Marked variability of melphalan plasma drug levels during regional hyperthermic isolated limb perfusion

Abstract Background Hyperthermic isolated limb perfusion (HILP) with melphalan as treatment for locally recurrent or in-transit malignant melanoma is frequently performed but the principle for calculating drug dosage remains poorly understood. Methods This study examined the pharmacokinetic profile of 14 consecutive patients to determine what variables were associated with toxicity and tumor responses. Results Marked fourfold variability was noted in patient plasma melphalan concentrations. We defined a factor—the ratio of estimated limb volume (Vesti) to melphalan volume of distribution (Vss), Vesti/Vss—that was much more strongly correlated with acute regional toxicity than either area under concentration-time curve or peak plasma concentration. In addition, we found that AUX2 was the best correlate of tumor response. Conclusions Pharmacokinetic evaluation of prospective HILP trials is critical to not only understand response and toxicity outcomes but also to potentially improve the therapeutic index of regional perfusion.