Background: To provide insight in the treatment variation of very-low-risk prostate cancer patients and to assess the role of hospital-related factors.
Methods: All patients diagnosed with very-low-risk prostate cancer (cT1c-cT2a, PSA < 10 ng/ml, Gleason score <7 and <3 positive cores) in 2015 and 2016 were identified through the population-based Netherlands Cancer Registry. Multilevel logistic regression analyses were performed to examine the crude and case-mix adjusted probability of immediate treatment vs. active-surveillance (AS) according to hospital of diagnosis and to evaluate the effect of patient-, tumour-, and hospital-related factors.
Results: In all, 2047 (85.4%) of the 2396 patients with very-low-risk prostate cancer were managed with AS. The crude proportion of patients with AS varied from 33.3 to 100% between hospitals. Case-mix adjusted probability varied from 71 to 97%. Tumour stage cT2a vs. cT1c (OR 2.0, 95%CI 1.1-3.6), two vs. one positive core (OR 2.8, 95%CI 1.6-4.7), diagnostic MRI (OR 2.8, 95%CI 1.5-5.2), discussion of a patient in a multi-disciplinary team (OR 2.2, 95%CI 1.1-4.5), discussion of treatment options with the patient (OR 3.3, 95%CI 1.5-7.4) and type of hospital (non-university referral hospital vs. community hospital: OR 0.5, 95%CI 0.2-0.9) were associated with immediate treatment.
Conclusion: The majority of Dutch very-low-risk prostate cancer patients is managed with AS but variation between hospitals exists. Part of the variation is explained by patient- and tumour characteristics but also hospital-related factors play a role. This implies that clinical practice could be improved.