Hepatocellular Carcinoma in Cirrhotic Versus Non-Cirrhotic Patients: A Retrospective Study of 483 Patients

: Background and Aim: Although cirrhosis is a classical risk factor for the development of hepatocellular carcinoma (HCC), its absence does not exclude this risk. We aimed to assess the clinical characteristics and outcomes of cirrhotic HCC (C-HCC) and non-cirrhotic HCC (NC-HCC) patients. Methods: Patients consecutively included in a prospective HCC cohort (University Hospital Bern) were analysed. They were categorised into two groups, based on the basis of histology or combined radiological and laboratory characteristics. Results: 20.4% of patients were NC-HCC. This group was characterized by a higher median age and a higher female prevalence compared to the C-HCC group. Non-alcoholic fatty liver disease (NAFLD) (25.7%) and HBV infection (14.9%) were the main risk factors in this group, whereas alcohol abuse (26%) and HCV (21.6%) in C-HCC, P<0.001. 19.4% of them were diagnosed during a screening programme. Resection was performed in 54.5% of NC HCC patients despite the advanced stage (BCLC stage B and C). No statistically significant difference in survival rate was observed between C and NC-HCC patients (24 months vs. 33.9 months, P=0.162). In multivariate analysis, in the NC-HCC group each unit increase in BMI was associated with mortality while liver transplantation and resection were positively associated with survival. In the C-HCC group, the BCLC stage C was negatively associated with survival while all the therapeutic lines were negative factors for mortality. Conclusion: NC-HCC hemochromatosis, Wilson disease, autoimmune hepatitis, alcohol abuse, primary biliary cholangitis, primary


Introduction
Classically hepatocellular carcinoma (HCC) develops on a cirrhotic liver. In the last few decades, it was reported that a certain number of patients develop HCC in non-cirrhotic liver (NC-HCC) [1]. Occurrence of NC-HCC ranges from 7 to 54%, with significant geographical and etiological variability [2]. Recent data suggests that obesity, diabetes, non-alcoholic liver disease (NAFLD), especially in its progressive form (i.e. NASH), are risk factors associated with HCC development [2][3][4][5]. This represents an alarming issue, as the NAFLD prevalence globally is high and increasing (from 15% in 2005 to 25% in 2010), while the global prevalence of NASH among biopsied NAFLD patients during the same timeframe has almost doubled [6]. Therefore, our aim is to evaluate the clinical features of NC-HCCs and to compare them with C-HCCs.

Materials and Methods
All patients diagnosed with HCC and treated at the Department of Visceral Surgery and Medicine, University Hospital of Bern (Inselspital), were asked to participate in the prospective HCC cohort from the University Hospital Bern (Bern HCC Cohort, BHC), Switzerland. The local ethics committee (Kantonale Ethikkommission Bern, Bern, Switzerland) approved collection of patient information and the study protocol (project-ID 2017-00957), which was consistent with the principles of the current version of the Declaration of Helsinki.
Patient characteristics such as demographics, Model for End-Stage Liver Disease (MELD) score, Child Pugh class, complications (ascites, encephalopathy, varices), performance status (Eastern Cooperative Oncology Group performance status 0-5), Barcelona Clinic Liver Cancer (BCLC) HCC stage at diagnosis (A-D), presence of portal vein thrombosis or extrahepatic metastasis, medical comorbidities, were collected on entry into the Bern HCC cohort, which have been previously described by several authors [7][8][9][10]. Patients were classified as noncirrhotic, based on one of the following criteria: histology either in biopsy within 1 year of HCC diagnosis or in a resection specimen in combination with the absence of radiological features of cirrhosis, or endoscopic lack of varices and thrombocytopenia (platelets <150) and radiological absence of splenomegaly (spleen >11 cm) or absence of radiological features of cirrhosis and FibroScan values < 12.5KPa [11].
Alcohol abuse was defined as a chronic alcohol intake exceeding 30 g/day in men and > 20 g/d in women or if a history of chronic alcohol abuse was positive in the past medical history [12]. NAFLD was considered based on either documented steatosis/steatohepatitis on liver biopsy or based on ultrasonographic features of fatty liver if all other known etiologies of liver disease could be ruled out. HCV status was determined by the presence of positive anti-HCV or HCV RNA tests detected any time before or after HCC diagnosis. HBV was defined by a positive surface antigen. Patients having none of the known risk factors (i.e. HCV, HBV, hemochromatosis, Wilson disease, autoimmune hepatitis, alcohol abuse, primary biliary cholangitis, primary sclerosing cholangitis were classified as idiopathic HCC).
The diagnosis and the treatment were performed in line with the current guidelines [13]. Patients were considered screened if they had previous negative screening test (imaging test +/-Alfa-fetoprotein) at 6 months interval within the 18 months preceding the diagnosis of HCC. Measurement of the tumor volume was calculated only for patients, who underwent resection, by using the following mathematical equation: 4/3 πr3, where r is the maximum radius of each tumor. In the case of multiple tumors, the TTV was calculated through the sum of each tumor nodule.

