Journal of Hepatology, 1992; 14: 183- 187 0 1992 Elsevier Science Publishers B.V. All rights reserved. 0168-8278/92/$05.00
183
HEPAT 00942
revalence, ris
clinical
features
F. Farina&, S. Fagiuoli’, N. De Maria’, M. Chiaramonte’, V. Aneloni*, S. Qngaro*, . Salvagnini’ and R. Naccarato’ ‘Cattedra Malattie Apparato Digerente, Istituto Di Medicina Interna, Universita’ Di Padova and %entro Traq%sionale @p-&de Civile Di PGdova, Padova, Italy (Receiwd
28 August 1990)
Recent reports indicate that hepatitis C virus (HCV) may play a role in the pathogenesis of hepatocellular carcinoma in cirrhotics. Using an ELISA test, we evaluated the prevalence of anti-HCV antibodies in 97 patients with hepatocellular carcinoma (HCC) in cirrhosis and in a group of 223 patients, including: 49 patients with HBsAg-positive chronic liver disease (CLD), 42 with alcoholic CLD, 110 with cryptogenic CLD and 22 with post-transfusional HBsAg-negative CLD. All diagnoses were histologically confirmed. Overall, anti-HCV-positive HCC were 64% of the total, with no statistically sign& cant difference with respect to CLD (60.9%). The prevalence of anti-HCV was higher in cryptogenic HCC (80%) than in HBsAg-positive (60%) or alcoholic HCC (42.8%) (p C 0.005). When HCC and cirrhosis of similar putative etiology were considered, anti-HCV prevalence was significantly higher in HCC than in cirrhosis only in the groups of patients with alcoholic liver damage (60% in HCC vs. 38% in cirrhosis, p < 0.005). In HBsAg-positive patients, anti-HCV prevalence was twice as high in HCC than in CLD, but the difference was not statistically significant. Overall, anti-HCV prevalence in HCC was significantly higher than in alcoholic or HBsAg-positive CLD (p c 0.001 and p < 0.01, respectively) but lower than in cryptogenic CLD (p < 0.001). Association between anti-HCV and anti-HBc was significantly more prevalent in patients with CLD than in those with HCC. From the clinical point of view, multifocal lesions and lack ofa history of previous liver disease were more frequent in anti-HCV-positive HCC (p < 0.05) than in HBsAg-positive or alcoholic patients. AntiHCV-positive HCC had a significantly worse proguosis @ < 0.005 with the Generalized Wilcoxon-Breslow test), when survival in the different subgroups was examined with life table analysis. No statistically significant differences were observed with respect to the remaining parameters considered. In conclusion, HCV infection appears to be a frequent event in HCC and CLD, particularly in patients with cryptogenic liver disease but, with the exception of patients with alcoholic liver damage, the prevalence of anti-HCV appears to be similar in HCC and cirrhosis. This could indicate that the infection plays an important role only in specific subgroups of HCC patients. Anti-HCV-positive HCC is frequently diagnosed in a more advanced stage and has a worse prognosis but, since the clinical history of this subgroup of patients frequently makes its debut with HCC, in our opinion the prognosis is worse due to late diagnosis. -
Although several factors have been linked to the development of hepatocellular carcinoma (HCC), thus far, V) has been considered t hepatitis B virus ( risk factor on the basis of evidence from epidemiology
(l-3), (6-8).
molecular biology (4,5) and biological studies However, HBV has never been shown to transf0 in cell cultures (9) and the data on the integration of viral DNA has never revealed the uniform presence of a
Correspondence: Dr. Fabio Farinati, Cattedra Malattie Apparato Digerente, niani 2,351OO Padua, Italy.
Istituto di Medicina Interna, Policlinico Universitario,
Via Giusti-
F. FARINATI
184 particular sequence or shown viral DNA fragments close to oncogenes or sequences considered to be relevant in terms of carcinogenesis (1 ,lO). Alcohol has also been considered to play a role in the development of liver callcer, although the association between alcohol and hepatocellular carcinoma is probably indirect (11,12). Morerecently, theavailability of a test which detects an antibody retogni&g an epitope directly linked to HCV, has prompted research on the possible association between HCV infection and HCC. A number of papers have been published (13-19), including a preliminary report from our group (XI), showing a prevalence of anti-HCV antibodies in HCC ranging from 35% to 79%, with higher values in areas which also have a high prevalence of HBV infection and HCC, such as the Mediterranean countries (21). The aim of this study was to confirm our preliminary results on a larger series and to ascertain whether anti-HCV-positive HCC presents particular clinical features and may be considered a specific clinical entity. The final goal of the study being to contribute to the understanding of the role of HCV in the development of HCC.
