| |
Hepatitis C Virus Seromarkers (viral load) and Subsequent Risk of Hepatocellular Carcinoma: Long-Term Predictors From a Community-Based Cohort Study
|
| |
| |
Download the PDF here
JOURNAL OF CLINICAL ONCOLOGY, OCTOBER 20 2010, Mei-Hsuan Lee, Hwai-I Yang, Sheng-Nan Lu, Chin-Lan Jen, Shiou-Hwei Yeh, Chun-Jen Liu, Pei-Jer Chen, San-Lin You, Li-Yu Wang, Wei J. Chen, and Chien-Jen Chen. From the National Taiwan University; National Taiwan University Hospital; Genomics Research Center, Academia Sinica; MacKay Medical College, Taipei; and Kaohsiung Chang-Gung Memorial Hospital, Kaohsiung, Taiwan
"The cumulative hepatocellular carcinoma risk increased from 1.1% for HCV RNA seronegative status to 6.4% for low HCV RNA levels and to 14.7% for high HCV RNA levels (P < .001). The cumulative risk also increased with elevated serum ALT levels from 1.7% for persistently ≤ 15 U/L to 4.2% for ever more than 15 U/L but never more than 45 U/L and to 13.8% for ALT ever ≥ 45 U/L (P < .001). Having HCV genotype 1 was associated with a higher cumulative hepatocellular carcinoma risk (12.6%) than not having HCV genotype 1 (4.5%; P < .001)"
"Cox proportional hazards models were used to examine the associations between newly developed hepatocellular carcinoma and serum HCV RNA and ALT levels and HCV genotypes after adjustment for age, sex, cigarettes smoking, alcohol consumption, body mass index, and history of diabetes"
"Our study suggests that serum HCV RNA level is a strong long-term (> 5 years) risk predictor for hepatocellular carcinoma. It may be used for clinical decision on early treatment by antiviral therapy.....Active HCV RNA replication is associated with elevated serum ALT levels, suggesting that HCV-induced liver disease may be a result of immunologically mediated injury after inflammatory response rather than direct cytopathic effect of the virus.....Patients with chronic HCV with elevated serum HCV RNA levels, elevated ALT levels, or HCV genotype 1 infection have an increased risk of hepatocellular carcinoma and need early interventions........potential possibility of irreversible liver damage as a result of treatment delay should be balanced......The risk stratification of patients with HCV based on serum HCV RNA and ALT levels and HCV genotype is useful for clinical recommendations adapted to health care resource availability. Patients seropositive for anti-HCV have an increased risk of hepatocellular carcinoma and need to be periodically observed. In particular, patients with HCV with elevated serum HCV RNA and ALT levels or HCV genotype 1 infection should be recommended for more intensive monitoring of liver disease progression and clinical management to lower the viral load."
"Serum HCV RNA levels, HCV genotype, and serum ALT levels were independent risk predictors of hepatocellular carcinoma in the multivariate analysis, as shown in Table 2. Elevated serum ALT levels were associated with an increased risk of hepatocellular carcinoma. Compared with participants with undetectable serum HCV RNA levels, who served as the referent group, the multivariate-adjusted hazard ratios of developing hepatocellular carcinoma were 4.71 (95% CI, 1.36 to 16.35) and 7.81 (95% CI, 2.34 to 26.11) for participants with low and high serum HCV RNA levels, respectively (P < .001 for trend). Participants with high serum HCV RNA levels had a significantly higher risk of developing hepatocellular carcinoma than those with low serum HCV RNA levels, showing a multivariate-adjusted hazard ratio of 1.99 (95% CI, 1.11 to 3.56; P = .02). Among participants with detectable serum HCV RNA level, infection with HCV genotype 1 was associated with an increased risk for hepatocellular carcinoma (multivariate-adjusted hazard ratio = 2.28; 95% CI, 1.10 to 4.70)."
