Denser ionisation clustering and complex DNA damage in proton Bragg peaks far exceeds that seen with conventional X-rays. This results in more efficient cell sterilisation, quantified by the relative biological effectiveness (RBE). Currently, a 1.1 RBE is used to determine the clinical proton doses by dividing the usual X-rays dose by this amount. This number, derived from short-term experiments, has been criticised as being irrelevant to late normal tissue (NT) effects following radiotherapy and included many control irradiations using lower voltage X-rays (with elevated RBE values) than those used in the clinic. In principle, an increased RBE could be used for each organ at risk, by using extensions of the clinically successful linear quadratic model.

Protons undoubtedly reduce or eliminate NT radiation dose in tissues distantly located from a tumour, but the necessity to include NT margins around a tumour can result in a higher volume of NT than tumour being irradiated. Deleterious side-effects can follow if the NT RBE exceeds 1.1, including in tissue very close to these margins and which are only partially spared.

Use of a constant 1.1 RBE can ‘overdose’ NT, which may require a greater dose reduction such as 1.2 in the brain; some tumours may be ‘under-dosed’ (since they might require a lesser or no reduction in dose). More sophisticated proton experiments show that RBE values of 1.1–1.5 and higher occur in some situations. There are now mathematical models of varying degrees of complexity that can estimate the RBE from the dose, LET and the low-LET radiosensitivities. True multidisciplinary cooperation is required to implement such new ideas in proton therapy in order to improve safety and effectiveness.