Patients with HCC who were followed-up in the Department of Gastroenterohepatology, Istanbul Faculty of Medicine, Istanbul University between January 2001 and August 2011 were included in this single center, retrospective cohort study. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and it was approved by the ethics committee of Istanbul Faculty of Medicine, Istanbul University. The diagnosis of HCC was based on the recommendations reported by the EASL panel of experts in 2001[6]. According to these recommendations; the diagnosis was established by histopathological and/or radiological criteria. CPS was calculated at baseline in cirrhotic patients as previously described[7]. BCLC stage was determined in every patient with HCC at initial diagnosis according to the extent of tumor, performance status, CPS, vascular invasion and extrahepatic spread[5]. We reviewed demographic, clinical and staging characteristics, laboratory data, etiology of primary liver disease, imaging characteristics and treatments of HCC patients. Number and size of nodules, total tumor diameter (TTD), type of tumor (single nodule, multinodular or diffuse-infiltrative), presence of major vascular involvement and extrahepatic metastasis were determined according to baseline imaging records. Survival data of all patients were updated as of November 2011.

All patients with a suspicion of primary liver cancer were evaluated by clinical, laboratory and imaging studies including a 4-phase computerized tomography (CT) scan or dynamic contrast enhanced magnetic resonance imaging (MRI). The diagnosis was made if there were radiologic hallmarks of HCC as arterial hypervascularity and venous/late phase wash out. In the absence of radiologic hallmarks of HCC or if findings were inconsistent on contrast enhanced CT or MRI, a biopsy was obtained and assessed by an expert hepatopathologist. Extrahepatic metastasis was screened by a contrast-enhanced chest CT and whole-body bone scan.

All patients with a diagnosis of HCC had baseline physical examination, and results of standard laboratory investigations including complete blood count, renal and liver function tests, screening tests for hepatitis viruses (hepatitis B surface antigen-HBsAg, hepatitis B core antibody-anti-HBc total and hepatitis C antibody-anti-HCV) and AFP. If HBsAg or anti-HCV was detected to be positive further tests [HBeAg, anti-HBe, HBVDNA (PCR), and anti-Delta total (plus HDVRNA when positive) or HCVRNA, respectively] were obtained. Most patients with HCC already had a diagnosis of a chronic liver disease, and remaining patients with unrecognized liver disease had undergone detailed evaluations to assess the presence of other etiologies including alcoholic liver disease, hemochromatosis, Budd-Chiari syndrome, non-alcoholic fatty liver disease, and autoimmune liver diseases. An upper gastrointestinal endoscopy was performed to evaluate the presence of esophageal or gastric varices in each patient.

Treatment of HCC was guided by BCLC classification, however most patients were listed for OLT using the expanded criteria after 2001 as previously described[8]. Treatment options included surgical resection, OLT, percutaneous ablation (RFA or ethanol/acetic acid ablation), TACE, Yttrium-90 radioembolization and systemic therapy using sorafenib. Curative partial hepatectomy was performed in patients with tumors confined to one lobe of the liver that shows no radiographic evidence of invasion of the hepatic vasculature, no evidence of portal hypertension and adequate liver functional reserve. All candidates for surgical resection underwent indocyanine green test to determine operative risk before hepatectomy. Percutaneous ablation was selected in patients who did not meet resectability criteria and had a single tumor ≤ 3 cm in diameter. Patients without a suitable living-donor were listed for OLT. Listed patients who had an anticipated time to OLT more than 6 mo underwent percutaneous ablation, TACE or Yttrium-90 radioembolization decided by physician’s discretion according to tumor characteristics and hepatic reserve. Patients with advanced stage HCC who were not candidates for curative treatments underwent TACE, Yttrium-90 radioembolization or sorafenib therapy. In terminal stage, patients were followed-up under natural course with best supportive care.

