Radiobiological Framework for the Evaluation of Stereotactic Radiosurgery Plans for Invasive Brain Tumours

This study presents a radiobiological formalism for the evaluation of the treatment plans with respect to the probability of controlling tumours treated with stereotactic radiosurgery accounting for possible infiltrations of malignant cells beyond the margins of the delineated target. Treatments plans devised for three anaplastic astrocytoma cases were assumed for this study representing cases with different difficulties for target coverage. Several scenarios were considered regarding the infiltration patterns. Tumour response was described in terms of tumour control probability (TCP) assuming a Poisson model taking into account the initial number of clonogenic cells and the cell survival. The results showed the strong impact of the pattern of infiltration of tumour clonogens outside the delineated target on the outcome of the treatment. The treatment plan has to take into account the existence of the possible microscopic disease around the visible lesion; otherwise the high gradients around the target effectively prevent the sterilisation of the microscopic spread leading to low probability of control, in spite of the high dose delivered to the target. From this perspective, the proposed framework offers a further criterion for the evaluation of stereotactic radiosurgery plans taking into account the possible infiltration of tumour cells around the visible target. 1. Introduction The aim of radiation therapy is to stop the tumour growth process with sparing of the normal tissues nearby. For stereotactic radiosurgery (SRS) this is achieved by delivering a highly conformal dose distribution to the target in one fraction. The relatively steep dose falloff around the target ensures the sparing of the normal tissue and/or the critical structures near the target and this is the core of the SRS concept. The evaluation of plans is currently performed as a function of the conformity of therapeutic isodoses to the defined target and the gradients outside the target. This approach intrinsically assumes that tumour cells are confined to the target volume and that there are no infiltrations in the normal tissues around this target or that the impact of the possible infiltrations outside the delineated target on the probability of eradication of the tumour is negligible. However, several of the brain tumours commonly treated with SRS are invasive and therefore the existence of tumour cells outside the tumour lesions that could be identified in diagnostic images cannot be excluded [1, 2]. From this perspective, the evaluation of the plans should be performed not only