Hepatocellular carcinoma (HCC) is the fifth most common malignancy in the world (1), and its incidence has been on the rise in recent years (2). Notably, HCC is increasingly detected and diagnosed at an early stage of the disease, however, the selection of optimal surgical treatments for patients with early-stage HCC remains controversial, especially for those with liver cirrhosis. Many guidelines recommend that liver resection (LR) and liver transplantation (LT) are considered as potentially curative therapies for patients with early-stage HCC (3, 4). For HCC patients with severe cirrhosis HCC and within Milan criteria, LT is widely accepted as the gold standard as it could eliminate both the tumors and cirrhotic liver which is prone to de novo recurrences of HCC. LT is not a conventional treatment option due to factors such as shortage of donor livers, the high associated cost and disease progression during the waiting period. Therefore, LR is still widely considered as the first option for the treatment of HCC patients without cirrhosis or Child-Pugh class A cirrhosis without portal hypertension (PH) (3, 5). In recent years, local ablations (LA), such as percutaneous microwave coagulation therapy(PMCT) and radiofrequency ablation (RFA), have been recommended as the first-line therapeutic options for patients with early-stage HCC and PMCT seems to show some advantages over RFA regarding efficacy, including better tumor control for perivascular HCCs, and better necrosis rate for cirrhotic liver (6–8). LR may sacrifice additional normal liver tissue which is critical for those with advanced liver cirrhosis, and may lead to severe surgical complications, such as liver failure (9). PMCT is easier to carry out, recovers faster, less invasive, and has a lower rate of liver decompensation in comparison with LR (10, 11). However, the degree of tumor necrosis and the necrosis range of LA are unsatisfactory. Whether LR or PMCT is an optimal option for early-stage HCC with different degrees of liver cirrhosis remains unclear.

Previous studies suggested that the prognosis of HCC depended not only on the treatment strategy and the tumor itself, but also on the underlying liver disease (12), such as chronic hepatitis or liver cirrhosis (1, 4). It has been reported that 60 to 90% of HCC patients in China have underlying liver cirrhosis (13, 14). Unsurprisingly, cirrhosis has been identified as one of the most important risk factors for the development of HCC (15), as well as one of the most important prognostic factors after surgical treatment of HCC (16). The severity of liver cirrhosis was proven to be closely related with the grade of portal hypertension (17), however, liver cirrhosis affected surgical outcomes of HCC independent of portal hypertension (18). Evaluating the status of underlying liver disease only by measuring Child-Pugh liver function and determining the “presence” or “absence” of cirrhosis is obviously unreasonable (14). Our previous studies indicated that the severity of liver cirrhosis significantly influences the short- and long-term surgical outcomes and there existed varied degrees of liver cirrhosis for HCC patients with Child-pugh A liver function and indocyanine green retention test at 15 minutes (ICG-R15) <10% (14, 18–21). Therefore, the severity of liver cirrhosis should be further sub-classified to form a reasonable surgical treatment plan with the aim of decreasing surgical complications and improving surgical outcomes (21).

Liver biopsy is considered as the gold standard for evaluating the severity of cirrhosis (22), but its invasiveness and sampling error preclude its preoperative application in HCC patients. Our previous study developed a clinical scoring system (CSS) as a non-invasive method for sub-classified the severity of liver cirrhosis in HCC patients (as described in Table 1 ) (23). The CSS exhibited high diagnostic accuracy in predicting the severity of cirrhosis, and its accuracy in predicting severe cirrhosis was 85.3%.

The purpose of this study was to compare the surgical efficacy between LR and PMCT for single HCC tumors ≤ 3cm in patients with Child-Pugh A liver function, and further explore the optimal choice of treatment for small and solitary HCC with different degrees of liver cirrhosis which was pre-operatively evaluated according to the CSS.

