Summary

Hepatocellular carcinoma (HCC) is one of the commonest malignancies worldwide. The incidence of HCC is particularly high in Asia, where chronic hepatitis B virus (HBV) infection and cirrhosis of liver are commonly seen. The present article summarizes the recent practice on HCC screening, including its benefits and cost-effectiveness in Hong Kong. Various investigation modalities used for pre-treatment work-up of HCC are reviewed. The five most widely used therapeutic options include surgical resection, local ablation therapy, intra-arterial treatment, systemic chemo-immunotherapy and liver transplantation, each of them having its own advantages and disadvantages. Recent advances in surgical techniques, including radiofrequency ablation, new approaches to hepatectomy and laparoscopic liver resection could have resulted in a much better oncological and survival outcome.

摘要

肝細胞癌(HCC)是全球最常見的惡性腫瘤之一。因為亞洲地區慢性乙型肝炎和肝硬化很常見，所以HCC在亞洲發病率特別高。本文總結HCC普查的發展情況，包括香港實施中的益處和經濟效益；討論各種HCC治療前檢查的方法和模式。本文探討五種常用治療方案，包括：i)外科切除；ii)部分切除；iii)動脈內治療；iv)全身免疫化療；v)肝移植的優劣之處。外科技術如放射頻率部分切除術；肝切除術的新方法和腹腔鏡肝切除手術，已大大改善腫瘤學結果和病人的預後。

---

Introduction

Hepatocellular carcinoma, being the commonest primary liver cancer, has become the second leading cause of cancer death in Hong Kong and China. The incidence is also increasing in the Western countries. Cancer statistics have shown that at least 1 million new cases of HCC occur annually.¹ Mortality from the disease remains high, with median survival time being 24.8 months after resection, and 5.8 months for symptomatically treated patients.² There is a strikingly uneven geographic distribution of HCC worldwide. The highest incidence, in the range of 10-20 per 100,000 population, occurs in South-East Asia (Taiwan, Korea, Thailand, Hong Kong, Singapore, Malaysia and southern China). Western countries, South America and Australia have a much lower rate of around 1-3 per 100,000 population.³

An aetiological association between hepatitis B virus and HCC has been established. Chronic infection with HBV imparts a 200-fold increased risk of developing HCC.⁴ Eighty percent of cases of HCC worldwide are estimated to be associated with HBV infection. Epidemiological data also supports a causal relationship between hepatitis C virus and HCC. About 70-80% of HCC patients have anti-HCV antibodies in southern Europe.⁵ Individuals co-infected with HBV and HCV seem to have even higher risk of developing HCC.

The present article reviews the current status of HCC screening, update on pre-treatment investigation modalities, commonly applied therapeutic options, and advances in surgical treatment for this highly lethal disease.

Screening of HCC

According to the World Health Organization (WHO), 10 criteria (Wilson) have to be fulfilled before a screening programme is considered ideal.⁶ These include:

1. the condition should be an important health problem;
2. accepted treatments are available;
3. facilities for diagnosis and treatment are available;
4. the condition can be recognizable in the latent/early stage;
5. suitable tests for screening are available;
6. the screening tests are acceptable to the population;
7. the natural history of the condition is well understood;
8. agreed policy on whom to treat is available;
9. the cost of diagnosis and treatment should be economically balanced with the whole medical expenditure; and
10. case-finding should be continued.

