Indicators and predictors of response to anti-viral therapy in chronic hepatitis C

Authors


Dr S. S. Lee, Liver Unit, Department of Medicine, University of Calgary, 3330 Hospital Drive NW, Calgary, Alta., T2N 4N1, Canada. E-mail: samlee@ucalgary.ca

Summary

The complications of chronic hepatitis C, including cirrhosis and hepatocellular carcinoma, are expected to increase dramatically world-wide over the next 10–20 years. Immunomodulatory/anti-viral therapy, employing interferon alfa both alone and in combination with ribavirin, affords the only effective treatment for hepatitis C. Accurate early prediction of response to interferon therapy may decrease or eliminate unnecessary or ineffective treatment, permit greater flexibility in tailoring therapy on an individual basis, and enhance the cost-effectiveness of treatment.

Liver biopsy provides valuable information about the baseline severity and subsequent progression of hepatitis C. Severe fibrosis or cirrhosis on the pre-treatment liver biopsy is associated with decreased response rates. The measurement of viral RNA levels and genotyping may be used to optimize individual patient treatment. Genotype non-1 and a low viral load are the most significant pre-treatment indicators of sustained virological response.

The most reliable predictor of a poor virological response is continued seropositivity for viral RNA during therapy. Therefore, a decision to stop or continue treatment can be based on a positive viral RNA test at 12 weeks for interferon-naive patients receiving interferon or pegylated interferon therapy.

 ‘The physician will manage the cure best who has foreseen what is to happen from the present state of matters…’ Hippocrates, On the Prognostics

Introduction

There are currently about 170 million people world-wide with chronic hepatitis C virus (HCV) infection, with about three million people infected in the USA and five million in Western Europe.1–4 Up to 20% of affected individuals will develop cirrhosis within 10–30 years, 1–4% of whom will develop hepatocellular carcinoma each year.2, 3 It has been projected that from 2010 to 2019, in the USA, 165 900 individuals will die from chronic liver disease, and 27 200 from hepatocellular carcinoma.4 Although the incidence of new HCV infections is falling, the enormous number of individuals with established HCV infection will generate increasing demand for health care support over the coming decades.

The accurate assessment of those patients most likely to respond to therapy will become increasingly important. Clinicians need to be able to predict treatment success and to determine the ideal duration of therapy for each patient. This article reviews the methods for the assessment of treatment response and the indicators for the prediction of therapeutic outcomes to interferon alfa (IFN-alfa)-based therapy.

Standardization of hepatitis c assessment

Virological and histological assessments are central to the diagnosis and management of hepatitis C. Although serological assays for initial screening and diagnosis have been standardized, variability in standards and results remains problematic amongst the various quantitative HCV RNA assay systems. In addition, the use of at least four different systems for measuring liver biopsy pathology precludes direct comparisons of histological findings between studies. The move to standardization has been slow, but will be achieved with the use of one preferred histological grading system, and the adoption of the newly established World Health Organization international standards in quantitative HCV RNA assays.

Serological assays

The main serological test in use today is the third-generation enzyme immunoassay (also termed enzyme-linked immunosorbent assay), which detects antibodies developed against several structural and non-structural antigenic determinants of HCV.5–7 This assay detects approximately 98% of infected individuals and is used as initial screening in immunocompetent patients and healthy blood donors; in immunosuppressed subjects, it can be used in conjunction with HCV RNA detection to reduce the false-negative rate.3

Molecular assays

Molecular assays characterize HCV and its genotypes by hybridization of test probes to specific RNA sequences.8 Qualitative detection of HCV RNA is based on reverse-transcription polymerase chain reaction methods. Several different tests have been developed, with considerable variability in results due to several factors (sample handling and storage conditions, different primers and reaction methods, and detection efficiency).9–12 Operator experience is important, as laboratories often show marked differences in accuracy,13 suggesting a need for simple, standardized testing. Tests in kit form may provide the solution.

