SEARCH

SEARCH BY CITATION

Keywords:

  • active surveillance;
  • disease progression;
  • early detection of cancer;
  • prostatic neoplasms;
  • survival

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

Study Type – Therapy (outcomes)

Level of Evidence 2b

What's known on the subject? and What does the study add?

Active surveillance aims to reduce overtreatment by selecting patients with low risk prostate cancer (PCa) based on favourable disease characteristics. However, most studies on active surveillance do not have long-term results available; in particular, data on patients with intermediate risk disease are lacking.

Our findings demonstrate that withholding radical treatment in men with low or intermediate risk screen-detected localized PCa leads to a substantial delay or even avoidance of radical treatment in a majority of men. Favourable disease-specific outcomes confirm the feasibility of active surveillance for low risk PCa and also support a role for active surveillance in selected patients with intermediate risk PCa.

OBJECTIVE

  • • 
    To assess the longer-term feasibility of active surveillance, we aimed to evaluate outcomes of patients with screen-detected localized prostate cancer (PCa) who initially elected to withhold radical treatment for either low or intermediate risk disease.

PATIENTS AND METHODS

  • • 
    All men underwent screening for PCa in the Rotterdam and Helsinki arms of the European Randomized Study of Screening for Prostate Cancer (ERSPC); eligible men were diagnosed with PCa prior to the establishment of the ERSPC-affiliated Prostate Cancer Research International: Active Surveillance (PRIAS) study (1994–2007) and were initially expectantly managed in the absence of a fixed follow-up protocol.
  • • 
    Low risk PCa was defined as clinical stage T1/T2, PSA ≤ 10 ng/mL, PSA density < 0.2 ng/mL/mL, Gleason ≤ 6 and maximum two positive biopsy cores, whereas PSA 10–20 ng/mL, Gleason score 7 and three positive biopsy cores were considered intermediate risk features.
  • • 
    Disease-specific, overall and treatment-free survival were analysed using the Kaplan–Meier and competing risks methods.

RESULTS

  • • 
    In all, 509 patients with PCa were eligible, of whom 381 were considered low risk and 128 intermediate risk.
  • • 
    During a median follow-up of 7.4 years, a total of 221 patients (43.4%) switched to deferred treatment after a median of 2.6 years.
  • • 
    The calculated 10-year disease-specific survival rates were 99.1% and 96.1% for low and intermediate risk patients, respectively (P= 0.44), and for overall survival 79.0% and 64.5%, respectively (P= 0.003).
  • • 
    Competing risks analysis showed similar results.

CONCLUSIONS

  • • 
    Withholding radical treatment in men with low to intermediate risk screen-detected PCa leads to a substantial delay or even avoidance of radical treatment and its potential side-effects in a majority of patients.
  • • 
    Disease-specific outcomes at 7.4 years of follow-up are favourable in low as well as intermediate risk patients.
  • • 
    This confirms the feasibility of active surveillance according to contemporary criteria, and also suggests a potential role for active surveillance in selected men with intermediate risk features.

Abbreviations
PCa

prostate cancer

AS

active surveillance

ERSPC

European Randomized Study of Screening for Prostate Cancer

PRIAS

Prostate Cancer Research International: Active Surveillance

RP

radical prostatectomy

CIF

cumulative incidence function.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The incidence of indolent prostate cancer (PCa) has increased substantially over the last two decades due to the widespread use of PSA as a marker for the early detection of PCa [1]. In response to the increasing overdetection that followed the rising incidence of PCa [2], active surveillance (AS) evolved as a treatment strategy for low risk PCa [3]. AS aims to prevent or at least delay treatment of presumed indolent tumours, thereby reducing potential side-effects from treatment and maintaining quality of life for these patients. AS programmes aim to select patients with low risk PCa by applying predefined inclusion criteria and a strict follow-up schedule, leaving the opportunity open to switch to deferred active therapy in the case of risk reclassification or signs of tumour progression without compromising the window of opportunity for curative treatment.

Most AS studies do not yet have long-term results available, and in particular outcomes after expectant management for men with screen-detected disease and more adverse features have not been widely reported. Information on the long-term outcomes of patients with expectantly managed PCa can contribute to expanding our knowledge and assessing the feasibility of AS as a treatment strategy for low and intermediate risk PCa in an era of widespread use of PSA testing for screening and early detection. We aimed to evaluate the outcomes in a cohort of patients with screen-detected PCa who initially elected to withhold radical treatment, and assessed differences in the outcomes of men who are assumed to be low risk according to AS criteria [4] and men with intermediate risk disease. Data on disease-specific mortality and risk of death from competing causes, active-therapy-free survival and metastases-free survival were recorded.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

All patients underwent screening for PCa between 1993 and 2007 in the Rotterdam or the Helsinki arm of the European Randomized Study of Screening for Prostate Cancer (ERSPC) and, after PCa diagnosis, initially elected a strategy of expectant management.

