Diagnostic and therapeutic management of hepatocellular carcinoma

Abstract

Hepatocellular carcinoma (HCC) is an increasing health problem, representing the second cause of cancer-related mortality worldwide. The major risk factor for HCC is cirrhosis. In developing countries, viral hepatitis represent the major risk factor, whereas in developed countries, the epidemic of obesity, diabetes and nonalcoholic steatohepatitis contribute to the observed increase in HCC incidence. Cirrhotic patients are recommended to undergo HCC surveillance by abdominal ultrasounds at 6-mo intervals. The current diagnostic algorithms for HCC rely on typical radiological hallmarks in dynamic contrast-enhanced imaging, while the use of α-fetoprotein as an independent tool for HCC surveillance is not recommended by current guidelines due to its low sensitivity and specificity. Early diagnosis is crucial for curative treatments. Surgical resection, radiofrequency ablation and liver transplantation are considered the cornerstones of curative therapy, while for patients with more advanced HCC recommended options include sorafenib and trans-arterial chemo-embolization. A multidisciplinary team, consisting of hepatologists, surgeons, radiologists, oncologists and pathologists, is fundamental for a correct management. In this paper, we review the diagnostic and therapeutic management of HCC, with a focus on the most recent evidences and recommendations from guidelines.

Key Words: Cancer; Chronic hepatitis; Cirrhosis; Hepatitis B virus; Hepatocellular carcinoma; Hepatitis C virus; Liver

Core tip: Hepatocellular carcinoma is an increasing health problem, representing the second cause of cancer-related mortality worldwide. The major risk factor for hepatocellular carcinoma (HCC) is cirrhosis. Early diagnosis is crucial for curative treatments. As a consequence, patients at risk of developing HCC should undergo surveillance programs in order to detect HCC in the initial stage. Surgical resection, radiofrequency ablation and liver transplantation are considered the cornerstones of curative therapy, while for patients with more advanced HCC recommended options include sorafenib and trans-arterial chemo-embolization.

INTRODUCTION

Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related deaths worldwide. More than 700000 new cases are diagnosed every year throughout the world and high incidence to mortality ratio (1.07) makes HCC the second most common cause of cancer-related deaths worldwide[1].

Although the majority of cases occur in Asia and Africa, the incidence has increased even in the developed world. The geographical variation in the incidence of HCC is mostly related with the different prevalence of major risk factors for HCC, such as hepatitis C virus (HCV) and hepatitis B virus (HBV) infection[2]. In developed countries, the epidemic of obesity, diabetes and nonalcoholic steatohepatitis (NASH) is also believed to contribute to the observed increase in HCC incidence[3]. However, the overriding risk factor for HCC, which is responsible for HCC in 80%-90% of cases regardless of etiology, is the presence of cirrhosis[4,5].

By recognizing the risk factors for HCC, high-risk groups can be identified and followed up with screening strategies. In fact, the management of high-risk patients with screening and surveillance has the real potential to detect HCC early and improve patient outcomes. When HCC is detected earlier, patients are candidates to receive curative treatments.

In this paper, we review the diagnostic and therapeutic management of HCC, with a focus on the most recent evidences and recommendations from guidelines.

DIAGNOSIS

Hepatic nodules can be detected on Ultrasounds (US), including contrast-enhanced US (CEUS), or on other noninvasive techniques, such as contrast-enhanced computerized tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET)-CT. The typical vascular profile of HCC on dynamic imaging is characterized by early arterial phase enhancement followed by loss of enhancement in the portal venous phase and delayed phase in comparison to the surrounding liver[6].

Moreover, molecular biomarkers could potentially be used for diagnosis, as well as prognostic evaluation and may help defining the individualized therapeutic approach to HCC. Figure 1 illustrates the diagnostic algorithms endorsed by the American Association for the Study of Liver Diseases (AASLD)[7].

Figure 1 American Association for the Study of Liver Diseases diagnostic algorithm for suspected hepatocellular carcinoma (adapted from[7]). CT: Computed tomography; MDCT: Multidetector CT; MRI: Magnetic resonance imaging; US: Ultrasound; HCC: Hepatocellular carcinoma.

US

Early diagnosis of HCC is important because, as expected, treatment is more effective when the tumor is small[8-10]. Dysplastic nodules (DNs) may develop into carcinoma[11]. Early detection of DNs with small areas of HCC is fundamental for effective treatment.

Cirrhotic patients are recommended to undergo HCC surveillance by abdominal US at 6-month intervals. However, the diagnosis of small HCC nodules may be challenging, as it is often difficult to differentiate benign from malignant lesions in the context of nodular cirrhosis; moreover, US depends on the operator and has limited sensitivity in obese patients. On the other hand, US is less expensive than other techniques and is radiation-free.

