Advertisement
UROLOGY
Advanced searchSave search
Oncology| Volume 118, P119-126, August 2018

Download started.

Ok

New Evidence for the Benefit of Prostate-specific Antigen Screening: Data From 400,887 Kaiser Permanente Patients

  • Paul F. Alpert
    Correspondence
    Address correspondence to: Paul F. Alpert, M.D., Department of Urology Academic Office, University of California Medical Center, 400 Parnassus Avenue, A633, San Francisco, CA 94143-0738.
    Affiliations
    Department of Urology, Kaiser Permanente Medical Center, San Rafael, CA

    Department of Urology, University of California San Francisco School of Medicine, San Francisco, CA
    Search for articles by this author
Open AccessPublished:May 14, 2018DOI:https://doi.org/10.1016/j.urology.2018.02.049
New Evidence for the Benefit of Prostate-specific Antigen Screening: Data From 400,887 Kaiser Permanente Patients
Previous ArticleHistologic and Oncologic Outcomes Following Liver Mass Resection With Retroperitoneal Lymph Node Dissection in Patients With Nonseminomatous Germ Cell Tumor
Next ArticleEditorial Comment

      Objective

      To investigate whether prostate cancer screening with prostate-specific antigen (PSA) is beneficial in reducing prostate cancer mortality, and to determine optimal screening intervals and age groups to be screened.

      Methods

      This is a retrospective cohort study of 400,887 men under age 80, with no history of prostate cancer, who had PSA testing at Kaiser Permanente Northern California in the 5 calendar years 1998-2002, and were followed up for 12-16 years. Subjects were stratified into 6 groups based on the screening interval, and into 7 groups based on age. Prostate cancer mortality rates for each of the 42 subgroups were calculated and compared.

      Results

      The data show that yearly PSA screening is beneficial, reducing prostate cancer deaths by 64% for men aged 55-75 years (95% confidence interval 50-78%, P <.001), and all-cause mortality by 24% (95% confidence interval 15%-34%, P <.001). This is the first study to evaluate various screening intervals and age groups, showing that yearly screening is the interval of choice. No benefit was found for screening at any interval for men under age 55.

      Conclusion

      Yearly PSA screening is highly effective in reducing both prostate cancer mortality and all-cause mortality in men with prostate cancer, and when combined with active surveillance to prevent overtreatment, lends support for PSA screening for men in good health aged 55-75.
      Prostate cancer is the most common malignancy in men in the United States, with over 167,000 new cases and almost 27,000 cancer deaths per year.
      • Siegel R.
      • Miller K.
      • Jemal A.
      Cancer statistics, 2017.
      CA Cancer J Clin. 2017; 67: 7-30
      When prostate-specific antigen (PSA) testing became available in 1986, it was hoped that through early detection, death from prostate cancer could be substantially reduced. But after 30 years, evidence for its usefulness has been questioned. In May 2012, the United States Preventive Services Task Force (USPSTF) downgraded their recommendation to Grade D, now advising against PSA screening in healthy men, concluding that PSA screening causes overtreatment and that the modest benefits of screening are outweighed by the harms.
      • Moyer V.A.
      Screening for prostate cancer: U.S. preventive services task force recommendation statement.
      Ann Intern Med. 2012; 157: 120-134
      There have been 2 large randomized prospective studies of PSA screening. The first was the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial published in 2009.
      • Andriole G.L.
      • Crawford E.D.
      • Grubb R.L.
      • et al.
      Mortality results from a randomized prostate-cancer screening trial.
      N Engl J Med. 2009; 360: 1310-1319
      Subjects were 76,693 men at 10 US centers, randomized either to annual PSA screening for 6 years or to “usual care.” At 13 years follow-up, the difference between the 2 groups was not significant.
      • Andriole G.L.
      • Crawford E.D.
      • Grubb R.L.
      • et al.
      Prostate cancer screening in the randomized Prostate, Lung, Colorectal, And Ovarian Cancer Screening Trial: mortality results after 13 years of follow-up.
      J Natl Cancer Inst. 2012; 104: 1-8
      However, the control group (“usual care”) in this study was highly contaminated, with more than 50% of the subjects having had PSA screening, making the lack of findings difficult to interpret.
      • Pinsky P.F.
      • Blacka A.
      • Kramer B.S.
      • Miller A.
      • Prorok P.C.
      • Berg C.
      Assessing contamination and compliance in the prostate component of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial.
      Clin Trials. 2010; 7: 303-311
      , 
      • Shoag J.E.
      • Mittal S.
      • Hu J.C.
      Reevaluating PSA testing rates in the PLCO Trial.
      N Engl J Med. 2016; 374: 1795-1796
      The second was the European Randomized Study of Screening for Prostate Cancer (ERSPC) that published their 11- and 13-year follow-ups in 2012,
      • Schröder F.H.
      • Hugosson J.
      • Roobol M.J.
      • et al.
      Prostate-cancer mortality at 11 years of follow-up.
      N Engl J Med. 2012; 366: 981-990
      and 2014.
      • Schröder F.H.
      • Hugosson J.
      • Roobol M.J.
      • et al.
      The European randomized study of screening for prostate cancer—prostate cancer mortality at 13 years of follow-up.
      Lancet. 2014; 384: 2027-2035
      Subjects were 182,162 men aged 50-74, from 7 European countries, randomized either to PSA screening every 4 years or to control. They found a 21% reduction in prostate cancer deaths in the screened group as compared to the controls. However; using a 2-year screening interval rather than the 4-year interval, the Göteborg subset of the ERSPC indicated a 44% reduction in prostate cancer deaths.
      • Hugosson J.
      • Carlsson S.
      • Aus G.
      • et al.
      Mortality results from the Göteborg randomised population-based prostate-cancer screening trial.
      Lancet Oncol. 2010; 11: 725-732
      This suggests that the choice of a shorter screening interval may have led to more powerful results.
      The present study re-examines the value of prostate cancer screening, and seeks to determine both the optimal screening interval, and the appropriate age groups to be screened.