With the exception of cost-effectiveness, which is still under debate, HCC satisfies nearly all these criteria for a good screening programme. This lethal disease has a high prevalence in endemic areas, with recognizable high risks groups (patients who are HBV or HCV carriers). Good screening tests are available (alpha-fetoprotein (AFP) and ultrasound) which are acceptable to the population. Diagnostic tests are also available (serological, radiological and histological) and curative treatment is possible especially when the condition is recognized at an early stage. However, whole-population screening for HCC is still not adopted for most governments including Hong Kong, as it is too expensive. A cost-effectiveness analysis on screening cirrhotic patients in the Western world has estimated that it will cost US$11,800-25,000 per tumour detected and US$26,000-113,000 per additional year of life gained.⁷ According to a study conducted in Hong Kong, the annual costs of detecting one HCC and one treatable HCC are US$1,167 and US$1,667, respectively,⁸ a figure much different from the Western calculation. Another cost-effectiveness analysis performed in Hong Kong has calculated the cost being HK$1.30 million for men and HK$3.56 million for women per life saved by 4-monthly ultrasonography and AFP screening.⁹

The combination of transabdominal ultrasonography (USG) and AFP remains the main screening method because they are easily assessable and non-invasive. Intrahepatic lesions as small as 1cm can be detected by USG, although differentiation from haemangiomas or regeneration nodules can be technically difficult. In a North American study, the sensitivity and specificity for USG is quoted to be 71.4% and 93.8%, respectively.¹⁰

Another commonly used screening tool for HCC is AFP, because it is easily available in most laboratories. One of the problems with tumour markers is to define its cut-off value for diagnosis. Most centres are using AFP of >500 ng/ml as a diagnostic level. Nevertheless, this is still unsatisfactory because early HCC patients can have a lower AFP level; whereas chronic hepatitis B or C patients with reactivation can sometimes reach a level of over 500ng/ml.¹¹ The higher the level of AFP set as cut-off value, the higher the specificity and the lower the sensitivity. After balancing sensitivity and specificity, an AFP level of 16 ng/ml was calculated to have the best discriminating power.¹² Most laboratories are using an AFP level of 20 ng/ml as the upper limit of normal, with a sensitivity and specificity of 60% and 90%, respectively.⁸

The optimal interval of surveillance is another practical issue for primary healthcare providers. Too short a screening interval is not cost effective. However, the interval should not be too long to allow the HCC to grow to such an extent that curative treatment is not possible. Theoretically, taking the growth rate of HCC into consideration, the optimal interval between screening is 4-6 months.⁸

The use of an intensive surveillance programme for hepatitis B carriers has been under investigation in Hong Kong.¹³ Patients with AFP>20ng/ml or with focal lesions identified by ultrasound were monitored more intensively using lipiodol CT scan followed by AFP and abdominal ultrasound every 3 months for 2 years and subsequently every 6 months. Liver biopsy was offered to patients with hypervascular lesions with corresponding lipiodol retention. The authors concluded that a high incidence of relatively small HCC may be detected using this intensive surveillance programme, which can be attributed to the detection of subclinical tumours by additional imaging techniques and more frequent surveillance of this high-risk group. However, the surgical resection rate was low, and they were unable to demonstrate a clinical benefit with early detection. So far, no prospective randomized control trial can confidently demonstrate that periodic screening for HCC is effective in reducing mortality.

Treatment modalities of HCC

Surgical resection remains the "gold standard" and the treatment of choice for patients with good liver reserve (see Table 1). However, only a small proportion of patients (10-30%) can be surgical candidates at the time of diagnosis.¹⁴ In the past, hepatectomy was a procedure with ultra-high risk; carrying up to 20% mortality rate in some public hospitals in Hong Kong around the mid-90's. This was mainly due to excessive blood loss and post-operative liver failure. Biliary complications, intra-abdominal sepsis and a big surgical wound also contributed to significant morbidity. After a liver resection with curative intent for HCC, the 5-year survival ranges from 26% to 50%,¹⁴ intra-hepatic tumour recurrence being the main cause of subsequent death. New advances in operative techniques and the use of the laparoscopic approach have proven benefits in the outcome of HCC patients in recent years (see below).

Liver transplantation is considered one of the most attractive treatment options for HCC. It removes both detectable and undetectable tumour nodules, together with the extra advantage of replacing the cirrhotic liver with a normal one, with a potential cure of its complications such as portal hypertension. The 5-year survival rate of liver transplantation for HCC has been shown to be better than resection, particularly in patients with a single tumour <5cm, or with multiple tumours <3 in number and each <3cm in diameter (Milan criteria).¹⁵ However, in Hong Kong, the persistent shortage of donor livers remains a major obstacle.