This study was conducted in accordance with the Helsinki Declaration and approved by the Medical Ethics Committee of Tongji Hospital Affiliated to Huazhong University of Science and Technology. All patients gave written informed consent for post-operative data analysis. A retrospective study was conducted on patients with HCC who received curative treatment in Tongji Hospital, Huazhong University of Science and Technology from January 2008 to December 2014. Patients who met the following criteria were included in the study:

The basic information of the HCC patients and preoperative examination results were obtained from the electronic medical record system, including platelet count, aspartate aminotransferase (AST), prothrombin time (PT), albumin, total bilirubin, alpha fetoprotein (AFP), hepatitis B and hepatitis C serological detection, ultrasonography (US), computed tomography (CT), magnetic resonance imaging (MRI).

All treatments followed the clinical guidelines for the treatment of HCC at the time. The decision for LR or PMCT was made based on the disease status (such as tumor location, surgical risk, feasibility of treatment) and the patients’ preferences. Surgical resection was routinely performed using an open method. PMCT was usually performed under ultrasound guidance.

Nowadays, with the widely use of surveillance programs for HCC in high-risk populations, patients are increasingly diagnosed as early-stage HCC. LR and heat-based local ablations (such as RFA and PMCT) are considered to be the best curative treatments for small HCC in Western countries (10, 24). LR is the most common method in clinical practice, but the resulting tissue trauma is relatively large, and RFA is certainly less invasive. Five randomized controlled trials compared the surgical efficacy of LR and RFA, but they reported conflicting results (25–29). Of them, three randomized controlled trials showed similar survival rates between LR and RFA (25, 27, 29), but in the other two studies, LR was more favorable than RFA in terms of OS and DFS (26, 28). PMCT and RFA have been recommended as first-line therapeutic options for patients with early-stage HCC and limited liver function or severe liver cirrhosis. However, it remains unclear whether MWA is superior to RFA in terms of RFS and OS for the treatment of small and solitary HCC (30, 31). In recent years, PMCT seemed to show some advantages over RFA regarding efficacy, including better tumor control for perivascular HCCs (6), and better necrosis rate for cirrhotic liver (7). Meanwhile, PMCT can overcome the “heat sink effect” which is an obvious disadvantage of RFA (32). A Meta-Analysis of randomized controlled trials indicated a similar efficacy and safety profile between PMCT and RFA, however, distant recurrence rate was significantly lower with PMCT (RR=0.60, 0.39–0.92) in comparison with RFA (8). In China, more than 80% of HCC patients are associated with varied degrees of liver cirrhosis (14). For HCC patients with advanced liver cirrhosis, preserving adequate liver tissue is more important for surgical outcomes, therefore, the authors should consider the effect of liver cirrhosis on tumor necrosis when performed local ablation. A recent study compared the efficacy of RFA and PMCT in achieving complete response in cirrhotic patients with early and very early HCC. The results indicated that PMCT was more efficient than RFA in achieving complete response in HCC nodules with 21 to 35 mm diameter in cirrhotic liver (recurrence rates of RFA and PMCT were 31.9% versus 13.5%, p = 0.019) (7). Potential reasons may be that PMCT induces higher temperature around the tumor in cirrhotic liver tissues (33), and RFA may result in a cold zone easily due to the slow warming of the target area when the ablative region is restricted to perivascular tissue (6). A recent study included a total of 144 eligible patients with small (≤ 3cm) and solitary perivascular (proximity to hepatic and portal veins) HCC who underwent RFA (N=70) or PMCT (N=74) and compared the therapeutic outcomes of these two treatments (6). The results indicated that a significantly higher local tumor progression rate was observed in the RFA group than the PMCT group (5-year local tumor progression rate: RFA 24.3% vs. PMCT 8.4%, p=0.030) (6). As the PMCT treatment showing more superiority, in this study, we compared the efficacy of LR and PMCT in early-stage HCC patients. Many centers have adopted LR as the first-line treatment of early-stage HCC. LR can not only remove the tumor but can also remove the blood vessels and the adjacent liver tissue invaded by the tumor, thus helping prevent recurrence and metastasis in the short term. LR is also the preferred method to the treatment of tumors on the surface or margin of the liver (34). However, LR sacrifices more healthy liver tissue and leads to more surgical complications for those with severe liver cirrhosis, requiring longer hospital stay and thus incurring higher costs (29). PMCT is increasingly accepted for the treatment of early-stage HCC due to its convenient operation, less invasiveness, low postoperative liver decompensation, quick recovery and short hospital stay (35). Previous studies suggested that PMCT had comparable surgical outcomes to LR for early-stage HCC patients, with a lower incidence of complications, cheaper medical cost and shorter hospital stay (10, 33). For HCC lesions with a size <3cm, Shi et al. (36) found that PMCT had similar surgical results to LR. However, some scholars also believe that LR offers significantly better 3-year OS and 5-year DFS than PMCT (35). Moreover, larger liver tumors (including metastases) may be more suitable for LR (34), because PMCT treatment requires multiple overlapping ablations, increasing the difficulty of treatment, so that the resulting ablation margin is not as accurate as resection (36). A recent study compared the therapeutic efficacy between the robot-assisted LR and PMCT. After propensity score matching, 3-year recurrence-free survival, OS and cancer-specific survival of the patients in both groups were comparable. PMCT showed advantages of less operation time, less blood loss and lower medical cost, while LR performed better in postoperative liver function. There was no significant difference in incidence of severe complications between two groups (10). Therefore, distinguishing HCC patients who would benefit most from the two treatments will be a crucial clinical challenge in the future.

