Cystic Brain Metastases: Characteristics and Optimal Management.

PURPOSE/OBJECTIVE(S) Among patients (pts) with brain metastases (BM), the population of pts with cystic brain metastases (CBM) is poorly described. CBM present unique treatment dilemmas for symptom and disease control. We aimed to better describe the characteristics and treatment outcomes of these pts. MATERIALS/METHODS We retrospectively analyzed medical records of all pts diagnosed with BM and referred to radiotherapy (RT) at our institution between August 2017 and June 2020. We included pts with at least one CBM, defined as a ring enhancing lesion, comprising at least 50% of the diagnosed lesion with a diameter > 0.5cm. We excluded pts previously treated with brain RT and those with leptomeningeal disease. The largest cystic lesion was chosen as the target lesion for assessment. RESULTS Of 452 pts with BM referred to RT, 121 (26.7%) had CBM according to our aforementioned definition. Among pts with CBM, NSCLC (55.4%) and breast cancer (13.2%) were the most common primaries. 72.7% were diagnosed synchronously with the diagnosis of metastatic disease. Symptomatic disease led to diagnosis in 71.1% of cases. 51 pts (42.1%) underwent surgery prior to RT while 63 (52.1%) received RT alone. When comparing the two groups, surgical pts had higher rates of symptomatic BM (86.3% vs. 61.9%, P = 0.004), fewer BM and larger maximal lesion diameter. 12 pts were not treated with RT and excluded from the survival and response analysis. Non-focal RT (whole brain or posterior fossa RT) was delivered to 37.6% and focal RT (radiosurgery or fractionated RT) to 62.4%. Focal RT pts had better performance status (PS), improved control of systemic disease and fewer BM (median - 2 vs. 5 lesions, P < 0.001). Median survival (MS) in the treated cohort was 17 months. 1-year and 2-year survival rates were 60.7% and 40.9%, respectively. Survival significantly improved with PS, younger age, fewer BM and upfront surgery (MS-not reached vs. 13.6mo, P = 0.03). RT method did not significantly affect survival (focal 21.5mo vs. non-focal RT 12.3mo, P = 0.114). Disease control was only assessed in pts that had at least one post-RT follow-up imaging (n = 89). Overall response rate [complete response (CR), partial response (PR) and stable disease] of the target lesion was 95.5% (CR-24.7%, PR-62.9%). Only four pts did not respond to RT. CR was more common in pts who underwent surgery (P = 0.012). Response rates were not influenced by RT method (P = 0.179). During follow-up 49.4% of pts showed brain disease progression, which was also not affected by the RT method. 15 pts developed brain radionecrosis (RN) (13.7%), with median time to RN 11.3 months. CONCLUSION CBM were diagnosed in about a quarter of the patients with BM, with NSCLC being the most common primary. These patients benefited from upfront surgery followed by RT. The RT method did not affect survival, response rate, or progression; therefore, the cystic nature of the disease should not influence the choice of RT modality. AUTHOR DISCLOSURE D. Limon: None. L. Ferro: None. T. Kohen: None. R. Grossman: None. I. Strauss: None.