Twenty-seven patients were suitable for hepatic resection and underwent surgery. Twelve patients who had 1 or 2 small (≤ 3 cm) nodules underwent percutaneous ablation with RFA, and 5 patients with a single nodule ≤ 2 cm underwent percutaneous acetic acid/ethanol injection. Two-hundred and sixty-seven patients who were ineligible for surgical resection or percutaneous ablation, but had tumor characteristics that are compatible with expanded criteria were listed for OLT. A total of 56 patients (47 within expanded criteria) with HCC underwent OLT during the follow-up period, due to the shortage of cadaveric organs or unavailability of a suitable living donor. The number of patients who underwent TACE was 172. Among them 90 patients underwent 1 session, 53 patients underwent 2 sessions, and 29 patients underwent ≥ 3 sessions of TACE. The distribution of treatment modalities in the remaining patients were as follows: 19 patients underwent Yttrium-90 radioembolization, 16 patients received systemic therapy with sorafenib, and 238 patients received no treatment until the end of the follow-up. Treatment characteristics of patients with HCC were summarized in Table 1.

Continuous variables are presented as mean ± standard deviation (SD) or median (range) while categorical variables were expressed as frequencies (%). Differences between frequencies were evaluated using Pearson χ² or Fisher’s exact test when necessary. Predictor variables of vascular invasion and extrahepatic metastasis were investigated by univariate and multivariate logistic regression analyses. Univariate Cox regression analyses were performed to find out factors associated with overall survival of patients with HCC. Variables which showed a significant influence (P < 0.05) were included in a multivariate Cox proportional hazard model to find out independent prognostic factors that affect overall survival. The results of the model were presented as a hazard ratio (HR) with 95% confidence interval (CI). We performed Kaplan-Meier analyses to determine cumulative survival probabilities and treatment effect on overall survival. Log-rank test was used for the statistical comparison of Kaplan-Meier curves. All statistical analyses were performed using IBM SPSS v20 (IBM SPSS Inc., Chicago, IL, United States). A two tailed P-value < 0.05 was considered statistically significant.

A total of 545 patients (449 male, mean age 59.5 ± 10) with HCC who were diagnosed and followed-up between January 2001 and August 2011 were included in the study. The diagnosis of HCC was established by CT or MRI in 459 patients, and the remaining patients underwent liver biopsy due to inconsistent findings in radiological examinations. The number of patients with underlying chronic liver disease was 532, 13 (2.3%) patients had normal liver without any identifiable risk factor for the development of HCC. 350 (66%) of patients with chronic liver disease were already aware of their underlying liver disease at the time of the diagnosis of HCC. However, only 110 patients (31.4%) were under regular follow-up with a combined use of scheduled liver ultrasonography and AFP measurement. The mean estimated duration of chronic liver disease was 69 ± 60 mo (range, 3-420 mo).

The number patients according to underlying etiology for chronic liver disease were as follows: 287 patients with chronic HBV, 120 patients with chronic HCV, 37 patients with chronic delta hepatitis, 10 patients with co-infection of HBV and HCV, 39 patients with cryptogenic liver disease, 21 patients with alcoholic liver disease, 10 patients with non-alcoholic fatty liver disease, 5 patients with autoimmune liver disease, 2 patients with Budd-Chiari syndrome and 1 patient with hemochromatosis. Among patients with chronic viral hepatitis, there were 24 patients who had significant alcohol consumption (> 210 g/wk) which may contribute to the severity of underlying liver disease. The majority of patients with chronic liver disease had cirrhosis at different stages (CPS A, 247 patients; CPS B, 140 patients; CPS C, 90 patients), and 68 patients (12.5%) were pre-cirrhotic based on clinical and/or histopathological examinations. Baseline patient characteristics are summarized in Table 2.

