DIAGNOSIS
Most HBP malignancies are initially detected on cross-section imaging studies. In most cases, a primary malignancy such as hepatocellular carcinoma (HCC) or pancreatic adenocarcinoma can be diagnosed with confidence on imaging alone when typical imaging features are demonstrated(6). However, tissue diagnosis is occasionally required when imaging features are equivocal, where liver metastases are present without an obvious primary source or when advanced molecular or mutation testing is necessary to guide subsequent systemic therapy including recruitment to clinical trials(7).
Specific biopsy techniques include percutaneous core biopsy for focal lesions or endobiliary forceps biopsies or cytology brushings for biliary strictures and cholangiocarcinomas (Fig. 1).
Percutaneous core biopsies of focal liver lesions are commonly used to identify metastases, most commonly from a lung, colorectal or breast primary. Present guidelines for HCC diagnosis relegate the use of biopsies only for lesions with indeterminate or conflicting imaging features. However, given the increasing knowledge of HCC molecular characteristics and subtypes with prognostic impact, there may be a greater role for biopsies in HCC diagnosis in the future(7).
Most percutaneous biopsies are performed under ultrasound guidance with the use of an outer coaxial needle and an inner core biopsy needle (typically 16-18 gauge) to minimise the number of capsule transgressions, bleeding complications and increase biopsy yield. It is the authors’ practice to administer local anaesthesia to the skin, tract and visceral capsule under real-time ultrasound guidance. Percutaneous core biopsies for focal lesions has a reported sensitivity of >90%. Lesions which are inconspicuous on ultrasound can be successfully biopsied with the aid of contrast-enhanced ultrasound with a sensitivity of 92.8% in one series(8).
Biliary strictures can be sampled from a percutaneous transhepatic or endoscopic approach (ERCP cytology brushings or EUS-FNA). The diagnostic yield of percutaneous transhepatic forceps biopsy is comparable to EUS-FNA with a reported sensitivity of 75% and specificity of 100% (9, 10) and can be performed at the time of percutaneous biliary drainage (Fig. 2).
Biopsy complications include tumour seeding and haemorrhage. Tumour seeding post-biopsy is low(11). In the setting of HCC, a 2007 meta-analysis found the median incidence to be 2.7%(12) and is lower at 1.1% from metastatic liver lesions(13). Major bleeding following percutaneous biopsy is rare with an incidence of 0.8-1.1%(14, 15). Post-biopsy tract embolisation with gelatin sponge or autologous blood clot has a role in reducing haemorrhage risk.
THERAPY
Surgical resection or liver transplantation are the preferred curative treatment options for patients with HPB malignancy. Locoregional IO therapies are primarily reserved for patients deemed non-surgical candidates due to comorbidities or disease status and involve thermal ablation, which is potentially curative for early-stage disease, or embolisation therapies for disease control(16).
Ablation
Thermal ablation techniques, utilising either radiofrequency (RFA) or microwave (MWA) energy to produce localized tumour heating and coagulative necrosis, are used to treat early-stage HCC (BCLC Stage A – single to 3 nodules) and liver metastases.
Ablation is performed by placement of an antenna in the epicenter of the lesion under imaging-guidance followed by selection of appropriate energy power and treatment time to achieve a zone of ablation incorporating the tumour (Fig. 3, Fig. 4). The efficacy between RFA and MWA (overall survival [OS], local recurrence and complication rates) are identical but MWA has largely replaced RFA in the liver due to higher intratumoral temperatures, better thermal convection, faster ablation times and less susceptibility to heat sink effects when tumours are located in proximity to large blood vessels which may result in incomplete ablation(18). The latest generation MWA systems are capable of ablation zones of between 4-5 cm per probe and multiple probes can be used to achieve a larger, overlapping ablation zone for larger tumours.
Current evidence has not conclusively shown superiority of surgical resection over ablation, but resection may have an advantage for larger tumours (3-5 cm) due to the greater risk of incomplete ablation(19). Individual patient and disease characteristics are probably of equal importance in the selection of an appropriate treatment modality.