Men aged 50–75 years were invited for PSA screening at 4-year intervals. A PSA threshold of 3.0 or 4.0 ng/mL was used as an indication for biopsy dependent on the centre, as was the additional use of findings at DRE and transrectal ultrasound [5,6]. In the case of abnormal findings, a lateralized sextant biopsy was performed that was assessed by expert genitourinary pathologists in both centres. In 2002, the Finnish centre changed its biopsy schedule to 10 or 12 cores [6]. Men in whom PCa was diagnosed were referred to the regular medical circuit for further management of their disease. Since all men were diagnosed prior to the establishment of the ERSPC-affiliated Prostate Cancer Research International: Active Surveillance (PRIAS) study [7], no fixed follow-up protocol was applied regarding PSA measurements, repeat biopsies and recommendations on when to switch to deferred radical therapy. Instead, the timing and method of follow-up was at the treating physician's discretion. Data on baseline characteristics, deferred treatment, distant metastases and causes of death were recorded. Deferred therapies were divided into three groups: radical prostatectomy (RP), radiation therapy and hormonal therapy. Therapies applied for benign prostatic hyperplasia were not considered as active treatment in the context of this paper.

Patients were considered low risk when contemporary PRIAS criteria were met: clinical stage T1c/T2, PSA ≤ 10 ng/mL, PSA density < 0.2 ng/mL/mL, Gleason score ≤ 6 and maximum two positive biopsy cores. These criteria aim to select small, localized, well-differentiated PCa [7]. Men with known positive lymph nodes or distant metastases at the time of diagnosis were not considered eligible. Men with PSA 10–20 ng/mL and/or Gleason score 7 and/or three positive biopsy cores were considered intermediate risk patients. Follow-up data were retrieved by chart review and mortality data were obtained by linkage with the national cancer registries. Causes of death were determined by an independent committee (Rotterdam) or either by death certificates or by an independent committee (Helsinki).

Active-therapy-free survival, metastases-free survival, disease-specific survival and overall survival were analysed using the Kaplan–Meier method. In addition, to avoid overestimation of disease-specific mortality, non-parametric competing risks regression analysis was performed [8]. Comparisons between the low and higher risk groups were made using the Mann–Whitney U test (continuous data) and the chi-squared test (nominal data). Statistical analyses were performed using SPSS statistical software (version 17.0; SPSS Inc., Chicago, IL, USA) and STATA version 11.0 software (College Station, TX, USA). All statistical tests were two-sided with P < 0.05 considered to be statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The overall cohort consisted of 509 patients, of whom 381 were considered low risk (median age 67.6 years) and 128 were considered intermediate risk (median age 67.4 years). Patients with low risk features had a lower PSA level, larger prostate volume, lower PSA density, more favourable T stage, lower Gleason score and fewer positive cores at diagnosis than intermediate risk patients. Study group characteristics at diagnosis are shown in Table 1.

Table 1.  Study group characteristics at diagnosis for 509 patients with screen-detected PCa who were initially expectantly managed
 Low risk (n= 381), median (25–75p)Intermediate risk (n= 128), median (25–75p) P Total (n= 509)
  • *

    Significant value (P < 0.05).

Age, years67.6 (64.2–71.3)67.4 (64.7–72.1)0.5267.6 (64.4–71.4)
PSA, ng/mL4.1 (3.2–5.0)5.3 (4.0–7.6)<0.001*4.3 (3.4–5.5)
Prostate volume, mL41 (33–51)31 (25–41)<0.001*39 (30–49)
PSA density, ng/mL/mL0.10 (0.07–0.13)0.20 (0.12–0.25)<0.001*0.11 (0.08–0.15)
T stage N (%) N (%)0.04* 
 T1c325 (85.3)104 (81.3) 429 (84.3)
 T2a48 (12.6)19 (14.8) 67 (13.2)
 T2b3 (0.8)5 (3.9) 8 (1.6)
 T2c5 (1.3)0 (0) 5 (1.0)
Gleason score  <0.001* 
 ≤6381 (100)99 (77.3) 480 (94.3)
 3 + 40 (0)25 (19.5) 25 (4.9)
 4 + 30 (0)4 (3.1) 4 (0.8)
Positive biopsy cores  <0.001* 
 1292 (76.6)54 (42.2) 346 (68.0)
 289 (23.4)35 (27.3) 124 (24.4)
 30 (0)39 (30.5) 39 (7.7)
Number of biopsy cores, mean (range)6.9 (3–13)7.1 (3–14)0.117.0 (3–14)

