Standardization of Hepatitis C Virus RNA Quantification

Transcription

1 Standardization of Hepatitis C Virus RNA Quantification JEAN-MICHEL PAWLOTSKY, 1,2 MAGALI BOUVIER-ALIAS, 1 CHRISTOPHE HEZODE, 3 FRANCOISE DARTHUY, 1 JOCELYNE REMIRE, 1 AND DANIEL DHUMEAUX 2,3 It was recently recommended that hepatitis C virus (HCV) RNA quantification be used to tailor the duration of combined interferon alfa (IFN- )/ribavirin therapy in patients infected by HCV genotypes 1, 4, and 5. This recommendation has been difficult to implement in the absence of standardized quantitative units for HCV RNA. The aim of this work was to define clinically relevant HCV RNA loads in standardized international units (IU), for use in routine clinical and research applications based on standardized quantitative assays. Two hepatitis C virus RNA quantitative assays were used: (1) the Superquant assay (National Genetics Institute, Los Angeles, CA), for which possibly relevant thresholds were established; and (2) the semi-automated Cobas Amplicor HCV Monitor assay version 2.0 (Cobas v2.0, Roche Molecular Systems, Pleasanton, CA) that measures HCV RNA loads in IU/mL. Quantification in the Cobas v2.0 assay was linear over the entire range of values tested, including viral loads higher than 850,000 IU/mL after 100-fold dilution. The accuracy and precision of the measures in IU/mL were satisfactory with Cobas v2.0. The results obtained with Superquant and Cobas v2.0 correlated (r.932; P <.0001). A value of 2,000,000 copies/ml (6.3 log 10 copies/ml) with Superquant was converted to nearly 800,000 IU/mL (5.9 log 10 IU/mL). In conclusion, all HCV RNA quantitative assays should give HCV RNA loads in international units and be validated with appropriate calibrated panels; 800,000 IU/mL in any of these assays should be used as the decision threshold to tailor the IFN- /ribavirin treatment duration in patients infected by HCV genotypes 1, 4, and 5. (HEPATOLOGY 2000;32: ) Advances in molecular biology based techniques have recently made it possible to detect and quantify viral genomes in human body fluids. This has helped to understand the pathogenesis of viral infections and to manage both untreated patients and patients receiving antiviral therapy, especially those Abbreviations: HCV, hepatitis C virus; RT-PCR, reverse transcription polymerase chain reaction; IFN-, interferon alfa; MU, million units; Eq, genome equivalent; WHO, World Health Organization; IU, international units. From the 1 Department of Bacteriology and Virology, 2 INSERM U99, and the 3 Department of Hepatology and Gastroenterology, Hôpital Henri Mondor, Université Paris XII, Créteil, France. Received May 5, 2000; accepted June 29, Address reprint requests to: Professor Jean-Michel Pawlotsky, M.D., Ph.D., Service de Bactériologie-Virologie, Hôpital Henri Mondor, 51 avenue du Maréchal de Lattre de Tassigny, Créteil, France. jean-michel.pawlotsky@hmn.ap-hop-paris.fr; fax: (33) Copyright 2000 by the American Association for the Study of Liver Diseases /00/ $3.00/0 doi: /jhep infected by human immunodeficiency virus, hepatitis B virus, and hepatitis C virus (HCV). 1-3 Initially, in-house reverse transcription polymerase chain reaction (RT-PCR) amplification-based techniques were used to quantify HCV RNA, but they lacked standardization, reproducibility, and accuracy. 4 Subsequent development of standardized commercial assays for HCV RNA quantification, namely the noncompetitive PCR-based Amplicor HCV Monitor technique (Roche Molecular Systems, Pleasanton, CA) and the branched-dna based signal amplification technique (Quantiplex HCV RNA; Bayer Diagnostics, Emeryville, CA), enabled HCV RNA to be quantified in any laboratory equipped for molecular biology based methods. The performance of these assays has recently been reviewed. 