Age-adjusted validation of the most stringent criteria for active surveillance in low-risk prostate cancer patients†
Presented as an abstract at the European Association of Urology Annual Congress, Barcelona, Spain, April 16-20, 2010 and at the Annual Meeting of the American Urological Association, San Francisco, California, May 29 to June 3, 2010.
The authors tested the performance of the currently used clinical criteria reported in populations studied by van den Bergh et al and Carter et al for the selection of patients with prostate cancer (PCa) for active surveillance (AS) according to age.
Data were analyzed from 893 patients who underwent with radical prostatectomy (RP). The authors investigated the rates of unfavorable PCa at RP (extracapsular extension, seminal vesicle or lymph node invasion, or Gleason score 7-10) in patients who fulfilled AS criteria according to age tertiles (ages ≤63 years, 63.1 to 69 years, and >69 years). Area under the plasma concentration time curve (AUC) analyses tested the criteria for predicting unfavorable PCa. Then, the patients were stratified according to the cutoff age of 70 years. Multivariate analyses were used to test the role of age in predicting unfavorable PCa.
The rate of unfavorable PCa characteristics was between 24% and 27.8%. In the van den Bergh et al population, after age 70 years, the rate of unfavorable PCa characteristics was 41% compared with 23.2% and 24.1% in patients in the previous age tertiles (ages ≤63 years and 63.1 to 69 years, respectively). In the Carter et al population, the rate of unfavorable PCa was 41.2% compared with 17.3% and 18.6% in the previous age tertiles (ages ≤63 years and 63.1 to 69 years, respectively). When the 70-year age cutoff was used, unfavorable PCa was identified in 17.9% to 23.6% of patients aged <70 years versus 4% to 41.2% of patients aged >70 years (all P < .001). AUC analyses revealed significantly lower performance in older patients. In multivariate analyses, after adjustment for prostate-specific antigen, prostate volume, and the number of cores, age represented an independent predictor of unfavorable PCa.
The currently used AS criteria performed significantly better for patients aged <70 years. The authors concluded that the current results should be taken into account when deciding whether to offer active surveillance to patients with low-risk PCa. Cancer 2012;. © 2011 American Cancer Society.
Prostate cancer (PCa) represents the most frequent neoplasm diagnosed in men in the United States.1 Recently published reports from the 2 major PCa screening programs have indicated that the rates of overdiagnosis and over treatment are high in patients with PCah.2, 3 Active surveillance (AS) initially was proposed as a reasonable treatment option for patients who were diagnosed with low-risk disease. The objective of AS is to decrease the rate of definitive therapy and its side effects in patients whose biologic tumor characteristics pose a minimal threat to their life expectancy. Several series have described the natural history of patients who undergo AS.4-6 To date, the results associated with this option are premature and controversial. Although a large proportion of patients who undergo AS have excellent oncologic outcomes,5-8 in these series, a consistent proportion of patients demonstrate disease progression and have to be treated after an initial period of observation.5, 9-12 Moreover, when patients exit the AS program and choose to undergo radical prostatectomy, high rates of unfavorable prostate cancer characteristics, such as extraprostatic extension and high Gleason score, are observed at pathologic evaluation, and these findings are particularly worrisome given the low volume and low-grade disease that these patients have at diagnosis.6, 13 Therefore, the major issue related to an initial surveillance policy is the possibility of losing the window of curability of the disease. These elevated rates of advanced prostate disease may be explained in part by the misclassification of patients with unfavorable PCa characteristics among AS candidates, which has been reported in several studies.14, 15 Those studies emphasized the need for more stringent criteria to correctly select candidates for AS.
Patient age represents a crucial issue in the treatment decision-making process for patients with PCa. Previous studies have indicated that there is a significant association between patient age and adverse PCa features.16 However, to our knowledge, the impact of age on the selection criteria for AS has never been assessed. In the current study, we tested the most stringent, currently used criteria for the selection of patients for AS, evaluating pathologic results from radical prostatectomy as an outcome according to patient age in a contemporary cohort of patients.