The various HCV genotypes can be differentiated using screening tests that detect individual mutations in the genome. Confirmatory tests then evaluate larger gene segments. Screening tests include restriction fragment length polymorphism analysis and reverse dot blot hybridization assays,14, 15 whereas the most accurate confirmatory tests are nucleotide sequencing and phylogenetic analysis of the E1 or NS5B genes.11, 16, 17

Quantitative tests of viral load are crucial for predicting the response to IFN and guiding the duration of therapy.18 Quantitative tests use two methods: target amplification methods (based on quantitative polymerase chain reaction) and signal amplification methods (such as the branched DNA assay). However, a direct comparison between assay systems is not feasible as they use different standards.19 Therefore, the new World Health Organization-sponsored international standard for HCV RNA assays will see all values reported as IU/mL.20

Liver biopsy

There are several different biopsy grading systems, but, regardless of which system is used, quantitative methods have two important advantages over descriptive definitions. First, they reduce variability in data interpretation. Second, they permit longitudinal analysis in individual patients, and the comparison of histological responses between patients.

The most widely used tissue pathology system is the Knodell histological activity index (HAI).21 This system allocates a numerical severity score to each of four sets of disease features: piecemeal, periportal, bridging and multi-lobular necrosis; intra-lobular degeneration and focal necrosis; inflammatory cell density in the portal tracts; and the degree of fibrosis and cirrhosis. The sum of these HAI scores falls within a range of possible values from zero (no detectable abnormalities) to 22 (marked multi-lobular necrosis with densely packed inflammatory cells and cirrhosis). Criticisms of this system are that the combined score does not separate the necro-inflammatory activity from the fibrotic architectural changes indicative of liver disease progression, and that piecemeal and bridging necrosis are included in the same category.22–24 In addition, considerable inter- and intra-observer subjective variability is possible.22

Therefore, alternative scoring systems have been devised. The Scheuer classification system assigns scores of zero (absent) to four (severe) in three separate categories: periportal/portal activity; lobular activity; and fibrosis/cirrhosis/architectural disruption.23 The Ishak system represents a modification of the Knodell HAI into: necro-inflammatory grading of four areas with scores totalling 18 and a separate staging of fibrosis up to six points for cirrhosis.24 Another alternative, the METAVIR system, was devised specifically for hepatitis C and consists of two categories: portal and lobular necro-inflammatory activity, scored from zero to three; and fibrosis/cirrhosis, scored from zero to four. The METAVIR system is widely used in Europe.25, 26

The comparison of data using these systems is made difficult because of their different scoring criteria, and their ability to predict long-term treatment outcomes remains undetermined. However, these problems do not preclude the evaluation of individual patients' disease status or histological outcomes of therapy within any single study. Inter-study comparisons of histological responses will require acceptance of a single grading system.

Indicators of therapeutic response

IFN-alfa administered as a single anti-viral agent, or in combination with ribavirin, is the mainstay of anti-viral treatment for chronic hepatitis C. At present, combination IFN-alfa with ribavirin is available in almost all countries, and newer drugs produced by protein pegylation of IFN, e.g. the linear 12-kDa peginterferon alfa-2b (Pegintron) and the branched 40-kDa peginterferon alfa-2a (Pegasys), have been developed. Recent studies have established these pegylated IFNs combined with ribavirin as the gold standard of treatment, with overall sustained virological response (SVR) rates of 54–61% reported.27–29 These figures represent a mixture of patients with HCV genotype 1 (42–51% response rates) and genotypes 2 and 3 (76–82% response rates).27–29 Unfortunately, because of the high cost or delays in approval/distribution, the pegylated IFNs are not universally available, and, in some countries, unmodified IFN-based treatment is still used. Therefore, the following sections review both standard and pegylated IFNs.

In clinical trials, the efficacy of these agents has been assessed using biochemical, virological and histological indicators of hepatic response. Although the optimal tool for predicting long-term improvement in hepatic function has yet to be identified, these indicators have provided a basis for evaluating the efficacy of IFN-based therapies.