The number of patients with a single tumor was 333 (61%), and the remaining patients had multinodular (191 patients, 35%) or diffuse HCC (21 patients, 3.9%). Median AFP level was 62 ng/mL (range, 1-223169 ng/mL), and 241 (44.2%) patients had an AFP level > 100 ng/mL. At the time of the diagnosis, the number of patients within Milan and expanded criteria were 247 (45%) and 307 (56%), respectively. The distribution of patients according to BCLC classification was as follows: BCLC 0, 14 patients (2.6%); BCLC A, 152 patients (27.9%); BCLC B, 105 patients (19.2%); BCLC C, 115 patients (21.1%); BCLC D, 159 patients (29.2%). Extrahepatic metastasis and macroscopic vascular invasion were diagnosed in 26 (4.8%) and 37 (6.8%) patients, respectively. The only predictor variable for the presence of extrahepatic metastasis at initial diagnosis was TTD (TTD ≥ 5 cm; OR = 3.19, 95%CI: 1.14-10.6, P = 0.029). Stage of liver disease, tumor type, HBV infection, number of nodules, presence of vascular invasion and AFP level did not predict extrahepatic metastasis (Table 3). Univariate logistic regression analyses showed that HBV infection, multinodular and diffuse-infiltrative HCC, TTD, and AFP level were associated with vascular invasion at initial diagnosis of HCC. At multivariate analysis, independent predictor variables of vascular invasion were found to be AFP > 200 ng/mL (OR = 2.95, 95%CI: 1.38-6.31, P = 0.005), TTD > 5 cm (OR = 3.14, 95%CI: 1.01-9.77, P = 0.047) and HBV infection (OR = 5.37, 95%CI: 1.23-23.39, P = 0.025) (Table 4).

Cumulative overall median survival was 16 (13-19) mo in the whole cohort (Figure 1A). The best survival outcome was achieved in patients with HCC who underwent surgical treatments as OLT and hepatic resection (HR = 0.07, 95%CI: 0.04-013, P < 0.001, Figure 1B). Treatment modalities including TACE, Yttrium-90 radioembolization, RFA were also found to be associated with improved overall survival (Table 5). Sorafenib therapy demonstrated a survival benefit, yet with a borderline significance. Univariate Cox regression analyses showed that tumor related factors associated with overall survival of patients with HCC were TTD, tumor-type, vascular invasion, extrahepatic metastasis, and AFP level (Table 5). Patient-related factors including age and gender did not have significant influence on overall survival, yet stage of liver disease significantly predicted overall survival. HBV (HR = 1.28, 95%CI: 0.98-1.66, P = 0.074) and non-viral etiologies (HR = 1.41, 95%CI: 1.002-1.989, P = 0.049) were found to have a borderline influence on patient survival compared to HCV. In multivariate Cox proportional hazard model, stage of liver disease, tumor-related factors including TTD, vascular invasion, extrahepatic metastasis, and AFP level retained significance regarding their influence on overall survival. Treatments including OLT/hepatic resection, RFA, TACE, and Yttrium-90 radioembolization were independently associated with improved overall survival). TACE and Yttrium-90 radioembolization provided a comparable survival benefit (Table 5).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Overall survival in patients with hepatocellular carcinoma. A: Overall median survival in the whole cohort; B: Survival curves was stratified by primary treatment modality.

Among patients with chronic liver disease (532 patients), regular follow-up screening for HCC was performed in 110 (21%) patients. The remaining patients either were not aware of the underlying liver disease or did not have adequate access to health care services due to social, economic or cultural issues. Patients who had regular follow-up and screening with AFP-ultrasonography were diagnosed at an earlier BCLC stage (stage 0-A; 63/110, 57% vs 135/422, 32%, P < 0.001). The number of patients within Milan (69/110, 63% vs 177/422, 42%, P < 0.001) and expanded criteria (85/110, 77% vs 220/422, 52%, P < 0.001) was significantly higher in patients who underwent regular screening follow-up compared to patients who did not. Patients within Milan or expanded criteria, patients who were diagnosed at earlier BCLC stages and patients who had regular follow-up screening for HCC had a significantly better survival (Figure 2; log-rank, P < 0.001).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Overall survival according to oncological stage and pre-diagnostic screening characteristics. Patients within Milan (A) or expanded criteria (B), patients who were diagnosed at earlier BCLC stages (C) and patients who had regular follow-up screening for HCC (D) had a significantly better survival (log-rank, P < 0.001). HCC: Hepatocellular carcinoma; BCLC: Barcelona Clinic Liver Cancer.

