Proffered Papers— Radiation Oncology

REAL-TIME READIOTHERAPY REVIEW OF A RANDOMIZED TRIAL OF PRE- AND POST-OPERATIVE RADIOTHERAPY OF LOCALIZED SOT-TISSUE SARCOMA OF THE EXTREMITY

Catton, CN, Goddard, K, O’Sullivan, B, James, K. Departments of Radiation Oncology, The Princess Margaret Hospital, Toronto, and The Cancer Control Agency of British Columbia. National Cancer Institute of Canada (NCIC) Clinical Trials Group, Kingston, Ontario

Purpose: To evaluate the process and results of real-time radiotherapy (RT) review of the NCIC SR-2 randomized trial of pre and post-operative RT for localized extremity soft-tissue sarcoma (STS).

Material and Methods: The trial opened in 1994 and closed in 1997 after the planned interim analysis showed a significant difference in outcome for the primary endpoint between the 2 treatment arms. 10 centers entered 189 patients. Review of RT plans was required within 3 fractions from the start of therapy. Copies of simulator films, isodose distributions, prescription and dose calculations, set-up photo and diagnostic images were required for the review, and were to be couriered to the review centre before treatment. Plans were evaluated for compliance to dose and fractionation, and that the CTV margins met minimum requirements, and were covered by the 95% isodose line, and that dose distributions were homogeneous to + 5%. A report of protocol compliance, or non-compliance with recommendations for plan modification was faxed to the submitting center and mailed to the central trial office (CTO). The final decision for any changes was left to the treating oncologist. Records of the reviews and completed treatments were analyzed to determine protocol compliance for RT given during the trial, and to determine the effectiveness of the real-time review process.

Results: 5 patients did not receive RT, leaving 184 for analysis. The trial review rate was 161/184 (87%), as 23 eligible cases were not reviewed. 153 of 161 treatments reviewed (95%) met protocol standards, including 5 plans modified because of the review. 8 cases (5%) were not modified as recommended. Reasons were: unknown (5 cases), treatment completed before the review (2 cases), disagreement about the location of the GTV (1 case). Review was completed within the first 3 fractions in 90/161 cases (56%). Reasons for review not done in real time were: sufficient data, but submitted late 56/71 (78%), incomplete data submitted, but updated late 6/71 (8%), reviewer late in submitting review 9/71 (12%). Review was not done because data was not submitted, or was lost en-route (15/23) or incomplete data was never subsequently updated (8/23).

Conclusions: Extremity STS is one of the most difficult sites to plan for RT, often requiring complex, and individualized plans, and RT review was an essential quality control component for the SR-2 trial. The trial achieved 87% of treatments reviewed, with 95% of these meeting protocol requirements. The real-time review rate was only 56%. Failure was usually due to the short time available for the collection and couriering of data between simulation and treatment. For future trials, on-line data transmission should improve efficiency and reduce the risk of data being lost in transit.13% of cases were not reviewed, and notification from CTO to reviewers and centers of upcoming reviews might lessen the rate of forgotten and lost submissions. CTO should ascertain that non-compliant plans are modified to meet protocol, or that an explanation for non-compliance is recorded.