1.1 THE NEED FOR GUIDELINES

Hepatitis C virus (HCV) is a major health care concern in the UK affecting some 200 000 to 400 000 individuals. The majority of these patients will have chronic HCV infection and many will develop chronic liver disease with the risk of developing cirrhosis and hepatocellular carcinoma (HCC). Successful treatment will arrest the progression of liver disease and so prevent the serious complications of chronic HCV infection. In addition, treatment will reduce the numbers of HCV infected individuals.

The relatively high cost of treatment enforces the need for a systematic approach for this condition so that resources are used most effectively. The development of clinical guidelines is important, as these will assist purchasing authorities, providers, clinicians, primary care groups, and patients in making decisions about appropriate treatment.

1.2 THE DEVELOPMENT OF GUIDELINES: RIGOUR OF DEVELOPMENT/SYSTEMATIC CRITICAL REVIEW OF THE LITERATURE

2.1 EPIDEMIOLOGY

2.2 NATURAL HISTORY OF HEPATITIS C VIRUS INFECTION

Subclinical HCV infection is the rule with only 10% of patients reporting an acute illness associated with jaundice. HCV infection rarely causes fulminant hepatitis45 ,46 but severe acute HCV infections have been reported in liver transplant recipients,19 patients with underlying chronic liver disease, and in patients coinfected with hepatitis B virus (HBV).47 Although the acute illness is usually mild, a high proportion of patients progress to chronic liver disease.8 In a study of 135 patients with PTH, 77% developed chronic disease and of the 65 patients with sequential liver biopsies, 32% had developed cirrhosis after a mean follow up of 7.5 years.16 Only 1% of these patients had histological remission with the remaining patients having chronic active (CAH) or chronic persistent hepatitis (CPH). However, in the same year Seeffet al published a long term follow up study of patients with post transfusion NANB hepatitis.48 A total of 568 patients with PTH and two control groups of 526 and 458 patients who had received transfusions without developing hepatitis were studied. After an average follow up of 18 years, mortality related to liver disease was 3.3% in PTH cases compared with 1.5% in the control groups and the majority of deaths occurred in patients with associated alcoholism. It appears therefore that most patients who develop progressive disease do so slowly.

In 1995 Tong et al published a study of 131 PTH cases referred to a centre between 1980 and 1994.49 A total of 101 patients underwent liver biopsy a mean of 22 years post transfusion. Twenty seven (20.6%) had chronic hepatitis, 30 (22.9%) had CAH, 67 (51.1%) had cirrhosis, and seven (5.3%) had HCC after mean time intervals from transfusion of 14, 18, 20, and 28 years, respectively. During the follow up period, 20 (15.3%) patients died, 19 (95%) from complications of cirrhosis or the development of HCC. Thus persistent post transfusion HCV infection does lead to progressive liver disease and in some patients death from related liver failure or the development of HCC although long term follow up studies are required to assess the contribution of HCV to morbidity and mortality. However, most studies have been conducted at referral centres, reflecting the severe end of the disease spectrum, so that the true numbers of patients with non-progressive or mildly progressive liver disease are unknown.

Viral factors associated with more rapidly progressive disease include high level viraemia,50 genotype 1 (especially 1b),51 and the degree of viral genetic diversity (quasispecies).52 ,53 Route of transmission may be important as patients infected via blood transfusion tend to have more histologically active liver disease.27 Other host factors such as immune deficiency,54 excess alcohol,55 ,56 and coinfection with HBV47 and HIV57 may also influence the rate of disease progression.

There is a variable rate of fibrosis progression with a median time from infection to cirrhosis of approximately 30 years (range 13–42 years).58 Independent factors associated with an increased rate of fibrosis progression include age at infection greater than 40 years, daily consumption of 50 g or more of alcohol, and male sex. There was no association between fibrosis progression and genotype.

In HCV associated compensated cirrhotics, five year survival is over 90% and 10 year survival 80%.59 A five year follow up showed that the risk of developing HCC was 7% (1.4% per year) and 18% decompensated. After decompensation, prognosis is poor with 50% survival at five years.

2.3 CLINICAL SPECTRUM OF DISEASE

Infection with the hepatitis C virus results in a variety of hepatic and extrahepatic diseases. In a minority of patients, infection results in an acute hepatitis with symptoms resembling other forms of acute hepatitis.4 The mean incubation period is seven weeks and symptoms, if present, last for 2–12 weeks. There are however few reliable studies on the natural history of acute HCV and although the minority of patients clear the virus, the precise numbers are not known.

Patients with chronic HCV often have no symptoms but may complain of non-specific symptoms such as fatigue, muscle aches, anorexia, right upper quadrant pain, and nausea. Symptoms and signs of chronic liver disease occur later in the disease. However, some patients with chronic HCV cirrhosis remain asymptomatic. Thus the presence of symptoms is a poor marker of the severity of liver disease.

2.4 EXTRAHEPATIC MANIFESTATIONS

HCV infections have been associated with a number of immunological disorders including autoimmune hepatitis, Sjogren's syndrome, lichen planus, thyroiditis, membranous glomerulonephritis, and polyarteritis nodosa.60 HCV is associated with essential mixed cryoglobulinaemia.61-63

Recognition of HCV involvement in disorders such as cryoglobulinaemia and idiopathic thrombocytopenic purpura64 will allow consideration of IFN therapy for these non-hepatic as well as hepatic diseases.

2.5 HEPATITIS C VIRUS AND ALCOHOL

There are high rates of HCV antibody positivity among alcoholic patients.65 ,66 Most antibody positive patients are also HCV RNA positive and some studies suggest higher levels of HCV RNA in this group of patients,67 although this remains to be confirmed. The presence of anti-HCV antibodies is associated with more severe liver disease in alcoholic patients.

