More Comments on Hepatic Artery Infusion

The July issue of ONCOLOGY contained an article by Alan Venook, MD, entitled “Update on Hepatic Intra-Arterial Chemotherapy.” The commentaries accompanying the article were written by Lawrence Leichman, MD, and Steven Stain, MD, and by Nancy Kemeny, MD. Harold J. Wanebo, MD, of the Department of Surgery at Roger Williams Hospital in Providence, Rhode Island, had also been invited to prepare a commentary on the article. Dr. Wanebo’s commentary arrived too late to be included in the issue containing Dr. Venook’s article, but we are pleased to publish Dr. Wanebo’s comments this month.—The Editors

Isolated liver metastases are the first signs of disease recurrence in about 25% of patients with colorectal cancer.[1] Approximately 25% of these patients will have synchronous metastases at the time of primary resection, and the remaining patients will have metachronous metastases. In the United States, approximately 140,000 to 150,000 new cases of colorectal cancer are detected each year.[2] Twenty-three percent of patients are expected to develop isolated liver metastases at first site of recurrence, and an estimated 37,500 would be eligible for some type of regional hepatic therapy. Approximately 9% of metastases are solitary and 20% are potentially resectable (7,500 cases), with the rest being candidates for other therapies. Thus, the impact of liver metastases is substantial.

Resection and Long-Term Control

Many studies have confirmed that resection of solitary metastases results in long-term control in about 25% to 30% of patients.[1,3-7] This generally can be achieved with relatively low mortality (around 5%) in most experienced centers. Although it is technically possible to resect multiple metastases, optimal results are achieved in patients with three or fewer metastases.[6]

Other factors, such as location and stage of the primary tumor and the disease-free interval after the primary resection (synchronous vs metachronous) have a variable impact on survival.[4] Lesion size, extent of resection, and the presence of bilobar vs unilobar metastases have controversial effects on survival, depending on the series examined.[3-7] The optimal margin of resection (greater than 1 cm) was shown in the Hepatic Registry Study and other reviews to affect outcome significantly. Patients with margins less than 1 cm had half the survival rate of those with margins greater than1 cm (27.6% vs 44%).[4]

Extrahepatic disease, including involvement of the porta hepatis nodes, generally precludes resection. Of significance is the potential for further relapse in 50% to 78% of patients, thus highlighting the need for additional therapy even in favorably resected patients.[4]

Alternatives for Unresectable Disease

Dr. Venook focus on the outcome of hepatic artery infusion (HAI) in patients with unresectable liver metastases. A review of five major randomized series thatcompared HAI with floxuridine (FUDR) vs systemic infusion of fluorouracil(Drug information on fluorouracil) (or FUDR) showed HAI to have a superior overall objective response rate of 41%, as compared with a rate of 14% with systemic chemotherapy, but only a suggestive survival trend of 16 vs 12.2 months (P = NS).[8-9]

Dr. Venook provides his own analysis of these data. The presence of high objective response rates has not translated into significantly improved survival (in contrast to most chemotherapeutic series, in which patients with high response rates usually have significantly better survival than do nonresponders). Perhaps the extent of disease and the presence of occult systemic metastases are underappreciated and may account for some of these differences. The liver itself, however, may not be adequately controlled in spite of more aggressive therapy.

The toxicity of previous infusion studies is also highlighted. Biliary sclerosis, a severe and frequently fatal consequence of HAI with high doses of FUDR, can be reduced by careful monitoring of liver function and appropriate dose reduction during therapy. The addition of dexamethasone(Drug information on dexamethasone) has been associated with improved response and perhaps a lowering of toxicity. A recent study by Kemeny et al at Memorial Sloan-Kettering Cancer Center combining FUDR, leucovorin, and dexamethasone in HAI treatment of liver metastases achieved a very high response rates, 71%, and a median survival of 23 months. Strict monitoring of liver toxicity reduced biliary sclerosis from 12% to 3%.[10,11]

Dr. Venook points out the need for careful surgical techniques to minimize problems with misperfusion during HAI. Assessment of arteriograms, attention to detail in catheter placement, and adequate regional devascularization of collateral channels along the hepatic artery, stomach, and duodenum will minimize misperfusion and the gastrointestinal (GI) toxicity of HAI. Careful ligation of vascular channels to the bile duct from the hepatic artery may also lower the biliary sclerosis resulting from high drug concentrations at the hepatic artery. We use on-table arteriography by direct injection into the hepatic artery catheter so as to ensure adequate placement in the gastroduodenal artery. This method permits adequate infusion of the drug into the proper hepatic artery, and prevents streaming into one or the other hepatic artery. It also demonstrates that infusion into the liver replicates the pattern seen on the preoperative arteriogram. The use of fluorescein(Drug information on fluorescein) to exclude perfusion of the stomach and duodenum is essential. Infusion via the sideport of the pump with technetium pertechnetate (nuclear scan) guarantees that only the liver is perfused with technetium-99m-macroaggregated albumin (TcMAA scan).[12]

One issue that is sometimes overlooked is the fact that vascular recanalization can occur, accounting for the appearance of symptoms of GI toxicity suggestive of misperfusion into the gut (easily tested by infusion of methylene blue, the TcMAA scan, or radioisotope or which can be seen by the endoscopist). These problems can be relieved by selective vascular occlusion performed by the interventional radiologist.

More Randomized Trials Needed

Lastly, Dr. Venook highlights the Cancer and Leukemia Group B (CALGB) protocol designed to decisively answer whether HAI is superior to systemic chemotherapy. The lack of a crossover arm may hint at a conclusive answer.

The adjuvant trial of HAI vs systemic chemotherapy after liver resection (Eastern Cooperative Oncology Group (ECOG)-Intergroup Study) is also mentioned. These studies, when completed, will yield important therapeutic information useful for conventional therapies, as well as starting points for new therapeutic initiatives.

The real message here is that clinical oncologists need to participate in ongoing randomized protocols, help complete them in a timely manner, and seek answers to these treatment modalities. Regional liver infusion provides an excellent model for testing new drug therapies and for exploring therapeutic opportunities with biological agents or gene-modified agents.

An extension of the latter is isolated regional perfusion of the liver. Dr. Alexander et al at Surgery Branch of the National Cancer Institute have demonstrated that isolated perfusion with melphalan(Drug information on melphalan) (Alkeran), with or without tumor necrosis factor, is associated with high antitumor effects and has a greater potential for producing a complete response.[13] Perhaps the achievement of a complete response can lead to long-term survival in many patients and may ultimately be coupled with the best arm of the CALGB study.