Correspondence University of Toronto, Chief, Division of Urology, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue #MG408, Toronto, ON M4N 3M5, Canada

Email laurence.klotz@sunnybrook.ca

Introduction

Why would any healthy person, newly diagnosed with cancer, elect not to have curative treatment? This reasonable question is often posed by patients who have newly diagnosed prostate cancer and are presented with the option of active surveillance. Communicating the answer to this question clearly is a crucial component of the active-surveillance strategy. The basis for this approach, the principles of patient selection and the indications for intervention during surveillance are reviewed in this article.

Active surveillance is an option for men with screen-detected, low-volume cancer. Patients with prostate cancer who fall into this category include men with a Gleason score of 6 or less, those with a prostate-specific antigen (PSA) value of 10 ng/ml or less and those with stage T1c or T2a disease. This definition encompasses about 45% of patients with newly diagnosed prostate cancer in a serially screened population. The concept is derived from the following five postulates: first, that screening for prostate cancer results in the detection of disease that in many patients is not clinically significant (i.e. if untreated, it would not pose a threat to health). Second, the patients who fall into this category can be identified with reasonable accuracy. Third, no treatment is minimal in terms of side effects and cost. Fourth, patients who are initially classified as low risk but who are reclassified over time as higher risk and treated radically are still cured in most cases. Finally, the psychological burden of living with untreated cancer has less impact on the quality of life than unnecessary, but curative, therapy.

Rational selection of patients for a surveillance strategy should be guided by these postulates. It is important to identify patients who have a low probability of disease progression during their lifetime according to clinical and pathologic features of the disease, patient age and comorbidity. Close monitoring of PSA levels and prostate pathology over time is essential. It is important that reasonable criteria for intervention are used, which will both identify patients with more-aggressive disease in a timely fashion and not result in excessive treatment, and to communicate appropriately with the patient, to reduce the psychological burden of living with untreated cancer. This article reviews the published data in the context of the five postulates above.

Postulate 1: detection through screening

Prostate-cancer screening results in the detection of disease that in many patients is not clinically significant. Screening for prostate cancer on the basis of prostate biopsy for men with elevated levels of serum PSA or an abnormal digital rectal examination results in many men whose lives are not at threat from prostate cancer being diagnosed with the disease. In the US, there are 2.74 million men who are aged 50–70 years and have a PSA level greater than 2.5 ng/ml. Catalona's group advocates using a PSA cut-off point of 2.5 ng/ml for biopsy in men aged 50 years or older.¹ If all American men in this age-group with a PSA level greater than 2.5 ng/ml were subjected to a biopsy, 775,000 cases of prostate cancer would be diagnosed this year in the US alone. This is 543,000 more cases than the 232,000 diagnosed in 2005, and 25 times the 30,350 men expected to die of prostate cancer each year in the US.² PSA screening has resulted in a progressive decrease in the extent of cancer present at the time of diagnosis (i.e. stage migration). As a consequence, the proportion of patients newly diagnosed with prostate cancer who fall into the 'favorable-risk' category has increased and now constitutes almost half of patients with the disease.³

The aggressiveness of cancer can be predicted by the use of existing clinical parameters. The parameters most widely used are the tumor grade or Gleason score, PSA level, and tumor stage. Favorable-risk prostate cancer is characterized by a Gleason score of 6 or less, a PSA level of 10 ng/ml or less, and stage T1c–T2a disease.⁴ Importantly, although patients with these characteristics have a much more favorable natural history and progression rate than those who have a higher Gleason grade or PSA level, in a few patients tumors will still progress to advanced, incurable prostate cancer and death.

Autopsy studies have demonstrated that prostate cancer typically begins in the third or fourth decade of life.⁵ In most patients, this means that there is a period of slow subclinical tumor progression from a microscopic focus of disease, which lasts approximately 30 years, followed by a period of clinical progression (potentially, but not invariably, leading to metastatic disease and death), which lasts about 15 years. The implication is that most patients have a long window of curability. This is particularly true for patients with favorable-risk, low-volume disease.