4 In recent years the use of HCV RNA quantification in clinical practice has been hindered for the following reasons: (1) HCV RNA load at a given time has no predictive value for the severity of HCV-induced liver injury and cannot be used as a prognostic indicator of the disease 5,6 ; (2) nonquantitative, qualitative HCV RNA detection assays remain more sensitive than quantitative assays and are thus routinely used to assess the virologic response of chronic hepatitis C to specific treatment 7 ; and (3) although pretreatment HCV RNA load is an independent predictor of sustained virologic responses to interferon alfa (IFN- ) treatment, 8-10 only one approved therapeutic strategy (IFN-, 3 million units [MU] 3 times per week for 6 to 12 months) has been available, hampering the use of pretreatment viral load to tailor treatment schedules. The situation has changed recently with the publication of 2 studies involving more than 1,700 treatment-naive patients with chronic hepatitis C and showing a significant benefit of combination therapy with IFN-, 3 MU 3 times per week subcutaneously, and ribavirin, 1,000 to 1,200 mg/d orally, compared with the same dose of IFN- alone. 11,12 In these reports, both pretreatment viral load and the HCV genotype were independent predictors of sustained HCV RNA clearance after combination therapy. 11,12 In addition, the rate of sustained virologic responses was identical after 24 weeks and 48 weeks of the IFN- /ribavirin combination, both in patients infected by genotypes 2 and 3 (whatever the pretreatment viral load) and in genotype 1 infected patients with low pretreatment viral loads (about 65% and 35%, respectively). 11 In contrast, genotype 1 infected patients with high pretreatment viral loads were more likely to be sustained virologic responders after 48 weeks than after 24 weeks of combined treatment (28% vs. 8%, respectively). 11 These results led to the following recommendations at the recent International Consensus Conference on Hepatitis C: naive patients infected by HCV genotypes 2 and 3, whatever their pretreatment viral load, and patients infected by HCV genotype 1 (and genotypes 4 and 5 in the absence of adequate data on these genotypes) with low

2 HEPATOLOGY Vol. 32, No. 3, 2000 PAWLOTSKY ET AL. 655 pretreatment viral loads should receive combination therapy for 24 weeks only, whereas naive patients infected by HCV genotypes 1, 4, and 5 with high pretreatment viral loads should be treated for 48 weeks, provided they are HCV-RNA negative at week 24 by qualitative PCR with a detection limit of the order of 100 copies/ml. 13 The cut-off point defining high and low pretreatment viral loads was established as the HCV RNA load discriminating best between sustained virologic responders and nonresponders in the 2 studies. 11,12 Both teams used the Superquant assay (National Genetics Institute, Los Angeles, CA), an inhouse competitive PCR-based technique using multiple PCR cycles performed in a central laboratory. 14 From these data the decision threshold was established at 2,000,000 copies/ml (i.e., 6.3 log 10 copies/ml). 11,12 More recently, the same authors reanalyzed their database and used the study group median viral load (i.e., 3,500,000 copies/ml or 6.5 log 10 copies/ml) as the decision threshold. 15 Unfortunately, these recommendations have been inapplicable on a global level, because most laboratories worldwide use commercial assays such as Amplicor HCV Monitor and Quantiplex HCV RNA. Indeed, in the absence of standardized HCV RNA quantification units, 1 copy/ml in Monitor, 1 genome equivalent (Eq)/mL in Quantiplex, and 1 copy/ml in Superquant do not represent the same amount of HCV RNA in a clinical sample, these units having been defined independently, with quantified standards of different natures, lengths, and sequences. In addition, the same units can also vary from one version of a commercial assay to the next. 16 Thus, 2,000,000 or 3,500,000 copies/ml in Superquant may not be 2,000,000 or 3,500,000 copies or Eq/mL in assays other than Superquant, meaning that these assays cannot be used to tailor the duration of combined treatment unless the exact correspondence between the various units has been established. 