Local ablation therapy is a type of treatment modality in which damaging agents are introduced directly into the neoplastic tissues. This minimally invasive therapy, commonly used in the form of percutaneous ethanol injection (PEI) or radiofrequency ablation (RFA) (see below), can be performed in repeated sessions with few complications. This is particularly good for small HCC. RFA has gradually gained more interest compared to PEI in view of its better overall survival demonstrated in a randomized trial.¹⁶ Other less popular methods of local ablative therapy used previously were microwave, laser, cryotherapy and acetic acid injection.

Intra-arterial treatment is the procedure in which chemotherapeutic and/or embolizing agents are given via the hepatic artery (Transarterial chemo-embolization, TACE). Two randomized controlled trials had shown TACE can improve survival.^17,18 A similar approach, also proven to be safe and effective in HCC, is by administering therapeutic doses of radioisotopes through the hepatic artery and hence irradiating the tumour internally.¹⁹ This technique, also termed transarterial radioembolization (TARE), has been used with iodine131-lipiodol¹⁹ and yttrium-90 microspheres.²⁰ However, when there is thrombosis of the main or ipsilateral portal venous system, or presence of Child's C cirrhosis, intra-arterial treatment is contraindicated because of the increased risk of liver failure.

Systemic chemotherapy is usually considered for patients who are unsuitable for any of the above treatments, particularly when regional lymph nodes are involved or the presence of extrahepatic metastasis. Most systemic chemotherapy regimens used for treatment of HCC have disappointing results. The most widely used agent being adriamycin. None of them has been shown to achieve a consistent response rate of >20%.²¹ Nevertheless, there were reported cases of complete pathological response with systemic combination chemoimmunotherapy for inoperable HCC.²² The combination therapy used consisted of cisplatin, interferon-alpha, adriamycin and 5-fluorouracil (a regimen known as PIAF). The radiological response rate was 26%, and among those responded to treatment, 10-20% of unresectable HCC were down-staged to resectable ones.^22,23

Decisions for selecting therapies and pre-treatment evaluation

The choice of treatment for HCC depends on patient's factors, such as age, performance status and co-existing medical diseases. The functional capacity of the liver as a whole can be assessed by the Child-Pugh score (serum bilirubin, albumin, prothrombin time, encephalopathy and ascites). Patients with very poor liver function (e.g. Child's C) should not be subjected to a major liver resection. The choice of treatment also depends on the disease factors, such as the size and number of focal lesions, and their positions in relation to other major structures. Major vessels invasion and overt extrahepatic metastasis would also significantly affect the management choice. Patient's own preference should also be respected.

A proper pre-treatment work-up is essential before a particular therapy is decided upon. This would include baseline blood tests including liver and renal function tests, a complete blood picture, clotting profile and hepatitis serology. AFP is commonly used for diagnosis as well as for monitoring response to treatment. A general anaesthetic assessment for patients with co-morbidities, with particular attention to those with increased cardiovascular or respiratory risks, is essential before a major liver resection is contemplated.

