Year XVII n.1/01

Introduction

The widespread adoption of ultrasound (US)-scan for screening high risk patients, has led to the identification of many patients with a small potentially treatable hepatocellular carcinoma (HCC) (1) (Figure 1 and 2). However, due to the lack of randomised controlled studies, it is not clear whether the mortality rate for HCC has been reduced in parallel. Such features of the natural history of HCC as the long-lasting subclinical incubation period of HCC that have been observed in many patients, as well as the large number of tumors which grow as a solitary mass seem to favour screening programs. Conversely, in many patients such aspects of HCC as multinodal onset of the tumor and great variations in the growth rates of single nodes, with doubling volume times from one to 20 months, may hinder the effectiveness of screening (2).

Population to be screened

As HCC is linked to environmental, dietary and life-style factors, epidemiological surveys were instrumental in identifying groups of individuals who are at risk for HCC. In two Consensus Conferences held in Anchorage (Alaska, USA) (3) and in Milan (Italy) (4) patients with cirrhosis, chronic carriers of hepatitis B surface antigen (HBsAg) and patients with rare metabolic liver diseases were identified as populations at risk for HCC. To improve cost-effectiveness of screening studies aimed at assessing individual’s cancer risk have been pursued. However, this is difficult because of individual variations in the metabolism of carcinogens, DNA repair capacities, genomic stability and inherited cancer predisposition.

Patients with cirrhosis

Four percent of all patients with cirrhosis due to chronic viral hepatitis or alcohol abuse develops HCC each year (1,5,6). HCC developed also in patients with chronic viral hepatitis without histologic evidence of cirrhosis, but at lower incidence rates. HCC developed in 110 Italian patients with cirrhosis out of a cohort of 417 followed since 1985. The annual rates of HCC development were 11% for the subgroup of men older than 53 yr with higher than 20ng/ml serum AFP levels. In a prospective study of 208 patients with compensated cirrhosis, HCC developed more often in patients with high liver cell proliferation than in those with low proliferation indices (annual rates: 5% vs 1%) (Figure 3). Thus, cirrhotic patients with exceptionally high risk of developing HCC can be identified to be enrolled in more cost-effective programs of HCC screening.

Chronic carriers of HBsAg

More than 250 million persons are persistently infected with HBV, world-wide (19). Epidemiological, clinical and experimental studies have established a strong link between chronic infection with this virus and HCC. HBV is, in fact, responsible for both genotoxic lesions of the liver cells and tumor promotion through increased liver cell proliferation associated with persisting hepatitis and does not necessarily require the step to cirrhosis to be oncogenic (20). In a prospective cohort study in Taiwan, the risk of HCC in 3,454 HBsAg carriers was 102 times greater than in 19,253 non carriers (14). However, the carriers at especially high risk for developing HCC were those with actively replicating HBV (HBeAg+/HBV-DNA+) and those with cirrhosis. Healthy carriers may develop HCC, but are at substantially lower risk (13). Often distinguishing between clinical subsets of HBsAg carriers may be difficult, unless the patients’ are periodically assessed with laboratory or histological investigations. The strong link between HBV and HCC has been further confirmed by the decrease of HCC among Taiwanese children that has been observed since the start of mass vaccination of all newborns against HBV (21).

Patients with rare metabolic diseases

Patients with porphyria cutanea tarda, genetic hemochromatosis, alpha-1-anti-trypsin deficiency, tyrosinemia and hypercitrullinemia are also at high risk for HCC. Patients with glycogenosis type I and III, Wilson’s disease and hereditary fructose intolerance may also develop HCC, but are at substantially lower risk (2). HCC has developed also in patients with primary biliary cirrhosis, probably reflecting treatment-related improvement in patient’s survival. HCC was also found in 64% of 160 Japanese patients and in 48% of 101 South African Blacks with Budd-Chiari syndrome (22,23).

