Computer modeling of factors that affect the minimum safety distance required for radiofrequency ablation near adjacent nontarget structures.

PURPOSE: To use computer modeling of radiofrequency (RF) ablation to evaluate the effects of (i) composition and varying perfusion of intervening tissue and (ii) electrode orientation and type on the required distance to avoid heating damage of adjacent nontarget structures.

MATERIALS AND METHODS: Systematic three-dimensional finite-element computer simulation of RF heating (6-20 minutes) was performed (3,128 simulations). The distance (5-25 mm) between the electrode and the potentially injured structure and tissue composition as layers of tumor/soft tissue, fat, and/or fluid was varied (thermal conductivity, 0.46, 0.23, and 0.7 W/m- degrees C; electrical conductivity, 0.5, 0.1, and 1 S/m, respectively). Varying perfusion (0-5 kg/m(3)-s), electrode orientation (parallel or perpendicular), and electrode type (ie, noncooled and internally cooled 3-cm single or 2.5-cm cluster) were also studied. The time required to reach various temperatures (eg, the time to reach 50 degrees C designated as t50) and the distances at which the temperatures occurred and the distances required to avoid threshold temperatures at the margin of adjacent structures were compared.

RESULTS: In all cases, increasing the amount of intervening fat increased t50 compared with tumor/soft tissue and/or fluid. With no perfusion, 9 mm of fat or 14 mm of tumor/soft tissue or fluid was required for perpendicular insertion (internally cooled single electrode) to prevent a temperature of 50 degrees C with 12 minutes of heating, compared with 12 mm of fat or 23 mm of tumor/soft tissue or fluid for parallel insertion. Less intervening fat was needed for noncooled electrodes (<8 mm parallel, <5 mm perpendicular), with more intervening tissue required for cluster electrodes (>13 mm) for an RF application of 20 minutes. Finally, the amount of intervening tissue required to prevent damage also decreased linearly with increasing perfusion for each tissue and electrode (r(2) = 0.74 for parallel; r(2) = 0.98 for perpendicular).

CONCLUSIONS: In the computer model described in the present study, thermal and perfusion characteristics between the electrode and adjacent nontarget structures (specifically the presence of fat) and the electrode characteristics themselves (including parallel versus perpendicular insertion) have been shown to affect the minimum safe distance required for the prevention of thermal injury.