In previous studies, worse liver function, liver cirrhosis and portal hypertension were considered as important comorbidities that makes HCC patients favorable candidates for PMCT, because surgical resection may cause more blood loss and more severe complications, such as liver failure (33, 37). A study by Chong et al. (37) evaluated the surgical outcomes of HCC patients treated with LR or PMCT and the role of Albumin-Bilirubin (ALBI) score in selecting surgical modalities. Propensity scoring matching analysis according to the ALBI grade was performed. LR offered better overall and disease-free survivals in patients with ALBI grade 1 whereas PMCT provided a significantly better overall survival (p=0.025) and a trend towards better disease-free survival (p=0.39) in patients with ALBI grade 2 or 3 (37). This study suggested the liver reserve function played a crucial role in the selection of LR or PMCT for early-stage HCC patients. Our previous study indicated that LR may provide better disease-free survival and lower recurrence rates than PMCT for single HCC ≤3 cm and Child-Pugh A cirrhosis, however, PMCT may provide therapeutic effects that are similar to LR for HCC patients with portal hypertension (33). In this study, portal hypertension was defined by an indirect predictor which was less accurate. Furthermore, we did not evaluate the importance of underlying cirrhosis. The severity of liver cirrhosis played an important role in the selection of surgical modality for HCC patients (14). Obvious histological variability in the same stage of liver cirrhosis exists in the majority of HCC patients. Moreover, clinical severity and prognosis within the same histological cirrhosis vary widely, and the risk and surgical outcomes of LR may be different (38). For example, there are still some advanced cirrhosis (F4B-F4C using Laennec scoring system) for compensated cirrhosis or Child-Pugh A cirrhosis; therefore, it should be emphasized these differences to make proper surgical decisions in clinical practice. Unfortunately, it is still unclear whether LR or PMCT is the better treatment option for a single small HCC tumor with different degrees of liver cirrhosis.