MATERIALS AND METHODS
Since September 2002, data from patients who underwent radical prostatectomy at our center have been collected prospectively in an institutional database. All patients are asked to provide informed consent to be included anonymously in future studies, and the prospective database was approved by the Institutional Review Board in 2002. For the purposes of the current study, only patients who had their biopsy and radical prostatectomy specimens analyzed at our institution were included in the analyses. Only patients who were diagnosed at extended biopsy (≥12 cores) were considered. Of 2431 patients who underwent radical prostatectomy at our institution between 2003 and 2009, we included only patients who underwent prostate biopsy in our center. Moreover, we excluded patients who had <12 cores at their initial biopsy and patients who were diagnosed with PCa after surgery for benign prostatic enlargement. Patients who had previously received radiotherapy or hormone therapy also were excluded. This resulted in 893 consecutive patients with complete and detailed clinical and pathologic information who underwent treatment between 2003 and 2009. Among these patients, we identified those who fulfilled the inclusion criteria for AS according to the criteria defined by Carter et al17 (originally proposed by Epstein et al18 to define indolent PCa) and by van den Bergh et al,12 respectively. The van den Bergh et al criteria for AS include clinical T1c/T2 tumor, a prostate-specific antigen (PSA) level ≤10.0 ng/mL, biopsy Gleason sum ≤3 + 3, PSA density <0.2 ng/mL per mL, and ≤2 cores positive for cancer.12 The Carter et al inclusion criteria for AS are clinical T1c tumor, biopsy Gleason sum ≤3 + 3, PSA density ≤0.15 ng/mL per mL, ≤2 cores positive for cancer, and ≤50% of any core involved with cancer.17 No patient underwent a restaging biopsy, and none of the included patients had been previously treated conservatively.
Clinical and Pathologic Evaluation
Clinical stage was assigned before surgery by the attending urologist according to the 2002 TNM staging system. PSA was always measured before digital rectal examination with the Elecsys Roche assay (Roche Diagnostics Corporation, Indianapolis, Ind). Prostate volume was measured by planimetric calculation during transrectal ultrasound. All biopsy and radical prostatectomy specimens were examined by 2 dedicated genitourinary pathologists. All prostatectomy specimens were processed according to the Stanford protocol.19 All patients underwent pelvic lymph node dissection with the removal of bilateral obturator and external lymph nodes.
Study Design and Statistical Analysis
We reviewed the contemporary criteria used for identification of AS candidates.20 First, among these criteria, we selected the 2 most stringent definitions, 1 from Europe and 1 from North America, and tested their ability to correctly exclude patients with unfavorable pathologic characteristics who are not currently considered suitable candidates for AS. Our statistical approach consisted of descriptive and categorical analyses in which the rates of pathologic unfavorable PCa characteristics, defined as the presence of either extracapsular extension, seminal vesicle invasion, lymph node invasion or a Gleason score of 7 to 10 in patients who initially fulfilled the AS criteria, were assessed. Second, the cohort of patients was divided according to age tertiles (ages ≤63 years, 63.1-69 years, and >69 years). We addressed the rates of unfavorable PCa characteristics with the same set of analyses according to patient age at treatment. Third, area under the plasma concentration time curve (AUC) analyses with 200 bootstrap resamples were performed to test the ability of criteria to predict unfavorable PCa at radical prostatectomy in each age category. Finally, we stratified patients according to the most informative age cutoff and performed multivariate analyses to test the role of age in predicting unfavorable PCa at pathology. The covariates included were preoperative PSA value, prostate volume, number of biopsy cores, and patient age (categorized according to the most informative cutoff age).
Statistical analyses were performed with the Statistical Package for the Social Sciences (version 17; SPSS Inc., Chicago, Ill) and S-PLUS (version 8; TIPCO Software Inc., Seattle, Wash). P values < .05 were considered statistically significant in all analyses.
The preoperative characteristics of 893 patients who underwent radical prostatectomy are listed in Table 1. The mean patient age was 65.7 years, and the median age was 66.5 years. The clinical tumor classification was T1c, T2, and T3 in 589 patients (66.1%), 287 patients (32.1%), and 16 patients (1.8%), respectively. The mean PSA level was 8.91 ng/mL, and the median PSA level was 6.35 ng/mL. The biopsy Gleason sum was ≤6, 7, and 8 to 10 in 581 patients (65.1%), 259 patients (29%), and 53 patients (5.9%), respectively.
