Timo Soeterik

39 Active Surveillance: Patient Selection less predictive compared to PSAD as an individual indicator. One out of three patients in the present cohort did not meet the PRIAS criterion PSAD <0.2 ng/ml/ml. This indicates that urologists may not regard this as an important AS eligibility criterion. Moreover, the vast majority of ongoing prospective AS studies (i.e. 7/9; 78%) did not include PSAD as a selection criterion. 13 However, given the strong association between PSAD and AS outcomes, the importance of PSAD as a predictor of AS outcomes might be currently underestimated. Furthermore, the higher number of patients not meeting the PSAD <0.2 ng/ml/ml criterion could be an alternative explanation for the strong association found for this clinical factor compared with the other PRIAS inclusion criteria. Because the numbers of patients were also lower in the other subgroups (e.g. patients with Gleason sum >6), this analysis might be susceptible to a higher probability of type II errors. In addition, owing to the limited statistical power, we cannot conclude that not meeting other criteria (including Gleason sum score ≤6 and clinical stage >T2), does not lead to a higher risk of unfavourable outcomes. We recommend further exploration of the potential correlations of the individual PRIAS criteria in AS cohorts including higher numbers of patients not fulfilling these criteria. Nevertheless, we regard our findings regarding PSAD to be valid, since they are in line with results of previous studies. For instance, in a population-based study of predictors of adverse pathology after radical prostatectomy among AS candidates, a PSAD >0.15 ng/ ml/ml was predictive for Gleason upgrading or pathological pT stage ≥T3 (OR 2.04, 95% CI 1.91 - 2.31). 14 In addition, another group found that PSAD >0.15 ng/ml/ml was strongly associated with adverse pathological findings on radical prostatectomy. 15 In four studies focusing on the association between baseline factors and disease progression in patients with PCa on AS, PSAD was also found to be a strong predictor for disease progression. 16–19 These previous results on the predictive value of PSAD for disease progression and prostatectomy outcomes in AS candidates, combined with our findings on the risk of metastasis, emphasise that PSAD should be taken into account when selecting patients with PCa for AS. The question remains whether or not we can safely expand the eligibility criteria for AS based on earlier results and our findings and, if so, to what extent this might be safe. From the present data, we conclude that AS is less safe for patients with a PSAD ≥0.2 ng/ ml/ml and >2 positive biopsy cores at baseline, since 12% of these patients eventually developed PCa metastases. In addition, since we found that 5% of patients with PSAD >0.2 ng/ml/ml developed metastasized PCa and thus exceeded the window of curability, we think that PSAD >0.2 ng/ml/ml should also be considered as a serious risk factor that should be taken into account during shared decision-making on treatment. The present study has several strengths, most notably being a multicentre study with a large population and describing the real-life AS situation for patients with PCa. However, some limitations also need to be addressed. First, most of the data were collected retrospectively, which may have led to information bias. For instance, there is a possibility that because of our retrospective evaluation patients were labelled as AS patients while they actually were on watchful waiting. However, the percentages of 2