Liver cancer: screening, diagnosis and management

Introduction

Hepatocellular carcinoma (HCC) is the most common form of primary liver cancer and is a growing global health concern. It is the sixth most common cancer worldwide, with 905,700 new cases in 2020, and is the fourth most common cause of cancer-related death, responsible for 830,200 deaths each year​​[1,2]​. 

The incidence of HCC is rising: new cases have increased by 160% over the past three decades in the UK alone, with around 17 new cases diagnosed per day​​[3,4]​. The average age at diagnosis is 69 years and the majority of cases occur in men​​[4]​. Prognosis remains poor, with five-year survival rates less than 20%; this is unchanged in the past decade despite advances in therapeutics​​[5]​. Considering the rise in cases and the potential burden of the condition for patients, it is important that pharmacists are aware of the signs of HCC and know how to appropriately refer.

This article will outline the importance of surveillance and how HCC is diagnosed, and will provide an overview of the current available treatment options, including the involvement of pharmacy teams where appropriate. 

Epidemiology

The epidemiology of HCC varies by region. Globally, viral hepatitis remains the most common cause of HCC​​[6]​​. Hepatitis B virus (HBV) is directly oncogenic and is the most significant risk factor for developing HCC — it is responsible for 50% of HCC cases​​[6]​. HBV is the main causative factor of HCC in East Asia, which has the highest HCC incidence in the world​[6]​. Hepatitis C virus (HCV) remains the most common aetiology of HCC in Western Europe, North America and Japan​[7]​. However, the attributable risk of HCV has decreased since the advent of antiviral medication, which results in sustained virological response (undetectable viral load six months after completion of treatment)​​[8]​. 

In the UK and United States, alcohol-related liver cirrhosis and non-alcoholic steatohepatitis (NASH) are significant causative factors for HCC. The incidence of NASH is rising as metabolic syndrome becomes a common phenomenon and 20% of HCC cases have NASH as the attributable factor in the UK and United States​​[9]​. 

Increasing age, male sex and smoking are also independent risk factors for HCC​​[1,10]​. Other, rarer aetiologies include alpha 1 antitrypsin disorder, haemochromatosis and primary biliary cholangitis​​[3]​.

Surveillance

Owing to poor prognosis and low surveillance uptake, HCC presents unique challenges to the NHS because half of UK cases are considered preventable​​[11]​. Around 90% of HCC cases have a background of cirrhosis or chronic liver disease, identifying individuals with these conditions as target ‘at risk’ populations for receipt of a surveillance test. In HCC, potential benefits of early detection are that curative therapies, such as resection or transplantation, could be offered sooner, resulting in improved survival rates​​[12]​.

Biannual ultrasound and alpha-fetoprotein (AFP) measurement are the diagnostic tests used for surveillance; they are widely available, affordable and have evidence supporting their utility. Research involving cohort studies and mathematical modelling conclude that HCC surveillance results in earlier diagnosis, receipt of curative treatments and improved survival​​[12]​​. Results from several studies have showed that HCC surveillance is cost-effective, although there is ongoing debate as to whether this is consistent across all aetiologies of CLD​​[13,14]​​. While recommendations regarding inclusion of AFP as part of HCC surveillance varies between guidelines, there is consensus among all professional bodies that ultrasound is required for HCC surveillance​​[15–17]​​. In the UK, biannual ultrasound is recommended, following European Association for the Study of the Liver guidelines (see Table 1​​[15–17]​​). 

There are significant challenges and controversies regarding HCC surveillance. Implementation and uptake of surveillance is suboptimal, with rates of surveillance uptake of less than 20% in the United States​[18]​. 

Low uptake is compounded by poor ultrasound sensitivity, which is 63% for lesions <2cm​​[19]​. The harms of surveillance relate to false positive or indeterminate results and subsequent tests, including further imaging, which confers radiation risks, or biopsy, which is highly invasive. In one study, physical harm (defined as any unnecessary test or procedure arising from a false positive or indeterminate ultrasounds) was experienced by 22.8% of patients​​[20]​. While most of these patients were classified as experiencing ‘mild’ or ‘moderate’ harm from repeated imaging, a small number underwent more invasive investigation including biopsy or angiography. 