"Multivariate-adjusted hazard ratios of developing hepatocellular carcinoma for combinations of serum HCV RNA and ALT levels and HCV genotype are listed in Table 3. Participants with detectable HCV RNA levels, elevated serum ALT levels, and genotype 1 HCV infection had the highest risk of hepatocellular carcinoma. Compared with participants with undetectable HCV RNA levels and low serum ALT levels, who served as the referent group, the multivariate-adjusted hazard ratio was as high as 21.87 (95% CI, 5.09 to 93.95)."
Abstract
Purpose Hepatitis C virus (HCV) contributes to one third of hepatocellular carcinoma cases worldwide. Long-term predictors for HCV-related hepatocellular carcinoma are essential for early intervention. Serum HCV RNA and ALT levels and HCV genotype were assessed for their predictability of hepatocellular carcinoma risk.
Methods A prospective cohort of 925 participants positive for antibodies against HCV and age 30 to 65 years was recruited and followed from 1991 to 2006. Serum HCV RNA and ALT levels and HCV genotypes at enrollment and during follow-up were examined. Newly developed hepatocellular carcinoma was identified by health examination and computerized linkage with national cancer registration and death certification profiles. Multivariate adjusted hazard ratios with 95% CIs were estimated using Cox regression models.
Results Fifty-five participants newly developed hepatocellular carcinoma during 8,476 person-years of follow-up, giving an incidence rate of 648.9 per 100,000 person-years. The cumulative hepatocellular carcinoma risk increased from 1.1% for HCV RNA seronegative status to 6.4% for low HCV RNA levels and to 14.7% for high HCV RNA levels (P < .001). The cumulative risk also increased with elevated serum ALT levels from 1.7% for persistently ≤ 15 U/L to 4.2% for ever more than 15 U/L but never more than 45 U/L and to 13.8% for ALT ever ≥ 45 U/L (P < .001). Having HCV genotype 1 was associated with a higher cumulative hepatocellular carcinoma risk (12.6%) than not having HCV genotype 1 (4.5%; P < .001).
Conclusion Elevated serum levels of HCV RNA and ALT and HCV genotype 1 infection are independent risk predictors of hepatocellular carcinoma. These findings have strong implications for the management of chronic HCV.
INTRODUCTION
Hepatitis C virus (HCV) infects approximately 170 million people worldwide.1 It contributes significantly to the increasing number of patients who undergo liver transplantation in the United States.2 Chronic HCV can progress to hepatocellular carcinoma,3 with a total of 195,000 cases worldwide.4 As a result of successful hepatitis B virus vaccination programs,5 HCV-related health burdens are emerging quickly in Asian countries where hepatitis B virus infection is endemic.6
Current guidelines in the management of chronic HCV recommend antiviral therapy for patients with detectable HCV RNA and liver histologic abnormalities without other contraindications.7 However, current treatment is not completely satisfactory.8 A recent new antiviral agent seems to improve the treatment response rate and allows a substantial reduction in treatment duration, but its adverse effects may reduce the adherence.9 Only a proportion of patients infected with HCV experience progression to hepatocellular carcinoma; thus, it is crucial to elucidate predictors capable to identify high-risk patients who need intensive care.
Serum HCV RNA level is an important seromarker for the diagnosis of active HCV infection and the monitoring of virologic response to antiviral therapy.10 HCV is classified into six major genotypes that may be used to determine the duration of therapy and the dosage of ribavirin.11 Both biomarkers have been recognized as prognostic factors for treatment,12 whereas their associations with hepatocellular carcinoma risk remain inconclusive.13-17 Serum ALT level, a biomarker for liver inflammation, is widely used in clinical practice.18 Its long-term predictability for hepatocellular carcinoma development in patients infected with HCV remains to be evaluated.
Early HCV infections are often asymptomatic until irreversible liver damage emerges. To identify high-risk population who should be recommended for clinical management, we conducted a community-based cohort study to evaluate the long-term predictability of serum HCV RNA and ALT levels and HCV genotypes for hepatocellular carcinoma development.
