Motion – The available treatments for hepatitis C are cost effective : Arguments for the motion

This article was originally presented at a symposium entitled, “Controversies in Gastroenterology”, sponsored by Axcan Pharma, Toronto, Ontario, April 8 to 10, 2002 Department of Medicine, University of California, San Francisco, California, USA Correspondence: Dr Norah Terrault, Department of Medicine, University of California, Box 0538, San Francisco, California 94143-0538, USA. Telephone 415-476-2227, fax 415-476-0659, e-mail noraht@itsa.ucsf.edu N Terrault. Motion – The available treatments for hepatitis C are cost effective: Arguments for the motion. Can J Gastroenterol 2002;16(10):705-709.


KEY ASPECTS OF COST EFFECTIVENESS ANALYSES AS THEY RELATE TO HEPATITIS C TREATMENT The perspective of cost effectiveness analyses
Most cost effectiveness analyses (CEAs) are undertaken from the perspective of society in general (ie, the public interest is served rather than that of any particular individual or group) (1).This is viewed as the most ethical approach, against which other perspectives can be judged.The societal perspective reconciles the health care needs of the population with the limitations to available resources, and is most able to determine the cost implications of competing health care strategies.On the other hand, such a perspective may not necessarily be readily usable by clinicians in the daily care of patients.Clinicians must determine whether the results of a CEA are valid and applicable to their own patients.

The 'reference' or 'base' case
The use of a 'reference case' is recommended, so that meaningful comparisons can be made among different CEAs (1).It should be representative of the target population for the intervention and provide details regarding the event pathway.Unfortunately, CEAs that address the treatment of hepatitis C virus (HCV) use different reference cases and, therefore, involve different patient populations.
The clinician cannot adequately interpret the results of a CEA without being aware of the target population.For example, because outcome measures, such as incremental cost effectiveness ratios (ICERs) and numbers of life-years gained, vary with the age of the target population, a CEA that involves 35-year-old patients (2) may not be comparable with one concerning patients who are 45 years old (3).Other parameters that can affect the results and that vary among different CEAs include indications for stopping treatment, retreatment of relapsers and nonresponders, and assumptions about patient compliance (Table 1).
The event pathway provides details about the progression of the disease and the effects of the intervention.The complexity of event pathways is determined by the number of transitions, from one health state to another, that are included.For example, the model of health states used by Bennett et al (2) was more detailed and incorporated more transition states than that employed by Younossi et al (3).
Finally, data regarding monetary and resource expenditures, therapeutic effectiveness and utility weights (qualityof-life [QOL] measures) are assigned to each of the events in the disease pathway.The validity of the CEA ultimately depends on the source and accuracy of these data.In Table 2, the ideal data sources (inputs) are compared with sources that have been used in the published CEAs on HCV treatment.
The characteristics of the target population, QOL estimates and costs of the intervention vary by country and locale.Therefore, country-or even locale-specific data should ideally be used.The perspective used and the sources of information regarding resource, utility and cost inputs for three recently published studies (4-6) comparing interferon (IFN) with IFN plus ribavirin are summarized in Table 3.In none of the CEAs were the data sources entirely countryspecific.Authors frequently cite data on therapeutic effectiveness from a variety of sources to increase the precision of the effectiveness estimates.For example, the Swedish study by Sagmeister et al (4) employed data from European, Canadian and American patient populations.It would be most appropriate to use country-specific sources of utility data, but this is not uniformly done; for example the QOL weights derived from an expert panel of American hepatol-  (16) 10-year-old child with chronic hepatitis Relapsers were not retreated and prognosis was the same as that for nonresponders Patients were 100% compliant There were no treatment-related complications ogists (as used by Buti et al [5]) may not be appropriate for a Spanish patient population.

Reporting of results: what is the outcome of interest?
Several outcome measures are widely used in the literature: • Quality-adjusted life-years (QALYs): This is regarded as the 'gold standard' for evaluating the effects of different health interventions.
• Life-years gained: This is often reported but, because it is not adjusted for the QOL, it distinguishes lifesaving from quality-enhancing effects of the intervention.
• ICER: This is the cost per QALY gained by the intervention compared with either an alternative or no intervention.An intervention is typically considered cost effective if it has an ICER of $50,000/QALY gained or less, although not all HCV studies have used this cut-off value.For example, Buti et al (5) decided that therapies with ICERs above ¤25,000/QALY gained were not cost effective.

Dealing with uncertainty
Variance in the estimates of disease prognosis, costs and resource utilization, and effectiveness of the intervention lead to uncertainties about its economic impact.The conventional method for dealing with these uncertainties is to undertake a sensitivity analysis, whereby the estimates of key components of the model are altered to assess their impact on the cost effectiveness ratios.The influence of demographic factors, such as age, sex or race, can also be investigated in this way.

CEAs OF ANTI-HCV THERAPY IN TREATMENT-NAIVE PATIENTS Interferon versus no treatment
Two studies (2,7) compared a six-month course of IFN with no treatment.Dusheiko et al (7) considered a hypothetical cohort of HCV patients aged 25 to 35 years, and found that the ICERs ranged from £420 to £525 (approximately US$630 to US$790).
Bennett et al (2) considered patients with mild histological disease at the time of treatment.The discounted ICERs varied significantly with the age of the patient: $530 for 20-year-olds, $7,100 for 50-year-olds and $62,000 for 70-year-olds.Moreover, the results were affected by the cost of IFN treatment and the rates of transition from mild to moderate disease and from moderate chronic hepatitis to cirrhosis.

IFN for six versus 12 months
Kim et al ( 8) compared six versus 12 months of IFN treatment and found that both were cost effective ($4,000 and $5,000/QALY gained, respectively) and comparable with other medical interventions.Even though 12 months of IFN was more efficacious (resulted in more QALYs and fewer deaths), six months of treatment was more cost effective (consumed fewer dollars per QALY gained).Age had a significant effect, in that 50-year-old patients with 'worst case scenario' factors had marginal cost effectiveness ratios exceeding $50,000/QALY gained.The cost and effectiveness of IFN therapy, the cost of treatment for decompensated cirrhosis, the QOL of patients with chronic hepatitis C, and the rate of disease progression all significantly influenced the results.
Younossi et al (3) compared IFN monotherapy in four different combination treatment strategies: • 12 months of combination therapy; • IFN followed by combination therapy for relapsers; • IFN followed by combination therapy for relapsers or nonresponders; • Combination therapy with the duration based on HCV genotype (48 weeks for genotype 1 and 24 weeks for other genotypes).
The fourth strategy was both the most effective and the most cost effective, with a discounted ICER of $7,500/QALY gained.Sensitivity analyses, involving the cost of combination therapy and the response and relapse rates, confirmed that the results were generally robust.However, the cost effectiveness ratios were sensitive to the effects of age and the effectiveness of combination therapy.
The cost effectiveness ratio increased to greater than $50,000/QALY gained if end-of-treatment responses decreased to less than 30%.
Buti et al ( 5) compared combination IFN therapy for 48 weeks versus combination therapy for 24 weeks, and found that it was associated with ICERs of ¤8515 and ¤15,891 per QALY gained, respectively.They found that the ICER for 48 weeks of combination therapy, compared with 48 weeks of IFN, was markedly affected by the presence of unfavourable treatment characteristics.For example, the ICER was ¤45 per QALY gained for young (under age 50 years) female patients with mild histological changes, low viral load and viral genotype 2 or 3, but increased to ¤52,169 per QALY gained for older male patients with more severe disease, high viral load and genotype 1.
Wong et al ( 6) compared combination therapy for 24 or 48 weeks with 24 and 48 weeks of IFN alone.They found that combination therapy had ICERs of $5,400 and $7,700 per discounted QALY gained, respectively, compared with 48 weeks of IFN monotherapy.Combination therapy for 24 weeks was cost saving compared with 48 weeks of IFN monotherapy for patients with HCV genotype 2 or 3. Sensitivity analyses involving sex, initial histology, viral load, the presence of genotype 1 and number of favourable response characteristics caused the ICER for 24 weeks of combination therapy to vary from $250 to $11,600/QALY gained compared with 48 weeks of IFN monotherapy.Similar sensitivity analyses showed that, compared with 24 weeks of therapy, 48 weeks of combination therapy had ICERs of $2,500 to $50,200/QALY gained.The longer duration therapy was inferior, however, for patients with genotype 2 or 3 and for those with four or five favourable response characteristics.
Sagmeister et al (4) compared 24-and 48-week courses of combination therapy with 48 weeks of IFN monotherapy.Their results were similar to those of other investigators (5,6).For patients with HCV genotype non-1, 24 weeks of combination therapy was the most cost effective.For patients with genotype 1, 48 weeks of combination therapy prolonged life expectancy for those who responded by 24 weeks, with a favourable cost effectiveness ratio of ¤7,135/QALY gained.Other treatment response factors did not significantly influence cost effectiveness.
Sennfalt et al (10) compared 24 and 48 weeks of combination therapy with 48 weeks of IFN.They found that the two combination therapy regimens were associated with ICERs of $1,400 and $6,000 per discounted QALY gained, respectively, compared with IFN monotherapy for genotype 1 patients.For these patients, 48 weeks of combination therapy had an ICER of $9,800 per discounted QALY gained, compared with 24 weeks of treatment.For patients with other HCV genotypes, 24 weeks of combination therapy was cost saving compared with 48 weeks of combination therapy, and had an ICER of $1,800/QALY gained compared with IFN monotherapy.Because only very limited sensitivity analyses were undertaken, it was difficult to assess the effects of age, treatment efficacy and the natural history of the disease on ICERs in this CEA.
Stein et al ( 9) evaluated combination therapy, with a duration of 24 weeks (for patients with viral genotype 2 or 3) to 48 weeks (for genotype 1) versus IFN monotherapy.They found that the discounted ICER per QALY gained for combination therapy, compared with IFN monotherapy, was £3,485.Discounted cost per QALY gained was less for women aged younger than 40 years with HCV genotype 2 or 3 and moderate hepatitis than for patients with less favourable response characteristics, and varied from £872 to £8,626.Medication costs were more important than assumptions about disease progression or the cost of treating hepatitis C disease in the sensitivity analyses.

CONCLUSIONS FROM AVAILABLE STUDIES
Several findings have emerged from CEAs of treatment for hepatitis C infection: • Marginal ICERs increase with the age of the patient, especially if there are unfavourable response characteristics or if IFN monotherapy is employed.
• The combination of IFN and ribavirin is more cost effective than IFN alone.
• It is useful to determine the HCV genotype when planning antiviral therapy: 24 weeks of combination therapy is more cost effective than 48 weeks for patients with genotype 2 or 3.
• The ICERs with combination therapy are very sensitive to the efficacy of treatment, which, in turn, is strongly affected by certain response characteristics.For example, in patients with unfavourable response characteristics (older male patient, genotype 1, high viral load and significant fibrosis [11]), combination therapy is associated with a high marginal ICER.

Future needs and unanswered questions
Pegylated IFN is more effective than the standard formulation and has recently been approved in North America and Europe for patients with HCV.When combined with ribavirin, it has become the treatment of choice for previously untreated patients with chronic HCV infection (12,13).
Because both treatment costs and response rates increase with the use of pegylated IFN, CEAs would assist health care providers in deciding when to offer this agent.There are no published CEAs that compare pegylated with standard IFN available as yet.
There is some evidence, from posthoc analysis of the results from one clinical trial, that adjustment of the dose of ribavirin according to the patient's weight may be advantageous.Additional efficacy data are required before CEAs of weight-based versus fixed-dose ribavirin, in combination with pegylated IFN, can be undertaken.
Even though a CEA has found that 24 weeks of standard IFN plus ribavirin therapy is more cost effective than 48 weeks of treatment in patients with genotype 2 or 3 HCV infection (6), the optimum duration of treatment based on genotype requires further clarification.In addition, future CEAs need to use data on 'early stopping' rules from recent clinical trials.For example, discontinuation of treatment at 12 weeks in virological nonresponders has been recommended in partials receiving pegylated interferon and ribavarin (12,14).
There are no published studies on the cost effectiveness of antiviral therapy for patients who have not responded to either IFN alone or in combination therapy.Only one study has been published that assessed the cost effectiveness of IFN plus ribavirin in patients who had relapsed after IFN monotherapy (15).Relapsers and nonresponders constitute a significant proportion of HCV patients in clinical practice.CEAs can assist in the evaluation of several treatment options, including retreatment with IFN and ribavirin; retreatment with pegylated IFN and ribavirin; and the use of either 'maintenance' IFN or pegylated IFN.These therapeutic strategies are being investigated in large clinical trials, the results of which are essential for future CEAs.

TABLE 1 Reference case and treatment-related assumptions used in cost effectiveness analyses on hepatitis C virus treatment (American studies only) Authors (reference) Reference case Treatment-related assumptions made by the authors
TerraultCan J GastroenterolVol16 No 10 October 2002 706 Wong et al (6) Not stated No histological benefits in virological nonresponders Target population the same as in Treatment discontinued at 12 weeks (interferon [IFN]) or at clinical trial: mean aged 42 years, 24 weeks (IFN plus ribavirin) in virological nonresponders 66% male, 95% without cirrhosis Sinha and Das

TABLE 2 Ideal data sources versus actual data sources for cost effectiveness analyses (CEAs) on hepatitis C virus (HCV) treatment Ideal data source Actual data source in published CEAs
Hepatitis C treatments and cost effectivenessCan J Gastroenterol Vol 16 No 10 October 2002 707