METHODS

All recipients of IFN-ribavirin-based HCV antiviral therapy who were followed-up at The Ottawa Hospital Viral Hepatitis Clinic from October 2003 through October 2006 were identified using an SPSS version 14.0 (SPSS) clinical database. Data were analyzed starting from October 2003 because ESA use to manage anemia induced by HCV treatment became the standard of care at the Ottawa Hospital Viral Hepatitis Clinic at that time. Of note, ESA use is generally initiated in patients with hemoglobin levels of <100 g/L when accompanied by qualitative complaints of fatigue. The rate of decrease and the absolute size of decrease in hemoglobin levels have little influence on the decision to initiate ESA use in our clinic. It is noteworthy that standard ribavirin and pegylated IFN doses, according to product monograms, were used. Similar doses were used, irrespective of HIV infection status. Post-HCV treatment data on specific adverse events, including myocardial infarctions, strokes, hospitalization for congestive heart failure, deep vein thrombosis, pulmonary embolisms, malignancies, and death, were collected by the Ottawa Hospital medical record review. Collection and use of these data for analysis were reviewed and approved by the Ottawa Hospital Research Ethics Board.

A descriptive analysis of IFN-ribavirin recipients stratified by ESA use was conducted. Variables potentially influencing the eventual use of ESA—including hemoglobin level, age, and weight at baseline; HIV serostatus; and HCV genotype—were assessed by multivariate logistic regression. Sustained virological response rates were compared by χ² analysis. Rates of specific adverse events during receipt of therapy and in the period after completion of treatment were compared by χ² analysis between ESA recipients and nonrecipients.

RESULTS

ESA use and efficacy in recipients of HCV treatment. A total of 179 courses of therapy were administered to 174 patients. Five patients received 2 courses of HCV treatment during the period of evaluation. For the purpose of the analysis, the most recent course of HCV therapy was considered. This represented a total of 4331 person-weeks of IFN-ribavirin exposure. Baseline characteristics are summarized in table 1.

Thirty-four patients initiated use of an ESA while receiving HCV therapy, for a total of 360 person-weeks of ESA exposure. The median time at initiation of ESA use was week 12 (range, week 2 to week 30). Key predictors of ESA use by multivariate logistic regression analysis included older age (P = .02), lower weight (P = .04), lower baseline hemoglobin level (P = .01), and infection with HCV genotype 1 or 4 (P = .01) (table 2). Sex, race, and HIV infection status were not found to predict use of (or access to) ESAs by univariate logistic regression analysis.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1.

Baseline characteristics of recipients of hepatitis C virus (HCV) treatment, according to use of an erythropoiesis-stimulating agent (ESA).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2.

Baseline predictors of use of erythropoiesisstimulating agents in the overall cohort by multivariate logistic regression analysis.

By week 4 of ESA use, 76% of recipients achieved a therapeutic response, defined as prevention of further decrease in hemoglobin level (i.e., a decrease of >5 g/L) after administration of an ESA. Eighty-eight percent of ESA recipients achieved a hemoglobin level of >110 g/L after initiation of ESA use. Variables including age, sex, weight, baseline hemoglobin level, slope of hemoglobin level decrease, HIV serostatus, IFN formulation (i.e., pegylated IFN or other IFN), HCV genotype, and biopsy stage and grade did not predict response to ESAs at week 4 of treatment. The overall sustained virological response rate was 54% in ESA recipients and 45% in nonrecipients (P = not significant). Sustained virological responses in patients with hemoglobin levels that fell below 100 g/L were obtained in 55% of ESA recipients and 50% of nonrecipients (P = not significant). Differential ribavirin dose reduction could potentially influence recovery of hemoglobin level and treatment effectiveness. Ribavirin dose reduction, a practice of last resort by clinic policy, occurred in 8 (24%) of 34 ESA recipients and 6 (4%) of 140 nonrecipients (P < .001, by χ² test). This may have resulted in a reduction in the effect size of ESA to enable an improved (increased) rate of sustained virological response.

ESA safety during and after HCV therapy. The mean duration of follow-up after HCV therapy was 13 months (range, 0–42 months; 2226 person-months in total). The mean duration among ESA recipients was 16 months (range, 0–42 months; 554 person-months), and the mean duration among nonrecipients was 12 months (range, 0–42 months; 1672 person- months).

No severe adverse events or deaths were identified that could be attributed to ESA use. No patient had a myocardial infarction, deep vein thrombosis, or pulmonary embolism during receipt of HCV therapy or in the period of evaluation after completion of therapy. No patient was hospitalized for congestive heart failure. No patients developed diabetes mellitus or renal disease during receipt of HCV therapy or in the evaluation period after its completion. One nonrecipient had a lacunar stroke 24 months after completion of HCV therapy. Three patients developed cancer. One Asian man infected with HCV genotype 3 received a diagnosis of hepatocellular carcinoma 2 weeks after discontinuation of HCV therapy. He was not exposed to an ESA. Two cases of cervical squamous cell carcinoma were identified: one woman not exposed to ESAs received a diagnosis 6 months after completing 28 weeks of HCV therapy, and the second woman received a diagnosis <1 month after completing HCV treatment and 3 months after receiving the last of 9 weeks of ESA use. The events rate was 0.0018 events per person-month in both groups.

