Progress toward Identifying Aggressive Prostate Cancer
http://www.100md.com
《新英格兰医药杂志》
Measurement of prostate-specific antigen (PSA) has profoundly affected virtually all clinical aspects of prostate cancer. A sharp increase in both the incidence of age-adjusted prostate cancer (about 100 percent) and the proportion of patients with early stages of the disease at the time of diagnosis (stage migration) has coincided with the advent of widespread PSA testing.1,2,3 Moreover, there have been substantial shifts toward the use of radical prostatectomy in younger men and men with lower pretreatment PSA levels and a rising incidence of nonpalpable lesions (tumor [T] stage 1c).3 Currently, less than 10 percent of men have distant metastases at the time of the initial diagnosis, and the proportion of patients offered local treatment with curative intent has increased substantially.
The decision regarding which local treatment to choose for prostate cancer is complicated by the multiplicity of options and the paucity of randomized trials. The choices range from watchful waiting to combined treatment. A substantial proportion of men in the age group most affected by prostate cancer die of other causes, yet the rate of death from prostate cancer remains high. In the United States alone, 82 men die of prostate cancer every day.
Nomograms, prognostic models, and artificial neural networks are among the tools frequently used by patients and physicians to help predict outcomes, choose a therapeutic intervention, and determine the prognosis.4,5,6 The first such nomogram, the Partin tables,7 provides estimates of the extent of local and regional disease at the time of prostatectomy on the basis of the pretreatment PSA level, the Gleason score of the biopsy sample, and the clinical stage. Nomograms and artificial neural networks predict the probability of biochemical relapse (as defined by increasing PSA levels) after radical prostatectomy on the basis of the pretreatment PSA level, percent positive biopsy, the Gleason score of the prostatectomy sample, the status of the surgical margins, and the presence or absence of involvement of seminal vesicles, capsular penetration, and lymph-node metastases.
Evidence of tumor outside the prostate gland and biochemical relapse may be indicative of incurable disease, but they are not necessarily predictors of death from prostate cancer. In this issue of the Journal, D'Amico et al.8 report that the rate of rise in the PSA level — the PSA velocity — during the year before radical prostatectomy predicts the risk of death from prostate cancer. They studied 1095 men who participated in a longitudinal community screening program and subsequently underwent radical prostatectomy with curative intent. The authors report that a PSA velocity of more than 2.0 ng per milliliter per year was an independent predictor of the risk of death from prostate cancer. Their observation is important, but a number of considerations merit discussion.
There are notable demographic differences between screened and unscreened populations of men. In the population screened by D'Amico et al., 71 percent had stage T1c disease and 84 percent had a Gleason score of 6 or less, as compared with respective values of 35 percent and 60 percent in a large, relatively recent registry of unscreened men.3 Furthermore, a considerable proportion of cancers diagnosed in other programs that actively screen for cancer are indolent, nonlethal forms of the disease that would otherwise not be diagnosed during the men's lifetime.2 In these men, treatment-related complications could exceed disease-related complications during the patient's lifetime.
D'Amico et al. suggest that determining the PSA velocity during the year before prostatectomy enhances the ability to identify men who may not require immediate treatment and who could be candidates for watchful waiting. As the authors point out, however, they did not investigate the comparative benefits and risks of treatment and watchful waiting in such men, and such factors can be defined only by carefully designed, randomized trials.
How much does the PSA velocity add to the predictions of outcome among poor-risk patients? In the study by D'Amico et al., only 28 percent of the men had a palpable tumor, only 6 percent had a preoperative PSA level of more than 10.0 ng per milliliter, and only 4 percent had a Gleason score of 8, 9, or 10 on needle biopsy. Interestingly, a Gleason score of 8, 9, or 10 on prostatectomy did not predict the risk of death from prostate cancer in a multivariate analysis; however, only 9 of 47 patients with such a Gleason score died of prostate cancer during a median follow-up of 4.8 years. Although the contribution of the PSA velocity to the prediction of the risk of death from prostate cancer in poor-risk men is suggestive, the numbers are small, the 95 percent confidence intervals are wide, and the duration of follow-up is relatively short. We question how much more we need to know about patients with a Gleason score of 8, 9, or 10 and a PSA level of more than 10.0 ng per milliliter to decide that they are in an unfavorable group who should be considered for clinical trials of aggressive combined treatment.
