Abstract
Background/Aim: This study aimed to retrospectively analyse adverse predictors to identify patients with huge hepatocellular carcinoma who were not appropriate candidates for hepatic resection. Patients and Methods: From 551 patients with hepatocellular carcinoma who underwent hepatectomy between 1992 and 2019, 92 were diagnosed with huge hepatocellular carcinoma (diameter >10 cm) and 115 were diagnosed with large hepatocellular carcinoma (diameter=5-10 cm). Clinical features and overall and disease-free survival rates were compared between the two groups. Results: Cumulative overall survival was significantly worse in the huge group than in the large group (p=0.035). In the huge group, multivariate analyses revealed that liver cirrhosis, multiple intrahepatic metastases (≥4), poor histological grade, and macroscopic portal vein invasion were significantly associated with poor prognosis. Conclusion: We identified four adverse predictors of survival and determined that patients with two or more predictors are not appropriate candidates for straightforward hepatic resection.
- Hepatocellular carcinoma
- liver resection
- mortality
- recurrence
- surgery
Hepatocellular carcinoma (HCC) is the fourth leading cause of cancer-related death after lung, colorectal, and stomach cancers (1). The prognosis of patients with HCC has improved with advances in imaging techniques and therapeutic options such as liver resection, liver transplantation, radiofrequency ablation, transhepatic arterial chemotherapy and embolization (TACE), external radiation, and molecularly targeted therapy (2). Clinical guidelines for HCC have been established in many countries and regions, and the appropriate treatment is selected using treatment algorithms based on performance status, liver function, extrahepatic metastasis, vascular invasion, tumour number, and tumour size (3).
However, huge HCC (≥10 cm) remains difficult to treat because it is frequently accompanied by portal vein tumour thrombus and/or multiple tumours (4-6). In our previous report, the overall survival rates for HCC at 1, 3, and 5 years following surgery were 41.0%, 29.3%, and 29.3%, respectively, and the median survival time (MST) was 10.1 months (7). Moreover, resection of huge HCC is often technically demanding and usually requires major hepatic resection, which may increase bleeding, transfusion, operation time, morbidity, and mortality (8). Previous studies have demonstrated that bleeding, transfusion, and a longer operative time led to a poor prognosis (9-11). Therefore, huge HCC is considered to have a poor prognosis not only because of oncological factors but also iatrogenic factors such as an invasive operation.
We occasionally encounter cases involving early relapse and early mortality due to recurrence after liver resection in huge HCC patients. Accordingly, we retrospectively analysed patient characteristics between large and huge HCC groups and determined the prognostic factors to identify patients with huge HCC who were not appropriate candidates for hepatic resection.
Patients and Methods
Patients and preoperative evaluation. Between January 1992 and December 2019, 551 patients with HCC underwent hepatic resection at Yokohama City University, Yokohama, Japan. Of these, 92 patients had HCC with a maximum tumour diameter of ≥10 cm (huge HCC group), and 115 had large HCC (tumour diameter between 5 and 10 cm; large HCC group). The institutional review board of Yokohama City University approved this study (approval no. F210900007), which conforms to the provisions of the Declaration of Helsinki. Informed consent was obtained from the patients and/or guardians.
The preoperative evaluation protocol included blood biochemistry, chest and abdominal radiography, electrocardiography, respiratory function test, indocyanine green (ICG) retention rate at 15 min, ultrasonography, contrast computed tomography from the chest to the pelvis, and/or contrast magnetic resonance imaging of the liver. The criteria for liver resection were determined according to the multiple regression equation: y 110 +0.942 × resection rate (%) + 1.36 × ICG retention rate at 15 min (%) + 1.17 × patient’s age + 5.94 × ICG maximal removal rate (mg/kg/min) (12). In cases where y <50, hepatectomy was performed, whereas in cases where y ≥50, we performed portal vein embolization followed by re-evaluation using these criteria 1 month later.
