Ablational Therapies for Hepatic Tumors

By John M. Kane III, MD

The treatment of primary and secondary tumors of the liver represents a significant challenge to the global health care community. As a consequence of viral hepatitis, aflatoxins and alcohol, hepatocellular carcinoma is one of the most common malignancies throughout the world. Affecting approximately one million people per year, survival has been rather dismal, with a 25-90% three-year mortality based on the stage of the tumor.1 In select patients, the only potentially curative therapies have been surgical resection or liver transplantation. In regard to secondary tumors, colorectal cancer remains the most frequent metastatic liver lesion in the United States. Liver resection is also potentially curative in certain patients, resulting in a 25-35% five-year survival. Unfortunately, many patients with primary or secondary tumors recur within the liver following resection and only a small proportion are candidates for reresection. Although usually only palliative in nature, many patients with metastatic neuroendocrine tumors also seek surgical treatment for symptomatic hepatic disease.

Rationale for Ablational Therapies

To understand the development of ablational therapies for hepatic tumors, one must appreciate the indications for and limitations of traditional surgical resection. Foremost, if hepatic resection of primary or metastatic tumors is undertaken for potentially curative intent, there must be no evidence of extrahepatic disease. Second, all hepatic disease must be "technically" resectable. Therefore, there should be no involvement of vital structures such as the hepatic artery, major bile ducts, or main portal vein. Third, there must be adequate functional hepatic reserve following resection. This is typically at least 20% of "normally" functioning liver parenchyma or more if there is hepatic dysfunction. Finally, given that liver resection is a major surgical undertaking, the patient should have minimal comorbid diseases in order to have an acceptable operative morbidity and mortality. Based on the above considerations, less than 25% of patients with primary or secondary liver tumors are candidates for surgical resection.

In contrast to surgical resection, the fundamental advantage of ablational therapies is the ability to preserve uninvolved hepatic parenchyma. Given that treatment is directed specifically toward the tumor, only a small rim of surrounding liver is destroyed. As the majority of hepatocellular carcinomas arise from a background of cirrhosis, maximal conservation of parenchymal volume is essential for maintaining an acceptable treatment-related morbidity and mortality. This approach is also well suited to the characteristics of metastatic tumors of the liver; multiple lesions, bilobar disease, central location, or tumor in close proximity to major hepatic structures. Patients who have undergone prior liver resection with limited hepatic reserve can also be treated for recurrent disease within the remaining parenchyma. An additional benefit of ablational therapies is that they are more amenable to a minimally invasive approach.

In summary, ablational therapies allow for greater flexibility in patient selection. Some patients with poor tumor characteristics, limited hepatic reserve, or significant comorbid diseases that would preclude surgical resection may now be considered for ablational modalities. This approach is also well suited for the palliation of symptomatic neuroendocrine liver metastases. Therefore, ablational therapies potentially increase the number of patients who may be considered for treatment of primary or secondary liver tumors. At the present time, the most commonly used ablational techniques are percutaneous ethanol injection, cryotherapy, and radiofrequency ablation.

Percutaneous Ethanol Injection

Percutaneous ethanol injection is the oldest and most straightforward of the ablational modalities and has been used primarily for hepatocellular carcinoma. Under ultrasound guidance, a needle is placed into the tumor and approximately 1-10 mL of 95% ethanol is injected. The ethanol diffuses throughout the tumor, but extratumoral spread is usually limited by the surrounding cirrhosis. Treatments are repeated over several weeks to effect complete destruction of the tumor. An alternative approach for larger or multiple tumors is a single session with large volume ethanol injection. However, this technique often requires general anesthesia secondary to the intoxicating effects of the ethanol. The tumoricidal effects of ethanol injection are via two mechanisms; direct cellular dehydration leading to coagulation necrosis and vascular endothelial damage with thrombosis and subsequent ischemia.2

Benefits of Injection

The most striking benefits of percutaneous ethanol injection are the simplicity and minimally invasive nature of the procedure. Although it can be performed at the time of laparotomy, it is usually performed via a percutaneous non-surgical approach. Unfortunately, percutaneous ethanol injection is not without drawbacks. It is usually contraindicated in the setting of significant coagulopathy or ascites, which are fairly common to this patient population. Complications of percutaneous ethanol injection include intraperitoneal hemorrhage, hepatic parenchymal infarction, portal vein thrombosis, and tumor seeding of the needle tract. Unique to percutaneous ethanol injection are intoxication secondary to the ethanol and occasional biliary fibrosis after accidental intraductal injection. Despite these risks, the safety profile of percutaneous ethanol injection is excellent. In more than 1000 patients treated with percutaneous ethanol injection in the published literature, morbidity was less than 2.5% with no treatment related mortality.2

There is an evolving body of data in support of percutaneous ethanol injection as a potentially curative therapy for hepatocellular carcinoma that is comparable to surgical resection. Onodera and colleagues reported on a small series of patients with early-stage hepatocellular carcinoma who were treated with various modalities, including surgical resection and percutaneous ethanol injection. There was no significant difference in the three-year survival rates between percutaneous ethanol injection (91%) and resection (54%).3 Kotoh and associates found no difference between surgical resection and percutaneous ethanol injection in regard to mortality (29% vs 39%) and recurrence (82% vs 70%) for solitary hepatocellular carcinomas less than 2 cm.4 There are three key points in this study which deserve mention. First, patients selected for percutaneous ethanol injection were usually those with poor hepatic function or advanced age. Therefore, these patients were more debilitated than the resection candidates. Second, all intrahepatic recurrences in both groups were at sites remote from the original tumor. Finally, the procedural cost (excluding hospitalization) was 18 times greater in the surgical resection group.