For HCC, a stepwise process of carcinogenesis has been proposed, involving a progression from regenerative nodules (RNs) to low-grade DNs or high-grade DNs to DNs with a focus of HCC and finally to HCC. This progression has been suggested to correlate with changes in the blood supply and perfusion of the nodules, which may be used to differentiate focal liver lesions[8,12-14]. Recently, SonoVue, a blood-pool marker used in CEUS, has been reported to help distinguishing RNs from small HCC based on the different enhancement pattern[15-18]. RN has an intranodular blood supply that is similar to the surrounding parenchyma. On the other hand, HCC usually exhibits an enhancement pattern in the arterial phase and washout in the late phase[19-26]. DN-HCC nodules have a mixed enhancement behavior, as they are composed of two different cells, high-grade HCC and atypical hepatic cells and their enhancement features are partially similar to HCC and partially similar to RNs[27]. Of interest, CEUS has been suggested to promote the diagnostic accuracy of biopsy, decreasing the false-negative rate for malignant lesions. In fact, CEUS may be used to identify the areas of viable tumor[27]. A biopsy of DN-HCC without CEUS guidance is more likely to give false-negative results, significantly affecting the possibility to early detect and treat HCC.

CT

CT is largely used in most centers to make the radiological diagnosis of HCC after a liver nodule is detected on US. Most centers use a four-phase multidetector CT (MDCT) scan, which consists of a non-enhanced phase, an arterial phase (which occurs 20-30 s after contrast injection), a portal venous phase (6580 s after contrast injection) and a delayed phase. On the four-phase CT, HCC classically appears as a hyper-attenuated lesion in the arterial phase, with loss of enhancement termed rapid washout in the portal venous and/or delayed phase. CT has high specificity but variable sensitivity for detecting HCC. In a systematic review, traditional spiral CT was reported to have a specificity of 93% but a sensitivity of only 68% in diagnosing HCC. A more recent review assessing the diagnostic accuracy of the 64-slice MDCT technology vs spiral CT found improved sensitivity (65%-79% compared to 37%-54%), with similar specificity (above 90%)[28]. However, sensitivity dropped to 33.45% for nodules smaller than 1 cm.

MRI

MRI is an appealing imaging technique, since it does not use ionizing radiation. MRI allows the differentiation between tumoral and normal liver parenchyma using magnetic fields, even without a contrast media[29]. Traditional dynamic contrast-enhanced MRI of the liver is performed using gadolinium chelates. In gadolinium-enhanced MRI, the typical HCC lesion is hyper-intense on T1-weighted images during the arterial phase and exhibits rapid washout during portal venous and delayed phases[12,30,31]. The sensitivity of standard gadolinium-enhanced MRI is around 90%, with a specificity of at least 95% for the detection of HCC greater than 2 cm in diameter[32]. Dynamic MRI appears superior to CT for the detection of HCC nodules[33,34], but its sensitivity is highly affected by the lesion size, being as low as 30% in the case of lesions smaller than 2 cm[35,36].

Specific contrast agents have been developed to improve the sensitivity of MRI for HCC, including the “dual contrast” agents (gadolinium-ethoxybenzyl-diethylenetriaminepentaacetic acid and gadobenate dimeglumine), which work both as markers of hepatobiliary excretion and vascularization. HCC nodules imaged with these contrast agents do not exhibit uptake, unlike benign nodules on the delayed phase[37]. While it appears performing similarly to MDCT for lesions larger than 2 cm, enhanced MRI might be more sensitive for lesions smaller than 1 cm[38].

Nuclear imaging

Both ¹⁸F-FDG and 11C-acetate PET imaging have been used for HCC detection and staging[39-41]. However, up to 40%-50% of HCC are not sensitive to ¹⁸F-FDG PET, because of their high expression of the glucose-6-phosphatase enzyme, which prevents intracellular accumulation of ¹⁸F-FDG[39]. On the other hand, 11C-acetate, which is believed to mainly participate in fatty acid synthesis in the liver, has been suggested to have increased sensitivity and specificity in comparison to ¹⁸F-FDG[39,41,42]. However, several studies have reported that 11C-acetate PET does not properly differentiate HCC from benign lesions[41,43-45], because the latter also accumulate 11C-acetate. Some recent studies[41,44-47] have suggested dynamic PET with kinetic modeling to be a promising tool to differentiate benign hepatic tumors from HCC.

Biomarkers

Α-fetoprotein (AFP) is the most widely used and broadly known biomarker for HCC, but its use as an independent tool for HCC surveillance is not recommended by current guidelines due to its low sensitivity and specificity. In the past, a significant concentration of AFP in the serum of a patient with liver cirrhosis and a suspicious mass in the liver larger than 2 cm was sufficient to diagnose HCC[48]. However, the current diagnostic algorithms endorsed by the AASLD and the European Association for the Study of the Liver strictly rely on typical radiological hallmarks in dynamic contrast-enhanced imaging apart from biomarkers[7,48,49].

Three serum biomarkers have been suggested as tools to determine the risk of liver cancer in high-risk populations worldwide: AFP, the ratio of lecithin-bound AFP to total AFP (AFP-L3), and des-gamma-carboxyprothrombin (DCP). However, most studies on the performance of biomarkers in HCC detection have not been performed in a surveillance setting but compared levels of predefined biomarkers in patients with HCC with a control group, in most cases represented by patients with chronic liver diseases. A randomized controlled study performed in a high-risk population in China showed that screening by AFP measurement led to earlier diagnosis of HCC but had no impact on mortality[50]. On the other hand, semiannual screening for HCC by AFP measurement in a population-based study in Alaska was effective in detecting HCC at early stages and significantly prolonged survival rates[51].

