Allen et al. (59) reported on pump complications in 544 patients treated at MSKCC between 1986 and 2001. The overall pump complication rate was 22%. These complications consisted of arterial thrombosis (6%), extrahepatic perfusion (3%), incomplete hepatic perfusion (2%), and hemorrhage (2%). However, the complications during the earlier half of the study period (1986-1993) were significantly

higher (25%) than the later half of the study time (1994-2001, 18%, P=0.05). The majority of complications were also salvaged, with 80% of pumps functioning for a minimum Inhibitors,research,lifescience,medical of 2 years. Overall pump failure rates were 9% at 1 year and 16% at two years. Hepatobiliary toxicity is the most serious and dose-limiting complication of HAI. It occurs at a higher incidence with FUDR (60). Elevation of serum transaminase Inhibitors,research,lifescience,medical levels is often the first sign of hepatotoxicity. Increases in alkaline phosphatase and Vemurafenib manufacturer bilirubin are more serious and show evidence of more significant hepatic

damage. Therefore, changes in liver functions should be monitored Inhibitors,research,lifescience,medical regularly during treatment with HAI FUDR. A dose-adjusting algorithm has been devised based on changes in liver function tests (61). Addition of dexamethasone to HAI of FUDR may reduce incidence of bilirubin elevation and also increase the rate of treatment response as demonstrated by Kemeny et al. (6,7). In the adjuvant pump studies at MSKCC, more than twofold increase in alkaline phosphatase levels was observed in 14% to 43% of patients. Total bilirubin elevation > 3.0 mg/dL was seen in 0 to 19%, and biliary stents were placed in 0 to 8%. A higher incidence of biliary toxicity was seen in Inhibitors,research,lifescience,medical the study where FUDR dose was 0.14 (as compared to 0.12 in the newer studies). In a new study that was recently published in JCO, patients

Inhibitors,research,lifescience,medical were randomized to receive Bev versus no Bev in addition to HAI + FOLFOX or FOLFIRI. In the group that received Bev, bilirubin ≥3 mg/dL was seen in 5 of 35 versus zero of 38 patients (P=0.02) and biliary stents were placed in four versus zero patients (P=0.05). This study was terminated early due to biliary toxicity. Biliary sclerosis is not observed with HAI of 5-FU (58) which tends to associate more with increased risk of myelosupression (58). Therefore, one logical approach to reduce biliary toxicity is to alternate between HAI FUDR and HAI 5-FU. Davidson et al. (62) used HAI FUDR at a dose of below 0.1 mg/kg/day for seven days followed by HAI boluses of 5-FU 15 mg/kg on days 14, 21, and 28. With this schedule, 12 (21%) patients had temporary liver enzyme elevations and only 2 patients (3.5%) developed biliary sclerosis. In another study, HAI FUDR was administered for seven days, and HAI 5-FU bolus was given on days 15, 22 and 29, with the cycle repeated every 35 days. None of the patients in this study had treatment terminated because of hepatobiliary toxicity (63).