CT imaging appearances in the early post-treatment period (Fig. 5), a spherical area of hypoenhancement corresponding to the ablation zone and in the very early post-treatment period, a hyperenhancing rim due to reactive tissue hyperaemia. Follow-up imaging shows a gradual decrease in size of the ablation zone with scarring. Incomplete treatment or recurrence is detected based on residual or new areas of enhancement on CT or MRI.
Chemo and radioembolisation
Transarterial chemoembolisation (TACE) is a locoregional treatment which involves precise intra-arterial delivery of chemotherapy and embolisation of tumour arterial supply to achieve combined cytotoxic and ischaemic effects leading to tumour necrosis. TACE is indicated in the treatment of intermediate stage HCC (BCLC Stage B) or unresectable colorectal liver metastases for disease control. TACE can also be used in the liver transplant setting to downstage a patient into transplant criteria eligibility and extend the patient’s time on the transplant waitlist(20). The best candidates for TACE are patients with solitary or limited multifocal HCC with preserved liver function (Child Pugh A or B) and no evidence of vascular invasion. TACE has been shown to improve survival of patients with HCC compared to best supportive management(21-23) and in patients with colorectal liver metastases after failure of systemic chemotherapy(24).
TACE can be further divided into conventional TACE (c-TACE) and drug-eluting-bead TACE (DEB-TACE) (Fig. 6, Fig. 7). c-TACE involves delivery of a liquid mixture of lipiodol with chemotherapy followed by arterial embolisation with gelatin sponge particles. DEB-TACE utilizes 75-300 micron embolic microparticles (e.g. DC Bead, BTG; Tandem, Boston Scientific) which gradually releases chemotherapy into the tumour. Both techniques are similar in efficacy but DEB-TACE is associated with better tolerability and fewer adverse events in comparative randomized-controlled trials(25, 26). TACE is delivered through a microcatheter advanced into the implicated hepatic arterial branches with delivery of the chemotherapy and embolic agent to the tumour. Embolisation works well in this setting as tumours preferentially derive vascular supply via the hepatic artery in comparison to normal liver tissue which is supplied predominantly by the portal vein (70%). Doxorubicin is commonly used as the chemotherapeutic agent for HCC (DEBDOX-TACE) whilst irinotecan is used for colorectal metastases (DEBIRI-TACE).
Transarterial radioembolization (TARE) or selective internal radiation therapy (SIRT) (Fig. 8) is a more recent locoregional treatment and is a form of brachytherapy delivered through Yttirum-90 containing resin (SIR-Spheres, Sirtex) or glass (TheraSphere, BTG) microspheres resulting in local emission of beta-radiation. The microspheres are delivered through a microcatheter from the hepatic artery to one liver lobe at a time. Although TARE is less ‘targeted’ compared to TACE, the technique is suited for multifocal lesions. As the embolisation effect is minimal due to the small particle size, it is suitable treatment option for patients with portal vein thrombosis or invasion where TACE is contraindicated. Both TACE and TARE have similar clinical outcomes for HCC and colorectal metastases(27-29).
Imaging post-locoregional therapy is based on the detection of residual or new arterial hyperenhancement to indicate residual or recurrent disease or lack of enhancement as an indicator of necrosis rather than lesion size. The modified Response Evaluation Criteria in Solid Tumours (mRECIST) and European Association for the Study of the Liver (EASL) guidelines have been published to specifically guide reporting following locoregional therapy(30, 31). Readers are also directed to an excellent review of post-locoregional therapy imaging by Young et al for further reading(32).
Portal vein embolisation (PVE)
In contrast to the aforementioned therapies, PVE has a role in curative-intent hepatectomy(33). Patients who are candidates for major liver resection may require PVE to increase the volume of the future liver remnant (FLR) when the FLR is too small to support essential liver function post-operatively. PVE is indicated when the FLR volume as measured on CT volumetry is between 25-30% of the original liver volume in healthy livers or between 35-40% in the presence of chronic liver disease such as cirrhosis, post-chemotherapy liver injury or cholestasis(34).