      Methods

      Project Design

      This study uses a retrospective cohort design. Although the nonrandomized nature of the design potentially produces complications in the data, various statistical methods were used to adjust for these complications. The benefit of the design is the ability to take advantage of the enormous databanks from large health plans like Kaiser Permanente, which care for large populations of patients over many years.

      Subjects and Eligibility Criteria

      Subjects consisted of all 400,887 men under 80 years of age, with no history of prostate cancer, who had PSA testing done at Kaiser Permanente Northern California in calendar years 1998-2002. Subject data were derived from the Kaiser Permanente Northern California electronic database, which includes information for all patients dating back to 1992. It contains demographic data, laboratory results, pathology results, and cancer treatment information. Mortality data were obtained from the Kaiser Permanente Division of Research mortality database that identifies death from in-system and out-of-system sources; the latter including Social Security death master file data and California State Department of Vital Statistics data. Data were collected on subjects through calendar year 2013.
      The 5-year study period 1998-2002 was selected because there were prior PSA data going back to 1992, allowing for 6-11 years of prior data to examine screening intervals. This study period also provided 12-16 years of follow-up on patients diagnosed with prostate cancer, an amount of time deemed reasonable to calculate mortality rates. In addition, more than 400,000 subjects under age 80 had PSA testing in this period, providing a large study population.

      Design for Data Analysis

      The 400,887 subjects were sorted into 42 subgroups: 6 groups based on PSA intervals (12-18 months, 18-24 months, 2-3 years, 3-4 years, 4-9 years, or no prior PSA) and 7 age groups (<50, 50-54, 55-59, 60-64, 65-69, 70-74, and 75-79). For each subgroup, a tally was made of the number of men tested for PSA, the number diagnosed with prostate cancer, the number who died of prostate cancer; and the number with prostate cancer who died of other causes. Differences between the groups were examined, and chi-square tests were used to assess significance. An alpha level of 0.01 was used for all statistical tests. Megastat was the statistical software utilized.
      The number of men needed to be screened (NNS) to prevent 1 death from prostate cancer was calculated as the inverse of the absolute risk reduction. The number of men needed to be diagnosed (NND) to prevent 1 death from prostate cancer is NNS times the prostate cancer incidence in the screened group.
      In addition to examining risk reduction for prostate cancer deaths, risk reduction was examined for all-cause mortality. As an inadvertent result of the retrospective design, subjects screened at yearly intervals had fewer nonprostate cancer deaths than unscreened subjects, suggesting a healthier population. In an attempt to compensate for this difference and equalize nonprostate cancer death rates for the 2 groups, the number of excess deaths from nonprostate causes in the unscreened group was subtracted from the all-cause deaths in that group. Although this mathematical adjustment excessively reduces the estimated difference in all-cause mortality between the screened and unscreened groups, it avoids the risk of overestimating the screening benefit. This is because healthier subjects who have not died of other causes are alive and, thus, still at risk for dying of prostate cancer. For the same reason, it should be noted that this health disparity also has the potential effect of underestimating the magnitude of the observed differences between screened and unscreened groups for prostate cancer-specific mortality.