Portal vein tumour thrombus, hepatic vein thrombus, and inferior vena cava tumour thrombus were not considered absolute contraindications for hepatic resection if the tumour thrombus could be eliminated. Four or more intrahepatic metastases were not considered an absolute contraindication for liver resection. We aggressively performed liver resection for large or huge HCCs with four or more intrahepatic metastases if all tumours could be eliminated. Moreover, in cases wherein all tumours could not be removed, we performed reductive hepatic resection for large or huge main tumours and subsequently performed TACE or hepatic arterial infusion chemotherapy (HAIC) for residual tumours (13). Following surgery, adjuvant HAIC was administered to HCC patients with portal vein invasion or intrahepatic metastases between 2000 and 2013 (14).
Curative resection was defined as complete microscopic removal of the tumour. All patients received a monthly follow-up from the operator for the first 2 years and every 3 months thereafter. During every visit, serum alpha-fetoprotein (AFP) and PIVKA-II levels were monitored. Abdominal computed tomography (CT) was performed every 4 months during the second year and every 6 months thereafter, or when AFP or PIVKA-II levels were elevated. Recurrence was diagnosed based on typical imaging findings on CT scans.
Surgical techniques. Our standard skin incision technique is a median abdominal incision and a right transverse abdominal incision extending from the lower edge of the median abdominal skin incision to the mid-axillary line (the so-called inverted L-shaped incision) (7). When we removed the tumour thrombus from the suprahepatic inferior vena cava, we added a median sternotomy extension of the inverted L-shaped incision, which has been described in detail elsewhere (15). In patients with a particularly large right lobe tumour, we used the anterior approach, wherein parenchymal transection was performed without earlier mobilisation of the liver and extrahepatic control of the right hepatic veins (16). In almost all hepatic resections, Pringle’s manoeuvre or hemivascular occlusion was applied. In all primary cases, intraoperative cholangiography and bile leakage tests were routinely performed. Drains were placed in all cases and removed when the drainage was serous and not stained with bile, usually around postoperative day 2.
Statistical analysis. All variables are expressed as means±standard deviation or medians with ranges, and they were compared using the Mann–Whitney U-test. Categorical variables were compared using the chi-squared test or Fisher’s exact test. Overall survival (OS) and disease-free survival (DFS) rates were calculated using the Kaplan– Meier method, and differences between groups were compared using the log-rank test. Multivariate analysis was conducted using a Cox proportional hazards model with stepwise selection of variables. All statistical analyses were performed using the IBM SPSS Statistics software package (version 26.0 for Windows; IBM, Armonk, NY, USA). Statistical significance was set at a p<0.05.
Results
Comparison of clinicopathological findings between large and huge HCC groups. In total, 115 patients with large HCC and 92 with huge HCC underwent liver resection. We compared the clinicopathological findings between the two groups (Table I). The patients in the huge HCC group were significantly younger than those in the large HCC group (p=0.028). The sex ratio, serum albumin, serum total bilirubin, ICG retention ratio at 15 min, hepatitis B virus (HBV) positive ratio, number of non-B non-C patients, Child–Pugh class, background liver status, tumour differentiation status, rate of macroscopic portal vein invasion, number of ruptured tumours, rate of reductive hepatic resections, rate of adjuvant HAIC, incidence of complications, 30-day mortality rate, and hospital stay were not significantly different between the two groups. The platelet count was significantly higher in the huge group than in the large group (p<0.001). The prothrombin time-international normalised ratio was significantly higher in the huge group than in the large group (p=0.007). Serum AFP and PIVKA-II levels were significantly higher in the huge group compared to the large group (p=0.001, p<0.001). Hepatitis C virus (HCV)-positive patients were significantly fewer in the huge HCC group than in the large HCC group (p=0.023). The number of tumours (p=0.002), rate of major hepatic vein invasion (p<0.001), rate of noncurative resection (p<0.001), operation time and blood loss (p<0.001), and 90-day mortality rate (p=0.037) were significantly higher in the huge HCC group compared to the large group.