Biochemical response indicator

Early IFN studies defined efficacy as the normalization of serum alanine aminotransferase (ALT) levels at 6 months after the end of treatment.2, 3, 30–32 Although longer intervals have been reported, biochemical or virological relapses generally occur within 6 months of completing therapy.3, 30–32 Generally, the rapid normalization of ALT is a good indicator of response to therapy, but this is not a robust predictor of sustained virological or histological response.2, 3, 30, 33 In some patient populations, such as patients with the non-1 HCV genotype or those with low baseline serum HCV RNA levels, ALT normalization and accompanying virological responses have been increased by extending therapy for an additional 12 months.34, 35

Virological response indicator

Following the discovery and genetic characterization of HCV,36, 37 techniques for the detection and quantification of serum HCV RNA became available. Accordingly, virological response became the principal end-point in the clinical evaluation of IFN therapy. An end-of-treatment virological response is defined as undetectable serum HCV RNA at the end of the treatment period. SVR is defined as undetectable HCV RNA 6 months after stopping treatment.2, 3, 38

Histological response indicator

Recent American,3, 39 European,2 Canadian40 and British41 hepatitis C consensus guidelines all recommend pre-treatment liver biopsies. Liver biopsy remains the standard criterion for assessing hepatic improvement in chronic hepatitis C.42–44 In general, a histological response to IFN is defined as improvement in the degree of inflammation or fibrosis between baseline and post-treatment biopsies. Most studies have used a Knodell HAI score decrease of at least two points to define a histological response.21, 42 It is important to note, however, that the degree of histological improvement that ultimately predicts a clinically significant long-term improvement in liver function has yet to be established. The ongoing Hepatitis C Anti-Viral Long-Term Treatment Against Cirrhosis trial may help to establish histological response parameters for predicting long-term outcome.45

Correlation between indicators of response

Although biochemical, virological and histological indicators are used to define improvement in liver function, a strong positive correlation between them is not present. Although a virological response to IFN is usually accompanied by ALT normalization, a biochemical response is not always associated with a virological response (discordance). End-of-treatment response assessments indicate that many patients with normal ALT levels remain viraemic, and are at risk of ongoing liver damage or later relapse.3, 46 Similarly, biochemical response does not correlate strongly with histological improvement. Patients with normal ALT levels may have histological abnormalities, whilst a small percentage of patients who lack virological or biochemical responses may still show some histological benefit from treatment.47 It is not clear whether long-lasting histological improvement can be achieved in these latter patients.47–49 Nevertheless, it has been suggested that some patients who do not achieve biochemical and virological responses may have a decreased incidence of serious liver disease (including hepatocellular carcinoma) after re-treatment with IFN.50 The utility of end-of-treatment ALT and HCV RNA levels as surrogate markers of liver disease activity in individual cases is therefore somewhat limited.3, 42

SVR is the most reliable serological predictor of hepatic improvement in patients treated with anti-viral therapy.47, 51–54 In a re-analysis of data from three pivotal IFN/ribavirin trials,38, 55, 56 Poynard et al. examined fibrosis scores in patients with significant baseline fibrosis (METAVIR F2/F3/F4), comparing those with and without SVR.57 These investigators found a significantly greater reduction in fibrosis scores in patients with SVR. Up to 76% of patients without SVR (non-responders and relapsers) exhibited worsening histological scores at follow-up, compared to less than 10% of patients with SVR. In addition, of those patients without SVR who exhibited histological improvement, there was no difference in improvement between those treated with IFN alone vs. IFN/ribavirin. This probably reflects the known anti-inflammatory effects of IFN, independent of its anti-viral effect.58

