Current trends suggest that we are undertreating men with high-risk prostate cancer. The risk of dying from high-risk prostate cancer is halved by radical local treatment (whether it is surgery or radiation treatment), as compared with observation or ADT alone,[2, 3, 24-27] but it is unknown which is the best local treatment. This is a controversial question and was identified by the Institute of Medicine as one of the top 25 issues to be resolved by comparative effectiveness research.
Outcomes of radical prostatectomy versus observation
Multiple large studies have shown a survival benefit of radical prostatectomy over observation, especially in men with high-risk disease. Data from the SEER database, the SPCG-4 and most recently the PIVOT have all shown the benefits of radical prostatectomy over observation with an absolute risk reduction for 10–12 years prostate cancer-specific survival in the range of 7.6–11.2%, and a number needed to treat of 13.[25, 26] Vickers et al. used data from SPCG-4 to create a statistical model to calculate the decrease in risk of death associated with surgery. For a 65-year-old man with high-risk disease, the absolute 10-year risk reduction in prostate cancer mortality attributable to radical prostatectomy was 17.2%. All these benefits are an order of magnitude greater than for low-risk disease.
High PSA does not exclude RP
A high PSA should not exclude men from radical prostatectomy. Pretreatment PSA can reflect not only cancer, but benign prostatic hyperplasia and chronic inflammation. Furthermore, a Mayo Clinic study found that the 10-year PSA recurrence-free survival, metastasis-free survival and cancer-specific survival was 40%, 83% and 90%, respectively, in men with a serum PSA level of 50–99 ng/mL, and 36%, 74% and 79%, respectively, in those with a PSA level of ≥100 ng/mL. These data further show that surgery can cure a significant proportion of men with very high-risk prostate cancer.
Radiation treatment in high-risk disease
Several well-conducted randomized trials, recently reviewed elsewhere,[32, 33] show that EBRT combined with neoadjuvant and adjuvant ADT is superior to EBRT alone. The converse is also true; men with high-risk localized prostate cancer have improved local control and survival when EBRT is added to ADT compared with ADT monotherapy.[2, 3, 27] These studies show, similar to other solid cancers, that superior local control improves cancer outcomes. Importantly, androgen suppression must be of 2–3 years duration to get the maximal benefit.[34-36] Specifically, in men with high-risk prostate cancer, Bolla et al. reported that the 10-year disease-specific mortality was 30% in men who received EBRT alone, compared with just 10% in men who received EBRT plus 3 years of ADT. There is debate about the optimal period of adjuvant hormones, but the AUA and EAU prostate cancer guidelines support RT with neoadjuvant and adjuvant ADT for at least 2 years.[38, 39] The exemplary effort of the radiation oncology community in this space shows that EBRT combined with ADT is superior to either monotherapy regimen, and has led to RT being recommended as the standard of care in the EAU guidelines, with surgery restricted only for “highly selected” patients. Although these studies show the critical role for local control in improving cancer outcomes, they do not infer, in any way, superiority of radiotherapy over surgery in achieving optimal local control.
Radical prostatectomy versus radiation treatment in men with high-risk prostate cancer
In the absence of a large randomized trial that is required to definitively determine which of these two effective treatment modalities is the best for men with high-risk prostate cancer, we are beholden to the immeasurable flaws and biases of the retrospective analyses. There are retrospective analyses that favor RP,[13-18] some that favor RT[19, 20] and several that found no difference between modalities including the only prospective, albeit small, randomized trial in Japanese men.[6, 21, 22]
Several large retrospective analyses favor RP over RT. Abdollah et al. used the SEER database to identify 68 665 patients with localized prostate cancer treated with radical prostatectomy or radiotherapy between 1992 and 2005. Propensity-score matching was used to minimize potential bias related to treatment assignment. In the subgroup of men with high-risk prostate cancer, the 10-year cancer-specific mortality rate was 6.8% for RP versus 11.5% for RT. Multivariable analyses allowing for age and comorbidities confirmed that RP provided superior cancer-specific mortality. Cooperberg et al. compared the risk-adjusted mortality outcomes of 7538 men treated by RP, EBRT or ADT from the CaPSURE registry where over 10% had high-risk disease. Although the median follow up was less than 5 years, there was a significant survival benefit in men with high-risk prostate cancer treated by RP compared with either EBRT or ADT.
Zelefsky et al. used competing risk regression analysis controlling for patient age, clinical stage, serum PSA level, biopsy Gleason score and year of treatment to show that RP was associated with a nearly 10% lower risk of progression to metastasis, and a lower cancer specific mortality as compared with EBRT in 409 men with high-risk prostate cancer. Tewari et al. showed in 453 men with Gleason 8 disease that after a median follow up of over 4 years that the risk of cancer-specific death was 13.4% in those that had RP, 16.8% for RT and 43% for observation. In another study of 528 men with high-risk disease, Albertsen et al. showed that after adjusting for pretreatment Gleason score, PSA, clinical stage, age at diagnosis and comorbidities, the 10-year cancer-specific survival was 90%, 80%, and 70% for surgery, radiation and observation, respectively.
Boorjian et al. reported a retrospective analysis of RP versus RT on high-risk men from the Mayo Clinic. After a median follow up of 10 years for RP (n = 1238) and 6 years for EBRT plus ADT (n = 609), no difference in 10-year disease progression or CSS rates was observed. However, RP had better OS rates (77%) compared with men treated by EBRT and ADT (62%) or EBRT alone (52%). Whether this effect on OS was a result of patient selection or the effects of ADT is unknown.
In contrast, a study by Grimm et al. using strict criteria for study inclusion, which meant including just 17% of 848 papers identified regarding treatment for prostate cancer, found that combination regimens of ADT, EBRT and brachytherapy had higher PSA progression-free survival than surgery, EBRT or brachytherapy alone for high-risk patients. However, the far more important measures of cancer-specific and overall survival were not assessed.
The only published prospective randomized trial is a multicenter Japanese study in a small number of men (n = 95) with high-risk disease treated with neoadjuvant diethylstilboestrol and then randomized to RP (n = 46) or RT (n = 49). After a median follow up of 8.5 years, there was no significant difference between biochemical relapse, clinical progression or CSS. However, it must be recognized that the RT group received no adjuvant ADT and were treated with 40–50 Gy to the pelvis followed by a 20-Gy boost to the prostate, both of which would be considered suboptimal by today's standards.
All these retrospective analyses have incalculable confounders related to different definitions of high-risk prostate cancer, non-randomized and heterogeneous patient populations, non-uniform outcome measures, limited patient numbers, and different treatment regimens (particularly pertinent to the evolution of EBRT regimes). These studies highlight the need for randomized trials to determine whether men with high-risk prostate cancer are better served with surgery, or conversely, have unfavorable results attributed to RT been a result of selection bias with surgery reserved for men in better health.[12, 23]