A meta-analysis on the performance of AFP in diagnosing HCC included seven studies and revealed a pooled sensitivity of 66% with a specificity of 86% [area under curve (AUC) = 0.87][52]. In a further meta-analysis including ten studies the pooled sensitivity of AFP for the diagnosis of HCC was 51.9%, with a specificity of 94% (AUC = 0.81)[53]. A major drawback of AFP as a surveillance tool is that serum levels are influenced by the activity of the underlying liver disease and therefore increased in patients with elevated alanine aminotransferase (ALT) levels, even in the absence of HCC, as shown in the HALT-C trial[54]. Furthermore, HCC biology is quite heterogeneous, with only a proportion of patients with HCC having elevated AFP serum levels, leading to low sensitivity of the marker. As a consequence, new complementary markers have been studied. The clinical utility of high-sensitivity AFP-L3 (hs-AFP-L3) in early prediction of HCC development in patients with chronic HBV or HCV infection was recently evaluated in a large Japanese study. Even in subjects with low AFP levels and without suspicious ultrasound findings, an elevation of hs-AFP-L3 was an early predictor of HCC development: in fact, hs-AFP-L3 increased in 34.3% of patients one year prior to diagnosis of HCC[55,56]. Numerous studies have investigated the performance of other markers, including α-l-fucosidase[57], glypican-3 (GPC-3), insulin-like growth factor[58], vascular endothelial growth factor (VEGF), or Dickkopf-1[59], Golgi protein 73 (GP73), interleukin-6 (IL-6) and squamous cell carcinoma antigen (SCCA)[60]. In a study comparing 144 patients with HCC to 152 patients with cirrhosis and 56 healthy controls, GP73 had a sensitivity of 62% and a specificity of 88% at a cut-off of 10 relative units[61]. Another study, including 4217 subjects (789 with HCC), revealed a sensitivity of 74.6% and a specificity of 97.4% at a cut-off of 8.5 relative units[62]. Using different cut-off values, IL-6 sensitivity ranged from 46% to 73% with a specificity of 87% to 95%[60,63,64]; in a large study including 961 patients, SCCA had a sensitivity of 42% and specificity of 83% using a cut-off of 3.8 ng/mL[60,65].

Serum IL-17 levels have been reported to be elevated in HCC patients[66]. In a retrospective study, Liu et al[67] found that plasma IL-17 concentration had a sensitivity of 74.3% and specificity of 75.6% (AUC = 0.86) at the cut-off value of 4.23 ng/L; however, the diagnostic accuracy of IL-17 was lower than AFP, which had a sensitivity and specificity of 100% and 66% respectively, at the cut-off value of 10.25 mg/L (AUC = 0.96).

Osteopontin, an integrin-binding glycol-phosphoprotein, was investigated in seven studies summarized in a meta-analysis[52]. The pooled sensitivity of osteopontin for HCC was 86% with a specificity of 86%, showing a diagnostic accuracy similar to that of AFP; however, further validation studies are needed before recommending the use of this biomarker in clinical practice.

Some studies have investigated the combined diagnostic performance of the three more validated non-invasive biomarkers used in HCC, namely AFP, AFP-L3 and DCP. By comparing 164 European patients with HCC to 422 subjects with chronic liver disease, a significant increase in AFP serum levels was shown in those with advanced HCC and viral hepatitis, while DCP was more frequently elevated in those with early-stage and NASH-associated HCC. Neither of the two parameters, if taken alone, could independently identify more than 30% of patients with HCC but combination of AFP (cut-off 10 ng/mL) and DCP (cut-off 5 ng/mL) showed a sensitivity of 55% for early stage HCC and 78% for all stages[68]. A further increase in sensitivity (up to 84%) was observed by adding AFP-L3[69]. The additional use of clinical variables, like age and gender, further improved the performance of the model[70,71].

A number of signal transduction pathways have been recognized as critical players in the pathophysiology of hepatocarcinogenesis, including the Wnt/β-Catenin pathway, the p53 pathway, the tumor suppressor retinoblastoma protein pRb1 pathway, the mitogen-activated protein kinase pathway, the Ras pathway, JAK/STAT signaling, mechanisms of cellular stress response, like heat shock proteins, epidermal growth factor receptor and transforming growth factor-β signaling[72,73].