Studies conducted with pegylated IFN have also demonstrated a correlation between SVR and an improvement in histological score. A comparative study of IFN-alfa-2b and peginterferon alfa-2b found that patients achieving SVR had at least a two-point decrease in HAI score.59 Conversely, patients who relapsed after a virological response at the end of treatment and those without SVR had, on average, a decrease in HAI score of less than one point from baseline.59 In a study comparing low- and high-dose peginterferon alfa-2b (0.5 µg/kg or 1.0 µg/kg) plus ribavirin, all patients exhibited some evidence of histological improvement from baseline, but not greater than a two-point decrease in HAI score.27 In patients with HCV cirrhosis who were treated with peginterferon alfa-2a (180 µg or 90 µg) or IFN-alfa-2a, and attained SVR, the corresponding histological responses (more than a two-point decrease in HAI score) were 88%, 100% and 80%, respectively.60 Interestingly, histological improvements were also observed in a substantial number of patients without SVR (33%, 26% and 35%, respectively), probably reflecting the aforementioned anti-inflammatory effect of IFN. Thus, biochemical and virological parameters may not adequately predict histological improvement during and after treatment, especially in virological non-responders or relapsers. This is particularly the case for cirrhotic patients, in whom IFN therapy may decrease fibrosis without concomitant virological responses.

Pre-treatment predictors of response

The impact of a wide range of viral and patient factors on response has been evaluated to determine whether a successful outcome can be predicted before initiating IFN therapy. Overall, the most consistent pre-treatment predictors of SVR to IFN-based therapy are genotype and viral load. HCV genotypes 2 and 3 are the most susceptible to IFN therapy. Conversely, genotype 1 is the most difficult to eradicate, even with the addition of ribavirin (Table 1).38, 55, 56, 61–66

Table 1.  Sustained responses to interferon therapy according to hepatitis C virus genotype
ReferenceRegimen and treatment durationGenotype and percentage of patients achieving sustained virological response
1Non-1
  1. IFN, standard interferon; MIU, million international units; tiw, three times a week.

McHutchison et al.55IFN + ribavirin × 48 weeks2866
IFN + ribavirin × 24 weeks2669
IFN × 48 weeks729
IFN × 24 weeks216
Shiratori et al.61IFN 6 MIU tiw × 6 months1553 (type 2)
IFN 9 MIU tiw × 6 months2976 (type 2)
Martinot-Peignoux et al.623 or 5 MIU tiw × 3–12 months970
Weiland et al.633 or 5 MIU × 24–60 weeks13 (1b)34 (non-1b)
Bell et al.653 MIU tiw × 6 months5 (1a)15 (2 and 3a)
6 MIU tiw × 6 months2 (1a)29 (2 and 3a)

A low baseline serum viral load (< 2 million copies/mL or < 800 000 IU/mL) is associated with a significantly higher probability of achieving SVR following IFN-based therapy.38, 55, 56, 61–66 SVR rates with IFN/ribavirin were increased 1.5-fold after 24 weeks of treatment in patients with < 2 million copies/mL, compared with patients with higher viral loads.38

Although genotype non-1 and viral load appear to be the strongest prognostic indicators of SVR to IFN-based therapy, other regimen- and patient-specific factors have also been identified. For standard IFN or IFN/ribavirin therapy, the absence of cirrhosis, mild or absent portal fibrosis, age < 40 years, shorter disease duration, low hepatic iron concentration, low HCV genome heterogeneity (quasi-species) and mutations in the NS5A gene of HCV have been reported to be associated with a higher probability of SVR.38, 55–57, 62, 65–72 Weaker associations have also been reported for gender (women responding better than men),38, 73, 74 levels of liver enzymes, such as aspartate aminotransferase and gamma-glutamyl transpeptidase,67 and body weight.75

Similarly, non-1 genotype and a low viral load are the most consistent factors associated with SVR to both peginterferon alfa-2b76 and peginterferon alfa-2a60, 77, 78 monotherapies. With peginterferon alfa-2b monotherapy, no other factors were found to be significantly associated with SVR, whereas, with peginterferon alfa-2a monotherapy, other significant associations included the absence of cirrhosis/bridging fibrosis, ALT levels greater than three times the upper limit of normal, body weight < 85 kg, body surface area< 2 m2, age < 40 years and an HAI score of > 10 (Table 2).60, 77, 78

Table 2.  Pre-treatment predictive factors for peginterferon alfa-2a (40 kDa) 78
Predictive factorOdds ratio (95% confidence interval) P value
  1. ALT, alanine aminotransferase; HAI, Knodell histological activity index; ULN, upper limit of normal range.

