This study was supported in part by the 21C Frontier Functional Human Genome Project from the Ministry of Science & Technology, Korea (FG-08-12-05).
Sustained low hepatitis B viral load predicts good outcome after curative resection in patients with hepatocellular carcinoma
Article first published online: 28 JUN 2010
© 2010 Journal of Gastroenterology and Hepatology Foundation and Blackwell Publishing Asia Pty Ltd
Journal of Gastroenterology and Hepatology
Volume 25, Issue 12, pages 1876–1882, December 2010
How to Cite
An, H. J., Jang, J. W., Bae, S. H., Choi, J. Y., Cho, S. H., Yoon, S. K., Han, J. Y., Lee, K. H., Kim, D. G. and Jung, E. S. (2010), Sustained low hepatitis B viral load predicts good outcome after curative resection in patients with hepatocellular carcinoma. Journal of Gastroenterology and Hepatology, 25: 1876–1882. doi: 10.1111/j.1440-1746.2010.06416.x
- Issue published online: 23 NOV 2010
- Article first published online: 28 JUN 2010
- Accepted manuscript online: 28 JUN 2010 12:00AM EST
- Accepted for publication 14 June 2010.
- hepatitis B virus;
- hepatocellular carcinoma;
- viral load
Background and Aim: Little is known about the role of hepatitis B virus (HBV) factors in the long-term prognosis of hepatocellular carcinoma (HCC) after resection. The objective of the present study was to identify the changing patterns of HBV levels and its effect on outcome after resection.
Methods: This study recruited 188 patients with HBV-related HCC who underwent curative resection. Among the 188 patients, 115 were alive without recurrence at 12 months, and had serial measurements of viral levels.
Results: The mean age was 53 years and the mean follow-up period was 48.5 months. With multivariate analysis, tumor size > 5 cm (P = 0.047), Child-Pugh class B (P = 0.017), vascular invasion (P = 0.028), and HBV DNA > 104 copies/mL at the time of resection (P = 0.003) were independently predictive of HCC recurrence for the entire population. For the 115 patients with serial measurements of viral levels, tumor size > 5 cm, HBV DNA > 104 copies/mL at resection, and the absence of sustained HBV DNA level < 104 copies/mL, the presence of cirrhosis, and elevated aminotransferase levels (> 40 IU/L) were marginally or significantly associated with HCC recurrence and overall survival. However, on multivariate analysis, sustained HBV DNA level < 104 copies/mL was the only factor for both low recurrence (P = 0.002; odds ratio [OR] 3.13; 95% confidence interval [CI] 1.55–6.35) and longer survival (P = 0.002; OR 3.76; 95% CI 1.61–8.78).
Conclusions: A high HBV replication state is among the most important predictors of adverse outcome after resection of HBV-related HCC. The sustained suppression of HBV below 104 copies/mL is a strong protective factor for long-term recurrence-free and overall survival.
Hepatitis B virus (HBV) is one of the most important risk factors for the development of hepatocellular carcinoma (HCC). Although there have been great advances in treatment modalities, HCC is still a devastating disease. In Korea, liver cancer is the 2nd most common cause of all cancer-related deaths and the leading cause of death in individuals in an age range of 40–50.1 One of the reasons for the dismal prognosis of HCC is the high tendency for recurrence after curative therapy. For example, the recurrence rate after curative surgery has been reported to be as high as 70% at 5 years,2 and the recurrence rate was reported to be 60% and 73–86% at 3 and 5 years after radiofrequency ablation, respectively.3 To date, no effective pretreatment or adjuvant modality has been established yet in the surgical treatment of HCC.
