| |
Strong Association of Hepatitis C Virus (HCV) Infection and Thrombocytopenia:
|
| |
| |
Implications from a Survey of a Community with Hyperendemic HCV Infection
Clinical Infectious Diseases 2004;39:790-796
Chong-Shan Wang,1,2 Wei-Jen Yao,3 Shan-Tair Wang,2 Ting-Tsung Chang,4 and Pesus Chou5
1A-Lein Community Health Center, Kaohsiung County, 2Institute of Public Health, Department of Medicine, College of Medicine, 3Department of Nuclear Medicine, and 4Institute of Public Health, Department of Internal Medicine, Division of Gastroenterology, National Cheng Kung University Hospital, Tainan, and 5Community Medicine Research Center and Institute of Public Health, National Yang-Ming University, Taipei, Taiwan
ABSTRACT
The role of hepatitis C virus (HCV) infection in thrombocytopenia (defined as a platelet count of <100,000 platelets/L) is unknown. Our aim was to study the association between HCV infection and thrombocytopenia in a community where hepatitis B virus (HBV) and HCV infections are hyperendemic.
A community-wide survey of subjects 40 years old who had undergone a preventive health examination between April 1997 and July 2000 in A-Lein Township, Kaohsiung County, Taiwan. Serum blood platelet counts and HBV surface antigen (HBsAg) and antibody to HCV (anti-HCV) levels were measured. Abdominal sonography was performed on viral hepatitispositive participants.
Among the 1690 subjects, 70% were seronegative, 17.4% were anti-HCV positive, 9.2% were HBsAg positive, and 3.4% were coinfected with HCV and HBV. The mean platelet count in subjects with anti-HCV (180,000 platelets/L) was lower than in those with HBsAg (201,000 platelets/L) and in those without anti-HCV and HBsAg (234,000 platelets/L) (P < .001). The prevalence of thrombocytopenia was 1.3% among seronegative subjects, 1.9% among HBsAg-positive subjects, 5.2% among coinfected subjects, and 10.2% among anti-HCVpositive subjects. Multiple logistic regression analysis revealed that anti-HCV positivity (odds ratio, 6.0; 95% confidence interval, 3.211.2), an alanine aminotransferase level of 40 U/L, and age of 65 years were significantly associated with thrombocytopenia. The prevalence of thrombocytopenia among anti-HCVpositive subjects increased as the severity of liver disease increased, but, in HBsAg-positive subjects, thrombocytopenia presented only in those with advanced liver disease.
HCV infection is strongly associated with thrombocytopenia, which is correlated with hepatocellular damage and hepatic fibrosis. It is advisable to further check the hepatic condition of the patient, especially for HCV infection, if thrombocytopenia is present.
BACKGROUND
Thrombocytopenia is a common clinical problem that is discovered in daily laboratory testing and often leads to further examinations. The most well-known causes of thrombocytopenia are infectious diseases, such as HIV infection or viral infections in children. Hepatitis C virus (HCV) infection, a progressive liver disease, is the major cause of chronic liver disease in the world, but it is not listed as a cause of thrombocytopenia in textbooks or review articles. Recently, however, at least 3 studies have reported that HCV infection is associated with thrombocytopenia, but they involved either hospital-based populations or had small sample sizes, or they might have confounded HCV infection with HIV infection. To date, no community-based study has illustrated the distribution of thrombocytopenia and investigated its relationship to HCV infection or to HCV liver disease severity.
Our long-term clinical observation of a community with hyperendemic HCV and hepatitis B virus (HBV) infections led us to suspect that HCV infection might be strongly associated with thrombocytopenia. Therefore, this study aimed to elucidate this relationship in a community-based population comprising a wide spectrum of patients among whom liver disease due to HBV and HCV infection varied in severity.
AUTHOR DISCUSSION
This study in a community with hyperendemic HBV and HCV infection shows that HCV infection, but not HBV infection, is strongly associated with thrombocytopenia. Participants with HCV infection had a lower mean platelet count than did seronegative and HBsAg-positive participants. Moreover, thrombocytopenia is also highly associated with older age, elevated ALT levels, and abdominal sonography-determined liver disease severity.