Currently, a substantial proportion of patients with unfavorable pathological findings at radical prostatectomy or with biochemical relapse receive immediate androgen-deprivation treatment, despite the lack of conclusive data that demonstrate a survival benefit. Information regarding a uniform policy for the implementation of systemic treatment in these kinds of patients was not provided by D'Amico et al. For this reason, the effect of systemic treatment after disease progression on the seven-year mortality figures presented by D'Amico et al. remains an unresolved and potentially confounding variable.
The rapidly growing body of information regarding the predictive value of a measure of PSA dynamics (PSA doubling time or PSA velocity) suggests that this approach will become critical in predicting prostate-cancer–specific survival. Our own experience at the Johns Hopkins Hospital suggests that in patients with biochemical relapse after primary treatment, the Gleason score, the time of the relapse, and the PSA doubling time independently predict the probability of distant metastasis.9 An updated analysis found that the PSA doubling time overrides the two other variables; Kaplan–Meier estimates of 10-year cancer-specific survival were 93 percent among patients with a PSA doubling time of 10 months or more, as compared with 58 percent among men with a doubling time of less than 10 months.9 Our experience is consistent with the report of D'Amico et al.10 that a doubling time of three months or less in patients with a biochemical relapse after prostatectomy or radiation was associated with an increased risk of death from prostate cancer. Furthermore, D'Amico et al.11 reported a median survival of 5 years among patients with a PSA doubling time of less than 12 months after radiation therapy. These survival figures are similar to the median survival among patients with early bone metastases who are receiving hormonal therapy12 and suggest that in men with a biochemical relapse, the PSA doubling time identifies those with systemic disease who are most likely to die of prostate cancer.
In their current report, D'Amico et al. provide evidence that the preoperative PSA velocity predicts the risk of dying of prostate cancer and that this measurement, together with other clinical and pathological data, may be used to enhance the identification of aggressive prostate cancer. Assessment of PSA dynamics may eventually be the key factor in selecting patients with disease for which expectant management may be suitable.
Source Information
From the Johns Hopkins Medical Institutions, Baltimore.
References
Stanford JL, Stephenson RA, Coyle LM, et al. Prostate cancer trends 1973-1995, SEER program. Bethesda, Md.: National Cancer Institute, 1999. (NIH publication no 99-4543.)
Etzioni R, Penson DF, Legler JM, et al. Overdiagnosis due to prostate-specific antigen screening: lessons from U.S. prostate cancer incidence trends. J Natl Cancer Inst 2002;94:981-990.
Cooperberg MR, Broering JM, Litwin MS, et al. The contemporary management of prostate cancer in the United States: lessons from the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE), a national disease registry. J Urol 2004;171:1393-1401.
Ross PL, Gerigk C, Gonen M, et al. Comparisons of nomograms and urologists' predictions in prostate cancer. Semin Urol Oncol 2002;20:82-88.
Vergouwe Y, Steyerberg EW, Eijkemans MJC, Habbema JDF. Validity of prognostic models: when is a model clinically useful? Semin Urol Oncol 2002;20:96-107.
Schwarzer G, Schumaker M. Artificial neural networks for diagnosis and prognosis in prostate cancer. Semin Urol Oncol 2002;20:89-95.
Partin AW, Mangold LA, Lamm DM, Walsh PC, Epstein JI, Pearson JD. Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium. Urology 2001;58:843-848.
D'Amico AV, Chen M-H, Roehl KA, Catalona WJ. Preoperative PSA velocity and the risk of death from prostate cancer after radical prostatectomy. N Engl J Med 2004;351:125-135.
Partin AW, Eisenberger MA, Sinibaldi V, Humphreys E, Mangold LA, Walsh PC. Prostate specific antigen doubling time predicts for distant failure and prostate cancer specific survival in men with biochemical relapse after radical prostatectomy. Prog Proc Am Soc Clin Oncol 2004;23. abstract.
D'Amico AV, Moul JW, Carroll PR, Sun L, Lubeck D, Chen MH. Surrogate end point for prostate cancer-specific mortality after radical prostatectomy or radiation therapy. J Natl Cancer Inst 2003;95:1376-1383.