Overall survival and DFS according to tumour size. The median follow-up period was 17.3 months (range=0.3-184.5 months) in the large HCC group and 30.6 months (range=2.4-195.1 months) in the huge HCC group. The overall survival rates at 1, 3, and 5 years were 87.2%, 59.3%, and 45.7%, respectively, in the large HCC group and 68.8%, 44.2%, and 33.9%, respectively, in the huge HCC group. MST was 72.4 and 61.3 months in the large and huge HCC groups, respectively. Overall survival was significantly worse in the huge HCC group than in the large HCC group (p=0.035; Figure 1a).
The DFS rates at 1, 3, and 5 years were 55.7%, 31.3%, and 22.3%, respectively, in the large HCC group and 36.9%, 19.5%, and 12.0%, respectively, in the huge HCC group. The DFS rate was significantly worse in the huge HCC group than in the large HCC group (p=0.002; Figure 1b).
Univariate and multivariate Cox regression analysis predicting overall survival. The clinical and histopathological characteristics of the 92 patients in the huge HCC group are summarised in Table II. Univariate and multivariate analyses revealed that liver cirrhosis, multiple liver tumours (≥4), poor histological grade, and main portal vein invasion were independent prognostic factors for overall survival.
Overall survival for patients according to the presence of four prognostic factors. Figure 2 shows the cumulative survival rates according to the presence of four prognostic factors. The cumulative overall survival rates at 1, 3, and 5 years were 92.4%, 73.5%, and 58.4%, respectively, for patients without all four prognostic factors, with MST of 71.7 months. For patients with one prognostic factor, these rates were 69.0%, 32.1%, and 22.0%, respectively, with MST of 41.0 months. For patients with two or more prognostic factors, these rates were 11.8%, 0%, and 0%, respectively, with MST of 8.2 months. Thus, the cumulative overall survival was significantly worse for HCC patients with two or more prognostic factors than for those without any prognostic factor (p<0.001) or one prognostic factor (p=0.001).
Discussion
In this study, we retrospectively compared the clinical features, OS rates, and DFS rates between the huge (≥10 cm) and large (5-10 cm) HCC groups. Patients with multiple tumours (≥4), major hepatic vein invasion, and noncurative resection were significantly more frequent in the large HCC group. AFP and PIVKA-II levels were significantly higher in the huge HCC group than in the large HCC group, whereas OS and DFS were significantly worse in the huge HCC group than in the large group. These results agree with those of previous reports (7, 17, 18). In contrast, some studies have reported no significant differences in OS or DFS between patients with large HCC and those with huge HCC because of improvements in surgical techniques and multidisciplinary treatment (11, 19). Furthermore, Alleman et al. have reported that median survival and 5-year survival were better for patients with huge HCC than for those with large HCC, possibly because of less aggressive tumour biology (20). However, a recent meta-analysis reported that 5-year survival was worse for HCC patients than for non-HCC patients, and the authors confirmed the benefits and feasibility of hepatectomy for HCC (21). Although these results remain controversial, select patients with huge HCC may be candidates for hepatectomy, similar to patients with large HCC.
Of late, liver resection for huge HCC has been reported to be safe. The estimated blood loss reportedly ranges from 480 to 4354.1 ml, with morbidity rates of 10.9%-72.7% and surgery-related mortality rates of 0%-11.1% (7, 9, 11, 16-18, 22-27). In a recent meta-analysis, the rates of treatment-related complications and mortality were reported to be 42.7% and 4.2%, respectively (21). Wang et al. reported no significant differences in the rates of treatment-related complications and mortality between huge HCC and non-huge HCC groups (21). Furthermore, Zhu et al. reported that severe complications were not more frequent after liver resection than after local regional therapy such as TACE for huge HCC (28). In our series, the estimated blood loss was 2440.0 mL, the morbidity rate was 20.7%, and the mortality rate was 1.1%. With regard to the 30-day mortality, one patient died on postoperative day 10 because of post-hepatectomy liver failure following massive postoperative bleeding from a hepatic artery aneurysm. Although the estimated blood loss was significantly higher and the operation time was significantly longer in the huge HCC group than in the large HCC group, there was no significant difference in morbidity and 30-day mortality between the huge and large HCC groups. Accordingly, hepatic resection for patients with huge HCC may be a safe treatment modality with low mortality and morbidity rates, similar to those for patients with large HCC.