Data on a large group of patients in Connecticut, US assessed by watchful waiting have been reported at 20 years.⁶ These data confirm the powerful predictive value of the Gleason score. In the pre-PSA screening cohort, 23% of untreated patients with a Gleason score of 6 died of prostate cancer within 20 years. Of the patients with a Gleason score of 7, about 65% died of prostate cancer within 20 years. In addition, the authors recently reanalyzed prostate pathology slides that were initially read between 1990 and 1992, using contemporary Gleason scoring.⁷ The study demonstrated clearly that there has been a shift in grade interpretation during the past 15 years. This shift is largely the result of the removal of Gleason scores 2–4 from the classification, the inclusion of small proportions of high-grade cancer into the grading system, and the reclassification of cribriform-pattern cancer from a Gleason score of 3 to a Gleason score of 4. Many cancers diagnosed 20 years ago as a Gleason score of 6 would be classified as a Gleason score of 7 today. The Connecticut study results probably represent a 'worst case' scenario for the expected mortality from untreated cancer with a Gleason score of 6. This means that the mortality from untreated, non-screen-detected prostate cancer in patients with contemporary Gleason scores of 6 might be as low as 10% at 20 years.

Postulate 2: patient selection for surveillance

The patients who fall into the active-surveillance category can be identified with reasonable accuracy. The 'gold standard' for clinically insignificant prostate cancer, used by virtually all clinicians who attempt to predict minimal disease according to clinical parameters, is a radical prostatectomy specimen containing less than 0.5 ml of prostate cancer with a Gleason score of 6 or less. Stamey et al. devised this reference standard after examination of prostate glands obtained from 139 consecutively sampled radical cystoprostatectomy specimens, of which 55 (40%) had incidental prostate cancer.⁸ The authors concluded that only tumor volumes above the 92^nd percentile (0.5–6.1 ml) were clinically significant, according to the assumption that the clinically significant cancer rate and clinical prevalence were both 8%. This approach is arbitrary and concerning.

Many groups have reported the incidence of insignificant disease using this definition.^{9, 10, 11, 12, 13, 14, 15, 16, 17, 18} The incidence varies from 30% in patients with stage T1c disease, as reported by Epstein et al.,⁹ to values as low as 9–12%.¹⁹ The clinical criteria for predicting minimal disease include a Gleason score of 6 or less, a low PSA density of <0.15, a total extent of cancer involved of less than 3 mm, an involvement of less than one-third of cores and an extent of core involvement of <50% (Table 1). Importantly, the Epstein criteria allow up to 50% involvement of individual cores, which represents much more substantial disease than a few microfoci. As mentioned above, this definition is derived from a pathologic end point. The definition of insignificant cancer as less than 0.5 cc of low-grade disease has never been validated in a trial with a clinical end point. Substantial study data, including the results of the Prostate Cancer Prevention Trial (PCPT)²⁰ and the incontrovertible ratio between the current lifetime probability of diagnosis (about 1 in 6) and death (1 in 40) of 7:1 suggest that the Epstein criteria understate the proportion of patients who harbor prostate cancer that would not pose a threat to their life (i.e. about six out of seven patients). So, patients fulfilling the Epstein criteria for insignificant disease (i.e. a Gleason score of 6 or less, less than one-third of positive cores and an involvement of 50% or less of individual cores) represent optimal candidates for active surveillance. The criteria for tumor volume can be relaxed for patients over the age of 65 years, and patients over 75 years might be candidates if they have a Gleason score of 7 (3 + 4).

Table 1 Use of pathologic and prostate-specific-antigen parameters to predict 'insignificant' prostate cancer.

Full tableFigures & Tables indexDownload Power Point slide (238K)

Postulate 3: adverse effects and cost considerations

No treatment is minimal in terms of adverse effects and cost. All widely accepted, local, definitive treatments for prostate cancer, including surgery, brachytherapy, external-beam irradiation and cryosurgery, induced erectile dysfunction in a considerable proportion of patients, and these treatments were associated with serious adverse effects in several patients.

Postulate 4: patient-risk reclassification

Choo et al.^{21, 28} and Klotz et al.²⁹ were the first groups to use a prospective active-surveillance protocol that incorporated a selective, delayed intervention for the patient subset with rapid PSA or grade progression on repeat biopsy. The eligibility criteria included patients with stage T1c or T2a prostate cancer who had a Gleason score of 6 or less and a PSA level of 10 ng/ml or less. For patients over 70 years of age, these criteria were relaxed to include a Gleason score of 7 (3 + 4) or less and/or a PSA value of 15 ng/ml or less. The cohort of active-surveillance patients comprised 299 patients.²⁹ The median age was 70 years (range 49–84 years); 80% of patients had a Gleason score of 6 or less, and the same proportion had a PSA level greater than 10 ng/ml (median 6.5 ng/ml). The median follow-up period was 72 months, during which time 101 (34%) patients came off active surveillance, whereas 198 patients remained on active surveillance. Of patients discontinuing surveillance, the reason was rapid biochemical progression in 15%, clinical progression in 3%, histologic progression in 4%, and patient preference in 12%. There was no correlation between grade progression and PSA doubling time. At a median follow-up of 7 years (range 2–11 years), the overall survival was 85% and disease-specific survival was 99%.