D'Amico AV, Cote K, Loffredo M, Renshaw AA, Schultz D. Determinants of prostate cancer-specific survival after radiation therapy for patients with clinically localized prostate cancer. J Clin Oncol 2002;20:4567-4573.
Eisenberger MA, Blumenstein BA, Crawford ED, et al. Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med 1998;339:591-598.(Mario Eisenberger, M.D., )
The decision regarding which local treatment to choose for prostate cancer is complicated by the multiplicity of options and the paucity of randomized trials. The choices range from watchful waiting to combined treatment. A substantial proportion of men in the age group most affected by prostate cancer die of other causes, yet the rate of death from prostate cancer remains high. In the United States alone, 82 men die of prostate cancer every day.
Nomograms, prognostic models, and artificial neural networks are among the tools frequently used by patients and physicians to help predict outcomes, choose a therapeutic intervention, and determine the prognosis.4,5,6 The first such nomogram, the Partin tables,7 provides estimates of the extent of local and regional disease at the time of prostatectomy on the basis of the pretreatment PSA level, the Gleason score of the biopsy sample, and the clinical stage. Nomograms and artificial neural networks predict the probability of biochemical relapse (as defined by increasing PSA levels) after radical prostatectomy on the basis of the pretreatment PSA level, percent positive biopsy, the Gleason score of the prostatectomy sample, the status of the surgical margins, and the presence or absence of involvement of seminal vesicles, capsular penetration, and lymph-node metastases.
Evidence of tumor outside the prostate gland and biochemical relapse may be indicative of incurable disease, but they are not necessarily predictors of death from prostate cancer. In this issue of the Journal, D'Amico et al.8 report that the rate of rise in the PSA level — the PSA velocity — during the year before radical prostatectomy predicts the risk of death from prostate cancer. They studied 1095 men who participated in a longitudinal community screening program and subsequently underwent radical prostatectomy with curative intent. The authors report that a PSA velocity of more than 2.0 ng per milliliter per year was an independent predictor of the risk of death from prostate cancer. Their observation is important, but a number of considerations merit discussion.
There are notable demographic differences between screened and unscreened populations of men. In the population screened by D'Amico et al., 71 percent had stage T1c disease and 84 percent had a Gleason score of 6 or less, as compared with respective values of 35 percent and 60 percent in a large, relatively recent registry of unscreened men.3 Furthermore, a considerable proportion of cancers diagnosed in other programs that actively screen for cancer are indolent, nonlethal forms of the disease that would otherwise not be diagnosed during the men's lifetime.2 In these men, treatment-related complications could exceed disease-related complications during the patient's lifetime.
D'Amico et al. suggest that determining the PSA velocity during the year before prostatectomy enhances the ability to identify men who may not require immediate treatment and who could be candidates for watchful waiting. As the authors point out, however, they did not investigate the comparative benefits and risks of treatment and watchful waiting in such men, and such factors can be defined only by carefully designed, randomized trials.
How much does the PSA velocity add to the predictions of outcome among poor-risk patients? In the study by D'Amico et al., only 28 percent of the men had a palpable tumor, only 6 percent had a preoperative PSA level of more than 10.0 ng per milliliter, and only 4 percent had a Gleason score of 8, 9, or 10 on needle biopsy. Interestingly, a Gleason score of 8, 9, or 10 on prostatectomy did not predict the risk of death from prostate cancer in a multivariate analysis; however, only 9 of 47 patients with such a Gleason score died of prostate cancer during a median follow-up of 4.8 years. Although the contribution of the PSA velocity to the prediction of the risk of death from prostate cancer in poor-risk men is suggestive, the numbers are small, the 95 percent confidence intervals are wide, and the duration of follow-up is relatively short. We question how much more we need to know about patients with a Gleason score of 8, 9, or 10 and a PSA level of more than 10.0 ng per milliliter to decide that they are in an unfavorable group who should be considered for clinical trials of aggressive combined treatment.
Currently, a substantial proportion of patients with unfavorable pathological findings at radical prostatectomy or with biochemical relapse receive immediate androgen-deprivation treatment, despite the lack of conclusive data that demonstrate a survival benefit. Information regarding a uniform policy for the implementation of systemic treatment in these kinds of patients was not provided by D'Amico et al. For this reason, the effect of systemic treatment after disease progression on the seven-year mortality figures presented by D'Amico et al. remains an unresolved and potentially confounding variable.