HCC is a leading cause of cancer-related death with curative treatment options limited to orthotopic liver transplantation, surgical resection and local ablation. Several prognostic scoring systems were developed to predict the prognosis for patients with HCC, and to individualize treatment by matching best therapeutic option with the patient who is most likely to benefit. Nevertheless, no classification is completely satisfactory because of many other risk factors which also influence patient survival.

Primary treatment modality, which must be among the most important determinants of patient outcome, has not been evaluated as a prognostic indicator in relation to other determinants of survival until now. Therefore, the authors investigated the association between established prognostic factors of HCC and treatment to show how chosen treatment modality affects the prognosis.

Primary objective of the study was to define potential factors that have influence on prognosis, specifically to determine survival benefit associated with primary treatment modality of HCC in a real-life setting. Secondary objective of the study was to find out the relationship between pre-diagnosis screening characteristics, clinical stage of the disease at diagnosis and overall survival.

In the present study, the authors investigated clinical, etiological, and prognostic features in a large, single-center cohort of patients with HCC who were diagnosed, treated and followed-up in the last decade. The diagnosis was established by histopathological and/or radiological criteria that was based on the recommendations reported by the EASL panel of experts in 2001. The authors reviewed demographic, clinical and staging characteristics, laboratory data, etiology of primary liver disease, imaging characteristics and treatments of HCC patients. Number and size of nodules, total tumor diameter (TTD), type of tumor, presence of major vascular involvement and extrahepatic metastasis were determined according to baseline imaging records. Univariate and multivariate Cox regression analyses were performed to find out factors associated with overall survival of patients with HCC.

A total of 545 patients with HCC who were diagnosed and followed-up between January 2001 and August 2011 were included in the study. Predictor variables of vascular invasion and extrahepatic metastasis were investigated by univariate and multivariate logistic regression analyses. The authors showed that HBV infection, multinodular and diffuse-infiltrative HCC, TTD, and AFP level were associated with vascular invasion at initial diagnosis of HCC. At multivariate analysis, independent predictor variables of vascular invasion were found to be AFP > 200 ng/mL, TTD > 5 cm and HBV. The only predictor variable for the presence of extrahepatic metastasis at initial diagnosis was TTD. Stage of liver disease, tumor type, HBV infection, number of nodules, presence of vascular invasion and AFP level did not predict extrahepatic metastasis. The best survival outcome was achieved in patients with HCC who underwent surgical treatments as OLT and hepatic resection. Treatment modalities including TACE, Yttrium-90 radioembolization, RFA were also found to be associated with improved overall survival. Ethanol/acetic acid ablation was not associated with any survival benefit. Systemic treatment with sorafenib was associated with a survival advantage at a borderline significance compared with no treatment, which can be explained by low number of patients receiving systemic therapy. Patients who had regular follow-up and screening with AFP-ultrasonography were diagnosed at an earlier BCLC stage and had a significantly better survival.

It has been known that liver functional reserve, tumor extension and alfa-fetoprotein level are among the most important determinants of patient survival. The authors showed that in addition to patient and tumor related factors, initial choice of treatment is a strong and independent predictor of survival. Survival benefit of non-curative treatments including transarterial chemoembolization and Yttrium-90 radioembolization has been an area of uncertainty. Transarterial chemoembolization and Yttrium-90 radioembolization provided a significant and comparable survival benefit in patients with hepatocellular carcinoma in the real-life setting.

Primary modality of treatment for hepatocellular carcinoma is a major determinant of patient survival that should be incorporated while estimating prognosis in the future trials evaluating benefits of investigational new drugs.