Table 1. Preoperative Characteristics of 893 Patients Who Underwent Radical Prostatectomy, Stratified According to Age Categories
|No. of patients (%)||893||302 (33.8)||310 (34.7)||281 (31.5)|| |
|Clinical stage: No. (%)|| || || || ||.13|
| T1c||589 (66.1)||212 (70.2)||205 (66.1)||172 (61.2)|| |
| T2||287 (32.1)||88 (29.1)||98 (31.6)||101 (35.2)|| |
| T3||16 (1.8)||2 (0.7)||6 (1.9)||8 (2.8)|| |
|PSA, ng/mL|| || || || ||.37|
| Mean [median]||8.91 [6.35]||7.59 [6.20]||10.24 [6.19]||8.85 [6.60]||.37|
| Range||1.04-220.0||1.29-220.0||1.28-178||1.04-85.28|| |
|Prostate volume, mL|| || || || ||<.001|
| Mean [median]||55.4 [50.0]||50.1 [45.0]||56.1 [51.0]||60.3 || |
| Range||10-220||11-220||11-171||10-200|| |
|PSA density, ng/mL/cm3|| || || || ||.62|
| Mean [median]||0.19 [0.13]||0.17 [0.13]||0.2 [0.13]||0.19 [0.13]|| |
| Range||0.02-11.3||0.02-1.47||0.02-11.3||0.03-5.4|| |
|Total no. of cores|| || || || ||.36|
| Mean [median]||17.4 ||17.8 ||17.5 ||17.3 || |
| Range||12-24||12-24||12-24||12-24|| |
|No. of positive cores|| || || || ||.08|
| Mean [median]||5.9 ||5.5 ||5.9 ||6.3 || |
| Range||1-24||1-20||1-24||1-22|| |
|Biopsy Gleason sum: No. (%)|| || || || ||<.001|
| ≤6||581 (65.1)||216 (71.5)||202 (65.2)||163 (58.0)|| |
| 7||259 (29)||81 (26.8)||83 (26.8)||95 (33.8)|| |
| ≥8||53 (5.9)||5 (1.7)||25 (8.1)||23 (8.2)|| |
Of 893 patients who were included in the study, there were 302 patients aged ≤63 years, 310 patients ages 64 to 60 years, and 281 patients aged ≥70 years after dividing them according to adjusted tertiles. Table 1 also lists the preoperative patient characteristics according to age categories. No statistically significant differences were observed in terms of clinical tumor classification, preoperative PSA, PSA density, total number of biopsy cores, or number of positive biopsy cores (all P ≥ .1) between the 3 groups. There were significant differences in terms of prostate volume (mean, 50.1 mL, 56.1 mL, and 60.3 mL for patients ages ≤63 years, 64-69 years, and ≥70 years, respectively; P < .001). Moreover, a statistically significance difference was observed in biopsy Gleason sum, because younger patients overall had less aggressive tumors (P < .001).
Of all patients, 162 patients (18.1%) and 129 patients (12.9%) fulfilled the criteria proposed by van den Bergh et al and Carter et al, respectively (Table 2). For these 2 criteria, these numbers were 69 patients (42.6%) and 52 patients (40.3%) aged ≤63 years, respectively; 54 patients (33.3%) and 43 patients (33.3%) ages 64 to 69 years, respectively; and 39 patients (24.1%) and 34 patients (26.4%) aged ≥70 years, respectively.
Table 2. Preoperative Characteristics of Patients Who Underwent Radical Prostatectomy and Could Have Been Selected for Active Surveillance
|No. of patients||162||65||54||29|| ||129||52||43||34|| |
|Clinical stage: No. (%)|| || || || || || || || || ||NA|
| T1c||134 (82.7)||58 (84.1)||45 (83.3)||31 (79.5)|| ||129 (100)||52 (100)||43 (100)||34 (100)|| |
| T2||28 (17.3)||11 (15.9)||9 (16.7)||8 (20.5)||.8||0 (0)||0 (0)||0 (0)||0 (0)|| |
|PSA, ng/mL|| || || || ||.9|| || || || ||.5|
| Mean [median]||5.77 [5.87]||5.78 [5.9]||5.76 [5.66]||5.75 [6.16]|| ||6.32 [5.88]||3.89 [5.86]||5.91 [5.56]||6.74 [6.28]|| |
| Range||1.18-9.83||1.29-9.83||2.01-9.73||1.18-9.20|| ||1.18-24.28||1.34-24.28||2.01-12.70||1.18-15.80|| |
|Prostate volume, mL|| || || || || || || || || || |
| Mean [median]||63.2 [57.5]||55.58 [54.0]||69.3 [63.5]||68.18 ||.009||72.13 ||69.94 ||72.7 [62.5]||79.35 ||.3|
| Range||13-190||20-130||16-170||13-190|| ||13-220||30-220||16-170||13-190|| |
|PSA density, ng/mL/cm3|| || || || ||.1|| || || || ||.3|
| Mean [median]||0.10 [0.1]||0.11 [0.10]||0.09 [0.08]||0.09 [0.1]|| ||0.09 [0.09]||0.09 [0.1]||0.09 [0.08]||0.09 [0.1]|| |
| Range||0.03-0.25||0.03-0.20||0.04-0.02||0.04-0.25|| ||0.03-0.25||0.03-0.14||0.04-0.15||0.04-0.25|| |
|Total no. of cores|| || || || ||.09|| || || || ||.09|
| Mean [median]||18.8 ||19.2 ||19.5 ||17.3 || ||19.22 ||19.52 ||20.09 ||17.68 || |
| Range||12-24||12-24||12-24||12-24|| ||12-24||12-24||12-24||12-24|| |
|No. of positive cores|| || || || ||.01|| || || || ||.03|
| Mean [median]||1.67 ||1.57 ||1.67 ||1.85 || ||1.61 ||1.52 ||1.58 ||1.79 || |
| Range||1-2||1-2||1-2||1-2|| ||1-2||1-2||1-2||1-2|| |
| Biopsy Gleason sum: No. (%)|| || || || ||NA|| || || || ||NA|
| ≤6||162 (100)||65 (100)||54 (100)||29 (100)|| ||129 (100)||52 (100)||43 (100)||34 (100)|| |
Of all 893 patients, at radical prostatectomy, 111 patients (12.4%) had extracapsular extension, 65 patients (7.3%) had seminal vesicle invasion, 55 patients (6.2%) had lymph node involvement, and 484 patients (54.2%) had Gleason sums ≥7 or higher. The overall rate of unfavorable PCa characteristics was 56.2% (502 patients). The mean number of lymph nodes removed in the overall population was 17.04, and the median number of lymph nodes removed was 16. When considering only patients who fulfilled the van den Bergh et al criteria, 7 patients (4.3%) had extracapsular extension, 2 patients (1.2%) had seminal vesicle invasion, and 43 patients (26.5%) had a Gleason sum ≥7. No patients had lymph node involvement. In the van den Bergh et al cohort, the overall rate of unfavorable PCa characteristics was 27.8% (45 patients). Finally, when only patients who met the Carter et al criteria were considered, 4 patients (3.1%) had extracapsular extension, 1 patient (0.8%) had seminal vesicle invasion, and 30 patients (23.3%) had a Gleason sum ≥7 or higher. No patients had lymph node involvement. In the Carter et al cohort, the overall rate of unfavorable PCa characteristics was 24% (31 patients).
Table 3 lists the rates of extracapsular extension, seminal vesicle invasion, and Gleason sum ≥7 stratified according to age categories in tertiles. In the van den Bergh et al population, after age 70 years, the rate of unfavorable PCa characteristics was 41% compared with 23.2% and 24.1% for patients in the first and second age tertiles (all P < .001), respectively. Similarly, in the Carter et al population, the rate of unfavorable PCa characteristics was 41.2% compared with 17.3% and 18.6% for patients in the first and second age tertiles (all P < .001), respectively.
Table 3. Rates of Pathologic Unfavorable Prostate Cancer Characteristics Stratified According to Age Categories
|Van den Bergh 200912 [n=162]|| || || || |
| Pathologic Gleason sum 7-10||43 (26.5)||16 (23.2)||12 (22.3)||15 (38.5)|
| Extracapsular extension||7 (4.3)||1 (1.4)||2 (3.7)||4 (10.3)|
| Seminal vesicle invasion||2 (1.2)||0 (0)||1 (1.9)||1 (2.6)|
| Lymph node involvement||0 (0)||0 (0)||0 (0)||0 (0)|
| Overall unfavorable||45 (27.8)||16 (23.2)||13 (24.1)||16 (41)|
|Carter 200217 [n=129]|| || || || |
| Pathologic Gleason sum 7-10||30 (23.3)||9 (17.3)||8 (18.6)||13 (38.2)|
| Extracapsular extension||4 (3.1)||1 (1.9)||1 (2.3)||2 (5.9)|
| Seminal vesicle invasion||1 (0.8)||0 (0)||0 (0)||1 (2.9)|
| Lymph node involvement||0 (0)||0 (0)||0 (0)||0 (0)|
| Overall unfavorable||31 (24)||9 (17.3)||8 (18.6)||14 (41.2)|
The AUC analyses evaluating the performance characteristics of the van den Bergh et al and Carter et al criteria for AS revealed a steep decrease in the performance characteristics of such definitions in the highest age tertile (aged ≥70 years; all P < .001; Mantel-Haenszel test). Specifically, in patients aged ≤63 years, the AUCs were 61% and 60.3%; in patients, ages 64 to 69 years, the AUCs were 61.4% and 60.8%, respectively; and, in patients aged ≥70 years, the AUCs were 57.3% and 56.3%, respectively, according to the criteria proposed by van den Bergh et al and Carter et al.