17 To permit universal standardization of HCV RNA quantification units, a recent collaborative study established the World Health Organization (WHO) International Standard for HCV RNA quantification. 18 A lyophilized genotype 1 sample was accepted as the candidate standard and was assigned a titer of 10 5 international units (IU)/mL. The International Standard consists of a batch of vials, each vial containing 50,000 IU in lyophilized form. 17 This standard has been used to calibrate HCV RNA concentration panels, and by industrial manufacturers to express HCV RNA loads in IU/mL in their assays. Roche Molecular Systems recently released the first semi-automated HCV RNA quantification assay giving results in IU/mL, the Cobas Amplicor HCV Monitor assay version 2.0 (Cobas v2.0). According to the manufacturer, this assay has a dynamic range of quantification between 600 IU/mL and 850,000 IU/mL. In the present study, we evaluated the performance of the Cobas v2.0 assay in terms of linearity and accuracy of quantification, and used it to calculate HCV RNA thresholds in IU/mL for tailoring the duration of IFN- /ribavirin therapy. The aim was to allow any standardized assay using IU/mL to be used routinely in clinical practice and to allow comparisons between results obtained with different assays. MATERIALS AND METHODS Materials. We used a panel of HCV-positive plasmas containing increasing concentrations of genotype 1b HCV RNA and calibrated to the WHO HCV RNA standard, the Nucleic Acid Panel HCV-RNA (NAP HCV-RNA; AcroMetrix, Berkeley, CA). In addition, 44 patients with chronic hepatitis C and HCV genotype 1b infection were enrolled in a clinical trial aimed at studying early viral replication dynamics during treatment with IFN-, 3 MU 3 times per week or daily, given alone or in combination with ribavirin, 1,000 to 1,200 mg/d orally (Pawlotsky JM, et al., manuscript in preparation). Blood was taken before treatment, every 4 hours for 12 hours, every 6 hours for 36 hours, and every 12 hours for 48 hours, then at day 14, week 4, week 12, and week 24. The pretreatment samples from the 44 patients were used in this study. To cover a wide range of HCV RNA titers, all the available samples taken during therapy from the 20 of 44 patients who responded to therapy (viral load decrease of more than 1 log 10 copies/ml in the Superquant assay) were also studied. A total of 348 blood samples were studied. Design of the Study. To determine the linear range of quantification, the accuracy of quantification in international units, and the standardization of the Cobas v2.0 assay, each concentration of the NAP HCV-RNA was tested 4 times in 4 different experiments with this assay, and the means and standard errors of the means were calculated. The Cobas v2.0 assay was then used together with Superquant to quantify HCV RNA in the 348 blood samples to study the relationship between the 2 assays and to establish the conversion equation from Superquant copies to international units; the equation was then used to convert the clinically relevant thresholds established with Superquant into international units. Finally, the NAP HCV-RNA was tested 4 times in 4 different experiments with the current version of the branched DNA-based assay (Quantiplex HCV RNA version 2.0) to assess the linearity of quantification in this assay and to establish the correspondence between Quantiplex genome equivalents and international units. Methods. All sera were prepared, aliquoted, and stored at 80 C within 3 hours of sampling. Never-thawed aliquots were used for the experiments. All samples were tested with both Superquant and Cobas v2.0. Both data sets were log-transformed for further analysis. HCV RNA was quantified with the Superquant assay as previously described. 