AIM: To evaluate clinical outcomes of patients that underwent surgery, transarterial embolization (TAE), or supportive care for spontaneously ruptured hepatocellular carcinoma (HCC).

METHODS: A consecutive 54 patients who diagnosed as spontaneously ruptured HCC at our institution between 2003 and 2012 were retrospectively enrolled. HCC was diagnosed based on the diagnostic guidelines issued by the 2005 American Association for the Study of Liver Diseases. HCC rupture was defined as disruption of the peritumoral liver capsule with enhanced fluid collection in the perihepatic area adjacent to the HCC by dynamic liver computed tomography, and when abdominal paracentesis showed an ascitic red blood cell count of > 50000 mm³/mL in bloody fluid.

RESULTS: Of the 54 patients, 6 (11.1%) underwent surgery, 25 (46.3%) TAE, and 23 (42.6%) supportive care. The 2-, 4- and 6-mo cumulative survival rates at 2, 4 and 6 mo were significantly higher in the surgery (60%, 60% and 60%) or TAE (36%, 20% and 20%) groups than in the supportive care group (8.7%, 0% and 0%), respectively (each, P < 0.01), and tended to be higher in the surgical group than in the TAE group. Multivariate analysis showed that serum bilirubin (HR = 1.09, P < 0.01), creatinine (HR = 1.46, P = 0.04), and vasopressor requirement (HR = 2.37, P = 0.02) were significantly associated with post-treatment mortality, whereas surgery (HR = 0.41, P < 0.01), and TAE (HR = 0.13, P = 0.01) were inversely associated with post-treatment mortality.

CONCLUSION: Post-treatment survival after surgery or TAE was found to be better than after supportive care, and surgery tended to provide better survival benefit than TAE.

Hepatocellular carcinoma (HCC) is the worldwide health problem and the third leading cause of cancer-related death globally[1-3]. Despite recent considerable advances in the understanding of tumor biology and the continued progression and development of diagnostic and therapeutic tools[4-7], the overall prognosis of HCC remains disappointing. In particular, due to its hypervascularity, HCC can exhibit rapid progression with direct invasion of surrounding tissues or it can invoke spontaneous tumor rupture[8]. HCC rupture is one of the life-threatening complications of HCC, and therefore, the most efficient treatment modality should be selected and rapidly applied to patients with ruptured HCC.

The incidence of spontaneous HCC rupture has decreased due to the earlier detection of HCC. Nevertheless, its incidence has been reported to be high as 3%-15% and its in-hospital mortality rate to range from 25% to 75% in the acute phase[9-12]. Open surgery was the main method used to treat HCC rupture from the 1960s to the 1980s[13-15]. Recently, survival benefit by transarterial embolization (TAE) has been reported[16-18]. However, to the best of our knowledge, no definite recommendation has been issued regarding optimal treatment of HCC rupture, and the comparative survival benefits of surgery and TAE remain unclear.

Therefore, in this retrospective study, we undertook to evaluate survival outcomes according to treatment modalities, that is, surgery, TAE, or supportive care, in patients with a spontaneously ruptured HCC, and sought to identify the factors that predispose post-treatment mortality in these patients.

We have shown here that overall survival rates of patients with ruptured HCC is significantly higher in patients with surgery or TAE than in those with supportive care, and tended to be higher in patients with surgery than in those with TAE. Furthermore, high serum bilirubin and creatinine levels, and vasopressor requirement were found to be significantly associated with post-treatment mortality. To our knowledge, this is the first study to investigate the survival benefit of surgery vs TAE. Although several previous studies have evaluated the therapeutic efficacy of surgery or TAE, in patients with spontaneous HCC rupture, direct comparative information is little available regarding survival outcomes after surgery and TAE in such patients.

Spontaneous HCC rupture is likely to occur in patients with advanced staged HCC with reported incidences of 10.0% in Japan[15], 12.4% in Thailand[10], and about 3.0% in the United Kingdom[26]. In the present study, the estimated incidence of spontaneous HCC rupture was 3.5%, and the overall 1-mo mortality was a high as 64% in our cohort, which are the similar to the outcomes of previous studies[10,11,27]. However, despite its high mortality, survival benefits for surgical treatment[13-15] and for TAE[16-18] have been reported in patients with a spontaneous HCC rupture. Likewise, in the present study, the overall survivals of patients with surgery or TAE were significantly higher than that in those with supportive care, especially in those in a hemodynamically stable state with a low serum bilirubin level, and good renal function. However, to date, there has been a dearth of reliable clinical evidence on the merits of surgical treatment versus those of TAE, in the context of survival benefit in patients with a spontaneous HCC rupture. Therefore, the present study may provide useful information for clinicians to determine the most appropriate treatment option for spontaneously ruptured HCC.

