Materials/Methods: We retrospectively searched for patients treated with CSI for medulloblastoma between 1999 and 2011. The following items were collected: demographic information, stage, risk group, treatment position, dose, fractionation, and position shifts based on weekly port ﬁlms. Results: Of 46 patients identiﬁed, 23 (50%) were treated in the prone position, and 23 (50%) supine. The median age at completion of RT was 7.7 years. Thirty-two patients (70%) were categorized as standard-risk, and 14 as high-risk. Standard-risk patients received 23.4 Gy in 1.8 Gy fractions to the craniospinal axis, followed by either a posterior fossa boost of 12.6 Gy and tumor bed boost of 19.8 Gy, or solely tumor bed boost of 32.4 Gy. High-risk patients were treated with 36 Gy to the craniospinal axis, fol- lowed by a tumor bed boost of 19.8 Gy. Opposed lateral cranial ﬁelds were matched to the superior border of the spinal ﬁelds. Posterior fossa and tumor boosts were delivered using IMRT. Port ﬁlms for the craniospinal ﬁelds were checked weekly. There were a total of 138 ﬁlms from week 1, 132 in week 2, 126 in week 3, and 39 from week 4 of treatment. Seven out of 72 (9.7%) of cranial port ﬁlms in the supine position were rejected, compared to 25 of 73 (34%) of those in the prone position (p < 0.01). Twenty of 72 supine upper spine ﬁlms were rejected, compared to 16 of 72 (22%) in the prone position (p Z 0.44). Finally, 13 of 73 lower spine ﬁlms were rejected compared to 9 of 73 prone ﬁlms (p Z 0.36). The proportion of rejected cranial ﬁlms in the supine versus prone positions by week of treatment were as follows: week 1 (13% vs. 43%, p Z 0.02), week 2 (9% vs. 24%, p Z 0.18), week 3 (0% vs. 30%, p Z 0.01), week 4 (40% vs. 44%, p Z 0.88). The most common reasons for rejection of cranial port ﬁlms were to correct head tilt and to increase the margin on the cribriform plate and temporal lobes. There was no signiﬁcant difference between the proportion of rejected port ﬁlms in the prone and supine positions for spinal ﬁelds, for any week of treatment. The proportions by week for the upper spine were: week 1 (13% vs. 22%, p Z 0.45), week 2 (35% vs. 29%, p Z 0.67), week 3 (17% vs. 25%, p Z 0.28). The proportions by week for the lower spine were as follows: week 1 (22% vs. 9%, p Z 0.23), week 2 (17% vs. 19%, p Z 0.89), week 3 (13% vs. 10%, p Z 0.72). The most common reason for rejection of spine ﬁelds is adequacy of margin laterally at the spinal border. Conclusions: CSI in the supine position allows for increased reproduc- ibility of patient positioning for treatment of the cranial ﬁelds compared to treatment in the prone position. Author Disclosure: J. Verma: None. B.S. Teh: None. E.B. Butler: None. M. South: None. A.C. Paulino: None. 160 A Consensus Atlas for Whole Ventricular Irradiation for Pediatric Germ Cell Tumors: Survey Results and Guidelines R. Mailhot, 1 R. Rotondo, 2 E. Murphy, 3 P.A. Caruso, 2 B. Fullerton, 2 N.J. Tarbell, 2 T.I. Yock, 2 T.E. Merchant, 4 B. Donahue, 5 and S.M. MacDonald 2 ; 1 Washington University in St. Louis School of Medicine, Saint Louis, MO, 2 Massachusetts General Hospital, Boston, MA, 3 Cleveland Clinic, Cleveland, OH, 4 St. Jude Children’s Research Hospital, Memphis, TN, 5 Maimonides Medical Center, New York, NY Purpose/Objective(s): Future pediatric cooperative group protocols for CNS germ cell tumors, including non-germinomatous germ cell tumors and germinoma will require 3-dimensional planning and include the whole ventricle volume (WVV) as the initial target volume for all patients enrolled. Because these volumes are complex and no formal guidelines exist for their delineation, developing a WVV contour atlas would be helpful to improve consistency in targeting, quality assurance and protocol performance. We sought to develop a contouring atlas based on consensus deﬁnitions from cooperative group pediatric radiation oncologists. Materials/Methods: A survey with 21 questions was developed and distributed to a group of pediatric radiation oncologists. Survey