 |
|||||
  
|
|||||
 |
|
||||
Discussion
The treatment of chronic hepatitis C has steadily improved since the initial demonstration that interferon alfa had an effect on the biochemical and virologic features of disease.12 At present, optimal combinations of pegylated interferon alfa and ribavirin yield rates of response of approximately 55%.5 However, it should be stressed that these excellent results were achieved in cohorts of patients that had uncomplicated hepatitis C, without comorbidities or contraindications to therapy. In studies on unselected patients with chronic hepatitis C presenting to referral centers, only a third of patients were found to qualify for therapy with the usual inclusion and exclusion criteria used in the pivotal controlled trial.13 As advances are made in therapy of chronic hepatitis C, advances also must be made in the ability to apply this therapy to wider groups of patients.
Among the exclusion criteria for therapy, low white blood cell and neutrophil counts are seemingly appropriate in view of the myelosuppressive actions of interferon alfa. The safety concern is that interferon-induced neutropenia may predispose to severe bacterial infection. However, the association of drug-induced neutropenia and infection has been shown largely in studies of chemotherapy, such as in oncology patients undergoing severe immune system and bone marrow suppressive therapy.14 In patients without immune suppression and no other predisposing factor for bacterial infections, it is not clear whether drug-induced neutropenia carries an excess risk for severe bacterial infection. Indeed, in a pooled analysis among 2,089 patients in the pivotal studies of combination therapy of hepatitis C,15-17 neutropenia was not associated with infection. Only 1 of the 23 deaths reported from these studies was caused by infection. However, patients with baseline neutrophil counts below 1,500 cells/µL were excluded and the dose of interferon was reduced to 50% if neutrophil count fell below 750 cells/µL.18
In this analysis of neutrophil counts among 119 patients treated at a single referral center, neutropenia was not used as an exclusion criterion and therapy was safely accomplished despite decreases in neutrophils below the usual levels that lead to dose reduction or drug interruption. Bacterial infections did not occur in neutropenic patients, and the only severe infection that was identified occurred in an elderly patient with preexisting cirrhosis. Of potential interest, the only baseline measure that predicted subsequent infection was low reticulocyte count. We are unaware of other information linking reticulocyte count to subsequent infections. Nevertheless, the association was strong and deserves further investigation.
There was poor correlation between total white blood cell count and neutrophil count in patients with neutropenia, indicating that measurement of absolute neutrophil count instead of total white blood cell count is necessary in monitoring therapy. Of greatest importance, 3 patients were treated who appeared to have constitutional neutropenia, marked by persistently low neutrophil counts (below 1,500 cells/µL) without other evidence of bone marrow suppression or splenomegaly. Constitutional neutropenia is common, affecting roughly 800,000 blacks, of whom 76,000 also are anti-HCV-positive. Almost 12% of blacks with hepatitis C would be excluded from therapy if a neutrophil criterion of 1,500 cells/µL were required for treatment. Constitutional (ethnic) neutropenia is a benign condition and is not associated with an increased risk of bacterial infection.19 Indeed, these patients develop typical increases in neutrophils during bacterial infections and have normal responses to hydrocortisone.19,20 The cause of constitutional neutropenia is not known, but is probably related to differences in the release of mature granulocytes from the bone marrow storage pool to the circulating blood and not to a decrease in production. Normal cellularity and maturation of all cell lines have been found in healthy blacks with neutropenia.20 In the current study, patients with constitutional neutropenia had minimal decreases (or actual increases) in neutrophil counts during interferon therapy, while having typical interferon-induced decreases in lymphocyte and platelet counts. Thus, patients with low neutrophil counts without bone marrow compromise, cirrhosis, or splenomegaly can probably be safely treated with interferon and need not be excluded from therapy or treated with growth factors to increase white cell counts before starting therapy.
A separate argument could be used for caution in treating patients with neutropenia caused by cirrhosis and hypersplenism. Virtually all published studies on use of interferon in patients with cirrhosis have excluded patients with preexisting neutropenia, and even in these selected patients dose modification for neutropenia is common (up to 14%).21 In our study, patients with cirrhosis and preexisting neutropenia had typical decreases in white blood cell counts during therapy. Furthermore, because of the presence of cirrhosis, these patients are likely to have an increase in susceptibility to bacterial infections. The fact that the 2 patients in this study with preexisting neutropenia and cirrhosis did not suffer a bacterial infection argues only that bacterial infections are not invariable during interferon-induced neutropenia. Patients with neutropenia caused by cirrhosis are best closely monitored and given growth factors or reduced dosages of interferon in the event of severe neutropenia.