The importance of alcohol in chronic HCV infection was shown in the recent study of Poynard et al showing that daily consumption of more than 50 g of alcohol is associated with an increased rate of fibrosis progression.58

2.6 HEPATITIS C VIRUS AND HEPATOCELLULAR CARCINOMA

HCV infection is associated with a large proportion of HCCs. In southern Europe and Japan, 50–75% of HCCs are associated with HCV.65 ,68-70 HCV may cause HCC as a consequence of cirrhosis or as a result of chronic necroinflammation rather than having any direct carcinogenic effects. Unlike HBV, HCV does not integrate into the host's DNA. The majority, if not all, of patients with HCV associated HCC have established cirrhosis. Both HBV coinfection and excess alcohol seem to have an additional effect on the development of HCC.71 ,72

The natural history of disease progression is slow in HCV related liver disease with estimates of 20–30 years' duration of infection prior to the development of HCC.49 In patients with established cirrhosis the rates of development of HCC range between 1% and 7% per year.59 ,73 The role of antiviral therapy in preventing the development of HCC in HCV infected cirrhotics is controversial.73

3.1 DIAGNOSTIC SEROLOGICAL ASSAYS

The discovery of HCV in 198974 led to the development of an antibody diagnostic assay based on viral recombinant peptides. The first generation tests incorporated a fused antigen of human superoxide dismutase (SOD) and HCV polypeptide (C100–3) used in an enzyme linked immunosorbent assay (ELISA).75 The first generation assay lacked sensitivity and specificity prompting the development of second generation assays incorporating antigens from the nucleocapsid (C22) and NS3 (C33) genomic regions. Third generation assays (ELISA-3) have since been introduced incorporating antigens from the putative nucleocapsid, NS3, NS4, and NS5 regions. ELISA-3 tests have a sensitivity of 97% and have shortened the mean time to seroconversion by 2–3 weeks.76 ELISA-3 tests are now the most widely used screening tests for HCV77 ,78 but despite the improved specificity, confirmation of positive results is still required as a significant proportion of positive tests will represent false positive results. The false positive rate is especially important in low prevalence settings where the number of false positives may exceed the number of true positives.

A positive ELISA test in a patient with chronic liver disease is probably enough to diagnose HCV infection and a confirmatory antibody test may not be needed. Confirmatory PCR testing of serum for HCV RNA is suggested for this group of patients.

3.2 CONFIRMATORY ASSAYS

By immobilising HCV antigens on to nitrocellulose strips, recombinant immunoblot assays were developed (for example, Chiron RIBA, Chiron Diagnostics, Emeryville, California, USA) for confirmation of positive ELISA results. A first recombinant immunoblot assay (RIBA-100) was developed with separately immobilised C100-3, 5-1-1, and SOD antigens.

Second generation RIBA tests were developed with antigens from nucleocapsid (C22) and NS3 (C33) in addition to C100-3 and 5-1-1. Both chimpanzee79 ,80 and human studies81-84 have suggested that second generation tests allow earlier detection of HCV infection in acute cases and are more frequently positive in chronic cases. A positive second generation RIBA result is associated with HCV viraemia by PCR in 88–98% of cases.85-87

A positive RIBA test is associated with reactivity with two or more of the antigens, and in the majority (63%) of cases88reactivity to all four antigens is detected. An indeterminate result shows reactivity to any one antigen. Several studies have shown that reactivity with c100-3 or 5-1-1 alone is rarely associated with PCR positivity and can be regarded as falsely positive.86 ,88-90 The majority of patients with lone antibody to c33 and about half of those with antibody to c22 will be PCR positive and therefore represent true positive results.86 ,88 ,89 ,91 ,92

Third generation RIBA tests have been developed incorporating synthetic C22 and C100-3, recombinant C33, and a recombinant NS5 antigen expressed in yeast to replace 5-1-1. This later version has been shown to be positive in most RIBA-2 indeterminate cases90 ,93and to correlate better with HCV viraemia.94 However, despite the improved sensitivity of this test, indeterminate results have been observed and HCV RNA is detected in 58% of these cases.95 Thus patients with indeterminate RIBA-3 results must be evaluated for evidence of viral replication and liver disease.

Following a positive antibody test, patients should be referred to the nearest specialist service for further clinical assessment. Specialist clinicians will be responsible for the care of these patients and will ensure some uniformity of approach while facilitating data collection, audit, and research.

3.3 THE POLYMERASE CHAIN REACTION

Initial PCR for HCV detection used primers derived from heterogeneous non-structural regions of the virus. The development of primers from the highly conserved 5′ non-coding region greatly enhanced the detection of HCV RNA by PCR.96 The sensitivity of PCR detection was further enhanced by the development of PCR primers producing shorter PCR products.96 The sensitivities of most PCR assays is in the range of 500–1000 equivalents per ml.

Direct detection of the virus using PCR is needed in patients recently infected with the virus and in immunosuppressed individuals who may be antibody negative. In addition, PCR is useful for determining the status of patients with indeterminate antibody profiles and for monitoring antiviral therapies. The sensitivities and specificities of the commercially available PCR tests are very high. Intermittent viraemia is unusual in patients untreated with IFN so enhancing the significance of a negative PCR result.97

3.4 LIVER TESTS

The use of routine liver tests to screen for chronic HCV infection is of limited value as about 50% of HCV infected (anti-HCV and PCR positive) patients will have normal transaminase values. Despite normal liver tests these viraemic patients should not be considered “healthy carriers” as the majority will have histological evidence of necroinflammatory liver disease with or without cirrhosis.98 Other studies have shown that transaminase levels can be helpful in predicting severity of liver disease, with higher levels associated with more advanced histology, but that they are of limited value in an individual patient.99 The value of monitoring transaminases is limited with levels fluctuating from normal to abnormal over time.