et al.
tients with chronic liver disease (CLD), subgrouped as follows: 49 patients with HBsAg CLD; 42 with alcohol abuse-related CLD; and 132 with cryptogenic (no viral infection markers, no signs of auto-immune disease or of hemochromatosis, Wilson, a,-antitrypsin deficiency) or post-transfusional non-A, non-B CLD. In all these patients the presence of HCC was ruled out on the basis of US and a-fetoprotein findings and/or a 1 year follow-up. The following anamnestic parameters were recorded: age at diagnosh; male/female ratio; length of history of CLD before diagnosis; history of blood transfusion or surgery; type of presentation (as a result of screening in CLD or occasional finding in patients not known as harbori:ig CLD); diameter and number of the neoplastic mass(ec,); a-fetoprotein levels; and survival (in months), resulting from follow-up at out-patients clinic or information fro:m the death registers of the town of residence. Data were statistically analysed using Student’s t-test, analysis of variance (Anova one-way), chi-square and (for evaluation of survival) actuarial survival curves, by life table analysis, according to the Generalized Savage (Mantel-Cox) and Generalized WiIcoxon (Breslow) methods.
Patients and Methods Ninety-seven patients with histologically confirmed HCC, prospectively collected, were included in the study. Diagnosis was made using US-guided fine needle cytology (21-22 gauge) or microhistology (18-20 gauge). All patients had associated cirrhosis. Mean age of patients was 61 years, range 41-76, male/female ratio 5:l. The mean age was slightly lower in males (60) than in females (67). In all the patients, HBV markers (HBsAg, antiHBs, anti-HBc) were sought using conventional laboratory tests (EIA, Abbot Lab, Chicago, IL). Patients with a routine alcohol consumption of more than 80 g/day were considered alcohol abusers. HBsAg-negative patients, with no signs of autoimmune disease, hemochromatosis or storage hepatic diseases, and lacking history of alcohol abuse, were considered as having ‘cryptogenic’ HCC. Anti-HCV antibodies were tested by the Ortho HCV antibody ELISA System, in duplicate, with a cut-off of 0.400 in optical density (OD). A smzller series of sera (corresponding to 30 patients, affected by both CLD and HCC) werealsochecked by the Recombinant Immunoblot ASsay (RIBA) and a 98% overall concordance with the results of the ELISA system was obtained. The mean value of anti-HCV was 1.744, with a range of 0.6->2.000. No correlation was found between anti-HCV and globulin levels (r = 0.1854,P = n.s.) or with sample storage duration. a-Fetoprotein levels were determined by an immunoradiometric assay (IRMA). We also studied 223 pa-
Results Anti-HCVprevalence The overall prevalence of anti-HCV antibodies in WCC
patients was 64% (62197). When patients were subgrouped according to the putative etiology of their disease, the prevalence was significantly lower in HBsAgpositive HCC (42.8%) patients than in alcohol abuse-related (60%) or cryptogenic HCC (80%) patients (p c 0.005 by chi-square trend) (Fig. 1). The overall prevalence of anti-HCV antibodies in HCC patients was not significantly different from the findings in the 223 CLD patients (60.9%). When the subgroups of HCC and CLD patients with the
80%
60%
40%
20%
OX TOTAL
U‘COHOL HCC
m
cRt-Fr0G.
cm0s1s
Fig. 1. Prevalence of anti-HCV in HCC and cirrhosis of comparable etiology. *p C 0.005; +p-=f 0.05.
ANTI-HCV POSITIVE HEPATOCELLULAR
CARCINOMA
185
TABLE 1 Anti-HCV and anti-HBc in HBsAg-negative I-ICC and cryptogerw i “I.D Anti-HCV+ Anti-HBc+
HCC CLD
--
Anti-HCVAnti-HBc-
Anti-I-XV+ Anti-HBc-
-
Anti-HCVAnti-HBc+ -.
-
No.
%
No.
%
No.
%
NO.
