Purpose: The Pediatric Oncology Group Protocol 9404 was a prospective clinical trial of two forms of chemotherapy in childhood T-cell acute lymphoblastic leukemia and advanced stage T-cell lymphoblastic non-Hodgkin's lymphoma. The protocol called for prophylactic C1 whole brain external beam irradiation, 18 Gy in 2 Gy/fraction for 9 fractions. We hypothesized that a correlation would be found between the number of children irradiated on protocol by an institution and the compliance rate of that institution with radiotherapy quality assurance (QA) guidelines. We also hypothesized that QA compliance would improve as the study progressed. Methods and Materials: We scored the radiation dose as a minor deviation from protocol guidelines if the dose to the prescription point differed from the protocol by 6-10%, and a major deviation if it differed from protocol by >10%. Treatment volumes were scored as a minor deviation if the margins were less than specified or the fields were excessively large. A major deviation was defined as the transection of a potential leukemia-bearing volume such as would be caused by blocking the cribriform plate, optic nerve, or temporal lobe. When the treating physician submitted a treatment plan and simulator film at the initiation of therapy to the Quality Assurance Review Center (QARC), a rapid turn-around review of the plan and suggestions for improvement was provided. At the end of therapy, all simulator and port films were reviewed at the QARC. Results: We reviewed the data from 353 patients treated at 73 institutions in the United States, Canada, and Europe. Of these patients, 2% (n = 7) were not assessable for QA because of incomplete information. Minor quality deviations were found in 27.7% of patients (n = 98) and major deviations in 7.9% (n = 28). The frequency of major deviations for institutions placing 1-4 patients on study was 11% vs. 5.5% for institutions placing 2:5 patients (p < 0.09). The frequency of minor deviations was 28.6% for institutions placing 1-4 patients on study vs. 27.1% for institutions placing >5 patients (p not significant). The frequency of major deviations fell over time (1996-1997, 15.5% vs. 1998-2001, 4.7%, p < 0.001). The frequency of minor deviations did not (1996-1997, 29.9% vs. 1998-2001, 26.9%, p not significant). Conclusion: For a relatively simple radiotherapy field, with clearly written protocol guidelines and rapid turn-around corrections from QARC, the rate of minor deviations was no different between institutions placing 1-4 patients on study and those placing ≥5. A trend (p < 0.09) was noted, however, for major deviations to decrease as a function of institutional experience, as well as over time (p < 0.001), supporting the validity of the hypothesis that pediatric clinical experience matters in QA for C1 whole brain leukemia radiotherapy. (C) 2002 Elsevier Science Inc.
