How does the use of specific reference ranges as opposed to a single reference range make the PSA assay a better test for males of all ages?

Angela M. Davis
October 4. 1996

Case Scenario:

A 44 yo Black Male presents as a new patient to the Medicine clinic with no significant past medical history. ROS is basically negative. Family hx reveals father with prostate cancer.

Clinical Bottom lines:

1. 1993 incidence of prostate cancer 160,000 cases.

2. 317,000 new cases and 41,000 deaths predicted for 1996.

3. Most common cancer and second leading cause of death among males in the U.S.
Annual Risk of dying is 1/1000 and lifetime risk is 3/100.

4. 10 year survival 75% when confined to prostate . 55% with regional extension, and 15% with metastasis.

5. Autopsy revealed 30% of males > 50 have prostate cancer.

6.  Possible risk factors: family hx. race( highest among Scandinavian and African
American males), socioeconomic, cadmium exposure, zinc deficiency, tobacco use,
h/o vasectomy.

7. 5 year survival of African American with all stages of prostate cancer is 62%
compared to 72% in White males. African American males felt to have larger volume
of latent prostate cancer.

The Evidence

Sensitivity and Specificity of the Serum PSA Test According to Race, Age Group, and Serum PSA Level in the Entire Study Population

PSA Level
(ng/mil)		 AGE
  40-49 yr 50-59 yr 60-69 yr 70-79 yr
  Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity
Blacks
1.0 100.0 74.3 100.0 45.7 100.0 32.6 100.0 20.6
2.0 94.3 93.2 100.0 72.6 100.0 38.6 100.0 48.3
3.0 81.4 96.6 98.2 81.9 98.7 71.5 100.0 60.4
4.0 48.6 98.6 94.7 88.4 98.1 79.0 98.6 73.1
5.0 40.0 98.6 80.7 92.0 93.6 82.5 95.8 75.5
6.0 26.2 98.6 54.4 94.7 92.9 86.6 94.4 79.4
7.0 17.1 99.3 35.1 96.2 69.9 89.4 90.1 83.4
8.0 8.7 99.3 21.9 97.2 57.5 91.4 90.1 88.1
9.0 7.1 99.3 13.2 97.5 34.0 92.7 67.6 90.2
10.0 3.7 99.3 8.8 97.7 19.9 93.7 53.5 92.6
11.0 2.9 99.3 6.1 98.2 16.0 94.5 46.5 93.7
12.0 0.0 99.7 2.5 98.5 9.6 95.5 29.6 94.7
13.0 0.0 99.7 0.9 98.7 7.7 96.0 22.5 95.5
14.0 0.0 99.7 0.9 99.0 7.1 96.6 16.9 96.6
15.0 0.0 99.8 0.9 99.1 4.7 97.2 10.6 97.4

 

PSA Level
(ng/mil)		 AGE
  40-49 yr 50-59 yr 60-69 yr 70-79 yr
  Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity Sensitivity Specificity
Whites
1.0 99.3 76.0 99.7 52.8 99.7 86.1 99.2 31.3
2.0 98.4 94.6 99.4 83.4 98.4 66.0 98.0 55.4
3.0 84.7 100.0 97.8 93.2 96.5 31.2 97.4 70.0
4.0 52.0 100.0 74.8 97.4 91.1 93.0 92.1 81.2
5.0 38.7 100.0 46.6 98.7 71.0 97.1 90.0 90.8
6.0 25.4 100.0 26.1 99.1 52.2 98.8 88.7 96.3
7.0 20.7 100.0 11.8 99.1 28.0 100.0 65.2 98.8
8.0 16.4 100.0 9.3 100.0 11.3 100.0 46.5 99.8
9.0 12.5 100.0 7.8 100.0 6.7 100.0 31.2 99.9
10.0 10.8 100.0 7.5 100.0 3.1 100.0 22.8 100.0
11.0 9.3 100.0 5.3 100.0 2.5 100.0 15.9 100.0
12.0 8.4 100.0 4.3 100.0 1.3 100.0 11.5 100.0
13.0 7.3 100.0 4.0 100.0 1.1 100.0 7.7 100.0
14.0 6.3 100.0 3.1 100.0 0.8 100.0 7.2 100.0
15.0 5.2 100.0 2.5 100.0 0.6 100.0 6.6 100.0

Age-Specific Reference Ranges for the PSA Test, Based on the 5th Percentile of the Distribution of PSA Levels in the Patients, According to Race

AGE (YR) WHITES BLACKS
  ng of PSA/ml
40-49 0.0-2.5 0.0-2.0
50-59 0.0-3.5 0.0-4.0
60-69 0.0-3.5 0.0-4.5
70-79 0.0-3.5 0.0-5.5


Comments:

  1. Subjects were all from Walter Reed Army Hospital with random selection of white males via computer selection, not Black males. 162 subjects were excluded for questionable reasons.
  2. Patient sample excluded those with prostate cancer and prostatitis, but included those with prostatism.
  3. Study was compared to 1993 Mayo Clinic study of 2119 White males and a 1995 Univ. of Michigan/Sapporo Medical College study of 335 Japanese males. Both studies included males from 40-79 yo.
  4. All important options and outcomes such as DRE alone, Transrectal ultrasound and biopsy were not considered.
  5. An explicit and sensible process was not used to consider the relative value of different outcomes. all patients did not have biopsies. All patients were not accounted for. No consideration for increased PSA secondary to recent ejaculation, acute urinary retention. Studies suggest waiting 6 weeks to check PSA after prostatitis, massage biopsy, or TURP.
  6. This study was somewhat ineffective because many selection biases and one can not clearly interpret the calculated sensitivity and specificity of the reference ranges. A long-term prospective study with less selection bias is necessary.

References:

  1. Morgan, Ted, Jacobsen, S.J., et al. Age Specific Reference Ranges for Serum Prostate Specific Antigen. New England Jour of Medicine 1996. 335: 304-310.
  2. Woolf, S.H., Screening for Prostate Cancer with PSA. New England Jour of Medicine 1995. 333:1401-1405.
  3. Kramer, Barnett, Brown, Martin, et al. Prostate Cancer Screening: What We Know and What We Need to Know. Annals of Internal Medicine 1993. 119:914-923.
  4. Prorok, Phillip, Connor, Robert. Statistical Considerations in Cancer Screening Programs. Urologic Clinics of North America 1990, 17:699-708.