Background: Hyperthermic isolation limb perfusion (HILP) with tumor necrosis factor alpha (TNFalpha) and IFNgamma was pioneered by Liénard and Lejeune in 1988. The TNFalpha was empirically employed at a dosage of 3-4 mg, that is ten times the systemic maximum tolerated dose (MTD). After eighteen years from its first clinical application, more than 300 patients have been treated. The aim of this study is to clarify two major arguments: the TNFalpha dose and eligibility criteria for patient selection.
Patients and methods: A phase I-II study has previously been conducted in 20 patients with in-transit melanoma metastases using a combination of melphalan and TNFalpha at dosages ranging from 0.5 to 3.3 mg. Twenty patients were treated and a complete pathological response of 70% was recorded, with no correlation between tumor response and TNFalpha. The dose of 1 mg of TNFalpha provided the best results regarding efficacy and toxicity. On the basis of this results a large phase II SITILO study was undertaken. Patients with stage IIIA - IIIAB (presence of in transit metastases and/or regional node involvement) were considered eligible; a total of 113 patients were enrolled in the study. The disease was bulky (> 10 nodules or fewer nodules with a diameter > or = 3 cm) in 42.5% of the patients and unresectable in 33%. Forty patients were treated with a TNFalpha dosage > 1 mg and 73 with 1 mg. All the patients were submitted to HILP via axillary and iliac vessels for tumor of upper and lower limb, respectively. TNFalpha was injected in the extracorporal circuit at the pre-established dose, followed after 30 minutes by melphalan (13 and 10 mg/L of limb volume for upper and lower limbs, respectively).
Results: A grade 1 and 2 limb toxicity was found in 52.9% and 30.1% of the patients, respectively, 5.5% of patients exhibited a grade 3 and 4, whereas grade 5 limb toxicity was not found. The complete and partial responses were 63% and 24.5%, respectively, with an objective response of 87.5%. We tried to correlate the typed tumor response (CR or not CR) and the TNFalpha dosage < or = 1 mg or > 1 mg, but no statistically significant difference was found between the two groups. The bulky disease was the only prognostic factor able to influence the tumor response.
Conclusion: Only patients with bulky melanoma disease can benefit from HILP with TNFalpha at a low dose of 1 mg.