The association between HCV infection and thrombocytopenia is unclear, but hepatic fibrosis might be central to it. In the current study, the prevalence of thrombocytopenia increased remarkably along with liver disease severity among participants with HCV infection; the prevalence was about 9 times higher among subjects with chronic hepatitis, cirrhosis, or hepatoma than among those with a normal or fatty liver. Even in participants with HBV infection, thrombocytopenia was found only in those with cirrhosis and hepatoma. In addition, the liver is the main site for the production of thrombopoietin, the dominant cytokine for controlling the development of megakaryocyte and platelet production. Thrombopoietin levels and platelet counts are highly correlated with liver-function impairment and the severity of hepatic fibrosis in chronic HCV infection. Other studies have shown increasing thrombopoietin levels and platelet counts after IFN therapy in patients with HCV infection or who have undergone liver transplantation. This indicates that thrombocytopenia in persons with HCV infection might be strongly associated with disease activity and long-term progression. In the current study, 2.3% of the anti-HCV-positive participants with normal liver sonographic findings, 5.1% of those with a fatty liver, and 20.3% of those with chronic hepatitis also had thrombocytopenia. Whether patients with thrombocytopenia need more-aggressive antiviral treatment to prevent progression to cirrhosis and hepatoma than do those with normal platelet counts merits further study. It is also potentially difficult to decide whether to administer IFN therapy to individuals with a very low platelet count.
Serum ALT levels are well-known indicators of hepatocellular damage. In the current study, an elevated serum ALT level was strongly associated with thrombocytopenia, even after adjusting for HBsAg positivity, anti-HCV positivity, and other risk factors. This indicates that thrombocytopenia might be highly correlated with hepatocellular damage. Therefore, it is reasonable to check the liver condition of patients with thrombocytopenia to determine whether they have viral hepatitis or other liver-associated diseases. In the current study, elevated serum ALT levels were at least 5 times more prevalent among individuals with HCV infection than among those who were seronegative but were only 1.8 times more prevalent among those with HBV infection. Elevated ALT levels are good predictors of fibrosis progression in chronic HCV infection. This indicates that individuals with HCV infection might have more hepatic inflammation and might be more likely to progress to advanced liver disease than those with HBV infection. Whether decreasing ALT levels in persons with thrombocytopenia will increase platelet counts and delay or prevent the progression of liver disease needs further study.
Our study also suggests that older persons (>65 years of age) are 4 times more likely than persons in other age groups to have thrombocytopenia. Among anti-HCV-positive individuals, older persons were 3 times more likely than persons in other age groups to have thrombocytopenia. This might be related to poor compensation for platelet production, especially in individuals with severe liver disease. Another possibility might be that older individuals in an area of HCV and/or HBV hyperendemicity have been infected with HCV and/or HBV longer and, therefore, are more likely to have significant liver impairment. The prevalence of thrombocytopenia in older persons with HCV infection was 6.6 times greater than that among older persons without HCV infection. Because most older individuals with anti-HCV and thrombocytopenia are asymptomatic, it would be better to check the status of HCV infection and to determine platelet counts for patients who are from communities of hyperendemicity or who are in high-risk groups before prescribing drugs that will have a tendency to induce hemorrhage or before performing invasive or surgical procedures.
The present study had some limitations. First, we did not examine drug-induced thrombocytopenia or autoimmune thrombocytopenia, both of which are very rare. Second, we did not conduct PCR for detection of HCV RNA to elucidate the relationship between the actual viral status or past infection, because 10%-15% of anti-HCV-positive participants have resolved infections, and their infections do not progress to chronic liver disease. The strength of the association between HCV infection and thrombocytopenia would have increased after removing participants from the study who tested positive for anti-HCV and negative for HCV RNA. Third, the severity of liver disease was classified by sonographic findings but was not confirmed by liver biopsy results. To perform such a procedure in a community-wide study was not possible because of ethical issues, risk of complications, and cost. Sonography can provide a noninvasive prediction of liver histological findings that has a high sensitivity (89%100%) and specificity (89%-93%), is a feasible practice in the community, and is appropriate for the screening and follow-up observation of patients with chronic liver disease or hepatic fibrosis.
In conclusion, this study of a community in which HBV and HCV infection is hyperendemic demonstrated that HCV infection is strongly associated with thrombocytopenia and that thrombocytopenia is also strongly associated with hepatocellular damage and hepatic fibrosis. This means that it is advisable to further check the hepatic condition, especially for HCV infection, in patients with thrombocytopenia. Whether thrombocytopenia is a good indicator for predicting the progression of viral hepatitis needs further study.