3.5 LIVER HISTOLOGY

Liver biopsy is usually performed before initiation of antiviral treatment and remains the most accurate measure of the extent of liver disease. Liver biopsy is usually done in patients with evidence of chronic HCV infection with abnormal transaminases who are being considered for antiviral therapy. In addition, histological information is useful when other diagnoses such as alcohol induced liver disease are being considered.

The role of liver biopsy in patients with normal transaminases is less clear. Several studies have shown that patients with normal transaminases often have evidence of significant liver disease on liver biopsy. In one study, 11% of 54 patients with CAH or active cirrhosis had normal alanine aminotransferase (ALT) levels,100 and in another more than 50% of patients with CAH, CPH, or cirrhosis had normal ALT levels.101 Liver biopsy may be considered in HCV positive patients with normal liver function tests (LFTs) and positive for HCV RNA who are being considered for treatment.

In a further study the use of clinical parameters to predict cirrhosis was found to be inaccurate with a correct diagnosis in less than one third of cases.102 In the absence of a less invasive measure of fibrotic liver disease, liver histology remains the gold standard for the assessment of the severity of liver disease.

The biopsy appearance at presentation does not predict the rate of disease progression in an individual non-cirrhotic patient but biopsies taken every 2–3 years may be useful in predicting outcome if there is progressive accumulation of fibrous tissue.

Some patients will test positive for antibody to HCV, have abnormal LFTs, but will be PCR negative: these patients should be screened for other liver diseases including autoimmune hepatitis and haemochromatosis. Anti-HCV positive patients found to be PCR negative with normal ALT levels should probably be followed up annually until the natural history (virological and biochemical relapse rate) is better known: liver biopsy may be recommended if there is a return of viraemia or a flare up of liver enzymes.

Liver biopsy is probably not indicated after a course of treatment in the majority of patients. A repeat liver biopsy at a remote time interval will provide information on disease progression in both responders and non-responders but the precise timing is not clear.

3.6 ASSESSMENT OF VIRAEMIA

Measuring the level of HCV RNA in blood samples has been widely reported with some studies showing varying levels with changes in LFTs103 and others suggesting stable levels in individual patients prior to treatment.97 However, the role of HCV quantitation in determining disease course remains unclear but the results of recent trials suggest that levels of viraemia are important in tailoring IFN/ribavirin combination therapy. The level of HCV viraemia can be measured by quantitative PCR103-105 or by signal amplification techniques such as branched DNA assay.

3.7 GENOTYPING

Analysis of the conserved 5′NCR allowed the distinction of three major groups, types 1, 2, and 3. Analysis of samples from around the world led to the discovery of other genotypes. Type 4 HCV was found predominantly in the Middle East and Egypt,106 and type 5 sequences were found only in South Africa.107 ,108 More recently type 6 has been described from Hong Kong.109

Phylogenetic analysis of the NS5 region has allowed the classification of HCV into six major genetic types and a number of subtypes. So far there has been no overlap in sequence variability between the different classes with nucleotide homologies of 88–100% between isolates, 74–86% between subtypes, and 56–72% between types.

Typing can be performed in several ways, either serologically with specific peptide ELISAs (serotyping)110 or by analysis of PCR products. The latter can be carried out by direct sequencing,111 ,112 with type specific primers113 on the basis of restriction fragment length polymorphisms114 or with sequence specific DNA probes (genotyping).115 Genotype helps predict the rate of disease progression116 ,117 and response to antiviral treatment.118 ,119

4.1 COUNSELLING

The diagnosis of HCV causes considerable anxiety to patients and it is therefore essential that all patients receive adequate counselling from a health carer with knowledge and experience in this field. The natural history, treatment options, and likelihood of success should be discussed. Patients should be reassured that HCV infections are not usually associated with other infections such as HBV or HIV. Although the precise role of sexual transmission remains to be established,26 ,32 ,120 ,121 because up to 5% of spouses of infected patients are infected,30 ,31 ,122 couples in new relationships should be advised to use barrier contraception. In established relationships the small risk of transmission should be explained and the couple should be reassured and left to decide whether to change their sexual practices. The risk of vertical transmission seems to be low (<6% of children becoming HCV positive) unless the mother is HIV positive or has a particularly high level viraemia.34-36 Mothers should be advised that breast feeding is probably safe and that so far HCV RNA has not been demonstrated in breast milk.38 ,39

Patients must also be screened for their suitability to receive IFN and ribavirin therapy. The decision to treat must be taken jointly by the physician and patient, based on careful consideration of a number of different factors.

Patients should probably not be offered IFN if there is a history of depressive illness, psychosis, untreated autoimmune thyroid disease, neutropenia and/or thrombocytopenia, organ transplantation other than liver, symptomatic heart disease, decompensated cirrhosis, uncontrolled seizures, or evidence of ongoing alcohol or intravenous drug abuse. Patients should have access to reliable refrigeration to store IFN and be able and willing to make regular clinic visits. Adequate warning should be given of the usual initial effects of IFN (fever and malaise) and in particular absence from work may be necessary during the early stages of treatment. Administration prior to sleep with a predose of 0.5–1 g of paracetamol in the initial weeks of treatment may reduce “flu-like” symptoms associated with initial IFN therapy. Severe side effects from either IFN or ribavirin are infrequent but they may be reversible and dose modulation may successfully reduce the occurrence of side effects while maintaining therapy. Women should be advised not to conceive during a course of IFN.

Ribavirin is contraindicated if there is evidence of end stage renal failure, anaemia, haemoglobinopathy, severe heart disease, uncontrolled hypertension, and for women, where appropriate pregnancy (a pregnancy test prior to treatment is advisable), or no reliable method of contraception. Both men and women should be advised to avoid conception during and for six months after IFN/ribavirin combination treatment.

