Systematic Review of Single Large and/or Multinodular Hepatocellular Carcinoma: Surgical Resection Improves Survival

Background: The role of surgical resection for patients with single large (≥ 5cm) and/or multinodular (≥ 2) hepatocellular carcinoma (HCC) is still controversial. This systematic review was performed to evaluate the safety and efficacy of resection for patients with single large and/or multinodular HCC. Materials and Methods: Databases (the PubMed, Web of Science, Embase, and Cochrane databases) were systematically searched to identify relevant studies exploring the safety and efficacy of resection for single large and/or multinodular HCC, published between January 2000 and December 2014. Perioperative morbidity and mortality, overall survival, and disease-free survival of the resection group were calculated. In addition, these outcome variables were also calculated for the control group in the included studies. Results: One randomized controlled trial and 42 non-randomized studies involving 9,580 patients were eligible for analysis. Eight (1,594 patients) of the 43 studies also reported the outcomes of transarterial chemoembolization (TACE). Although 51.4% of patients featured cirrhosis, 90.7% of them demonstrated Child-Pugh A liver function in the resection group. The median rates of morbidity (24.5%) and mortality (2.5%) after resection were significantly higher than that of TACE (11.0%, P <0.001; 1.9%, P <0.001). However, patients who underwent resection had significantly higher median one, three-, and five-year overall survival (76.1%, 51.7%, and 37.4%) than those who underwent TACE (68.3%, 31.5%, and 17.5%, all P <0.001). The median 1, 3-, and 5-year DFS rates after resection were 58.3%, 34.6%, and 24.0%, respectively. Conclusions: Although tumor recurrence after resection for patients with single large and/ or multinodular HCC continues to be a major problem, resection should be considered as a strategy to achieve long-term survival.


Introduction
Hepatocellular carcinoma (HCC) is the sixth most common cancer among malignant tumors and the third most common cause of cancer-related deaths (Siegel et al., 2014).Its incidence is increasing in many countries.Recent advances in diagnostic methods and the widespread application of screening programs in high-risk populations have facilitated the detection of early-stage HCC.However, a substantial proportion of patients still present with single large (≥ 5 cm) or multinodular tumors (≥ 2).Surgical resection is considered as the first-line treatment for early-stage HCC, with five-year overall survival (OS) up to 67% (Lim et al., 2012).The prognosis of patients with single large and/or multinodular HCC is generally poor.Instead of curative treatment strategies, locoregional therapies, particularly transarterial chemoembolization (TACE) have been used primarily in selected patients
According to these treatment guidelines (Bruix et al., 2014;EASL-EORTC. 2012), the role of resection is mainly fit for patients with early-stage HCC and those who have preserved liver function.The restrictive treatment criteria for resection may result in some candidates who are suitable for resection receiving TACE.Although TACE is suggested to be the first-line treatment for patients with single large and/or multinodular HCC, some retrospective studies with large sample size in different countries have already indicated that resection could offer low mortality and favorable survival benefits (Chen et al., 2006;Hsu et al., 2012;Torzilli et al., 2013;Zhong et al., 2014a).These results revealed that resection should be considered as an optional treatment for single large and/or multinodular HCC.Moreover, the only randomized controlled trial in this field also demonstrated that resection is superior to TACE for patients with HCC beyond Milan criteria (Yin et al., 2014), Therefore, the question of whether resection is acceptable and applicable for patients with single large and/or multinodular HCC remains controversial.
Here we perform a systematic review for the evidence published from 2000 to 2014 on outcomes of resection in patients with single large and/or multinodular HCC and preserved liver function.In addition, in order to compare the efficacy of resection with the standard therapy of TACE for patients with single large and/or multinodular HCC, we also reviewed the efficacy of TACE among the included studies.

Search strategy
Literatures of the PubMed, Web of Science, Embase, and Cochrane databases were systematically searched using the following search terms to identify studies in December 2014: "hepatocellular carcinoma" or HCC or "liver cancer" or "liver neoplasm" or "liver tumor" or "primary liver carcinoma", single large or giant or huge or multinodular, "surgical resection" or "hepatic resection" or hepatectomy or "curative resection" or "liver resection" or surgery.The Cochrane systematic review methodology was used for this review.

Eligibility criteria
Original studies were included if they reported the outcomes of initial resection in patients with primary single large and/or multinodular HCC with preserved liver function.Only studies published in English between January 2000 and December 2014 may be eligible included.Patients who underwent explicit non-curative or palliative treatments were excluded.Namely, only patients with primary HCC after potential curative resection were included into analysis.For duplicated publications, only those with largest sample size or longer follow-up were included.For reducing the selective bias and measure bias, only studies with a sample size of more than 50 were included.The references of retrieved articles were manually searched to locate other potential relevant studies.The outcomes included perioperative mortality and morbidity, OS, or disease-free survival (DFS).

Data extraction
Two investigators (XDY and JHZ) independently and critically appraised the eligibility of relevant studies based on the inclusion criteria through scrutinizing the title and abstract of each record.Discrepancies were resolved by a third investigator (JC).Three authors (XDY, JHZ, JC) extracted the following data from included studies: patient demographics, disease characteristics, perioperative morbidity and mortality, OS, DFS, and tumor recurrence.Due to the clinical heterogeneity among different studies and the lack of control group, we did not perform metaanalysis.

Data analysis
Continuous variables were calculated using SPSS software (version 16.0, SPSS Inc., Chicago, IL) and expressed as median (range).All the analysis was carried out by Excel 2013.

