, has been shown to underestimate viral loads in HCV infections with genotypes 2 and 4,6,12 compared with other methods. In the present study, 65% of the discrepant samples genotyped were types 2,, or 4. Since these types constitute only 2% of the clinical samples reported by our laboratory, these results indicate that American Society of Clinical Pathologists Am J Clin Pathol 2000;114:180-187 185

Erali et al / EVALUATION OF QUANTITATIVE HEPATITIS C VIRUS ASSAY detection of these genotypes is much improved in the version 2.0 assay. This conclusion also is supported by a recent study that extensively evaluated the ability of the AMPLICOR version 2.0 assay to detect various HCV genotypes. 1 The incomplete detection of various genotypes in the version 1.0 assay may have been related to the efficiency of PCR amplification in the 5 untranslated region of the HCV genome. While this is considered a fairly well-conserved region of the genome, sequence variations among the different genotypes exist. The efficiency of PCR amplification may be affected by these sequence variations. The addition of dimethyl sulfoxide by Roche to the PCR reagent mixture in the version 2.0 AMPLICOR HCV MONITOR Test seems to have allowed improved and equivalent amplification of the various genotypes. The ability of the AMPLICOR HCV MONITOR Test, version 2.0, to more accurately detect genotypes will contribute to a better understanding of the importance of genotype on viral load and its relationship to disease and treatment outcome. The COBAS AMPLICOR HCV MONITOR Test, version 2.0, and the Chiron QUANTIPLEX HCV RNA 2.0 bdna Test generated fairly comparable results over the narrow range evaluated. Since the QUANTIPLEX assay has a lower limit of detection of 200,000 Eq/mL and the AMPLICOR assay has an upper limit of detection of 1 million copies per milliliter, the range that could be evaluated was rather narrow. However, these data suggest that modifications made to the COBAS AMPLICOR HCV MONITOR Test, version 2.0, reduce the differences in quantitation previously observed between the 2 assay platforms. Improvements in the COBAS AMPLICOR HCV MONITOR Test have resulted in an average 5-fold increase in the quantitative result for a given clinical sample and in more uniform detection of different viral types. These changes underscore the need for standardized calibration of all qualitative and quantitative HCV molecular assays. Recent studies examining the outcomes of patients undergoing combination interferon alfa and ribavirin treatment have relied on HCV measurements using a multicycle RT-PCR method that uses gel electrophoresis and Southern blot end detection. 14 These studies ascribe a clinical significance to HCV RNA levels that are greater than or less than 2 million copies per milliliter. Lack of standardized calibration of HCV RNA quantitative units makes comparison of clinical studies difficult and limits the interpretation of clinical hypotheses generated by different study groups. Nonstandardization also creates confusion for clinicians who use a different HCV quantitative assay system to monitor their patients when following clinical recommendations from the literature. An important question is whether the 2 million copies per milliliter reported in the ribavirin-interferon trials are equivalent to the copies per milliliter reported in other assays, including the widely used Roche test. In addition, it is also important to ask whether it is appropriate to assign 2 million copies per milliliter as a special cutoff with unique clinical significance or whether there may be a gradual gradient of outcomes associated with viral titer. The accuracy of different assay systems in the broad ranges above and below the 2-million-unit cutoff must be evaluated. An international calibration standard for HCV RNA is being established to provide reference material that could be used for the standardization of various HCV RNA assays. 