Comparison of short- and long-term outcomes of patients with hepatocellular carcinoma undergoing surgical resection, local ablation, angiological treatment or palliation: Single center results over 35 years

Hepatocellular carcinoma (HCC) is one leading cause of cancer mortality often presenting at inoperable stage. The aim of this study was to examine and compare surgically resected, locally ablated, angiologically treated and palliatively treated HCC patients’ short- and long-term outcomes in a single center over 35 year period. All HCC diagnosed in Oulu University Hospital between 1983-2018 were identified from hospital records (n=273). Patients underwent hepatic resection (n=49), local ablation (RF, laser ablation or PEI; n=25), angiological treatments (TACE, TAE and SIRT; n=48) or palliative treatment (chemotherapy, best supportive care; n=151). Primary outcomes of the study were postoperative complications within 30 days after the operation, and short-(30- and 90-day) and long-term (1, 3 and 5-year) survival. Results were adjusted with sex, age, comorbidities, cirrhosis, Child-Pugh index points, ASA status, year of operation and stage.


Abstract
Background Hepatocellular carcinoma (HCC) is one leading cause of cancer mortality often presenting at inoperable stage. The aim of this study was to examine and compare surgically resected, locally ablated, angiologically treated and palliatively treated HCC patients' short-and long-term outcomes in a single center over 35 year period.

Methods
All HCC diagnosed in Oulu University Hospital between 1983-2018 were identified from hospital records (n=273). Patients underwent hepatic resection (n=49), local ablation (RF, laser ablation or PEI; n=25), angiological treatments (TACE, TAE and SIRT; n=48) or palliative treatment (chemotherapy, best supportive care; n=151). Primary outcomes of the study were postoperative complications within 30 days after the operation, and short-(30-and 90-day) and long-term (1, 3 and 5-year) survival. Results were adjusted with sex, age, comorbidities, cirrhosis, Child-Pugh index points, ASA status, year of operation and stage.

Results
Surgically resected patients were younger than patients in other groups. Recurrence and local recidives occurred more often in local ablation group and in angiological treatment group (p<0.001). Surgical resection rate was 17.9%. Overall complication rates in surgical resection, local ablation and angiological group were 71.5%, 32.0% and 58.3%, (p<0.001).

Conclusions
Based on our study on Northern Finland population, the surgical resection of HCC seems to be the most effective treatment considering long-term survival and tumor recurrence after adjustment for confounding factors.
Background Hepatocellular carcinoma (HCC) is the sixth most common cancer worldwide and it is the fourth most common cause of cancer mortality 1 Globally, hepatitis B and C viral infections are most common underlying causes of HCC, especially in eastern countries 2 . In western countries, heavy alcohol consumption is a major cause of liver disease, which can lead to cirrhosis and HCC 3 4 . Surgical resection is the first-line therapy for single HCC of any size, when hepatic function is preserved, and sufficient remnant liver volume is maintained. 5 Only 20-30% of the HCC are resectable. 6 Liver transplantation is considered the first-line therapy for HCC within Milan Criteria unsuitable for resection 5 , but the availability of transplantation is limited. 6 Radiofrequency ablation (RF) in single tumors 2 to 3 cm is an alternative to surgical resection based on technical factors (location of the tumor), hepatic and extrahepatic conditions. 5 Percutaneous ethanol injection (PEI) is an option in some cases where RF is not technically feasible due to localization of the tumor. 5 Other options for non-resectable HCC are transarterial chemoembolization (TACE) and molecular targeted therapies 5 .
In resections, < 30% morbidity and < 3% mortality rates have been reported 5 . Morbidity and mortality rates are higher in patients with cirrhotic liver. 7 In RF, complication rates vary from 0 to 6.1% 8 and perioperative mortality rates ranges from 0 to 1.8%. 9 In TACE, complication rates vary from 25 to 45% 10 and overall mortality rates are around 0.6% 11 .
The aim of this study was to examine and compare results of surgical resection, local ablation (RF and PEI), angiological treatment (TACE, transarterial embolization (TAE), selective internal radiation therapy (SIRT)) and palliative treatment in HCC regarding complication rates, and short-and long-term outcomes. The study population consisted of Northern Finland population, where alcohol plays a major role behind the etiology of HCC.
All patients were treated in Oulu University Hospital.