%
11/62 42!106
18* 40*
9135 19126
26” 73*
51162 64llCb6
82” 60”
2613.5 7J26
74** 27**
*p < 0.005. “p
183
HEPAT 00942
revalence, ris
clinical
features
F. Farina&, S. Fagiuoli’, N. De Maria’, M. Chiaramonte’, V. Aneloni*, S. Qngaro*, . Salvagnini’ and R. Naccarato’ ‘Cattedra Malattie Apparato Digerente, Istituto Di Medicina Interna, Universita’ Di Padova and %entro Traq%sionale @p-&de Civile Di PGdova, Padova, Italy (Receiwd
28 August 1990)
Recent reports indicate that hepatitis C virus (HCV) may play a role in the pathogenesis of hepatocellular carcinoma in cirrhotics. Using an ELISA test, we evaluated the prevalence of anti-HCV antibodies in 97 patients with hepatocellular carcinoma (HCC) in cirrhosis and in a group of 223 patients, including: 49 patients with HBsAg-positive chronic liver disease (CLD), 42 with alcoholic CLD, 110 with cryptogenic CLD and 22 with post-transfusional HBsAg-negative CLD. All diagnoses were histologically confirmed. Overall, anti-HCV-positive HCC were 64% of the total, with no statistically sign& cant difference with respect to CLD (60.9%). The prevalence of anti-HCV was higher in cryptogenic HCC (80%) than in HBsAg-positive (60%) or alcoholic HCC (42.8%) (p C 0.005). When HCC and cirrhosis of similar putative etiology were considered, anti-HCV prevalence was significantly higher in HCC than in cirrhosis only in the groups of patients with alcoholic liver damage (60% in HCC vs. 38% in cirrhosis, p < 0.005). In HBsAg-positive patients, anti-HCV prevalence was twice as high in HCC than in CLD, but the difference was not statistically significant. Overall, anti-HCV prevalence in HCC was significantly higher than in alcoholic or HBsAg-positive CLD (p c 0.001 and p < 0.01, respectively) but lower than in cryptogenic CLD (p < 0.001). Association between anti-HCV and anti-HBc was significantly more prevalent in patients with CLD than in those with HCC. From the clinical point of view, multifocal lesions and lack ofa history of previous liver disease were more frequent in anti-HCV-positive HCC (p < 0.05) than in HBsAg-positive or alcoholic patients. AntiHCV-positive HCC had a significantly worse proguosis @ < 0.005 with the Generalized Wilcoxon-Breslow test), when survival in the different subgroups was examined with life table analysis. No statistically significant differences were observed with respect to the remaining parameters considered. In conclusion, HCV infection appears to be a frequent event in HCC and CLD, particularly in patients with cryptogenic liver disease but, with the exception of patients with alcoholic liver damage, the prevalence of anti-HCV appears to be similar in HCC and cirrhosis. This could indicate that the infection plays an important role only in specific subgroups of HCC patients. Anti-HCV-positive HCC is frequently diagnosed in a more advanced stage and has a worse prognosis but, since the clinical history of this subgroup of patients frequently makes its debut with HCC, in our opinion the prognosis is worse due to late diagnosis. -
Although several factors have been linked to the development of hepatocellular carcinoma (HCC), thus far, V) has been considered t hepatitis B virus ( risk factor on the basis of evidence from epidemiology
(l-3), (6-8).
molecular biology (4,5) and biological studies However, HBV has never been shown to transf0 in cell cultures (9) and the data on the integration of viral DNA has never revealed the uniform presence of a
Correspondence: Dr. Fabio Farinati, Cattedra Malattie Apparato Digerente, niani 2,351OO Padua, Italy.
Istituto di Medicina Interna, Policlinico Universitario,
Via Giusti-
F. FARINATI
184 particular sequence or shown viral DNA fragments close to oncogenes or sequences considered to be relevant in terms of carcinogenesis (1 ,lO). Alcohol has also been considered to play a role in the development of liver callcer, although the association between alcohol and hepatocellular carcinoma is probably indirect (11,12). Morerecently, theavailability of a test which detects an antibody retogni&g an epitope directly linked to HCV, has prompted research on the possible association between HCV infection and HCC. A number of papers have been published (13-19), including a preliminary report from our group (XI), showing a prevalence of anti-HCV antibodies in HCC ranging from 35% to 79%, with higher values in areas which also have a high prevalence of HBV infection and HCC, such as the Mediterranean countries (21). The aim of this study was to confirm our preliminary results on a larger series and to ascertain whether anti-HCV-positive HCC presents particular clinical features and may be considered a specific clinical entity. The final goal of the study being to contribute to the understanding of the role of HCV in the development of HCC.