In a large study of percutaneous ethanol injection for hepatocellular carcinoma by Livraghi and coworkers, five-year survival was 37-40% for single lesions less than 5 cm, 30% for lesions greater than 5 cm, and 26% for multiple tumors.2 They also found that five-year survival decreased with worsening Child class (A, 47%; B, 29%; C, 0%). They noted that most of the Child A patients died of tumor progression, in contrast to complications of cirrhosis in the Child C patients. Therefore, percutaneous ethanol injection may have a limited effect on survival in patients with hepatocellular carcinoma and advanced hepatic disease. Finally, in a review of the available literature on surgical resection vs. percutaneous ethanol injection for single hepatocellular carcinomas less than 5 cm, the five-year survivals were comparable at 49% vs. 48%, respectively.2


Knowledge of the detrimental effects of freezing on both normal and malignant tissues dates back to ancient times. Early attempts at tumor cryotherapy were limited by the inability to obtain the extremely low temperatures necessary for complete tissue destruction. The advent of liquid nitrogen-cooled cryoprobes has resulted in more predictable and uniform tumor freezing. At the temperatures achieved with modern cryoprobes (-196°C), there is formation of both intracellular and extracellular ice crystals that produce direct cellular destruction and local tissue ischemia.

At the present time, cryotherapy is usually performed via an open technique. As with surgical resection, the liver is completely mobilized and vascular control is obtained. Using intraoperative ultrasound, the target lesions and their relationships to the major intrahepatic structures are visualized. Under real time ultrasound guidance, cryoprobe placement is performed using a modified Seldinger technique. A needle is placed into the lesion, followed by a guidewire, dilator, and finally the cryoprobe. To minimize complications, it is essential that the cryoprobe passes through a portion of uninvolved hepatic parenchyma prior to entering the tumor. For larger lesions, multiple probes can be placed simultaneously. The liver must also be insulated from the surrounding intra-abdominal structures to prevent inadvertent local injury. During freezing, the progression of the peritumoral "ice ball" is monitored by ultrasound. The goal is for the outer rim of the ice ball to exceed the tumor by at least 1 cm. For lesions adjacent to major vascular structures, there is occasionally incomplete freezing due to a "heat sink" effect from the rapid flow of warm blood. This effect can be minimized with temporary hepatic inflow occlusion at the time of freezing. The tumor is allowed to thaw and the freeze-thaw cycle is often repeated for a total of two to three times.

Given the large size of the cryoprobe, bleeding from the probe tract can occasionally be problematic. Cold-induced injury may occur to either the bile ducts or poorly insulated extrahepatic structures. Other complications include pleural effusion, cold-induced arrhythmias, and biloma. One of the most feared acute complications is a major fracture of the liver while the parenchyma is still frozen. Additional complications unique to cryotherapy include a transient postoperative thrombocytopenia, myoglobinuria occasionally associated with acute renal failure, and a syndrome of "cryoshock" consisting of multiorgan failure, coagulopathy, and disseminated intravascular coagulation.5 A world survey by Seifert and Morris revealed an overall mortality of 1.5% for hepatic cryotherapy.5 Although cryoshock was observed in only 1% of all patients, it accounted for 18% of all treatment-related deaths. It has been suggested that the development of cryoshock may be associated with multiple freeze-thaw cycles.

Recipients of Cryotherapy

The results of cryotherapy for primary and metastatic liver tumors are somewhat clouded by the fact that this modality is almost always reserved for patients with surgically unresectable disease. Consequently, this select group should have more advanced disease and a poorer prognosis as compared to surgical candidates. In addition, cryotherapy has often been used in combination with more traditional treatments such as resection, chemotherapy, or chemoembolization. In a study of the role of cryotherapy in primary liver malignancies, the one-, three-, and five-year survival rates for cryotherapy alone in 78 patients were 64%, 40%, and 27%, respectively.6 For the treatment of hepatic colorectal metastases, Tandan and associates critically reviewed the available literature comparing cryotherapy to surgical resection.7 The median follow-up for the cryotherapy studies was 12-29 months with overall and disease-free survivals of 33-64% and 22-29%, respectively. In contrast, several large studies of surgical resection consistently had 20-40% five-year survivals. Given that the primary goal of cryotherapy for hepatic neuroendocrine metastases is palliation, results have been more promising. In a small study by Siefert and associates, 92% of patients with neuroendocrine metastases treated with cryotherapy were alive and mostly asymptomatic at a median follow-up of 13.5 months.8 There was only one death which was not treatment or tumor related.