Only 3 out of 299 patients had died of prostate cancer at the time of writing this review. All three patients had PSA doubling times of greater than 2 years, and death occurred 3.0, 5.1 and 5.2 years after diagnosis. All three patients exhibited the same pattern of clinical progression: initial favorable prognostic factors; a rapid rise in PSA level, which led to treatment at 6, 9 and 11 months, respectively, after the initial diagnosis; and a further progressive rise in PSA level and clinically apparent bone metastases within 1 year of treatment, leading to androgen-deprivation therapy. All three patients died within 3 years of the initiation of hormone therapy. This very rapid progression after diagnosis indicates that these patients had occult metastases at the time of initial disease presentation and their outcome would not have been altered by earlier treatment. In the Swedish trial, there was almost no difference between the two groups before 5 years had elapsed.³⁰

In the Choo et al. study, the median PSA doubling time, calculated by logarithmic regression, was 7 years.²¹ In total, 22% of patients had a PSA doubling time of greater than 3 years, whereas 42% had a PSA doubling time of over 10 years, which indicates an indolent course of disease in these patients. The Gleason score remained stable in 92% of patients; only 8% of patients demonstrated a marked rise in their Gleason score, classed as an increase of at least two points. In this group, 29 patients (10% of the cohort) had a radical prostatectomy as a result of a short PSA doubling time or tumor-grade progression. All these patients had an initial Gleason score of 5 or 6, a PSA value greater than 10 ng/ml and a tumor stage of pT1–2 at study entry. The final pathology was stage pT2 in 18 patients, stage pT3a in 11 patients, stage T3c in 1 patient and node-positive in 1 patient. Of the 18 patients with a PSA doubling time of greater than 3 years, only 7 patients had positive margins. This result suggests that even among the worst subset of the cohort, that is to say those reclassified as higher risk over time, the majority could be cured by delayed therapy.

Prostate-specific-antigen doubling time or velocity as a criterion for intervention

In the study by Klotz et al. a PSA doubling time of greater than 3 years was used as the threshold for intervention.²⁹ The PSA doubling time was calculated by modeling the natural logarithm of PSA (ln[PSA]), using a generalized linear mixed model, derived from a multivariate analysis of the surveillance cohort. In this model, the baseline PSA value, tumor grade and patient age predicted for the subsequent PSA doubling time. Thus, the model enables correction for these variables.

Using this approach, the proportion and frequency of stable, untreated patients who would have been offered treatment according to the following PSA doubling time thresholds for radical intervention was calculated: a PSA threshold level of 10 ng/ml for patients with an initial PSA level of less than 10 ng/ml; a PSA threshold level of 20 ng/ml; a linear regression of ln(PSA) versus a doubling time of less than 3 years for all PSA values; ln(PSA) versus a doubling time of less than 3 years using the first and last PSA levels on record; an actual PSA velocity of greater than 2 years during the past year; a calculated PSA velocity of greater than 2 years; and a general linear mixed model of ln(PSA). In this analysis, 134 patients remained on surveillance, with a minimum of 2 years' follow-up. As of January 2007, the median follow-up was 5.8 years (range 2.0–10.5 years). No patient died of prostate cancer or had metastatic disease; 14 patients (10.4%) died of other causes. The proportion of patients who would have received treatment if various triggers were used is listed in Table 2. It was concluded that patients followed up on surveillance may be overtreated if the threshold for intervention is a PSA level of greater than 10 ng/ml, a PSA level greater than 20 ng/ml, a PSA doubling time derived from the first and last values, or an actual or calculated PSA velocity of greater than 2 years.

Table 2 Results of various prostate-specific-antigen threshold levels for radical intervention in the Sunnybrook 'stable' cohort.

Full tableFigures & Tables indexDownload Power Point slide (230K)

Carter et al. have used a similar approach in 81 men, but subjected the patients to annual prostate biopsies.³¹ The threshold for active intervention was determined solely by the detection of disease progression on surveillance biopsy samples of the prostate (i.e. a Gleason score of 4 or greater, more than two cores positive for cancer, or more than 50% involvement of any core with cancer) and not on the basis of changes in serum PSA levels alone. Of the 25 men who exhibited progression, 13 underwent radical prostatectomy; 12 (92%) of these 13 men had curable cancer. A current controversy in the field relates to the degree to which PSA kinetics serve as a reliable measure for intervention and whether more-frequent biopsy, with intervention at the first sign of histologic progression, is superior. This question will probably be resolved by further experience.

