Radiofrequency ablation: variability in heat sensitivity in tumors and tissues.

PURPOSE

To characterize the thermal dosimetry (ie, heating profile) of radiofrequency ablation (RFA) in multiple ex vivo tissues and in vivo tumor models.

MATERIALS AND METHODS

RFA was performed for 3-24 minutes in ex vivo bovine livers (n=20), porcine kidneys (n=20), and turkey muscles (n=20) and in vivo canine venereal sarcomas (n=8). RFA was performed by using 1 and 3-cm long tips internally cooled electrodes. In addition, RFA was performed in in vivo R3220 rat mammary adenocarcinomas (n=36) and human renal cell carcinomas in nude mice (n=6) by using 1-cm monopolar electrodes. Continuous temperature monitoring was performed at multiple depths to calculate thermal dosimetry, reported as the area under the curve (AUC). Cumulative equivalent minutes at 43 degrees C (CEM43) were used for the critical ablation margin. Data were compared with analysis of variance and regression analysis.

RESULTS

For each tissue and/or tumor type, statistically significant temperature differences (up to 14 degrees) were observed at the ablation margin (P<.01). Temperature was dependent on the procedure duration. For 10-minute treatments, temperatures were significantly higher in the kidney compared with the R3230 tumor (72 degrees C+/-2.2) (P<.01) and lower in R3230 tumor (41.6 degrees C+/-1.4) (P<.05) but were similar for liver and muscle (51.6 degrees C+/-1.6 and 54.1 degrees C+/-1.8, respectively). Thus, a wide range of ablative temperatures were observed (41.0 degrees C+/-0.7 to 76.7 degrees C+/-1.9), with coagulation diameter correlating logarithmically with radiofrequency duration and AUC (R2=0.85-0.95). The CEM43 demonstrated an extreme range of values (10(11)).

CONCLUSION

The results of the study demonstrate a wide range of thermal sensitivity to RFA among commonly investigated tissues and tumor models, suggesting that further characterization of tissue-specific end points (ie, the duration and end temperature of ablation) is likely warranted. The AUC showed good correlation with ablation sizes, but the CEM43 proved unworkable given an extreme range of values for RFA.