Screening tests

Screening tests must be simple, safe, acceptable to the patients and effective. Both serum alphafetoprotein (AFP) and abdominal ultrasound (US) fulfill these requisites, though they differ in terms of sensitivity and cost. The normal range for serum AFP is 0 to 20 ng/ml for the healthy adults. Serum levels of 400 ng/ml are very suggestive of HCC. However, two thirds of patients with small HCCs have less than 200 ng/ml and more than 30% of patients with HCC do not have abnormal circulating levels of AFP even in advanced stages (1,6,8). Moreover, between the range of 20 and 200 ng/ml, patients with chronic liver disease and false positive results due to hepatitis flares outnumber those with HCC. In patients with borderline elevations of serum AFP, the microheterogeneity of the sugar component of AFP can be assessed by lectin affinity electrophoresis coupled with antibody-affinity blotting. In a prospective study of 361 cirrhotics, 33 showed elevated atypical AFP 3 to 18 months before HCC was detected by imaging techniques (24). To minimise the false results of AFP determinations, cirrhotics are being surveilled by means of real-time US.

As a general rule, a lesion seen as a discrete node in the liver should be presumed to be a preneoplastic lesion or HCC, and should be investigated accordingly. Most HCCs can be detected as a hypo-dysechoic mass (1,6,8).

Screening strategies

Screening strategies are adopted according to the local epidemiological and socioeconomic situation. In population-based studies, most individuals are asymptomatic, and therefore the screening intervals can be easily standardised. Serum AFP is the screening method of choice because these studies include thousands of persons and relatively few patients with liver disease at risk of false positive results with AFP. Because of the high risk of false results, AFP is not appropriate for screening cirrhotics. Abdominal US is the screening and surveillance method of choice. In fact, in 33-60% of the cirrhotics studied by US, HCC would have gone undetected if the patients had been screened by AFP only (1,6,24). The 6-month interval was thought to be the convenient screening interval for cirrhotic patients whose risk of developing HCC is approximately 30 times greater than that of healthy carriers of HBsAg.

Selection of patients to treat

Treatment options have been selected according to the presence or absence of cirrhosis, number and size of tumors, and degree of hepatic deterioration. Staging is a crucial variable in treatment outcome, since many therapeutic failures have resulted from incorrect patient selection. When patients are scanned by biphasic spiral computed tomography (CT) during the arterial phase highly vascularised tumors appear against a background of relatively unhenanced liver that is primarely enhanced during the late portal vein phase (25). In a multicenter study in Milan (26), 43% of 178 patients with early detected cancer had multinodular HCCs. However, most tumors smaller than two cm are hypovascular and therefore escape detection with such gold standard staging techniques as biphasic spiral-CT. For staging clinical status, the Child-Pugh scoring system provides accurate estimates of patient survival. The three year survival of untreated patients with a small tumor and well-compensated cirrhosis was approximately 25% (27).

Patients with normal livers

Hepatic resection is the primary option for patients with HCC who present with normal livers and well-preserved hepatic function. In two studies (28,29), the cumulative five-year survival for 128 such patients treated in 2 Centers with hepatic resection was approximately 45%.

Patients with cirrhosis and a small tumor

The functional capacity of the liver not involved by HCC is the major factor in these patients’ prognosis. Thus, hepatic resection was substantially successful in patients with small tumors and well preserved hepatic function. More recently, in the wake of substantial improvements in transplantation technology and better understanding of the natural history of HCC, liver transplantation has gained more popularity as the real curative treatment for patients with HCC and cirrhosis.

Orthotopic Liver Transplantation (OLT)

Liver transplant eliminates both detectable and undetectable tumor nodes and all the preneoplastic lesions in the cirrhotic liver. Removal of the diseased liver also reduces the risk of morbidity and mortality from portal hypertension. Opposing these “pros” for OLT are several important “cons”, for example, shortage of donated organs, high costs of the procedure, stringent criteria for selection of patients, high risk of early tumor recurrence due to faulty staging of the disease and immunosuppression, and recurrence of hepatitis. One major obstacle to the interpretation of OLT results are the large differences between transplantation centers in terms of time-lag between candidacy and operation. Between January 1988 and June 1994, the five-year survival of 537 patients with HCC (7% of total) who were given transplants in 82 European Centers was 39%. This included the 54.5% survival of 176 patients in whom cirrhosis was the primary indication for OLT and 45.5% of 361 cirrhotics in whom HCC was the primary indication. In Milan, 48 consecutive patients with viral cirrhosis and a single < 5 cm tumor or fewer than 3 < 3 cm nodes were treated by OLT (30). The 4-year actuarial survival was 92% for the 35 patients who were confirmed at the operation as having met the selection criteria, compared to 60% for the 13 patients who did not, because they were found to have ancillary nodes. Although survival of transplanted patients seems to be largely influenced by tumor size and number, there is no general agreement on the ideal tumor size that entails the least risk of recurrence, mostly because small tumor volume does not mean an early biological stage for all cases. Indeed, vascular invasion by the tumor and perihepatic lymph nodes can occur even in patients with small HCCs (28, 29). Unfortunately, vascular invasion can be assessed only during the operation and lymph nodes can be precisely assessed only during laparoscopy or laparotomy. The outcome of OLT may be influenced by recurrence of viral hepatitis, since infection of the graft may facilitate rejection and re-establish the oncogenic potential of the liver. The efficacy of interferon against hepatitis C is under evaluation. For hepatitis B, hyperimmune gamma-globulins and the nucleoside analogue Lamivudine are protective but costly (31,32).

