Fw: NATAP: Peg/RBV-Rate of Infections Nor Neutropenia-Related
- NATAP http://natap.org/
Rate of Infectious Complications during Interferon-Based Therapy for Hepatitis C Is Not Related to Neutropenia
Clinical Infectious Diseases June 15, 2006;42:1674-1678
Curtis L. Cooper, Saif Al-Bedwawi, Craig Lee, and Gary Garber
University of Ottawa, Division of Infectious Diseases, The Ottawa Hospital-General Campus, Ottawa Health Research Institute, Ottawa, Ontario
The relationship between infectious complications and neutropenia was evaluated in recipients of interferon-based therapy for hepatitis C followed at The Ottawa Hospital Viral Hepatitis Clinic from June 2000 to May 2005. One hundred ninety-two patients received 211 courses of therapy (5707 person-weeks of therapy). No patients received granulocyte colony-stimulating factor. Sixty-seven infectious complications occurred in 57 patients (1.17 infections per 100 person-weeks of therapy). The median time to infection was 17 weeks after the start of therapy. Age, sex, weight, race, human immunodeficiency virus status, stage and grade of biopsy, and type of interferon were not correlated with infection rate by Cox regression analysis. The rates of total, fungal, viral, and bacterial infections did not correlate with nadir neutrophil count or magnitude of decrease from baseline. Neutrophil count is not correlated with infection rate in recipients of interferon-based therapy for hepatitis C. Reduction in interferon dose and/or dosing with granulocyte colony-stimulating factor in those with neutropenia is not supported by this analysis.
Infectious complications occurring in patients receiving IFN-based therapy for hepatitis C virus (HCV) infection are not well described. Clinical trial results describe severe infectious adverse events but often omit less serious infectious complications [1-4]. It is unclear what factors predict increased risk of infection for those receiving therapy. For example, it is reasonable to speculate that infectious complications are more frequent in persons with advanced liver disease, but the relative risk compared with those without cirrhosis is not well quantified.
Neutropenia is a common complication of IFN-based therapy for HCV infection. Although the relationship between IFN-induced neutropenia and risk of infection in this instance is unclear, reduction of IFN dose is often mandated in clinical trials [2, 5] and recommended in clinical practice  for those experiencing treatment-related decreases in neutrophil count. Data supporting therapeutic dose reduction in this clinical context are lacking. Furthermore, dose reduction of any sort must be well justified, because appropriate dosing with both pegylated IFN and ribavirin is critical to achieving maximal rates of sustained virological response . To address these issues, we conducted a retrospective study exploring the relationships between rates of infection and absolute neutrophil count and between risk of infection and absolute neutrophil count in patients with HCV infection who were receiving IFN-based therapy at a referral tertiary care viral hepatitis clinic.
All recipients of IFN-based antiviral therapy for HCV infection and who were followed for the entire course of treatment at the Ottawa Hospital Viral Hepatitis Clinic (Ontario, Canada) from June 2000 to May 2005 were identified using a computerized database. This database contains demographic and clinical information pertaining to all patients evaluated in this clinic. Collection and use of this data for analysis was reviewed and approved by The Ottawa Hospital Research Ethics Board.
Treatment-related information, including HCV load (IU/mL; Cobas Amplicor HCV Monitor 2.0; LaRoche) and genotype (Inno-LiPA; Innogenetics), type of IFN-based therapy (i.e., IFN-a2b, pegylated IFN-a2a, or pegylated IFN-a2b), and duration of therapy, are contained in this database. Laboratory measures, including alanine aminotransferase, aspartate aminotransferase, and hemoglobin levels, as well as neutrophil and platelet counts at baseline and during the course of therapy, including follow-up at 1 and 6 months after completion of therapy, are collected. Information pertaining to HIV serostatus, hepatitis B virus infection, and liver biopsy data are contained in the database. Information about the use of erythropoietin is captured. Of note, it is not standard practice to use granulocyte colony-stimulating factor (G-CSF) in our clinic. Reduction of IFN dose because of clinical adverse effects or laboratory abnormalities is rarely practiced because of the adverse effect of this on treatment outcome.
