Marieke van Son

35 RADICAL VS. FOCAL TREATMENT FOR LOCALIZED PROSTATE CANCER in favour of EBRT in our study. First, EBRT data was collected in a retrospective manner, while focal and LRP data were collected prospectively. Second, unknown or unmea- sured confounders may have distorted results. Although EBRT patients had higher PSA and higher-grade disease, they may have had smaller tumours or longer PSA doubling time, potentially indicating less aggressive disease. Third, as discussed above, the widespread use of neo-adjuvant ADT among EBRT patients may have substantially improved FFS within the available medium length follow-up. For the focal group, estimated FFS seemed to decline faster beyond six years fol- low-up in both the three-way and two-way Kaplan Meier curves. Although this estima- tion is limited by smaller numbers of patients at risk at later time points, this may reflect emergence of residual cancer cells in the treated area or de novo lesions emerging in untreated tissue. This requires further research. We selected patients with NCCN low- to intermediate-risk disease, assuming eligi- bility for both radical treatment and FT. Besides active treatment, current guidelines however recommend offering active surveillance (AS) to patients with (very) low-risk disease 33-35 . Following general AS eligibility criteria (Gleason score ≤6, clinical T1c or T2a/b and PSA ≤10 ng/mL, not taking into account PSA density or number of positive cores) 36 , 222/1360 (16.3%) of patients in our study could have been offered AS. This is important because it is generally agreed that FT should only be considered in men who are likely to benefit from active treatment. Nonetheless, the only randomized focal study available compared focal ablation (using vascular targeted photodynamic therapy [VTP]) to AS, randomizing 413 men. At four years, they concluded that conversion to radical treatment was less likely in the focal group (24% vs 53%), lowering the risk of treatment-related morbidity 37 . There has been criticism of this study recruiting men with very low risk disease and not incorporating a confirmatory MRI targeted biopsy when a lesion was seen prior to randomisation. As our primary outcome we studied the composite endpoint treatment failure, con- sisting of salvage treatment, metastatic disease, systemic treatment or progression to WW. Here, the frequently used endpoint biochemical progression-free survival is of limited value due to the lack of a biochemical failure definition after FT 38 . Although OS is the most valid and reliable endpoint, treatment failure serves as a clinically mean- ingful surrogate endpoint within the time frame of this study. We considered prostate bed EBRT after LRP as adjuvant treatment (i.e. part of primary treatment) when given as a consequence of rising PSA and positive surgical margins. Before LRP, patients are explained that surgery entails the risk of incomplete resection, which then requires ad- juvant radiotherapy. Therefore, we did not consider such adjuvant treatment as failure. In the same setting, we allowed one focal re-do as part of initial focal treatment. WW was added to the treatment failure definition to account for the fact that EBRT patients were older and more likely to have comorbidities, potentially preventing them from undergoing salvage treatment upon recurrence. Our study did not have comparative toxicity or patient-reported outcome data. Within randomized trials comparing radiotherapy and prostatectomy, no discernible differ- 2