Radiofrequency Ablation

Radiofrequency ablation of hepatic tumors has developed in response to the limitations of cryotherapy. The large size of the cryprobes, the problem of cold injury to adjacent structures, and the risk of hemorrhage have necessitated an open approach to therapy. In addition, it has been proposed that heat may be a more predictable tumoricidal insult as compared to cold. The effects of tissue heating are well described: irreversible protein denaturation above 49°C, coagulation at 70°C, and desiccation at 100°C.9 Radiofrequency ablation uses high frequency alternating current to generate tissue temperatures from 70°C to 90°C. Under ultrasound guidance, the tip of an electrode is placed into the tumor and current is applied for approximately 2-10 minutes. Given the small size of the electrodes, this technique can easily be performed percutaneously.

Maximal lesion size with monopolar radiofrequency ablation electrodes is limited to approximately 1.5 cm. Unfortunately, increasing the power applied to the electrode does not significantly increase lesion size as the flow of the current becomes limited by decreased conduction through desiccated tissues at the tip of the electrode. Therefore, several modifications of the technique have been developed to increase the maximal lesion size. These include inflow occlusion to minimize the "cold sink" effect of hepatic blood flow, multiple electrodes, cooling of the electrode tip, electrode tip irrigation with saline, and catheter tips with a deployable "umbrella" array of electrodes.10

Complications have been extremely rare with radiofrequency ablation. Given the small size of the probes and the coagulative nature of the therapy, the risk of hemorrhage is minimal. In a large study of 123 patients by Curley and coworkers, complications following radiofrequency ablation occurred in 2.4% of patients with no treatment-related mortality.11 Although radiofrequency ablation is relatively new, early results have been promising for hepatocellular carcinoma, and to a lesser extent, for metastatic tumors. A recent study by Rossi and colleagues used radiofrequency ablation to treat 39 patients with hepatocellular carcinoma and 11 patients with metastatic liver tumors.12 For the patients with hepatocellular carcinoma, median survival was 44 months, which corresponded to a 40% five-year survival. Of note, 41% of these patients developed recurrent tumor. However, only 13% of all recurrences were at the site of the previous radiofrequency ablation. For the patients with metastatic tumors, one patient remained disease free one year after treatment. In the study by Curley et al, post-treatment tumor recurrence was only 1.8% in 48 patients with hepatocellular carcinoma and 75 patients with metastatic tumors.11


At the present time, surgical resection remains the "gold standard" potentially curative therapy for primary and secondary liver tumors. However, the ablational modalities fill an important niche in the armamentarium of the physicians treating these patients. Due to poor location, underlying hepatic dysfunction, or comorbid diseases, only a minority of patients with hepatic tumors are considered candidates for surgical therapy. In addition, both primary and metastatic tumors have a high incidence of post-resection intrahepatic recurrences. Common to all of the ablational techniques is the ability to maximally preserve uninvolved liver parenchyma. They are also readily amenable to a minimally invasive approach. This allows for greater flexibility in patient selection for therapy. There is increasing evidence to suggest that percutaneous ethanol injection is comparable to surgical resection for the treatment of hepatocellular carcinoma. Although the role of cryotherapy and radiofrequency ablation in the treatment of hepatic tumors has not been completely defined, early results have been encouraging. Current interest in ablational techniques includes modification to laparoscopic approaches, ablation-assisted resection, and the use of microwave and laser catheters. With future technological innovations, the majority of patients with hepatic tumors may one day be successfully treated with non-resectional ablational techniques. (Dr. Kane is a Fellow in Surgical Oncology at Roswell Park Cancer Institute, Buffalo, NY.)


1. Carr B, Flickinger J, Lotze M. Hepatobiliary Cancers. DeVita V, Hellman S, Rosenberg S, eds. In: Cancer, Principles & Practice of Oncology, 5th ed. Philadelphia: J. B. Lippincott Co.; 1997:1087.

2. Livraghi T, Giorgio A, Marin G, et al. Radiology 1995;197:101.

3. Onodera H, Ukai K, Nakano N, et al. Cancer Chemother Pharmacol 1994;33(supp):S103.

4. Kotoh K, Sakai H, Sakamoto S, et al. Am J Gast 1994;89:194.

5. Seifert J, Morris D. World J Surg 1999;23:109.

6. Zhou X, Tang Z. Semin Surg Oncol 1998;14:171.

7. Tandan V, Harmantas A, Gallinger S. Can J Surg 1997;40:175.

8. Seifert J, Cozzi P, Morris D. Semin Surg Oncol 1998;14:175.

9. Scudamore C, Patterson E, Shapiro J, et al. J Invest Surg 1997;10:157.

10. Goldberg S, Solbiati L, Hahn P, et al. Radiology 1998;209:371.

11. Curley S, Izzo F, Deirio P, et al. Ann Surg 1999;230:1.

12. Rossi S, Di Stasi M, Buscarini E, et al. Am J Radiol 1996;167:759.