METHODS
Study Cohort Enrollment
Methods of enrollment and follow-up of this community-based prospective study have been described previously.19,20 The flow of study participants is shown in Figure 1. Among 89,293 invited residents living in seven townships in Taiwan, 23,820 provided written informed consent for interview, health examination, and blood collection. Participants who were seropositive for either anti-HCV or hepatitis B surface antigen at study entry were observed by abdominal ultrasonography and serologic tests until the end of 2005. Participants who were seropositive for hepatitis B surface antigen, patients with hepatocellular carcinoma and liver cirrhosis at study entry, or participants observed for less than 5 years were excluded in the analysis of risk predictors. The participants who developed hepatocellular carcinoma within 5 years after enrollment were consequently excluded because they were not at risk for hepatocellular carcinoma after 5 years of enrollment. In addition, participants who died within 5 years after enrollment would not develop hepatocellular carcinoma after 5 years of enrollment (at risk) and thus were also excluded. Finally, 925 participants who were seropositive for anti-HCV were included in the analysis. Cumulative risk of hepatocellular carcinoma of 18,172 participants who were seronegative for anti-HCV was also calculated for comparison purpose. The study protocol was approved by the Institutional Review Board of the College of Public Health, National Taiwan University (Taipei, Taiwan).
Interview and Blood Collection
All participants were interviewed using a structured questionnaire by public health nurses. Collected information included sociodemographic characteristics, dietary habits, cigarette smoking and alcohol consumption habits, and history of major diseases. All participants were invited for health examinations and blood tests regularly until the end of 2005. Blood samples were obtained from each participant using disposable needles and vacuum syringes. The serum samples were fractionated within 6 hours after collection and stored at -70°C until they were assayed.
Laboratory Examinations
Tests on hepatitis B surface antigen, serum ALT level, and anti-HCV were performed using commercial kits, and laboratory procedures were described previously.19 Serum samples positive for anti-HCV were further measured for HCV RNA by the COBAS TaqMan HCV test, v2.0 (Roche Diagnostics, Indianapolis, IN), with a linear range from 25 to 3.9 x 108 U/mL. Serum samples with detectable HCV RNA were examined for HCV genotypes by LightCycler-based (Roche Diagnostics) polymerase chain reaction and melting curve analysis, which could effectively differentiate different HCV genotypes by showing different melting temperatures. In our test, genotype 1 and non-1 status determination can achieve a 100% accuracy by the current design.21,22
To check the possible degradation of serum HCV RNA during long-time storage, adequate serum samples at last follow-up of 697 participants (75%) were retrieved for HCV RNA test. The median time interval between study entry and last follow-up was 11 years. Among these participants, 218 (31.3%) and 208 (29.8%) had undetectable HCV RNA at study entry and last follow-up, respectively. There was a significant correlation between HCV RNA levels at study entry and last follow-up (r = 0.76; P < .001). The mean, median, and quartiles of detectable serum HCV RNA levels at study entry were consistently lower than those at last follow-up by one log10, indicating a homogeneous decline of HCV RNA levels after long-term storage. In this analysis, the median of detectable serum HCV RNA levels (2.4 x 104 U/mL, equivalent to 3.5 x 105 U/mL in consideration of the degradation from long-term frozen storage) at study entry was used to classify participants with high and low HCV viral load.
Ascertainment of Newly Developed Hepatocellular Carcinoma
Patients with newly developed hepatocellular carcinoma were identified by follow-up health examination and computerized linkage with National Cancer Registration profiles from January 1, 1991 through December 31, 2006. To ensure complete ascertainment, computerized linkage with National Death Certification profiles was also used to identify deaths from hepatocellular carcinoma. Hepatocellular carcinoma was confirmed in patients by histopathology, two imaging techniques (abdominal ultrasonography, angiogram, or computed tomography), or one imaging technique plus a serum α-fetoprotein level of ≥ 400 ng/mL.23
Statistical Analysis
Because the assumption of linear increase in hepatocellular carcinoma risk with the continuous increase in serum HCV RNA level might not be justifiable, we used the median level as the cutoff point to assure an equal number of participants allocated in the low and high serum HCV RNA groups. Because serum ALT levels at study entry and follow-up examination may fluctuate over time, we combined serial measurements of serum ALT levels in the following groups to predict the risk of hepatocellular carcinoma: persistently ≤ 15 U/L, ever more than 15 U/L but never more than 45 U/L, and ever ≥ 45 U/L.