Surgical indications for HCC with four or more intrahepatic metastases remain controversial, and we aggressively performed liver resection for large or huge HCCs with four or more intrahepatic metastases if all tumours could be eliminated. Moreover, we have performed reductive hepatic resection for large or huge main tumours followed by HAIC or TACE for residual tumours in cases where all tumours could not be eliminated by liver resection (13). According to the clinical practice guidelines for HCC in Japan (29), TACE, HAIC, or molecularly targeted therapy is recommended. However, large HCCs are frequently accompanied by multiple tumours. In our series, 31.5% of the huge HCC patients had ≥4 intrahepatic metastases. Generally, these cases are contraindicated for surgery because of poor long-term outcomes. However, several authors have performed reductive hepatic resection in combination with multidisciplinary treatment for huge HCC with multiple intrahepatic metastases (30-34). Inoue et al. (31) have reported the usefulness of reduction surgery followed by repeated TACE for patients with huge HCC and multiple intrahepatic metastases when the patients had good liver function, and the total volume of metastatic nodules was <10% of the remnant liver parenchyma. Moreover, we revealed the usefulness of reduction surgery followed by HAIC (13). In our previous report, the overall survival rates at 1, 3, and 5 years after reduction surgery were 73.2%, 38.7%, and 38.7%, respectively, although that study included non-huge HCC. In the present study, reductive hepatic resection was not a negative predictive factor for OS in multivariate analysis for large HCCs. These results may be attributed to the multidisciplinary treatment of residual tumours. Therefore, reduction surgery followed by multidisciplinary treatment such as TACE and HAIC may offer the possibility of long-term survival for select patients with huge HCC accompanied by multiple intrahepatic metastases.
There is currently no effective neoadjuvant or adjuvant therapy for HCC; however, some authors have recommended perioperative therapy for huge HCCs (35, 36). Hwang et al. (36) strongly recommended transarterial chemoinfusion or TACE for huge HCC with satellite nodules at 1 month after surgery. We performed adjuvant HAIC for HCC patients with portal vein invasion or intrahepatic metastases after surgery from 2000 to 2013 (14). Therefore, in the present study, 17.4% of the large HCC patients received adjuvant HAIC. Our previous study (14) has suggested that adjuvant HAIC reduces the number of recurrent nodules in the remnant liver; therefore, it may also be useful for curatively resected huge HCC with intrahepatic metastases.
In the huge HCC group, we experienced early mortality due to recurrence of HCC after surgery, and such cases are not appropriate candidates for hepatic resection. The 90-day mortality rate for huge HCC was 7.6%, which was significantly higher than that for large HCC (1.7%, p=0.037). In the huge HCC group, we experienced seven cases of 90-day mortality, including one case of 30-day mortality. Six of the seven patients died of HCC, and five of these six patients were discharged after surgery and could visit the hospital as outpatients, whereas one could not be discharged because of liver dysfunction due to intrahepatic multiple recurrences of HCC. Two of the six patients underwent noncurative resection, and small intrahepatic metastases remained in the remnant liver. The other four patients showed early multiple recurrences in the remnant liver. In all six cases, the tumours showed rapid growth after surgery, and we could not perform any treatment because of poor liver function. Therefore, surgery should be avoided in patients with rapidly growing tumours or patients whose liver function is not expected to improve sufficiently after surgery.