Imaging techniques commonly used in diagnosing HCC include transabdominal ultrasound followed by a contrast triphasic computed tomography (CT). Blood flow to HCC is derived predominantly from the hepatic artery and tends to enhance during the arterial phase after contrast infusion. Therefore, HCCs show a typical hypervascular pattern, with clear-cut enhancement predominantly in the arterial phase, rapid wash-out in the portal venous phase, and hypoattenuating in the delayed phase. Large lesions may show a mosaic pattern with multinodularity within. Areas of necrosis and haemorrhage can also be defined. The addition of contrast agents such as intra-arterial carbon dioxide and helium microbubbles in ultrasonography also shows promise in improving accuracy.²⁴ Magnetic resonance imaging (MRI) has similar sensitivity and specificity when compared to triphasic CT scan. Positron emission tomography (PET) scan has limited use as a diagnostic tool for HCC, because well or moderately differentiated HCC may not generate a high level of metabolism requirement compared to that of surrounding tissues. It is reserved for the evaluation of extrahepatic spread. Hepatic angiogram (HAG) with lipiodol has been shown helpful in diagnosis, because of the highly vascular nature of the tumour. Dynamic liver function study such as indocyanine green (ICG) clearance is performed in many centres prior to a major liver resection. By measuring the percentage of ICG retention in the body 15 minutes after injection (R15), this has been claimed to be a reliable test in predicting mortality after hepatectomy.²⁵ CT volumetry, by calculating quantitatively the residual liver volume, has its value in predicting post-operative liver failure and is particularly helpful in cirrhotic patients. Liver biopsy is uncommonly performed nowadays, especially when surgery is planned, as it carries a low but possible risk of bleeding and needle tract seeding. This need not be performed under circumstances in which the diagnosis of HCC is almost certain after clinical, biochemical and radiological evaluation.

Recent advances in surgical treatment of HCC

Three areas of technical advances regarding surgical treatment of HCC in the past decade have resulted in dramatic decline in the operative morbidity and mortality.

