Page 82 - EASL POSTGRADUATE COURSE
P. 82
from HCC is higher than for other types of cancer (e.g. 4.5 in men with a BMI of 35kg/m2) [6]. The
cumulative incidence of HCC is three times higher in patients with T2DM than in patients without
T2DM [7]. There are strong pathophysiological mechanisms linking obesity, T2DM and NAFLD.
It will be difficult and probably of limited clinical relevance to disentangle these three conditions to
estimate their independent roles as risk factors for HCC.
HCC in NAFLD: often without cirrhosis
Beside a higher prevalence of obesity and T2DM, NAFLD patients with HCC have a lower prevalence
of cirrhosis than patients with HCC and another underlying liver disease. This is an important
characteristic of HCC in NAFLD which has been reported in multiple publications. In a comparison
of HCC in patients with MetS and HCC in patients with chronic liver disease, the background liver
was significantly more often free of significant fibrosis (F0-F2) in the former group than in the latter
(65% vs. 26%, respectively, P<0.001) [8]. A study of NASH patients with HCC found that 39% and
70% of males and females, respectively, had cirrhosis [9]. In another study the proportion of cirrhotic
patients was only 63% in the 596 NAFLD patients with HCC, compared with 81% of 161 autoimmune
hepatitis patients with HCC, 88% of 166 primary biliary cirrhosis-patients with HCC and 80% of
1,423 alcoholic liver disease patients with HCC [10]. In a US national cohort of 1,500 patients who
developed HCC from Veterans Administration hospitals, cirrhosis was less common in NAFLD-related
cases (58.3%) compared with alcoholic liver disease- or HCV-related HCCs [11].
HCC in NAFLD: implications for management
The incidence of HCC in NASH cirrhosis appears to be ~2%/year [1], which is above the 1.5% suggested
cut-off for the implementation of surveillance [12]. When considering patients with advanced cirrhosis
the five-year cumulative incidence of HCC is 7.6%, so just in this range, and in this population, HCC
is the leading cause of death, emphasising the importance of surveillance [13]. However, a significantly
higher percentage of patients with NAFLD-related HCC did not receive HCC surveillance in the
three years before their HCC diagnosis, compared with alcohol- or HCV-associated HCC patients.
Furthermore, 62% and 78% (P<0.01) of NAFLD-related HCC patients and HCV-related HCC
patients, respectively, received HCC-specific treatment [11].The fact that fewer patients with NAFLD-
related HCC are enrolled in a surveillance programme might be explained by the following 3 factors: 1)
as reviewed above, many of these patients do not have a cirrhosis, the stage which sets the implementation
of surveillance; 2) ultrasonography, the screening test, can be difficult in NAFLD patients due to their
obesity; however, no alternative strategy has been adequately tested; 3) co-morbidities or advanced age
makes the discovery of HCC irrelevant. This last aspect also contributes to the fact that these patients
receive less HCC-specific treatments.
Since many patients with NAFLD develop HCC before reaching a cirrhotic stage, one needs to better
stratify the patients at risk. This strategy should only enrol in surveillance programmes pre-cirrhotic
patients at risk. Presently, there are no parameters permitting the implementation of such a strategy. It is
possible it will include genetic testing, such as determination of the PNPLA3 polymorphism. Genotype
frequencies were significantly different between 100 NAFLD-HCC cases and 275 NAFLD-controls
(P=0.0001), with enrichment of the rs738409 minor (G) allele. Carriage of each copy of the allele
conferred an additive risk for HCC, with GG homozygotes exhibiting a 5-fold increased risk over CC.
When compared to the UK general population the risk-effect was even more pronounced (GG vs. CC:
OR, 12.19) [14].
The therapeutic options for patients with NAFLD are the same as those for patients with other underlying
liver disease [15]. Resection might be more frequently considered since these patients are often not
cirrhotic. Comorbidities have to be taken into account, particularly CVD and liver steatosis, which
negatively impact liver regeneration. Sorafenib is the only indicated drug for HCC patients who are
eligible for systemic, targeted therapy. There is currently no evidence that NAFLD-patients with HCC
may benefit from such therapy. In terms of prevention, several studies have suggested that metformin
lowers the risk of HCC in T2DM patients [16].
