Preclinical assessment of a noncooled MR thermometry–based neurosurgical laser therapy system

Abstract

OBJECTIVE MRI-guided laser interstitial thermal therapy (MRgLITT) has recently gained interest as an ablative stereotactic procedure for intractable epilepsy, movement disorders, and brain tumors. Conventionally, a LITT system consists of a laser generator and cooled laser applicator, which is a fiber optic core surrounded by a sheath through which cooled fluid is pumped. However, this footprint can make the system bulky and nonmobile, limit the maximum depth of targeting, and increase the chances of breakdown. Herein, the authors conduct a preclinical assessment of a noncooled MRgLITT system in a porcine model. METHODS Three-tesla MRI was used to guide the in vivo placement of noncooled laser applicators in the porcine brain. The study consisted of a survival arm and terminal arm. The laser was activated at a power of 4–7 W for ≤ 180 seconds. Temperature changes were monitored using the MR thermometry software ThermoGuide in the survival arm (n = 5) or both ThermoGuide software and adjacently inserted thermal probes in the terminal arm (n = 3). Thermal damage was determined by the software using the temperature-time relationship of cumulative equivalent minutes at 43°C (CEM43). Temperatures calculated by the software were compared with those recorded by the temperature probes. The dimensions of thermal damage thresholds (TDTs; 2–9, 10–59, 60–239, ≥ 240 CEM43 isolines) given by MR thermometry were compared with the dimensions of irreversible damage on histopathological analysis. RESULTS There was a strong correlation between temperature recordings by ThermoGuide and those by thermal probes at both 4 mm (r = 0.96) and 8 mm (r = 0.80), with a mean absolute error of 0.76°C ± 2.13°C and 0.17°C ± 1.65°C at 4 and 8 mm, respectively. The area of 2–9 CEM43 was larger than the area of irreversible damage seen on histopathological analysis. The dimensions of the 10 and 60 CEM43 correlated well with dimensions of the lesion on histopathological analysis. A well-defined border (≤ 1 mm) was observed between the area of irreversible damage and healthy brain tissue. CONCLUSIONS This preclinical assessment showed that the noncooled LITT system was able to precisely reach the target and create well-defined lesions within a margin of safety, without any adverse effects. MR thermometry software provided an accurate near-real-time temperature of the brain tissue, and dimensions of the lesion as visualized by the software correlated well with histopathological findings. Further studies to test the system’s efficacy and safety in human subjects are in progress.