1.	Improvement in hepatectomy technique The mortality rate for liver resection 20 years ago was >10%. Nowadays, surgery is much safer, with an acceptable overall mortality rate of <5%.26 The Liver Cancer Study Group in Japan reported the largest series of resected HCC in 6785 cirrhotic patients between 1988 and 1999. The 1-, 3-, 5-, and 10-year survival rates were 85%, 64%, 45% and 2% respectively.27 To perform safe liver resections, specific problems are addressed in the following: Selection of cases by careful pre-operative assessment is essential, as discussed in the above section. Most centres will include a dynamic liver function study by ICG clearance test and/or liver volume measurement by a volumetric CT scan. The extent of resection has always been a difficult intra-operative decision for liver surgeons. On one hand, resection of all malignant tissues (including satellite nodules) with an adequate margin is tempting for effective clearance; on the other hand, leaving enough non-tumourous liver parenchyma is essential to avoid post-operative hepatic failure. The segmental approach of hepatectomy is based on the liver's anatomical description by Couinaud, which has been widely practiced in Europe. This can provide a safe and radical resection, allowing a larger remaining liver volume. Studies have shown that significantly better overall and disease free survival can be achieved by anatomical resections of small solitary HCC, compared to limited resections, without increase in operative risks.28 Another important advancement in liver surgery is the use of intra-operative ultrasound (IOUS). This can readily allow in-situ visualization of the liver anatomy, tumour location, as well as small tumour nodules that have escaped from pre-operative imagings. The segmental or subsegmental portal venous drainage areas of the segment containing the tumour are identified, allowing a safe anatomical resection to be carried out after careful determination of the transection plane. Limiting blood loss and transfusion are essential in all liver resections. Intermittent inflow occlusion with periods of 15 minutes clamping and 5 minutes unclamping can be well tolerated.29 On the other hand, inflow occlusion is abandoned by some liver surgeons, especially on cirrhotic livers, where ischaemia/reperfusion injury is poorly tolerated. Techniques for parenchymal transection are being improved throughout the years, aiming to achieve a precise transection plane, minimize tissue necrosis as well as careful identification of bile ducts and vessels. The ultrasonic dissector (Cavitron Ultrasonic Surgical Aspirator, CUSA) serves to divide and remove liver parenchyma, exposing vascular structures and bile ducts. The harmonic scalpel has a haemostatic effect, based on the principle of sealing blood vessels with denatured protein. The TissueLink dissecting sealer, based on a localized radiofrequency device, can coagulate up to 6-8mm thick liver tissue. This saline-cooled radiofrequency coagulation device can therefore effectively achieve intra-operative haemostasis and facilitates liver parenchymal transection. The importance of post-operative critical care and pain management should never be under-estimated. Admission to the intensive care unit for at least a day is recommended, especially after a major resection, for close monitoring of cardiovascular status, metabolic disturbance as well as early detection of post-operative haemorrhage. The use of inotropic agents has been significantly minimized nowadays with improved intra-operative haemostastic techniques. Management of pain by patient-controlled analgesia (PCA) is now frequently employed.
2.	Radiofrequency ablation Radiofrequency ablation (RFA) has gained much attention in recent years. With a tremendous expansion in its application for patients with liver tumours, RFA has become the most popular type of local ablation therapy. It involves the localized application of thermal energy to destroy tumour cells. Alternating electric current in the range of radiofrequency (RF) waves (460kHz) is applied from a RF generator through a needle electrode placed directly into the tumour.30 Ionic agitation from alternate current causes tissue coagulation through localized frictional heating. Subsequent tissue desiccation increases impedance, which eventually decreases current flow, leading to automatic "roll-off" of the cycle. It is most commonly performed by percutaneous route under local anaesthesia; but it can also be performed via laparoscopic approach or open surgery under general anaesthesia. Its advantage over percutaneous ethanol injection (PEI) has been demonstrated in a randomized controlled trial, in terms of significantly better overall survival.16 Generally, RFA is indicated in HCC patients where the tumour is unresectable, either due to poor liver reserve, or multifocality of the disease. Tumours less than 5cm are considered suitable for RFA, because complete coagulative necrosis may not be achievable in tumours bigger than 5cm. In such cases, multiple ablations with pre-procedural mapping for overlapping zones is recommended. RFA is particularly applicable in small HCC, where a high complete tumour necrosis rate of 90-100%, and a low local recurrence rate of 3.6% can be achieved.31 HCC located at the subcapsular region, perivascular tumours and centrally located lesions in close proximity to major bile ducts are considered less desirable for RFA. The procedure has also been applied to patients with intrahepatic HCC recurrence, when repeat resection is not favourable. It has also been reported to be effective in the acute management of ruptured HCC, with complete haemostasis and tumour ablation both achieved in one goal.32 While RFA for small HCC has been proved favourable in the short run, randomized trials comparing RFA to resection for long term results are still ongoing. A recently published retrospective study suggested that RFA may offer similar long term results to surgical resection for single nodule HCC, although the groups were not truly comparable.33 The risks of RFA should not be under-estimated. Mulier reported the mortality and morbidity rates of 3670 patients who received RFA to be 0.5% and 8.9%, respectively.34 Deaths were due to sepsis, liver failure, cardiac complications, peritoneal haemorrhage and bile duct stricture. Major complications included abdominal bleeding (1.6%), intra-abdominal sepsis (1.1%), bile duct injury (1%), liver failure (0.8%), pulmonary complications (0.8%), skin burn at ground pad site (0.6%), hepatic vascular injury (0.6%), visceral damage (0.5%), myoglobinuria (0.2%) and renal failure (0.1%). Patients with advanced cirrhosis34,35 and inexperience of the operator (<50 cases)35 were risk factors for developing complications. Careful patient selection and meticulous RFA techniques need to be emphasized in order to minimize complications.
3.	Laparoscopic liver resection More than 10 years have elapsed since Gagner reported the first laparoscopic hepatectomy for liver tumour.36 This technique is still in its developing stage, mainly related to technical difficulties connected to exposure of the liver and controlling haemorrhage. The lack of dedicated tools, the fear of gas embolism and the risk of tumour dissemination may have also slowed down the diffusion of this technique. Out of the relatively small series of experience, the conversion rate of around 7% was reported globally.37 The advantages of a minimally invasive approach to liver surgery are significant, especially in cirrhotic patients. Theoretically, this would include decreased postoperative pain, early mobilization and feeding, reduction in respiratory and thrombo-embolic complications, shorter hospital stay, earlier commencement of adjuvant therapy, reduction in intra-abdominal adhesions and also cosmetic advantages. Moreover, with less destruction of the abdominal wall porto-systemic collaterals, portal hypertension and ascites should be better controlled post-operatively.38 Data on short- and longer-term outcome of laparoscopic hepatectomy for HCC was provided by Shimada et al.39 Their study reported a better short-term outcome (in terms of hospital stay and complication rate) compared to conventional open hepatectomy; with similar overall and disease-free survival rate. Similar results were reported by Cherqui et al in 2003, with advantages in terms of morbidity, a same or better 3-year survival and a similar recurrence rate.40 Laparoscopic hepatectomy in cirrhotic patients has been considered a safe procedure provided careful selection criteria are followed. Subcapsular, small lesions (maximum diameter 4-5cm), located superficially in the left lateral segments (segment II, III), segment IVb, or right inferior segments (segments V, VI), on a well-compensated cirrhotic patient (Child's A) constitute good indication for laparoscopic approach.38 Laparoscopic ultrasound, using flexible probes, is an indispensable tool to locate the tumour intra-operatively, and to study its relationships with regard to major vessels and bile ducts. The use of ultrasound scissors and blades (Ultracision/ harmonic scalpel), endo-stapler and fibrin glue has also resulted in shorter operative time, better haemostasis and reduction in bile leaks. Nevertheless, laparoscopic hepatectomy should be performed in highly specialized hepatobiliary centres and by surgeons with extensive experience.