Statistical Analysis
Data was expressed as median and interquartile range or means and standard deviations after testing the normal distribution. Normality of data was tested by Kolmogorov-Smirnov test. When comparing the baseline characteristics, the Student t-test and Mann-Whitney U test were used for continuous data, and Chi-Square test for categorical variables. Overall, survival was defined as the time from HCC diagnosis to the time of death, the time of last follow-up evaluation or the date of data censoring. The Kaplan-Meier survival curves and log-rank tests were used to compare survival rates between the C-HCC and NC-HCC groups and between screened and non-screened C-HCC and NC-HCC group. Univariate and multivariate logistic regression analyses were performed to evaluate patient characteristics (i.e. sex, age, etiology of underlying liver disease, surveillance, extrahepatic metastasis, portal vein tumoral invasion, BCLC stage, therapies) associated with risk of mortality. A P value < 0.05 was considered statistically significant. All analyses were performed with SPSS version 25 (SPSS, Chicago, IL).

Results
Between the 1 st of January 2010 and the 31 st of August 2020, 483 HCC patients were included in the cohort study, of which 101 (20.9%) presented NC-HCC. Comparison of the baseline characteristics between the two groups revealed several significant differences (Tables 1 & 2). Values are presented as n (%), mean (SD, standard deviation), median (IQR, interquartile range).
The p values apply to the cirrhotic versus non-cirrhotic group. P value was assumed to be significant when <0.05. N represents the patients' number with available data.
HCV: Hepatitis C Virus; HBV: Hepatitis B Virus; NAFLD: Non-Alcoholic Liver Disease; Other* include: hemochromatosis, primary biliary cholangitis, and autoimmune hepatitis; ECOG PS: Eastern Cooperative Oncology Group Performance Status; MELD: Model for End-stage Liver Disease.  Table 2). NC-HCC patients who underwent liver resection had a TTV significantly higher than their counterparts (P<0.001).

I Survival in Cirrhosis Versus No Cirrhosis HCC Patients
Over a median follow-up of 69.4 months, the median survival of NC-HCC patients was longer as compared to C-HCC patients, without reaching the statistical significance (34 months vs. 24 months, P=0.162) (Figure 1).  (Figure 2).

II Risk Associated with Mortality in Cirrhotic and Non-Cirrhotic HCC Patients
In the survival analysis both HCC group types involve different risk factors. In a univariate Cox regression analysis of the C-HCC group, the presence of portal vein thrombosis, extrahepatic metastases were independently associated with mortality, whereas BCLC stage (0-B) and treatment had a positive effect on survival. In the NC cohort, a unit increase in BMI was found as an independent factor for mortality, while BCLC stage (A-B) and treatment were associated with a good outcome (Table 3).   In the multivariate analysis of the C-HCC group, only BCLC stage C was associated with mortality, while all the therapeutic lines were associated positively with a good outcome. In the NC-HCC, each unit increase in BMI is associated negatively with survival while therapy (transplant, resection) was associated positively with survival (Table 4).