Ruptured HCC is a catastrophic disorder characterized by fatal complications, such as, coagulopathy, hemodynamic instability, or liver insufficiency. Thus, treatment should be considered carefully based on adequate information. As has been found in previous studies[13-15], surgical treatment was found to provide significant survival benefit as compared with supportive care in the current study. Furthermore, surgical group had relatively better hepatic function reserve, smaller tumors, and smaller numbers of tumors than the supportive care groups, although bias might have been introduced by selection for surgery. Nonetheless, the cumulative overall survival rate was higher in surgical group than in the supportive care group, and surgical treatment was found to be independent predictor of post-treatment survival by multivariate analysis. Although not all patients could have undergone surgery due to a poor hepatic function or an unstable vital status, surgical hemostasis can be considered if hepatic dysfunction or hemodynamic instability can be maximally corrected immediately after initial rupture of a hepatoma. Furthermore, post-surgical complications need to be considered before treatment decision-making despite the absence of an immediate severe complication after surgical intervention in the present study.

It has been reported that TAE is effective in achieving immediate hemostasis for ruptured HCC. In the present study, the overall survival rate was better in the TAE group than in the supportive care group. Advanced angiographic techniques enable the tumor location, active bleeding focus, and portal vein patency to be assessed, but life-threatening complications, such as, liver failure can developed after TAE, at rates ranging from 12% to 34%[12,17]. In the present study, post-TAE liver failure and technical failure for immediate hemostasis was observed in 6 (24%) and 5 (20%) patients of the TAE group, respectively, which suggests that TAE should be selectively administered in patients with good reserved hepatic function, tolerable coagulopathy, and a patent main portal vein.

In terms of comparison of survival benefits between the surgical and TAE groups, the current study failed to show a significant difference, although the cumulative overall survival rate tended to be higher in the surgical treatment group. Although TAE is less invasive than open surgical hemostasis, we suppose that open surgical treatment offer a higher chance of successfully achieving hemostasis by removing the bleeding focus or by allowing complete ligation of feeding artery. In previous studies, the 30-d mortality rate after TAE group has been reported to be lower than that after open surgical group[12,17,28]. However, in the present study, cumulative overall 1-mo survival rates were statistically similar between two groups. Moreover, at 1 mo after operative treatment, cumulative survival was clinically higher in the surgical group than in the TAE group despite the statistical insignificance (Figure 4). However, it should be borne in mind that this lack of significance may have been due to small patient numbers. Therefore, large number of patients who underwent surgical treatment need to be evaluated in the comparative study in the future.

Patients with a poor liver function reserve cannot tolerate surgical resection or aggressive angiographic intervention. Therefore, CTP class or MELD score, which reflect reserved hepatic function, could be important pretreatment factors. However, in the present study, of variables comprising the CTP class, only serum bilirubin was found to be independently associated with post-treatment survival. Likewise, of the variables comprising the MELD scores, only serum bilirubin and creatinine were found to be independent factors. Therefore, we estimated individual variables of CTP class or MELD score to avoid overestimation of the other factors in the multivariate analysis of post-treatment survival in patients with spontaneous HCC rupture.

Elevated serum creatinine and vasopressor requirements are likely to reflect multiorgan failure[29], and decreased effective circulating volume or the presence of superimposed infection may induce alterations in organ perfusion and in hemodynamic stability. Moreover, inflammatory reactions triggered by hepatocellular necrosis after HCC rupture may contribute to liver insufficiency, and subsequent multiorgan failure. Although decreased serum albumin level and a prolonged prothrombin time suggest reduced liver synthetic function, they can be corrected by albumin and coagulation factor replacement. On the other hand, serum bilirubin level cannot be rapidly and artificially corrected immediately after parenchymal liver damage. Therefore, serum bilirubin may be an independent factor of patient survival, unlike serum albumin or INR. Furthermore, the importance of serum bilirubin level in patients with HCC rupture has been previously reported[17,30]. However, during treatment decision-making, age, INR, albumin, and tumor size and number may also be clinically important variables despite their lack of statistical significance in the present study.

Our study has several limitations. First, the study is inherently limited by its retrospective study design. However, we enrolled all eligible patients. Second, the varied clinical and tumor statuses of patients and the critical disease status prevented randomization, and probably introduced bias. Furthermore, it takes long time to collect the prospective database of patients due to low incidence of the disease. Third, the absolute number of patients who underwent surgery was small, and therefore, there might be no significant difference in cumulative survival rate between surgery and TAE groups. Accordingly, we suggest a large-scale study be conducted to confirm our study.

In conclusions, the present study suggests that the post-treatment outcomes of surgery or TAE are better than that of supportive care in patients with spontaneous HCC rupture, and that surgical hemostasis might provide better survival benefit than TAE. However, we advise that serum bilirubin, creatinine, and hemodynamic status should be considered during treatment decision making. Regardless of its shortcomings, we believe that the present study would provide important information that aids decision making in patients with spontaneous HCC rupture.