Patients' characteristics and disease characteristics
Characteristics of the included patients and details of the disease characteristics were shown in Table 1.Four studies (Wang et al., 2008;Ho et al., 2009;Nojiri et al., 2014;Yin et al., 2014) mainly described patients with multiple tumors.And the rest studies presented patients with single large with or without multinodular tumors.In total, 9580 patients from 43 different papers were reviewed.The sample size ranged from 50 to 1143.The majority were males from 59.0 to 93.0 (median 81.2) percent.The median rate of patients with hepatitis B virus infection and hepatitis C virus infection were 68.3% (range,10.0-93.0%)and 17.9% (range, 1.0-74.0%),respectively.The median rate of cirrhosis was 51.4% (range, 0-89.0%).However, most patients were with Child-Pugh A liver function (median, 90.7%; range, 58.0-100%).

Discussion
Most western hepatobiliary surgeons do not recommended resection for patients with single large and/or multinodular HCC, even if those with Child-Pugh A liver function.Instead, palliative treatment with TACE is recommended as the first-line therapy for these patients.High rate of hospital mortality and low rate of DFS are two main reasons which limited the extensive usage of resection.
With the improvement of perioperative care and surgical technique, zero hospital mortality rates can be achieved in some big liver centers (Imamura et al., 2003, Jarnagin et al., 2002).Our results demonstrated that the median hospital mortality after resection was 2.5% in patients with single large and/or multinodular HCC.The result of mortality was lower than that (4.0%) reported in a meta-analysis of 69 studies in which patients in various stages of HCC were treated with resection (Ramacciato et al., 2012).The second attention is the low rate of DFS after resection.In patients with early stage HCC, the fiveyear median DFS is 37% (Lim et al., 2012).However, the five-year median DFS (24.0%) in our study is comparable to that (range, 15.0-35.0%)reported in a systematic review of 22 studies in which patients with huge (> 10 cm) HCC were treated with resection (Tsoulfas et al., 2012).In our study, patients with single large and/or multinodular HCC who underwent resection have significantly higher five-year median OS than those who underwent TACE (P<0.001).Though TACE is considered as standard treatment modality for single large and/or multinodular HCC in western official guidelines, TACE-related mortality also should not be ignored, with a median rate of 2.4% reported by meta-analysis (Marelli et al., 2007).Moreover, the five-year OS is less than 17% in patients with huge (> 10 cm) HCC after TACE therapy (Poon et al., 2000;Huang et al., 2006).
Large-volume units with fully equipped and experienced in the management of complicated HCC demonstrated that tumor size does not influence patients' survival, although more complex surgical techniques are required for giant tumors (Young et al., 2007;Yang et al., 2009;Zhong et al., 2013;Zhang et al., 2014;Zhong et al., 2014b;Zhong et al., 2015).Resection of large HCC without macrovascular invasion has been achieved comparably favorable outcomes with small tumors, which could not be achieved through other palliative treatments (Yeh et al., 2003;Ariizumi et al., 2013).Eight of the included retrospective studies also revealed that resection provided superior outcomes than TACE in patients with single large and/or multinodular HCC (Lin et al., 2010;Luo et al., 2011;Hsu et al., 2012;Jin et al., 2014;Lei et al., 2014;Liu et al., 2014;Yin et al., 2014;Zhong et al., 2014a).Moreover, the recent randomized controlled trial which involving 173 patients with resectable multiple HCC beyong Milan criteria also revealed that resection was associated with significantly better OS than TACE (P<0.001) (Yin et al., 2014).Therefore, tumor size and tumor number should not be considered as contraindication to resection.
Although postoperative tumor recurrence remains high, our findings showed that resection is reasonable and associated with survival benefits.First, with the improvement of skillful surgical techniques and perioperative care, the rates of perioperative morbidity and mortality are acceptable.With regard to the tumor recurrence after resection, repeated resection or radiofrequency ablation may be available for some patients.In addition, postoperative effective adjuvant treatment options for reducing risk of recurrence also improve patients' survival (Zhong et al., 2012, Zhong et al., 2014c, Zuo et al., 2015).However, the durable long-term survival outcomes of a treatment strategy combining these modalities have not been fully assessed, and ongoing trials targeting the population to benefit from the multidisciplinary management are still awaited.
The majority of the published experiences concerning the resection treatment in patients with single large and/or multinodular HCC are reported by Asian countries, where a great number of patients are with chronic hepatitis B liver disease.However, in western countries, more patients are with hepatitis C virus infection and cirrhosis, which are more likely to develop tumor recurrence because of strong hepatic inflammatory activity and progressive liver disease (Fong ZV and Tanabe KK. 2014).This may help to explain why the Barcelona Clinic Liver Cancer classification system has been approved as guidance for HCC treatment algorithms by the European Association for the Study of Liver and the American Association for the Study of Liver Disease, but still be debated by the main Asian associations for the study of liver diseases (Kudo et al., 2011).Therefore, external validation is keenly needed from different study groups.
There are some limitations in this review which must be considered.First, substantial clinical heterogeneity among the included patients owing to the regional differences may limit the ability of the results to be expanded and extrapolated.Second, the favorable results of resection might partly be due to a high selection of patients with a well preserved liver function and limited intrahepatic tumor spread.Thirdly, width recruitment periods of this study may also produce bias because of the surgical technique improvement in recent decade.Randomized trials with large sample size are therefore required to further examine the role of resection in these patients.
In conclusion, available studies showed that surgical resection provides long-term survival in patients with single large and/or multinodular HCC and preserved liver function.However, the strategy of resection should be carried out with caution in selective patients because of the unignored perioperative morbidity and mortality.

Figure 1 .
Figure 1.PRISMA Flow Diagram for Selection of Articles Included; HCC, Hepatocellular Carcinoma

Table 3 . Comparison of OutcomesRegarding Single Large or Multinodular Hepatocellular Carcinoma Undergoing Surgical Resection or Transarterial Chemoembolization
, patients who undergoing SR included in our review; †, patients who undergoing TACE reported among eight included studies *