Postulate 5: the psychological burden of untreated cancer

The psychological burden of living with untreated cancer has less impact on the quality of life than unnecessary, but curative, therapy. The psychological effects of living for many years with untreated cancer are unknown; however, the evidence indicates it is the diagnosis of cancer that takes a psychological toll, whether patients are treated curatively or not. A companion study to the randomized trial of surgery versus watchful waiting in Sweden described above found no significant psychological difference at 5 years between the two study arms.³² The occurrence of worry, anxiety or depression was equal between the treatment and the observation arms. Surveillance is clearly stressful for some men; however, concern about PSA recurrence is common among both treated and untreated patients. Patients who are educated to appreciate the indolent natural history of most good-risk prostate cancers might avoid much of this anxiety. Further quality-of-life studies focusing on this issue are clearly warranted. Our follow-up strategy for managing patients with active surveillance and selective, delayed intervention is described in Box 1.

Conclusions

A rational approach to active surveillance involves patient selection for initial surveillance on the basis of a Gleason score of 6 or less, a PSA level 10 ng/ml or less, and stage T1c–T2a tumors. For men younger than 60 years, the Epstein criteria of the involvement of less than one-third of cores, and no more than 50% involvement of individual cores, are also warranted. For men older than 75 years or those with a life expectancy of less than 10 years because of comorbidity, the PSA threshold level can be greater than 10 ng/ml or the Gleason score can be 7 (3 + 4).

The second component of surveillance is patient selection for radical intervention after a period of observation. Our approach has been to use a PSA doubling time of less than 3 years (i.e. applies to 20% of patients) or grade progression to a Gleason score of 7 (4 + 3) or higher (i.e. applies to 5% of patients). We have used the generalized linear mixed method, incorporating ln(PSA) to calculate the PSA doubling time. Using a measure of a PSA velocity of greater than 2.0 ng/ml/year might result in overtreatment of stable patients. In our cohort, 50% of stable patients (100% of whom remain untreated and free of disease progression) would have had radical intervention during surveillance with use of this criterion.

It should be emphasized that these are guidelines, not rules; clinical judgment is required. For example, progression to a Gleason score of 3 + 4 in a young patient might be considered an indication for intervention by many practitioners and might warrant continued observation in an elderly individual or a patient with considerable comorbidity. The approach of active surveillance affords the opportunity for low-risk patients at a very low risk of prostate cancer mortality to avoid the adverse effects of radical treatment. This approach seems safe and requires validation in a large-scale randomized study.

Acknowledgments

Désirée Lie, University of California, Irvine, CA, is the author of and is solely responsible for the content of the learning objectives, questions and answers of the Medscape-accredited continuing medical education activity associated with this article.