Hepatic resection

Liver transplantation cannot be offered to all patients with cirrhosis who are found to harbour a small HCC. Thus, in many countries hepatic resection remains the primary therapeutic option for these patients. Since the functional capacity of the remaining liver is a major factor affecting prognosis for patients undergoing hepatic resection, limited hepatic resections (segmentectomy and subsegmentectomy) are the technical procedures of choice. Since 1983 the widespread adoption of intraoperative US has changed the outlook of this treatment. The best results in terms of both short-term and long-term survival were for patients with single tumors less than 2 cm in diameter and well preserved hepatic function. In 347 Japanese patients, the 5-year survival rate was as high as 60.5% (33), with very low mortality rates (0-5%). The 3-year cumulative survival was 50% for 78 Japanese patients with single tumors and Child’s A status, 35% for 26 with Child’s B status and 0 for three with Child’s C status (34). For 72 European patients, these figures were 51% for Child’s A patients and 12% for Child’s B-C (35). In patients with Child-Pugh A cirrhosis, portal hypertension is the most reliable predictor of survival after resection. None of the 14 operated patients with less than 10 mmHg hepatic venous pressure gradient had had unresolved hepatic decompensation compared to 11 of the 15 patients with higher gradients (36). Thus, resection is definitely contraindicated for patients with deteriorated cirrhosis or severe portal hypertension in view of the high operative risk and short life expectancy. In Barcelona, the intention-to-treat analysis of patients treated by transplantation or resection demonstrated better survivals for resected patients with compensated liver and mild portal hypertension compared to similar patients who were transplanted in the last years when the waiting list rose from 62 days to 162 days, on average (37).

Patients with advanced disease

Patients could be refused surgery because of advanced age, deteriorated liver function, large tumors, tumors localised in strategic positions or associated clinical conditions which contraindicate surgery.

Percutaneous interstitial treatments

US-guided interstitial treatments include tumor injection with absolute ethanol, 50% acetic acid or hot saline, or tumor thermoablation with radiofrequency, microwaves or laser. Most treatments were carried out with intratumor ethanol injection (PEI), which causes extensive coagulative necrosis of the tumor cells, and thrombosis of the tumor vessels, and is well tolerated (Figure 4 and 5). Once more, survival was largely influenced by liver function, size and number of tumors. The 5-year survival of 293 Italian patients with Child’s A cirrhosis and a small HCC was 47%, compared to 29% for 149 with Child’s B cirrhosis (38). Life expectancy of Child’s A patients with a small tumor treated by PEI appeared to be as good as that of similar patients treated with hepatic resection, and associated with a low risk of severe complications (1.7%) and mortality (0.1%). In 60 randomly selected patients with tumors smaller than 3 cm and compensated cirrhosis, injection with 50% acetic acid was superior to ethanol in terms of 2-year cancer-free survival (92% vs 63%, P=0.02), being particularly active in patients with hypervascular tumors (39). Radiofrequency thermoablation is safe and convenient in patients with compensated cirrhosis and a small HCC. In a prospective randomised study of 86 patients with compensated cirrhosis and small HCC, radiofrequency was superior to PEI in terms of complete tumor necrosis (90% vs 80%), and numbers of treatments (1.2 vs 4.8), but it caused more complications (9.5% vs 0) (40).