Liver cirrhosis, multiple intrahepatic metastases (≥4), poor histological grade, and macroscopic portal vein invasion were found to be prognostic factors for OS in this study, and it is suggested that patients with two or more of these prognostic factors are not appropriate candidates for hepatic resection. Major portal vein invasion, multiple intrahepatic metastases (≥4), and poor histological grade have previously been recognised as poor prognostic factors after liver resection for patients with huge HCC (9, 17, 36-38). Some papers have described the usefulness of prognostic nomograms for postoperative recurrence and mortality for patients with huge HCC and reported that vascular invasion and histological grade were prognostic factors of huge HCC (39, 40). Furthermore, cirrhosis has been reported as a poor prognostic factor for patients with huge HCC (17, 41); in patients with cirrhosis, liver regeneration after hepatectomy is significantly slower and less complete (42). Therefore, treatment may be delayed or impossible when early recurrence occurs after liver resection in patients with cirrhosis. In the present study, the overall survival rates at 1, 3, and 5 years for patients without any risk factor were 96.4%, 74.3%, and 69.0%, respectively, with MST of 71.7 months. For patients with one risk factor, these rates were 79.9%, 46.1%, and 29.6%, with MST of 41.0 months. These results are acceptable when compared with results from other institutes (9, 10, 18,19, 21-23, 27, 36, 37, 43). However, the overall survival rates at 1, 3, and 5 years for patients with two or more risk factors were 11.8%, 0%, and 0%, respectively, with MST of 8.2 months. These results suggest that patients with two or more adverse predictors have rapid tumour growth or insufficient improvement in liver function after surgery; therefore, they are not appropriate candidates for hepatic resection.
TACE, HAIC, and molecularly targeted therapy have been reported as alternative treatment options for unresectable HCC. The overall survival rates at 1, 3, and 5 years for patients with huge HCC with multiple intrahepatic metastases (≥4) were 33.3%, 11.1%, and 0%, respectively (44), whereas those for patients with huge unresectable HCC at 1 and 2 years were 29% and 14%, respectively (45). These results are not satisfactory but may be superior to the results for surgical resection in patients with two or more prognostic factors. Recently, the usefulness of atezolizumab combined with bevacizumab for patients with unresectable HCC has been reported (46). The trial included patients with macrovascular invasion of the main portal trunk or the portal vein branch contralateral to the primarily involved lobe, bile duct invasion, or at least 50% hepatic involvement (or any combination of these three features) (46). Therefore, atezolizumab combined with bevacizumab may be effective in patients with two or more prognostic factors. If these treatments eliminate the poor prognostic factors and allow hepatic resection, even huge HCCs with two or more prognostic factors may be good candidates for surgical resection. Taking all points into consideration, huge HCCs with two or three prognostic factors are considered borderline resectable tumours. These patients may be candidates for neoadjuvant treatment before surgical resection.
This study had several limitations. First, as with any retrospective study, selection bias was inevitable. Second, the number of patients was small, and all patients were treated at a single institution. Third, the study did not evaluate some important preoperative variables; therefore, multicentre studies should be conducted. Finally, because of the long period of patient recruitment, the treatment for recurrence and adjuvant therapy may have differed depending on the period, and the types of treatment available may have affected the prognosis.
In conclusion, hepatic resection can be performed safely and effectively for select HCCs measuring ≥10 cm in diameter. We identified four adverse predictors of survival: liver cirrhosis, multiple intrahepatic metastases (≥4), poor histological grade, and major portal vein invasion. Our findings suggest that patients with huge HCC accompanied by two or more of these adverse predictors are not appropriate candidates for straightforward hepatic resection.
Acknowledgements
We are grateful to Dr. Gakuryu Nakayama and Dr. Keiichi Yazawa for their assistance with this study.
Footnotes
Authors’ Contributions
TK drafted this manuscript. TK, KT, and YS were the main surgeons. RM, KS, MD, S-CL, YY, YH, and IE assisted during the surgery. KS, S-CL, and IE reviewed and modified the manuscript. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors have no conflicts of interest to declare.
- Received December 7, 2021.
- Revision received March 11, 2022.
- Accepted March 14, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).