Gene expression profiling of peripheral blood mononuclear cells using microarrays and bioinformatics-driven data analysis identified a blood-based signature of three genes, namely Chemokine (C-X-C motif) receptor 2 (CXCR2), C-C chemokine receptor type 2 (CCR2) and E1A Binding Protein P400 (EP400), able to predict HCC with a sensitivity of 93% and a specificity of 89%[74]. High-throughput metabolomics technology with the comprehensive analysis of small molecular metabolites may identify serum metabolic profiles to be used as biomarkers in HCC diagnosis. Molecular signatures may help to distinguish dysplastic nodules from well-differentiated HCC. In Asian and Western patients with HCV infection, specific gene signatures have been reported to accurately reflect the pathological progression of disease from cirrhosis to dysplasia to early and advanced HCC[75,76]. Moreover, a three-gene set including glypican3 (GPC3; 18-fold increase in HCC, P = 0.01), LYVE1 (12-fold decrease in HCC, P = 0.0001) and surviving (2.2-fold increase in HCC, P = 0.02) had an accuracy of 94% to distinguish DNs from early HCC in HCV-related cirrhosis[77]. Heat shock protein 70 and cyclase-associated protein 2 are other tissue biomarkers potentially useful to in the diagnosis of HCC[78,79].

As for novel biomarkers, microRNA (miRNA) have received particular attention[80]. miRNAs are small non-coding and evolutionary conserved RNA molecules that serve as posttranscriptional regulators of mRNA expression and interfere with mRNA translation to protein[81]. miRNAs are able to conserve their function into the cell by regulating the expression of a target population of molecules; moreover, they can be released from the cell both in combination with other proteins or as a free molecule[82-88].

Differences in miRNA expression patterns in several malignant conditions, including HCC, have been found[89-91]. In particular, three miRNAs, miR-122, miR-192 and miR-199a/b-3p, account for more than 70% of total miRNA released by normal liver tissue[91]. In HCC, a broad spectrum of changes in microRNAoma has been reported[91-94], suggesting that miRNAs may potentially become valid biomarkers in HCC. To improve the diagnostic utility of miRNAs in HCC, Li et al[95] performed deep sequencing in pooled samples from patients with chronic HBV patients, HCC and controls with and without cancer. They recognized a pattern of 6 miRNA differentially expressed in patients with HCC. The use of three miRNAs (miR-25, miR-375 and let7f) had a sensitivity of 97.9% and a specificity of 99.1% to discriminate between controls and HCC patients. Of interest, the use of two miRNAs (miR-10a and miR-125b) could adequately discriminate the cohort with chronic HBV and HBV-associated HCC with an AUC of 99.2% (sensitivity 98.5% and specificity 98.5%)[95]. In another study, a panel of 7 miRNAs (miR-122, miR-192, miR-21, miR-223, miR-26a, miR-27a and miR-801) provided a high diagnostic accuracy for the identification of HBV-related HCC[96], with a sensitivity of 81.8% and a specificity of 83.5%, independently of disease stage. However, the expression of selected miRNA was analyzed using RT-PCR, which may be critical for clinical translation of these findings[96]. miR-21, which is the most frequently deregulated miRNA in cancer, was found at higher level both in sera and plasma of HCC patients[97,98], while other studies showed no significant differences[99,100]. Similarly, miR-122, the most abundant miRNA in the liver, was also found at high level in sera of HCC patients[98,99]. Other inflammatory conditions of the liver, such as acute and chronic hepatitis and NASH may strongly influence miR-122 levels[101,102]. In this setting, further studies are required to establish the capability of these biomarkers to discriminate between chronic liver diseases and HCC.

STAGING

A number of staging systems have been used for HCC, even if the Barcelona Clinic Liver Cancer (BCLC) staging system is the most extensively used in clinical practice. BCLC staging system includes the evaluation of tumor stage, cirrhosis stage, functional performance status (PS) and it links staging with a treatment algorithm[103]. Moreover, the BCLC staging system was endorsed by both the American and European liver society and validated in European and American cohorts[104,105].

Early stage HCC (stage 0) has the best prognosis and is characterized by the presence of one lesion smaller than 2 cm in diameter, with no evidence of vascular invasion, in patients with stable cirrhosis (Child-Pugh class A).

Patients with stage A HCC could present with either a solitary lesion or up to three lesions of less than 3 cm in diameter. These patients have relatively preserved liver function (Child-Pugh class A or B) and good functional status (PS 0-2). The 5-year survival rate is 50%-75%; as reported in Figure 2, treatment may be different based on the presence of portal hypertension, the degree of liver dysfunction and other comorbidities.

Figure 2 Barcelona Clinic Liver Cancer staging system and treatment strategy for hepatocellular carcinoma (adapted from Ref. [49]). M: Metastasis classification; N: Node classification; PS: Performance status; RFA: Radiofrequency ablation; TACE: Transarterial chemoembolization.

Patients with intermediate stage HCC (stage B) have Child-Pugh class A or B cirrhosis, good functional status (PS 0) and multinodular HCC, with no evidence of vascular invasion. Patients with evidence of vascular invasion or extra-hepatic spread have advanced stage HCC (stage C). These patients typically have worse functional status (PS 1 or 2).

Patients with terminal stage HCC (stage D) present with decompensated cirrhosis (Child-Pugh class C), poor functional status (PS > 2), and advanced tumor growth (vascular invasion and/or extra-hepatic spread). Unfortunately, these patients receive no benefit from the currently available therapies, and survival is usually around 3 mo. Figure 2 illustrate the BCLC staging system and treatment strategies for HCC.

