Utility of Short Initial MRI Brain in Brain Metastases Patients Treated With Stereotactic Radiosurgery

Abstract

The optimal timing of interval MRI surveillance following stereotactic radiosurgery (SRS) to detect most new lesions while asymptomatic and treatable by additional radiosurgery is unknown. As frequent MRI follow-up is an important patient burden and contributor to cost of care for brain metastases, we aimed to characterize the utility of early initial post-treatment MRI brain surveillance and to identify patient characteristics that might be used to risk-stratify patients who may benefit from early surveillance scans.An IRB-approved retrospective single-institution cohort of patients with brain metastases treated with frameless SRS was analyzed. Baseline demographics, treatment characteristics, and timing of follow-up MRI findings were reviewed in the electronic medical record. Fisher's exact test and Pearson's Chi-square test were used to compare proportions. Mann-Whitney-U test was used to compare continuous variables.One hundred consecutive patients treated from January 2018 - March 2020 were included, with plan for 6-week follow-up MRI post-treatment. Thirteen lacked follow-up MRI brain, leaving 87 for analysis. Median time to MRI was 6.7 weeks, with 78% of scans obtained less than 8 weeks post-SRS. Twenty-four (28%) patients had findings concerning for new or progressive disease on follow-up MRI. Nine were symptomatic. Of the 24, ten (42%) received SRS, 1 (8%) received whole brain radiotherapy (WBRT), 1 (4%) received surgery, 1 (4%) started systemic therapy, 7 received a short interval MRI scan, and 3 (12%) went to hospice. Of the 10 who received SRS, 3 were symptomatic, 6 had lesions ≥ 5 mm, and all 10 had lesions ≥ 5 mm. Of the 7 recommended for short interval MRI, 1 received SRS, 1 received WBRT, and 5 had no further evidence of disease progression, resuming routine surveillance. Comparing patients with suspicious findings to those with stable disease on follow-up MRI, there was a statistically significant difference (Table) in the distribution of histology (P = 0.02) and performance status (P = 0.047), but no difference in number of lesions, extra-cranial disease control, prior receipt of SRS, or use of targeted therapy or immune checkpoint inhibitors.A small but significant number of brain metastases patients treated with SRS may have additional therapy based on initial short interval surveillance MRI scans. Further investigation in larger cohorts may help identify patient characteristics that predispose to benefit from short interval MRI scans, and determine the value of earlier detection of new lesions compared with a longer interval to first scan.