Median follow-up was 7.4 years (25–75 percentile 5.0–9.6 years) after PCa diagnosis. A total of 221 patients (43.4%) switched to deferred active therapy after a median of 2.6 years, while treatment was avoided, at least until last known follow-up, for a median of 6.8 years (25–75 percentile 4.5–9.1 years) in the other 288 men (56.7%). In the low risk group 152 patients (39.9%) underwent treatment, while in the intermediate risk group 69 patients (53.9%) were eventually treated (P= 0.006). The number of patients per deferred treatment option is shown in Table 2. Men in the low risk group received significantly more radiation therapy (P= 0.01) and less hormonal therapy (P= 0.01) than the intermediate risk group. The 10-year active-therapy-free survival rate was 49.7% and 30.3% in the low and intermediate risk groups respectively (log-rank P= 0.005; Fig. 1).

Table 2.  Study group characteristics during follow-up for 509 patients with screen-detected PCa who were initially expectantly managed
 Low risk (n= 381), N (%)Intermediate risk (n= 128), N (%) P Total (n= 509), N (%)
  • *

    Significant value (P < 0.05).

Deferred active therapy152 (39.9)69 (53.9)0.006*221 (43.4)
 Radical prostatectomy32 (21.1)18 (26.1)0.4150 (22.6)
 Radiotherapy104 (68.4)35 (50.7)0.01*139 (62.9)
 Hormonal therapy16 (10.5)16 (23.2)0.01*32 (14.5)
Distant metastases1 (0.3)3 (2.3)0.02*4 (0.8)
Overall death60 (15.7)38 (29.7)0.001*98 (19.3)
PCa death3 (0.8)2 (1.6)0.445 (1.0)
 Median (25–75p)Median (25–75p) Median (25–75p)
Follow-up time, years7.5 (4.9–9.6)7.2 (5.3–9.9)0.697.4 (5.0–9.6)
Follow-up PSA measurements,N10 (6–14)10 (6–15)0.5710 (6–14)
Time to active therapy, years2.5 (1.2–4.9)2.7 (1.1–4.6)0.902.6 (1.1–4.8)
Time to metastases, years3.5 (–)8.4 (7.4–11.7)0.187.9 (4.4–10.8)
Time to overall death, years5.3 (3.1–8.6)6.8 (4.7–8.9)0.056.0 (3.6–8.6)
Time to PCa death, years2.9 (0.4–5.8)8.1 (7.5–8.6)0.085.8 (1.7–8.1)
image

Figure 1. Kaplan–Meier curve showing active-therapy-free survival, stratified by risk group.

Download figure to PowerPoint

Distant metastases were found in one low risk patient and three intermediate risk patients, leading to 10-year metastases-free survival rates of 99.7% and 96.4%, respectively (log-rank P= 0.03).

Overall 98 patients (19.3%) died during follow-up, of whom five (1.0%) died of PCa. The overall death rate was higher in the intermediate risk group (29.7% vs 15.7%; P= 0.001). PCa was considered the cause of death in three men in the low risk group and two in the intermediate risk group (P= 0.44). Study group characteristics during follow-up and a case description of patients in whom PCa was recorded as the cause of death are shown in Tables 2 and 3, respectively.

Table 3.  Case description of five patients in whom PCa was recorded as their cause of death
PatientTumour characteristics at diagnosisTime to active treatment, year (therapy type)Time to PCa death, year
Age, yearClinical stagePSA, ng/mLProstate volume, mLBiopsy result, positive/total coresGleason scoreRisk group
  1. RTx, radiotherapy; HTx, hormonal therapy. *Died during treatment of an abdominal aneurysm that was coincidentally found during the evaluation for PCa. †Received initial expectant management because of comorbidity, received hormonal therapy due to urinary retention, and despite no clinical evidence of progressive disease at time of death PCa was recorded on his death certificate. ‡Received initial expectant management because of comorbidity.