TREATMENT

Several therapeutic options are available for HCC, depending on HCC stage, liver function, comorbidities, and local clinical expertise. A multidisciplinary team, consisting of hepatologists, surgeons, radiologists, oncologists and pathologists, is fundamental for a correct management.

Surgery

The evaluation of liver functional reserve before hepatectomy is fundamental the maximum amount of liver mass that can be safely removed: on the one hand, liver functional overestimation may lead to hepatic failure; on the other hand, poor resection may significantly increase the risk of early recurrence of HCC.

The most important methods to assess liver function before surgery are the galactose tolerance test, ⁹⁹mTc-galactosyl human serum albumin liver scintigraphy and the indocyanine green (ICG) test. Makuuchi’s selection criteria for hepatectomy rely on three factors, ascites, serum bilirubin and ICG retention rate at 15 min (ICGR15)[106]. Patients are considered eligible for liver resection if they have no ascites and if serum bilirubin is ≤ 2 mg/dL. Patients with total bilirubin of 1.1-1.9 mg/dL can undergo partial liver resection; for patients with serum bilirubin ≤ 1 mg/dL, the extent of resection is based on ICGR15: (1) resection of 2/3 of the total liver volume (TLV) (e.g., right lobectomy) in patients with normal ICGR15 of < 10%; (2) resection of 1/3 of the TLV (e.g., left lobectomy) in patients with ICGR15 of 10%-19%; (3) resection of 1/6 of the TLV in patients with ICGR15 of 20%-29%; and (4) limited resection or enucleation in patients whit ICGR15 ≥ 30%.

A surgical mortality rate of 0% has been reported in 1056 consecutive hepatic resections performed in accordance with these criteria[107].

In patients with portal venous invasion[108], the area supplied by the portal vein branches should be systemically removed as much as possible within the acceptable range of liver function. In this contest, systematic subsegmentectomy has been developed to overcome the potential incompatibility between the attempt to remove cancer and the need to preserve liver function[109]. Tumor stage, tumor size, number of tumors and capsule formation predict recurrence-free survival. Moreover, vascular invasion is a poor indicator of long-term survival[110]. In one study, risk factors for early recurrence (within 2 years after surgery) were non-anatomical resection, microscopic vascular invasion, and AFP ≥ 32 ng/mL[111]. Another retrospective study confirmed the association between the type of surgical approach and the outcome, showing that the cumulative survival rate was significantly higher after anatomical resection compared to non-anatomical resection[112].

As reported above, one crucial issue is the determination of the adequate liver remnant volume after hepatectomy. In normal livers, it is important to preserve the 20%-40% of the TLV or the standard liver volume (SLV)[113-120]. Anderson et al suggested that the smallest adequate liver remnant volume should be ≥ 20% of the SLV in patients with no underlying chronic liver disease[114,121]. However, HCC usually develops in patients with chronic hepatitis or cirrhosis, who are at risk of hepatic failure in case of insufficient liver remnant volume after hepatectomy. In this setting, portal vein embolization (PE) may prevent hepatic failure, because the portal vein branches are blocked to induce compensatory hypertrophy in the remnant liver area[122]. Three-dimensional CT scan allows an accurate determination of the position of major blood vessels and the tumor, as well as resection margins, and liver remnant volume[123].

Perioperative complications include bile leakage, hemorrhage and intra-abdominal abscesses[124,125]. Intraperitoneal drainage is necessary for monitoring and treatment of these complications, even if the Center for Disease Control and Prevention guidelines do not recommend routine drainage in elective hepatectomy. If drainage is required, a closed suction drain should be used and placed through a separate incision distant from the operative one. Moreover, the drainage should be removed as soon as possible[126]. These recommendations have been validated in several studies[127-133]. Moreover, in a randomized clinical trial, subcutaneous drainage was not effective in preventing surgical site infections[134]. Hepatic failure and disseminated intravascular coagulation are other postoperative complications. In one study, the authors evaluated the efficacy of steroids to improve liver function after hepatectomy[135]. They found that serum bilirubin levels were significantly lower in the steroid group on post-operative day (POD) 2 compared with the non-steroid group. The postoperative time courses of bilirubin, IL-6 and the C-reactive protein level were significantly lower whereas the prothrombin level was significantly higher in the steroid arm. No differences in the proportion of patients with complications and the length of hospital stay were reported between the two groups. To unify the definition of post-hepatectomy liver failure (PHLF), the International Study Group of Liver Surgery proposed defining PHLF as an increased international normalized ratio and concomitant hyperbilirubinemia on or after POD 5[136]. PHLF seems to predict the incidence of complications and mortality better than the 50-50 criteria [i.e. prothrombin time (PT) < 50% and serum bilirubin > 50 mmol/L][137] and MELD score[138].

Liver transplantation

Liver transplantation has become a feasible alternative for many patients with HCC, given the advances in surgical techniques and immunosuppression.