Factors contributing to tumor recurrence after liver resection include old age, male gender, heavy alcohol consumption, high α-fetoprotein (AFP) level, low platelet count, low albumin level, large tumor size, multiple tumors, the presence of cirrhosis, poor differentiation of the tumor, microsatellite lesions, vascular invasion, positive resection margins, intrahepatic metastasis, and high Child-Pugh score.4–6 In addition to these demographic and clinicopathologic variables, some studies have recently noted that certain viral factors, such as high HBV viral load and hepatitis B e antigen (HBeAg) seropositivity are also related to tumor recurrence after resection.7–12
Among the viral parameters, high HBV DNA load has often been reported to have a correlation with higher postoperative recurrence,7,10–12 but its impact on postoperative survival remains largely uncertain.8–10 The interpretation of these results, however, is complicated by previous use of antiviral agents, population heterogeneity, and different types of curative modality among studies. Moreover, one of the major drawbacks of the studies is that the HBV DNA levels were only analyzed at one point in time, without documentation of changing patterns of viral levels on multiple examinations. In fact, HBV DNA levels are highly fluctuating during the natural course of chronic HBV infection. Hepatitis B viral levels at a single time point may not accurately reflect the overall viral and inflammatory activity in the liver of each patient. It is therefore necessary to assess a change in HBV DNA levels over time to achieve a better understanding of the true effect of the HBV load on long-term outcomes. Unfortunately, there have been only two studies that have attempted to observe the changing levels of HBV DNA at multiple time points in patients undergoing resection.10,12
To explore this issue, the present study aimed to investigate the serial HBV DNA levels as well as the HBV DNA levels at the time of resection in patients with HBV-related HCC who underwent surgical resection. From a statistical analysis, we found the definitive role of continuous viral suppression on recurrence and overall survival of HBV-related HCC.
Between January 1998 and December 2007, 277 patients underwent curative surgical resection for HBV-related HCC at the Catholic Medical Center (Incheon St. Mary's Hospital, Seoul St. Mary's Hospital, The Catholic University of Korea) in South Korea. Out of 277 patients, 31 patients who had started antiviral therapy before surgery and 58 patients who were not assessed for HBV DNA levels at resection were excluded from the analysis. The remaining 188 patients were all positive for hepatitis B surface antigen (HBsAg) and negative for anti-HCV and anti-HIV antibodies. None of them had a history of habitual alcohol drinking of more than 40 g per day on average for more than 10 years. We thus performed an analysis of those 188 consecutive patients retrospectively. The diagnosis of HCC was based upon elevated serum AFP levels (> 200 ng/mL) with typical radiologic features, including computed tomography (CT), hepatic angiography, or magnetic resonance imaging. Once HCC was diagnosed, 127 patients (67%) received at least one session of loco-regional therapies, such as transarterial chemoembolization, percutaneous ethanol injection therapy, and radiofrequency ablation, before liver resection. Every specimen obtained from surgery was confirmed to be HCC histologically. Curative resection was defined as a complete resection of all macroscopically evident tumors with a surgical margin more than 5 mm by pathologic examination.
Initial work up and follow up
History, physical examination and baseline laboratory tests were obtained from 188 patients. All patients were tested at baseline for HBsAg, antibody to HBsAg, HBeAg, antibody to HBeAg (Abbott Laboratories, Abbott Park, IL, USA), complete blood cell counts, serum alanine aminotransferase (ALT), albumin, bilirubin, prothrombin time, and AFP. Between 1998 and 2002, serum HBV DNA was quantified using the Digene Hybrid Capture II assay (Digene; sensitivity > 105 copies/mL). The quantification of low-level HBV viremia was done using the Cobas Amplicor HBV Monitor Test (Roche Diagnostics; sensitivity > 200 copies/mL) or dual target polymerase chain reaction previously performed in our laboratory (sensitivity > 1 × 103 copies/mL).13 From 2003 to the end of the study follow-up, serum HBV DNA levels were measured using the VERSANT 3.0 branched DNA assay (Bayer HealthCare LLC, NY, USA; sensitivity > 2 × 103 copies/mL) or the Real-Q HBV quantification kit (BioSewoom Inc., Seoul, Korea; sensitivity > 56 copies/mL).
The resected tumor was examined both microscopically and macroscopically by a certified pathologist. The histologic grade of the tumor was classified into well, moderate, and poor differentiation according to the Edmondson classification.14 The maximal diameter of the tumor was considered to be the tumor size. The presence of capsulation, vascular invasion, and microsatellite lesion in the resected tumor, along with the degree of cirrhosis in the non-cancerous tissue, were also evaluated.