In this study, we aimed to elucidate the long-term predictability of HCV seromarkers. Because serum HCV RNA and ALT levels may be affected (probably lowered) by the end-stage liver diseases, the predictability of these seromarkers may be underestimated. Therefore, we analyzed the risk starting at the fifth year to avoid the biased estimation. The person-years of follow-up were calculated for each person as the time from the fifth year after enrollment to the date of hepatocellular carcinoma identification, the date of death, or December 31, 2006, whichever came first. Incidence rates of hepatocellular carcinoma per 100,000 person-years were calculated by dividing the number of patients with newly developed hepatocellular carcinoma by person-years of follow-up. Kaplan-Meier method curves were used to depict cumulative risks for hepatocellular carcinoma by follow-up years, and the significance of their differences were assessed using the log-rank test.
Cox proportional hazards models were used to examine the associations between newly developed hepatocellular carcinoma and serum HCV RNA and ALT levels and HCV genotypes after adjustment for age, sex, cigarettes smoking, alcohol consumption, body mass index, and history of diabetes. Hazard ratios with 95% CIs were estimated to assess the magnitude of the associations between risk predictors and hepatocellular carcinoma. The proportionality assumption of Cox models was examined, and the assumption was not violated. All analyses were performed using the SAS statistical software package (version 9.1; SAS Institute, Cary, NC).
RESULTS
Newly developed hepatocellular carcinoma occurred in 55 of 925 participants who were seropositive for anti-HCV during 8,476 person-years of follow-up. Among the patients who developed hepatocellular carcinoma, 38 (69%) were confirmed by histopathologic examination.
The incidence rates of hepatocellular carcinoma by sex, age, serum ALT and HCV RNA levels, and HCV genotypes are listed in Table 1. The incidence rates were similar in men and women (641.8 and 659.1 per 100,000 person-years) and increased with increasing age at study entry. The incidence rates per 100,000 person-years were 163.2, 347.4, and 1,256.8 for serum ALT levels of persistently ≤ 15 U/L, ever more than 15 U/L but never more than 45 U/L, and ever ≥ 45 U/L, respectively; the incidence rates were 107.0, 629.9, and 1207.7 for undetectable, low, and high serum HCV RNA levels at study entry, respectively. HCV genotype 1 was associated with a higher incidence rate of hepatocellular carcinoma than genotype non-1 (1,035.0 v 430.5 per 100,000 person-years, respectively).
The cumulative risks of hepatocellular carcinoma after the fifth year of enrollment by baseline serum HCV RNA levels, serum ALT levels, and HCV genotypes are shown in Figure 2. At the end of follow-up, the cumulative risk was only 0.4% for participants who were seronegative for anti-HCV. There was an increasing cumulative risk of hepatocellular carcinoma for anti-HCV- seropositive participants with undetectable, low, and high serum HCV RNA levels (1.1%, 6.4%, and 14.7%, respectively; P < .001 for trend). Among participants who were seropositive for anti-HCV, the cumulative risks were 1.7%, 4.2%, and 13.8% for serum ALT levels of persistently ≤ 15 U/L, ever more than 15 U/L but never more than 45 U/L, and ever ≥ 45 U/L, respectively (P < .001 for trend).
Among participants with detectable serum HCV RNA levels, the cumulative incidence was 12.6% for HCV genotype 1 and 4.5% for genotype non-1 (P < .001).