These guidelines apply to adults over 18 years of age; the upper age limit where treatment should be given is unclear but perhaps 65 or 70 years would be reasonable.

The management of patients needs to take into account the differing patient groups according to their transmission routes. These groups include current and ex-injecting drug users, blood or blood product recipients (some identified by HCV lookback), and those with unknown or unapparent transmission (for example, sexual, vertical, household, or occupational). These patient groups may well need different approaches to the way they are managed.

For many current IVDUs there are multiple contraindications to therapy, in addition to concern that continued or recommencing intravenous drug use will result in reinfection with HCV. Some patients may respond to a multidisciplinary approach and antiviral treatment may be considered within the context of detoxification and rehabilitation programmes. Many patients with a history of injecting drugs fail to reattend for follow up after the initial diagnosis of HCV.123 Although current IVDUs should not be treated, some patients on oral methadone and particularly those individuals who are committed to rehabilitation programmes may be considered for treatment.

The group of patients infected by blood products is similar to that infected by blood but may already be affected by the dissemination of HBV or HIV. There is debate about the health impact of HCV on patients with genetic clotting disorders as well as the need for intervention, including liver biopsy, monitoring, or non-intervention.124 The risks and costs of performing liver biopsy may be greater than in other groups of patients and in many of these patients the consequences of the clotting disorder or of coinfection with HBV or HIV are more of a health concern than chronic HCV. The management of patients coinfected with HIV is also controversial, particularly in view of recent developments in antiviral therapy.

An estimated 2–5% of chronic HCV infected patients have no behavioural risks but all of the above routes may be the source of possible transmission along with others such as tattooing and poor sterilisation of reused medical instruments.

Significant numbers (up to 40%) of patients do not accept treatment or complete the full course of treatment or follow up,123particularly those patients with a history of injecting drugs. Even in those with moderately severe disease, a significant number (42%) did not want to undergo treatment.

The role of patient support groups at the local and national level is to be encouraged and perhaps facilitated in the primary care setting. The British Liver Trust organise a national support group network and help in starting up new groups (telephone 01473–276326).

4.4 INTERFERON MONOTHERAPY

Numerous studies have now been published to evaluate different IFNs, dosing regimens, and response definitions. The disparate study designs and data analysis make interpretation of the results and comparison with other studies difficult. Few trials have included more than 100 patients per treatment group.125

The goal of treatment is the achievement of sustained (24–48 weeks post treatment cessation) transaminase and virological response (PCR negative) with histological improvement. Most of the treatment trials have used similar doses of between 1 and 3 million units (MU) of IFN three times a week for periods of 3–6 months. A dose of 3 MU is more efficacious than 1 MU.126 In addition, only those patients receiving 3 MU had significant improvements in liver histology. Albertiet al have shown that 6 MU three times a week leads to a higher proportion of patients with normal ALT at the end of treatment compared with those treated with 3 MU three times a week.127 Another study using 10 MU three times a week suggested that sustained response rates could be as high as 50% although there is a greater risk of treatment failures due to side effects.128

Longer treatment regimens of 12 or 18 months also resulted in greater numbers of sustained responders. In one study with a three year follow up period, treatment for 48 weeks led to a sustained biochemical response in 57.1% of patients compared with 15.4% in patients treated with the same dose for 24 weeks.129 One large trial studied 329 patients treated initially with 3 MU three times a week for six months and then randomised to a further one year of 3 MU or 1 MU three times a week or no further treatment.130 A total of 303 patients were randomised and the study end points were normalisation of ALT at the end of treatment during a follow up period of 19–42 months and improvement in histology at the end of treatment. Patients treated with 3 MU for 18 months were more likely to have normal ALT at the end of treatment (p=0.008), during follow up (p=0.02), and to have improved histological activity scores at the end of treatment (p=0.02).

The majority of patients (>90%) with sustained response seem to maintain normal ALT with negative HCV-RNA in prolonged follow up (1–6 years).131 The histological appearances also improve and in some patients the liver becomes normal.

Poynard at al published a meta-analysis of more than 100 randomised IFN trials125 in 1996. The study analysed placebo controlled trials as well as trials using different IFN regimens. Trials were included if they were clearly randomised, using IFN alone, and were using at least one of the following clinical end points: normalisation of ALT during and at the end of treatment (complete ALT response), sustained ALT normalisation (sustained ALT response 6–18 months post treatment cessation), and improvement in histological lesions when biopsy after treatment was compared with biopsy before treatment. Trials were only included if the dose of IFN was at least 3 MU three times per week with a duration longer than six months. All patients were IFN naive.

4.5 INTERFERON/RIBAVIRIN COMBINATION THERAPY

Ribavirin is a nucleoside analogue that is well absorbed orally and has broad antiviral activity against a variety of DNA and RNA viruses. Ribavirin is administered in doses of 1000–1200 mg/day depending on body weight (above/below 75 kg).

4.6 WHO TO TREAT—ACCORDING TO BIOPSY

The decision of whether to treat is complex. As the treatment is relatively expensive and may not cure most cases, patients need to be selected as those most likely to respond to treatment and also those in whom the impact of treatment is greatest, in terms of halting disease progression and preventing complications. Decisions about treatment should be made after liver biopsy has been performed and patients classified into mild, moderate, or severe disease categories according to histological appearances. Histological appearances are classified as mild if the fibrosis score (stage) is less than or equal to 2/6, and if the necroinflammatory score (grade) is less than or equal to 3/18. If the fibrosis score is 3–5/6 and/or the necroinflammatory score is greater than 3/18, the appearances are described as moderate. If the fibrosis score is 6/6, the biopsy is cirrhotic irrespective of necroinflammatory score.

All liver biopsies should be examined by a histopathologist with experience in liver pathology and who can apply the recently reformed grading and staging scores.146 Some pathologists prefer to base the assessment of severity of hepatitis on individual components of the grading system. In this case mild hepatitis can be defined as having scores for interface hepatitis and for lobular hepatitis of 0 or 1 out of 4. Confluent necrosis should be absent (score 0). Any grade of portal inflammation is acceptable.

In patients with mild slowly progressive disease, it may be best to withhold treatment until more efficacious treatments are available. Others would regard this as the best time to treat, perhaps resulting in higher numbers of responders, and others would argue that the virus rather than the disease process needs to be treated and so all infected patients need to be considered for treatment.

4.7 PREDICTORS OF RESPONSE

Several factors have been implicated but their accuracy in predicting a response in individual patients has been poor. However, some physicians exclude patients from treatment if they have one or more of the pretreat- ment markers associated with a reduced likelihood of response in an attempt to improve response rates and efficiency of antiviral treatment.

4.7.1 Pretreatment factors

Both host and viral factors have been identified by either univariate or multivariate analyses. One initial study suggested a more favourable outcome in young females and in patients with lower pretreatment ALT levels.126 Other studies have failed to link female sex to better response but have confirmed that younger patients tend to respond more favourably127 and that pretreatment ALT and γ-glutamyltransferase levels tend to be lower in responders.153 The better response in younger patients may reflect a shorter duration of infection in association with less severe histological lesions. Indeed, several studies have shown that absence of cirrhosis and low fibrotic histological scores are associated with better treatment outcomes.119 ,127 One study reported the biochemical response rates in patients treated with IFN for 12 months according to the presence or absence of cirrhosis. Within a six month follow up period, 5.3% of cirrhotics compared with 40.5% of non-cirrhotics showed sustained normal ALT levels.149Other studies have shown that the hepatic iron content of non-responders (1156±283 μg/g dry weight) tends to be higher compared with responders (638±118 μg/g dry weight; p<0.05).154

Viral factors thought to be important in determining treatment response include viral genotype, level of viraemia, and level of viral heterogeneity. Improved responses are found in those patients infected with HCV 2118 ,155 or HCV 2 and 3 compared with patients infected with genotype 1.117 In a review of 15 IFN trials, sustained response was seen in 18.1% of HCV 1 infected patients compared with 54.9% of patients infected with other genotypes.156 However, the positive predictive value and accuracy of genotype in predicting sustained response was shown to be fairly poor (55% for predictive value, 72% for accuracy).

The Benelux study reported on 350 patients randomly assigned to standard IFN monotherapy (3 MU three times a week for 24 weeks) or titrated treatment (6 MU three times a week for eight weeks followed by dose reduction based on ALT levels). A total of 319 patients were evaluable for at least six months of follow up and overall the sustained response rate was no better than 14%. The titrated regimen was no better than standard therapy but in multivariate analysis by logistic regression, infection with HCV 2 or 3 were independent predictors of sustained biochemical response.157

Pretreatment levels of viraemia have been studied in relation to response to IFN monotherapy. One study using a competitive PCR quantitation method revealed that sustained responders had significantly lower HCV RNA levels compared with non-responders.158 Other researchers have used the branched DNA assay (Chiron Corporation) to measure HCV viraemia. Lauet al found that patients with sustained response to IFN had mean viraemia levels of 0.35×10⁶genomes/ml, partial responders with relapse had mean viraemia levels of 1.6×10⁶ genomes/ml, and non-responders had a mean viraemia level of 3.1×10⁶ genomes/ml.159However, recent doubts have been expressed as to the reliability of the bDNA assay, as the sensitivity seems to vary according to genotype.

In the meta-analysis,125 the predictive value of low HCV RNA level for a sustained response was 51%, with an accuracy of 68%. Thus level of viraemia may be an important pretreatment variable but better and more reliable methods and studies are needed before this can be used in the decision making process.

Sequence analysis of the E2/NS1 region of HCV by analysing multiple clones from different patients has shown that the degree of variability in this region correlates with response to IFN.160 ,161 Responders show little or no sequence diversity in this region compared with non-responders who seem to be infected with a large heterogeneous pool of HCV variants. Viral heterogeneity may reflect higher replication rate, longer duration of infection, heterogeneous infecting inoculum, or differential host immune responses. Most recently, a so-called IFN sensitivity determining region has been described within the NS-5 region of the virus.162

Similar pretreatment factors have been analysed in the more recent combination treatment trials (see tables 2, 3). In the IFN/ribavirin combination study of Poynard and colleagues,144 logistic regression revealed five factors associated with a favourable response: genotype 2 or 3, viral load less than 2 million copies per ml, age less than 40 years, minimal fibrosis on biopsy, and female sex (fig 1). Although the difference between 24 or 48 weeks of combination treatment barely reached significance in McHutchinson's trial, it seems that those patients with factors associated with poor response such as genotype 1 and high level viraemia will benefit most from 48 weeks of treatment. In patients infected with genotype 1, the sustained response rate increased from 16% for 24 weeks to 28% for 48 weeks of treatment. In those patients infected with other genotypes, the response rates were equally good for 24 (69%) and 48 (66%) weeks of treatment. Similar findings were shown when patients were discriminated by HCV viraemia with improved sustained response rates for 48 weeks of treatment (36%) in high level viraemia (>2 million copies per ml) compared with 24 weeks (27%). In patients with low level viraemia, there was no benefit in prolonging treatment to 12 months (24 weeks of treatment 42%; 48 weeks of treatment 43%).

View this table:

• View inline
• View popup

Table 2

Sustained responses (SR; PCR negative 24 weeks after treatment cessation) according to genotype in naive patients taken from the two multicentre randomised controlled trials published in 1998144 ,145

View this table:

• View inline
• View popup

Table 3

Sustained responses (SR; PCR negative 24 weeks after treatment cessation) according to baseline viraemia (expressed as copies per ml) in naive patients taken from the two multicentre randomised controlled trials published in 1998144 ,145

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Hepatitis C virus (HCV) algorithm. Sustained response defined as negative polymerase chain reaction (PCR) six months after treatment cessation. Data derived from Poynard and colleagues144 and McHutchinson and colleagues,145 as shown in the review by Weiland.177 IFN, interferon; ALT, alanine aminotransferase; US, ultrasound; AFP, α fetoprotein.

4.8 SIDE EFFECTS OF TREATMENT

4.9 RECOMMENDED TREATMENT REGIMEN AND MONITORING

Monitoring patients during therapy is extremely important, requiring regular clinical examination, psychological assessment, urinanalysis, serum chemistry, blood counts, and thyroid function tests. Pregnancy tests should be performed prior to treatment and patients advised not to conceive while on treatment and for at least six months after combination therapy.

We recommended using alfa-IFN 3 MU three times per week for 12 months if used as monotherapy. Patients should be tested at three months and those failing to respond (biochemically and virologically) should stop therapy and be considered for further treatment trials of combination therapies. Combination therapy should be prescribed as IFN 3 MU three times per week with ribavirin 1000 mg/day for patients weighing 75 kg or less and 1200 mg for those weighing more than 75 kg. Treatment should not be stopped at three months irrespective of biochemical or virological response but in those patients with treatment planned for 12 months (genotype 1) persistence of viraemia at six months may prompt treatment cessation.

Patients should be seen weekly for the first four weeks of IFN/ribavirin combination treatment so that blood counts are performed to look for haemolysis. Thereafter patients should be seen at monthly intervals until six months and then every three months until therapy is finished if treatment is given for a full year. Patients need continued support and encouragement and side effects must be monitored. At each visit full blood counts, renal, thyroid, and LFTs should be taken. Serum for PCR should be taken at three months of treatment and if possible at every three months of therapy. In patients treated with IFN monotherapy a positive PCR test of serum at three months should prompt treatment withdrawal. The response to combination therapy should be assessed by PCR testing at six months in patients with genotype 1; treatment should be stopped if the PCR remains positive.

Dose reductions may be necessary for side effects but if possible the course of IFN or IFN/ribavirin should be completed. Follow up testing, with serum ALT and serum RNA, should be performed six and 12 months after treatment. A negative PCR test 24 weeks after treatment cessation defines a sustained treatment response and in the majority of cases will remain negative in prolonged follow up. We do not recommend routine follow up liver biopsy.

4.10 TREATMENT OF NON-RESPONDERS AND RELAPSERS AFTER INTERFERON MONOTHERAPY

The response rate for a second course of IFN is extremely poor in patients failing to respond to an initial course of IFN (abnormal ALT at the end of therapy). In a review of 13 studies including 591 non-responders, the sustained response rates to a second course were 1–3%.165

However, the results of a second course of IFN in patients relapsing after therapy are more encouraging. The rates of sustained response were 15% for patients receiving a second course of 3 MU for six months, 29% in patients receiving >3 MU for six months, and as high as 43% in patients treated for 12 months or longer. In studies that analysed virological response, the presence of a negative PCR test at the end of the first course of treatment was highly predictive of a sustained response after retreatment. The combined analysis of five studies showed a sustained response in 56% of 145 patients rendered PCR negative at the end of the first course of treatment compared with 2.9% of 206 who remained PCR positive but with normal ALT levels at the end of the first course of therapy.165 There appears little point in retreating patients who relapse after an initial high dose IFN course as sustained response rates are low (<5%).

More recently, IFN/ribavirin combination studies have also revealed improved response rates for a further course of combination therapy after either a failed course of IFN monotherapy or in patients who relapse after treatment. In an Italian study of 96 non-responders, HCV RNA was undetectable at the end of treatment in 27% of patients treated with a further six month course of combination therapy compared with 7% in those treated with IFN monotherapy (p<0.05).166 However, at six months after therapy the response rate had fallen to 15%. In a Spanish study the six month post therapy response was 10% for a mixed group of non-responders and relapsers.167 In another study with equal numbers of non-responders (24) and relapsers (24), HCV was not detectable six months after treatment in 20.8% of those treated with combination therapy compared with 4.2% treated with a further course of IFN monotherapy.168 However, in this study 25% of patients required dose reduction and 12.5% of patients were withdrawn due to intolerable side effects.

The International Hepatitis Interventional Therapy Group published their randomised study of 345 relapsers treated with a further six month course of IFN/ribavirin or IFN monotherapy.169 The primary end point was the absence of HCV RNA in serum at the end of therapy and six months after the cessation of therapy. At the end of treatment 82% (141/173) versus 47% ( 80/172) were PCR negative for the combination versus monotherapy groups, respectively (p<0.001). At six months after treatment cessation the rates of PCR negativity were 49% versus 5% (p<0.001). In this study the safety profiles of the two treatment regimens were similar, with discontinuation of therapy in 6% of the combination group and 3% of the IFN monotherapy group.

4.11 COST EFFECTIVE ANALYSIS

The long term benefit of antiviral therapy is difficult to prove currently due to the relatively long natural history of HCV and the fact that patients need to be followed up for many years after successful viral eradication. So far studies have shown that sustained viral response (that is, negative PCR at 24 weeks after treatment cessation) is maintained for several years and that these viral responses are accompanied by improvements in histological scores of necroinflammation. It seems reasonable to suggest that sustained virological responders will derive long term benefit and will not progress to end stage liver disease as frequently as those failing to respond. Those studies showing reduced rates of HCC in treated patients support this view although the results of further studies are needed to clarify this issue. While monitoring and treating patients costs money, far greater health care related costs are incurred when patients develop cirrhosis and its complications, including variceal bleeding or HCC.170 However, disease progression is not universal and treatment may be unnecessary in patients who do not develop HCV related complications in their lifetime.