We repeated the analyses stratifying patients according to the age limit of 70 years. The rates of unfavorable PCa characteristics at radical prostatectomy were 23.6% and 17.9% for patients aged <70 years who fulfilled the van den Bergh et al and Carter et al criteria, respectively. Conversely, these rates were as high as 41% and 41.2%, respectively, for patients aged ≥70 years (all P < .001).
Finally, we performed a multivariate logistic regression analysis addressing the risk of unfavorable PCa characteristics at radical prostatectomy in patients who fulfilled the criteria proposed by van den Bergh et al and Carter et al (Table 4). After adjustment for preoperative PSA value, prostate volume, and the number of biopsy cores, age represented an independent predictor of unfavorable PCa in both populations (OR, 2.36; P = .03 in the van den Bergh et al population; OR, 3.05; P = .01 in the Carter et al population).
Table 4. Multivariate Logistic Regression Analysis Predicting the Presence of Unfavorable Prostate Cancer Characteristics at Radical Prostatectomy
|No. of biopsy cores||0.96||.27||0.94||.12|
|Age ≥70 y||2.36||.03||3.05||.01|
AS represents an emerging treatment option for patients with low-risk PCa. Despite the lack of definitive data of patients on surveillance, several series from the United States and Europe have produced excellent results in terms of disease-specific survival.5, 10, 21, 22 However, a consistent proportion of patients progress and need active treatment despite having favorable disease characteristics at diagnosis. More important, it was demonstrated recently that up to 50% of patients who are treated after an initial period of observation develop biochemical recurrence,22 thus questioning the ability to maintain the window of curability in patients who have low-risk PCa characteristics at diagnosis. It remains to be demonstrated whether these findings are related to a rapid increase of PCa aggressiveness or whether these patients were not staged correctly at diagnosis. To date, there is a consistent body of evidence demonstrating that a non-negligible proportion of patients who are suitable for AS according to the most commonly adopted criteria do have unfavorable PCa characteristics discovered (extracapsular extension, seminal vesicle invasion, positive surgical margins, and lymph node involvement) if they undergo radical prostatectomy.14, 15, 23 It has been demonstrated that the strictest criteria used for AS lower these rates, but Gleason score upgrading represents a major issue, because roughly 25% of patients who are candidates for surveillance actually harbor Gleason scores ≥7.23-26 These patients have a high probability of progressing; therefore, a search for new criteria has been advocated. However, there are several features that have never been tested in the application of the AS criteria, and age represents a crucial issue when selecting patients for active treatment versus surveillance. In the current study, we decided to test the hypothesis that the most stringent criteria used for the selection of patients for AS may perform differently in different age categories. First, we observed that 23.3% and 27.1% of patients who were suitable for AS according to the criteria proposed by Carter et al and van den Bergh et al, respectively, had unfavorable PCa characteristics discovered at radical prostatectomy. The most common unfavorable PCa characteristic was significant Gleason score upgrading, whereas extracapsular extension and seminal vesicle invasion were extremely rare. It is interesting to note that, when patient age was taken into consideration, the rate of unfavorable PCa characteristics at radical prostatectomy increased significantly with age in patients who may have been considered preoperatively for AS. In particular, the rates of unfavorable PCa characteristics were 17.9% to 23.6% in patients aged <70 years versus 41% to 41.2% in patients aged >70 years (all P < .001). Moreover, multivariable logistic regression analyses indicated that patient age at diagnosis represented an independent predictor of unfavorable PCa characteristics in both the van den Bergh et al and Carter et al populations. Finally, the AUC analyses targeting the prediction of unfavorable PCa characteristics according to age categories revealed significantly lower performance of the tested criteria in older men.