Embolization redirects portal vein flow towards the residual segments, promoting liver growth and thus optimising FLR prior to hepatectomy (Fig. 9). PVE can be performed via an ipsilateral (to the planned resection side) or contralateral (via the FLR) approach (Fig. 10). Following percutaneous transhepatic access into the portal vein, individual portal vein segments are selectively embolised using a mixture of embolic agents to achieve stasis (Fig. 11, Fig. 12). Commonly used embolic materials include microparticles such as PVA, gelatin sponge, embolisation coils, glue and vascular plugs. Preoperative PVE has a high clinical success rate with a mean FLR hypertrophy rate of 37.9%(34).
PALLIATION
IO has a rapidly expanding role in the palliative management of patients with advanced-stage cancer, often with complex post-surgical anatomy and medical comorbidities; increasing anaesthesia and surgical risk(35, 36). The minimally invasive nature of IO offers suitable alternatives to maximise quality of life and reduce the length of hospitalisation.
Percutaneous transhepatic cholangiography (PTC), biliary drainage and stenting
Biliary obstruction with obstructive jaundice is a frequent presentation of advanced HPB malignancy and can arise from tumour or extrinsic nodal compression as well as in-situ tumour growth in the case of cholangiocarcinoma. Most cases of biliary obstruction are now diagnosed on cross-sectional imaging with ultrasound or CT. PTC is a well-established procedure, first conceptualized more than fifty years ago which maps out the biliary anatomy and is now most commonly performed during percutaneous transhepatic biliary drainage (PTBD) to relieve biliary obstruction(37)(Fig. 13). Although endoscopic (ERCP) management of biliary obstruction has seen an increased role in recent years, PTBD remains indispensable in the treatment of biliary hilar strictures, post-operative anatomy limiting endoscopic access (e.g. following gastrectomy, Whipple resection) or where endoscopic attempts have failed. Potential complications associated with PTBD include malposition or migration of biliary drains, vascular injury during percutaneous access and peri-procedural sepsis(38).
Percutaneous biliary stenting to internalize biliary drainage often follows an initial trial of PTBD as a more permanent solution and increases patient comfort. Self-expanding metal stents are preferred over plastic stents in cases of malignancy due to superior patency, lower likelihood of reintervention and lower risk of migration(39) (Fig. 14). In addition, hilar strictures which present an anatomical challenge, can be managed with hilar-reconstruction techniques including kissing, Y or trifurcation stents(39) (Fig. 15, Fig. 16). The Achilles heel of biliary stenting is limited long-term primary patency rates which is on average about 1 year as a consequence of tumour ingrowth. Polytetrafluroethylene-covered stents were once thought to be more effective in resisting tumour ingrowth(39, 40). However, a 2014 randomised controlled trial cast doubt upon the superiority of covered stents(41) and a recent meta-analysis of 14 trials found no significant difference between covered and uncovered metal stents in terms of primary patency and stent dysfunction(42).
Coeliac plexus neurolysis
The coeliac plexus, the largest sympathetic plexus, relays nociceptive impulses from the stomach to the proximal transverse colon. Neurolysis of the coeliac plexus with a percutaneous injection of pure ethanol alleviates pain originating from the upper abdominal viscera. This technique has been widely used since its introduction in 1914(43). Image guidance can be provided by fluoroscopy, ultrasound, computer tomography (CT) or magnetic resonance imaging (MRI); CT is now the preferred technique given the superior anatomical detail achievable and greater independence from operator skill(44).
CT guided percutaneous coeliac plexus nerve block is an established therapeutic choice in the treatment of intractable pain in the setting of upper abdominal malignancy. Coeliac plexus neurolysis is an effective technique (Fig. 17), found to have lasting analgesic effects in 70-90% of patients with upper abdominal malignancies(45). Neurolysis allows for reduction in the chronic use of high-dose opioid analgesia and the associated adverse effects; improving the quality of life(46, 47). There is a low complication rate associated with this procedure(45, 48).