      Definition of Terms

      Screening PSA: a PSA to test for prostate cancer in the absence of signs or symptoms.
      PSA for cause: a PSA to test for prostate cancer when there are signs or symptoms of possible disease.
      Marker PSA: the PSA test done during the 1998 through 2002 period. If there were more than 1 PSA in this period, and one of them led to a cancer diagnosis, that 1 was designated the Marker PSA. If there were multiple tests and no cancer diagnosis, the PSA in this period that provided the shortest PSA interval was chosen as the Marker PSA.
      Prior PSA: the most recent PSA done 12 or more months before the Marker PSA. These were all treated as screening PSAs, and subjects were assigned to groups based on the PSA Interval (defined in the following).
      PSA Interval: the time interval between the Marker PSA and the Prior PSA.
      In text and tables, notation was simplified in reference to intervals.
      12-18 months = 12 to <18 months
      18-24 months = 18 to <24 months
      2-3 years = 2 to <3 years
      3-4 years = 3 to <4 years
      4-9 years = 4 to <9 years
      No Prior PSA Group: subjects who had no PSA done before the Marker PSA. This was considered to be the unscreened group.

      Example

      A 62-year-old male had a screening PSA in August 1998 which had not led to a diagnosis of prostate cancer. In October 1999, he had another screening PSA = 4.3. A biopsy in December 1999 showed prostate cancer, Gleason 3 + 3.
      Marker PSA: October 1999. This is the date of PSA leading to biopsy and diagnosis.
      Prior PSA: August 1998. This is the most recent PSA >12 months before diagnosis.
      PSA Interval: 14 months.

      Determining Screening PSA vs PSA for Cause

      A problem with the Kaiser database is the inability to distinguish between PSA tests done for screening vs those obtained for signs and symptoms. Although the problem potentially affects the results in 3 ways, it is possible to calculate the direction of the effect and to ascertain, as discussed in the following, that in each situation, the effect is to underestimate, rather than inflate, the differences between the PSA screened vs unscreened subjects.
      Situation 1: Prior PSA
      It was assumed that all of these had been done for screening, as they all had a screening interval of at least 12 months, suggesting benign findings at the time of the prior PSA.
      Errors will be PSAs for cause mislabeled as “screening.”
      Effect: these PSAs for cause will increase the mortality statistic of the screened group, and decrease the potential mortality difference between screened and unscreened groups. The potential effect would be to understate the magnitude of the observed differences between screened and unscreened groups.
      Situations 2 and 3: Marker PSAs
      Screened groups: includes all subjects who had a prior PSA. For subjects in these groups, it was assumed that all PSAs were for screening.
      Errors would be any PSAs done “for cause” that were mislabeled as for “screening.”
      Effect: overestimate the mortality statistic of the screened group.
      No Prior group: for subjects in this group, it was assumed all PSAs were “for cause.”
      Errors would be any PSAs done for “screening” and mislabeled as “for cause.”
      Effect: underestimate the mortality statistic for No Prior (unscreened) group.
      Net Effect: in each of the above 2 situations, the errors would potentially decrease the calculated difference in mortality between the screened and unscreened groups.

      Results

      Treatment Modalities

      Of the 400,887 men studied, 8542 had a biopsy-proven prostate cancer diagnosis during the 5-year period selected for study. Table 1 shows the initial treatment modality for the diagnosed subjects. During the 12-16 years of follow-up, 770 died of prostate cancer, 2512 died of other causes, and 5260 patients remained alive.
      Table 1Initial treatment modality for cancer patients
      Primary TreatmentNumber of SubjectsDied of CaP (n)Percent Died of CaP (%)Median PSA at DiagnosisMean Gleason at DiagnosisMean Age at Diagnosis
      Surgery22691215.3%6.26.362.7
      Radiation37452576.9%7.56.467.1
      Palliation only114928224.5%15.46.969.3
      No treatment13791108.0%6.96.167.2
      Total group85427709.0%7.46.466.3
      CaP, prostate cancer.