Serum HCV RNA levels, HCV genotype, and serum ALT levels were independent risk predictors of hepatocellular carcinoma in the multivariate analysis, as shown in Table 2. Elevated serum ALT levels were associated with an increased risk of hepatocellular carcinoma. Compared with participants with undetectable serum HCV RNA levels, who served as the referent group, the multivariate-adjusted hazard ratios of developing hepatocellular carcinoma were 4.71 (95% CI, 1.36 to 16.35) and 7.81 (95% CI, 2.34 to 26.11) for participants with low and high serum HCV RNA levels, respectively (P < .001 for trend). Participants with high serum HCV RNA levels had a significantly higher risk of developing hepatocellular carcinoma than those with low serum HCV RNA levels, showing a multivariate-adjusted hazard ratio of 1.99 (95% CI, 1.11 to 3.56; P = .02).
Among participants with detectable serum HCV RNA level, infection with HCV genotype 1 was associated with an increased risk for hepatocellular carcinoma (multivariate-adjusted hazard ratio = 2.28; 95% CI, 1.10 to 4.70).
Multivariate-adjusted hazard ratios of developing hepatocellular carcinoma for combinations of serum HCV RNA and ALT levels and HCV genotype are listed in Table 3. Participants with detectable HCV RNA levels, elevated serum ALT levels, and genotype 1 HCV infection had the highest risk of hepatocellular carcinoma. Compared with participants with undetectable HCV RNA levels and low serum ALT levels, who served as the referent group, the multivariate-adjusted hazard ratio was as high as 21.87 (95% CI, 5.09 to 93.95).
DISCUSSION
In this community-based long-term follow-up study, elevated serum HCV RNA and ALT levels and genotype 1 HCV infection were found to be independent risk predictors of hepatocellular carcinoma 5 or more years before its diagnosis. To ensure that our study participants received standard care, those who had abnormal serum levels of ALT and α-fetoprotein or abnormal ultrasonographic findings were referred to medical centers for further management. However, patients with chronic HCV in Taiwan rarely received antiviral treatment with interferon because of its high cost and adverse effects until October 2003, when patients with abnormal serum ALT levels (> 82 U/L) and moderate fibrosis proven by liver biopsy could be reimbursed for treatment by the National Health Insurance. Therefore, this cohort study may be considered a natural history study of chronic HCV.
The risk of hepatocellular carcinoma for anti-HCV-seropositive participants with undetectable serum HCV RNA levels was much lower than that for participants with detectable levels. This finding suggests that spontaneous seroclearance of HCV RNA is favorable, implying that patients with chronic HCV may be benefited from lowering serum HCV RNA levels to undetectable by effective treatment.24,25 Among participant with undetectable serum HCV RNA levels, the incidence rate of hepatocellular carcinoma was moderately high for those who ever had elevated serum ALT levels. This suggests that serum ALT level is an important seromarker in the clinical management for this group of patients who need to be monitored periodically.
Only 22% of participants who were seropositive for anti-HCV had high serum HCV RNA levels (≥ 105 U/mL) in this study. Compared with clinical patients enrolled onto another study,26 the serum HCV RNA levels of patients with chronic HCV in our study were relatively low. This may be a result of a higher proportion of asymptomatic participants and the possible degradation of serum HCV RNA levels after long-term storage. From the tests of serum HCV RNA levels at the last follow-up, we found a homogeneous degradation of serum HCV RNA levels. There might be a one log10 decline in serum HCV RNA level. The median value (2.4 x 104 U/mL) was equivalent to 3.5 x 105 U/mL in consideration of the degradation from long-term frozen storage.
Previous studies failed to find the association between serum HCV RNA levels and risk of hepatocellular carcinoma,15,17,27 which might have resulted from their methods of HCV RNA detection and study design. Individuals with low-level viremia could not be distinguished without a sensitive assay.15,17 Moreover, it is considered difficult to assess the predictability of serum HCV RNA levels for hepatocellular carcinoma at its diagnosis because late-stage liver disease status may interfere with RNA replication.27 Our study suggests that serum HCV RNA level is a strong long-term (> 5 years) risk predictor for hepatocellular carcinoma. It may be used for clinical decision on early treatment by antiviral therapy.
