374. The Effect of Sustained HBV Viremia on the Recurrence of
Hepatocellular Carcinoma After Curative Resection
S. Ahn; B. Kim; J. Kim; D. Kim; Y. Paik; K. Lee; B. Moon; C. Chon; Y. Moon; K. Han
Background
An elevated serum HBV DNA
level is a risk factor for developing hepatocellular carcinoma (HCC). Rather
than the instantaneous serum HBV DNA, the duration of viremia is more important
in carcinogenesis. Nevertheless, most investigations have evaluated the DNA
level at study entry as a risk factor. We assessed the effects of sustained HBV
viremia on the recurrence of HCC after curative resection.
Methods
A total of 230 patients undergoing curative resections between 2000 and 2006 were initially included. Patients who received antiviral therapy at diagnosis or during follow-up, or experienced recurrence within 6 months after resection were excluded. In addition, those who had fluctuating DNA levels (cutoff value: 100,000 copies/mL) were also excluded. Ultimately, 157 patients were enrolled. The serum HBV DNA level, biochemical tests, alpha-fetoprotein, and liver dynamic computed tomography were obtained at 3-month intervals after surgery. The primary endpoint was the recurrence of HCC.
Results
With a median follow-up duration of 33.4 months, 75 (47.7%) of 157 patients experienced recurrences. Eighty-nine (non-viremia group) had consistently negative serum HBV DNA levels (<100,000 copies/mL), while 68 (viremia group) were consistently positive (>100,000 copies/mL). No significant differences occurred in age, sex, liver functions, histology, tumor markers, or TNM staging (American Joint Commission on Cancer, 6th ed.) between the two groups. Patients in non-viremia group had a lower 5-year recurrence rate (54.7%) than those in viremia group (72.8%) (p=0.041). On multivariate analysis, tumor stage (p=0.001) and sustained HBV viremia (p=0.041) were identified as independent factors predicting recurrence. HBeAg status and serum aminotransferase levels had no effect on the clinical outcome.
Conclusions
Sustained HBV viremia, in addition to tumor stage, may be a prognostic factor for the recurrence of HCC after curative resection. Therefore, to prevent long-term recurrences, antiviral therapy could be considered to lower the serum HBV DNA load.
376.
Early Diagnosis Panel of Serum Biomarkers of HBV-Induced Hepatocellular
Carcinoma During 3
Years Follow-up
C. Wang; B. Wang; H. You; X. Ou
Objective
To investigate the combined value of alpha fetoprotein(AFP), des-γ-carboxy prothrombin (DCP), α-feto-protein-L3 (AFP-L3), γ-glutamyl transpeptidase-II (GGT-II),CA125,CA19-9 in patients with liver cirrhosis who had been followed for the development of hepatocellulor carcinoma (HCC). And using SELDI-TOF-MS to investigate proteomic markers of these patients.
Patients and Methods
Eighty-one patients with HBV-induced liver cirrhosis were monitored by serum AFP, DCP, AFP-L3, CA125, CA19-9, GGT-II and ultrasonography every 6 months for 3 years. Proteomic spectra by SELDI-TOF-MS were generated in 10 patients who were diagnosed as HCC during following up. At the same time, all the serum markers were measured in 99 patients with HBV-induced HCC, 30 with secondary liver metastasis, and 20 healthy controls.
Results
During 3 years following up, 12 (15.9%) of the 81 patients with HBV-induced liver cirrhosis were diagnosed as HCC. Each marker’s positive rate in patients who developed HCC during follow-up before HCC was diagnosed showed in table. The diagnostic value of DCP was the highest, with 75.8%sensitivity, 88.6%specificity, and 81.4%accuracy; The accuracy of other markers was reduced in order of AFP, AFP-L3, CA125, GGT-II, and CA19-9. Logistic regression analysis identified DCP, GGT2, AFP, and CA125 as independent markers for HCC. When compared the accuracy with the combination panel of all the serum markers, AFP, DCP plus CA125 showed the highest of 88.2% while the accuracy of AFP alone was only 76.2%. Spearman correlated analysis showed that DCP and CA125 were related with tumor size. DCP, CA125 and CA19-9 were related with Portal vein thrombosis, but no markers were related with extrahepatic metastasis. In addition, the proteomics analysis showed the 4350.8m/z protein increased from cirrhosis to HCC, while the level of the 6860.5, 7571, 15125, 15893 m/z proteins decreased from cirrhosis to HCC.
