Long-Term Blood Brain Barrier Permeability Changes After Radiosurgery and Immunotherapy for Melanoma Brain Metastases

Abstract

In the era of immunotherapy (IT) and longer patient survival, differentiating radiation necrosis (RN) from recurrent or progressive tumor (RPT) is a major diagnostic challenge due to similar appearance on MRI. Ktrans is a value calculated from dynamic contrast enhanced (DCE) MRIs correlating with vascular permeability, such as the breakdown of the blood brain barrier (BBB). The value of ktrans can be elevated above baseline (<0.04 min-1) with either RPT (>3.6 min-1) or RN. The purpose of this study was to investigate the pattern of ktrans changes in patients receiving SRS and IT with the hope of utilizing ktrans to help distinguish RN from RPT. We retrospectively reviewed DCE brain MRIs of 10 patients with melanoma brain metastases (MBM) following treatment with SRS and IT. 3D T1-weighted spoiled gradient echo DCE perfusion imaging was performed on a 3-Tesla MRI. Ktrans was measured using FDA approved software (OleaSphere, Olea Medical Solutions Inc Cambridge, MA). A region of interest was placed on T1 enhancing tissue on ktrans maps fused with T1 post-contrast images. A mean ktrans was assigned to each lesion on each scan over time, and the highest (mean) value was used for this analysis. A binary logistic regression to evaluate the post-treatment correlation of the ktrans value with RN was performed. A median of 4.5 cycles of ipilimumab were delivered (range 2-14) with a median of 21Gy (16-24Gy). A median of 2.5 DCE images per patient were reviewed (range 1-7), and these were acquired at a median of 17.9 months after SRS (range 0.9-87 months). The median follow up for this group of 10 patients was 32 months (range 5.4-80 months), and the mean survival was 63 months (95% CI 44-82).The median ktrans(mean) for all patients who received SRS/IT was found to be elevated above baseline at 0.35 min-1 (range 0.13-4.57 min-1). Six of 10 patients also had radiographic evidence of RN [3/6 pathologically proven] diagnosed at a median of 11.9 months after SRS (range 3.6-65.1months). The peak ktrans was seen at 24.7 months after SRS in patients with RN vs 17.9 months in those without RN (p=0.156). Patients with RN had peak ktrans(mean) value of 0.42 min-1 vs 0.285 min-1 in those without RN (p=0.12). This small, preliminary study shows that ktransmean for patients treated with SRS and IT continues to evolve after radiographic diagnosis of RN and peaks at 24.7 months after SRS. The value of ktrans trends higher in those with RN (0.42 min-1 vs 0.285min-1), but it was not statistically significant. This is the first study to investigate the long-term change in ktrans (permeability) after radiosurgery and immunotherapy. This information may be valuable when deciding on the management of enlarging brain metastases in the setting of immunotherapy.