14 Briefly, HCV RNA was extracted from serum and a complementary DNA template was produced. The complementary DNA was then amplified with primers from the 5 noncoding region of the HCV genome in four separate PCR reactions with 25, 30, 35, and 45 cycles. The HCV RNA detection limit in this assay was 100 copies/ml (2.0 log 10 copies/ml). HCV RNA was quantified with the Cobas v2.0 assay according to the manufacturer s instructions. HCV RNA was extracted, and both reverse transcription and amplification were performed with primers from the 5 noncoding region of the HCV genome. The quantitation standard and sample DNAs were detected by hybridization to a biotin-labeled probe. The extraction step was manual, whereas reverse transcription, amplification, amplicon detection, and calculation of international units were done automatically by the Cobas Amplicor device (Roche Molecular Systems). The HCV RNA detection limit in this assay is 600 IU/mL (2.8 log 10 IU/mL). As recommended by the manufacturer, samples with values exceeding 850,000 IU/mL were considered to be above the linear range of the assay and were retested after 100-fold dilution. Only the latter result was taken into account. The lot number of the kits used in this study was A HCV RNA was quantified with Quantiplex HCV RNA version 2.0 according to the manufacturer s instructions. The HCV RNA detection limit in this assay is 200,000 Eq/mL (5.3 log 10 Eq/mL). Statistical Analysis. The relationship between the results of the 2 assays was tested by linear regression analysis, yielding the conversion equation. RESULTS The NAP HCV-RNA was tested 4 times in 4 different experiments with the Cobas v2.0 assay, and the means and standard errors were calculated. As recommended by the manufacturer, the sample containing 2,000,000 IU/mL (6.3 log 10 IU/ ml) of the standard was tested neat and after 1/100 dilution in the Cobas v2.0 assay. As shown in Fig. 1, the Cobas v2.0 assay

3 656 PAWLOTSKY ET AL. HEPATOLOGY September 2000 FIG. 2. Relationship between Cobas v2.0 and Superquant HCV RNA values in the 348 blood samples. FIG. 1. Cobas v2.0 quantification of a panel of samples containing increasing HCV RNA concentrations calibrated with the WHO HCV RNA standard (NAP HCV-RNA). Each concentration was tested 4 times in 4 different experiments. The means 1 standard deviation are presented. was linear over the entire range of values tested. The accuracy of the measures in IU/mL relative to the panel was excellent between 50,000 IU/mL (4.7 log 10 IU/mL) and 2,000,000 IU/mL (6.3 log 10 IU/mL), and satisfactory below 50,000 IU/ ml. Table 1 shows the coefficients of variation and the standard deviations observed with each panel concentration. Both were satisfactory to ensure that the assay has sufficient precision to ensure that 0.5 log 10 (i.e., 3-fold) differences in viral load are significant. As shown in Fig. 2, the results obtained in the Superquant and Cobas v2.0 assays correlated (r.932; P.0001). Linear regression analysis yielded the following conversion equation (for HCV genotype 1): log 10 IU/mL (Cobas v2.0) log 10 copies/ml (Superquant) This equation was used to convert the Superquant thresholds forwarded for tailoring the treatment duration (Table 2). 