Conclusions
DCP and GGT-II emerged earlier in HCC patients than other markers. The best combination of the markers for early diagnosis of HBV-induced HCC was AFP, DCP plus CA125.
|
Markers |
18 months pre-HCC |
12 months pre-HCC |
6 months pre-HCC |
0 months pre-HCC |
Total developed HCC patients(n) |
|
DCP |
2 (16.7%) |
4 (33.3%) |
6 (50%) |
9 (75%) |
12 |
|
AFP(>15ng/ml) |
0 |
4 (33.3%) |
8 (66.7%) |
8 (66.7%) |
12 |
|
AFP(>200ng/ml) |
0 |
0 |
3 (25%) |
5 (41.7%) |
12 |
|
AFP-L3 |
0 |
2 (16.7%) |
4 (33.3%) |
6 (50%) |
12 |
|
GGT-II |
2 (16.7%) |
4 (33.3%) |
6 (50%) |
8 (66.7%) |
12 |
|
CA125 |
0 |
0 |
1 (12.5%) |
2 (16.7%) |
12 |
|
CA19-9 |
1 (12.5%) |
3 (25%) |
6 (50%) |
7 (58.3%) |
12 |
S. Ura; M. Honda; T. Ueda; R. Nishino; H. Takatori; M. Nakamura; S. Kaneko
Background and Aims
MicroRNAs (miRNAs) are small non-coding RNAs that represent an important mechanism for posttranscriptional gene silencing and have recently been recognized to play important roles in tissue development, cell differentiation and proliferation and apoptosis processes in human diseases and cancers. We have previously reported gene expression profiles of chronic hepatitis (CH)-B and CH-C differ (Gastroenterology.2001, 120, 955, Hepatology. 2006, 44, 1122). This study examined miRNA expression profiling in chronic viral hepatitis (CH) and hepatocellular carcinoma (HCC) liver tissues.
Methods
We applied real-time detection (RTD)-PCR to detect miRNAs. We used Looped-primer RTD-PCR with TaqMan® MicroRNA Assays Human Panel Early Access Kit (Applied Biosystems), which contains a 180-assay panel covering most known human miRNAs. We analyzed tumor and non-tumor liver tissues from 12 patients with HBV-related HCC (HCC-B), 14 patients with HCV-related HCC (HCC-C) and 9 patients with normal liver (N). Candidate target genes by miRNAs were predicted by Miranda Pro 3.0. To evaluate the regulatory effect of miRNAs on target genes in these samples, whole gene expression profiling was obtained from these samples using an in-house cDNA microarray comprising 9614 clones. Data analysis was performed by BRB Array Tool (NCBI).
Results
Hierarchical clustering analysis (unsupervised learning methods) and supervised learning methods using compound covariate predictor revealed significant classification of N, CH and HCC respectively (N vs. CH: 97% accuracy, p<0.001; CH vs. HCC: 79% accuracy, p<0.001). Differences were larger between N and CH than between CH and HCC. Interestingly, miRNA profiles differentiated CH-B and CH-C, and HCC-B and HCC-C (CH-B vs. CH-C: 92% accuracy, p<0.001; HCC-B vs. HCC-C: 77% accuracy, p=0.02), although statistical values for HCC-B and HCC-C were less. Out of differentially expressed miRNAs among these classifications (p<0.01), 18 miRNAs from N vs. CH and 8 miRNAs from CH and HCC regulated target gene expression in host tissue samples, as determined by group comparison using whole gene expression profiling data from cDNA microarrays. Similarly, 4 miRNAs from CH-B vs. CH-C and 2 miRNAs from HCC-B vs. HCC-C were determined and expressions could differentiate HBV- or HCV-infected liver.