References

1. Nadler RB et al. (2005) Use of 2.6 ng/ml prostate specific antigen prompt for biopsy in men older than 60 years. J Urol 174: 2154–2157 | Article | PubMed |
2. Welch HG et al. (2005) Prostate-specific antigen levels in the United States: implications of various definitions for abnormal. J Natl Cancer Inst 97: 1132–1137 | PubMed |
3. Jemal A et al. (2007) Cancer Statistics, 2007. CA Cancer J Clin 57: 43–66 | PubMed | ISI |
4. D'Amico AV et al. (1998) Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. J Insur Med 30: 204–205 | PubMed |
5. Sakr WA et al. (1993) The frequency of carcinoma and intraepithelial neoplasia of the prostate in young male patients. J Urol 150: 379 | PubMed | ISI | ChemPort |
6. Albertsen P et al. (2005) 20-Year outcomes following conservative management of clinically localized prostate cancer. JAMA 293: 2095–2101 | Article | PubMed | ISI | ChemPort |
7. Albertsen PC et al. (2005) Prostate cancer and the Will Rogers phenomenon. J Natl Cancer Inst 97: 1248–1253 | PubMed |
8. Stamey TA et al. (1993) Localized prostate cancer: relationship of tumor volume to clinical significance for treatment of prostate cancer. Cancer 71 (Suppl): S933–S938 | Article |
9. Epstein JI et al. (1994) Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 271: 368–374 | Article | PubMed | ISI | ChemPort |
10. Epstein JI et al. (1993) Correlation of pathologic findings with progression after radical retropubic prostatectomy. Cancer 71: 3582–3593 | Article | PubMed | ChemPort |
11. Irwin MB and Trapasso JG (1994) Identification of insignificant prostate cancers: analysis of preoperative parameters. Urology 44: 862–867 | Article | PubMed | ISI | ChemPort |
12. Cupp MR et al. (1995) The volume of prostate cancer in the biopsy specimen cannot reliably predict the quantity of cancer in the radical prostatectomy specimen on an individual basis. J Urol 153: 1543–1548 | Article | PubMed | ISI | ChemPort |
13. Goto Y et al. (1996) Distinguishing clinically important from unimportant prostate cancers before treatment: value of systematic biopsies. J Urol 156: 1059–1063 | Article | PubMed | ISI | ChemPort |
14. Epstein JI et al. (1993) Is tumor volume an independent predictor of progression following radical prostatectomy? A multivariate analysis of 185 clinical stage B adenocarcinomas of the prostate with 5 years of followup. J Urol 149: 1478–1481 | PubMed | ISI | ChemPort |
15. Epstein JI et al. (1998) Nonpalpable stage T1c prostate cancer: prediction of insignificant disease using free/total prostate specific antigen levels and needle biopsy findings. J Urol 160: 2407–2411 | Article | PubMed | ISI | ChemPort |
16. Noguchi M et al. (2001) Relationship between systematic biopsies and histological features of 222 radical prostatectomy specimens: lack of prediction of tumor significance for men with nonpalpable prostate cancer. J Urol 166: 104–109 | Article | PubMed | ISI | ChemPort |
17. Augustin H et al. (2003) Insignificant prostate cancer in radical prostatectomy specimen: time trends and preoperative prediction. Eur Urol 43: 455–460 | Article | PubMed | ISI |
18. Anast JW et al. (2004) Relating biopsy and clinical variables to radical prostatectomy findings: can insignificant and advanced prostate cancer be predicted in a screening population? Urology 64: 544–550 | Article | PubMed | ISI |
19. Graif T et al. (2007) Under diagnosis and over diagnosis of prostate cancer. J Urol 178: 88–92 | Article | PubMed |
20. Thompson IM et al. (2003) The influence of finasteride on the development of prostate cancer. N Engl J Med 349: 215–224 | Article | PubMed | ISI | ChemPort |
21. Choo R et al. (2001) PSA doubling time of prostate carcinoma managed with watchful observation alone. Int J Radiat Oncol Biol Phys 50: 615–620 | Article | PubMed | ChemPort |
22. Roemeling S et al. (2007) Active surveillance for prostate cancers detected in three subsequent rounds of a screening trial: characteristics, PSA doubling times, and outcome. Eur Urol 51: 1244–1250 | Article | PubMed |
23. Kakehi Y (2003) PSA DT in Japanese active surveillance cohort (48 patients). Jpn J Clin Oncol 33: 1–5 | Article | PubMed |
24. Teahan SJ and Klotz LH (2006) Current role of prostate-specific antigen kinetics in managing patients with prostate cancer. BJU Int 97: 451–455 | Article | PubMed |
25. Egawa S et al. (2000) Use of pretreatment prostate-specific antigen doubling time to predict outcome after radical prostatectomy. Prostate Cancer Prostatic Dis 3: 269–274 | Article | PubMed | ISI |
26. McLaren DB et al. (1998) Watchful waiting or watchful progression? Prostate specific antigen doubling times and clinical behavior in patients with early untreated prostate carcinoma. Cancer 82: 342–348 | Article | PubMed | ChemPort |
27. D'Amico AV et al. (2004) Preoperative PSA velocity and the risk of death from prostate cancer after radical prostatectomy. N Engl J Med 351: 125–135 | Article | PubMed | ISI | ChemPort |
28. Choo R et al. (2002) Feasibility study: watchful waiting for localized low to intermediate grade prostate carcinoma with selective delayed intervention based on prostate specific antigen, histological and/or clinical progression. J Urol 167: 1664–1669 | Article | PubMed | ISI |
29. Klotz L (2005) Active surveillance for prostate cancer: for whom? J Clin Oncol 23: 8165–8169 | Article | PubMed | ISI |
30. Bill-Axelson A et al. (2005) Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med 352: 1977–1984 | Article | PubMed | ISI | ChemPort |
31. Khan MA et al. (2003) Expectant management of localized prostate cancer. Urology 62: 793–799 | Article | PubMed |
32. Steineck G et al. (2002) Quality of life after radical prostatectomy or watchful waiting. N Engl J Med 347: 790–796 | Article | PubMed | ISI |

Contact the journal about this article

Subject areas under which this article appears: Surgical Oncology | Radiotherapy | Screening