15 Standardization of quantitative and qualitative molecular infectious disease assays should be a high priority for developers of commercial tests and for those developing tests using in-house reagents and formats. Reporting in standardized units with the benefit of validation studies that examine test performance for the widely used test assays is a necessary precedent for appropriate understanding and interpretation of results generated using different methods. From the 1 ARUP Institute for Clinical and Experimental Pathology, Salt Lake City, UT, and the 2 Department of Pathology, University of Utah Health Sciences Center, Salt Lake City. Supported in part by Roche Molecular Systems, Pleasanton, CA. Address reprint requests to Ms Erali: ARUP Institute for Clinical and Experimental Pathology, 500 Chipeta Way, Salt Lake City, UT 84108. Acknowledgments: We thank Stewart Wood for technical assistance. References 1. Simmonds P, McOmish F, Yap PL, et al. Sequence variability in the 5 non-coding region of hepatitis C virus: identification of a new virus type and restrictions on sequence diversity. J Gen Virol. 199;74:661-668. 2. Simmonds P, Mellor J, Sakuldamrongpanich T, et al. Evolutionary analysis of variants of hepatitis C virus found in South-East Asia: comparison with classifications based upon sequence similarity. J Gen Virol. 1996;77:01-024.. National Institutes of Health Consensus Development Conference Panel statement: management of hepatitis C. Hepatology. 1997;26(suppl 1):2S-10S. 4. Hawkins A, Davidson F, Simmonds P. Comparison of plasma virus loads among individuals infected with hepatitis C virus (HCV) genotypes 1, 2, and by Quantiplex HCV RNA assay versions 1 and 2, Roche Monitor assay, and an in-house limiting dilution method. J Clin Microbiol. 1997;5:187-192. 5. Jacob S, Baudy D, Jones E, et al. Comparison of quantitative HCV RNA assays in chronic hepatitis C. Am J Clin Pathol. 1997;107:62-67. 6. Lunel F, Cresta P, Vitour D, et al. Comparative evaluation of hepatitis C virus RNA quantitation by branched DNA, NASBA, and monitor assays. Hepatology. 1999;29:528-55. 7. Trabaud M-A, Bailly F, Si-Ahmed SN, et al. Comparison of HCV RNA assays for the detection and quantification of hepatitis C virus RNA levels in serum of patients with chronic hepatitis C treated with interferon. J Med Virol. 1997;52:105-112. 186 Am J Clin Pathol 2000;114:180-187 American Society of Clinical Pathologists

Clinical Chemistry / ORIGINAL ARTICLE 8. McHutchison JG, Gordon SC, Schiff ER, et al. Interferon alfa-2b alone or in combination with ribavirin as initial treatment for chronic hepatitis C. N Engl J Med. 1998;9:1485-1492. 9. Davis GL, Lau JYN. Factors predictive of a beneficial response to therapy of hepatitis C. Hepatology. 1997;26(suppl 1):122S-127S. 10. Mahaney K, Tedeschi V, Maertens G, et al. Genotypic analysis of hepatitis C virus in American patients. Hepatology. 1994;20:1405-1411. 11. Orito E, Mizokami M, Nakano T, et al. Serum hepatitis C virus RNA level as a predictor of subsequent response to interferon-alpha therapy in Japanese patients with chronic hepatitis C. J Med Virol. 1994;44:410-414. 12. Tong CYW, Hollingsworth RC, Williams H, et al. Effect of genotypes on the quantification of hepatitis C virus (HCV) RNA in clinical samples using the Amplicor HCV Monitor Test and the Quantiplex HCV RNA 2.0 Assay (bdna). J Med Virol. 1998;55:191-196. 1. Mellor J, Hawkins A, Simmonds P. Genotype dependence of hepatitis C virus load measurement in commercially available quantitative assays. J Clin Microbiol. 1999;7:2525-252. 14. Tong MJ, Blatt LM, McHutchison JG, et al. Prediction of response during interferon alfa 2b therapy in chronic hepatitis C patients using viral and biochemical characteristics: a comparison. Hepatology. 1997;26:1640-1645. 15. Saldanha J, Lelie N, Heath A, and the WHO Collaborative Study Group. Establishment of the first international standard for nucleic acid amplification technology (NAT) assays for HCV RNA. Vox Sang. 1999;76:149-158. American Society of Clinical Pathologists Am J Clin Pathol 2000;114:180-187 187