Discussion
In our HCC cohort, the demographic characteristics (age, gender prevalence), risk factors, tumoral stage at diagnosis and treatment approach differed between NC-HCC and C-HCC. We found no statistically significant difference between the overall survival of NC-HCC and C-HCC patients. In survival analysis both HCC groups involved different risk factors. Similar to several previous studies, our NC-HCC group was characterised by an older age and a higher females prevalence compared to the C-HCC group [2,4,[14][15][16]. However, this observation was not confirmed by other studies [17,18]. The main etiological factor in the cohort seems to be reason for the conflicting results, thus the cohorts in which HBV was the main risk factor had a younger population and higher male proportion, while the prevalence of NAFLD was associated with older age and female prevalence [19].
Despite extensive tests, we could not establish the underlying liver disease in 14.9% of NC-HCC and 2.3% of C-HCC cases. Several pathologies have been recognised to be associated with HCC in the subgroup of patients for whom no other classic risk factor could be identified. Previous studies have found that obesity and type 2 diabetes are independent risk factors for HCC occurrence [20][21][22][23]. `These risk factors were present in 32% and 40%, respectively of NC-HCC without a classical risk factor. In addition, hypothyroidism, porphyria or other environmental factors have been proposed as possible associated conditions for HCC occurrence [24][25][26]. In our NC group two patients had hypothyroidism as the only associated pathology. In cirrhotic livers, the "burned-out NASH" was reported as the cause be the for the lack of histological changes [27].
The positive impact of HCC surveillance in cirrhotic patients on an earlier diagnosis, assignment to curative therapy and a long-term survival is universally accepted. A matter of debate still is the benefit of a surveillance programme in NC patients outside the actual EASL guidelines (i.e. chronic HBV at intermediate or high risk of HCC and non-cirrhotic F3 patients, regardless of etiology). Based on EASL recommendation, 19.4% of NC-HCC patients in our cohort were diagnosed during a screening programme [28,29]. However, it is important to mention that the rate of adherence to a screening programme in cirrhotic population was low, only 43.5%. The relative low adherence to surveillance (52%) is a serious problem and it was reported by Zhao C in recent metanalysis [30].
In terms of survival, in our cohort the difference of survival time between the two groups did not reach the statistical significance. In previous studies, several groups reported that the NC-HCC group had a better survival rate compared to C-HCC group but are also groups that did not find any significant difference [16,31]. In contrast with the C-HCC group, in which the screened population had a significantly better survival rate, in the NC-HCC group, we found no statistically significant difference. In this context and considering that the cirrhotic patients showed a relative low adherence to the screening programme, the adherence of NC patients is questionable. In line with previous studies, in our cohort despite an advanced stage at diagnosis, the preserved liver function, allowed to be performed more aggressive therapies with a positive impact on survival [31]. This suggests indirectly that for NC-HCC patients, BCLC staging is not suitable, since a high number of NC-HCC patients could benefit from a more aggressive approach [32]. However, this advantage of a preserved liver function seems to be limited by the associated comorbidities (i.e. obesity).
While the current study included a relatively large number of patients, it is limited by its retrospective nature and single center data. Moreover, since the diagnosis of HCC do not require a biopsy from the tumorous and non-tumorous liver tissue as part of the standard work-up, the number of patients with histology from non-tumorous tissue was rather small. However, a relative high percentage of NC-HCC patients was assigned to resection, which allowed us to have hepatic tissue available. Moreover, in our hospital during percutaneous ablation, biopsy is often performed. Despite available hepatic tissue, a certain degree of misclassification between ASH and NAFLD could not be avoided. Nevertheless, this study provides important overview on the differences between NC-HCC and C-HCC patients in a European cohort.
In conclusion, NAFLD represents an important risk factor for the occurrence of HCC in non-cirrhotic livers. The lack of a screening programme for NC-HCC population results in an incidental and late HCC diagnosis. However, in the absence of cirrhosis, the majority of these patients are able to undergo curative treatment with no statistically significant difference in survival rates between groups.