Indian Journal of Cancer Medknow Publications on behalf of Indian Cancer Society ISSN: 0019-509X EISSN: 1998-4774 Vol. 48, Num. 3, 2011, pp. 339-344

Indian Journal of Cancer, Vol. 48, No. 3, July-September, 2011, pp. 339-344

Original Article

Transarterial chemoembolization for hepatocellular carcinoma: Significance of extrahepatic collateral supply

SB Paul¹, SR Gamanagatti¹, A Mukund¹, SZ Abbas¹, SK Acharya²

¹ Department of Radiodiagnosis, All India Institute of Medical Sciences, New Delhi, India
² Department of Gastroenterology, All India Institute of Medical Sciences, New Delhi, India
Correspondence Address: S K Acharya, Department of Gastroenterology, All India Institute of Medical Sciences, New Delhi, India, subratacharya2004@yahoo.com

Code Number: cn11089

PMID: 21921335

DOI: 10.4103/0019-509X.84941

Abstract

Keywords: Extrahepatic supply, hepatocellular carcinoma, multiphasic computed tomography, percutaneous acetic acid ablation, transarterial chemoembolization

Introduction

Hepatocellular carcinoma (HCC) is one of the leading causes of worldwide cancer mortality, with an estimated 1 million deaths annually and a five-year survival rate of less than 5% . ^[1] More than 80% of the cases are encountered in Asia and Africa. ^[2] Screening high-risk patients of cirrhosis for the development of HCC is a norm practiced in the developed countries to detect cases of early HCC so that they are amenable to the curative treatment modalities. On the contrary, in developing countries like India, most of the patients are diagnosed with advanced disease at presentation for which palliative therapy forms the mainstay of treatment. ^[3],[4] Transarterial chemoembolization chemoembolization (TACE) is an established treatment modality in which chemotherapeutic agent/s are directly injected into the artery supplying the HCC and thereafter the artey is embolized using gelatin sponge particles. TACE is the most preferred treatment for unresectable HCC. ^[5],[6]

Hepatocellular carcinoma is a highly vascular tumor supplied by the hepatic artery and it is precisely this anatomic configuration that is made use of in the procedure of TACE. TACE is aimed at delivering the chemotherapeutic drugs to the hypervascular tumor via the feeding hepatic artery with subsequent embolization of the feeding arterial supply resulting in extensive necrosis of the large vascularised HCC. However, certain HCCs are prone to develop additional extrahepatic collateral arterial supply (ECS) depending on various factors like their location and size. ^{[7],[8],[9],[10],[11]} Such a situation may lead to the need of performing additional chemoembolization through the ECS in order to achieve complete local tumor response. This study aims to highlight this important aspect for effective management of HCC by TACE and to identify these difficult cases for effective tumor control.

Materials and Methods

Between 2001 and 2008, patients of HCC reporting to the liver clinic of our hospital were subjected to a detailed clinical and radiological examination. Clinical evaluation included detailed history and physical examination. Investigations included complete blood count, liver function and serum alpha-fetoprotein (AFP).

Radiological workup comprised of an abdominal ultrasonogram (US) performed on a Philips HDI 5000 unit using a 3.5 MHz curvilinear transducer. Multiphasic computed tomography (MPCT) of the liver was done at the time of diagnosis in all the patients using a subsecond helical CT scanner (Somatom plus 4, Siemens, Erlangen, Germany). Non-contrast CT of the liver was done first. This was followed by the scanning of the liver after the administration of 100-120 ml of iodinated intravenous contrast media at a rate of 3 ml/sec using a pressure injector (Medrad, Medrad, PA, USA). Data acquisition was done in the arterial, venous and delayed phases at 20 sec, 60 sec and 2 min respectively after the contrast injection.