In 1996, Mazzaferro et al[139] defined the so-called Milan criteria, which identified as eligible for transplantation patients with solitary lesions < 5 cm in diameter and those with up to 3 lesions, each one < 3 cm in diameter. Similar survival rates in patients with tumors < 3 cm have been reported by the Bismuth group[140]. The Milan criteria have been accepted worldwide to identify patients which can be safely tranplanted. The limited number of available organs is the main limitation for this procedure. Yao et al[141,142] demonstrated that patients with a single lesions ≤ 6.5 cm, or up to three lesions each one ≤ 4 cm with a cumulative diameter ≤ 8 cm had surgical outcomes similar to those transplanted on the basis of Milan criteria. Tumor histology has an important impact on post-transplantation survival, with better outcome in patients with well-differentiated tumors[143,144]. The availability of transplantable grafts remains the critical issue for all patients awaiting liver transplantation, considering that time is a major determinant of overall survival[145-153]. Living donation can be a good choice for transplantation in patients with HCC because the transplant can be planned with an optimal timing to both assess the tumor aggressiveness and minimize the risk of recurrence[154-160]. Another factor that can affect the risk of recurrence after transplantation is the use of immunosuppressive agents. Sirolimus, a bacterial macrolide with immunosuppressive and antineoplastic properties, which inhibits IL-2-mediated lymphocyte proliferation, seems to decrease metastatic tumor growth and angiogenesis in the liver. It was demonstrated that the administration of post-transplant sirolimus, within a steroid-free protocol and a low tacrolimus target, was associated with decreased risk of tumor recurrence and no significant increase in the risk of infection and hepatic artery thrombosis[161-165].

Non-surgical management

Among non-surgical approaches, percutaneous ethanol injection (PEI), microwave ablation (MWA) and percutaneous radiofrequency ablation (RFA) represent the three most widely used techniques for the treatment of HCC less than 5 cm in diameter and/or with less than 3 tumoral lesions.

In RFA, electrical current is applied via an electrode resulting in resistive heating and tissue hyperthermia[166]. Tissues adjacent to the electrode are the most effectively heated[167-169]. The mechanism of cytotoxicity in RFA depends on tissue impedance, with power deposition hindered in regions of high tissue impedance, such as the surrounding lung or tissue adjacent to the electrode, that has undergone water vaporization due to rapid heating[166,169]. Multiple engineering designs have been developed to overcome the limitations caused by tissue impedance, including multi-tined electrodes to expand the contact surface area, saline injection, and internal cooling. Moreover, RFA requires the placement of grounding pads on the patient to close the electrical circuit, and skin burns related to the pads have been reported[170,171]. However, in clinical practice skin burns are rare, considering that larger grounding pads are usually used to improve the dispersion of thermal energy[172]. RFA efficacy may be limited by the “heat sink” effect, consisting in heat dissipation resulting from blood flow. This effect is more marked for lesions close to the liver hilum[173]. There are several reports on RFA use in both primary and metastatic liver tumors. In a Cochrane database analysis, including 11 randomized clinical trials, Weis et al[174] analyzed a total of 1819 participants with HCC with the primary outcome of overall survival, comparing RFA to hepatic resection[175-177], PEI[178-183], MWA[184], and percutaneous laser ablation (PLA)[185]. The authors concluded that hepatic resection was superior to RFA in terms of survival, even if RFA might be associated with fewer complications and shorter hospital stay. Moreover, RFA was associated with better survival than PEI, whereas there was no evidence of significant differences between RFA and MWA or PLA.

In the study by Lee et al[186], patients undergoing surgical resection were younger and had better liver function reserve and PS than those receiving RFA. When accounting for these differences using propensity score analysis, RFA was superior to surgery for patients with small HCC and Child-Pugh Turcotte score of 5.

MWA relies on the direct application of an electromagnetic field, which causes dielectric hysteresis, leading to local tissue hyperthermia[187]. MWA is able to penetrate through several tissues, including those with high impedance[166,187]. High tissue temperatures can be achieved with MWA, with increased efficacy as compared to RFA[188]. Given the efficacy profile and the shorter time required to achieve ablation, the use of MWA has gradually increased for the treatment of both primary and metastatic tumors of the liver. Ding et al[189,190] studied 198 patients with HCC, all in BCLC Stage A meeting Milan criteria and did not find any difference between RFA and MWA in terms of disease-free survival, cumulative survival, and complication rates. Similar results have been reported in other cohorts[184,191].

PEI involves the direct instillation of ethanol into tumors, which results in coagulative necrosis. The technique is relatively simple and inexpensive. However, in clinical practice PEI is limited by poor and irregular distribution of ethanol within the tumor and diffusion into the adjacent normal tissues. Even if some studies with PEI reported favorable outcomes after a long-term follow up (greater than 15 years), most evidences suggest that RFA is associated with better overall survival than PEI[192-194].