DISCUSSION

Anemia may occur in patients receiving HCV antiviral treatment as a result of impaired compensatory reticulocytosis due to IFN-induced bone marrow suppression and dose-dependent ribavirin hemolysis. Concurrent medications, comorbidities, coinfections, nutritional deficiencies, and substance abuse may further compound the problem [4–6, 12]. Therapy with erythropoietin, a glycoprotein growth factor and the primary stimulus of erythropoiesis, is a commonly used intervention to circumvent this complication in the HCV-infected population [12, 13]. ESAs, such as darbopoetin alfa (Aranesp; Amgen) and epoetin alfa (EPREX; Ortho Biotec), are often used for the management of chronic renal failure-associated anemia and nonmyeloid malignancies when the anemia is due to chemotherapy [14].

Concerns have arisen because of the higher rates of adverse events in ESA recipients in whom treatment led to hemoglobin levels of >120 g/L [8–10]. One prospective trial involving 1432 patients with chronic renal failure randomized to receive epoetin alfa, with a target hemoglobin level of either 113 or 135 g/L, demonstrated significantly increased rates of death, myocardial infarction, stroke, and hospitalization for congestive heart failure during a median period of 16 months in patients whose hemoglobin levels were targeted to the higher value [8]. Similarly, another randomized, controlled trial involving 939 women with metastatic breast cancer who were receiving chemotherapy randomized the patients to receive either weekly doses of epoetin alfa or placebo for up to 1 year. This study was terminated early because interim results showed an increased mortality rate at 4 months and a higher rate of fatal thrombotic events in the epoetin arm. At the time of study termination, the 12-month survival rate was also significantly lower in the epoetin arm [9]. As a consequence of these results and others [10], ESAs are no longer indicated for the treatment of anemia in patients with nonmyeloid malignancies when the anemia is due to the disease itself [4].

To our knowledge, our study is the first to explore the relationship between ESA use for IFN-ribavirin HCV treatment- induced anemia and rates of cardiovascular events, thrombosis, malignancy, and death. The mean follow-up period after ESA exposure in our HCV-infected population was 13 months, which is much longer than the follow-up period after ESA use in the evaluation of other disease states described above. We did not identify an increased risk of these adverse events during receipt of ESA or after completion of therapy. Despite the lack of evidence linking increased rates of adverse events with ESA use in this population, our personal experience is that some practitioners are extrapolating results from studies involving patients with cancer or chronic renal failure to the HCV-infected population. This may be inappropriate, given the inherent differences in these populations.

There are increasing data indicating that erythropoietin receptors are present in several cancer-cell lines, which may account for the increased rates of complications observed in patients with cancer who receive ESAs [15–17]. Malignancy itself is known to induce a hypercoagulable state. Patients with renal disease generally have more comorbidities than do patients with normal renal function, including cardiovascular disease and diabetes mellitus, and therefore would be expected to have higher rates of cardiovascular events and hospitalizations for congestive heart failure. Concurrent medications may also have altered pharmacokinetic profiles in patients with certain types of renal disease. Furthermore, in our HCV clinic, we use ESAs to maintain hemoglobin levels that will avoid the need for transfusions and that will diminish symptoms. We do not have targeted hemoglobin levels of >120 g/L, as was the practice in some studies that reported increased adverse event rates [8–10]. Thus, our patients clearly differ from those in the aforementioned studies. Withholding ESAs for HCV treatment-induced anemia may result in more patients requiring dose reductions or abandonment of HCV treatment altogether. This is worrisome, given the compelling relationship between suboptimal HCV treatment dosing and poor sustained virological responses.

Our results demonstrating that older age, lower weight, lower baseline hemoglobin level, and infection with HCV genotype 1 or 4 are predictors of ESA use make intuitive sense. Patients who are older, who weigh less, and who have lower baseline hemoglobin levels are more likely to have other comorbidities, to take more medications, and to have suboptimal nutritional intake. The higher dosage requirements of ribavirin for the treatment of infection with HCV genotype 1 or 4 likely account for the association between infection with these genotypes and greater use of ESAs.

Despite older age and a higher frequency of ribavirin dose reduction, a trend toward higher rates of sustained virological response among ESA recipients, compared with among nonrecipients, was observed in our evaluation. This finding is plausible because ESA recipients would likely have their anemia corrected and thus be able to continue to receive HCV treatment. We acknowledge that the finding of higher rates of sustained virological response among people who received ESAs may be influenced by survival bias (i.e., those individuals with significant adverse effects to IFN-ribavirin early in the course of treatment or those who did not achieve adequate early virological responses would have discontinued HCV medications prematurely). Consequently, they would not go on to develop profound anemia or to require ESAs. There are other limitations to consider when interpreting these results. A retrospective cohort design was employed, which is subject to issues related to incomplete data for some parameters.We retrieved data from a computer system that captures information from 2 major Ottawa hospitals. If our patients experienced adverse events or died at another hospital in the region, this information may not have been captured. The interaction between ESA use and other risk factors for cardiovascular events and malignancy, including age and smoking, was not considered in this analysis but merits further evaluation.

Despite these concerns, our evaluation validates the efficacy of ESAs in correcting anemia, thereby enabling patients to complete full courses of HCV therapy and receive full doses of medication. We demonstrate that ESA use is not associated with increased risk of cardiovascular events, thrombosis, malignancy, or death in an HCV-infected population, either during receipt of therapy or in the period after completion of HCV treatment. Although studies with larger sample sizes and with additional evaluating criteria for ESA use are desirable, the present study supports the safety of ESA use in patients with HCV treatment-induced anemia.