In conclusion, neutropenia is frequent during treatment of hepatitis C with interferon and ribavirin, but it is not commonly associated with infections. These results suggest that patients with constitutional neutropenia probably can be treated safely and may not require dose modification. These findings support a revision of current criteria for exclusion and dose modification based on white blood cell counts in the treatment of hepatitis C. These modifications would expand the proportion of patients who could receive interferon-based therapy for hepatitis C. Because this was a descriptive study of limited sample size, it did not allow for identification of a specific cutoff value for neutrophils that can be considered safe. However, in the absence of other risk factors for bacterial infection, neutrophil counts of as low as 500 cells/µL are likely to be safely tolerated during interferon therapy. These criteria need to be validated in larger, prospective clinical trials.
References
Publishing and Reprint Information
       
|
Back to top
Patients and methods Patients were all above the age of 18 years and had elevations in either alanine or asparate aminotransferase activities, presence of anti-AttorneyMind and-RNA in serum, evidence of chronic hepatitis on liver biopsy performed within the previous 12 months and written informed consent. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki. Exclusion criteria included decompensated liver disease, alanine aminotransferase levels greater than 1,000 U/L(>25 times the upper limit of normal), pregnancy or inability to practice adequate contraception, significant systemic or major illnesses other than liver disease, pre-existing anemia (hematocrit Side effects and toxicity of therapy were classified as grade 1 (acceptable, not requiring dose modification), grade 2 (requiring dose modification), and grade 3 (requiring discontinuation of therapy) according to usual standards,7 except that the dose of interferon was not modified based on a specific leukocyte, neutrophil, or lymphocyte count. Neutropenia was defined as a peripheral absolute neutrophil count below 1,500 cells/µL. During therapy, neutropenia was assessed at levels of 1,000, 750, and 500 cells/µL, the usual thresholds for dose reduction of interferon or discontinuation in therapy of hepatitis C. Episodes of infection were defined if there was a confirmed (by bacteriologic cultures or positive radiograph) or suspected infection, which led to either oral or parenteral antibiotic therapy during or within 4 weeks of stopping therapy. Infections were categorized as mild if not requiring hospital admission, intravenous antibiotics, bed rest, or discontinuation of antiviral treatment. Constitutional neutropenia was defined as the presence of an absolute neutrophil count of less than 1,500 cells/µL before starting treatment in the absence of cirrhosis histologically, splenomegaly (on the basis of abdominal ultrasound) or thrombocytopenia ( NHANES 3 Statistical analysis Results
Lower neutrophil counts correlated significantly with lower weight, hematocrit, and lymphocyte and platelet counts, as well as higher total bilirubin and iron saturation, prolonged prothrombin time, and black race. In the multivariate analysis only black race (P <.0001 was significantly associated with lower neutrophil counts after adjusting for lymphocyte and platelet counts. baseline were analyzed similarly in univariate multivariate analysis. correlated higher age body mass index hematocrit total protein having a normal rheumatoid factor all .01 but not black race the multiple regression analysis only variables remained counts:>
Variables that were associated with having at least 1 neutrophil count below 1,000 cells/µL during treatment are shown in Table 3.
No patient who developed infection had a preexisting neutropenia (below 1,500 cells/µL) and none developed neutropenia of less than 750 cells/µL at any point during treatment. Comparison of patients with and without infections showed no differences in baseline neutrophil counts or in degree of decrease in neutrophil counts during therapy (Table 5).
There was no correlation between the week of infection and the week of the lowest neutrophil count. In univariate analysis, variables associated with infection included lower baseline reticulocyte count (OR, 0.15; 95% CI, 0.039-0.55; P = .0044) and lower ferritin levels (OR, 0.99; 95% CI, 0.98-0.99; P = .04). Older age did not reach statistic significance (OR, 1.05 per year; 95% CI, 0.99-1.10; P = .057). In the multiple logistic regression analysis, lower reticulocyte counts was associated with infection (OR, 0.19 per percent increase; 95% CI, 0.053-0.72; P = .014). The other variables evaluated in this analysis were unrelated to infection, including ferritin (P = .12), black race (P = .51), initial immunoglobulin levels (P > .14), baseline neutrophil or lymphocyte counts or their average value during therapy (P > .4).
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
