PII S0360-3016(02)02752-9 PHYSICS CONTRIBUTION IMPROVING THE CONSISTENCY IN CERVICAL ESOPHAGEAL TARGET VOLUME DEFINITION BY SPECIAL TRAINING PATRICIA TAI, M.B.,* JAKE VAN DYK, M.SC.,* JERRY BATTISTA,PH.D.,* EDWARD YU,PH.D., M.D.,* LARRY STITT, M.SC.,* JON TONITA, M.SC., † OLUSEGUN AGBOOLA, M.D., ‡ JAMES BRIERLEY, M.D., § RASHID DAR, M.B.,* CHRISTOPHER LEIGHTON, M.D.,  SHAWN MALONE, M.D., ‡ BARBARA STRANG, M.D., ¶ PAULINE TRUONG, M.D.,* GREGORY VIDETIC, M.D.,* C. SHUN WONG, M.D., § REBECCA WONG, M.B., # AND YOUSSEF YOUSSEF, M.D.** *Department of Oncology, London Regional Cancer Center, London, Ontario, Canada; † Epidemiology Department, Allan Blair Cancer Center, Regina, Saskatchewan, Canada; ‡ Department of Radiation Oncology, Ottawa Regional Cancer Center, Ottawa, Ontario, Canada; § Department of Radiation Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada;  Department of Radiation Oncology, Windsor Regional Cancer Center, Windsor, Ontario, Canada; ¶ Department of Radiation Oncology, Hamilton Regional Cancer Center, Hamilton, Ontario, Canada; # Department of Radiation Oncology, Toronto-Bayview Regional Cancer Center, Toronto, Ontario, Canada; **Department of Radiation Oncology, Kingston Regional Cancer Center, Kingston, Ontario, Canada Purpose: Three-dimensional conformal radiation therapy requires the precise definition of the target volume. Its potential benefits could be offset by the inconsistency in target definition by radiation oncologists. In a previous survey of radiation oncologists, a large degree of variation in target volume definition of cervical esophageal cancer was noted for the boost phase of radiotherapy. The present study evaluated whether special training could improve the consistency in target volume definitions. Methods and Materials: A pre-training survey was performed to establish baseline values. This was followed by a special one-on-one training session on treatment planning based on the RTOG 94-05 protocol to 12 radiation oncologists. Target volumes were redrawn immediately and at 1–2 months later. Post-training vs. pre-training target volumes were compared. Results: There was less variability in the longitudinal positions of the target volumes post-training compared to pre-training (p < 0.05 in 5 of 6 comparisons). One case had more variability due to the lack of a visible gross tumor on CT scans. Transverse contours of target volumes did not show any significant difference pre- or post-training. Conclusion: For cervical esophageal cancer, this study suggests that special training on protocol guidelines may improve consistency in target volume definition. Explicit protocol directions are required for situations where the gross tumor is not easily visible on CT scans. This may be particularly important for multicenter clinical trials, to reduce the occurrences of protocol violations. © 2002 Elsevier Science Inc. Esophageal cancer, Target volume definition, Training, Quality assurance, Three-dimensional (3D) radiation treatment planning. INTRODUCTION Accurate and reproducible target volume definition is a prerequisite for three-dimensional conformal radiation therapy (3D-CRT). The use of 3D-CRT reduces the dose to the normal tissues and may allow dose escalation to the tumor, thus possibly improving the clinical outcome. This approach requires precise definition of the target volume. In 1996, our team completed a trans-Canada survey that revealed a large degree of variation of target volume definition of cervical esophageal tumor among 48 radiation oncologists (1). The data were obtained by measuring the transverse diameter and length of volumes hand-drawn by the responding radiation oncologists on the computerized tomography (CT) scan films using a fine-tip pen. For the sites of brain, bladder, and prostate cancers, other investigators also found a great variation in target volume delineation (2– 4). The hypothesis of this study is Reprint requests to: Patricia Tai, Radiation Oncology Depart- ment, Allan Blair Cancer Center, 4101 Dewdney Avenue, Regina, SK S4T 7T1 Canada. Tel: (306) 766-2206; Fax: (306) 766-2845; E-mail: ptai@scf.sk.ca Presented in part at the first UK Radiation Oncology Conference (UKRO 1), 23–25 April 2001, York, UK. Supported by Research Fellowship Award from Cancer Care Ontario, Division of Research and Education. Acknowledgment—The authors thank Barbara Barons for her sec- retarial assistance and Matthew Schmid for his assistance with graphs. Received Jul 24, 2001, and in revised form Jan 10, 2002. Accepted for publication Jan 16, 2002. Int. J. Radiation Oncology Biol. Phys., Vol. 53, No. 3, pp. 766 –774, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/02/$–see front matter 766