October 1995

Systemic Leakage and Side Effects of Tumor Necrosis Factor α Administered Via Isolated Limb Perfusion Can Be Manipulated by Flow Rate Adjustment

Author Affiliations

From the Departments of Intensive Care and Anesthesiology (Drs Sorkin, Halpern, Setton, and Rudich), Surgery B and C (Drs Abu-Abid, Lev, Gutman, Aderka, and Klausner and Mr Kudlik), and Nuclear Medicine (Mr Bar-On), Ichilov Hospital, Tel Aviv (Israel) Sourasky Medical Center; and Sackler Faculty of Medicine, Tel Aviv University, and the National Unit of Orthopedic Oncology, Soroka Medical Center, Beersheba, Israel (Dr Meller).

Arch Surg. 1995;130(10):1079-1084. doi:10.1001/archsurg.1995.01430100057012

Background:  The tolerated systemic dose of recombinant tumor necrosis factor α (rTNF-α) is very limited, since its administration leads to a severe septic shock—like condition. Its implementation in isolated limb perfusion (ILP) for metastatic melanoma or advanced soft-tissue sarcoma confined to the limb facilitates doses of rTNF-α 10 times higher than the systemic tolerated dose. However, with the traditional high flow rate used in ILP, systemic leakage and side effects are not eliminated.

Objective:  To determine if a lower perfusion flow rate would reduce leakage and consequently toxic effects.

Methods:  Isolated limb perfusion was performed for melanoma and soft-tissue sarcoma confined to the limb using a flow rate of 869 ±122 mL/min in nine patients (group 1) and a lower rate of 286±62 mL/min in six patients (group 2).

Results:  The systemic leakage rate was 12.5%±2.9% in group 1, compared with 2.3%±1.0% in group 2 (P=.003). Peak TNF-α levels were 29 000±2700 pg/mL in group 1, higher than 1580±1355 pg/mL in group 2 (P=.02). The tachycardia, hypotension, increased cardiac output, decreased systemic vascular resistance, bilirubinemia, elevation of liver enzyme levels, hypocholestrolemia, thrombocytopenia, and prolongation of prothrombin and partial thromboplastin times all observed in group 1 were significantly attenuated or eliminated in group 2. The limb Po2, Pco2, pH, and viability remained similar in both groups. Also, the tumor response rate remained high and was unaffected by the decrease in flow rate.

Conclusions:  Decreasing perfusion flow rate during ILP results in diminished leakage of TNF-α Consequently, the systemic hemodynamic, metabolic, and hematologic toxic effects are virtually abolished.(Arch Surg. 1995;130:1079-1084)