All patients were followed up at 3-month intervals until 5 years after liver resection, then at 4 to 6-month intervals thereafter; AFP levels and liver dynamic CT scans were assessed at each follow-up examination. HBeAg and HBV DNA were checked at the time of resection, and thereafter at 3 to 6-month intervals, or more frequently if necessary. During the follow-up, antiviral therapy was administered for patients with increasing HBV DNA and elevated ALT levels. Antiviral drugs were not used for patients with detectable HBV DNA and ALT levels < 80 IU/L who were not willing to receive, because the Korean health insurance policy for antiviral therapy has been strictly confined to viremic carriers with elevated ALT levels > 80 IU/L, irrespective of cirrhosis or HCC.
When intrahepatic recurrence was suspected, hepatic angiography was additionally performed to confirm this. Recurrence was defined as a radiologically evident new lesion, which includes enhanced or vascular staining area in the liver, or new lesions developing at an extrahepatic organ.
To explore the impact of an evolutionary change in the HBV DNA level following resection on postoperative outcome, 115 evaluable patients who survived without evidence of recurrence within 12 months were additionally analyzed. According to serial HBV DNA levels, these patients were divided into three different groups. Patients with sustained viral levels above 105 copies/mL before and after resection were categorized into the high viremia group, while patients with sustained viral levels below 104 copies/mL were categorized into the low viremia group. The remaining patients were categorized into the fluctuating HBV DNA group.
Categorical variables were compared by Cox regression analysis. Disease-free and overall survival rates were calculated by the Kaplan–Meier method, and differences in survival between groups were compared by the log-rank test. Statistical significance was defined by a P-value less than 0.05. Variables that were shown to be significant in the univariate analysis were further included stepwise into the multivariate analysis. Statistical analysis was carried out using SPSS 15.0 for Windows (SPSS, Inc. Chicago, IL, USA).
Table 1 shows baseline characteristics of the entire study population. The mean age was 53.4 ± 9.5 years, and 148 (78.7%) patients were males. Forty-nine (26.1%) patients had HBeAg, and the median HBV DNA level of the patients was 52.33 (< 0.056–> 105) × 103 copies/mL. One hundred sixty-eight (89.4%) patients were classified into Child-Pugh class A and the remaining patients into Child-Pugh class B. The median tumor size was 3.6 (1.0–11.0) cm. A single tumor was found in 165 (87.8%) patients and 23 (12.2%) patients had multiple tumors.
|Characteristics||(n = 188)|
|Age (years)||53.4 ± 9.5|
|HBV DNA (× 103 copies/mL)||52.33 (< 0.056–> 105)|
|HBV DNA (copies/ml)|
|> 104/≤ 104||94/94|
|ALT (IU/L)||37 (2–562)|
|Total bilirubin (mg/dl)||1.02 ± 1.39|
|Albumin (g/dl)||3.79 ± 0.52|
|PT (%)||91.0 ± 16.7|
|AFP level (ng/ml)||38.9 (0.54–75,800)|
|> 20/≤ 20||109/79|
|Tumor size (cm)||4.7 ± 3.3|
|Median (range, cm)||3.6 (1.0–11.0)|
|Preoperative treatment (TACE/PEI)||115/3|
Nineteen (10.1%) patients received antiviral therapy during the follow-up after resection. Eighteen patients received lamivudine, and the remaining one patient started entecavir. Of the 13 patients evaluable for antiviral response, 10 showed sustained viral suppression up to the end of follow-up, whereas the remaining 3 had lamivudine resistance and showed a decrease in HBV DNA levels (< 105 copies/mL) after adefovir rescue therapy.
Predictive factors for recurrence and overall survival after resection in the whole study population
During the median follow-up of 24.5 (1–132) months, 95 (50.5%) developed recurrences. Seventy-four patients developed intrahepatic recurrence alone, 14 patients had extrahepatic recurrence without intrahepatic recurrence, and 7 patients had concurrent intrahepatic and extrahepatic recurrences. Among those patients who recurred, 37 patients died from disease progression and 13 patients died from hepatic failure. The median time to recurrence was 9.6 months. The 3- and 5-year disease-free survival rates for the whole population were 46.4% and 36.2%, respectively, and the corresponding rates for overall survival were 59.9% and 43.1%, respectively (Fig. 1).
