Purpose: Describe cluster models, normal-tissue complication probability models in which both the number and the spatial location of radiation-sterilized functional subunits play a role in defining complication probability.
Methods and materials: Computer simulation was used to determine the maximum size cluster of sterilized subunits associated with a given dose distribution. Complications were associated with large clusters.
Results: Cluster models showed a volume effect, as increasing effect for constant dose when the volume increased or constant effect when the dose was reduced with increasing volume. Cluster models gave similar results to existing models when tissues were irradiated uniformly. With inhomogeneous dose distributions, on the other hand, different spatial distributions of "hot spots" may lead to different predictions of complication probability by cluster models. The result was that a higher complication probability resulted when hot spots were contiguous (clustered) than when they were dispersed, even if both situations are characterized by the same dose-volume histogram. A potential advantage of cluster models is to provide an easy, internally consistent way to predict complications arising from the inhomogeneous dose distributions that sometimes arise with intensity-modulated radiotherapy.
Conclusion: Cluster models offer a new way to quantify complication probability in treatment situations in which a wide variety of hot-spot distributions occur.