| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Department of Surgery, Beth Israel Medical Center, Albert Einstein College of Medicine, New York, New York.
Correspondence: Address correspondence to: Ronald S. Chamberlain, MD, MPA, FACS, Beth Israel Medical Center, Albert Einstein College of Medicine, 10 Union Square East, Suite 2M, New York, NY 10803; Fax: 212-844-8401; E-mail: rchamber{at}chpnet.org
In this months issue of the Annals of Surgical Oncology, Cheng et al.1 describe their success at placing hepatic arterial infusion (HAI) pumps in 38 consecutive patients using an entirely laparoscopic approach. In addition to placement of the pump, 68% of patients (N = 26) also had laparoscopic radiofrequency ablation (RFA, N = 24), liver resection, or both (N = 2). The 38 patients in the study group were identified from an initial cohort of 56 patients who were determined to have unresectable metastatic colorectal cancer based on tumor number, bilobar distribution, proximity to major vascular, or biliary structures precluding a negative margin resection, or high risk surgical intervention. After preoperative imaging studies failed to identify extrahepatic disease in any of the 56 patients, all had diagnostic laparoscopy and laparoscopic ultrasound. Of these patients, 32% (N = 18) were found to have histopathologic confirmation of peritoneal disease (N = 18) or lymph node disease (N = 8) that led to planned HAI placement being aborted. The authors do not tell us what imaging studies were used to determine resectability or extrahepatic disease preoperatively; however, a 32% rate of occult M1 disease, although high, is within the range of that reported by others.2 Despite having little to do with the premise of this paper, these results alone should provide additional motivation to all surgeons who do major hepatic resections (particularly for metastatic disease or foregut tumors) to consider initial laparoscopic evaluation to exclude occult M1 disease prior to laparotomy.
Although Cheng et al.1 are not the first to describe a laparoscopic approach to HAI pump placement, their review of 38 consecutive cases, performed by well-recognized expert laparoscopic surgeons, provides a comprehensive view of the technical challenges, overall feasibility, safety, and outcome of this approach. Most notably, HAI pump placement was accomplished in 100% of patients in whom it was attempted. According to their protocol, all patients had preoperative celiac and mesenteric angiography to provide a road map for pump placement. Twenty patients had "normal arterial anatomy" in which a common hepatic artery (CHA) (from the celiac axis) gave rise to a single gastroduodenal artery (GDA) that was cannulated for pump placement. Interestingly, 18 patients had aberrant arterial anomalies: 10 accessory or replaced arteries, a left hepatic artery arising from the CHA, or a variety of other anomalies. Despite these anomalies, the authors were to able to cannulate the GDA in 16 of 18 patients using a combination of preoperative embolization or intraoperative ligation of the accessory or replaced vessels; in two instances, cannulation of the LHA was required. Although not the focus of this paper, the authors remind those surgeons not well initiated in HAI pump placement that nearly all replaced or accessory vessels can be ligated with the expectation that the remaining arterial inflow will provide complete hepatic perfusion.
Although the authors of this paper are clearly talented laparoscopic surgeons and have demonstrated the feasibility of this operation, the bigger question is "so what?" By asking "so what," I am not referring to whether the average surgical oncologist will ever be able to perform this procedure. In fact, I remain confident that over time, as most have mastered laparoscopic RFA alone, they will master laparoscopic HAI pump placement as well. I am also not questioning whether the average surgical oncologist would have the tremendous patience necessary to learn and perform this operation, although a mean operating room time of 5 hours and 37 minutes, with a range between 3 hours and 40 minutes to a high of 11 hours and 46 minutes, does give me pause. The learning curve is likely steep and it is unclear what diminution in operating room time can be expected. Rather, when I ask "so what," I raise a more global question as to what the role of HAI pump placement will be in the foreseeable future for the patient group in this study, namely, patients with unresectable hepatic colorectal metastases. If, as I believe, the proved role of HAI pump in this group is tentative at best, then the question must be asked whether it is worth retraining a whole discipline to learn to perform this operation laparoscopically.
Allow me to briefly put forward my argument. Although readily adopted by a core group of surgical and medical oncologists, little to no level one evidence supports the notion that HAI therapy results in improved disease-free survival, overall survival or quality of life in patients with unresected colorectal hepatic metastases. Indeed, although it has been 35 years since Sullivan and Zurac3 first reported the use of prolonged ambulatory hepatic infusion of floxuridine in patients with hepatic metastases from colorectal cancer, the role HAI can play, if any, in the treatment of metastatic colorectal cancer is far from resolved. Although a superior hepatic response rate has been reported in a variety of single institutional studies (up to 80%),4,5 a metaanalysis of published HAI trails showed a 41% response rate for hepatic infusion therapy, with no survival advantage.6,7 Although surgical refinements in HAI therapy have solved many of the technical issues and reduced pump complication rates, simultaneous advances in systemic chemotherapy for metastatic colorectal cancer have also occurred. Indeed, contemporary treatment of metastatic colorectal cancer with fluorouracil and irinotecan (FOLFIRI) or oxaliplatin (FOLFOX) is associated with hepatic response rates of between 34% and 51% in most studies, which is superior or at least equivalent to that shown in the HAI metaanalysis. Equally as exciting, recent publications using newer systemic chemotherapies have demonstrated clinically important responses to the point of converting patients who are unresectable to resectable.7–9
In conclusion, it seems unlikely that we will ever be able to define a precise role for HAI in patients with unresectable hepatic colorectal metastases because the "ideal" chemotherapeutic cocktail with which to compare it remains a moving, albeit exciting, target. I must resuscitate the old surgical axiom that "just because we can, doesn’t mean we should." In treating patients with hepatic metastases from colorectal cancer, the focus should always be on the patient, not on the liver. Hepatic response rates for FOLFOX and FOLFIRI regimens alone are well within that reported for most HAI trials and this should give pause to any surgeon considering placing HAI pumps in this patient population. Despite what may be great proficiency at pump placement by many experienced surgeons, "an operation not worth doing is not worth doing well."
Received for publication April 5, 2004. Accepted for publication April 12, 2004.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
