Despite clear problems with the overtreatment of indolent prostate cancer, there remains a significant percentage of men still being diagnosed with clinically high-risk disease who require aggressive treatment. Optimal management in these patients remains controversial, with strong advocates for radical prostatectomy (RP), radiotherapy (RT), androgen deprivation therapy (ADT), and, increasingly, a multimodal approach. In addition, novel systemic agents are being explored in the adjuvant or neoadjuvant setting. Surrounding these difficult questions regarding treatment are 3 fundamental issues: 1) does clinical staging accurately reflect the true state of disease; 2) does local control actually improve cure, and 3) which form of local control is actually best.
The current literature supports the need for local control. The Scandinavian Prostate Cancer Group Study Number 4 (SPCG-4) trial randomized patients to RP or observation and demonstrated a survival advantage for RP. Although this trial was not limited to patients with high-risk disease, the majority of patients likely had high-risk or intermediate-risk disease; only 12% were classified with clinical T1c disease, whereas 49% had pT3 disease, 38% had a Gleason score of ≥ 4 + 3, and 20% had a prostate-specific antigen (PSA) level > 20 ng/mL. In contrast, the PIVOT (Prostate Cancer Intervention versus Observation) trial did not demonstrate a survival advantage for the cohort as a whole. However, in subset analysis of patients with clinically high-risk disease, surgery was favored, with a hazards ratio (HR) of 0.4 (95% confidence interval [95% CI], 0.16-1.00; P = .04). Two randomized trials in the radiation literature have demonstrated the advantage of local control in patients with high-risk disease. Warde et al randomized 1205 patients with locally advanced (T3 or T4) or high-risk organ-confined disease (T2 with either a PSA level > 40 ng/mL or a PSA level > 20 ng/mL and a Gleason score ≥ 8) to lifelong ADT or ADT with RT. The results demonstrated an improved 7-year survival in the combined therapy group, with an HR of 0.77 (95% CI, 0.61-0.98; P = .033). In a second study by Widmark et al, a total of 875 men with locally advanced (T3, PSA level < 70 ng/mL, N0, and M0) were randomized to ADT or combined ADT and RT. They demonstrated an improved 10-year prostate cancer-specific mortality (PCSM) with a relative risk of 0.44 (95% CI, 0.30-0.66; P < .0001). In sum, the benefit of local control in patients with high-risk disease is clear.
Both RP and RT are effective treatment options. In a contemporary study, Stephenson et al demonstrated that men treated with RP for high-risk disease had a 15-year PCSM rate of 19% compared with 31% mortality from competing causes. Bolla et al reported a 10-year PCSM rate of approximately 30% and 10%, respectively, for patients with high-risk prostate cancer undergoing RT alone or in combination with ADT. Although the addition of ADT to RT has demonstrated a survival benefit, the exact duration of therapy continues to be studied.[10, 11] Still, as many as 40% of patients in the United States do not undergo treatment with curative intent but rather undergo primary or palliative ADT alone. This occurs despite population-based studies in patients with high-risk disease demonstrating a PCSM rate of 64% within 8 years among men managed with noncurative intent and improved survival in men treated with RP and RT.[3-6]
The more difficult question to answer is which treatment is better because to the best of our knowledge there are no good randomized data comparing RP with RT. Paulson et al demonstrated a survival advantage for surgery but the study size was small, there were concerns regarding trial design, and, lastly, the RT techniques studied are no longer current. A more recent study by Giberti et al failed to demonstrate a benefit with either RP or RT; however, their trial was limited to patients with low-risk disease. Large retrospective trials have demonstrated a consistent survival benefit for surgery. Kibel et al demonstrated improved disease-specific survival (HR, 1.66; 95% CI, 1.05-2.63) and overall survival (HR, 1.71; 95% CI, 1.40-2.08) for surgery compared with RT. Similar studies by Zelefsky et al, Albertsen et al, and Merglen et al all demonstrated similar improvements in metastasis-free survival, prostate cancer-specific survival, and overall survival.[16-18] Although these studies all controlled for important confounding variables including disease-specific variables (stage and grade) and patient-specific variables (ethnicity and comorbidity), it is impossible to control for all factors. Therefore, it remains plausible that the differences in outcome are due to unmeasured confounding variables. In addition, since patients undergoing RP who are found to have high-risk disease appear to benefit from adjuvant or salvage RT, it is likely that some of the benefit to surgery is due to subsequent treatment with RT.19
Because treatment selection in patients with prostate cancer is intimately linked to clinical stage of disease, correct patient staging is imperative to selecting appropriate therapy. However, in a large retrospective study, Ward et al noted that 27% of patients with high-risk disease were overstaged based on RP pathology. A similar finding was reported by Epstein et al, who found that 51% of patients with a biopsy Gleason score of 8 and 31% with a biopsy Gleason score of 9 to 10 had a Gleason score ≤ 7 at the time of RP.