Conclusions
Expression of miRNAs clearly differentiated N, CH and HCC as well as HBV- and HCV-infected liver. Functionally active miRNAs determined by combined gene expression data can reveal pathogenesis and identify therapeutic targets for liver disease.
N. Park; B. Park; J. Shin; S. Jung; C. Nam; Y. Na; J. Seo; H. Shin; S. Lee; D. Lee; S. Kim; J. Kim; Y. Chung
Background
DNA repair mechanisms are important for maintaining DNA integrity and preventing carcinogenesis. Genetic variations in DNA repair genes are thought to modulate DNA repair capacity and are suggested to be related to hepatocellular carcinoma (HCC) risk.
Methods
We investigated the effects of the XRCC3 Thr241Met and ERCC5/XPG His1104Asp polymorphisms on the risk of developing HCC in a case-control study. The XRCC3 Thr241Met and ERCC5/XPG His1104Asp polymorphisms were assayed by PCR-RFLP method for 541 cases with HCC and 405 controls without HCC among hepatitis B virus (HBV) carriers. Allele frequencies of polymorphism were compared between patients with HCC and patients without HCC among HBsAg positives by logistic regression.
Results
The observed genotype frequencies of polymorphisms in both cases and controls conformed to the Hardy-Weinberg equilibrium. The frequencies of the ERCC5/XPG His/His, His/Asp and Asp/Asp genotypes of among cases (21.6, 63.9 and 14.5%, respectively) were significantly different from those among controls (9.6, 74.6 and 15.8%, respectively; P<0.001). The ERCC5/XPG His/His genotype was more frequent in cases than in controls, whereas the His/Asp genotype was less frequent in cases than in controls (P<0.001 and P<0.001, respectively). With the adjustment for age and gender, the ERCC5/XPG His/His genotype was associated with a significantly in the risk of HCC (adjusted OR, 2.62; 95% CI, 1.75-3.91; P<0.001) with the combined His/Asp and Asp/Asp genotypes as the reference. However, the XRCC3/XPB Thr241Met genotype was not associated with HCC risk (adjusted OR, 1.24; 95% CI, 0.65-2.37; P=0.52).
Conclusions
These results suggest that the ERCC5/XPG His1104Asp polymorphism contributes to genetic susceptibility to HCC, while XRCC3 Thr241Met polymorphism is not associated with HCC risk, in Korean patients with chronic HBV infection.
400. Hepatitis B Virus X Gene Mutations Are Common in Patients Both With or Without
Hepatocellular Carcinoma and Further Mutations Occur When Hepatocellular
Carcinoma Develops
C. Tan; P. Chang; G. Lim; P.
Ooi; Y. Chung; K. Chow; W. Chow
Background
Most cases of hepatocellular carcinoma (HCC) worldwide are associated with chronic hepatitis B virus (HBV) infection. The hepatitis B x (HBx) gene is believed to be involved in the molecular pathogenesis of HBV-related HCC where mutations of the HBx gene are very common. However, little is known about the extent of mutation of the HBx gene in a similar group of individuals with chronic HBV infection but without HCC. Also, it is not clear if further mutations in the HBx gene occur before HCC develops.
Aims
(1) To compare the prevalence of HBx gene mutations in a cohort of patients with HBV-related HCC against age and HBe status-matched HBV controls without HCC.
(2) To investigate the prevalence and concordance of HBx gene mutations in the sera, HCC and non-HCC liver tissues of patients with HBV-related HCC.
Methods
The study (HCC) group comprised 16 consecutive patients with HBV-related HCC who underwent resection. The control group (2:1 ratio, N=32) comprised age and HBe status-matched non-HCC patients seen in our hepatitis B outpatient clinic. All the patients had perinatally-acquired HBV infection and none had received treatment for HBV. Nested PCR of the full HBx genome was performed on the sera of all cases and PCR products were sequenced to look for mutations. Nested PCR was similarly performed on tumor and non-tumor tissues of the HCC group and PCR products sequenced.