Study design
This study was a retrospective cohort study in a single institution tertiary care hospital in Northern Finland. The study population consists of 273 patients with hepatocellular carcinoma diagnosed in Oulu University Hospital between January 1983 and March 12,

Data collection
The patients were identified from archives using ICD-10 code C22.0& (hepatocellular carcinoma). Inclusion criteria was histological verification of HCC. All diagnoses were confirmed with histological examination either from surgical resecate or biopsy sample.
The clinical data was collected from Oulu University Hospital patient records. Of 273 patients, 49 underwent surgical resection, 25 RF, laser ablation or PEI, 48 were treated with TACE, TAE or SIRT and 151 were treated with palliative treatment or best supportive care. Some patients received more than one treatment presented in Figure 1. Twenty-five patients were excluded from the study due to lack of information from the patient files or because of the indefinite diagnosis. Four groups were formed: 1) Surgical resection, 2) Local ablation (RF or PEI), 3) Angiological group (TACE, TAE, SIRT) and 4) Palliative treatment (chemotherapy or best supportive care). Study groups were formed according to the most radical treatment, for example if HCC was surgically resected and received also RF, patient was included in surgical resection group. The complications were classified primarily with Accordion Severity Grading System 12 and secondarily with Clavien-Dindo classification system 13 .

Outcomes
Primary outcomes of the study were postoperative complications within 30 days after the operation, and short-(30-and 90-day) and long-term (1, 3 and 5-year) survival.

Statistical analysis
Mann-Whitney U-test was used to compare differences between two independent groups with continuous variable. For categorical data-analysis we used χ2-test and Fisher-tests.
The threshold for significance was set at P < 0.05. In all continuous variables, median and interquartile range is presented. For survival data, Kaplan-Meier with log-rank test was used. Cox-regression analysis was used to analyze survival in three treatment groups adjusting with the following covariates: sex, age, comorbidities, cirrhosis Child-Pugh index, ASA status, year of operation and stage. Complications were classified as minor and major based on Clavien-Dindo classification system 13 and Accordion Severity Grading

Preoperative features of the study groups
Baseline characteristics of the study groups are presented in Table 1 Comorbidity Index in the four groups was 1, 3, 2 and 1, respectively. Liver cirrhosis was present in 33%, 56%, 40% and 33%. According to Child Pugh Classification, class B/C was present in surgical resection group in 4.1%, in local ablation group 20%, in other invasive treatment group 4.2% and in palliative group 29.8%. Most common ASA status in all four groups was grade III, including 59% in the resection group, 80% in local ablation group, 65% in angiological group and 67% in palliative group.

Tumor features
Tumor stage I was the most common in the first three groups, including 79.6% in the resection group, 76.0% in local ablation group and 47.9% in angiological group. In palliative group 58.3% had tumor stage III or IV. Tumor size was bigger in resection group than in local ablation group (median 50 mm vs 30 mm, p<0.001). No difference in size was observed between resection and angiological group. For other parameters and between group comparisons, see Table 2.
In local ablation group the most common treatment was RF with 17 (68.0%) patients.
Eight (16.7%) patients had readmission in 30 days. Postoperative features are presented in Table 2.

Postoperative complications
Overall complications occurred more frequently in resection group than in local ablation group (71.5% vs 32.0%, p<0.001). There was no significant difference between resection group and angiological group in number of complications (71.5% vs 58.3%, p=0.116).

Short-and long-term outcomes
In resection group 27 (55.1%) patients had tumor recurrence during follow-up, of which 2 (4.1%) patients had local recidive. In local ablation group, 15 (60.0%) patients were diagnosed with tumor recurrence, of which local recidive occurred with 8 (32.0%) patients, of which five patients were treated with RF and three with PEI. In angiological group, 38 (79.2%) patients were diagnosed with tumor recurrence, of which local recidive occurred with 25 (52.1%) patients (Table 2).

Survival trends over time
To analyse survival trends over time, we divided groups (resection, local ablation, angiological, palliative) into further two equal sized cohorts based on year of operation.
Cut-off years were 2000 for resection, 2012 for local ablation, 2011 for angiological and 2011 for palliative group. We observed no statistically significant differences over time inside any of the groups separately. If treatment groups were combined, disease-specific survival in old cohort at 1-, 3-and 5-years were 50.8%, 27.6% and 16.5%. Respective survival rates in new cohort were 58.0%, 40.6% and 37.2% (p=0.035 between groups at 5years).