82 Postgraduate Course Syllabus • Metabolic Liver Disease
cumulative incidence of HCC is three times higher in patients with T2DM than in patients without
T2DM [7]. There are strong pathophysiological mechanisms linking obesity, T2DM and NAFLD.
It will be difficult and probably of limited clinical relevance to disentangle these three conditions to
estimate their independent roles as risk factors for HCC.
HCC in NAFLD: often without cirrhosis
Beside a higher prevalence of obesity and T2DM, NAFLD patients with HCC have a lower prevalence
of cirrhosis than patients with HCC and another underlying liver disease. This is an important
characteristic of HCC in NAFLD which has been reported in multiple publications. In a comparison
of HCC in patients with MetS and HCC in patients with chronic liver disease, the background liver
was significantly more often free of significant fibrosis (F0-F2) in the former group than in the latter
(65% vs. 26%, respectively, P<0.001) [8]. A study of NASH patients with HCC found that 39% and
70% of males and females, respectively, had cirrhosis [9]. In another study the proportion of cirrhotic
patients was only 63% in the 596 NAFLD patients with HCC, compared with 81% of 161 autoimmune
hepatitis patients with HCC, 88% of 166 primary biliary cirrhosis-patients with HCC and 80% of
1,423 alcoholic liver disease patients with HCC [10]. In a US national cohort of 1,500 patients who
developed HCC from Veterans Administration hospitals, cirrhosis was less common in NAFLD-related
cases (58.3%) compared with alcoholic liver disease- or HCV-related HCCs [11].
HCC in NAFLD: implications for management
The incidence of HCC in NASH cirrhosis appears to be ~2%/year [1], which is above the 1.5% suggested
cut-off for the implementation of surveillance [12]. When considering patients with advanced cirrhosis
the five-year cumulative incidence of HCC is 7.6%, so just in this range, and in this population, HCC
is the leading cause of death, emphasising the importance of surveillance [13]. However, a significantly
higher percentage of patients with NAFLD-related HCC did not receive HCC surveillance in the
three years before their HCC diagnosis, compared with alcohol- or HCV-associated HCC patients.
Furthermore, 62% and 78% (P<0.01) of NAFLD-related HCC patients and HCV-related HCC
patients, respectively, received HCC-specific treatment [11].The fact that fewer patients with NAFLD-
related HCC are enrolled in a surveillance programme might be explained by the following 3 factors: 1)
as reviewed above, many of these patients do not have a cirrhosis, the stage which sets the implementation
of surveillance; 2) ultrasonography, the screening test, can be difficult in NAFLD patients due to their
obesity; however, no alternative strategy has been adequately tested; 3) co-morbidities or advanced age
makes the discovery of HCC irrelevant. This last aspect also contributes to the fact that these patients
receive less HCC-specific treatments.
Since many patients with NAFLD develop HCC before reaching a cirrhotic stage, one needs to better
stratify the patients at risk. This strategy should only enrol in surveillance programmes pre-cirrhotic
patients at risk. Presently, there are no parameters permitting the implementation of such a strategy. It is
possible it will include genetic testing, such as determination of the PNPLA3 polymorphism. Genotype
frequencies were significantly different between 100 NAFLD-HCC cases and 275 NAFLD-controls
(P=0.0001), with enrichment of the rs738409 minor (G) allele. Carriage of each copy of the allele
conferred an additive risk for HCC, with GG homozygotes exhibiting a 5-fold increased risk over CC.
When compared to the UK general population the risk-effect was even more pronounced (GG vs. CC:
OR, 12.19) [14].
The therapeutic options for patients with NAFLD are the same as those for patients with other underlying
liver disease [15]. Resection might be more frequently considered since these patients are often not
cirrhotic. Comorbidities have to be taken into account, particularly CVD and liver steatosis, which
negatively impact liver regeneration. Sorafenib is the only indicated drug for HCC patients who are
eligible for systemic, targeted therapy. There is currently no evidence that NAFLD-patients with HCC
may benefit from such therapy. In terms of prevention, several studies have suggested that metformin
lowers the risk of HCC in T2DM patients [16].
82 Postgraduate Course Syllabus • Metabolic Liver Disease
