Predicting final lesion characteristics during MR-guided focused ultrasound pallidotomy for treatment of Parkinson's disease.

OBJECTIVE: Magnetic resonance-guided focused ultrasound (MRgFUS) ablation of the globus pallidus interna (GPi) is being investigated for the treatment of advanced Parkinson's disease symptoms. However, GPi lesioning presents unique challenges due to the off-midline location of the target. Furthermore, it remains uncertain whether intraprocedural MR thermometry data can predict final lesion characteristics. METHODS: The authors first performed temperature simulations of GPi pallidotomy and compared the results with those of actual cases and the results of ventral intermediate nucleus (VIM) thalamotomy performed for essential tremor treatment. Next, thermometry data from 13 MRgFUS pallidotomy procedures performed at their institution were analyzed using 46 degrees C, 48 degrees C, 50 degrees C, and 52 degrees C temperature thresholds. The resulting thermal models were compared with resulting GPi lesions noted on postprocedure days 1 and 30. Finally, the treatment efficiency (energy per temperature rise) of pallidotomy was evaluated. RESULTS: The authors' modeled acoustic intensity maps correctly demonstrate the elongated, ellipsoid lesions noted during GPi pallidotomy. In treated patients, the 48 degrees C temperature threshold maps most accurately predicted postprocedure day 1 lesion size, while no correlation was found for day 30 lesions. The average energy/temperature rise of pallidotomy was higher (612 J/ degrees C) than what had been noted for VIM thalamotomy and varied with the patients' skull density ratios (SDRs). CONCLUSIONS: The authors' acoustic simulations accurately depicted the characteristics of thermal lesions encountered following MRgFUS pallidotomy. MR thermometry data can predict postprocedure day 1 GPi lesion characteristics using a 48 degrees C threshold model. Finally, the lower treatment efficiency of pallidotomy may make GPi lesioning challenging in patients with a low SDR.