In this retrospective study, we found that patients with a single HCC tumor ≤3cm and Child-Pugh class A liver function had longer DFS and OS with LR than with PMCT. At the same time, univariate and multivariate analyses also confirmed that liver cirrhosis was an independent critical risk factor for OS and DFS. In our study, about 83.5% of HCC patients had liver cirrhosis, while previous study had shown that HCC patients with liver cirrhosis had significantly worse long-term outcomes after surgical resection than patients without cirrhosis (39). It is increasingly being recognized that the presence of underlying cirrhosis is closely related to the recurrence of HCC after resection (40). Therefore, the underlying cirrhosis should be considered as a key factor affecting the surgical outcomes (14). Special attention should be paid to assessing the severity of cirrhosis before treatment decisions are made, and there are no recommendations for the treatment of single and small HCC in patients with different degrees of liver cirrhosis. In the clinical settings, the severity of liver cirrhosis may be accurately evaluated pre-operation by liver biopsy, however, the clinical feasibility of this method was severely limited owing to its invasiveness which prevented the collection of histologic information from all the HCC patients at any given time. Thus, there is an urgent need to find a non-invasive method to sub-classify liver cirrhosis that could faithfully evaluates the degrees of liver cirrhosis and that is also essential for the individualization of surgical modalities. By using several simple clinical parameters, we developed a new clinical scoring system to predict the histological severity of cirrhosis with a diagnostic accuracy of more than 80%, especially for those with severe liver cirrhosis (85.3%) (23). The preoperative evaluation of the degrees of liver cirrhosis using this non-invasive method may provide us a strong evidence for the selection of surgical treatments.

Combined with this clinical scoring system, we found that when the score ≥4, the OS and DFS between the two groups was similar. However, the occurrence of surgical complications was more frequent in the LR group than those in the PMCT group (18.8% vs. 4.6%, p<0.001). PMCT is a relatively less invasive procedure that preservers more liver parenchyma, has a lower cost, shorter hospital stay, and better reproducibility, with a lower risk of postoperative complications. For HCC patients with advanced liver cirrhosis, preserving adequate liver tissue is more important for surgical outcomes. LR may result in increased liver function impairment, increased risk of postoperative liver decompensation and death, and therefore higher costs and longer hospital stays due to the complicated surgical procedure and the sacrifice of more non-neoplastic liver parenchyma (25, 41). It should be pointed out that PMCT-treated HCC patients had worse clinical presentations than those treated with LR although there is no statistical difference between the two groups, therefore, PMCT should be considered as the preferred option, and priority can be given to HCC patients who were contraindicated for surgery due to age, comorbidities or background liver damage. According to the results of this study, PMCT is therefore a better treatment option for HCC patients with severe cirrhosis and tumor size ≤ 3 cm, but we should also consider factors such as the location of the tumor, the experience and proficiency of the operator when choosing PMCT.

Finally, several limitations should be considered in this study. Firstly, due to the small total sample size, the statistical power was low and there is a relatively high possibility of selection bias. Secondly, most of the patients in this study had underlying cirrhosis due to hepatitis B, which is clearly not representative of all HCC patients. Thirdly, some PMCT patients had no pathohistological results, so the well-known factors affecting prognosis such as the degree of tumor differentiation and microvascular invasion could not be analyzed. Lastly, this study was a retrospective single-center study, and multi-center clinical trials should be conducted in the later stage to provide stronger evidence and obtain more accurate results.

In conclusion, this retrospective study shows that LR provides better OS and better DFS than PMCT in patients with a single HCC ≤3cm and Child-Pugh class A liver function irrespective of liver cirrhosis. However, for those patients accompanied with severe liver cirrhosis, PMCT may provide a therapeutic effect similar to that of LR and may represent an optimal option, with less invasive, lower medical cost and complications, which should be considered as a superior option. Large number and multicenter randomized controlled trials should be performed to verify our conclusions in the future.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Medical Ethics Committee of Tongji Hospital Affiliated to Huazhong University of Science and Technology. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

E-lZ made the concept and design. JiangL, H-sT, JianL, and W-qW acquired and analyzed the data. JiangL drafted the manuscript. E-lZ critically revised the manuscript for the final version. W-wS and JiangL conducted the statistical analysis. All authors contributed to the article and approved the submitted version.

This work was supported by the grants from National Natural Science Foundation of China (81902839).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.