Laser-Interstitial Thermal Therapy as a Novel and Effective Treatment in Radiation Necrosis Following Stereotactic Radiosurgery to the Brain

Abstract

Laser-interstitial thermal therapy (LITT) has been proposed as an alternative treatment to surgery for radiation necrosis (RN) in patients treated with stereotactic radiosurgery (SRS) for brain metastases. The advantage of LITT is its minimal invasiveness and ability to provide same-procedure pathologic confirmation of RN. The present study sought to retrospectively analyze LITT outcomes in patients with RN from SRS. This was a single-institution retrospective study of 31 patients (32 cases) from 2011-2018 with pathologically-proven RN after SRS for brain metastases (n=29) or proximally treated extracranial lesions treated with external beam radiotherapy (n=2). A laser ablation system was used for all LITT cases. Same-day biopsy was performed in all cases. Same-day contrast-enhanced MRI was performed to create T1-weighted volumetric images. Patients were prospectively followed with FACT-Br, EQ5D, HVLT and clinical history and examination as part of an IRB-approved multi-institutional database. Imaging and clinical outcomes were determined via the EMR. Not all participants had quality of life (QOL) data at each visit and thus repeated measures linear models were used with visit, treated as a categorical measure, being the only covariate. From these models, adjusted means, standard errors and tests comparing visits to pre-LITT were generated. Kaplan-Meier method was used to estimate time to failure and overall survival. All analyses were done using a data management and decision management product. Median time from radiation to LITT was 12.6 months. Median SRS dose prior to RN was 20 Gy (IQR: 17-22). Primary cancers included lung (41%), non-melanomatous skin/head and neck (25%), breast (16%), melanoma (6%) and other (12%). Completion rates of questionnaires were 15/32 for EQ5D, 23/32 for FACT-Br, 8/32 for HVLT. Median volume of LITT target volume was 3.4 cc (IQR: 2.2-4.6 cc). Twenty-four of 32 (75%) LITT patients successfully tapered off steroids within 1 month. Three, 6 and 9 month freedom from RN recurrence was 95.8%, 91.0%, and 91.0%. Median survival of this population was 2.1 years from LITT. Two patients had recurrent RN, at 3.4 months (3.2 cc) and 1.9 months (2.9 cc). They were treated successfully with bevacizumab and with craniotomy, respectively. Mean KPS score decreased from baseline (84) to 3 month follow-up (75) (p=0.03). Mean Fact-BR scores, stratified by emotional, functional, physical and social/family well-being scores, were stable from baseline (6.0, 17.2, 6.4 and 22.0) to 3 months (4.9, 15.7, 9.4 and 21.9) (p=0.32, 0.32, 0.07 and 0.95, respectively). Mean HVLT-R Total Recall scores were stable from baseline (20.6) to 3 months (18.4) (p=0.09). Mean EQ5D Aggregate scores were stable from baseline (7.1) to 1 month (7.6) (p=0.25). LITT represents a safe and durably effective treatment option for RN in the brain. Patient reported outcomes showed no severe declines after LITT.