168T1c3.8252/63 + 3Low0.5*
263T1c3.1402/63 + 3Low0.8 (RTx)5.8
371T1c7.8571/123 + 3Low0.3 (HTx)2.9
474T2b4.3313/63 + 3Intermediate7.3 (HTx)8.6
563T1c4.0Unknown1/63 + 3Intermediate2.7 (RTx)7.5

The 10-year overall survival rate was 75.1% in the complete study cohort, and 79.0% and 64.5% in the low and intermediate risk groups, respectively (P= 0.003, Fig. 2). The 10-year disease-specific survival rate overall was 98.4%; this rate was 99.1% and 96.1% in the low and intermediate risk group, respectively (P= 0.44, Fig. 3). The cumulative incidence function (CIF) for PCa and other cause mortality, computed using competing risks analysis, showed similar results to the Kaplan–Meier method. The CIF after 10 years for disease-specific mortality in low and intermediate risk patients was 0.9% and 2.9% (P= 0.80), respectively (overall 1.5%). The CIFs in these groups were 20.1% and 32.5% (P= 0.06) for other cause mortality, respectively (overall 23.5%).

image

Figure 2. Kaplan–Meier curve showing overall survival, stratified by risk group.

Download figure to PowerPoint

image

Figure 3. Kaplan–Meier curve showing PCa-specific survival, stratified by risk group.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

We assessed the outcomes of patients initially managed expectantly with either low or intermediate risk PCa to evaluate the long-term feasibility of AS in a screen-detected population. Compared with men with low risk PCa at diagnosis, intermediate risk men, by definition, had more unfavourable disease characteristics at diagnosis. Deferred therapy, distant metastases and overall deaths occurred more often in the group with intermediate risk features. However, withholding radical treatment in low and intermediate risk screen-detected PCa leads to a substantial delay or even avoidance of therapy in a majority of men. Eventually, 43.4% of all patients received deferred treatment after a median of 2.6 years, while active therapy was avoided in 56.6% for a median of at least 6.8 years. Moreover, no difference was found in disease-specific survival between the two groups. The overall PCa-specific mortality was low, with a 10-year disease-specific survival rate of 98.4% for the entire cohort, despite a significant share (25.1%) of intermediate risk patients. After accounting for other cause mortality, competing risks analysis showed comparable results for overall and disease-specific mortality.

Previously, other studies have shown favourable outcomes of conservatively managed patients with low and intermediate risk disease. A recent report from a population-based nationwide cohort study in Sweden [9] showed a 10-year disease-specific mortality of 3.6% in a group of 2021 expectantly managed men (i.e. including both AS and watchful waiting) who had a mean age of 64.7 years and who were followed for a median of 8.2 years. The 10-year disease-specific mortality rates for low and intermediate risk men in that cohort were 2.4% and 5.2% respectively, while the overall mortality rate reached 20.4%. The corresponding rates for 10-year disease-specific mortality in our cohort are even more favourable: 0.9% and 3.9% for low and intermediate risk disease and 1.6% for the overall cohort, which might be due to the small number of events in both studies, the more stringent definition of low and intermediate risk PCa that we applied, or the difference in the number of screen-detected cases. Moreover, only two patients who died of PCa received deferred treatment with curative intent and potentially missed the window of curability because of initial expectant management, which is only 0.4% of our cohort and 2% of the overall deaths.

Another study presented results from the Surveillance, Epidemiology, and End Results (SEER) programme [10] of men who had a median age of 78 years with T1 or T2 PCa in whom curative therapy was not attempted and of whom 42% had received androgen deprivation therapy. The 10-year disease-specific mortality was 8.3–9.1% for well and intermediate differentiated tumours, while the overall mortality rate was 51.7% after a median follow-up of 8.3 years. Although these data were obtained in the PSA era and already show substantially lower mortality rates compared with the control arm of the Scandinavian Prostate Cancer Group 4 (SPCG-4) study [11,12], the overall health and the age of the patients in this study are not representative for the majority of patients who opt for AS in contemporary practice. Recently, results from the randomized Prostate Cancer Intervention vs Observation Trial (PIVOT) were presented [13], showing that RP did not reduce disease-specific survival more than observation in patients with low risk PCa after 10 years, which also suggests a favourable natural course of disease in patients who are considered to be at low risk.