Genotype other than 14.11 (2.90–5.86)0.0001
Serum viral load < 2 million  copies/mL3.25 (2.30–4.64)0.0001
Absence of cirrhosis or  bridging fibrosis3.34 (2.07–5.46)0.0001
Serum ALT > 3 × ULN2.33 (1.62–3.36)0.0001
Body weight < 85 kg1.63 (1.12–2.37)0.0106
Age < 40 years1.52 (1.07–2.17)0.0196
HAI score > 101.58 (1.02–2.47)0.0416

Interestingly, with peginterferon alfa-2a therapy, baseline necro-inflammatory markers (HAI score of > 10, ALT greater than three times the upper limit of normal) have been found to be of prognostic value, a situation analogous to that for IFN therapy of hepatitis B virus infection. In contrast, IFN/ribavirin therapy appears to be equally effective in those patients with slightly elevated ALT (between normal and 1.3 times the upper limit of normal),79 and even normal ALT,80 compared to those with significantly elevated ALT levels.

Genotype and baseline viral load are also associated with response to combination peginterferon alfa-2b + ribavirin therapy.27 After 24 weeks of treatment at two different dosages of peginterferon alfa-2b (1.5 µg/kg or 1.5 µg/kg followed by 0.5 µg/kg) plus ribavirin, genotype 2/3 patients exhibited an increase in SVR of approximately two-fold compared with genotype 1 patients, and patients with a viral load of < 2 million copies/mL showed a response 1.5–2-fold better than cases with a high viral load.27 Logistic regression analysis demonstrated five baseline variables associated with SVR: genotype non-1, low viral load, lighter body weight, younger age and absence of bridging fibrosis/cirrhosis.27

The recently reported results of a peginterferon alfa-2a + ribavirin study offer some interesting differences.28 Multivariate analysis of the variables associated with SVR in that study found only three significant associations: genotype non-1 (odds ratio, 3.25), age < 40 years (odds ratio, 2.60) and body weight < 75 kg (odds ratio, 1.91).28 With this form of pegylated IFN + ribavirin, viral load and the presence of significant fibrosis/cirrhosis are not important factors associated with virological response. The reasons for this disparity in results between the two types of pegylated IFN combined with ribavirin remain unclear at present.

Despite the identification of significant prognostic indicators for IFN-based therapies, their application does not accurately predict SVR in all patients. Although they provide a basis for discussion about possible therapeutic outcomes, predicting the response of an individual patient before initiating IFN therapy remains difficult. Withholding IFN treatment because of one or more adverse pre-treatment indicators may exclude from treatment patients who would, in fact, respond to therapy.66 Conversely, prolonging treatment for patients with fewer than four favourable prognostic factors may be unnecessary, because these patients may respond to shorter (24 week) treatment periods.1 Thus, more accurate pre-treatment predictors of response to IFN-based therapies are still required.

Predictors of response after initiation of therapy

Although biochemical and virological indicators provide a means to assess the end-of-treatment response in the management of chronic hepatitis C, the potential of these indicators to predict outcomes earlier in treatment is still under investigation. Given the possibility for re-emergence of HCV following inappropriate discontinuation of IFN therapy, early response indicators must be valid measures of actual outcomes. The National Institutes of Health (NIH) treatment guidelines from 1997 recommended that IFN therapy should be stopped if ALT normalization was not observed after 12 weeks of treatment.39 However, this recommendation is now clearly outdated: studies have shown that the use of 12-week ALT levels to predict the therapeutic response to IFN will result in the discontinuation of therapy in approximately 20% of patients who would have achieved SVR if treatment had been continued.81

Since the issuance of the 1997 NIH guidelines, several studies have demonstrated that HCV RNA assays provide a more reliable predictor of SVR than do ALT levels (Table 3).67, 81–83 Unlike serum ALT measurements, HCV RNA analysis is not prone to wide fluctuations in values obtained at 4, 8 or 12 weeks of therapy.81, 82 Thus, HCV RNA status, even at 4 weeks, is a more accurate predictor of SVR than ALT levels at 4, 8 or 12 weeks.