11,12 A serum HCV RNA level of 2,000,000 copies/ml (6.3 log 10 copies/ml) in the Superquant assay was equivalent to 831,763 IU/mL, i.e., approximately 800,000 IU/mL (5.9 log 10 IU/mL). A serum HCV RNA level of 3,500,000 copies/ml (6.5 log 10 copies/ml) in the Superquant assay was equivalent to 1,318,257 IU/mL, i.e., approximately 1,300,000 IU/mL (6.1 log 10 IU/mL). To avoid errors related to possible nonlinear quantification in the Superquant assay at the upper and lower ends of the scale, linear regression analysis was performed after removing values for samples with RNA titers below 4.5 log 10 copies/ml (31,622 copies/ml) or higher than 6.5 log 10 copies/ml (3,162,280 copies/ml) in Superquant. Although the resulting equations were different, conversion always yielded the same IU/mL thresholds (data not shown). Table 3 and Fig. 3 show the distribution of the Superquant HCV RNA titers in samples with values more and less than 800,000 IU/mL (5.9 log 10 IU/mL), and more and less than 1,300,000 IU/mL (6.1 log 10 IU/mL), respectively. Interestingly, 800,000 IU/mL and 1,300,000 IU/mL in Cobas v2.0 classified the samples roughly identically as being above or below the respective corresponding thresholds in Superquant, a result not surprising because the difference between these 2 thresholds is very small (0.2 log 10 IU/mL, whereas viral load differences of less than 0.5 log 10 IU/mL are not usually taken into account because they may be caused by intrinsic variability of quantitative assays). A few samples were classified differently by the 2 assays (Table 3 and Fig. 3), a finding again likely caused by intrinsic variability of quantitative assays in samples with viral loads close to the thresholds. Among the 44 pretreatment samples tested, i.e., samples representative of those requiring HCV RNA testing before antiviral treatment, 5 samples (11%) with values below 800,000 IU/mL in Cobas v2.0 had values above 2,000,000 copies/ml in Superquant; 3 samples (7%) with values below 800,000 IU/mL in Cobas v2.0 had values above TABLE 1. Coefficients of Variation and Log 10 Standard Deviations Obtained After Testing a Panel Containing Increasing HCV RNA Concentrations Calibrated With the WHO HCV RNA Standard (NAP HCV-RNA) NAP HCV-RNA in IU/mL (log 10 IU/mL) 500 (2.7) 5,000 (3.7) 50,000 (4.7) 200,000 (5.3) 500,000 (5.7) 2,000,000 (6.3) Neat 2,000,000 (6.3) Diluted CV 100.7% 64.2% 27.6% 10.7% 20.0% 29.1% 18.3% log 10 SD NOTE. Each concentration was tested 4 times in 4 different experiments with the Cobas v2.0 assay. Coefficients of variation (CV) were calculated from the nontransformed data and standard deviations (SD) were calculated for the log 10 -transformed data.

4 HEPATOLOGY Vol. 32, No. 3, 2000 PAWLOTSKY ET AL. 657 TABLE 2. HCV RNA Thresholds Possibly Relevant to Tailoring the Duration of Combined Interferon/Ribavirin Therapy, in Superquant Copies/mL and in IU/mL Clinically Relevant Thresholds in Superquant Copies/mL (log 10 Copies/mL) Clinically Relevant Thresholds in IU/mL (log 10 IU/mL) 2,000,000 (6.3) 800,000 (5.9) 3,500,000 (6.5) 1,300,000 (6.1) 3,500,000 copies/ml in Superquant; and 4 samples (9%) with values below 1,300,000 IU/mL in Cobas v2.0 had values above 3,500,000 copies/ml in Superquant. This suggests that values close to the decision threshold should be considered cautiously when making clinical decisions, because they may be slightly under- or overestimated by any HCV RNA quantitative assay. The NAP HCV-RNA was tested 4 times in 4 different experiments with the Quantiplex HCV RNA 2.0 assay. Quantiplex version 2.0 was linear between 200,000 (5.3 log 10 IU/mL) and 2,000,000 IU/mL (6.3 log 10 IU/mL) (data not shown). The sample containing 50,000 IU/mL (4.7 log 10 IU/mL) appeared to be close to the detection limit of the assay, because it was weakly positive in only 1 of the 4 runs. The samples containing less than 50,000 IU/mL were negative in this assay. Linear regression analysis yielded the following conversion equation for the expression of Quantiplex HCV RNA 2.0 results in IU/mL in the corresponding interval (r 1.000; P.02): log 10 IU/mL log 10 Eq/mL Using this equation, 5.9 log 10 IU/mL (800,000 IU/mL) is equivalent to 6.7 log 10 Eq/mL or 5,110,870 Eq/mL, which approximates to 5,000,000 Eq/mL in Quantiplex version 2.0, and 6.1 log 10 IU/mL (1,300,000 IU/mL) is equivalent to 6.8 log 10 Eq/mL or 6,309,573 Eq/mL, which approximates to 6,300,000 Eq/mL in Quantiplex version 2.0. DISCUSSION Efforts to standardize HCV RNA quantification techniques started early in their development, 19,20 but the WHO initiative is the first attempt to standardize HCV RNA quantification units, 18 a mandatory step for reliable routine use of HCV RNA quantification in clinical studies and in patient management. Our results with the NAP HCV-RNA, a panel of HCV genotype 1 standards calibrated with the WHO standard, show that HCV RNA quantification in international units with the Cobas v2.0 assay is both linear over a wide range of values and accurate. The use of international units to express HCV RNA loads thus appears feasible and, in our opinion, should be extended to all HCV RNA quantification assays. Thus, although correspondence equations can be deduced from panels tested with various assays using their own quantitative FIG. 3. Ability of a threshold of 800,000 IU/mL (5.9 log 10 IU/mL) (A) and of a threshold of 1,300,000 IU/mL (6.1 log 10 IU/mL) (B) in Cobas v2.0 to discriminate among samples with more or less than 2,000,000 copies/ml (6.3 log 10 copies/ml) (solid line), and more or less than 3,500,000 copies/ml (6.5 log 10 copies/ml) (dashed line) in Superquant, respectively. units (as we did with the Quantiplex assay version 2.0), nonstandardized units must now be replaced by international units. Based on convincing clinical evidence, 11,12 the International Consensus Conference on Hepatitis C recommended in 1999 that 2,000,000 copies/ml (6.3 log 10 copies/ml) in the Superquant assay be used as a decision threshold to tailor the duration of combined interferon/ribavirin therapy in patients infected by HCV genotypes 1, 4, and This recommendation has not been easy to apply, because of the lack of standardization of HCV RNA quantitative units among different assays and between successive versions of the same assay. 4,16 TABLE 3. Ability of Thresholds of 800,000 IU/mL (5.9 log 10 IU/mL) and 1,300,000 IU/mL (6.1 log 10 IU/mL) in Cobas v2.0 to Discriminate Among Samples With More or Less Than 2,000,000 Copies/mL (6.3 log 10 Copies/mL) or 3,500,000 Copies/mL (6.5 log 10 Copies/mL) in Superquant, Respectively SuperQuant <2,000,000 Copies/mL >2,000,000 Copies/mL <3,500,000 Copies/mL >3,500,000 Copies/mL Cobas v ,000 IU/mL (n 283) 250 (88%) 33 (12%) 267 (94%) 16 (6%) Cobas v ,000 IU/mL (n 65) 11 (17%) 54 (83%) 19 (29%) 46 (71%) Cobas v2.0 1,300,000 IU/mL (n 292) NC NC 275 (94%) 17 (6%) Cobas v2.0 1,300,000 IU/mL (n 56) NC NC 12 (21%) 44 (79%) NC: not calculated.

5 658 PAWLOTSKY ET AL. HEPATOLOGY September 2000 We show here that the corresponding threshold of 2,000,000 copies/ml in international units is 800,000 IU/mL (5.9 log 10 IU/mL). This threshold can therefore be used to tailor the treatment duration with any assay quantifying HCV RNA loads in international units, provided that the assay has been correctly calibrated and validated with appropriate standardized panels. As shown in Fig. 3, several samples with values below 800,000 IU/mL in Cobas v2.0 contained more than 2,000,000 copies/ml in Superquant, whereas several samples with values above 800,000 IU/mL contained less than 2,000,000 copies/ml in Superquant. This was not surprising: neither assay can be considered as a gold standard, and both have intrinsic variability related to their respective precision, reproducibility, and accuracy. 