      Assessment of Data Completeness

      Of the 8542 subjects diagnosed with prostate cancer, 76.4% were Kaiser Permanente members at their death or at 12-16 years of follow-up. This 76.4% included 7.7% dead of prostate cancer, 23.5% dead of other causes, and 45.3% alive at 16 years. Another 8.5% of subjects were reported dead after leaving Kaiser, resulting in complete outcome data for 85% of subjects. The remaining 15% were no longer Kaiser members at 16 years of follow-up, had not been reported dead in any databases, and were assumed to be alive. Given the possibility of unreported deaths in this last group of 15%, it is important to assess its potential impact. If deaths in the non-Kaiser group were to occur in the same proportion as in subjects who remained Kaiser members, 9.6% (rather than 8.5%) of nonmember subjects would be expected to be dead, and 14% (rather than 15%) would be expected to be alive at 16 years of follow-up. This suggests that approximately 1% of our data might be missing due to unreported deaths for subjects dying after leaving Kaiser.

      Effect of Screening by Age Groups

      Table 2 shows all of the data stratified by PSA test interval and age group. For ease of comparison, Table 3 details the fourth column of Table 2, and shows the prostate cancer mortality rates per 100,000 subjects tested by PSA interval and by age group. The last column shows combined data for subjects aged 55-74 years. Of note is the finding that of the 400,887 men tested during this period, 154,125 or 38% were 54 or younger, equally divided between those under 50, and those 50-54, and there is no demonstrable benefit to screening men less than 55 years of age.
      Table 2Data for all subjects by PSA test interval and by age group
      Interval to Prior PSASubjects With PSA Testing Done in 1998-2002Subjects Dying of CaPCaP Deaths per 100,000 SubjectsSubjects Diagnosed with CaPCaPs per 100,000 Subjects
      Subjects aged <50
       12-18 mo53310023421
       18-24 mo3640009247
       2-3 y31651328253
       3-4 y1881008425
       4-9 y18851537371
       No Prior61,2551423112183
       Total77,1571621167219
      Subjects aged 50-54
       12-18 mo13,56753794688
       18-24 mo774833850641
       2-3 y642523152803
       3-4 y338912926761
       4-9 y34580022632
       No Prior41,8512559286679
       Total76,9683647530689
      Subjects aged 55-59
       12-18 mo18,2938442291252
       18-24 mo87627801501712
       2-3 y65363461241897
       3-4 y3574384591651
       4-9 y4815483931931
       No Prior25,296411624621826
       Total67,276669811171660
      Subjects aged 60-64
       12-18 mo20,446231124342123
       18-24 mo7825121792543246
       2-3 y5413101851923547
       3-4 y27924143993546
       4-9 y393182041303307
       No Prior17,068583405733357
       Total57,47511720416822926
      Subjects aged 65-69
       12-18 mo20,960281346663177
       18-24 mo6856142043475061
       2-3 y435592072144914
       3-4 y2177115051205512
       4-9 y3289133951444378
       No Prior13,375856366464830
       Total51,01216031421374189
      Subjects aged 70-74
       12-18 mo18,105542986223436
       18-24 mo5173183482835471
       2-3 y3181216601946099
       3-4 y16991810591126592
       4-9 y2829176011204242
       No Prior8738788934665333
       Total39,72520651917974524
      Subjects aged 75-79
       12-18 mo13,698513724002920
       18-24 mo4022153731634053
       2-3 y2602166151104228
       3-4 y14528551573926
       4-9 y2603269991114264
       No Prior6897537682713929
       Total31,27416954011123556
      Total of all subjects400,88777019285422131
      PSA, prostate-specific antigen.
      Table 3Prostate cancer deaths rates per 100,000 subjects tested, stratified by interval to prior PSA and by age group
      Death Rates per 100,000 Subjects Tested
      Age Groups<5050-5455-5960-6465-6970-7475-7955-74
      Interval to prior PSA
       12-18 mo03744112134298372145
       18-24 mo03880179204348373185
       2-3 y323146185207660615221
       3-4 y029841435051059551351
       4-9 y53083204395601999283
       No Prior2359162340636893768406
       Total214798204314519540255
      Comparison of 12-18 mo Interval vs. No Prior
       12-18 mo03744112134298372145
       No Prior2359162340636893768406
       % Change−100%−38%−73%
      Reached statistical significance of P ≤.01.
      		−67%
      Reached statistical significance of P ≤.01.
      		−79%
      Reached statistical significance of P ≤.01.
      		−67%
      Reached statistical significance of P ≤.01.
      		−52%
      Reached statistical significance of P ≤.01.
      		−64%
      Reached statistical significance of P ≤.01.
       P value.27.32<.001<.001<.001<.001<.001<.001
      Comparison of 18-24 mo Interval vs. No Prior
       18-24 mo03880179204348373185
       No Prior2359162340636893768406
       % Change−100%−35%−51%−47%−68%
      Reached statistical significance of P ≤.01.
      		−61%
      Reached statistical significance of P ≤.01.
      		−51%−54%
      Reached statistical significance of P ≤.01.
       P value.36.47.08.03<.001<.001.11<.001
      Comparison of 2-3 year Interval vs. No Prior
       2-3 y323146185207660615221
       No Prior2359162340636893768406
       % Change0%−48%−72%−46%−67%
      Reached statistical significance of P ≤.01.
      		−26%−20%−46%
       P value.40.37.26.07<.001.22.43.35
      Comparison of 3-4 year Interval vs. No Prior
       3-4 y029841435051059551351
       No Prior2359162340636893768406
       % Change−100%−51%−48%−58%−20%0%−28%−14%
       P value.51.48.26.08.47.51.38.09
      Comparison of 4-9 year Interval vs. No Prior
       4-9 y53083204395601999283
       No Prior2359162340636893768406
       % Change0%−100%−49%−40%−38%−33%0%−30%
       P value.40.15.19.17.47.51.27.30
      * Reached statistical significance of P ≤.01.
      For the <50-year-old group, χ2 (5, N = 77,157) = 3.51, P = .62.
      For the 50- to 54-year-old group, χ2 (5, N = 76,968) = 4.04, P = .54.
      For the 5 groups of 55 and over, the rate ratio (RR) of prostate cancer mortality between 12-18 months screening and no prior screening was highly significant by chi-square. The 95% confidence intervals (CIs) are also shown.
      55-59 years old, RR 0.27 (95% CI −0.09 to 0.63, P <.001), or a 73% relative risk reduction.
      60-64 years old, RR 0.33 (95% CI 0.04-0.62, P <.001), or a 67% relative risk reduction.
      65-69 years old, RR 0.21 (95% CI −0.02 to 0.44, P <.001), or a 79% relative risk reduction.
      70-74 years old, RR 0.33 (95% CI 0.10-0.57, P <.001), or a 67% relative risk reduction.
      75-79 years old, RR 0.48 (95% CI 0.19-0.78, P <.001), or a 52% relative risk reduction.
      And for the combined groups:
      55-74 years old, RR 0.36 (95% CI 0.22-0.50, P <.001), or a 64% relative risk reduction.