Transcatheter Arterial

Chemo-Embolization (TACE)

Transcatheter arterial embolization of HCC is possible because the liver has a dual blood supply while HCC is supplied virtually only from the hepatic artery. TACE through the femoral artery leads to ischemic necrosis of the tumor and makes hepatic arterial injection of antitumor agents possible, giving higher local concentrations of drugs with fewer systemic side effects. TACE of the proximal hepatic artery (conventional TACE) has been widely employed in Eastern and Western countries as an alternative to hepatic resection and has now been improved, as segmental or subsegmental TACE. The procedure is contraindicated for patients with venous tumor supply (hypovascular tumors), advanced liver deterioration, complete thrombosis of the portal vein trunk, renal failure or extrahepatic metastases. In the last decade, three randomised controlled studies of TACE and one randomised controlled study of transarterial embolization (TAE, without chemotherapy) treatment of patients with unresectable HCC have been conducted (41-44) Conversely, non-controlled studies of segmental TACE in 63 Japanese patients with Child-Pugh A cirrhosis and small HCC reported four-year survivals comparable to those for similar patients treated with resection or PEI (45). Thus, TACE needs to be reassessed for treating patients with compensated cirrhosis and a small vascularised HCC.

Other treatments

Systemic chemotherapy has been widely used to treat inoperable HCC, but the response rate is very low (20%) (46). The possible sex hormone-dependence of HCC and the presence of tumor hormone receptors have suggested a potential for hormonal manipulation of tumor growth, particularly using anti-estrogens. However, in two large studies of 120 and 477 patients with inoperable HCC, but less deteriorated liver functions, treatment with the anti-estrogen tamoxifen did not improve survival or the quality of life, compared to controls (47,48). In a prospective randomised study of 58 patients with advanced HCC, treatment with subcutaneous octreotide 250 mcg twice daily significantly improved survival at 12 months (49). However, the sample size of the study was not adequate for preventing a type II error in the assessment of the difference between treatment and no-treatment. The conventional method of external irradiation is not effective against HCC. Using three-dimensional radiation planning (conformal radiotherapy) the beam scatter can be minimised to deliver the therapeutic dose, making selective irradiation of the liver possible. Local radiation was performed safely in patients with Child A cirrhosis and smaller than 8 cm tumor, with a partial response rate of 64% (50). With proton radiation therapy, a large amount of radiation is focused only on the lesion, and the exposure of surrounding non-tumoral liver can be limited. Of 83 patients so treated only 19% had a complete response without any appreciable effect on survival. In the near future, the gene therapy may offer the chance to cure many patients suffering from Hepatocarcinoma. The principle consists in transfecting the neoplastic cells with a virus capable of introducing a gene which facilitates the suicide of the cell, or therapeutic genes which can make the neoplastic hepatocytes responsive to antiviral drugs.

Conclusion

Early diagnosis of Hepatocarcinoma is possible, but we do not know whether screening reduces the mortality rate for this tumour, in that we lack checked clinical research works. Without doubt, screening can offer advantages, such as an improvement in the therapeutic response by patients who have benefited from an early diagnosis, for some of whom radical treatments can be applied, as well as the reassurance of patients with negative diagnostic tests. All this involves a saving in resources. However, screening can also involve disadvantages, such as inflicting a longer “morbility” on patients with a prognosis which cannot be influenced by treatment, giving false reassurance to patients with untruly negative results and unnecessary morbility to patients with untruly positive results. These disadvantages involve an increase in costs. Patients with cirrhosis or congenital metabolic diseases involving a risk of Hepatocarcinoma need to be checked more frequently (at six-month intervals) than HBsAg carriers with no signs of hepatic illness (at twelve-month intervals). Surgical resection is the selected treatment in the few patients suffering from tumour in a “healthy” liver. Orthotopic liver transplantation continues to be the best solution for patients with cirrhosis and a tumour smaller than 5 cm in size. However, owing to the lack in organ donors, many centres continue to regard surgical resection as the first “therapeutic” option. Palliative treatments, such as percutaneous injection of alcohol and radiofrequency, are indicated for patients with a well-balanced liver suffering from a neoplastic disease which cannot be surgically treated. In many patients, the most effective approach is given by a combined resort to these treatments, that is targeted ablation therapies and resectional surgery or transplantation.

Fig. 1

Fig. 2

Fig. 4

Fig. 5