Top Line: Can the risk of recurrence after prostate radiation be dynamically predicted after each post-treatment PSA?
The Study: We have plenty of risk stratification models for prostate cancer prior to treatment. But how can we assess the risk of recurrence after treatment? This study leveraged data from the CHHiP trial to create a model that dynamically predicts the risk of recurrence after definitive radiation based on initial risk factors and post-treatment PSA kinetics. As a reminder, CHHiP randomized 3216 men with prostate cancer to conventional fractionation (74 Gy in 37 fractions) or one of two hypofractionation schedules (60 Gy in 20 fractions or 57 Gy in 19 fractions). Those with intermediate and high risk disease received ADT. The 60/20 arm was deemed non-inferior to conventional fractionation. For the purposes of the current study, only patients who received ADT were included although that was the vast majority of participants (95%). Overall 20% experienced some form of recurrence including biochemical recurrence (18%), clinical recurrence (2%), or prostate cancer death (0.03%). Overall, the model performed fair in predicting recurrence (AUC 0.7). However, its performance improved every year after treatment, increasing to ~0.8 by year 5. Because we all love a good cutpoint, the authors generated PSA cutpoints of 0.23 at 3 years, 0.34 at 4 years, and 0.41 at 5 years below which there was a <5% risk of recurrence by 8 years. To our knowledge, this model is not available online for clinical use, but in practice, such a model could be used to personalize the intensity of follow up or initiate early workup for recurrence.
TBL: Post-treatment PSAs can be used to dynamically update an individual patient’s risk of recurrence after definitive radiation and ADT for prostate cancer. | Parr, Int J Radiat Oncol Biol Phys 2023
The Study: We have plenty of risk stratification models for prostate cancer prior to treatment. But how can we assess the risk of recurrence after treatment? This study leveraged data from the CHHiP trial to create a model that dynamically predicts the risk of recurrence after definitive radiation based on initial risk factors and post-treatment PSA kinetics. As a reminder, CHHiP randomized 3216 men with prostate cancer to conventional fractionation (74 Gy in 37 fractions) or one of two hypofractionation schedules (60 Gy in 20 fractions or 57 Gy in 19 fractions). Those with intermediate and high risk disease received ADT. The 60/20 arm was deemed non-inferior to conventional fractionation. For the purposes of the current study, only patients who received ADT were included although that was the vast majority of participants (95%). Overall 20% experienced some form of recurrence including biochemical recurrence (18%), clinical recurrence (2%), or prostate cancer death (0.03%). Overall, the model performed fair in predicting recurrence (AUC 0.7). However, its performance improved every year after treatment, increasing to ~0.8 by year 5. Because we all love a good cutpoint, the authors generated PSA cutpoints of 0.23 at 3 years, 0.34 at 4 years, and 0.41 at 5 years below which there was a <5% risk of recurrence by 8 years. To our knowledge, this model is not available online for clinical use, but in practice, such a model could be used to personalize the intensity of follow up or initiate early workup for recurrence.
TBL: Post-treatment PSAs can be used to dynamically update an individual patient’s risk of recurrence after definitive radiation and ADT for prostate cancer. | Parr, Int J Radiat Oncol Biol Phys 2023