Because AS is a relatively new treatment strategy, most prospective studies on the subject do not have data on long-term outcomes available. Klotz et al. [14] did publish results of an AS cohort comprising low and intermediate risk PCa cases (71% vs 29%) with a median follow-up of 6.8 years in men with a median age of 70.3 years. They described a 10-year disease-specific survival rate of 97.2%, while the 10-year overall survival rate was 68%. These survival rates are quite comparable but slightly lower than the rates in our cohort, which is most probably due to the higher median age in their series. The median follow-up of other prospective AS studies is still too short to evaluate disease-specific outcomes. The proportion of patients switching to deferred treatment in contemporary prospective AS cohorts varies between 14% and 35% [4,14–18]; the higher proportion of patients undergoing active therapy in our cohort might result from the lack of a follow-up protocol. Nevertheless, it was shown that the majority of treated patients switch to active treatment during the first 3 years after diagnosis, which is a common observation in most AS series and can, at least partly, be explained by risk reclassification of the disease [19].

Outcomes in intermediate risk patients in our cohort were quite favourable, showing similar disease-specific survival rates compared with the low risk cohort, despite less favourable baseline characteristics. Previously, Cooperberg et al. [15] also found favourable results in their AS cohort (n= 466) of patients with low risk and intermediate risk PCa; although intermediate risk patients had more adverse features at diagnosis, they did not undergo more active therapy or experience more progression. None of the men undergoing deferred RP were node positive and none experienced biochemical progression within 3 years. These results show that selected candidates with intermediate risk features can be followed on AS, preserving favourable long-term prognosis. In addition, in the current cohort, Cox regression analysis correcting for curative treatment (RP or radiation therapy) showed no difference between the risk groups regarding PCa mortality (data not shown). Moreover, it should be noted that no distinction could be made between men on watchful waiting and men on AS. In watchful waiting no follow-up protocol is used and the intent for curative treatment, if necessary, is lacking. Therefore, it can be expected that a true AS cohort with strict follow-up of their PSA kinetics and tumour characteristics with deferred active treatment on indication might show even better outcomes than the ones presented here.

To identify low risk PCa, we used PRIAS study criteria [4,7]. We believe the favourable results in this report confirm the feasibility of AS according to these criteria, but we are aware of the fact that inclusion and follow-up criteria still need further validation through outcome data with longer follow-up. Furthermore, we found that selected men with intermediate risk PCa may be appropriate candidates for AS as well. The higher overall mortality rate and use of hormonal therapy in the intermediate risk group indicate that the overall health in this group was worse than in the low risk group. And although the life expectancy in our cohort (median age 67 years) is longer than our median follow-up, the low number of adverse events over a 10-year time frame in the intermediate risk group supports the idea of less stringent criteria for AS in screen-detected men over 70 years or with significant comorbidities. Moreover, even in the case of deferred treatment, quality of life is likely to be maintained by prolonged therapy-free survival [20,21] and patients might profit from improved treatments or other potential advances in medical care.

This study has some limitations. Since no protocol was applied for surveillance, follow-up might have been less strict than in contemporary AS cohorts and patients who were actually candidates for watchful waiting very probably also form part of this cohort. However, more frequent follow-up such as is applied in contemporary AS protocols might have led to earlier detection of progression and even better outcomes in some men. Furthermore, potential variation in pathology reporting over time could have occurred due to the shift in Gleason scores [22,23], which could have influenced clinical outcomes. Follow-up was limited to 7.4 years and, regarding the median life expectancy for men in our cohort, longer follow-up remains useful to confirm our findings.

In conclusion, withholding radical treatment for men with low or intermediate risk screen-detected localized PCa may lead to a substantial delay or even avoidance of radical treatments and their side-effects. Favourable disease-specific outcomes over a 10-year period confirm the feasibility of AS for low risk PCa and also open the way for implementation of this method for intermediate risk disease for selected patients. The choice of therapy should always take comorbidity, patient age and patient preferences into account. Regarding the considerable difficulties in implementing trials randomizing for radical treatment and AS, answers to important questions on long-term mortality, and inclusion and follow-up criteria for AS, are more likely to come from prospective AS studies such as PRIAS.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The ERSPC is supported by grants from the Dutch Cancer Society (KWF 94-869, 98-1657, 2002-277, 2006-3518), the Netherlands Organization for Health Research and Development (002822820, 22000106, 50-50110-98-311), the 6th Framework Program of the EU: P-Mark: LSHC-CT-2004-503011, Beckman Coulter Hybritech Inc. and Europe against Cancer (SOC 95 35109, SOC 96 201869 05F02, SOC 97 201329, SOC 98 32241). The ERSPC received Erasmus MC and Ministry of Health institutional review board approval.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES