Growth factors during HCV therapy may be “cost-effective”, but are they “effective”?


  • Andrew J. Muir M.D., M.H.S.,

    1. Duke Clinical Research Institute and Division of Gastroenterology, Duke University Medical Center, Durham, NC
    Search for more papers by this author
  • John G. McHutchison M.D.

    Corresponding author
    1. Duke Clinical Research Institute and Division of Gastroenterology, Duke University Medical Center, Durham, NC
    • Duke Clinical Research Institute, P.O. Box 17969, Durham, NC 27715
    Search for more papers by this author
    • fax: 919-668-7164

  • See Article on Page 1598.

  • Potential conflict of interest: Dr. McHutchison and Dr. Muir are consultants for, are on the speakers' bureau of, and received grants from Schering Plough. They received grants from Roche, GlaxoSmithKline, and Pfizer.

Over the last 2 decades, there have been important advances in the care of patients with hepatitis C virus (HCV) infection. The last major treatment breakthroughs were the approvals of peginterferon-α-2b in 2001 and peginterferon-α-2a in 2002. While we await the results of ongoing trials of promising novel antivirals, considerable efforts have gone toward improving the response of current therapies.


HCV, hepatitis C virus; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year; SVR, sustained virologic response; G-CSF, granulocyte-colony stimulating factor.

We have consistently seen that compliance during therapy and the dose of ribavirin are strong predictors of response for genotype 1–infected patients.1, 2 The challenging side effects of both peginterferons and ribavirin also result in high rates of dose reductions and discontinuations in clinical trials and in practice. To combat the hematologic side effects of these medications, “growth factors” have been employed for the last several years. A retrospective database analysis of a large health benefits company reported that among patients receiving peginterferon-α-2a and ribavirin, 16.7% of patients received erythropoietin while 12.9% received granulocyte-colony stimulating factor (G-CSF).3 An obvious drawback is the substantial cost of these medications, especially when combined with the expense of HCV therapy.

With these issues in mind, Del Rio and colleagues present a cost-effectiveness analysis in this issue of HEPATOLOGY that examines the use of commercially available erythropoietins for the treatment of anemia in patients undergoing HCV treatment.4 Although the anemia from ribavirin is secondary to hemolysis, erythropoietins are used to stimulate red blood cell production and enable the patient to compensate for the anemia. The authors ultimately determined that darbepoetin use is a cost-effective strategy. Compared to a ribavirin dose-reduction strategy, the incremental cost-effectiveness ratio (ICER) for darbepoetin was $34,793 for patients with genotype 1 and $33,832 per quality-adjusted life year (QALY) for patients with genotype 2 or 3. These costs fell below the accepted ICER threshold of $50,000/QALY, although treatment with the more expensive erythropoietin-α was not cost-effective ($60,600 per QALY for patients with genotype 1 and $64,311 per QALY for patients with genotype 2 or 3 infection).

When evaluating a cost-effectiveness analysis, the model must include all relevant strategies and outcomes, and the process used to assemble the evidence into probabilities needs to be explicit and sensible.5 One challenge for these authors was the small number of randomized trials examining this issue. The three major trials addressing the use of these agents in HCV treatment are summarized in Table 1. Two of these trials did not report virologic outcomes but suggested that the additional use of erythropoietin resulted in the delivery of higher doses of ribavirin, maintenance of higher hemoglobin levels and improved health-related quality of life.6, 7 The most recent trial did not find that the use of growth factors improved sustained virologic response (SVR) with standard doses of peginterferon-α and ribavirin.8 Only when growth factors were combined with higher ribavirin doses was an improvement in response rate observed.

Table 1. Published Randomized Trials of Erythropoietin in HCV Therapy
StudyRandomizedPlacebo-controlled, double-blindTreatment armsAnemia definitionMean Hb increase (epoetin versus SOC), g/dLRibavirin dose maintained (epoetin versus SOC)Duration of study observationSVROther Endpoints
  1. Abbreviations: Hb, hemoglobin; SOC, standard of care; SVR, sustained virologic response; NR, not reported; QOL, quality of life; EVR, early virologic response; ETVR, end of treatment virologic response.

Dieterich et al.6yesno1. epoetin-αHb < 12+2.8 versus +0.4,83% versus16 weeks after epoetinNRMean change in
 2003  2. SOC dose reduction g/dLP < .0001 54% P - .022 initiated  ribavirin dose
Afdhal et al.7 2004yesyes1. epoetin-α 2. placeboHb < 12 g/dL+2.2 versus+0.1, P < .000188% versus 60%, P < .000116 weeks after epoetin initiatedNRQOL
Shiffman et al.8 2005yesno1. epoetin-α 2. SOC doseHb < 12 g/dLNRNR72 weeks1.24% 2.27%EVR, ETVR, mean ribavirin dose
    reduction 3. high-dose ribavirin + epoetin-α    3.45% (P = .05) 

The impact of growth factors on SVR is especially important because the cost-effectiveness model described by Del Rio and colleagues assumed that patients receiving growth factors for anemia would have similar treatment responses to patients without ribavirin dose reductions in the major clinical trials of peginterferon-α and ribavirin.1, 2, 9 No trial to date has proven this assumption. The importance of this issue is highlighted by the fact that Del Rio's sensitivity analyses determined that the SVR was critical to the conclusions of the cost-effectiveness model. If the SVR was <50%, the ICER rose above the acceptable $50,000/QALY threshold. The Shiffman study's best strategy with growth factors combined with high-dose ribavirin only led to a SVR rate of 45%, which is similar to the SVR observed in registration trials with standard doses and no growth factors. Lower response rates are often observed when clinical trial data is translated into clinical practice, but these results ultimately mean that the Shiffman study may not support the use of growth factors given the SVR threshold of 50% declared by Del Rio.8

In addition to concerns about the efficacy, these limited data sets have hindered the ability to establish safety in the HCV population. Red cell aplasia has been reported during erythropoietin therapy in at least one hepatitis C patient, and rapid rises in hemoglobin or targeting higher hemoglobin levels have been associated with an increased tendency to thrombotic episodes, a concern that is reflected in the package labeling for these agents.10 The absence of large randomized trials and the low likelihood that these trials will ever be performed means we have limited understanding of the frequency of these side effects.

In this era of evidence-based medicine, the assertion that growth factor use in this setting is appropriate is rather perplexing. The analysis by Del Rio and colleagues ultimately recommends the strategy with darbepoetin, but no prospective trials of darbepoetin use among patients receiving HCV therapy have been published. Typically, we would prove that a strategy was safe and effective before determining if the strategy was cost-effective. In keeping with this dilemma, the most recent published HCV practice guidelines and technical review were also unable to assign solid recommendations for growth factor use in HCV patients during treatment, based on the available evidence.11, 12

So where do we go from here? With relatively common use of growth factors in practice, can we expect further studies regarding their use? The Shiffman study has so far only been reported as an abstract and will require closer scrutiny when published, but this study would seem to highlight the need for further investigation.8 The analysis by Del Rio and colleagues is rather limited in scope and uses the reference case of a Caucasian patient who is naïve to HCV therapy. The response rate of African Americans is well below the SVR threshold of 50% discussed in their sensitivity analysis. This analysis also does not address some of the more challenging patient groups. Patients co-infected with HIV/HCV and those with recurrent HCV after liver transplantation appear especially likely to experience anemia in the course of HCV treatment. Previous studies for both groups have generally used conservative doses of ribavirin, yet dose reductions and omission of ribavirin doses have been common.13–16 Erythropoietin use has been permitted in these studies, but no trial has definitively answered if growth factors improve virologic outcomes. The low-dose ribavirin strategy opposes our experience with HCV mono-infected patients. Higher doses of ribavirin would likely require growth factors to maintain the ribavirin dosing, and such studies are clearly needed in these patient groups. If and when such studies are conducted, we would propose an algorithm such as that detailed in Fig. 1, which borrows from the experience of the previous studies in HCV mono-infected patients.6, 7 An alternative algorithm for practitioners that is clinically useful and more conservative has been proposed and published by the Veterans Health Affairs Pharmacy Benefits Management Strategic Healthcare Group and Medical Advisory Panel (

Figure 1.

A proposed algorithm for the management of high-dose ribavirin-induced anemia and use of erythropoetin agents.

Although anemia has received the most attention and investigation, neutropenia and thrombocytopenia are other challenges during HCV antiviral therapy. Patients with portal hypertension may have both conditions, which can prevent a patient from receiving HCV therapy. Both neutropenia and thrombocytopenia can also lead to dose reductions and discontinuation of peginterferon-α, and the end result is lower SVR rates. G-CSF has also become widely used in HCV therapy.3 No randomized trial, however, has examined the effectiveness and safety of this medication in patients with HCV. Van Thiel and colleagues studied 30 consecutive patients with advanced hepatitis C treated with daily interferon and either G-CSF or dose reduction.17 G-CSF effectively increased the neutrophil count, and the group receiving G-CSF achieved SVR of 53%. Thrombocytopenia was initially treated with IL-11 (Neumega), and platelet levels increased from a mean of 143 × 103/μL at baseline to 198 × 103/μL at week 12 of treatment (P < .001) in one study.18 However, edema was a common concern raised with the medication, and patients with portal hypertension will remain at risk for edema. More recently, a promising thrombopoietin agonist was reported to be effective at increasing platelet counts in 24 of 28 patients with pre-existing significant thrombocytopenia, a maneuver that subsequently allowed initiation of antiviral therapy and the chance of obtaining an SVR in patients that were not candidates for therapy.19 As is the case with anemia, clear guidelines on appropriate initiation of therapy and monitoring parameters will be key to the development of these and future biological agents in HCV treatment guidelines. Given the potential expense of these medicines, they will also need to improve clinical outcomes including SVR rates, and other potential endpoints such as fewer infections with G-CSF therapy, and fewer bleeding complications with thrombopoietin agonists.

We want our patients to feel better. We want them to tolerate HCV therapy better, and growth factors appear to do that in the short term by improving blood counts, medication doses, and health-related quality of life during therapy. Is this level of “efficacy” sufficient to warrant widespread use of growth factors in the setting of HCV therapy?20 From a societal perspective, the resource utilization and economic effects of these therapies must be carefully considered, but from a practitioner and patient standpoint, therapies that make patients feel better in the short term seem to be a clinically relevant and important approach to patient management. Both viewpoints from either end of this spectrum seem justifiable, depending on the frame of reference. However, we believe enthusiasm will be dampened if this added expense does not result in improved virologic outcomes. Before we can recommend the widespread use of these therapies, we must therefore know that they affect SVR. We need appropriately powered large-scale clinical trials to address these endpoints, as well as to understand the key issues in terms of a large safety database, and the robustness of the effects on quality of life and adherence to HCV therapy. We simply do not know the answers to these questions at this time. If these medications are effective, the analysis by Del Rio and colleagues would suggest this is a cost-effective strategy. But the question remains, are growth factors effective?