Treatment with IFN and ribavirin is associated with high initial costs, and studies need to address the relative cost effectiveness of treatment in terms of years of life gained, quality of life, indirect costs including loss of productivity for sick patients, and projected health care costs for the management of the complications of chronic HCV such as the development of cirrhosis and HCC.

The costs of implementing a management policy involve staff, laboratory, and drug costs. Patients attending clinics regularly will use medical, nursing, secretarial, and other staff time, although many will attend general medical or perhaps gastroenterological/hepatological clinics. In larger centres there may be specialist clinics with attending clinical nurse specialists. The laboratory costs will include routine blood testing, analysis of liver biopsy specimens, and PCR testing (cost ∼£40–50), genotype testing and viral load measures. The cost of one year's IFN at 3 MU three times weekly is approximately £2500 rising to £5000 for six months' combination therapy and £10 000 for those requiring 12 months of IFN/ribavirin combination treatment.

Economic analysis must play a part in planning the management of HCV but due to the long course of the disease many of the important questions about cost effectiveness of therapy cannot be answered. Instead, researchers have made use of modelling using Markov simulations to predict the likely outcomes and thus assess the impact of therapy, such as the one produced by Dusheiko and Roberts.171 This model analysed two hypothetical groups of patients, one treated and one untreated, followed for 30 years. The costs estimated were only those direct costs associated with hospital follow up and treatment, and those indirect costs including quality of life, time off work, and so on were not assessed. Thus the value of intervention in this model was underestimated.

Not surprisingly the costs in the treated group were higher in the first year while treatment is given and follow up is intense. However, in subsequent years the costs in the treated cohort were less than the untreated group because fewer entered the later expensive stages of the disease. Although IFN monotherapy is expensive it is not out of line with many other health care interventions, with a discounted cost per year of life saved ranging from £2142 to £8555. This compares with figures of £32 000 for cholesterol reduction in patients with coronary heart disease,172 £6315 for home dialysis,173 and £3135 for enalapril treatment of chronic heart failure.174

The model was based on information available in the early 1990s and as more data become available about response to treatment and pathogenicity of the various groups it should be possible to produce models that more accurately reflect disease patterns. Wonget al have assessed the cost effectiveness of IFN/ribavirin combination versus IFN monotherapy using recent controlled trial data.175 In comparison with 12 months of IFN monotherapy, the model indicated that combination therapy for six or 12 months should increase life expectancy and is cost effective with a marginal cost effectiveness ratio of $2100 and $2300 per discounted quality adjusted life year gained, respectively. The Scottish Health Purchasing Information Centre have developed a spreadsheet that can be used to estimate costs and is available from the group.176Their cost per life year saved estimates were £3000 (1000 mg ribavirin) and £3500 (1200 mg ribavirin) for six months and £6000 and £6700 for 12 months compared with no treatment. Once again these figures are within the range of other accepted NHS activities. Any increase in targeting therapy to particular groups, such as genotypes, must be balanced with the costs of screening procedures and the sensitivity and specificity of these procedures. Later delivery of therapy to a group of progressors is likely to be more cost effective, as the benefits are likely to be felt sooner.

More information is needed and the results of future studies should include indirect costs to patients, primary care, and community services, and improved outcome measures such as quality of life measures should be incorporated into the analysis.

DIAGNOSIS

• Patients with suspected HCV infection should be tested for anti-HCV by an up to date (currently third generation) ELISA test. (Recommendation grade C.)

• All patients with positive antibody tests and those patients thought to be at risk of HCV infection despite negative or indeterminate serological tests should undergo PCR testing of serum. A positive result confirms current viraemia whereas a negative test suggests non-viraemic infection, transient absence of viraemia or recovered infection, a level of viraemia below the detection limit of the assay, or may reflect a non-specific ELISA result. (Recommendation grade B.)

• Patients with positive ELISA but negative PCR should therefore be tested with recombinant immunoblot assay to confirm antibody status. (Recommendation grade B.)

• A qualitative PCR test is recommended in immunodeficient patients with suspected HCV infection. (Recommendation grade B.)

• The results of routine liver tests correlate poorly with both necroinflammatory and fibrosis scores found on liver biopsy. (Recommendation grade B.)

• Liver biopsy is valuable for assessing status of liver inflammation, potential progression of fibrosis, and the presence or absence of cirrhosis. To clarify these, and to assess suitability for treatment, liver biopsy is recommended for patients found to be viraemic, whether or not liver function tests are abnormal. Standard histological scoring systems by a suitably experienced pathologist should be used to encourage uniformity of histological reports. The risks and benefits of liver biopsies must be fully discussed with the patient. (Recommendation grade B.)

• Measurement of HCV RNA concentrations in serum and determination of HCV genotype are recommended and should be used to determine the duration of treatment (see later). (Recommendation grade A.)

COUNSELLING REGARDING TRANSMISSION

• Patients should be counselled on the implications of HCV positivity and advised on the risks of infectivity.

 — the natural history is slowly progressive (median time to cirrhosis 28–32 years). (Recommendation grade A.)

 — HCV positive patients should not donate blood, organs, tissues, or semen. (Recommendation grade C.)

 — the risk of sexual transmission is small (maximum 5% but possibly much less). (Recommendation grade B.) There is insufficient evidence to firmly recommend barrier contraception in stable monogamous relationships but is strongly advised for HCV infected patients with multiple sexual partners. (Recommendation grade C.)

