A Multi-Institutional Experience of Isolated Limb Infusion: Defining Response and Toxicity in the US

Abstract

Isolated limb infusion (ILI) is a minimally invasive approach for treating in-transit extremity melanoma, with only two US single-center studies reported. Establishing response and toxicity to ILI as compared with hyperthermic isolated limb perfusion is important for optimizing future regional chemotherapeutic strategies in melanoma. Patient characteristics and procedural variables were collected retrospectively from 162 ILIs performed at 8 institutions (2001 to 2008) and compared using chi-square and Student's t-test. ILIs were performed for 30 minutes in patients with in-transit melanoma. Melphalan dose was corrected for ideal body weight (IBW) in 42% (n = 68) of procedures. Response was determined at 3 months by Response Evaluation Criteria in Solid Tumors; toxicity was assessed using the Wieberdink Limb Toxicity Scale. In 128 evaluable patients, complete response rate was 31%, partial response rate was 33%, and there was no response in 36% of patients. For all patients (n = 162), 36% had Wieberdink toxicity grade >or=3 with one toxicity-related amputation. On multivariate analysis, smaller limb volumes were associated with better overall response (p = 0.021). Use of papaverine in the circuit to achieve cutaneous vasodilation was associated with better response (p < 0.001) but higher risk of grade >or=3 toxicity (p = 0.001). Correction of melphalan dose for ideal body weight did not alter complete response (p = 0.345), but did lead to marked reduction in toxicity (p < 0.001). In the first multi-institutional analysis of ILI, a complete response rate of 31% was achieved with acceptable toxicity demonstrating this procedure to be a reasonable alternative to hyperthermic isolated limb perfusion in the management of advanced extremity melanoma.