Diagnosis

Clinical presentation

The clinical presentation of HCC can vary greatly. Patients with early-stage HCC are frequently asymptomatic, with radiological diagnosis made during image surveillance. In late-stage disease, given the majority of HCC occurs in chronic liver disease, patients often present with evidence of decompensated liver disease, including jaundice, encephalopathy, ascites and variceal bleeding. Patients may also experience right upper quadrant pain and weight loss, in particular with evidence of sarcopenia (i.e. loss of muscle mass)​​[3]​​. 

Radiological diagnosis

Radiology is the backbone of HCC diagnosis, exploiting the vascular properties of the liver. The liver receives dual blood supply from the portal vein and hepatic artery, paired with the bile duct as part of the portal triad (see Figure 1​[21]​). During carcinogenesis there is increased oxygen requirement as precancerous dysplastic nodules progress to HCC, leading to secretion of angiogenic growth factors, such as vascular endothelial growth factor (VEGF)​[22,23]​. This results in new unpaired arteries supplying HCC tissue, with progressive loss of portal venous supply (see Figure 2​[24]​).

Dynamic contrast-enhanced imaging

The differentiation in tumour and background liver blood supply can be utilised using dynamic contrast-enhanced imaging, with CT and MRI, the gold standard imaging modalities for HCC diagnosis​​[15,16]​. Although ultrasound is favourable for HCC surveillance, it has a low sensitivity and positive predictive value compared with contrast-enhanced imaging​​[25]​​. In some expert centres, contrast-enhanced ultrasound (CEUS) can be employed for HCC diagnosis, especially in cases of inconclusive CT/MRI​​[26]​. However, given the lack of high-quality prospective studies assessing CEUS and its inferior sensitivity for smaller HCC lesions, it is not the recommended first-line imaging modality for confirming diagnosis​​[15,16]​.

The hallmark radiological features of HCC are hyperenhancement during the hepatic arterial phase, followed by washout during the portal venous and delayed phases (see Figure 3​[27]​). For dynamic contrast-enhanced imaging, CT and MRI utilise iodine-based and gadolinium-based agents, respectively. Traditional extracellular gadolinium-based agents are exclusively renally excreted and not absorbed by hepatocytes, highlighting HCC arterial hyperenhancement and portal venous washout with liver parenchyma hyperenhancement, like CT imaging. However, hepatobiliary agents, such as godoxetic acid, are absorbed intracellularly by functioning hepatocytes, leading to enhancement of liver parenchyma. These hepatobiliary agents allow for further differentiation of malignant tissue from background liver during the hepatobiliary phase, 20 minutes after contrast injection (see Figure 4​[28]​), compared with extracellular gadolinium-based agents. Inclusion of hepatobiliary phase imaging has been shown to improve the detection of early and atypical HCC lesions compared with contrast-enhanced CT and conventional MRI​​[25,29,30]​.

The ‘Liver Imaging Report and Data System’ (LI-RADS) diagnostic algorithm has been developed to determine the probability that a given lesion is HCC​​[31]​. Table 2 and Table 3 show the major and ancillary CT/MRI imaging features in the LI-RADS algorithm.

Histological diagnosis

Owing to its distinct radiological features, HCC is the only solid organ tumour that does not mandate tissue confirmation for diagnosis​[17,32,33]​. Lesions displaying important hallmarks of HCC on imaging (i.e. arterial phase hyperenhancement, non-peripheral washout and capsule enhancement) is sufficient for non-invasive diagnosis. Pathological confirmation with liver biopsy is recommended in cases of non-cirrhotic HCC or where the diagnosis of HCC remains uncertain​[15–17]​. There is a perceived risk of complications caused by liver tumour biopsies, including bleeding and tumour seeding. Although tumour seeding rates vary between 1.8% to 4.0%, there is no reported evidence of an impact on patient survival​[34]​. As such, liver biopsy should not be precluded in cases of diagnostic uncertainty and should be encouraged to improve further molecular and genomic understanding of HCC​[35]​. 

Serum biomarkers

AFP is the only biomarker to fulfil the standardised biomarker validation​[36,37]​. AFP has been used as a biomarker in HCC surveillance. Although AFP shows a high specificity for HCC, its sensitivity is poor, missing more than half of early and small HCCs​[38]​. 