Diagnostic criteria followed for HCC were the modified European Association for Study of Liver (EASL) criteria. ^[12] Contrast-enhanced multiphasic magnetic resonance imaging (MRI) was done if necessary for diagnosis. Staging of HCC was done based on the Barcelona clinic liver cancer (BCLC) staging. ^[13]

BCLC-B/C patients of unresectable HCC fulfilling the following inclusion criteria-having underlying Child's A/B cirrhosis, a normal main portal vein, less than 50% involvement of the liver by the tumor-were included in the study and subjected to TACE. Patients with co-morbid illness like coronary artery disease, congestive heart failure, chronic renal failure, etc., previous history of encephalopathy/upper gastrointestinal bleed in the last six months or having extrahepatic disease were excluded from the study. Ethics clearance for TACE has been taken from the hospital committee for the ongoing trials at our centre. An informed written consent was obtained prior to the procedure of TACE in all the patients.

Technique of transarterial chemoembolization

TACE was performed through the trans-femoral route. Superior mesenteric artery and celiac axis arteriogram was obtained to begin with. Selective cannulation of the feeding hepatic artery to the tumor was done using 5F coeliac catheter and 0.0038-in. 'J' tip terumo guide wire. The catheter was placed in the artery as close as possible to the tumor. The chemotherapeutic drugs (Doxorubicin 50 mg, Cisplatin 100 mg) with 10 ml of iodinated non-ionic contrast media and 20 ml of iodized oil (lipiodol) were delivered through this cannulated feeding hepatic artery. Subsequently, this feeding artery was embolized using gelatin sponge pledgets. This is done to occlude the arterial supply temporarily so that the injected drug mixture can have a prolonged stay in the region of the tumor, facilitating better local effect of the drugs on the tumor and causing tumor necrosis. It also prevents the drugs from moving into the systemic circulation. Intra-arterial lidocaine (10 mg) was given between 10-ml aliquots of chemoembolization material to reduce pain after embolization.

If the tumor was very large, located on the surface of the liver, had an exophytic component or was seen to be infiltrating the parietes or the adjacent major organs, the post-embolization run was studied for the extent of the retained lipiodol. In situations of incomplete lipiodol retention, an active attempt was made to look for the presence of the ECS and its cannulation in the same sitting. The expected extrahepatic source to the tumor depended upon the anatomical location of the tumor. Selective angiograpy was then performed. The extrahepatic collateral/s was/were cannulated actively. An attempt was made to go very superselective (as distally as possible in the collateral) during the course of the cannulation to avoid injury to the non-target organs. Chemoembolization through the collateral supply was performed in addition to the same procedure performed through the hepatic artery in these patients.

In cases where the ECS cannulation failed despite laborious attempts, an alternative locoregional treatment therapy of percutaneous acetic acid ablation (PAI) of the tumor supplied by the extrahepatic source was performed in a separate sitting subsequently. PAI was done under ultrasound guidance using local anesthesia. Total dose of acetic acid to be injected was calculated based on the size of the mass (three times the largest diameter of the mass). Of the total calculated dose, about 2-3 ml of acetic acid was injected into the tumor in one session. For multiple sessions, not more than two sessions per week were done. The local response was assessed by MPCT at one month following ablation and the decision to inject more of acetic acid was based on the local response on MPCT.

Follow-up and assessment of treatment response

All patients were followed up prospectively after one month following the procedure of TACE. They were subjected to a detailed clinical examination, serum alpha fetoprotein (AFP) estimation [(only for those who had elevated AFP at diagnosis (>10 ng/ml)) and radiological test of MPCT to assess the local tumor response. Assessment of response to treatment was made in terms of complete response, partial response and disease progression on MPCT performed after one month of the procedure of TACE. In cases of partial response of the tumor, if the lipiodol-covered tumor showed any defect, or there was any local recurrence in the region of the periphery of the tumor, presence of ECS to the tumor was further suspected. Additionally, if on follow-up a sustained elevation of serum AFP was observed despite complete tumor response on MPCT, a high suspicion of the presence of the ECS to the HCC was considered. In all such cases, repeat TACE was done with a special attempt to look for the extrahepatic sources of supply to the tumor. A gap of four to six weeks was given between before repeating the session of TACE.