      Number Needed to be Screened or Diagnosed

      The NNS to prevent 1 death from prostate cancer over an average of 14 years of follow-up is easily calculated from data in Table 3. For the 65- to 69-year-old group, there are 636 deaths per 100,000 screened in the No Prior group, and 134 deaths per 100,000 screened at 12-18 months. The absolute reduction is 502 deaths (636 − 134) per 100,000 screened. To save 1 life, (100,000 screened)/(502 deaths prevented) = 199 men screened for each life saved (95% CI 155-228, P <.001).
      The NND to prevent 1 death from prostate cancer is similarly calculated from data in Table 2, Table 3. In the example earlier, 3177 cancers will be detected among the 100,000 men, aged 65-69, screened at 12-18 months intervals, while preventing 502 deaths. NND = (3177 cancers diagnosed)/(502 lives saved) = 6.3 cancers diagnosed for each life saved (95% CI 4.9-8.8, P <.001).

      Effect of Screening Intervals on Prostate Cancer Mortality

      To determine the most effective screening interval, each of the intervals was compared to no prior testing. The lower portion of Table 3 shows that the efficacy of screening drops off rapidly for intervals greater than 12-18 months. In addition, it can be seen that statistically significant values (highlighted by asterisks) only occur for subjects aged 55-79 for the 12-18 group, for subjects aged 65-74 for the 18-24 months interval group, and for subjects aged 65-69, only for the 2-3 years interval group. For the combined groups aged 55-74, there was a 64% reduction in cancer mortality with 12-18 months screening. This dropped to 54% with 18-24 months screening, and to 46% with 2-3 years screening, although this latter value did not achieve statistical significance.