Table 3.  Reliability of hepatitis C virus (HCV) RNA and alanine aminotransferase (ALT) tests, performed at various time points during treatment, in predicting sustained response to interferon (IFN) therapy
ReferenceFactor measuredTime point% of patients
HCV RNAALT
  1. NSR, no sustained response (non-responder or relapser); SVR, sustained virological response.

Gavier et al.67Predictive value for SVR 4 weeks50.041.0
12 weeks43.834.1
 Predictive value for NSR 4 weeks95.378.3
12 weeks93.077.8
Tong et al.83Predictive value for NSR12 weeks98.092.0
Misidentification of eventual  responders as non-responders  from 12-week test6.733.0
Brouwer et al.81Predictive value for SVR 6 months' treatment 4 weeks36.031.0
 8 weeks25.0
12 weeks26.0
12 months' treatment 4 weeks31.025.0
 8 weeks21.0
12 weeks21.0
Predictive value for NSR 6 months' treatment 4 weeks99.096.0
 8 weeks98.0
12 weeks97.0
12 months' treatment 4 weeks97.091.0
 8 weeks91.0
12 weeks91.0
Odds ratio for correct identification 6 months' treatment 4 weeks46.311.5
 8 weeks18.3
12 weeks13.610.4
12 months' treatment 4 weeks3.4
 8 weeks2.9
12 weeks2.7

Although the normalization of ALT levels after 8 weeks is a significant prognostic indicator of response to IFN/ribavirin therapy in IFN-naive patients, studies show that the HCV RNA status after 24 weeks of treatment is a more reliable predictor of SVR with this therapeutic regimen.2 Persistence of viraemia after 24 weeks of IFN/ribavirin therapy is the strongest predictor of non-response, correctly identifying 99% of patients who will not attain SVR.73, 84 Therefore, a minimum 24-week treatment regimen is recommended for all patients receiving IFN/ribavirin therapy.73

Although it is apparent that viral clearance (i.e. undetectable HCV RNA) is the most reliable initial indicator of sustained biochemical and virological response, just how early an accurate prediction for virological response can be made is still not settled. Civeira and Prieto, reviewing 18 studies involving 988 patients, concluded that undetectable levels of HCV RNA after 4 weeks of standard IFN monotherapy correlate with a subsequent SVR in about 50% of patients (positive predictive value), whereas the presence of HCV RNA at 4 weeks correctly predicts failure to achieve SVR (negative predictive value) in over 97% of patients.82 The 12-week HCV RNA values have a similar negative predictive value but a slightly less positive predictive value, and so these authors concluded that the optimum time point to check the viral response is at week 4.82 Some of these contentions were supported by a prospective study in which 71% of IFN-treated patients exhibiting early (4 weeks) clearance of HCV RNA went on to achieve SVR.85 Early viral clearance at 4 weeks may also be predictive of treatment outcome with IFN/ribavirin therapy: one study showed a positive predictive value of 46% with undetectable HCV RNA at that time point.81 Similar trends have been observed in recent studies examining the predictive value of virological responses measured at 2 and 4 weeks of IFN/ribavirin therapy on the achievement of SVR in non-responders or relapsers to previous IFN monotherapy,86, 87 and in those with genotype 1 infection.85 Despite the potential usefulness of these very early time points, the 12-week responses proved to have better predictability values in the above studies.85–87

Similar to unmodified IFN, the likely response of an individual patient to peginterferon alfa-2a (40 kDa) is indicated by the HCV RNA level. An early viral response, defined as the undetectability of HCV RNA or > 2-log drop in HCV RNA after 12 weeks of therapy, was the most accurate indicator of SVR,78 when compared with corresponding values obtained at 4, 8 or 24 weeks. Specifically, although the positive predictive values of such HCV RNA changes ranged between 46 and 77%, the negative predictive value was highest (98%) at the 12-week time point.78 In clinical practice, this means that the decision to continue or stop treatment can be made at the 12-week time period. In other words, a patient with an early viral response has a reasonable chance of attaining SVR, and therefore should continue treatment, whereas a patient without such a viral response by week 12 has only a 2% chance of achieving SVR, and thus should stop treatment.