Of the 15 potential variables listed in Table 2, abnormal ALT (> 40 U/L), Child-Pugh class B, positive HBeAg, serum HBV DNA levels of > either 105 or 104 copies/mL at resection, vascular invasion, presence of cirrhosis, and tumor size > 5 cm were significantly or marginally associated with higher tumor recurrence after resection. On multivariate analysis using the Cox regression model, Child-Pugh class B (P = 0.017; odds ratio [OR] 4.10; 95% confidence interval [CI] 1.28–13.12), serum HBV DNA levels > 104 copies/mL at resection (P = 0.003; OR 3.16; 95% CI 1.48–6.77), vascular invasion (P = 0.028; OR 2.29; 95% CI 1.10–4.80), and tumor size > 5 cm (P = 0.047; OR 2.42; 95% CI 1.01–5.79) remained significant as independent risk factors for an increased risk of HCC recurrence after resection in the whole population.
|P-value||OR (95% CI)||P-value||P-value||OR (95% CI)||P-value|
|Age > 50 years||0.329||0.508|
|ALT > 40 IU/L||0.067||1.14 (0.58–2.25)||0.707||0.003||1.30 (0.60–2.81)||0.499|
|Child-Pugh class B||0.041||4.10 (1.28–13.12)||0.017||0.001||4.91 (1.74–13.82)||0.003|
|AFP level > 20 ng/mL||0.378||0.021||1.08 (0.45–2.61)||0.862|
|HBeAg seropositivity||0.003||2.22 (0.94–5.24)||0.068||0.028||2.99 (1.24–7.20)||0.014|
|HBV DNA > 104 copies/mL†||0.001||3.16 (1.48–6.77)||0.003||0.042||2.40 (0.50–11.49)||0.276|
|HBV DNA > 105 copies/mL†||0.013||1.80 (0.85–3.80)||0.127||0.065||1.69 (0.68–4.19)||0.259|
|Tumor differentiation‡||0.230||0.032||1.31 (0.53–3.29)||0.559|
|Vascular invasion||0.055||2.29 (1.10–4.80)||0.028||0.004||1.18 (0.47–2.94)||0.729|
|Microsatellite lesion||0.464||0.018||2.97 (1.22–7.25)||0.017|
|Cirrhosis||0.035||1.03 (0.38–2.81)||0.950||0.019||1.92 (0.69–5.38)||0.213|
|Tumor size > 5 cm||0.058||2.42 (1.01–5.79)||0.047||0.010||5.41 (2.25–13.04)||0.001|
For overall survival after resection, Child-Pugh class B (P = 0.003; OR 4.91; 95% CI 1.74–13.82), HBeAg seropositivity (P = 0.014; OR 2.99; 95% CI 1.24–7.20), microsatellite lesions (P = 0.017; OR 2.97; 95% CI 1.22–7.25), and tumor size > 5 cm (P = 0.001; OR 5.41; 95% CI 2.25–13.04) were identified as independent factors for poor survival in the multivariate analysis (Table 1). However, neither HBV DNA level of 104 nor 105 copies/mL at resection was significant as a prognostic factor for overall survival.
Predictive factors for recurrence and overall survival after resection in patients surviving more than 1 year
To explore the impact of an evolutionary change in HBV DNA levels on long-term prognosis, 115 evaluable patients without recurrence at 1 year who had serial measurements of viral levels were analyzed. Univariate analysis showed that serum HBV DNA levels > 104 copies/mL at resection, fluctuating or sustained high viremia on serial monitoring, and tumor size > 5 cm were associated with higher tumor recurrence. Elevated ALT levels (> 40 IU/L) as well as these three variables were associated with poor survival (Table 3). In the multivariate analyses, serial patterns of HBV DNA levels was identified as the only factor for both postoperative recurrence (P = 0.002; OR 3.13; 95% CI 1.55–6.35) and survival (P = 0.002; OR 3.76; 95% CI 1.61–8.78) (Table 3).