The use of external beam radiation therapy (SBRT) in treatment of liver tumors has been traditionally limited by the overall low tolerance of liver tissue to radiation[195]. In fact, radiation produces tumoral killing by transferring energy within atoms, determining the generation of reactive oxygen species with subsequent direct and indirect DNA and cellular damage. The final step is the generation of double-strand DNA breaks, leading to tumor cell death. Radiation can achieve excellent tumor control when delivered to ablative doses[196]. Maximum dose is limited by the radiation tolerance of the surrounding normal liver tissue and adjacent organs. Particularly, radiation-induced liver disease is a complication typically manifesting with the triad of anicteric hepatomegaly, ascites, and elevation of alkaline phosphatase. Imaging techniques, breathing motion control and advances in radiation machines technology permit accurate localization of hepatic tumors and help directing radiation to the tumor while minimizing exposure of surrounding normal liver[197,198]. The size and number of lesions that can be targeted, as well as the radiation dose that can be delivered, depends on normal liver reserve and estimated risk of liver complications. As expected, patients with reduced liver function require dose reduction[199]. Similarly, patients with Child Pugh class B cirrhosis may require dose reduction, while those with Child Pugh class C cirrhosis are not usually eligible for this type of treatment.

Another radiation-based technique is high-dose rate (HDR) CT-guided interstitial brachytherapy[200-202]. Radiation is delivered using an iridium-192 source as a single fraction. The advantage of this technique is a greater protection of the surrounding healthy liver compared to external radiation techniques. A prospective phase II trial[203] showed encouraging results for patients with large tumors near the hilum, using average dose of 17 Gy. Mearini et al[204] reported favorable outcomes of 35 patients with HCC (tumor size 5-12 cm), treated with HDR brachytherapy. At 12 mo, local control was 93% and no major toxicity was reported.

High-intensity focused ultrasound (HIFU) incorporates multiple ultrasound beams produced by piezoelectric or piezoceramic transducers directed into a three-dimensional focal point[205]. Ultrasound beams are both thermally ablative and cause cavitations to the underlying tissues. Coupling of the ultrasound source and the patient is achieved through a degassed water bath in order to have minimal reflection or absorption of the soundwaves prior to reaching the focal point. The patient is required to minimize movements during the procedure and the focal zone is shifted step by step to cover the area of interest for ablation. The safety and efficacy of HIFU was evaluated in several studies[206-212]. Ng et al[208] reported on a series of 49 patients with HCC (median tumor size 2.2 cm, range, 0.9-8 cm) and concluded that HIFU was effective for those who were not surgical candidates. He reported 1- and 3-year overall survival rates of 87.7% and 62.4%, respectively[209]. Similar data were published by Wu et al[210], with overall survival rates of 86.1%, 61.5%,and 35.3% at 6, 12, and 18 mo, respectively. Cheung et al[205] reported on the outcomes of HIFU for the treatment of HCC before liver transplantation in 10 patients as compared to 29 patients who received transarterial chemoembolization and found that HIFU was effective (90% had complete response, 10% partial response), with none of the patients on the liver transplant list (n = 5) dropping out[206].

Irreversible electroporation (IRE) is an apparently non-thermal technique in which the direct placement of electrodes creates a pulsed direct current, inducing cytotoxicity in tumor cells by altering transmembrane potentials, which irreversibly disrupt cell membrane integrity[213]. IRE requires the position of at least two applicators in parallel to create ablation zones in the range of 1.5-2 cm per electrode pair[214]. The zone of ablation created by IRE is dependent on multiple factors[213,215], such as electrode spacing and relative position, active tip length, pulse number and duration, and applied voltage. Because of these factors, IRE results more technically challenging than other locally ablative techniques. Moreover, the current generated by IRE causes whole-body muscle contractions and general anesthesia, requiring the use of neuromuscular blockage. In addition, IRE can induce cardiac arrhythmias, though this complication can be avoided with the use of cardiac synchronization of the administered pulses to the complete refractory period of the cardiac cycle[216]. IRE has a theoretical safety advantage as compared to other locally ablative techniques in the treatment of tumors close to structures susceptible to thermal injury, such as major bile ducts. In addition, because of the reduction in the “heat-sink” effect, IRE is potentially more effective for tumors next to major vessels, especially for smaller lesions, and showed excellent local tumor control at 3-6 mo, but high recurrence rates after 12-18 mo[215,217-220].

Cryoablation involves the direct application of a cryoprobe into the tumor. The thermal contact with the tumor results in ice-crystal development and osmotic shock. One recognized advantage of cryoablation is that the zone of ablation is readily visible (“iceball”) using CT scan, US, or MRI monitoring, allowing for precise targeting of the ablation area[221]. Moreover, multiple probes can be used simultaneously to create larger ablation zones and shorten procedural times. Despite the technical advantages of cryoablation, its use has been limited by the safety profile. Cryo-shock is an uncommon but potentially life-threatening complication, characterized by thrombocytopenia, acute renal failure, adult respiratory distress syndrome and disseminated intravascular coagulopathy[221]. In a meta-analysis comparing cryoablation to RFA in the treatment of unresectable HCC[222], RFA was superior, particularly in terms of complication rates and local tumor recurrence.

Percutaneous laser ablation (PLA) involves the direct deposition of laser light via fiber-optic applicators to induce tissue hyperthermia in tumors. The thin flexible fiber-optic delivery fibers allow for safer and technically easier approaches to tumors[223]. Moreover, feedback and dose-planning systems allow a good control of ablative zones and consequently low complication rates. However, it has been suggested that PLA has some limitations in achieving complete tumor ablation as compared to other locally ablative therapies[185,224,225].

