Hepatocellular Carcinoma

• Hepatocellular carcinoma (HCC) is fairly uncommon in Europe and North America but worldwide it is the commonest cause of death from malignancy with an estimated one million cases a year. In Europe there is 30 times the chance that a liver tumour will be secondary rather than primary.¹
• In the UK there are about 1,500 deaths per year from the disease.²
• 90 to 95% of patients have cirrhosis.²
• Most occur in Asia and sub-Sahara Africa where the incidence can be as high as 1 in 1,000.

Causes

• The prevalence generally follows that of hepatitis B³ that affects 400 million people worldwide.
• Hepatitis C affects 170 million worldwide. It also increases the risk of carcinoma⁴ and there is a lifetime risk of about 5% that carcinoma will develop, usually around 30 years after infection.
• Co-infection with both hepatitis B and C is usually said to increase the risk still further⁵ but this is disputed.⁴
• Of the various causes of cirrhosis, hepatocellular carcinoma is most often associated with haemochromatosis, alpha-1-antitrypsin deficiency, tyrosinaemia, chronic active hepatitis, hepatitis B or C and alcoholic cirrhosis. It is less common in primary biliary cirrhosis and is rare in Wilson's disease.
• The 5-year cumulative risk in hepatitis B is 15% in high endemic areas and 10% in the West.⁵
• There is a 5-year cumulative incidence of 21% in hereditary haemochromatosis.⁵ The risk also rises where liver storage of iron is excessive as in thalassaemia.
• In alcoholic cirrhosis, the 5-year cumulative risk is 8% and in subjects with advanced biliary cirrhosis it is 4%.⁵
• It may also be associated with a high concentration of aflatoxins in food. These tend to occur in Africa and the Philippines. Aflatoxins from Aspergillus flavus damage DNA and cause mutations of the p53 gene.
• Contraceptive steroids used for over 8 years may increase the risk up to 4-fold, as may the use of anabolic steroids but the absolute risk is very small.
• In 25% of cases there is no predisposing factor. In Western societies there is a history of alcohol abuse in half of cases and as many as 50% of alcoholics may have subclinical tumours at autopsy. The risk appears greater in those who stop drinking but this may be because those who continue to drink die before the cancer develops.

The age of onset is usually in the 7th decade in Western societies but in endemic areas it is most often in the 4th or 5th decade. In those societies the male preponderance is about 8:1 but in Western societies it is around 3:1. The British Society of Gastroenterologists (BSG) believes that the disease will become more frequent in the future, largely because of hepatitis C,² but unless the habits of many young adults are rectified, there may also be a significant increase due to a steep rise in alcoholic cirrhosis in both men and women.

75% of tumours are multifocal in origin. Metastasis is a late event and occurs to lung, portal vein, periportal nodes, bones, or brain.

Presentation is usually with symptoms of advancing cirrhosis and liver failure.

• Pruritis
• Splenomegaly
• Bleeding oesophageal varices
• Cachexia
• Hepatic encephalopathy
• Abdominal distension from ascites
• Right upper quadrant pain is less common

• Jaundice
• Hepatomegaly may be stony hard, with an arterial bruit
• Ascites
• Spider naevi
• Peripheral oedema
• Periumbilical collateral veins

• LFT, FBC, INR, Calcium and glucose
• Alpha-fetoprotein is elevated in 75% and the higher the level the worse the prognosis
• Ultrasound, CT scan, MRI scan
• Chest x-ray may show a raised right hemidiaphragm or lung metastases
• Percutaneous liver biopsy
• Laparoscopy to inspect the liver and direct biopsy

• Cirrhosis
• Cholangio-carcinoma
• Primary lymphoma of the liver
• Metastatic carcinoma (30 times as common in Europe as HCC)

Histology varies from well-differentiated to anaplastic but this is not considered in staging as it does not affect prognosis. A unified system of staging is desirable as this enables meaningful comparisons to be made between trials. It is also important that the system used should accurately reflect prognosis. A number of different systems exist and there seems little uniformity.⁶

The Okuda system⁷ is often mentioned but dates back to 1985. This has been largely superseded by staging based on TMN classification.

• T1 single tumour of 2cms diameter or less with no vascular invasion
• T2 single tumour of less than 2 cms but with vascular invasion or multiple tumours of 2cms or less in one lobe only without vascular invasion or a single tumour over 2cms without vascular invasion
• T3 solitary tumour over 2cms with vascular invasion or multiple tumours in one lobe only with vascular invasion
• T4 multiple tumours in more than one lobe or involvement of a major branch of the portal or hepatic vein
• N0 no nodes involved
• N1 regional nodes only
• M0 no distant metastasis
• M1 metastasis outside the liver.
• T,N or M may have a suffice of x to indicate that it was not possible to assess that parameter

From this is derived the staging:

Before treatment of the primary tumour, complications must be treated. These include ascites tapping, treating encephalopathy or spontaneous bacterial peritonitis and banding varices.