The study by Abern et al in this issue of Cancer22 adds significantly to the literature by demonstrating that pathologic downstaging occurs more frequently that previously reported and patients who were downstaged had survival outcomes that were similar to those of patients with intermediate-risk and low-risk disease in prior studies. They defined clinically high-risk prostate cancer as that with a PSA level > 20 ng/mL, a biopsy Gleason score ≥ 8, or clinical classification of ≥ T3. They identified at total of 3930 patients undergoing RP between 1988 and 2011 at 6 Veterans Affairs medical centers, 527 of whom had clinical high-risk prostate cancer. At the time of RP, 206 of 527 patients (39%) did not have high-risk disease on the basis of final pathology, as defined by the absence of a Gleason score ≥ 8, classification of ≥ T3, or positive lymph nodes. Not surprising, patients who had lower-risk disease were less likely to have a biochemical disease recurrence (HR, 0.34), develop metastasis (HR, 0.17), or die of prostate cancer (HR, 0.17). The 10-year PCSM rate in men with favorable final pathology was 98%. This is nearly identical to that reported by Stephenson et al for patients with intermediate-risk disease. Men with only 1 high-risk clinical feature were more likely to demonstrate favorable pathology (42% vs 8%; P < .01) and a preoperative PSA level > 20 ng/mL was less predictive than RP stage or grade for oncologic outcomes.
A key benefit of RP in the men who were downstaged was the relatively low rate of ADT. All patients in this patient population would require concomitant ADT and RT. However, the rate of ADT was 37% in the overall patient cohort and only 17% in the group of patients with favorable pathology. A benefit is derived by avoiding the side effect profile of ADT. Adverse effects of ADT including hot flashes, decreased libido, fatigue, decreased muscle mass, decreased bone density, increased total body fat, and possibly harmful cardiovascular effects have been described.[23-26] Based on their findings, Abern et al concluded that upfront RP in patients with clinically high-risk disease is indicated because it provides accurate pathologic stage and grade. It is important to note that surgery can identify a substantial subset of men with favorable features in whom additional therapy is not indicated. Even patients with true high-risk disease were likely to benefit because they could receive tailored additional adjuvant therapy such as RT or ADT.[19, 27]
In the future, the improved ability to identify high-risk disease without surgery would be a clear benefit. Novel magnetic resonance imaging (MRI) or biomarker panels may be the key. The incorporation of endorectal coil MRI in the clinical staging of patients with prostate cancer may decrease both overstaging and understaging. Endorectal coil MRI is reported to have widely ranging sensitivities (51%-89%) and specificities (67%-87%) for the detection of prostate cancer.[28, 29] This variation is even more pronounced when using MRI to detect T3 disease, with sensitivities reported to range from 13% to 95% and specificity from 49% to 97%.[28-31] This variation has been attributed to reader experience. Prostate cancer biomarker testing, such as prostate cancer antigen 3 (PCA3) and the TMPRSS2-ERG fusion gene, are being incorporated into clinical staging. Currently, the data regarding the ability of these markers to significantly detect aggressive prostate cancer remains mixed and thus they have not achieved widespread clinical use.[34-36] Until MRI and biomarker panels become more refined in the detection of aggressive prostate cancer, clinical overstaging in patients with high-risk disease will continue to occur and, as Abern et al demonstrated, RP can be used with favorable outcomes in this population.