The 12-week predictability with combination pegylated IFN + ribavirin therapy is also excellent. The available evidence indicates that the negative predictive values are optimum at 12 weeks (approximately 97–99%) for both peginterferon alfa-2a + ribavirin28 and peginterferon alfa-2b + ribavirin (Davis GL, personal communication, 2002). Thus, as with pegylated IFN monotherapy, the decision to stop or continue therapy for pegylated IFN + ribavirin combination treatment can be made at 12 weeks. This strategy has recently been recommended by the NIH 2002 Consensus Development Conference.3

Recent studies with peginterferon alfa-2a + ribavirin have further refined treatment strategies. First, only patients with genotype 1 require ‘full’ treatment with 48 weeks of peginterferon alfa-2a and high-dose ribavirin (1000 or 1200 mg daily); genotype 2/3 patients require only 24 weeks of peginterferon and low-dose ribavirin (800 mg daily).29 Second, as 97% of patients with genotypes 2/3 show an early viral response, it is probably more efficient to simply treat all genotype 2/3 patients for 24 weeks without checking HCV RNA at week 12. In genotype 1 patients, the negative predictive value of 98% of a 12-week early viral response argues strongly for stopping treatment at this time in those who fail to achieve such an early response. This differential strategy for genotype 1 vs. genotype 2/3 patients has recently been recommended by French hepatitis C consensus guidelines.88

The rate of viral decline in serum may offer another means to predict therapeutic outcome with IFN therapy. In a preliminary study comparing standard IFN alfa-2a with peginterferon alfa-2a, a rapid viral response (minimum 2-log decline in viral titre in the initial 4 weeks of therapy, followed by an additional minimum 2-log decline in viral titre until week 8) was the most predictive indicator of SVR (odds ratio, 38.7), independent of any other baseline factor except viral load.89 In patients achieving this rapid viral response, subsequent viral breakthrough was less common, and SVR more frequent, with peginterferon alfa-2a than with IFN alfa-2a (68% vs. 51%; P < 0.001). The percentage of patients who did not exhibit a rapid viral decline, but later attained SVR, was very low for both treatments (5% with peginterferon alfa-2a and 1% with IFN alfa-2a).

It may even be possible to use the kinetics of early viral clearance within the first few days to predict treatment outcomes. HCV RNA concentrations in patients who attain SVR have been shown to decrease quite rapidly over the first 2 days of standard IFN therapy, followed by a slower rate of decline,89 a second-order rate function. Based on this kinetic profile, Magalini et al. have found that failure to achieve an 85% reduction in HCV RNA level after only 3 days of IFN therapy accurately predicts the persistence of viraemia at 4 weeks, and hence a reduced chance of SVR in this sub-group (24%) of patients.90 Whether a similar predictability of such early responses applies to pegylated IFNs, with or without ribavirin, requires further study.

Conclusions

IFN-based therapy remains the only effective treatment for chronic hepatitis C. Virological indicators of response are pre-eminent, and standardization of the HCV RNA assays and reporting will be immensely useful in this regard. Several pre-treatment and on-treatment predictors of therapeutic responses to IFN have been identified, helping to individually tailor therapy at an early stage. Together with early viral clearance, a non-1 genotype and low pre-treatment viral load appear to be the best predictors of sustained biochemical and virological response to IFN-based therapy. The assessment of fibrosis by liver biopsy is also important, because the presence of severe fibrosis or cirrhosis correlates with a lower response rate.

The development of novel pegylated IFNs, such as peginterferon alfa-2b and peginterferon alfa-2a (40 kDa), has re-defined the management of hepatitis C. These new drugs produce greater sustained biochemical and virological responses than their unpegylated counterparts, and may be effectively combined with ribavirin to increase SVR rates. However, pre-treatment and on-treatment predictability factors for virological response to the pegylated IFNs appear to be generally similar to those of unmodified IFN-alfa.

Acknowledgement

Dr Lee is supported by a Senior Scholarship Award from the Alberta Heritage Foundation for Medical Research.

Ancillary