 — transmission from mother to child is rare (maximum of 6%) but transmission rates are higher in HIV positive mothers. (Recommendation grade B.)

 — breast feeding is not contraindicated. (Recommendation grade C.)

 — household contacts should avoid third party contact with blood by not sharing toothbrushes and razors, and by covering open wounds. (Recommendation grade C.)

 — standard precautions for the prevention of transmission to medical personnel and patients is mandatory in health care settings. (Recommendation grade C.)

 — needle exchange programmes in drug addicts may help reduce parenterally transmitted infection. (Recommendation grade C.)

• Current IVDUs should not be treated although in selected cases ex-IVDUs taking regular oral methadone may be considered for treatment. (Recommendation grade C.)

TREATMENT—GENERAL MEASURES

• Patients should be advised that excess alcohol consumption (>50 g/day) appears to hasten the progression of disease. (Recommendation grade B.)

• Consideration should be given to entering patients with established cirrhosis into surveillance programmes for HCC, if their general state of health is sufficiently good that emerging cancers could be appropriately treated. (Recommendation grade C.)

• Patients must be screened for their suitability to receive IFN and ribavirin, with criteria which includes proven viraemia and abnormal liver histology. (Recommendation grade C.)

• IFN and ribavirin are currently the only licensed treatments for HCV in the UK. • IFN/ribavirin combination is the treatment of choice for IFN naive patients. (Recommendation grade A.)

• IFN/ribavirin combination is also recommended for those patients relapsing after IFN monotherapy. (Recommendation grade A.)

• IFN monotherapy should be considered for those patients in whom ribavirin is contraindicated. (Recommendation grade C.)

• The role of pegylated interferon remains unknown. (Recommendation grade C.)

TREATMENT—INTERFERON MONOTHERAPY

• We recommend IFN monotherapy should be initiated at a dose of 3 MU three times per week by injection. (Recommendation grade B.)

• IFN monotherapy should be continued for 12 months unless there is evidence of failure to respond (see below). (Recommendation grade B.)

• There is no evidence to suggest that one type of alfa-IFN is superior to another (alfa-2b, alfa-2a, alfa-n1, and consensus interferon (CIFN)). (Recommendation grade B.)

TREATMENT—INTERFERON/RIBAVIRIN COMBINATION THERAPY

• Recent results of large randomised controlled studies have shown improved response rates for IFN/ribavirin combination therapy in IFN naive and relapsers compared with IFN monotherapy. (Recommendation grade A.)

• Combination therapy consists of IFN at standard doses (usually 3 MU three times per week) with ribavirin 1000 mg/day for patients weighing 75 kg or less and 1200 mg for those weighing more than 75 kg. (Recommendation grade A.)

• In viraemic patients, the decision to offer treatment should be influenced by the histological findings. (Recommendation grade B):

 — treatment can be reasonably withheld in patients with mild disease (see text) but they should be followed to see if there is evidence of progressive liver disease by the use of repeated biopsy after an interval. (Recommendation grade C.)

 — treatment should be offered to those patients shown to have moderate disease. (Recommendation grade C.)

 — cirrhotic patients respond less well to IFN monotherapy but sustained responses have improved with IFN/ribavirin combination treatment. There is no conclusive evidence that treatment in this group of patients delays progression of liver disease or the development of HCC. (Recommendation grade B.)

• Treatment should not be withheld on the basis of genotype analysis or the measurement of HCV RNA levels. (Recommendation grade B.)

• The duration of combination treatment depends on the genotype and level of viraemia. (Recommendation grade A.)

• Patients infected with non-HCV 1 (mostly genotype 2 or 3) should be treated for six months irrespective of the level of viraemia. (Recommendation grade A.)

• Patients infected with genotype 1 and low level viraemia (<2 million copies per ml) should be treated for six months whereas 12 months' treatment is recommended for those infected with genotype 1 and high level viraemia (>2 million copies per ml). (Recommendation grade A.) If HCV quantitation is not available treatment is recommended for 12 months in HCV 1 infected patients. (Recommendation grade A.)

• Patients unlikely to respond to IFN monotherapy can be identified at three months by persistent elevation of serum transaminase levels and the persisting presence of HCV RNA by PCR in serum. (Recommendation grade B.) If ALT levels are normal or HCV RNA negative (or both) at three months, treatment should be continued for the full duration (12 months). (Recommendation grade B.) In patients with initially normal ALT levels, failure to become RNA negative at three months suggests longer treatment will be ineffective. (Recommendation grade B.)

• The recommendation that early treatment response can be used to predict sustained response does not apply to patients receiving combination therapy. (Recommendation grade A.) In those patients receiving 12 months of IFN/ribavirin combination therapy a positive PCR at six months is an indication to stop treatment. (Recommendation grade C.)

• Although transient or mild side effects are common during IFN monotherapy, serious toxicity requiring reduction in dose or cessation of treatment occurs in 5–10% of patients during treatment. (Recommendation grade A.)

• Withdrawal from IFN/ribavirin combination therapy occurs more often with 10–20% of patients requiring a reduction in dose or cessation of combination therapy. (Recommendation grade A.)

• Patients with a combined biochemical and virological response at the end of IFN monotherapy, who relapse in follow up over the next year, have a significant chance of a sustained response after further treatment with IFN/ribavirin. (Recommendation grade A.)

• Patients with a biochemical but not virological response during initial treatment with IFN monotherapy are unlikely to have a sustained response to further treatment with IFN/ribavirin. (Recommendation grade A.)

• There is continuing development in the treatment of patients with HCV infection. In particular, the role of pegylated interferon and re-use of weight adjusted doses of interferon will shortly be established. The guidelines will need regular and frequent review. (Recommendation grade C.)