Results
In the HCC group, 9/10 had mutations in HBx gene in the sera (6 were PCR negative) compared to 19/21 in the control group (11 were PCR negative), p=NS. The most common mutations in both groups were similar, ie H86R and I127T. All tumor tissues (16/16) yielded mutations in the HBx gene compared to 13/16 in non-tumor tissues. Concordance of the signature HBx gene mutations between tumor and non-tumor tissues was 43.8%(7/16) and 12.5%(2/16) between sera and tumor tissues (p<0.05).
Conclusions
HBx gene mutations are common in the serum of patients with chronic HBV infections with or without HCC. However, in patients with HCC, HBx gene mutations are more common in the liver tissues and further mutations may occur when they integrate into liver tissues, leading to development of HCC.
401. Trends in the Incidence of Hepatocellular Carcinoma in
M. L. Fernandes; Y. Chan; S.
Lim
Background and Aims
To describe the trends in the incidence
of hepatocellular carcinoma (HCC) in the three main ethnic groups in
Methods
The incidence of HCC was obtained from
the Singapore Cancer Registry. Data on the prevalence and incidence of
hepatitis B and hepatitis C was obtained from the Ministry of Health
Results
Between 1968 and 2002, 2913 cases of HCC
were registered by the Singapore Cancer Registry. The age standardized incidence
(ASR) of HCC in the population fell by 58% in men and 47% in women from
1968-1972 to 1998-2002 from 17.1 to 7.1 per 100,000 men (p = 0.004) and from
2.8 to 1.5 per 100,000 women (p=0.01). Amongst the three main ethnic groups,
the incidence for HCC was highest in Chinese compared with Malays (p=0.048) and
Indians (p=0.023). The overall prevalence of hepatitis B in
Conclusions
There has been a significant decrease in
the incidence of HCC in
402. The Changing Epidemiology and Etiology of Hepatocellular
Carcinoma from 1969 Through 2006 in
M.
Connelly; M. Bruce; N. Kassebaum; L. Bulkow; M. Snowball; B. J. McMahon
Alaska Native people have an increased
rate of primary hepatocellular carcinoma (HCC) compared to the overall
Methods
We reviewed all cases of viral and
non-viral associated HCC in Alaska Native persons from 1969 through 2006. Using
ICD-9 codes from the
Results
Over the study period, 115 cases of HCC were identified in 110 persons. The age distribution for HCC was bimodal with peaks in the second and sixth decades of life (early peak due to HBV). The overall HCC rate was 3.63 per 100,000 and did not change significantly over the study period (p=0.54). However, HCV-associated HCC increased from zero cases prior to 1985 to 23 since 1995, representing 30% of viral associated HCC (p<0.01). The rates of HBV-associated HCC varied widely over time (1.26-3.79 per 100,000) but did not show a significant trend over the study period (p=0.52). The age distribution of HBV-associated HCC demonstrated a shift towards presentation later in life. From 1969 to 1999, 29% of HBV associated HCC cases were under 20 years of age; after 2000, no cases were under age 20 (p=0.028). Non-viral HCC rates ranged from 0.3 to 2.04 cases per 100,000 but demonstrated no significant trend over time (p=0.90). Mean BMI was greater in non-viral HCC compared to viral HCC (38.8 vs 26.6, p=0.002).
Conclusions
HCC rates in Alaska Natives remained stable over the study period, but the epidemiology and etiology are changing. HCV has emerged as an important cause of HCC since 1995. Two decades after mass hepatitis B immunization, the HCC age distribution has shifted to cases presenting later in life. This is consistent with the presence of an aging HBV-infected population with few new chronically infected young persons entering the population. The significantly increased BMI in non-viral HCC cases compared to viral HCC cases suggests nonalcoholic fatty liver disease as a likely etiology.
