Background: The optimal dose of TNF-α delivered by isolated limb perfusion (ILP) in patients with locally advanced soft tissue sarcoma is still unknown.
Patients and methods: Randomised phase II trial comparing hyperthermic ILP (38–40°) with melphalan and one of the four assigned doses of TNF-α: 0.5 mg, 1 mg, 2 mg, and 3/4 mg upper/lower limb. The main end point was objective tumour response on MRI. Secondary end points were histological response, rate of amputation and toxicity. Resection of the remnant tumour was performed 2–3 months after ILP. The sample size was calculated assuming a linear increase of 10% in the objective response rates between each dose level group.
Results: One hundred patients (25 per arm) were included. Thirteen per cent of patients had a systemic leakage with a cardiac toxicity in six patients correlated with high doses of TNF-α. Objective tumour responses were: 68%, 56%, 72% and 64% in the 0.5 mg, 1 mg, 2 mg and 3 or 4 mg arms, respectively (NS). Sixteen per cent of patients were not operated, 71% had a conservative surgery and 13% were amputated with no difference between the groups. With a median follow-up of 24 months, the 2 year overall and disease-free survival rates (95% CI) were 82% (73% to 89%) and 49% (39% to 59%), respectively.
Conclusion: At the range of TNF-α doses tested, there was no dose effect detected for the objective tumour response, but systemic toxicity was significantly correlated with higher TNF-α doses. Efficacy and safety of low-dose TNF-α could greatly facilitate ILP procedures in the near future.