Response of TACE on MPCT

The response of the tumor to the treatment of TACE was assessed and categorized into the following:

a) Complete response was labeled on MPCT when the tumor was fully covered with lipiodol and had no enhancing viable tissue in the arterial or venous phase of the MPCT.
b) Partial response was present when the tumor area was partly covered with lipiodol and the enhancing viable tumor was present in the arterial/venous phase of MPCT.
c) Disease progression was seen when either the treated tumor (s) showed increase in size with enhancing viable components or fresh lesions appeared at other sites in the hepatic parenchyma other than the previous existent lesions.

Results

A total of 85 patients of HCC were subjected to 147 sessions of TACE. Of these, eight cases were found to have additional ECS. This source of supply was from the following arteries; inferior phrenic artery (n0 = 1), intercostals ( n = 1), internal mammary artery ( n = 1), omental arteries ( n = 1), gastroduodenal artery (GDA) ( n = 3), branch of the superior mesenteric artery (SMA) (n = 1).

TACE could be successfully performed through the extrahepatic supply in five cases [Figure - 1] and [Figure - 2]. Post-procedure dual-phase CT at one month showed complete local response to treatment in all these five cases. However, there was difficulty in cannulation in the remaining three cases due to multiple reasons hindering proper cannulation of the feeding artery. In one case, there were numerable feeding collaterals from multiple sources. Chemoembolization through one source, that is, the branch from the right inferior phrenic artery in addition to the hepatic artery could be undertaken. However, cannulation of the additional small feeding collaterals from the intercostal arteries could not be performed as they had a very long, narrow and tortuous course and due to repeated attempts developed spasm of the artery and thus the procedure could not be accomplished. In the other two cases, there was failure to cannulate the collateral supply due to the presence of an unfavorable anatomy of the feeding artery (narrowing at the origin of the artery with extremely tortuous, narrow caliber distally). This resulted in incomplete treatment of the tumor in all these three patients and thus the treatment of the tumor was completed by an alternative available locoregional therapy of PAI and complete response of the tumor was achieved [Figure - 3].

Response of the treatment was assessed by serum AFP in 3/8 patients who had elevated AFP at the time of TACE. Of these three patients, one patient showed decline in serum AFP at one month while the remaining two developed fresh lesions and hence the AFP level got further elevated, indicating disease progression.

The procedure of TACE through the ECS was well tolerated by our patients. Only mild post-procedure complications like pain abdomen, fever, nausea and vomiting constituting the post-embolization syndrome were encountered. These were self-limiting and subsided in about three to four days following the procedure. No non-target complication was encountered in our five cases.

Discussion

TACE is the most commonly used treatment modality for unresectable HCC. ^[5],[6] This procedure entails the delivery of the mixture of the chemotherapeutic drugs and lipiodol followed by the embolic agents into the vascular liver tumor through its intrahepatic arterial supply resulting in the ischemia and necrosis of the viable, hypervascular tumor. Occasionally, HCCs may also have additional ECS. TACE is a locoregional therapy and the precise aim is to produce local tumor control. Therefore, the presence of additional ECS to the tumor is of immense importance for treatment planning. For TACE to be effective, not only the hepatic arterial branches but also these additional collateral vessels have to be adequately embolized so that the entire tumor is treated.

The index of suspicion should be high in cases where the MPCT shows the following features - large surface masses, tumors having an exophytic component or in contact with bare area of the liver or large tumors with direct invasion into the adjacent organs or extracapsular infiltration. ^{[7],[8],[9],[10],[11],[14]} Large tumors are usually fed by multiple feeding arteries-not only hepatic arterial branches but also extrahepatic collaterals and this makes the procedure more laborious and technically difficult for achieving complete necrosis of the large tumors by TACE. ^[10],[11] The prevalence of ECS at the initial chemoembolization session in a tumor less than 4 cm in diameter is less than 3% and this increases to 63% when the tumor is larger than 6 cm in diameter. ^[8]

Additionally, if the MPCT following TACE shows a peripheral defect in the retained lipiodol deposition within the tumor or the appearance of a local recurrence at the peripheral portion of the treated tumor, there would be a high possibility of the presence of the extrahepatic source to the HCC. Moreover, the index of suspicion should always be high in patients with sustained elevation of serum AFP level despite adequate embolization of the hepatic artery. Rarely, hypertrophied extrahepatic collateral vessels, especially the right inferior phrenic artery may be seen during the arterial phase of MPCT.