403. Prognostic Factors Affecting 3 Years Survival in Patients with
Hepatocellular Carcinoma (HCC)
M.
Strazzabosco; M. De Giorgio; S. Vezzoli; R. Bezzo; M. Luca; G. Verga; S.
Okolicsanyi; M. Valsecchi; S. Galimberti; M. Colledan; S. Fagiuoli
HCC is heterogeneous cancer with different risk factors, and response to therapy. Overall survival at 3 years is poor, but a subset of patients shows a survival, comparable to that achieved after liver transplantation. The ability to identify these patients would facilitate decision making and organ allocation.
Aims
To evaluate the influence of pre-treatment parameters on three-year survival in a cohort of non-transplanted HCC patients.
Methods
Out of 400 patients with HCC diagnosed according to the EASL criteria treated and prospectively followed-up at a single referral centre, we selected 228 consecutive patients that did not undergo liver transplantation and were either alive (group A, 91 pts) or dead (group B, 137 pts) three years after diagnosis. Patients were treated, with: resection (26 pts in group A, 27 pts in group B), ablative treatments (49 in A and 28 patients in B), TACE (13 pts in A and 28 in B) while 3 patients in A and 55 in B remained untreated because of refusal of therapy, co-morbidities or advanced disease. Qualitative and quantitative variables were first analyzed with the likelihood χ2 test qualitative or Wald test, respectively, and then with a logistic regression model (multivariate analysis, with stepwise method).
Results:
Median survival was 53,4 months (range: 36,2-105,2) in A and 10,7 months (range: 0-35,5) in B. At univariate analysis, ascites, encephalopathy, portal vein thrombosis, HCV, HBsAb/HBcB, overall tumour burden, number of nodules, ALT, ALP, γGT, Bilirubin, Albumin, PT-INR, Hematocrit, AFP, MELD, Child Pugh, BCLC, TNM and CLIP were significantly different between A and B (p< 0,05). After multivariate analysis, tumor burden, (Odds Ratio:1,332), bilirubin(OR:3,053, portal vein thrombosis (OR:13,208) and HbsAb/HbcAb (OR:3,201) were significant independent predictors of survival. These were used to generate a logistic regression model able to predict the probability of death and survival 3 years after HCC diagnosis. The predictive ability expressed as the Area Under the Curve of the ROC Curve was 83,4%, showing a good accuracy of the model. Depending on the cut-off probability, sensitivity (survival shorter than 3 years correctly classified) of 90.91% and 1-specificity (survival longer than 3 years incorrectly classified) of 16.95% were reached.
Conclusion
Prognosis of patients with HCC can be predicted using a mathematical model factoring tumor burden, macrovascular invasion, liver function and previous contact with HBV as independent factors. The ability of the model to stratify patients according to their estimated prognosis three years after the diagnosis may be of help in decision making and organ allocation.
J. Cheong; S. Cho
Backgrounds/Aim
The outcomes of hepatitis B virus (HBV) infection are probably related to host immune factors. Interleukin-18 (IL-18) plays significant roles in immune defense. This study was undertaken to investigate the association between persistence of HBV infection or HCC occurrence and single nucleotide polymorphisms (SNPs) of IL-18 gene.
Methods:
Between March 2002 and December 2002, a total of 1,050 Korean patients were enrolled in three different groups; ‘HBV clearance (n=320)’ and ‘chronic hepatitis (CH) / liver cirrhosis (LC) (n=637)’ and ‘hepatocellular carcinoma (n=93)’ groups. We assessed polymorphisms at four polymorphic sites in IL-18 gene at position -9731G>T, -9212G>C, -140C>G, +4861A>C in study subjects. Results: IL-18 -9212C allele (OR = 0.25, P = 0.01), -140G allele (OR = 0.36, P = 0.02) and +4861C allele (OR = 0.25, P = 0.01) were significantly associated with HBV clearance in a recessive model. Additionally, IL-18 -9212G allele (OR = 2.18, P = 0.01 in co-dominant model, OR = 2.11, P = 0.03 in dominant model), -140C allele (OR = 1.97, P = 0.03 in co-dominant model, OR = 1.98, P = 0.04 in dominant model) and +4861A allele (OR = 1.91, P = 0.04 in co-dominant model) showed susceptible effect on the HCC development.