Infections are routinely evaluated by history and physical examination and documented on the patient chart at each clinic visit. For the purposes of this analysis, infections were initially identified by clinical chart assessment by a single reviewer (S.A.-B.) and confirmed by a second reviewer (C.L.C.). In instances of classification uncertainty, the final determination was made by consensus. Infectious adverse events were classified as severe if death, hospitalization, or severe disability resulted. Patients receiving only a portion of their therapy at our clinic were not included in this analysis, because it was not possible to accurately identify all treatment-related infections for these persons.
Descriptive analysis was conducted. Dichotomous variables were compared by x2 analysis. The rate of infection was calculated by taking the total number of infections and dividing by the person-weeks of IFN-based therapy received. Specific rates of infection were also calculated for viral, fungal, and bacterial infections. Rates of infection were compared by age group, sex, weight, race, HIV serostatus, stage and grade of the liver biopsy score (Metavir), and type of IFN therapy, using Cox regression analysis. The rates of infections were also evaluated between those with neutrophil count nadirs of <1000 cells/uL and <750 cells/uL. The absolute size of decrease from baseline in neutrophil count was evaluated as a dichotomous (decrease of 1000, 2000, and 3000 cells/uL from baseline) and as a continuous variable. An exploratory analysis of the effect of erythropoietin on neutrophil count and rate of infection was performed. The analysis was conducted using SPSS software, version 11.0 (SPSS).
One hundred ninety-two patients received 211 courses of therapy, representing 5707 person-weeks of therapy. Eight additional patients were excluded from analysis, because they were not cared for at our clinic for the entire course of treatment. The characteristics of this population are described in table 1. All but 1 subject received ribavirin in combination with IFN therapy.
Sixty-seven infectious complications occurred in 57 patients (1.17 infections per 100 person-weeks of therapy). The median time to infection was 17 weeks of therapy (first and third quartiles, 7 and 22 weeks). Forty percent of infections involved the respiratory tract. The majority of other infections were cutaneous (16%), were based in the oral cavity (15%), were genitourinary (12%), or were gastrointestinal (12%) (figure 1). Twenty-eight (42%) of 67 were microbiologically confirmed or suspected bacterial infections (7 cases of bronchitis and/or pneumonia, 6 cases of wound and/or cellulites, 4 cases of dental infection, 4 cases of urosepsis). Twenty-eight (42%) of 67 infections were laboratory-confirmed or suspected to be of viral etiology (17 upper respiratory tract infections, 6 cases of viral gastroenteritis, 3 cases of cutaneous herpes simplex virus). Ten (15%) of the 67 infections were fungal in origin, including 6 cases of oral thrush and 4 mucocutaneous fungal infections. Three patients required oral antifungal therapy, and the others responded to topical management. The proportion of fungal infections was similar among patients with HIV-HCV infection (2 [7%] of 29 courses of treatment) and patients with HIV infection alone (8 [4%] of 182 courses of treatment) (P = .6). One case of pinworm was identified. Four bacterial infections (6%) were severe enough to require hospitalization, intravenous antibiotic therapy, and/or discontinuation of therapy. Severe infections included 3 cases of pneumonia identified at weeks 13, 19, and 20. Therapy was discontinued as a consequence in 2 of these 3 instances. In the fourth case of bacterial infection, a male with known compensated cirrhotic liver disease before initiation of treatment was admitted to the intensive care unit with new-onset ascites, group B streptococcal bacteremia, and respiratory failure after 11 weeks of therapy. HCV therapy was discontinued without achieving a sustained virological response. The patient recovered and is now awaiting liver transplantation.