Conclusion

Although HCC is a common and usually fatal illness, whole population screening is not an option because this would be too expensive. Theoretically, the cost-effectiveness should increase in areas with high prevalence of HCC, especially when this is limited to high risks patients, such as hepatitis B or C carriers. Therefore, screening for this group of individuals who can afford to pay is justified. Transabdominal ultrasound together with AFP (using 16 ng/ml as upper limit of normal) every 4-6 months remain the most popular means of screening, despite the fact that their sensitivities are not satisfactory. Better tumour markers are therefore in need for screening of HCC.

The five well recognized treatment modalities for HCC include surgical resection, transplantation, local ablation, transarterial treatment and systemic chemo-immunotherapy, each has its own advantages and disadvantages. Selection for the most suitable therapeutic option depends on patient's factors as well as disease factors (Figure 1). Careful pre-treatment evaluation is necessary in order to improve survival as well as oncological outcome. This would at least include careful assessment of the liver function by Child-Pugh score and dynamically by ICG study. Transabdominal ultrasound followed by a good quality triphasic contrast CT scan are necessary for accurate tumour localization and examination of the disease extent.

Recently, the overall outcome for hepatectomy has significantly improved, with a mortality rate dropped to <5% in many centres. Radiofrequency ablation has gained tremendous popularity, especially for small HCC. It has a wide range of applications and has proven advantages over percutaneous ethanol injection, in terms of survival. Nevertheless, meticulous RFA techniques should be emphasized in order to minimize possible complications. Laparoscopic liver resection for HCC is technically demanding but feasible, with desirable outcomes, provided that careful selection criteria are followed. It should be performed by experienced surgeons in specialized centres.

Key messages

1. Hepatocellular carcinoma (HCC) is the second leading cause of cancer death in Hong Kong. It has aetiological association with hepatitis B and hepatitis C infection.


2. Whole population screening would be too expensive. Screening for high risk individuals who can afford to pay is justified, by 6-monthly transabdominal ultrasound and AFP.


3. The 5 recognized treatment modalities for HCC are surgical resection, local ablation, transarterial treatment, systemic chemo-immunotherapy and liver transplantation.


4. Decisions for selecting therapies for HCC should take into consideration the patient's factors, tumour's factors and expertise of the centre.


5. Recent advances including new hepatectomy techniques, radiofrequency ablation and minimally invasive surgery have resulted in much better oncological and survival outcome for HCC patients.

---

---

References