The correlation of the MPCT and angiographic findings is of immense importance. If a tumor observed at CT is not demonstrated on hepatic angiography, an active effort should be made to look for the extrahepatic collateral feeders to the tumor depending upon its location. When the tumor staining on hepatic angiogram has a focal defect or a focal iodized oil retention defect post iodized oil infusion, alternative feeder vessels are a strong possibility.

Multiple causes of ECS to the HCC have been enumerated. In the past, it was considered to occur due to the occlusion of the hepatic artery by surgical ligation (not performed these days), postoperative omental or peritoneal adhesions following repeated TACE with resultant occlusion of the peripheral hepatic arteries, arterial dissection, or tumor location in the bare area and suspensory ligaments. ^{[8],[14],[15]} Our patients were allocated TACE as the primary treatment and it was observed that the location of the tumor (surface tumor) and presence of the exophytic component of the mass were the likely contributing factors. Similar observations have been made by Kim and Chung et al.[7],[8]

It has been shown that superselective TACE with an attempt to deliver the drug very close to the tumor has a better outcome than whole-liver or lobar TACE. Delivering the therapeutic agent directly to the target lesions minimizes embolization of non-targeted normal liver parenchyma. ^[16],[17] Hence, a similar principle holds true in cases of TACE performed through the ECS to the tumor. This superselective cannulation is even more challenging in the setup of extrahepatic collaterals where the feeders are invariably unusually long, tortuous and narrow making it further difficult to negotiate very close to the tumor. Such vessels are many a time difficult to negotiate despite the use of small-caliber microcatheters. To add to the agony, these collateral vessels are also liable to go into spasm following repeated cannulation attempts (as seen in our cases) resulting in the procedure being futile. We failed to cannulate three such patients despite the use of microcatheters and thus, had to resort to the alternative treatment modality of PAI for achieving complete local tumor control. The procedure has a risk of complications because of the inadvertent embolization of non-target branches. ^{[18],[19],[20]} The majority of the collateral vessels supplying the HCC are of small caliber and are usually branches of arteries which supply the critical areas such as skin, gastrointestinal tract, spinal cord, gall bladder, and diaphragm. The complications of non-target embolization depend on the location of the collateral supply. Skin erythema and gastrointestinal erosion, ulceration, or perforation may occur following non-target embolization of gastric, omental branches. Inadvertent embolization of cystic artery may lead to cholecystitis, whereas skin necriosis, of the cystic artery, whereas skin necrosis, may arise when the internal mammary, intercostal, or lumbar artery are embolized and very rarely paraplegia may result from the inadvertent embolization of spinal branches arising from intercostal or lumbar collateral vessels. Non-target embolization of the inferior phrenic artery may result in shoulder pain and pleural effusion.

Hence, while doing TACE through these collateral branches, one should be very superselective and use microcatheter systems to avoid risk of embolization of non-target branches. We did not encounter any significant complications in the five cases we embolized.

There are a number of limitations of this procedure as well. TACE performed through the extrahepatic collateral arteries may not be successful in the following situations: (a) when a large number of collaterals are seen supplying the large tumor, it may be practically impossible to embolize each one of them successfully, (b) drug administration through the microcatheter may be insufficient to cover the large-sized tumor at one time. In such situations, repeat TACE is not a solution, rather, it would be appropriate to treat this portion of the viable tumor with an alternative treatment modality like PAI if possible.

To conclude, identification of the presence of an ECS to large unresectable HCCs is of immense importance to increase the therapeutic efficacy of TACE, and in identifying patients of HCC who would eventually benefit from an alternative locoregional therapy like PAI.

References

Copyright 2011 - Indian Journal of Cancer

The following images related to this document are available:

Photo images