Session Item

Most normal tissue complication probability (NTCP) models are derived from outcomes following photon-based radiotherapy. However, due to the fundamentally different properties between photons and protons, such as reduced dose bath and a higher relative biological effectiveness (RBE), NTCP models for photons might not be applicable to proton therapy (PT). Furthermore, most NTCP models are based on dose/volume parameters only (e.g. the probit model), but it is being recognised that also non-dose/volume related factors may be important predictors as well. The aim of this study was therefore to develop multivariate NTCP models for gastrointestinal (GI) and genitourinary (GU) morbidity for prostate patients treated with PT.

Dose volume histogram data (DVH), for the rectum, rectal wall, bladder, and bladder wall from 1151 prostate cancer patients treated with passive scattering PT between 2006 and 2010, as well as various patient characteristics and pre-treatment factors were analysed. Prescribed target doses were 78-82 Gy (RBE=1.1) in 2 Gy fractions. A least absolute shrinkage and selection operator (LASSO) logistic regression analysis was used to analyse the effects of the studied factors on rectal and bladder morbidities. Two prospectively scored alternative grade 2 late rectal bleeding (LRB) endpoints (CTCAE v3.0) were studied: Grade 2A+2B (GR2A2B) was classified as medical and procedural (n=184 (16%)), and Grade 2B (GR2B) was classified as procedural only (n=63 (5%)). For the bladder, prospectively scored Grade 3+ GU morbidity was used as endpoint (n=51 (4%)). In the multivariate analysis we accounted for collinearity of the DVH parameters. The discriminating ability of the model was described by the area under the curve (AUC), and model calibration was evaluated by the Hosmer-Lemeshow (HL) test.

In the final multivariate models, both GR2A2B and GR2B (Fig. 1) LRB was associated with age, anti-coagulant use, rectum volume, rectum D5%, and V75 of the rectum wall. The Grade 3+ GU endpoint was associated with anti-coagulant use, prostate volume, bladder D5%, equivalent uniform dose (EUD) (n=0.18) of the bladder, and V75 of the bladder wall. The AUC and HL test values were 0.66 and p = 0.49 for the GR2A2B LRB model, 0.73 and p = 0.35 for the GR2B LRB model, and 0.64 and p = 0.09 for the Grade 3+ GU morbidity model (Fig. 2), indicating good model fits.

Multivariate NTCP models were developed for both LRB and GU morbidity following passive scattered PT for prostate cancer. All endpoints were associated with many of the same predictors. In particular, anti-coagulant use, V75 of either the full structure or wall, as well as D5%, were important prognostic factors in this cohort. Consequently, these predictors may be used to optimise proton treatment planning.