Optimization of intensity-modulated 3D conformal treatment plans based on biological indices.

To overcome the limitations of the intensity modulation optimization techniques based on dose criteria, we introduce a method for optimizing intensity distributions in which we employ an objective function based on biological indices. The objective function also includes constraints on dose and dose-volume combinations to ensure that the results are consistent with the physician's judgement. We apply a variant of the steepest-descent method to optimize the objective function. The method is three-dimensional and incorporates scattered radiation in the optimization process using an iterative scheme employing the pencil beam convolution method. Previously we had shown that the inverse technique of obtaining optimum intensity distributions, for which the objectives are defined in terms of a desired uniform dose to the target volume and desired upper limits of dose to normal organs, produces satisfactory approximations of the desired dose distributions for prostate plans. However, for lung, the performance of this technique was considerably inferior. Our conclusion was that, in general, it is not sufficient to specify the objectives of optimization purely in terms of a desired pattern of dose and that the objectives should also incorporate biology, perhaps in the form of biological indices. We demonstrate that the biology-based approach produces lung plans that are superior to those produced when only dose-based objectives are used. For the treatment of prostate, the two methods produce comparable dose distributions.