Conclusions
This study suggested that SNPs at various sites of IL-18 gene were associated with HBV clearance and HCC occurrence.
J. Cheong; K. Kwack; S. Lee;
J. Kang; S. Cho
Hepatitis B virus infection (HBV) associated chronic liver disease (CLD) and hepatocellular carcinoma (HCC) are major causes of death in HBV endemic areas. Expression of collagens is related progression and suppression in CLD and/or HCC. We investigated whether polymorphisms of collagen type III, alpha 1 (COL3A1) associated with genetic susceptibility of CDL and HCC.
Methods
Three hundred and forty-two patients (230 CLD, 112 HCC) and 117 healthy controls were performed association study using Illumina’s SNP chip by allele specific extension method. Selected polymorphisms and haplotypes were analyzed by logistic regression controlling age and gender as covariables. In statistic analysis, SNP CHA-1 was high significantly association with control verse CLD, control verse HCC, and control verse CLD with HCC in allele model, codominant, dominant (p=0.01 to 0.002). Haplotype 1 (ht1), ht 2, and ht 5 significant associated with to control verse CLD (p=0.021, p=0.030 p=0037, respectively), control vs HCC (ht 1: p=0.046), and control verse CLD with HCC (ht 1: p=0.013 and ht 5: p=0.046).
Conclusion
In conclusion, we showed that SNP CHA-1 polymorphism and ht1 may increase the clearance of HBV infection, whereas ht2, ht5 may increase the susceptibility to disease progression, by using SNP chip analysis.
P. L. Almasio; T. Stroffolini; M. Persico; G. Gaeta; S. Bruno
The impact of previous exposure to HBV as a risk factor associated with tumor occurrence in subjects with HCV-related cirrhosis has not been fully investigated. We assessed whether serum anti-HBc is associated with HCC development in HCV-RNA+ve, HBsAg -ve patients with cirrhosis.
Methods
The study group included 693 patients (432 men, mean age 54.8 years) with HCV-related histologically-proven cirrhosis and anti-HBc testing, surveilled with ultrasound examinations every six months, was analyzed. All patients were treated with interferon (IFN) between January 1992 and December 1997. Independent predictors of HCC were assessed by Cox multiple regression analysis.
Results
Mean follow-up was 96.2 months; 303 (43.8%) patients were anti-HBc seropositive. Anti-HBc+ve patients were more often males (67% vs. 58.7%, p 0.01) and had lower transaminase levels (3.3 vs. 3.8 u.l.n., p 0.004) than anti-HBc-ve patients. Incidence rates of HCC per 100 person-years were 1.84 (95% C.I. 1.34-2.47) in the anti-HBc +ve patients and 1.86 (95% C.I. 1.41-2.42) in the anti-HBc-ve ones (Rate Ratio: 0.9; 95% C.I. 0.65-1.52) (Figure 1). In 599 non responder patients by Cox multiple regression, there was no association of serum anti-HBc with HCC development (H.R. 0.92; 95% C.I. 0.61-1.41), or liver-related deaths (H.R. 0.88; 95% C.I. 0.54-1.42). Only in the subgroup with sustained viral response we observed a slight increase of incidence of HCC per 100 person-years in anti-HBc +ve patients as compared to negative ones (1.17, 95% C.I. 0.32-2.98 vs. 0.64, 95% C.I. 0.13-1.87).
Conclusions
Serum anti-HBc+ve/HCV+ve cirrhosis treated with IFN irrespective to the achievement of response did not show an increased risk of HCC in comparison with anti-HBc-ve subjects. These results indicate that in every-day clinical practice patients with HCV-related cirrhosis and past exposure to HBV infection do not require a tailored surveillance for HCC.