4 Nevertheless, the Cobas v2.0 assay was accurate within the range of values covering the 800,000- IU/mL threshold (Fig. 1), suggesting that this value can be used with confidence in this assay. When making therapeutic decisions, however, one should bear in mind that HCV RNA loads may be slightly over- or underestimated in any HCV RNA quantification assay, raising the issue of misclassification of HCV RNA loads close to the decision threshold. After reanalyzing their initial database, the investigators of the 2 studies on IFN- /ribavirin treatment of naive patients with chronic hepatitis C recently suggested to use the study group median viral load (i.e., 3,500,000 copies/ml in Superquant) to predict sustained virologic responses instead of 2,000,000 copies/ml. 15 We show here that the corresponding value in international units is 1,300,000 IU/mL (6.1 log 10 IU/ ml). It should be stressed, however, that the difference between these 2 thresholds is very small (0.2 log 10 IU/mL). Viral load changes of less than 0.5 log 10 IU/mL (i.e., 3-fold) during patient follow-up are not usually taken into account because they may be caused by intrinsic variability of quantitative assays. 4 This is supported by data in Table 3 and Fig. 3, showing that 800,000 IU/mL and 1,300,000 IU/mL in Cobas v2.0 classified the samples roughly identically as being above or below the respective corresponding thresholds in Superquant. Based on these data, we therefore recommend that 800,000 IU/mL be used as the decision threshold in any standardized assay using international units. Thus, according to the recommendation of the International Consensus Conference on Hepatitis C, 13 patients infected by HCV genotypes 1, 4, or 5 with viral load values below 800,000 IU/mL before treatment should be treated for 24 weeks only. In these patients, other nonvirologic parameters such as gender, age, and the fibrosis score on liver biopsy may also point to a need for prolonged therapy if HCV RNA is undetectable at week 24 of therapy. 15 In contrast, patients infected by HCV genotypes 1, 4, or 5 with more than 800,000 IU/mL before treatment should always be treated for 48 weeks, provided that they are HCV RNA negative at week 24 of therapy. Standardization of HCV RNA quantification units opens the way to the widespread use of HCV RNA quantification as a treatment guide in chronic hepatitis C. Two potential applications appear particularly promising: (1) HCV RNA load can be used to tailor treatment schedules when it is a predictor of a sustained virologic response to therapy. It must be stressed that the 800,000 IU/mL threshold can only be used for tailoring the duration of combination therapy with IFN-,3MU3 times per week, and ribavirin, 1,000 to 1,200 mg/d. The relevant decision thresholds will have to be determined prospectively for all new therapeutic schedules using these and/or other drugs. It is therefore particularly important that future therapeutic trials be designed not only to identify predictors of sustained viral clearance, but especially to establish their practical value in individual treatment tailoring to improve the results of HCV therapy. (2) HCV RNA quantification can also be used to monitor early viral replication dynamics. This recently threw light on the mechanisms underlying the antiviral actions of drugs such as IFN- and ribavirin (Pawlotsky JM, et al., manuscript in preparation). 21,22 Early viral dynamics also appear to be strongly predictive of the response to therapy and might help to tailor treatment to the true antiviral effect during therapy (Neumann AU, et al., manuscript in preparation). In conclusion, our results suggest that all HCV RNA quantitative assays should now give HCV RNA loads in IU/mL and be validated with appropriate calibrated panels. They also suggest that 800,000 IU/mL should be used as the decision threshold to tailor the duration of the currently approved treatment schedule with IFN- plus ribavirin in patients infected with HCV genotypes 1, 4, and 5. Acknowledgment: The authors are grateful to Dianne Young, Karen Gutekunst, and Beverly Dale (Roche Molecular Systems, Pleasanton, CA), Paul D. Neuwald (AcroMetrix, Berkeley, CA), and Françoise Huisse (Bayer Diagnostics, Puteaux, France) for kindly providing the reagents used in this study. REFERENCES 1. Mellors JW, Munoz A, Giorgi JV, Margolick JB, Tassoni CJ, Gupta P, Kingsley LA, et al. Plasma viral load and CD4 lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 1997;126: Coffin JM. HIV population dynamics in vivo: implications for genetic variation, pathogenesis, and therapy. Science 1995;267: Pawlotsky JM. Hepatitis C: viral markers and quasispecies. In: Liang TJ, Hoofnagle JH, eds. Hepatitis C. New York: Academic Press 2000 (in press). 