responses were obtained from 49 participants and reported. Results: More than 50% of respondents were in favor of including the third ventricle, fourth ventricle, suprasellar cistern, and pineal cistern in the WVV contour; structural inclusion frequencies are displayed in the table below. IMRT was the preferred modality for treatment. Marked variability existed in responses concerning ﬁeld arrangements. An atlas was generated and guidelines developed on the basis of the survey results and subsequent discussion. Contours were delineated by a radiation oncologist and reviewed by other radiation oncologists and a neuroradiologist and neuroanatomist. Written guidelines were established. Conclusions: The survey results demonstrated consensus concerning structural conclusions but conﬁrmed variability among best practices for contouring, thus highlighting the need for consensus guidelines and a use of a contouring atlas. It is expected that the atlas generated will improve target volume consistency in cooperative group protocols. Author Disclosure: R. Mailhot: None. R. Rotondo: None. E. Murphy: None. P.A. Caruso: None. B. Fullerton: None. N.J. Tarbell: None. T.I. Yock: None. T.E. Merchant: None. B. Donahue: None. S.M. MacDonald: None. 161 Proton Radiation Therapy for High-risk Pediatric Medulloblastoma: Outcomes for Very Young Children Treated With Upfront Chemotherapy R.B. Jimenez, 1 T.I. Yock, 2 N. Depauw, 2,3 J. Adams, 2 D. Ebb, 4 B. Fullerton, 2 N.J. Tarbell, 2 and S.M. MacDonald 2 ; 1 Harvard Radiation Oncology Program, Boston, MA, 2 Massachusetts General Hospital, Department of Radiation Oncology, Boston, MA, 3 Center for Medical Radiation Physics, University of Wollongong, Wollongong, New South Wales, Australia, 4 Massachusetts General Hospital, Department of Pediatric Hematology and Oncology, Boston, MA Purpose/Objective(s): Medulloblastoma and supratentorial PNET (SPNET) are common childhood brain tumors. Historically, very young age has been associated with a poor prognosis. Additionally, some studies indicate that the sequencing of radiation and chemotherapy affects disease control. The purpose of this study is to report the early outcomes for very young children with medulloblastoma treated with upfront chemotherapy followed by proton radiation therapy (3D-CPT). Materials/Methods: All patients under 60 months with medulloblastoma or SPNET treated with chemotherapy prior to 3D-CPT from 2002-2010 included. All patients underwent maximal surgical resection, high-dose chemotherapy, and adjuvant 3D-CPT with either craniospinal irradiation (CSI) followed by involved ﬁeld radiation therapy (IFRT) or IFRT alone. Results: Fifteen patients (median age at diagnosis 37.1 months, range 23-55 months) were treated with high dose chemotherapy and 3D-CPT. Twelve of 15 received high-dose chemotherapy with stem cell rescue on or per a Head Start or COG protocol. Twelve of 15 received CSI + IFRT and the remaining 3/15 received IFRT alone. Median time from surgery to the initiation of radiation was 193 days (range 50-301 days). Median CSI dose was 23.4 Gy (RBE) (range 18-30.6); median boost dose was 54.0 Gy (RBE). Thirteen of 15 patients had medulloblastoma; 2/15 patients had SPNET. 10/15 had unfavorable characteristics including large cell/anaplastic histology, M2-4 disease, and/or subtotal resection. At a median of 31 months from completion of radiation (range 12-93 Oral Scientiﬁc Abstract 160; Table Structure Inclusion and Deﬁnition of WVV Yes (%) No (%) Unsure (%) Ependymal surfaces? 37 (75.6) 6 (12.2) 6 (12.2) Third ventricle? 49 (100) 0 (0) 0 (0) Fourth ventricle? 40 (81.6) 7 (14.3) 2 (4.1) Suprasellar cistern? 42 (85.7) 4 (8.2) 3 (6.1) Pineal cistern? 42 (85.7) 4 (8.2) 3 (6.1) Foramen of Magendie? 23 (47.9) 16 (33.3) 9 (18.8) Foramina of Luschka? 18 (37.5) 21 (43.8) 9 (18.8) Structure Inclusion and Deﬁnition of WVV ventriculostomy Yes (%) No (%) Only w/third ventriculostomy Pre-pontine cistern? 6 (13.9) 23 (53.5) 14 (32.6) Volume 84  Number 3S  Supplement 2012 Oral Scientiﬁc Sessions S65