Only 25% of tumours or less are amenable to surgery. The rest are too advanced for resection at the time of diagnosis or there are other factors that mitigate against surgery such as liver failure. Overall the 5 years survival after resection is only 40%. A systematic review of controlled trials¹² found that of treatments other than surgery, current treatments are ineffective or minimally and uncertainly effective. Embolization, tamoxifen and IFN are minimally or uncertainly effective but can deserve further assessment by larger and methodologically more sound randomized trials. The only realistic option for cure of the disease is surgery for small tumours.²

A study from Japan¹³ found that only 70 of 1,172 referrals (6%) could be classified as early but of those patients, the 5 years survival was 93%.

• Chemotherapy is usually with doxorubicin, fluouracil or cisplatin but response rates are under 10%.
• Chemoembolisation involves inserting emboli loaded with anti-mitotic drugs into the hepatic artery. The tumour area is the target and the rest of the liver is spared as it depends primarily on the portal vein for its blood supply. Mitomycin, doxorubicin and cisplatin are used and response rates may be as high as 60 to 80%.
• Liver transplantation for resectable or unresectable tumours even with cirrhosis may be feasible for small lesions without metastasis. Operative mortality is often 10 to 20%. Metastasis after transplantation occurs in 30 to 40% of patients and after transplantation, 5-year survival is less than 20%. Limited supply of organs and long waits make this an unrealistic option for most patients. Transplantation has the potential to treat both the tumour and possible end-stage liver disease from cirrhosis.
• Injection of ethanol or acetic acid, heat from radiofrequency, microwave, or laser ablation, or cryoablation with liquid nitrogen may be used on tumours less than 4 or 5 cm in diameter.
• Radiotherapy is limited by radiation hepatitis.
• Gene therapy may even have a place in the future.¹⁴

Prognosis is generally poor except in the rare early cases. In advanced cases, life expectancy is a few months.³ Even in those amenable to surgery, cirrhosis has an adverse effect, reducing survival at 5 years from 68 to 44%, at 7 years from 57 to 32% and at 9 years from 57 to 22%.¹³

There are data as to survival in untreated patients with HCC, showing that the natural history is highly variable.¹⁵ The major factors influencing overall survival are severity of underlying liver dysfunction and tumour size at initial detection. The state of liver dysfunction is graded by the Child-Pugh system of A to C. This classification is explained in the article on cirrhosis. Between 50% and 90% of patients with Child-Pugh A cirrhosis will survive a year untreated compared with only 20% with Child- Pugh C. Small HCCs at presentation have relatively long tumour doubling times, and overall survival with tumours of less than 5 cm was 81% at one year and 17% at three years with no therapy. This suggests that if earlier diagnosis can be made, the opportunity for intervention may be greater.

As the prognosis is so good in early disease¹⁶ and so bad at the later stages when diagnosis is usually made, it seems an obvious candidate for screening those at risk. Screening patients with chronic viral hepatitis and cirrhosis has resulted in detection of smaller lesions in the liver. Some of these are small carcinomas, but many are not malignant or are potential precursors to malignancy. Neither radiology nor biopsy can confidently distinguish between these possibilities when the lesions are smaller than about 1.5 cm in diameter. Therefore, diagnostic algorithms must be developed to assist in investigating these small lesions and deciding when they are truly malignant, when treatment is necessary, or whether a watchful waiting course is appropriate.¹⁷

The result of screening has been very disappointing but the BSG makes the following recommendations:

• Surveillance using abdominal ultrasound and α-fetoprotein (AFP) estimation can detect HCC of a smaller size than those presenting without screening.
• The only potentially curative therapies depend on detection of small HCC.
• Despite the above, there are no data confirming that these advantages in detection of earlier lesions produce an improvement in long term survival or cost saving.
• Surveillance for hepatocellular carcinoma should be considered in the following high risk groups:
  • Males and females with established cirrhosis due to hepatitis B virus, particularly those with ongoing viral replication.
  • Males and females with established cirrhosis due to Hepatitis C.
  • Males and females with established cirrhosis due to hereditary haemochromatosis.
  • Males with alcohol related cirrhosis who are abstinent from alcohol or likely to comply with treatment.
  • Males with cirrhosis due to primary biliary cirrhosis.
  • The risk of HCC development in cirrhosis due to autoimmune hepatitis, primary sclerosing cholangitis in both sexes, and alcoholic and primary biliary cirrhosis in women is generally low. Non-cirrhotic HCCs do occur in viral cirrhosis but the absolute risk is low.
• If surveillance is offered, it should be six monthly abdominal ultrasound assessments in combination with serum AFP estimation.
• If surveillance is offered, patients should be aware of the implications of early diagnosis and the lack of proven survival benefit.

In the UK, up to 40% of cases present with HCC as the first indication of underlying liver disease, in distinction to countries such as Japan where 80% of HCC are detected at an asymptomatic stage by screening of patients with known cirrhosis.²

Prevention is the best approach. Widespread use of hepatitis B vaccine and prevention of hepatitis B will reduce the frequency of Hepatitis B³ and with it cirrhosis and hepatocellular carcinoma. It may be regarded as the first anti-cancer vaccine. There is no vaccine against hepatitis C but treatments including interferon alpha may have a beneficial effect.³ There is no animal vector of hepatitis B or C.

A more sensible approach to alcohol consumption would also be beneficial.