to 184cm3 with a median value of 132cm3. These variations took into account differences in CTV delineation, TPS programand the expansion volume restricted by anatomical barriers like sustentorial volume or bone structures. The different CTV werecompared also to a “gold standard CTV” producing by a radiologist as expert.

Conclusions: This study demonstrated significant interphysician variability in producing Clinical Target Volume which is amajor step for 3DCRT planning process. This level of difference is high. These differences render necessary a clear explanationof different target volumes, a prospective control of quality procedure of each contoured CTV with definition of minor andmajor deviations. This asurance quality control will be extended to comparisons of dosimetric planning using DV Histogramsand to evaluate the intraphysician variability. Our ability to define correctly CTV in prospective multicentric protocol is oneof major steps for improving our results with 3DCRT for low-risk medulloblastoma.

255 Concurrent Chemotherapy and Low-Dose Craniospinal Irradiation Followed by Conformal Posterior FossaTumor Bed Boost for Average Risk Medulloblastoma: Efficacy and Patterns of Failure

J. Douglas1,2, J. Barker1, R. Geyer2, S. Lindsley1, R. Ellenbogen3

1Department of Radiation Oncology, University of Washington, Seattle, WA, 2Department of Pediatric Oncology,Children’s Hospital and Regional Medical Center, Seattle, WA, 3Department of Neurosurgery, Children’s Hospital andRegional Medical Center, Seattle, WA

Purpose/Objective: Concurrent chemotherapy and low-dose craniospinal irradiation (CSI) for children with average-riskmedulloblastoma is appropriate treatment following radical resection. The optimal treatment volume for the final high-doseposterior fossa boost (PFB), however, remains controversial. We reviewed our experience using a 3-D conformal PFB includingonly the tumor bed plus margin as the final treatment volume.

Materials/Methods: From 1994-2001, 34 children meeting the eligibility criteria for this retrospective review were treated atour institution. Patients included in this analysis were diagnosed with a pathologically-proven, non-metastatic (M0) primaryposterior fossa medulloblastoma, were treated with curative intent, and had adequate follow-up and available records for ourreview. Definitive treatment for these children included initial radical resection followed by concurrent vincristine and low-dose(2340 cGy) CSI; the tumor bed and/or posterior fossa was subsequently boosted to a median total dose of 5580 cGy (range5400-5940 cGy). When the PFB volume included the entire posterior fossa, lateral cranial fields were used (3 children); whenthe PFB volume included the tumor bed plus margin, multifield conformal radiotherapy was used (31 children). Followingradiotherapy completion, all patients completed additional maintenance combination chemotherapy. The median age atdiagnosis was 84 months (range 36-181 months), and the median follow-up for all patients was 23 months (range 3-87 months).

Results: Three-year actuarial progression-free survival was 90.5%. Sites of relapse for the 4 children who have developedprogressive disease were exclusively nonprimary; specifically, there were no isolated posterior fossa failures. Relapses wereobserved in 1 of 3 children treated with whole posterior fossa boost and in 3 of 31 patients treated with conformal posteriorfossa boost (p�0.31). Concurrent chemotherapy and low-dose CSI were associated with treatment-related nausea (44%),esophagitis (15%), and neutropenia (6%); however, no acute toxicity � grade 2 was seen among these children.

Conclusions: Treatment with a limited, 3-D conformal PFB following concurrent chemotherapy and low-dose CSI is effectiveand is not associated with high rates of relapse or local failures. This approach warrants a randomized trial as proposed by theChildren’s Oncology Group.

256 Craniospinal Radiation in the Treatment of Biopsy Proven Intracranial Germinomas: the Children’sHospital of Philadelphia (CHOP)/Hospital of the University of Pennsylvania (HUP) Experience

A. Maity1, L.N. Sutton2, H.G. Shu1, A. Janss3, J.B. Belasco3, L. Rorke4, P.C. Phillips3, J.W. Goldwein1

1Department of Radiation Oncology, Hospital of the University of Pennsylvania , Philadelphia, PA, 2Department ofNeurosurgery, Children’s Hospital of Philadelphia, Philadelphia, PA, 3Department of Pediatric Oncology, Children’sHospital of Philadelphia, Philadelphia, PA, 4Department of Pathology, Children’s Hospital of Philadelphia, Philadelphia,PA

Purpose/Objective: The best treatment for intracranial germinomas remains controversial with a wide range of opinionsregarding optimal radiation fields and doses and the role of chemotherapy. While many oncologists have moved towards usingfocal radiation therapy, the policy at our institution has still been to use craniospinal radiation because of the excellent resultswe have obtained. This study reports on our 25 year experience with this approach.

Materials/Methods: From 9/76 to 5/01, 40 patients with intracranial germinomas were treated at CHOP/HUP; however, 39 ofthese 40 received craniospinal radiation and were therefore further evaluated for this study. All tumors were initially imagedby CT (n�15) or MRI (n�24). Dissemination of disease was assessed radiologically by myelogram and/or MRI (n�32) andcytologically by examination of CSF fluid (n�35). All tumors were subsequently biopsied to establish the diagnosis. In twopatients the pathology was suspicious for mixed histology; but all other cases appeared to be pure germinomas. CSF and/orserum alpha-fetoprotein (AFP) levels were available on 25 patients and beta-human chorionic gonadotrophin (b-HCG) levelson 23 patients. No patient had an elevation of AFP, but five had an elevated b-HCG The median whole brain dose was 36 Gy(range 18–44.2 Gy), the median total dose to intracranial disease 50.4 Gy (range 44-55.8 Gy), and the median spinal dose 30.6Gy (range 18-40 Gy). Sites of spinal disease were boosted to a total dose ranging from 39.6-50.2 Gy. In general. patients withintracranial germinomas have not been treated with chemotherapy (CMT), but in two cases CMT was given because of thepossibility of a mixed germinoma. In addition, three other patients, all treated prior to 1992, also received CMT.

Results: The median age of patients at diagnosis was 15 years (range 8-43) with 9 females and 30 males. Nine out of 32 (28%)patients had evidence of spinal leptomeningeal seeding by radiologic imaging. Four out of 35 (11%) patients had positive CSFcytology. Of these four, only two had positive radiologic imaging. Overall, a total of 11 out of 36 patients (31%) had someevidence of spinal dissemination. With a median follow-up of 6.8 years (range 0.5-20.2 years), there have been no relapses. This

149Proceedings of the 44th Annual ASTRO Meeting