Marieke van Son

55 MRI-GUIDED ULTRAFOCAL HDR-BRACHYTHERAPY Even more so, inadequate patient selection could be the reason behind our poor tumour control, with 7/9 out-of-field recurrences. First, systematic biopsies could have led to under sampling of clinically significant tumours. According to the PRECISION trial, the diagnostic yield of MRI-targeted biopsies is higher than TRUS-guided sys- tematic biopsies(25), which means we potentially missed higher-risk disease. Even with systematic biopsies, almost half of this cohort already had Gleason sumscore 7. Gleason grade is a well-established independent predictor of BDFS, with grade 3+4 and 4+3 tumours at biopsy corresponding to 82-91% and 65-85% 4 to 5-year BDFS, respectively(26,27). In our cohort, 6/10 biochemical recurrences occurred in patients with Gleason 7 tumours. Second, for low/intermediate-risk patients (predominant Gleason pattern <4), the EAU-guidelines do not advise additional imaging for staging purposes in the primary setting, due to a lack of evidence(28). This means there is a risk of missing intrapros- tatic multifocality, although even mp-MRI has a false negative rate of 10-20% regard- ing clinically significant disease(29). Well-designed controlled trials regarding PET/CT imaging for nodal and metastatic staging are lacking, but evidence is increasing that a more sensitive metastases detection can be achieved than with classical bone scan and abdominopelvic CT(30). Overall, a combination of more extensive diagnostic staging and further dose esca- lation could be the next step in evaluating the validity of ultrafocal HDR-brachytherapy. Another point of discussion is the interpretation of the Phoenix definition for failure after FT, which leaves more biologically active untreated prostate tissue than whole- gland treatment. There is a need for a tailored definition for BF after FT. An alternative outcome measure is to consider only progression beyond curative treatment options as failure. A sub-analysis using this definition for disease-free survival offers a different perspective on tumour control. As recommended by an FT consensus meeting(31), considering successful focal salvage treatment as no failure yields a BDFS of 73% (54- 97%). Taking into account both successful focal and whole-gland salvage treatments, BDFS increases to 91% (79-100%) (Supplementary figure 4, available online at https:// doi.org/10.1016/j.ijrobp.2019.03.032 ). Importantly, focal and whole-gland salvage ra- diotherapy were well tolerated without exacerbated toxicity (grade 2 urinary frequency among both patents who underwent whole-gland salvage, deterioration of erectile function in 1 patient who was still potent). With the absence of increased toxicity and stable patient-reported QoL after primary FT, a strategy of repeated treatments could favour FT over a primary whole-gland approach, especially with increased adoption of ultrafocal targeting. The explorative univariable Cox-analysis showed a higher risk of first BF with in- creasing age. We do not have a definitive explanation for this association. Age did not correlate with other prognostic variables such as PSA, PSADT and AJCC stage group. With only 10 BF events, there is a possibility of a type I error. A disadvantage of ultrafocal HDR-brachytherapy is the labour-intensiveness. Total procedure time is currently 3-4 hours and this will inherently remain time-consuming, 3