The addition of AFP to ultrasound for surveillance improves sensitivity for early HCC detection compared with ultrasound alone, but at the cost of a lower specificity​[39]​. As such, the use of AFP in surveillance and diagnosis of HCC is not actively endorsed in current practice, although it is still adopted in some centres​[15,16]​. AFP has a high utility in predicting survival and cancer recurrence post liver transplant​[40,41]​. Elevated AFP can therefore be used as selection criteria for liver transplant​[41–43]​. Other biomarkers, such as AFL-L3, an isoform of alpha-fetoprotein, and des-gamma-carboxy prothrombin (DCP), have been proposed in liver surveillance and prognostication but lack sufficient validation for clinical use​[44]​.

Staging and treatment strategies

Once HCC is diagnosed, as with other cancers, prognostic assessment and classification is required to enable selection of appropriate treatment options. In most patients with HCC, the coexistence of cancer and liver cirrhosis can complicate both the assessment of prognosis and the selection of treatments​[15]​. One of the most commonly used staging systems is the Barcelona-Clinic Liver Cancer (BCLC) staging system, which incorporates both prognostic and treatment dependent variables and has incorporated advances in management including the newer systemic therapies and ablation for very early cancers​[15,45]​. The modified BCLC staging system and treatment strategy, which is endorsed by the European Association for the Study of the Liver, is shown in Figure 5​[45]​.

Resection

Resection of solitary tumours has good results with 80–90% five-year survival; it remains the optimal treatment for larger solitary lesions or more peripheral lesions that are laparoscopically accessible and not amenable to interventional radiology treatments as targeting their blood supply is more technically challenging​[46,47]​.

In patients with non-cirrhotic HCC, complication rates can be low​[48]​. Resection of HCC in NAFLD/NASH is associated with increased postoperative complications, although mortality is relatively low and long-term survival may be higher than in resections performed in the context of chronic viral hepatitis​[49]​.

In cirrhotic livers there are three main limitations to consider for resection. The first is the functional status of the liver and whether enough parenchyma will remain postoperatively to enable good liver function. The second is the bleeding risk conferred by the presence of clinically relevant portal hypertension (CRPH). The third is the extent of the hepatectomy required and the degree of surgical invasiveness (open versus laparoscopic)​[15]​. 

These factors contribute to the risk of hepatic decompensation post-operatively and require individualised risk assessment. Multiparametric models incorporating the presence of portal hypertension, MELD score and extent of hepatectomy can predict both the likelihood of decompensation and liver-related mortality​[50]​. Consensus guidelines are that the ideal patient has a solitary tumour and preserved liver function in the absence of CRPH​[15]​. However, improvements in pre-, intra- and peri-operative care have led to experienced centres carrying out procedures in carefully selected patients that do not fulfil these criteria with good results​[15]​.

Transplant

The ‘Milan Criteria’ was developed in 1996 and defined a maximum criteria for liver transplant of one tumour less than 5cm or three tumours less than 3cm, and no evidence of microvascular invasion​[51]​. A small prospective study of 48 patients demonstrated that survival at one and five years was 84% and 74%, respectively, with a recurrence rate of 8%, and in the subgroup of patients that were shown definitively to be within Milan Criteria had four-year survival of 95%. The criteria from this small single-centre study have since been validated in larger cohorts with similar outcomes for survival​[52–54]​.

Locoregional therapies

Ablation

Aside from transplantation and resection, radiofrequency or microwave ablation is the third potentially curative option for HCC​[15,33,55]​. Unlike resection, ablation can be undertaken in patients with moderately decompensated liver disease as the risk of provoking further decompensation appears lower​[33,56]​. Ablation is recognised as providing equivalent survival at a lower cost than resection for small lesions <2cm; above 2cm recurrence rates are higher​[57–61]​. 

Guidelines suggest that ablation should be considered for small lesions, usually those less than 3cm, including multiple small, discrete lesions — traditionally three or less — or larger lesions up to 5cm on occasion​[15,33,55]​. It can be challenging to obtain a sufficient treatment margin for larger lesions and it is recognised that such treatment is unlikely to be curative owing to high rates of local recurrence. This may be owing to marginal microsatellites that are not ablated in large tumours with smaller margins, and these can cause early recurrence​[62,63]​. 

As well as radiofrequency and microwave ablation, there are other ablation techniques being investigated, including cryoablation, irreversible electroporation and laser ablation, but these are not yet in widespread use​[64]​.

Transarterial chemoembolisation, transarterial radioembolisation and stereotactic body radiotherapy 

Patients with preserved liver function and multi-nodular disease, those not suitable for ablation or resection for technical reasons (for example, size, difficulty accessing the tumour, or close proximity of the tumour to adjacent structures such as the vasculature or the gall bladder which may be damaged during an ablation), can be considered for alternative locoregional therapies​[15,33,55]​. Unlike ablation, these alternative therapies are rarely curative. These treatments are not suitable for those with more advanced cirrhosis — for example, those with a Child-Pugh score ≥B8 — as these treatments can provoke hepatic decompensation negating the benefit from reduced tumour burden​[15,29,36,55,64,65]​. 

The most widely used locoregional therapy, aside from ablation, is transarterial chemoembolisation (TACE). In conventional TACE, a lipiodol-based emulsion is prepared with an embolising agent — for example, a contrast agent — and suffused with a cytotoxic drug, most commonly doxorubicin, cisplatin or irinotecan. This preparation is then selectively injected into the arterial blood vessel supplying the tumour during angiography; exploiting the dual blood supply of the liver and targets the tumour while minimising collateral damage to the healthy liver tissue. The liver parenchyma receives blood from both the hepatic artery and portal vein; however, HCC are almost entirely supplied from the hepatic artery, hence arterial (chemo)embolisation will predominantly affect cancer tissue sparing healthy tissue with dual supply from the portal vein. Cancer tissue is degraded by both the loss of its blood supply, and by the local delivery of the cytotoxic agent. Specialist pharmacists are integral to this process, as the success of TACE relies on an efficacious chemo-embolic agent balancing effective embolisation with drug delivery​[66]​.

An alternative approach to TACE is to perform bland embolisation (transarterial embolisation [TAE]), which involves embolising the tumour’s arterial blood supply without using a cytotoxic agent. TAE aims to treat the target lesion purely by occluding its blood supply. Other alternative techniques include drug-eluting bead TACE (DEB-TACE), which uses microspheres carrying the cytotoxic agent to effect sustained local drug delivery, and transarterial radioembolisation (TARE) using a radioactive embolising agent instead of a conventional cytotoxic agent​[66]​. 

Stereotactic body radiotherapy (SBRT) is an emerging technique that can deliver focused, high-dose radiotherapy to tumours​[64]​. The role of SBRT is not yet established but early evidence suggests it might offer comparable local control to radiofrequency ablation, although lower survival has been seen​[55,67,68]​. SBRT appears unlikely to displace ablation as the first-line curative treatment for small tumours but is likely to have a role alongside TACE in more advanced disease. For example, SBRT can be safely used in patients with portal vein thrombosis, or to treat tumours adjacent to the biliary system, unlike TACE​[69,70]​.

Locoregional therapies, including TACE and radiofrequency ablation, increasingly allow tumours to be ‘downstaged’ such that they meet Milan criteria for transplantation​[71,72]​. It is likely that this role will expand in the future, as transplant criteria evolve​[15,33,55]​. 

Systemic therapy 

Patients with diffuse, infiltrative liver disease, extrahepatic spread, or portal vein invasion (BCLC-C) are not suitable for surgery or locoregional therapy and should be considered for systemic therapy​[15,33,55]​. For many years, sorafenib, an oral multiple tyrosine kinase inhibitor (TKI), was the sole agent approved for use in HCC — offering a median overall survival of around 10 months compared with 8 months in untreated patients​[15]​. However, the combination of atezolizumab, a monoclonal antibody against programmed cell death-ligand 1 (PD-L1), and bevacizumab, a monoclonal antibody that inhibits VEGF A, has recently been shown to be superior to sorafenib with a median overall survival of 19 months — this combination, known as ‘atezo-bev’, is currently the preferred regimen​[55,73,74]​. Durvalumab, a monoclonal antibody against PD-L1, after a priming dose of tremelimumab, a monoclonal antibody against cytotoxic T-lymphocyte-associated protein 4, may also be superior to sorafenib​[75]​. Hence immunotherapy is now considered the first line systemic therapy for HCC​[55]​. 