|P-value||OR (95% CI)||P-value||P-value||OR (95% CI)||P-value|
|Age > 50 years||0.090||0.342|
|ALT > 40 IU/L||0.112||1.11 (0.29–4.24)||0.875||0.024||1.05 (0.27–4.15)||0.944|
|Child-Pugh class B||0.704||0.987|
|AFP level > 20 ng/mL||0.281||0.572|
|HBV DNA > 104 copies/mL†||0.006||2.43 (0.25–23.25)||0.442||0.023||4.10 (0.14–125.0)||0.410|
|Sustained low viremia < 104 copies/mL (absent)||0.001||3.13 (1.55–6.35)||0.002||0.010||3.76 (1.61–8.78)||0.002|
|Tumor size > 5 cm||0.030||2.57 (0.57–11.68)||0.221||0.044||3.05 (0.40–23.25)||0.282|
Tumor recurrence and overall survival according to the serial patterns of HBV DNA levels
The cumulative tumor recurrence according to the pattern of HBV viral status after liver resection is shown in Figure 2. There was a significant difference in disease-free survival between the three groups. Patients with a continuous low viral level (< 104 copies/mL) had a lower tumor recurrence rate than those with a continuous high or fluctuating viral level (log-rank test, P < 0.001; Fig. 2). During the median follow-up of 41.7 (11.7–132.1) months, the recurrence rates after resection in patients with sustained low, fluctuating, and sustained high viral loads were 12.5%, 20.3%, and 60.9% at 3 years, and 21.5%, 47.9%, and 75.2% at 5 years, respectively (Fig. 2).
There was also a stepwise increase in the overall survival rate from the sustained high, fluctuating, and sustained low viremia groups. The overall survival rate was significantly higher in patients with sustained low viremia (log-rank test, P = 0.018; Fig. 3). The 3- and 5-year overall survival rates were 96.3% and 82.9% in patients with sustained low viremia, 81.2% and 59.6% in patients with fluctuating viremia, and 73.3% and 45.1% in patients with sustained high viremia, respectively (Fig. 3).
The present study focused primarily on the prognostic significance of the HBV viral load, particularly the changing patterns of HBV DNA levels, in clinical outcomes after resection of HCC. Previously, Kubo et al. first reported the association between HBV viral load and tumor recurrence after surgical resection in a retrospective review of 40 patients with HBV-related HCC.7 Another report by Kubo et al. indicated that a continuous high HBV DNA (> 5 mEq/mL) level was related to poor survival as well as a high recurrence rate.8 In contrast, a large observational study with 230 patients by Kim et al. failed to show the difference in survival between the sustained viremia (> 5 log copies/mL) and non-viremia groups despite the high recurrence rate in the sustained viremia group.10 In the present study analyzing patients by serial patterns of HBV DNA levels, we found a significant role of continuous viral suppression in both recurrence and overall survival of patients with HBV-related HCC.
The most significant finding of this study is that a sustained low HBV DNA level (< 104 copies/mL) during and after liver resection was strongly associated with prolonged survival, as well as a lower risk of recurrence on both univariate and multivariate analyses. Although there have been several reports that indicated a high tumor recurrence potential after resection in patients with high viremia,7–12 most of the studies have only considered HBV DNA levels at one point during the study period, which could not represent the true effect of potentially fluctuating HBV DNA levels over time on outcome. In the present study, we analyzed the HBV DNA level at the time of resection and the changing patterns of HBV DNA levels during the follow-up period as different prognostic parameters. As a result, only the sustained low viremia was statistically significant on multivariate analysis for recurrence and overall survival. Moreover, patients with a continuous low viremia had better disease-free and overall survival than those with a fluctuating viral level. Collectively, these results demonstrate that continuous HBV viral suppression has a greater impact on prognosis than the HBV DNA level at one point in time.