et al.
tients with chronic liver disease (CLD), subgrouped as follows: 49 patients with HBsAg CLD; 42 with alcohol abuse-related CLD; and 132 with cryptogenic (no viral infection markers, no signs of auto-immune disease or of hemochromatosis, Wilson, a,-antitrypsin deficiency) or post-transfusional non-A, non-B CLD. In all these patients the presence of HCC was ruled out on the basis of US and a-fetoprotein findings and/or a 1 year follow-up. The following anamnestic parameters were recorded: age at diagnosh; male/female ratio; length of history of CLD before diagnosis; history of blood transfusion or surgery; type of presentation (as a result of screening in CLD or occasional finding in patients not known as harbori:ig CLD); diameter and number of the neoplastic mass(ec,); a-fetoprotein levels; and survival (in months), resulting from follow-up at out-patients clinic or information fro:m the death registers of the town of residence. Data were statistically analysed using Student’s t-test, analysis of variance (Anova one-way), chi-square and (for evaluation of survival) actuarial survival curves, by life table analysis, according to the Generalized Savage (Mantel-Cox) and Generalized WiIcoxon (Breslow) methods.
Patients and Methods Ninety-seven patients with histologically confirmed HCC, prospectively collected, were included in the study. Diagnosis was made using US-guided fine needle cytology (21-22 gauge) or microhistology (18-20 gauge). All patients had associated cirrhosis. Mean age of patients was 61 years, range 41-76, male/female ratio 5:l. The mean age was slightly lower in males (60) than in females (67). In all the patients, HBV markers (HBsAg, antiHBs, anti-HBc) were sought using conventional laboratory tests (EIA, Abbot Lab, Chicago, IL). Patients with a routine alcohol consumption of more than 80 g/day were considered alcohol abusers. HBsAg-negative patients, with no signs of autoimmune disease, hemochromatosis or storage hepatic diseases, and lacking history of alcohol abuse, were considered as having ‘cryptogenic’ HCC. Anti-HCV antibodies were tested by the Ortho HCV antibody ELISA System, in duplicate, with a cut-off of 0.400 in optical density (OD). A smzller series of sera (corresponding to 30 patients, affected by both CLD and HCC) werealsochecked by the Recombinant Immunoblot ASsay (RIBA) and a 98% overall concordance with the results of the ELISA system was obtained. The mean value of anti-HCV was 1.744, with a range of 0.6->2.000. No correlation was found between anti-HCV and globulin levels (r = 0.1854,P = n.s.) or with sample storage duration. a-Fetoprotein levels were determined by an immunoradiometric assay (IRMA). We also studied 223 pa-
Results Anti-HCVprevalence The overall prevalence of anti-HCV antibodies in WCC
patients was 64% (62197). When patients were subgrouped according to the putative etiology of their disease, the prevalence was significantly lower in HBsAgpositive HCC (42.8%) patients than in alcohol abuse-related (60%) or cryptogenic HCC (80%) patients (p c 0.005 by chi-square trend) (Fig. 1). The overall prevalence of anti-HCV antibodies in HCC patients was not significantly different from the findings in the 223 CLD patients (60.9%). When the subgroups of HCC and CLD patients with the
80%
60%
40%
20%
OX TOTAL
U‘COHOL HCC
m
cRt-Fr0G.
cm0s1s
Fig. 1. Prevalence of anti-HCV in HCC and cirrhosis of comparable etiology. *p C 0.005; +p-=f 0.05.
ANTI-HCV POSITIVE HEPATOCELLULAR
CARCINOMA
185
TABLE 1 Anti-HCV and anti-HBc in HBsAg-negative I-ICC and cryptogerw i “I.D Anti-HCV+ Anti-HBc+
HCC CLD
--
Anti-HCVAnti-HBc-
Anti-I-XV+ Anti-HBc-
-
Anti-HCVAnti-HBc+ -.
-
No.
%
No.
%
No.
%
NO.
%
11/62 42!106
18* 40*
9135 19126
26” 73*
51162 64llCb6
82” 60”
2613.5 7J26
74** 27**
*p < 0.005. “p