To our best knowledge, our results offer a new perspective, because this is the first time that age categories have been taken into consideration in the selection of patients for AS. However, it was demonstrated previously that older patients may have higher rates of unfavorable PCa characteristics at radical prostatectomy relative to their younger counterparts. For example, Richstone et al16 reported a 45% rate of upgrading in patients aged >69 years versus 35.2% in younger patients (P = .001). Similarly, significant upstaging was observed in 40.2% of patients aged ≥70 years versus 29.3% of younger patients. Albertsen et al27 reported that patients aged ≥70 years who had a biopsy Gleason sum of 6 had a greater likelihood of dying from PCa than younger patients, in contrast to patients who had a biopsy Gleason sum between 7 and 10, who tended to be less likely to die from PCa the older they were at diagnosis. In a matched pairs analysis comparing 3 cohorts ages 46 to 55 years, 56 to 65 years, and >65 years, Magheli et al observed that younger patients had lower grade disease (P < .001) and lower rates of positive surgical margins (P = .035) and extraprostatic extension (P < .001), but those authors did not observe higher rates of lymph node involvement (P = .85) or seminal vesicle invasion (P = .56).28 In the current study, Kaplan-Meier analysis revealed no significant differences in biochemical recurrence across the age cohorts (log-rank P = .38). Therefore, we conclude that increased age is not associated with worse biochemical outcomes after radical prostatectomy; rather, age represents a surrogate for known predictors of biochemical recurrence after surgery. These data confirm our findings, because worse pathologic features were identified at pathology. Unfortunately, we do not have information on the biochemical recurrence rates of our patients to corroborate the findings of Magheli et al.28 Taken together, all of these studies suggest an association between age and prostate cancer aggressiveness. However, to our knowledge, the effect of age on AS selection criteria has never been tested before. We demonstrated that not all patients who are potential candidates for AS according to the most stringent selection criteria have the same probability of having indolent disease identified at pathology. In particular, older patients are affected more frequently by more aggressive disease at final pathology compared with their younger counterparts, even when they are affected by very-low-risk disease.
The results of our analysis may have several implications. First, because older patients are more frequently considered candidates for initial surveillance due to the slow progression of low-risk PCa, our data represent a warning about this policy, suggesting that older patients should be considered with greater caution before initiating an observational policy. Indeed, up to 41% of patients aged ≥70 years are at risk of being misclassified. This rate needs also to be counterbalanced by the life expectancy of such patients according to their comorbidity profile. Conversely, it should be noted that, in patients aged <70 years who fulfill the criteria proposed by van den Bergh et al and Carter et al, the rates of unfavorable PCa characteristics at radical prostatectomy were as low as 17.9% to 23.6%, suggesting that the staging is more accurate in younger patients. Finally, our data may be useful in further research to define age-adjusted criteria for the selection of patients for AS programs.
The strength of our study relies on the homogeneity of the pathologic results, on the high number of biopsy cores used in the diagnostic phase, and on the relatively limited time period of patient accrual (ie, our series should be considered contemporary), and, finally, on the great selection of the cohort of included patients. Conversely, our study has some limitations. First, the adoption of rigid inclusion criteria significantly reduced the number of patients that could have been included. However, we believe that the consideration of highly selected patients in our analyses improves the strength of our results. Second, a potential limitation of our study is related to the finding that, among older patients, only those with more aggressive disease received radical treatment, whereas those with less aggressive disease may have received other management options, such as surveillance or radiation therapy. This is indicated clearly in Table 1, which reveals that patients aged >70 years had lower rates of Gleason 6 scores. However, when we applied the AS criteria, we selected only patients with very-low-risk disease who were comparable to patients in the others age categories (Table 2). Third, although the collection of data was performed prospectively, the retrospective design of the study may be considered a limitation. Fourth, because upgrading and upstaging represent major problems for patients who choose AS, a restaging biopsy has been proposed and currently is being used in several series.29 Because they were candidates for surgical treatment, none of our patients underwent a restaging biopsy. Further studies analyzing the upgrading of patients who undergo restaging biopsies before AS are advocated to demonstrate whether this policy is able to overcome the commonly reported upgrading phenomenon. Finally, we believe that only studies evaluating the follow-up of patients who are managed initially with AS according to age will be able to demonstrate definitive outcomes, such as disease progression and cancer-specific mortality. However, as long as these outcomes are not definitive, staging studies like ours will need to be taken into account when surveillance protocols are proposed for patients with low-risk PCa.
In conclusion, roughly, 15% of patients who underwent radical prostatectomy at our institution may have been selected for AS protocols according to the criteria proposed by van den Bergh et al and Carter et al. Up to 25.9% of patients who can be considered for AS harbor unfavorable PCa characteristics when these criteria are applied. The performance of such recommendations decreases after age 70 years, because patients aged >70 years are at higher risk of harboring unfavorable PCa characteristics. The introduction of age-adjusted criteria seems to be important for optimizing the selection of patients with low-risk PCa.
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.