In respect of the cut-off value of viral load, we set HBV DNA < 104 copies/mL as a low viral load in this study. This is the lowest HBV DNA level ever suggested as a prognostic parameter for HCC recurrence. Many researchers have used a higher HBV DNA cut-off value (> 105 copies/mL) to differentiate between patients with high and low viremia.7–10 Those studies indeed showed a difference in tumor recurrence, but not in survival. In our results, we found that a lower HBV DNA cut-off of 104 copies/mL is superior to > 105 copies/mL in predicting outcomes after resection. It is therefore needed to suppress further HBV DNA to a lesser level in order to obtain better clinical outcomes after surgery. In fact, two studies have previously defined a border between high and low viremia as 104 copies/mL on this issue.11,12 However, these studies did not show any definitive significance of viral load on overall postoperative patient survival. The reason for this may be a relatively small number of enrolled patients, a short follow-up duration, and the lack of documentation of serial checks of HBV DNA levels. In the present study, using a low HBV DNA cut-off value of 104 copies/mL as well as serial measurement of viral levels in the enrolled patients, we demonstrated a clear survival benefit as well as low recurrence potential after resection in those patients with sustained low viral level. Furthermore, by stratified analyses of three viremic groups from serial measurement of viral levels, we also showed that the post-liver resection prognosis of the fluctuating HBV DNA group behaves largely in a fashion between the high and low viremia groups. Thus, the overall findings suggest that maintenance of the HBV DNA level as low as possible is essential for long-term disease-free survival in patients undergoing resection.
Given the strong association between high viremia and poor postoperative outcomes, it is anticipated that the use of antiviral therapy for the surgical population may provide a better postoperative outcome. Recently, some reports have explored the benefit of pre-operative antiviral therapy on prognosis after resection.12,15 However, each study offered varying results and failed to reach a solid conclusion favoring antiviral therapy in the setting of resection for HBV-related HCC. Long-term results have to be further evaluated and large, randomized prospective trials are expected.
The mechanism for recurrent carcinogenesis in the remnant liver after resection of HBV-related HCC is largely unclear. HBV can cause HCC directly by enhancing expression of HBx antigen, preS2 activator, or altering functions of NF-kB, TGF-β1, or α2-macroglobulin.16–18 Indirectly, active HBV replication can lead to hepatic inflammation, which induces ultimately fibrosis and carcinogenesis19,20 by triggering DNA polymorphism, chromosomal instability, immune responses, and upregulation of adhesion molecules.16,21–23 In the same way, a high HBV viral load leads to active inflammation in the remnant liver, and then causes tumor recurrence via intrahepatic metastases from primary tumor or multicentric carcinogenesis.7,15,24–26 Further studies are necessary to identify the conclusive role of a high viral load on the tumor recurrence pattern after resection.
There are some limitations in the present study. Since this was a retrospective design, more frequent and planned checking of clinical data was not performed for some patients. More than half of the patients received loco-regional therapies prior to liver resection, which could have affected the outcomes. The change in quantification assays for HBV DNA levels over the study period remains a consideration on the interpretation of the results. However, for each person, quantification for HBV DNA levels was done with the same method and measured serially on multiple occasions. We therefore believe that inter-assay variation had a minor role in the data results.
In conclusion, along with tumor and liver function parameters, HBV viral load plays a key role in determining overall survival as well as recurrence after surgery. Above all, maintenance of HBV DNA level < 104 copies/mL during and after liver resection was a strong protective factor for both tumor recurrence and overall survival for patients with HBV-related HCC. Continuous suppression of HBV DNA to the lowest possible level by using antiviral agents in patients undergoing surgical resection needs to be elucidated by large prospective studies.
- 1Deaths and Causes of Death in 2007: STATISTICS KOREA [Internet]. Korea: National Statistical Office; 2008 Sep 23. Cited 14 June 2010. Available from URL: http://kostat.go.kr/board_notice/BoardAction.do?method=view&board_id=144&seq=85&num=85&parent_num=0..
- 11Recurrence of hepatitis B-related hepatocellular carcinoma is associated with high viral load at the time of resection. Am. J. Gastroenterol. 2008; 103: 1663–73., , , , , .Direct Link:
- 13Lamivudine-resistance in patients with chronic hepatitis B and/or cirrhosis and detection of mutations in YMDD motif of hepatitis B virus genome. Korean J. Hepatol. 2001; 7: 15–33., , et al.