Discussion And Conclusions
We observed preferable long-term survival after surgical treatment of HCC when compared to other treatment modalities also after adjusting for confounding factors. Improvement in survival was observed over time. Based on resection rate and baseline characteristics, more patients might be treated with surgery in Northern Finland. where the number of patients is sufficient. Laparotomy was the standard surgical approach in our center. In guidelines hepatic resection is recommended to be performed via laparoscopic/minimally invasive approaches when possible. 5 Approach can cause confounding when comparing complication profiles to recent reports.
In surgically resected group, overall survival rates at 1-, 3 and 5-years were 85.1%, 59.0% and 51.2%. In previous studies, better survival rates has been reported in small (< 5cm) HCC. [14][15][16][17][18][19] . In these studies, the overall survival rates varied at 1-,3 and 5-years from 91.3% to 100.0%, from 73.4% to 92.2% and from 61.5% to 75.7%. [14][15][16][17][18][19] Lower overall survival rates have been reported in patients with cirrhosis varying from 41.0 to 79.0% at 5-years. 7,20,21 It is notable that in our study, the median tumor size in surgically resected group was 5.0 cm (IQR 3.5-10.0), with cirrhotic liver in one third of patients. Hepatic recurrence rates after surgical resection from 16.7% to 78.8% have been reported. 15 In surgically treated group, we reported 14 (28.6%) major complications and 21 (42.9%) minor complications, and two (4.1%) postoperative deaths. Liver resection remains a complex surgical procedure with reported major complication rates from 27.8% to 55.5% and mortality rated from 0.0% to 11.0%. 7,16,17,38 Cirrhosis and weak liver function associate to high mortality rates. 7 Perioperative mortality in cirrhotic patients should be less than 3% 5 and major morbidity less than 30% 522 . The complication profile in referenced studies in major complications was mainly similar to our study, with surgical site infection being the most common. In our study less ascites-, bile leakage -and pleural effusion-related complications were observed. 7,16,17 Complications in our study were more common after surgical resection when compared to RF, but no difference was observed between resection and angiological treatment, advocating the use of surgical treatment.
In literature, major complications in RF treated patients have been reported from 0.9% to 4.3%. 17,24-28,32,39,40 and mortality rates from 0.0% to 1.6%. 25,26,32,39,40 Reported complications rates following TACE are high (25-45%), with the majority being reversible elevations of hepatic transaminases and serum bilirubin. 10 . In our study, overall complications occurred in 28 (58.4%) patients. We did not observe any tumor needle seeding complications, which was reported in several studies. 24-26,28,32 The most common complications after TACE in our study was pain problems, organ site infection and sepsis, which were also detected in several referenced studies. 10 The length of hospital stay after surgical resection and other less invasive treatments was similar compared to previous reports. 7,16,17,38,41 We observed a significant rise in other treatment modalities than surgery, which can be due to multiple factors, for example the development of new therapies, histological and radiological examination and patient evaluation. In Finland, alcohol plays a critical role in etiology of cirrhosis and HCC, which is a known risk factor of surgery. 7

Ethics approval and consent to participate
The patient survival data was acquired from Statistics Finland. The use of patient data was approved by the Oulu University Hospital Ethics Committee and by the National Authority for Medicolegal Affairs (VALVIRA).

Consent for publication
The manuscript is approved for publication by all the authors.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. AFP, median (IQR) e 6.0 (3.0-191.0) 6.0 (3.0-9.0) 9.0 (3.5-261.0) a= Significant difference between resection group and local ablation group b= Significant difference between resection group and angiological group c= Significant difference between resection group and palliative group d= Significant difference between local ablation group and angiological group e= Significant difference between local ablation group and palliative group f= Significant difference between angiological group and palliative group Significant difference = P<0.050 Emergency patient 0 (0.0%) 0 (0.0%) 3 (6.3%) a= Significant difference between resection group and local ablation group b= Significant difference between resection group and angiological group c= Significant difference between resection group and palliative group d= Significant difference between local ablation group and angiological group e= Significant difference between local ablation group and palliative group f= Significant difference between angiological group and palliative group Significant difference = P<0.050 Table 3. Accordion Severity Grading System -based postoperative complications in patients with hepatocellular carcinoma after surgical resection, local ablation including RF, laser ablation and PEI and angiological group including TACE, TAE and SIRT. The numbers of resection, local ablation group and angiological group are presented in the same table with local ablation group in parentheses and angiological group in square brackets.
Pneu moni a 6 (1) [2] 12.2 (4.0) Dee p veno us thro mbo sis   Figure 1 Flow-chart presenting the given treatment in the four study groups of patients with hepatocellular carcinoma.

Figure 2
Trends in treatment modalities of hepatocellular carcinoma.

Figure 3
Disease-specific survival of hepatocellular carcinoma stratified by treatment modality.

Figure 4
Overall survival of hepatocellular carcinoma stratified by treatment modality.