Modelling of NTCP for acute side effects in patients with prostate cancer or brain tumours receiving proton therapy

The purpose of this study was to identify patients who likely benefit most from proton therapy (PT), based on the potential reduction of normal tissue complication probability (NTCP) compared to photon therapy. The NTCP models required for this comparison were developed using clinical data on acute side effects of prostate cancer and brain tumour patients having received PT.
In this study, 113 patients with primary brain tumours and 30 patients with adenocarcinoma of the prostate who had received PT were included. For the brain tumour patients, the radiation-induced acute side effects alopecia, erythema, pain and fatigue were considered. For prostate cancer patients, several gastrointestinal and genitourinary side effects were investigated. The occurrence of these side effects was correlated with different dose-volume parameters of associated organs at risk (OARs), such as skin and brain or rectum and bladder. The respective NTCPs were modelled by logistic regression. For every patient a volumetric modulated arc therapy (VMAT) photon treatment plan was retrospectively created. Differences in dosimetric parameters and NTCP between PT and VMAT plans were evaluated.
Significant correlations were found between acute side effects and dose to OARs. For example, occurrence of alopecia grade 2 and erythema grade ≥2 depended on dose-volume parameters in the high dose region of the skin (p<0.001). Proton plans showed significantly reduced low to intermediate dose volumes in all investigated OARs compared to VMAT plans (p<0.001). In the more relevant high dose volumes smaller differences between proton and photon treatment were found.
We found significant correlations between the occurrence of acute side effects and dose-volume parameters of associated OARs for patients with primary brain tumours or prostate cancer receiving PT. After inclusion of late side effects and validation in an external dataset, these NTCP models may be used to identify patients likely to benefit most from PT.