The rapidly growing body of information regarding the predictive value of a measure of PSA dynamics (PSA doubling time or PSA velocity) suggests that this approach will become critical in predicting prostate-cancer–specific survival. Our own experience at the Johns Hopkins Hospital suggests that in patients with biochemical relapse after primary treatment, the Gleason score, the time of the relapse, and the PSA doubling time independently predict the probability of distant metastasis.9 An updated analysis found that the PSA doubling time overrides the two other variables; Kaplan–Meier estimates of 10-year cancer-specific survival were 93 percent among patients with a PSA doubling time of 10 months or more, as compared with 58 percent among men with a doubling time of less than 10 months.9 Our experience is consistent with the report of D'Amico et al.10 that a doubling time of three months or less in patients with a biochemical relapse after prostatectomy or radiation was associated with an increased risk of death from prostate cancer. Furthermore, D'Amico et al.11 reported a median survival of 5 years among patients with a PSA doubling time of less than 12 months after radiation therapy. These survival figures are similar to the median survival among patients with early bone metastases who are receiving hormonal therapy12 and suggest that in men with a biochemical relapse, the PSA doubling time identifies those with systemic disease who are most likely to die of prostate cancer.
In their current report, D'Amico et al. provide evidence that the preoperative PSA velocity predicts the risk of dying of prostate cancer and that this measurement, together with other clinical and pathological data, may be used to enhance the identification of aggressive prostate cancer. Assessment of PSA dynamics may eventually be the key factor in selecting patients with disease for which expectant management may be suitable.
Source Information
From the Johns Hopkins Medical Institutions, Baltimore.
References
Stanford JL, Stephenson RA, Coyle LM, et al. Prostate cancer trends 1973-1995, SEER program. Bethesda, Md.: National Cancer Institute, 1999. (NIH publication no 99-4543.)
Etzioni R, Penson DF, Legler JM, et al. Overdiagnosis due to prostate-specific antigen screening: lessons from U.S. prostate cancer incidence trends. J Natl Cancer Inst 2002;94:981-990.
Cooperberg MR, Broering JM, Litwin MS, et al. The contemporary management of prostate cancer in the United States: lessons from the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE), a national disease registry. J Urol 2004;171:1393-1401.
Ross PL, Gerigk C, Gonen M, et al. Comparisons of nomograms and urologists' predictions in prostate cancer. Semin Urol Oncol 2002;20:82-88.
Vergouwe Y, Steyerberg EW, Eijkemans MJC, Habbema JDF. Validity of prognostic models: when is a model clinically useful? Semin Urol Oncol 2002;20:96-107.
Schwarzer G, Schumaker M. Artificial neural networks for diagnosis and prognosis in prostate cancer. Semin Urol Oncol 2002;20:89-95.
Partin AW, Mangold LA, Lamm DM, Walsh PC, Epstein JI, Pearson JD. Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium. Urology 2001;58:843-848.
D'Amico AV, Chen M-H, Roehl KA, Catalona WJ. Preoperative PSA velocity and the risk of death from prostate cancer after radical prostatectomy. N Engl J Med 2004;351:125-135.
Partin AW, Eisenberger MA, Sinibaldi V, Humphreys E, Mangold LA, Walsh PC. Prostate specific antigen doubling time predicts for distant failure and prostate cancer specific survival in men with biochemical relapse after radical prostatectomy. Prog Proc Am Soc Clin Oncol 2004;23. abstract.
D'Amico AV, Moul JW, Carroll PR, Sun L, Lubeck D, Chen MH. Surrogate end point for prostate cancer-specific mortality after radical prostatectomy or radiation therapy. J Natl Cancer Inst 2003;95:1376-1383.
D'Amico AV, Cote K, Loffredo M, Renshaw AA, Schultz D. Determinants of prostate cancer-specific survival after radiation therapy for patients with clinically localized prostate cancer. J Clin Oncol 2002;20:4567-4573.
Eisenberger MA, Blumenstein BA, Crawford ED, et al. Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med 1998;339:591-598.(Mario Eisenberger, M.D., )