Rates of infection did not differ with age, sex, weight, race, HIV status, stage and grade of liver biopsy, or type of IFN therapy when evaluated by Cox regression analysis (data not shown). The rate of infection did not differ among the 8% of patients who temporarily or permanently had their IFN dose reduced while undergoing treatment (OR, 0.48; 95% CI, 0.12-2.0; P = .3). The rate of infection was 1.26 infections per 100 person-weeks of therapy for those with biopsy-proven cirrhosis, compared with 1.27 infections per 100 person-weeks in noncirrhotic patients. Of note, all cirrhotic patients were classified according to the Child-Pugh scoring system as having class A at the start of therapy. The use of G-CSF for the prevention of infectious complications could not be evaluated, because no patients received this adjunctive therapy.
The mean (± SD) absolute neutrophil count decreased from a baseline count of 3800 ± 1700 cells/uL to a nadir of 1900 ± 1100 cells/uL. This nadir was reached by week 8 of therapy and remained relatively stable thereafter (figure 2). The total infection rate (figure 3) and rates of bacterial, viral, and fungal infections did not correlate with nadir neutrophil count (<1000 or <750 cells/uL) or with the magnitude of decrease from baseline (data not shown). A total of 26 (19%) of 138 pegylated IFN recipients and 5 (9%) of 54 standard IFN recipients had nadir neutrophil counts of <1000 cells/uL (P = .08). The total rate of infection among those with nadir neutrophil counts of <1000 cells/uL was 0.83 infections per 100 person-weeks of therapy per year, and the rate of infection among those with counts that decreased to <750 cells/uL was 0.75 infections per 100 person-weeks of therapy). Transient neutrophil counts of <500 cells/L occurred in 2 patients. Therefore, analysis of infection rates using this cutoff point was not possible. Neither of these patients developed infections while undergoing treatment. The incidence of severe infectious complications (n = 4) was 0.07 cases per 100 person-weeks of therapy. The neutrophil count never decreased to <1000 cells/uL during therapy for 3 of these 4 patients. In the fourth patient, the nadir neutrophil count was 700 cells/uL but was generally >1000 cells/uL during therapy. Pneumonia developed in this patient, who had a neutrophil count of 1200 cells/uL. In this cohort, <2% of subjects underwent IFN dose reduction in response to neutropenia. Therefore, we are confident that neutropenic patients received doses of IFN similar to those received by patients without neutropenia.
The influence of erythropoietin on rate of infection was evaluated. This medication was used during 27 courses of therapy. At the start of erythropoietin treatment, the median neutrophil count (± SD) was 1420 ± 913 cells/L. No consistent change in neutrophil count was observed with the use of this product. Ten (37%) of 27 erythropoietin recipients, compared with 47 (26%) of 184 nonrecipients, developed infectious complications (P = .21). The majority of infections (80%) in the erythropoietin group occurred before initiation of this medication.
Among recipients of high-dose chemotherapy for the management of hematologic malignancy, neutropenia is highly correlated with risk of infectious complications . In this patient group, G-CSF reduces the duration of neutropenia [8-10]. G-CSF has been demonstrated to be useful in reducing risk of infection in some [9, 11], but not all, studies [8, 10, 12]. This specific beneficial effect diminishes with increasing age. Overall mortality is generally not altered by the use of G-CSF [9, 13].
Neutropenia frequently complicates IFN-based antiviral therapy for hepatitis C . The relationship between neutropenia and infectious complications in this instance is not well described. In the absence of data, reduction in IFN dose and/or G-CSF dose for those who develop neutropenia while undergoing IFN-based therapy for HCV infection are often initiated in clinical practice, as suggested in treatment guidelines  and mandated in most clinical trials . Reduction of IFN dose is well accepted to have a detrimental effect on the rate of sustained virological response . Despite G-CSF use, there is little evidence that it results in increased sustained virological responses [16, 17] or a reduced rate of infection. Furthermore, it is expensive.