Active HCV RNA replication is associated with elevated serum ALT levels, suggesting that HCV-induced liver disease may be a result of immunologically mediated injury after inflammatory response rather than direct cytopathic effect of the virus. In this study, all of the ALT measurements at the study entry and follow-up examinations were included in the analysis. Elevated serum ALT levels were found to increase the risk for hepatocellular carcinoma, emphasizing the importance of monitoring ALT levels in the management of chronic HCV. The test of serum ALT level is relatively insensitive, and a substantial proportion of patients infected with HCV with normal ALT have been reported to have liver inflammation or significant fibrosis on liver biopsy.28 However, the value of this low-cost blood test is enhanced because of its long-term predictability for hepatocellular carcinoma in our relatively healthy and asymptomatic participants.
Among participants with detectable serum HCV RNA levels and normal ALT, patients infected with HCV genotype 1 had a much higher risk for hepatocellular carcinoma than those infected with HCV genotype non-1. Therefore, patients with chronic HCV with normal ALT should be considered for the HCV genotype test to guide clinical decisions. Most HCV genotype 1 in this study may be classified as HCV genotype 1b, which was found to be the most dominant subtype in Taiwan.29 Similarly, a previous study also indicated that patients with cirrhosis infected with HCV genotype 1b had significantly higher risk for hepatocellular carcinoma development.13 However, HCV genotypes were not found to be associated with serum HCV RNA level in our study. It is possible that some HCV variants may be associated with rapid liver disease progression, which needs further studies.
Additional risk factors, such as sex, habits of cigarette smoking or alcohol consumption, obesity, and history of diabetes, were assessed in this study. These risk factors were not found to be associated with hepatocellular carcinoma among patients with HCV. Previous studies indicated that male sex was associated with liver fibrosis progression or hepatocellular carcinoma.13,30 However, we did not find an association between sex and hepatocellular carcinoma. An effect of sex on hepatocellular carcinoma may result from the lifestyle variables such as alcohol consumption and cigarette smoking. Because these two risk predictors were adjusted in our study, the sex effect on hepatocellular carcinoma may be lowered. Most previous studies mainly enrolled patients from clinics or hospitals,13,30 but our study participants were enrolled from the community and were relatively healthy. The sex effect may be different in study populations at different clinical stage. Moreover, men had higher serum HCV RNA levels than women; thus, the sex effect on hepatocellular carcinoma in this study might be decreased after the adjustment for HCV viral load.
The major limitation of this community-based study was the lack of liver biopsy data; thus, data on advanced fibrosis or cirrhosis were not available. The occurrence of liver cirrhosis is an intermediate stage that occurs in the natural history of hepatocellular carcinoma among patients with chronic HCV.
Thus, information on liver cirrhosis was not included in the multivariate analysis. However, more than 80% of patients with newly developed hepatocellular carcinoma who were seropositive for anti-HCV in this study were found to have liver cirrhosis detected by abdominal ultrasonography and/or an increased ratio between serum levels of AST and ALT. In addition, the hepatocellular carcinoma risk might be better predicted if the exact time of HCV infection were available. Unfortunately, it could not be determined in this community-based study the same as in most previous studies.
Patients with chronic HCV with elevated serum HCV RNA levels, elevated ALT levels, or HCV genotype 1 infection have an increased risk of hepatocellular carcinoma and need early interventions. Unfortunately, patients with high HCV RNA levels or HCV genotype 1 have less favorable virologic response rates. Which patients with chronic HCV need antiviral treatment is still disputed, reflecting the complexities in the management of chronic HCV.31 Sustained response rate, treatment-related adverse effects, and potential possibility of irreversible liver damage as a result of treatment delay should be balanced. Our findings provide information that is helpful in the management of chronic HCV. The risk stratification of patients with HCV based on serum HCV RNA and ALT levels and HCV genotype is useful for clinical recommendations adapted to health care resource availability. Patients seropositive for anti-HCV have an increased risk of hepatocellular carcinoma and need to be periodically observed. In particular, patients with HCV with elevated serum HCV RNA and ALT levels or HCV genotype 1 infection should be recommended for more intensive monitoring of liver disease progression and clinical management to lower the viral load.
|
|
| |
| |
|
|
|
