A BRAIN-TUMOR DOSE-ESCALATION PROTOCOL BASED ON EFFECTIVE DOSE EQUIVALENCE TO PRIOR EXPERIENCE

Purpose: The current study describes the design of a dose escalation p rotocol for conformal irradiation of primary brain tumors that preserv es the safe experience of a previous, sequential dose escalation schem e while enabling the delivery of substantially higher effective doses to a central target volume. Methods and Materials: Normalized isoeffec tive composite dose distributions were formed for 20 patients treated on the original protocol (which specified three progressively smaller planning target volumes [PTVs]) using the linear quadratic model there corrected to equivalent 2 Gy fractions using alpha/beta = 10 Gy). The se distributions were investigated and a new protocol was designed to preserve a similar level of efficacy and lack of toxicity far the oute r volumes, but allowing a higher dose to the inner PTV. Treatment plan s were then investigated to determine if the objectives of the new pro tocol were achievable. In particular, plans that simultaneously achiev ed all biological treatment planning objectives tall fields treated ea ch day) were investigated. Finally, the success of the protocol design was demonstrated by analysis of the effective dose distributions of 1 0 patients treated using the new protocol. Results: The composite norm alized isoeffective minimum doses to the outer PTVs (PTV3 and PTV2) in the original protocol were close to 60 Gy and 75 Gy, respectively, an d these values are specified as the minimum doses to those volumes for the new protocol. Homogeneity requirements to maintain equivalence fo r the outer target volume domains are: not more than 25 % of [PTV3 exc lusive of PTV2] >75 Gy; and not more than 50% of [PTV2 exclusive of PT V1] >85 Gy. Treatment plans using multiple noncoplanar arrangements of beams and static intensity modulation treat all volumes at each sessi on. DVHs of the normalized isoeffective dose distributions reveal the equivalence of the new protocol plans to the sequential plans in the p revious protocol as well as the ability to achieve a higher dose of 90 Gy to the isocenter of PTV1 (+/- 5% homogeneity required). Conclusion : The ability to incorporate past experience through use of the linear quadratic model in the design of a new dose escalation protocol is de monstrated. (C) 1998 Elsevier Science Inc.