The immune suppression produced by high-dose chemotherapy used to manage hematologic malignancy is more broad-based than that of IFN. Furthermore, mucosal injury to the gut further compromises host defenses against infection. This likely explains the apparent difference in infection-related risk of neutropenia. Our study demonstrates that nadir neutrophil count and magnitude of decrease from baseline are not correlated with rate of infection in recipients of IFN-based therapy for HCV infection. With this is mind, the use of G-CSF to reduce the risk of infection in this instance is not supported. These findings are corroborated by others who likewise did not identify neutropenia as a risk factor for infections during treatment with IFN-a2b and ribavirin . In this study, a bacterial infection rate of 18% (22 of 119) was reported, which was a similar rate to that in our cohort (28 [13%] of 211 courses of therapy).
Our study was in partial agreement with that of Puoti et al. . In their analysis, 31 treatment-related infections were recorded in 255 patients treated with ribavirin plus standard (n = 103) or pegylated IFN (n = 152). This cohort was similar to our own in most key baseline characteristics (i.e., age, sex, alanine aminotransferase level, neutrophil count, advanced fibrosis [17% of patients vs. 26% of patients in our cohort]). None of the nonrespiratory infections were observed in patients with neutropenia, and it was not identified as a predictor of infection by multivariate regression analysis. Puoti et al.  reported that acute respiratory infections (n = 8) were associated with neutropenia, which is in contrast to our findings. As in our work, neutropenia (neutrophil count, <1000 cells/L) was more common among patients treated with pegylated IFNs (48% vs. 9%; P = .0009).
The use of erythropoietin to manage IFN- and ribavirin-induced anemia is supported . In addition, there is some evidence that thrombocytopenia may respond to erythropoietin . We conducted an exploratory evaluation to determine if erythropoietin also influences neutrophil count or function. No consistent reversal of neutropenia or stabilization of neutrophil count was observed in recipients of therapy for HCV infection. Only 2 infections were observed in patients while being treated with erythropoietin. However, small numbers as well as different times of initiation and duration of erythropoietin treatment preclude conclusion that rates of infection are lower in erythropoietin recipients.
These findings are generated from a comprehensive and complete database on all patients receiving IFN-based antiviral therapy for HCV infection at the Ottawa Hospital Viral Hepatitis Clinic and, therefore, are robust. Our cohort is ideal for evaluation of the relation between therapy, neutropenia, and infection, because no patients received G-CSF, and reduction of the IFN dose, considered an option of last resort, is rarely recommended by our clinicians. Nevertheless, we acknowledge several limitations. After the initiation of therapy, patients were seen in the clinic at weeks 2, 4, 6, and 8; thereafter, they were seen monthly until the completion of therapy and at months 1 and 6 after treatment was completed. It is possible that minor infectious complications were not identified if they occurred between these evaluation times. We are confident that a minimum of minor infectious complications was unidentified, because it is routine practice in our clinic to question patients about any infectious signs or symptoms occurring since their last evaluation. The patients were also asked to report any signs or symptoms of infection to our clinical nurse between visits. Although we assume that most upper respiratory tract and gastrointestinal infections are viral in nature, specimens for culture for viruses were not routinely collected. Duration of symptoms and severity of infection by quantitative determination were not evaluated, given the retrospective nature of this study. It is possible that these characteristics of infection may be influenced by nadir or absolute change in neutrophil count. A control group of patients with HCV infection followed at the same intervals but not receiving therapy for HCV infection would have provided a quantitative measure of the normal rate of infection, with which the results of this study could have been compared.
Despite these concerns, this analysis demonstrates that systemic antiviral therapy for HCV infection is associated with a low risk of infectious complications. The majority of these infections are minor. There is an absence of correlation between the neutrophil count and rate of infection in recipients of IFN-based therapy for HCV infection. Even in patients with neutropenia, the majority of infections occurring during therapy were of minor consequence and did not require hospitalization or modification of treatment for HCV infection. We conclude that, in most cases, reduction in IFN dose and/or the use of G-CSF in those developing neutropenia is not warranted.
NATAP nataphcv mailing list -- nataphcv@...
This is an annoucement-only mailing list. Do not reply.
To unsubscribe: send a blank email to nataphcv-request@... with a subject of unsubscribe.
For more information, see http://seven.pairlist.net/mailman/listinfo/nataphcv
[Non-text portions of this message have been removed]