• 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • br and the Malmo Preventive Project


    [2] and the Malmo Preventive Project [15].]FID$T1[Our study differs from these two studies with regards to not only race but also other key factors, including length of follow-up (one decade vs 2+ decades), outcome definition (aggressive vs metastat-ic/fatal disease), time period (pre-PSA vs post-PSA screening eras), and socioeconomic position/access to health care. To facilitate comparison, we have included unpublished results in Table 4 for “aggressive” PCa in PHS using the same definition and follow-up period as in the present analysis. Within PHS, the difference between the proportion of aggressive cases captured over 10 yr and that of lethal cases captured over 30 yr is not large; the proportions remain slightly lower than those in SCCS. This suggests that the higher sensitivity of baseline PSA observed in SCCS is more likely due to differences in race, time period, and characteristics of the study population rather than differ-ences in the outcome definition and length of follow-up. Based on this comparison, it appears that baseline PSA captures aggressive PCa at least as well in black men as in white men.
    In addition, it is noteworthy that baseline PSA levels among controls in our study were similar to those among white controls from other prospective nested case-control studies of baseline PSA levels in midlife (Tables 4 and 5)
    [2,15,16]. Only one prior study [11] reported results for black controls while also finding similar PSA distributions among black and white controls.
    The similarity in distributions of baseline PSA levels between black and white controls in nested case-control studies is in Imipenem to that in many cross-sectional studies, including an earlier study in SCCS participants [26], which found higher PSA levels in black men [27–33], particularly among older age groups. These studies measured PSA in men without clinical evidence of PCa at time of blood draw but did not assess eventual case/ control status over time. Thus, they include men with undiagnosed, prevalent PCa. National cancer statistics [34] and autopsy studies [6] show that black men have higher rates of disease at every age; therefore, cross-sectional studies would be expected to find higher PSA levels among black men due to a higher prevalence of latent disease. Nested case-control studies of PSA are less likely to include men with prevalent disease, as there is follow-up time after baseline during which prevalent disease might be diagnosed. The observations of similar midlife PSA between black and white men in this and other nested case-control studies [11,12,15,16] suggest that a risk-stratified screening strategy of PSA measured in midlife might use similar cut points for black and white men.
    While this is the largest prospective study exploring prediction of PCa by baseline PSA in a population of black men, our study is subject to limitations. Importantly, there exists an issue of “verification bias” due to presence of opportunistic PSA screening during the study period. However, results of PSA measurements made for this
    Table 4 – Comparison between studies of baseline prostate-specific antigen distribution and proportion of prostate cancer cases captured by percentile groups
    Southern Community
    Physicians’ Health Study [2]
    Malmo Preventative
    Cohort Study
    Age-specific PSA level a PSA level Lethal b prostate c prostate cancer,
    PSA level b
    Aggressive prostate
    Lethal ]FID$T9[prostate
    PSA]FID$T1[ = prostate specific antigen.
    a Aggressive prostate cancer:[7T$DIF]Gleason grade 7 and above, or AJCC Stage III or IV, or prostate cancer death.
    b Lethal prostate cancer: metastatic cancer at diagnosis or during follow-up, or prostate cancer death.
    c For the Physicians’ Health Study, results for aggressive cancer are previously unpublished and presented here for comparison with Southern Community Cohort Study results. These results are based on 10 yr of follow-up and aggressive prostate cancer as defined in note 1.
    Table 5 – Distribution of total PSA (ng/ml) by age group and race among controls from case-control studies of prostate cancera
    Total PSA, ng/ml
    Age group Race n Study 25th 50th 75th 90th Reference
    population percentile percentile percentile percentile
    Current study
    Current study
    Current study
    CHDS = Child Health and Development Study; PHS = Physicians’ Health Study; PSA = prostate-specific antigen; SCCS = Southern Community Cohort Study; VIP = Västerbotten Intervention Project. a PSA levels by race among controls from all nested case-control studies of baseline PSA that reported PSA levels by age group.