Retrospective evaluation of high-dose-rate brachytherapy multicriteria planning using physical dose versus radiobiological criteria for prostate cancer

Abstract

Abstract Radiobiological indices can provide insights into treatment efficacy beyond traditional physical dose metrics and potentially facilitate the comparison between various radiotherapy plans. This study investigates the integration of radiobiological indices with standard physical dose criteria to improve high-dose-rate (HDR) brachytherapy plan evaluation and selection process for the treatment of prostate cancers in a multicriteria optimization (MCO) framework. This is accomplished within the framework of a graphics processing unit-based multicriteria optimization algorithm, gMCO. 2000 Pareto-optimal plans for 200 patients were optimized for a 15 Gy HDR brachytherapy boost after external beam radiation therapy (44 Gy in 22 fractions). Tumour control probability (TCP), normal tissue complication probability (NTCP), and uncomplicated tumour control probability (UTCP) were calculated for each plan. Maximizing UTCP alone resulted in insufficient target coverage (target $$V_{100} < 90\%$$ ) according to clinical guidelines. Conversely, maximizing target coverage while meeting institutional criteria compromised UTCP significantly (reduction of about 0.09). Selecting plans that met all institutional criteria first, then maximizing UTCP, achieved a balanced compromise between tumour control and normal tissue safety. While combining UTCP and standard dose metrics based on dose-volume histogram (i.e., absorbed dose or physical dose constraints) with MCO can enhance brachytherapy plan optimization, exclusive reliance on standard TCP and NTCP models, using recommended parameters, yields clinically unacceptable plans.