      Effect of Screening on All-cause Mortality

      Risk reduction for all-cause mortality was examined in subjects aged 55-74, comparing 12-18 months screening and no prior screening. As discussed in the methods section, the 12-18 months group appeared much healthier, with 671 nonprostate cancer deaths per 100,000 men screened, compared to 1008 deaths in the No Prior group. If there were 217 fewer nonprostate cancer deaths in the No Prior group, the 2 groups would have equal nonprostate cancer death rates (see Table 4). If 217 deaths are subtracted from the all-cause deaths in the No Prior group, it is possible to get a corrected estimate of the all-cause death rate reduction. As seen in Table 4, there is a 24% reduction in all-cause deaths at 12-16 years of follow-up—from 1078 per 100,000 screened for the No Prior group to 816 for the 12-18 months group (95% CI 15-34%, P <.001).
      Table 4Data for subjects aged 55-74 by PSA test intervals 12-18 months and No Prior Corrected for Health differences to calculate all-cause mortality change
      Interval to Prior PSASubjects with PSA TestingSubjects Dying of CaPCaP Deaths per 100,000 SubjectsSubjects With CaP Dying of All CausesAll-Cause Deaths per 100,000 Subjects TestedSubjects With CaP Dying of Other CausesNon-CaP Deaths per 100,000 Subjects Tested
      12-18 mo77,804113145635816522671
      No Prior64,47726240691214146501008
      No Prior Corrected for Health64,477262406(912 − 217) = 6951078(650 − 217) = 433671
      % Change−64%−24%

      Discussion

      These data strongly support the hypothesis that PSA screening is beneficial, reducing prostate cancer deaths by 64% and all-cause mortality by 24% for men aged 55-74 years. They also strongly support the second hypothesis that the screening interval is critical and show that yearly screening is the interval of choice.
      The third hypothesis was that age would be a critical variable in determining the value of PSA screening. The data show a maximal benefit to men aged 55-74, with no benefit of screening for men under age 55.
      The magnitude of these results is compatible with other independent findings. The Göteborg subset of the ERSPC indicated a 44% reduction in prostate cancer deaths with 2-year screening.
      • Hugosson J.
      • Carlsson S.
      • Aus G.
      • et al.
      Mortality results from the Göteborg randomised population-based prostate-cancer screening trial.
      Lancet Oncol. 2010; 11: 725-732
      This is consistent with our data for 18-24 months and 2-3 years screening, which showed decreases of 54% and 46%, respectively in prostate cancer mortality when compared to no prior screening for men aged 55-74 (see Table 3). Data from the National Cancer Institute (NCI) showed that from 1993 to 2014, there was a 51% decrease in the prostate cancer death rate in the United States (from 39.3 to 19.1 deaths per 100,000).
      • Siegel R.
      • Miller K.
      • Jemal A.
      Cancer statistics, 2017.
      CA Cancer J Clin. 2017; 67: 7-30
      This is the same period in which widespread PSA screening was done. Although this does not prove a causal relationship, it is highly suggestive, and the improvements in radiation therapy and chemotherapy in this time period cannot account for the magnitude of this change. Since considerably fewer than 100% of eligible American men were regularly screened, to achieve a 51% reduction in mortality rates, the efficacy of PSA screening would have to be significantly greater than 51%, which is compatible with the findings of the present study.
      A concern that both the NCI and the ERSPC data raise is the increased number of clinically insignificant prostate cancers discovered with screening, an increase of approximately 30% in the NCI data.
      • Siegel R.
      • Miller K.
      • Jemal A.
      Cancer statistics, 2017.
      CA Cancer J Clin. 2017; 67: 7-30
      The unnecessary treatment of this group prompted the USPSTF to recommend against PSA screening in 2012. However, the overtreatment problem can be solved by utilizing active surveillance for appropriate low-grade, low-stage cancers. It is estimated that 36% of patients with a new diagnosis of prostate cancer are candidates for active surveillance, and that 30% of patients electing active surveillance will ultimately have treatment for progression.
      • Dall'Era M.A.
      • Albertsen P.C.
      • Bangma C.
      • et al.
      Active surveillance for prostate cancer: a systematic review of the literature.
      Eur Urol. 2012; 62: 976-983
      If the 36% of patients eligible for active surveillance elected this, and 70% never require treatment, then 25% of patients with a diagnosis of prostate cancer will never require treatment. This compensates for the approximately 30% increase in the rate of prostate cancer diagnoses and mitigates against the USPSTF's argument that screening causes too much unnecessary treatment and attendant morbidity.