4. Pawlotsky JM. Measuring hepatitis C viremia in clinical samples: can we trust the assays? HEPATOLOGY 1997;26: Puoti C, Magrini A, Stati T, Rigato P, Montagnese F, Rossi P, Aldegheri L, et al. Clinical, histological, and virological features of hepatitis C virus carriers with persistently normal or abnormal alanine transaminase levels. HEPATOLOGY 1997;26: Romeo R, Colombo M, Rumi M, Soffredini R, Del Ninno E, Donato MF, Russo A, et al. Lack of association between type of hepatitis C virus, serum load and severity of liver disease. J Viral Hepat 1996;3: Pawlotsky JM. Diagnostic tests for hepatitis C. J Hepatol 1999;31(Suppl 1): Martinot-Peignoux M, Marcellin P, Pouteau M, Castelnau C, Boyer N, Poliquin M, Degott C, et al. Pretreatment HCV RNA levels and HCV genotype are the main and independent prognostic factors of sustained response to alpha interferon therapy in chronic hepatitis C. HEPATOLOGY 1995;22: Nousbaum JB, Pol S, Nalpas B, Landais P, Berthelot P, Bréchot C. Hepatitis C virus type 1b (II) infection in France and Italy. Ann Intern Med 1995;122: Pawlotsky JM, Roudot-Thoraval F, Bastie A, Darthuy F, Rémiré J, Métreau JM, Zafrani ES, et al. Factors affecting treatment responses to interferon- in chronic hepatitis C. J Infect Dis 1996;174: Poynard T, Marcellin P, Lee SS, Niederau C, Minuk GS, Ideo G, Bain V, et al. Randomised trial of interferon alpha-2b plus ribavirin for 48 weeks or for 24 weeks versus interferon alpha-2b plus placebo for 48 weeks for treatment of chronic infection with hepatitis C virus. Lancet 1998;352: McHutchison JG, Gordon SC, Schiff ER, Shiffman ML, Lee WM, Rustgi VK, Goodman ZD, et al. Interferon alfa-2b alone or in combination with ribavirin as initial treatment for chronic hepatitis C. N Engl J Med 1998; 339: EASL International Consensus Conference on Hepatitis C. Consensus Statement. J Hepatol 1999;30: Tong MJ, Hwang SJ, Lefkowitch M, Lee SD, Co RL, Conrad A, Schmid P, et al. Correlation of serum HCV RNA and alanine aminotransferase levels in chronic hepatitis C patients during treatment with ribavirin. J Gastroenterol Hepatol 1994;9:

6 HEPATOLOGY Vol. 32, No. 3, 2000 PAWLOTSKY ET AL Poynard T, McHutchison J, Goodman Z, Ling MH, Albrecht J. Is an à la carte combination interferon alfa-2b plus ribavirin regimen possible for the first line treatment in patients with chronic hepatitis C? HEPATOLOGY 2000;31: Martinot-Peignoux M, Boyer N, Le Breton V, Le Guludec G, Castelnau C, Akremi R, Marcellin P. A new step toward standardization of serum HCV RNA quantification in patients with chronic hepatitis C. HEPATOLOGY 2000;31: Fang JWS, Albrecht JK, Jacobs S, Lau JYN. Quantification of serum hepatitis C virus RNA. HEPATOLOGY 1999;29: Saldanha J, Lelie N, Heath A. Establishment of the first international standard for nucleic acid amplification technology (NAT) assays for HCV RNA. Vox Sang 1999;76: Zaaijer HL, Cuypers HT, Reesink HW, Winkel IN, Gerken G, Lelie PN. Reliability of polymerase chain reaction for detection of hepatitis C virus. Lancet 1993;341: French Study Group for the Standardization of Hepatitis C Virus PCR. Improvement of hepatitis C virus RNA polymerase chain reaction through a multicentre quality control study. J Virol Methods 1994;49: Lam NP, Neumann AU, Gretch DR, Wiley TE, Perelson AS, Layden TJ. Dose-dependent acute clearance of hepatitis C virus genotype 1 virus with interferon alfa. HEPATOLOGY 1997;26: Neumann AU, Lam NP, Dahari H, Gretch DR, Wiley TE, Layden TJ, Perelson AS. Hepatitis C viral dynamics in vivo and the antiviral efficacy of interferon- therapy. Science 1998;282: