Results: In terms of under-predicting, correctly predicting, and over-pre- dicting survival, faculty results were 6%, 25%, and 69%, respectively; residents were 3%, 36%, and 61%, respectively; nurses were 4%, 25%, and 71%, respectively; and therapists were 0%, 6%, and 94%, respectively. Excluding the radiation therapists, the accuracy of predicting survival was not statistically different among the faculty, residents, and nurses (P Z 0.343). Regarding the top 3 reasons for their prediction, faculty ranked type of cancer, tumor burden, and performance status, residents ranked type of cancer, tumor burden, and pace of disease, nurses ranked tumor burden, type of cancer, and pace of disease, and therapists ranked per- formance status, type of cancer, and pace of disease. Faculty most commonly prescribed particular fractionation schedules based on life ex- pectancy, concern of toxicity, and a need for rapid response, in descending order. Conclusion: The ability to predict survival among patients receiving palliative radiotherapy is overall poor, with caregivers mostly over-pre- dicting survival. Optimizing fractionation scheme will require improved integration of clinical information and development of new tools predictive of survival. Author Disclosure: S. Aggarwal: None. N.D. Prionas: None. J.N. Carter: None. P. Pradhan: None. J.L. Bui: None. R. von Eyben: None. C.K. Ho: None. S.L. Hancock: None. S.G. Soltys: None. A.C. Koong: None. D.T. Chang: None. 336 Clinical Outcomes of Palliative Radiation Therapy for Pediatric Oncology Patients A.D. Rao, 1 S.R. Alcorn, 2 J. Moore, 3 and S.A. Terezakis 4 ; 1 Johns Hopkins University School of Medicine, Department of Radiation Oncology & Molecular Radiation Sciences, Baltimore, MD, 2 Johns Hopkins University, Baltimore, MD, 3 Johns Hopkins University School of Medicine, Baltimore, MD, 4 Johns Hopkins University School of Medicine, Department of Radiation Oncology and Molecular Radiation Sciences, Baltimore, MD Purpose/Objective(s): Limited data exist detailing palliative radiotherapy (RT) experiences pediatric oncology patients. We aim to characterize the practice, efficacy of symptom control, and toxicity of palliative radiation therapy delivered at a single institution. Materials/Methods: We performed a retrospective review of all pedi- atric patients age < 25 years treated with RT with palliative intent, defined as with the goal of symptom control or prevention of impending, life-threatening progression, in a single institution over a 6.5-year span from July 2009-January 2016. In several instances, single patients received palliative RT to more than one site; hence, more cases are reported than total patients. Response for each case was determined by patient report at post-RT clinical encounters, dated at time of encounter. Results: Of 385 total pediatric patients treated with RT, 43 patients (11%) were treated with palliative intent to a total of 76 separate sites. Most common malignant histologies included leukemia/lymphoma (33%), rhabdomyosarcoma (18%), neuroblastoma (12%), retinoblastoma (9%), Ewing’s sarcoma (8%), osteosarcoma (5%), and other (15%). Seventy-one sites were treated at a median of 27.9 months (range Z 0.7-158.4 months) after diagnosis. Median survival after completion of palliative RT was 3.5 months (95% CI Z 1.2 to 5.9 months). Indications for palliative RT were as follows: 63% pain, 8% cord compression, 8% intracranial symptoms, 8% respiratory compromise, 7% asymptomatic lesions risking impending symptoms, 4% post-laminectomy, 1% bowel obstruction, and 1% cosmesis due to a facial lesion. Of the 76 sites treated, response data were available after 59 (78%) of the treatments. Of these 59 treatments, 74% resulted symptomatic relief (44% complete and 30% partial response), of which 36% responded prior to completion of the palliative RT course. Overall median time to response was 22 days (range, 2-96 days). Most commonly prescribed palliative RT regimens were 3 Gy x 10 (20%), 2 Gy x 10 (16%), and 4 Gy x 5 (11%). Symptom stabilization was achieved in 17%, and continued clinical progression occurred in 8% of RT cases. An exception was leukemia and lymphomadthe most common malignant histologies in this cohortdfor which 100% of treatment courses led to clinical responses. Overall, RT was terminated early in 11% of treatment courses due to clinical deterioration. One patient experienced RTOG grade 3 acute anorexia with 50 Gy in 25 fractions to a 15 cm pelvic sarcoma invading the bilateral ureters and rectum; otherwise, the palliative regimens were well- tolerated. Conclusion: Palliative RT is a useful, safe, and effective tool for pediatric oncology patients. These data may be used for education of other health care professionals to illustrate its merits. Future research on minimizing dose and number of fractions necessary for symptomatic relief in pediatric patients is warranted. Author Disclosure: A.D. Rao: None. S.R. Alcorn: None. J. Moore: None. S.A. Terezakis: None. 337 The Impact of Instrumentation Upon Local Control Following Spine Stereotactic Radiosurgery J.A. Miller, 1 E.H. Balagamwala, 2 C.A. Berriochoa, 2 L. Angelov, 3 J.H. Suh, 4 M.B. Tariq, 1 K. Yang, 5 A. Magnelli, 4 S. Soeder, 2 A.M. Mohammadi, 3 T. Djemil, 2 E.C. Benzel, 3 and S.T. Chao 2 ; 1 Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland, OH, 2 Department of Radiation Oncology, Cleveland Clinic, Cleveland, OH, 3 Department of Neurosurgery, Cleveland Clinic, Cleveland, OH, 4 Cleveland Clinic, Cleveland, OH, 5 Cleveland Clinic Lerner College of Medicine, Cleveland, OH Purpose/Objective(s): Spine stereotactic radiosurgery (SRS) has emerged as a safe and effective treatment for spine metastases. However, it is un- known whether this highly-conformal technique is suitable at instrumented sites. We hypothesized that in-field spinal instrumentation is not associated with decreased local control following SRS. Materials/Methods: A retrospective cohort study of patients undergoing SRS for spine metastases was conducted. Patients with less than one month radiographic follow-up were excluded. Each patient with in-field spinal instrumentation was propensity-matched to two controls without instru- mentation on the basis of demographic, disease-related, dosimetric, and treatment-site characteristics. Standardized mean differences were used to assess balance between matched cohorts. The primary outcome was 12- month local control, which was assessed every three months with spinal MRIs. Lesions demonstrating any in-field progression were considered local failures. Secondary outcomes of interest were post-SRS pain flare, vertebral compression fracture (VCF), and grade 3 toxicity. Kaplan- Meier analysis was used to determine actuarial local control, which was compared across cohorts via a log-rank test. Multivariate Cox proportional hazards modeling was then used to identify independent predictors of local failure. Results: Among 650 candidate controls, 166 were propensity-matched to 83 patients with instrumentation. Excellent balance was achieved among baseline characteristics. The median prescription dose was 16 Gy. The 12- month local control was not significantly different between cohorts (69% [instrumentation] vs. 75% [control], P Z 0.84, HR Z 1.05, 95% CI [0.65- 1.67]). Median times to local failure were also not significantly different (32 months [instrumentation] vs. 31 months [control], P Z 0.84). The presence of spinal instrumentation did not contribute to a multivariate model predicting local failure. Independent predictors of local failure included radioresistant histologies (HR Z 5.15, P Z 0.03), controlled systemic disease (HR Z 0.46, P < 0.01), greater number of treated vertebrae (HR Z 1.37, P < 0.01), and minimum PTV dose (HR Z 0.87, P Z 0.02). The incidence of pain flare (11% vs. 14%, P Z 0.55), post-SRS VCF (17% vs. 22%, P Z 0.41), and grade 3 toxicity (1% vs. 1%, P Z 1.00) were similar between cohorts. Conclusion: In this propensity-matched analysis, local control and toxicity were similar among patients with and without in-field spinal instrumen- tation, suggesting that instrumentation does not significant detract from Volume 96  Number 2S  Supplement 2016 Oral Scientific Sessions S149