RESULTS
Mean platelet counts and prevalence of thrombocytopenia at baseline.
In the present study, 70.0% of the subjects were seronegative, 17.4% were anti-HCV positive, 9.2% were HBsAg negative, and 3.4% were coinfected. The mean platelet counts, which were rounded off, for anti-HCV-positive subjects (180,000 platelets/uL), coinfected subjects (179,000 platelets/uL), and subjects with an ALT level of >40 U/L (191,000 platelets/uL) were lower than those for the total study population (220,000 platelets/uL) (P < .001). The prevalence of a platelet count of <100,000 platelets/uL was 3.0% (51 of 1690 subjects) among the total study population, 1.2% among seronegative subjects, 1.9% among HBsAg-positive subjects, 5.2% among coinfected subjects, and 10.2% among anti-HCVpositive subjects. Univariate analysis revealed that participants with a platelet count of <100,000 platelets/L were more likely than those with a platelet count of >100,000 platelets/uL to be >65 years old (OR, 4.3; 95% CI, 2.010.7), to have an ALT level of >40 U/L (OR, 3.3; 95% CI, 1.86.1), and to have anti-HCV (OR, 8.8; 95% CI, 4.517.9) (P < .05). Sex, BMI (data not shown), habitual smoking, and habitual alcohol consumption were not significantly associated with thrombocytopenia.
Comparison of baseline data between seropositive and seronegative participants.
A comparison of the anti-HCVpositive participants with the seronegative participants showed that the anti-HCV participants had higher ALT levels (OR, 5.1; 95% CI, 3.86.8), but there were no significant differences with respect to age, sex, BMI, habitual smoking, or habitual alcohol consumption. A comparison of HBsAg-positive participants with seronegative participants showed that more HBsAg-positive participants were in the 40-46-year-old age group, had higher ALT levels (OR, 1.8; 95% CI, 1.12.7), and were habitual smokers (OR, 1.9; 95% CI, 1.22.9).
Comparison of baseline data of anti-HCVpositive participants to determine thrombocytopenia risk.
Univariate analysis of baseline data for the anti-HCVpositive participants revealed that those who were >65 years old (OR, 3.1; 95% CI, 1.29.5) or had an ALT level of >40 U/L (OR, 2.2; 95% CI, 1.05.4) were more likely to have thrombocytopenia. There were no significant differences with respect to sex, BMI, habitual smoking, or habitual alcohol consumption.
Multiple logistic regression analysis to determine thrombocytopenia risk.
After stratification by demographic data, hepatitis status, and health-related behavior in backward elimination from a multiple logistic regression analysis, anti-HCV positivity was revealed to be most strongly associated with thrombocytopenia (OR, 6.0; 95% CI, 3.211.2). ALT levels of >40 U/L (OR, 2.1; 95% CI, 1.13.9) and older age (>65 years; OR, 4.3; 95% CI, 2.09.5) were also strongly associated with thrombocytopenia.
Prevalence of thrombocytopenia, according to abdominal sonographydetermined liver disease severity, among viral hepatitispositive subjects.
Of the 507 seropositive participants, 468 (92.3%) had been checked with abdominal sonography. The prevalence of thrombocytopenia among subjects with anti-HCV increased from 2.3% among those with sonographic evidence of a normal liver (3 of 128 subjects) to 5.1% among those with a fatty liver (3 of 59), 20.3% among those with chronic hepatitis (12 of 59), and 31.8% among those with advanced liver disease (including liver cirrhosis and hepatoma; 7 of 22). The prevalence of thrombocytopenia among HBsAg-positive participants with sonographic evidence of either a normal liver, a fatty liver, and chronic hepatitis was 0%, compared with 33.3% among those with advanced liver disease (3 of 9 participants). The prevalence of thrombocytopenia among anti-HCV-positive participants with chronic hepatitis or advanced liver disease was also much higher than among those with a normal or fatty liver (OR, 9.2; 95% CI, 3.329.3). Both HBsAg-positive and anti-HCVpositive participants with advanced liver disease were associated with almost the same prevalence of thrombocytopenia (P > .1).
|
|
| |
| |
|
|
|
