This study comparing oncological outcomes after ten years of the three most established treatment options for localized prostate cancer patients shows that brachytherapy and external radiotherapy were not associated with decreased 10-year overall survival, but presented higher biochemical recurrence compared with radical prostatectomy in men with clinically localized prostate cancer. There have been other studies showing long-term results8,9,10 but, to our knowledge, this is the first one evaluating mortality, together with biochemical recurrence, in patients at low and intermediate prostate cancer risk.
Our results showing no differences in overall survival between radical prostatectomy and radiotherapy groups at ten years were consistent with the ProtecT randomized clinical trial all-cause mortality results between prostatectomy and external radiotherapy6. However, our findings differ from non-randomized studies, all of which showed significantly higher risk of death for radiotherapy groups than for radical prostatectomy: aHR of 1.7–1.58, 1.89, and 1.610 for brachytherapy; and aHR of 1.7–1.58, 1.959, and 1.910 for external radiotherapy. This higher mortality risk in radiation groups could be explained by differences in the sample characteristics, such as age9 or comorbidities11, between these studies and ours. In fact, the consistent demonstration in those studies of no higher prostate cancer‐specific mortality8,9,10 suggests that differences in all-cause mortality are associated with patients’ factors unrelated to prostate cancer. On the other hand, differences between surgical and radiation primary therapies in combined treatment with adjuvant androgen deprivation, associated with certain complications26,27, could partially explain this higher mortality risk.
We found relevant differences in prostate cancer-specific mortality among treatments at ten years of follow-up. Our results show higher statistically significant risk for external radiotherapy (aHR = 9.37, p = 0.015), but it was not significant for brachytherapy (aHR = 4.41, p = 0.120), compared to radical prostatectomy. In contrast, all the previous observational studies showed no significant higher risk of death due to prostate cancer, with smaller hazard ratios for brachytherapy (aHR of 1.39, 1.410, and 2.3 for low and 0.6 for intermediate tumoral risk8) and external radiotherapy (aHR of 1.79, 1.110, and 1.8 for both tumoral risk groups8). This difference in the magnitude of risk is explained by the remarkably low prostate cancer-specific mortality within the radical prostatectomy group in our study, close to zero, compared to the others: 1.8%8, 1.2%9, and 3.4%10. Underestimation by chance cannot be discarded, since there was only one death caused by prostate cancer among the 192 surgery patients, which occurred ten years and nine months after treatment. Differential misclassification between death by prostate cancer and death by other causes among treatment groups was minimized by obtaining cause of death through linkage with the National Institute of Statistics22.
In our study, brachytherapy showed lower risk of biochemical recurrence than external radiotherapy (29.1% vs 43.0%), consistently with previous findings (29% vs 67%)28, but higher risk than radical prostatectomy (29.1% vs 23.8%; and aHR of 1.93, in favor of radical prostatectomy). This latter finding clearly contrasts with previous publications10,12. On one hand, the meta-analysis of 8,385 patients12 obtained an odds ratio of 1.24 (95% CI 0.91–1.68) in favor of brachytherapy, but with considerable heterogeneity (I2 = 77%). On the other hand, the most recent observational study focused on intermediate risk10 also showed at ten years of follow-up a lower biochemical recurrence rate in brachytherapy (19.8%) than in radical prostatectomy (42.9%) after propensity scores adjustment. This rate in their radical prostatectomy group is twice that of our cohort (23.8%), mainly composed by low-risk prostate cancer patients. This suggests a relationship between tumoral risk and biochemical recurrence among surgery patients. In fact, in our study the rate for patients treated with radical prostatectomy was 12.1% in low risk and 32.7% in intermediate risk, the latter closer to the rate found in the Goy et al. study10.
The main limitation of this study is its observational design. The main concern is a possible treatment selection bias where, for example, brachytherapy is preferentially prescribed to patients with lower tumor risk, and surgery to younger patients13,14,15. In our cohort, the propensity scores balanced treatment selection bias15, except for Gleason score. Therefore, we constructed a final model adjusted with propensity scores and Gleason score. Furthermore, these results are consistent with those obtained through traditional adjustment in the Cox models (Supplementary Table S1). Secondly, results of prostate cancer-specific mortality should be interpreted with caution due to the low number of deaths attributed to prostate cancer in our cohort, which affects the statistical power. Thirdly, the results on biochemical recurrence could have been affected by missing data, since there were 132 patients (24.4%) with no PSA at 10 years of follow-up. Missing data ranged from 19.3% in radical prostatectomy to 27.7% in external radiotherapy, therefore the differences between radical prostatectomy and radiotherapy groups might be underestimated. However, no statistically significant differences were found between patients with and without PSA data at 10 years of follow-up (data not shown). Finally, the treatments were applied over one decade ago; since then, diagnostic techniques and treatments for prostate cancer have evolved. For instance, most new high-tech radiotherapy modalities seem to enhance patient survival and therapeutic gain29, but hypofractionation is still controversial30; therefore, further research in patients undergoing new modalities of treatment is needed.
The strengths of the study are certainly the large number of patients who completed the 10-year follow-up, which provides robust evidence on overall mortality and biochemical recurrence, especially for patients who underwent brachytherapy. Moreover, confirmation of all-cause mortality and prostate cancer-specific mortality with the National Institute of Statistics (INE) provided reliable and unbiased data for date and cause of death. Finally, our study’s outcome collection was in accordance with the current recommendations of the International Consortium for Health Outcomes Measurement (ICHOM) for localized prostate cancer, which were developed in 201531.
Further than survival and biochemical recurrence, it is important to consider other outcomes for treatment selection in localized prostate cancer. Acute complications and patient-reported outcomes are also included in ICHOM’s recommendations31. The results on survival and biochemical recurrence of the present study need to be balanced with results on patient-reported outcomes previously published16, showing that brachytherapy is the treatment option that causes the least negative impact. However, no single treatment can be considered the preferred strategy for managing all patients, and each treatment decision should be made jointly by patients and physicians5.
In conclusion, novel long-term results are provided on the effectiveness of brachytherapy to control localized prostate cancer during ten years after treatment, compared to radical prostatectomy and external radiotherapy. It presents high overall survival similarly to radical prostatectomy, but higher risk of biochemical progression. These results provide patients, clinicians, and health planners with valuable information, considered jointly with the other relevant outcomes, to make evidence-based decisions and facilitate shared clinical decision-making.