HCC is preferentially supplied by the hepatic arterial inflow, while the normal parenchyma is largely supplied by the portal vein. The trans-arterial chemo-embolization (TACE) procedure is based on these blood supply dynamics. TACE consists in the placement of an intra-arterial catheter in the vessels supplying the tumor, to deliver high concentrations of a chemotherapeutic agent (e.g., doxorubicin, cisplatin or mitomycin) along with an embolic agent, such as lipiodol gelatin sponge or polyvinyl alcohol particles, in order to achieve both targeted chemotherapy and reduction in arterial supply to the tumor.

Drug eluting beads TACE (DEB-TACE), are becoming largely popular because of the favorable safety profile. DEB-TACE delivers small beads, which have been saturated for several hours with chemotherapeutic drugs. The beads occlude the feeding vessels of HCC, while doxorubicin is progressively released, increasing chemotherapeutic concentrations locally and creating tumor necrosis. The choice of bead size, from 75 to 700 μm, depends on tumor size and the preferred level of concentration within the treated volume. The best results are achieved when chemoembolization is performed selectively to segmental or subsegmental arteries feeding the tumor[226]. TACE is considered the standard of care for intermediate stage HCC without vascular invasion or metastases. In several randomized controlled trials, TACE was associated with partial response in 15%-62% of patients, and improved survival[227-233]. Some studies have suggested that complete tumor ischemia may stimulate angiogenesis, resulting in an increased susceptibility to tumor growth rather than suppression. It has been therefore suggested to maintain arterial patency both to prevent this pro-angiogenic effect and to permit repeated treatments[234-237]. Side effects associated with both DEB-TACE and TACE include nausea, vomiting and right upper quadrant pain (post-embolization syndrome), doxorubicin-related cardiac toxicity, bone marrow aplasia, hepatic abscesses, cholecystitis[229,231,238]. Two randomized controlled trials demonstrated improved side effect profiles[239,240] with equivalent survival rates[235,239] and longer time to progression for DEB-TACE in comparison with conventional TACE[240]. A meta-analysis showed comparable tumor response rates[241].

Radioembolization is a modestly invasive, fluoroscopically guided and microcatheter-based technique, using either yttrium-90 (Y-90) embedded non-biodegradable glass microspheres (25 ± 10 μm) or Y-90 embedded non-biodegradable glass resin-based microspheres (29-35 μm). Radioembolization exploits the preferential arterial blood supply of HCC by delivering radiotherapy directly to the tumor and preserving the normal liver parenchyma. In the target lesion, Y-90 delivers tumoricidal doses of a pure high-energy beta emitter. Because of the short tissue penetration and half-life, Y-90 is an ideal radioisotope for intra-arterial radiotherapy. Patients who have intermediate/advanced BCLC stage HCC and who are not candidates for TACE due to portal vein invasion are ideally candidate to radioembolization with Y-90[242-244]. Radioembolization represents a suitable alternative to chemotherapy for patients with advanced HCC[245]. Moreover, Y-90 radioembolization can be proposed as a bridge to liver transplantation[139,246,247].

As for combinations therapies, one of the most studied approaches is represented by the association of RFA plus TACE. In fact, the decreased blood flow due to TACE reduces heat loss and improves the RFA margins. On the other hand, TACE enhances nearby control of satellite lesions[248]. Several meta-analyses have found that the combination of RFA and TACE is associated with improved survival in comparison with RFA alone, particularly for tumors larger than 3 cm in diameter[249-252]. Hyperthermia is able to potentiate the cytotoxic effect of radiation[253,254]. Additionally, in animal studies, the combined use of radiation and RFA resulted in improved tumor growth control compared with RFA alone[255,256]. The combination of thermal ablation with SBRT represents another encouraging option, even if more research is required to establish the most appropriate dosing and timing regimen[257].

Sorafenib is a small-molecule multikinase inhibitor, which blocks Raf kinase, VEGF receptor and platelet derived growth factor receptor (PDGFR). In two randomized, double-blinded, controlled, phase III clinical trials, the SHARP (Sorafenib HCC Assessment Randomized Protocol trial) and the Asia-Pacific (conducted in the Asia-Pacific region), sorafenib was associated with improved progression-free and overall survival in patients with advanced unresectable HCC[258] and currently represents a therapeutic option for patients who are not candidates for curative treatment or TACE.

CONCLUSION

HCC is a major global public health problem due to the rising incidence and high mortality in both developing and developed countries. An important point to be addressed is the promotion of preventive strategies, such as hepatitis B vaccination, and chronic hepatitis B and C treatment, in order to cut down the number of patients who may develop cirrhosis and potentially progress to HCC. Early diagnosis is crucial for curative treatments. As a consequence, patients at risk of developing HCC should be regularly followed up to diagnose HCC in the initial stage. Surgical resection, RFA and liver transplantation are considered the cornerstones of curative therapy, while for patients with more advanced HCC recommended options include sorafenib and TACE.

Unfortunately, most evidence comes from case series and retrospective studies. There is a need for larger, multicenter, randomized studies in order to define the most appropriate, evidence-based therapeutic approach to patients with HCC.