Hello, my name is Vyachak Chandra, and I'm with the University of Florida Department of Neurosurgery in Gainesville, Florida. I'd like to talk to you today about the long-term outcomes of patients with brain metastases treated with LINAC-based radiosurgery. A little bit of background information. Metastases are, as we all well know, the most common cause of malignant CNS tumors. Historically, patients survived anywhere from one month with no treatment to about four to six months with whole-brain radiotherapy treatment from the initial diagnosis of brain metastases. Stereotypic radiosurgery uses ionized radiation delivered to tissue in order to destroy it without the use of an actual scalpel, initially developed in 1949 by Lars Lexell in Sweden. He continued to modify these methods and eventually created the gamma knife in 1968. Two major forms of stereotypic radiosurgery are gamma knife and the linear accelerator. A gamma knife bundles over 200 rays from radioactive sources by collimation, focusing this in an isocenter, and then you focus that isocenter on a patient's lesion, while the linear accelerator works a little bit differently, emitting X-rays. Tertiary collimators bundle these X-rays and then focus the rays for use on the patient's lesion, and there's only one source of these X-rays, and an emitter is moved around an actual patient's head. CNS's most recent guidelines came out in 2019 and have not changed significantly since the previous guidelines in 2016 for the use of stereotypic radiosurgery. They answered four major questions. Number one, should new patients with newly diagnosed metastatic brain disease undergo SRS compared to other treatment modalities? There's really only level three evidence saying that SRS is recommended as an alternative to surgery when surgery would lead to neuro deficits, and SRS should be considered to support palliative care. Should SRS be used after open surgical resection of brain mats? And there's also level three evidence showing that, yes, SRS should be used to decrease local recurrence rates. Next, should SRS alone be used in the management of patients with one to four brain mats? There's level three evidence saying that for solitary brain mats, SRS should be used to decrease local progression, and for patients with two to four brain mats, SRS is recommended instead of whole brain radiotherapy when the cumulative volume is less than seven milliliters. Finally, is SRS useful in the management of patients with more than four brain mats? And there's level three evidence supporting that it is useful to improve median overall survival when the cumulative tumor volume of more than four metastases is less than seven milliliters. We then looked at an initial SRS study done by Dr. Friedman here at the University of Florida. Looking at the initial set of patients from May 1989 all the way up to February 2006, included about 619 total patients that obtained Flanag base SRS for brain metastases. These patients were followed at regular intervals with MRI scans, local control and regional control were observed and tabulated, as well as overall survival. Overall survival was determined by date of death obtained through the Social Security Death Index. The results of this study showed that mean survival was about 7.9 months, with one and two year survival rates of 36 and 14% respectively. Local control was achieved in 84% of all lesions treated. The one and two year actuarial local control rates were 82 and 72% respectively. For our own study, we wanted to look at the survival duration of patients treated with SRS for brain mats over the entire 30 year period. Two of our main hypotheses were that post-SRS survival would be increased over the past 14 years, aka after the previous study was completed, due to improved dosing protocol and improvements in the treatment of systemic cancers. Our other hypothesis was that post-SRS survival would be decreased over the past 14 years as patient selection for radiosurgery has evolved towards treating sicker patients. Our patient characteristics were similar compared to the previous study. Age, sex, KPS, type of primary tumor, whether they have systemic metastatic disease, timing of brain metastasis onset, RPA class, number and location of metastases, treatment volume, dosage, etc. In terms of our patients, we had about 1,100 total patients that had reliable death data that we were able to obtain from the Social Security Death Index that were able to be included in our study. Majority of these patients had non-small cell lung cancer as their primary and various RPA classes as you can see below. We also had information on what dosing of radiation therapy they received. Our statistical analysis used Cox regression models, in this case looking at the overall survival to obtain hazard ratios for which risk factors would be most statistically significant in terms of predicting or correlating with overall survival. In this case, we were able to show that systemic metastases had the highest hazard ratio when RPA class was excluded with a hazard ratio of 1.53 being statistically significant. In a similar analysis, age, KPS, primary control, and systemic metastases were excluded, and in this case, being a male had the highest hazard ratio for patients having worse overall survival as well as prior whole brain radiotherapy patients. When looking at the Cox regression models for loss of local control with RPA class excluded, melanoma was shown to have the highest hazard ratio and be statistically significant, basically telling us that having melanoma leads to faster loss of local control which correlates to other studies done prior, and similar analysis was also done when excluding age, KPS, primary control, and systemic metastases. When looking at the Kaplan-Meier curves for local control, on the left, you can see that the median time to loss of local control for the entire study group was 18.4 months, and on the right side, we stratified that by treatment era to see if there was any noticeable difference. In era 1, in black, from 1989 to 2006, we can see the median time to loss of local control was 16 months, and in red, the 2006 to 2019 data showed an increased time to loss of local control of 24.2 months. We also looked at tumor volume versus SRS dose in the first set, first era from 1989 to 2006. You can see in black, and era 2 in red. When looking at overall survival by RPA class, we saw similar data relative to other studies showing that RPA class 1 and 2 patients did better than RPA class 3 in terms of overall survival. On the left side, you can see our overall survival Kaplan-Meier curve for all of our brain metastasis patients treated with SRS, showing a median overall survival time of 11 months, and on the right, we stratified this data into two based on the previous paper of 1989 to 2006 in black, and 2006 to 2019, showing increased overall survival, 8.1 months versus 23.5 months, which was statistically significant. There were some limitations to this study in that it was at a single institution, just the University of Florida, retrospective in nature. We didn't have good information on functional status or the cause of death for each patient. We also didn't know much about the impact on survival based on the advances in systemic chemotherapy, and that was difficult to quantify. Finally, we didn't have up-to-date Social Security death index data, as this data is up-to-date up to February of 2014, but limited after that. Finally, in terms of future directions for this study, we wanted to look at SRS dosage and whether or not there's been an increase in associated complications. As evidenced by our study, there's been an increase in dosing regimens over the past 10 years. We would also like to try to obtain regional control data in order to be able to fully quantify this compared to a previous study. Finally, I'd like to say thank you to my mentor, Dr. Kelly Foote, as well as other mentors, Dr. Friedman and Dr. Frank Bova at the University of Florida for allowing me to participate in this study.

brain metastases

LINAC-based radiosurgery

long-term outcomes

stereotypic radiosurgery

2019 guidelines