Patients considered for systemic therapy must have relatively preserved liver function, usually adjudged by the widely used Childs-Pugh score. Atezo-bev is usually restricted to those with Childs-Pugh A cirrhosis, although sorafenib was historically used in selected patients with Childs-Pugh B cirrhosis — with significantly reduced survival​[15,23,55,76,77]​. Bevacizumab is associated with an increased bleeding risk; therefore, there is a low threshold for endoscopy screening in patients with evidence of portal hypertension​[78]​. Patients with clinically significant varices despite use of a non-selective beta-blocker or a programme of variceal band ligation should not commence atezo-bev​[79,80]​. Anticoagulant use, vascular (for example, recent stroke or myocardial infarction) and autoimmune disorders (for example, active ulcerative colitis), hypertension, and prior liver transplantation may all preclude treatment with atezo-bev​[79]​.

Sorafenib is used if atezo-bev is ineffective or not tolerated, or first line if atezo-bev is contraindicated. If patients tolerate but progress on sorafenib, regorafenib (an oral multi-kinase inhibitor) is indicated as second- or third-line therapy as it may extend survival​[81]​. Cabozantinib, a small-molecule TKI, offers an additional second- or third-line option, irrespective of a patient’s tolerance of sorafenib or prior use of a second-line option such as regorafenib​[82]​. Ramucirumab, an anti-angiogenic VEGF receptor-2 monoclonal antibody, has been shown to have survival benefit for those with an AFP >400ng/mL​[83]​. 

Lenvatinib, a multi-kinase inhibitor, is an alternative to sorafenib and is used in a similar manner (e.g. if atezo-bev is contraindicated). Lenvatinib is non-inferior to sorafenib but has not been tested against atezo-bev​[84]​. Likewise, durvalumab monotherapy is non-inferior to sorafenib​[75]​. However, it is not yet clear what the best second line options are following lenvatinib or durvalumab monotherapy, hence sorafenib is preferred in current treatment algorithms​[55]​. 

For each of these medications, there are attendant considerations around adverse drug reactions, monitoring and drug interactions. For example, the TKIs and the multi-kinase inhibitors can lead to hand-foot syndrome, checkpoint inhibitors such as atezolizumab can cause acute hepatocellular and cholestatic liver injury, and almost all of the systemic therapies can cause common adverse events such as diarrhoea, rash and fatigue​[73–75,81,82]​. Consequently, pharmacists and pharmacy technicians are involved when initiating therapy, and during treatment. The systemic therapy landscape has changed rapidly over the past five years and it is likely to continue to evolve as more therapies are demonstrated to effectively treat HCC. This will make the role of pharmacists increasingly important, as more drug options become available, and we are able to individualise care to account for patient characteristics and concomitant treatments for pre-existing illness.

Best supportive care

Patients with advanced liver disease, who are outside of transplant criteria and are unsuitable for systemic therapy, should be offered best supportive care. It is recognised that often their prognosis is limited more by their advanced liver disease than by their cancer, and their care should include optimisation of such​[15,33,55]​. This includes management of hepatic encephalopathy and ascites, alongside more traditional palliative measures, such as control of pain, nausea and agitation. The use of questionnaires as quality-of-life screening tools can highlight primary symptoms to address from a patient perspective​[85]​. Increasingly, the importance of palliative care in the management of chronic liver disease is being recognised, and specific guidance is being developed to assist clinicians managing these challenging patients​[86]​. Many of the medicines traditionally used to control distressing symptoms at the end of life are contraindicated in those with advanced liver disease or can cause adverse effects; for example, opiates can provoke hepatic encephalopathy​[87]​. Doses may need close adjustment and monitoring, and alternative medications may need to be identified — for example, shorter rather than longer acting benzodiazepines for agitation may be preferred to prevent accumulation​[88]​. 

Summary 

Pharmacists are closely involved at each stage of the patient’s journey with HCC. They are involved in optimising the treatment of underlying liver disease, including screening for drug interactions and adverse events in patients with chronic liver disease. The role of the pharmacy team is instrumental in monitoring immunosuppressant dosing in those undergoing transplant and in the preparation of chemo-embolic agents and immunotherapy.