      Conclusion

      These data show that yearly PSA screening is highly beneficial, reducing prostate cancer deaths by 64% for men aged 55-74 years, and reducing all-cause mortality in this group by 24%. Yearly screening is the interval of choice. No benefit was found for men under age 55. When combined with active surveillance to prevent overtreatment, these data lend support for yearly population-based PSA screening for prostate cancer for men aged 55-74 who are in good health.

      Appendix. Supplementary Data

      The following is the supplementary data to this article:

      References

        • Siegel R.
        • Miller K.
        • Jemal A.
        Cancer statistics, 2017.
        CA Cancer J Clin. 2017; 67: 7-30
        View in Article
        • Moyer V.A.
        Screening for prostate cancer: U.S. preventive services task force recommendation statement.
        Ann Intern Med. 2012; 157: 120-134
        View in Article
        • Andriole G.L.
        • Crawford E.D.
        • Grubb R.L.
        • et al.
        Mortality results from a randomized prostate-cancer screening trial.
        N Engl J Med. 2009; 360: 1310-1319
        View in Article
        • Andriole G.L.
        • Crawford E.D.
        • Grubb R.L.
        • et al.
        Prostate cancer screening in the randomized Prostate, Lung, Colorectal, And Ovarian Cancer Screening Trial: mortality results after 13 years of follow-up.
        J Natl Cancer Inst. 2012; 104: 1-8
        View in Article
        • Pinsky P.F.
        • Blacka A.
        • Kramer B.S.
        • Miller A.
        • Prorok P.C.
        • Berg C.
        Assessing contamination and compliance in the prostate component of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial.
        Clin Trials. 2010; 7: 303-311
        View in Article
        • Shoag J.E.
        • Mittal S.
        • Hu J.C.
        Reevaluating PSA testing rates in the PLCO Trial.
        N Engl J Med. 2016; 374: 1795-1796
        View in Article
        • Schröder F.H.
        • Hugosson J.
        • Roobol M.J.
        • et al.
        Prostate-cancer mortality at 11 years of follow-up.
        N Engl J Med. 2012; 366: 981-990
        View in Article
        • Schröder F.H.
        • Hugosson J.
        • Roobol M.J.
        • et al.
        The European randomized study of screening for prostate cancer—prostate cancer mortality at 13 years of follow-up.
        Lancet. 2014; 384: 2027-2035
        View in Article
        • Hugosson J.
        • Carlsson S.
        • Aus G.
        • et al.
        Mortality results from the Göteborg randomised population-based prostate-cancer screening trial.
        Lancet Oncol. 2010; 11: 725-732
        View in Article
        • Dall'Era M.A.
        • Albertsen P.C.
        • Bangma C.
        • et al.
        Active surveillance for prostate cancer: a systematic review of the literature.
        Eur Urol. 2012; 62: 976-983
        View in Article

      Article info

      Publication history

      Published online: May 14, 2018
      Accepted: February 9, 2018
      Received: October 4, 2017

      Footnotes

      Financial Disclosure: The author has no competing interests.

      Funding Support: Kaiser Permanente Northern California provided all the raw data for this study through the Kaiser Foundation Research Institute. There was no external financial support.

      Paul Alpert is the sole author of this manuscript, and the corresponding author.

      This study was approved by the Kaiser Permanente Northern California Institutional Review Board, which waived the requirement for informed consent since there was no patient contact, and no individual data was released.

      Identification

      DOI: https://doi.org/10.1016/j.urology.2018.02.049

      Copyright

      © 2018 The Author. Published by Elsevier Inc.

      User license

      Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) |
      How you can reuse Information Icon
      Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)

      Permitted

      For non-commercial purposes:

      • Read, print & download
      • Redistribute or republish the final article
      • Text & data mine
      • Translate the article (private use only, not for distribution)
      • Reuse portions or extracts from the article in other works

      Not Permitted

      • Sell or re-use for commercial purposes
      • Distribute translations or adaptations of the article

      Elsevier's open access license policy

      ScienceDirect

      Access this article on ScienceDirect
      We use cookies to help provide and enhance our service and tailor content. To update your cookie settings, please visit the for this site.
      Copyright © 2023 Elsevier Inc. except certain content provided by third parties. The content on this site is intended for healthcare professionals.


      RELX