The Prostate 31:139–141 (1997) EDITORIAL Prostate Cancer In African-American Men The recent New England Journal of Medicine article by Morgan et al.  raises several questions about the need for improved prostate cancer (PCa) control in African American men (AAM). AAM have up to two times the incidence and three times the death rate from prostate cancer of Caucasian men (CM) . PCa in AAM also appears to occur at younger ages. AAM have thus been designated as a distinct high-risk group and advised to have their annual cancer related check-up beginning at age 40, rather than age 50 as for the general population . From radical prostatectomy series, AAM have also been noted to have higher mean tumor volumes with higher PSA levels . The recent article by Morgan et al. focused primarily on the statistical manipulation of PSA values in an attempt to improve PSA diagnosis through age-specific reference ranges for AAM without sufficient consideration of the ultimate endpoint in early detection, namely, the reduction of disproportionate mortality. Age-specific reference ranges were initially developed by Oesterling et al.  to account for increasing average PSA by age groups, predominantly related to greater probability of gland enlargement from benign prostatic hyperplasia (BPH) and resultant greater PSA false positives. The purported benefit of age-specific reference ranges would be to allow greater detection sensitivity for younger men while increasing the specificity of PSA for older men. The initial agespecific PSA reference ranges  (Table I) were thus derived from only CM. Participation of AAM in early-detection projects has always been disproportionately low, with the greatest accrual rate of 7% attained in the American Cancer Society-National Prostate Cancer Detection Project (ACSNPCDP) . Without consideration of the significantly greater PCa incidence among AAM, focusing only on PSA sensitivity and specificity  may negate the greater mortality impact of cancers that might also be more aggressive. AAM have been noted to have 1.5 times the mortality rate when controlling for stage  in the Surveillance Epidemiology and End Results (SEER) cancer registry from the metropolitan Detroit area. A major difficulty in constructing age-specific reference ranges for AAM can be seen in the entire shift of focus by Morgan et al.  to produce 95% sensitivity vs. the original goal of 95% specificity in Oester© 1997 Wiley-Liss, Inc. ling’s original derivation of age-specific reference ranges  (Table I). Herein lies the statistical difference between PSA levels of AAM and CM when using the 95th percentile (i.e., mean PSA + 2 standard deviations) for patients without evidence of cancer. AAM and CM had nearly identical median PSA levels when controlling for age, but AAM had greater PSA variability with increasing age (i.e., larger standard deviations), producing markedly higher 95th percentiles for normal screening patients . These significantly higher age-specific reference ranges (95th percentiles) for AAM resulted in 41% of prostate cancer cases in AAM being missed vs. only 15–20% of cancers missed among CM. Therefore, the authors chose to emphasize the 95th percentile for cancer detection from their clinical cohort and obtained age-specific reference ranges nearly identical to the original values obtained by Oesterling  for CM with benign prostates. Interestingly, they did not emphasize that the same concept of 95th percentile cancer detection obliterates age-specific PSA ranges for CM over age 50 (Table I). Again, the inversion of emphasis from high specificity to high sensitivity will doubtless confuse many clinicians. Perhaps few will grasp the marked differences in early detection when age-specific reference ranges for AAM emphasize 95th percentile sensitivity and have nearly identical values for CM emphasizing 95 percentile specificity. Should we therefore call it age-sensitive reference ranges for AAM? TABLE I. Age-Specific Reference Ranges: Specificityand Sensitivity-Driven Differences* Specificity driven Sensitivity (95th percentile) Age range (years) CM— Oesterling et al.  AAM— Morgan et al.  CM— Oesterling et al.  AAM— Morgan et al.  40–49 50–59 60–69 70–79 2.5 3.5 4.5 6.5 3.4 6.5 11.3 12.5 2.5 3.5 3.5 3.5 2.0 4.0 4.5 5.5 *Note the similar PSA decision levels (boldface columns) for Caucasian men (CM) emphasizing specificity and those for AfricanAmerican men (AAM), relying on sensitivity. The sensitivitydriven column for CM also remains constant above age 50. 140 Littrup Fig. 1. Tumors having similar total volume and Gleason score could produce greater numbers of biopsies with individual Gleason scores $7 if the volume of the grade 4 component is larger in AAM. Perhaps the more interesting parameter for early detection outcomes is a test’s positive predictive value (i.e., PPV = true positives/[true positives + false positives]), insofar as this also reflects the incidence of the disease. Deriving the PPV for the early detection cohort in the Morgan et al.  study shows up to 50% higher PPV for AAM. We have also noted a similar pattern of significantly higher PPV for AAM in our large biopsy series (8), regardless of the PSA value or other diagnostic criteria. Significantly higher PPV for AAM therefore offers more insight into early detection, having greater cancer yield with lower false positives and lower screening costs among AAM. We have presented data  showing that screening costs for AAM will always be 25–50% less expensive than early-detection approaches for CM depending on the decision levels of PSA. Our large biopsy series suggests that the higher incidence of prostate cancer among AAM could have even greater mortality implications when we consider the types of cancers detected compared to the case with CM. The greater incidence of prostate cancer among AAM does not fully account for their disproportionate mortality when controlling for stage and suggests that prostate cancers in AAM have greater biologic potential. Delay in diagnosis, or lack of access to appropriate therapeutic options owing to socioeconomic factors, does not appear adequate to explain the 1.5 times greater mortality rate for AAM given similar-stage prostate cancer . Analogous to greater tumor vol- umes identified among AAM when controlling for PSA level , we have also noted significantly greater numbers of biopsy cores involved with cancer in AAM (P < 0.001) when a targeted sextant biopsy technique is utilized . The targeted sextant biopsy technique involves directing a biopsy within each sextant to the dominant region of suspicion as defined by hypoechoic change or foci of distinct increased color doppler flow. From comparison with our radical prostatectomy specimens, this leads to a more accurate representation of Gleason score with the eventual radical prostatectomy specimen . More importantly, significantly greater numbers of cancer cores among AAM had Gleason scores $7 (P < 0.001) and were distributed in a significantly greater multisextant pattern than with CM. In other words, the greater numbers of cancer cores seen among AAM appear to be related to more diffusion or greater volume of higher grade cancer among AAM (Littrup, unpublished data). The concept of the relative distribution of a grade $4 component has been proposed by McNeil  as a potent predictor of prostate cancer progression. Although the total volume of potentially curable prostate cancer has been estimated at 3–5 cc, the critical tumor mass for developing lymph node metastases has been estimated at 3.2 cc of grade $4 component within a cancer . Our biopsy data are being finalized for publication and suggest that the greater numbers of higher grade cancer seen on biopsy could account for Editorial the disproportionate mortality among AAM. We therefore proposed the grade-volume hypothesis of greater volumes of higher grade tumor in AAM (Fig. 1) to emphasize potential differences in not only incidence but also the types of cancers detected among AAM. Early detection for AAM utilizing any appropriate initiation point of PSA for subsequent follow-up or biopsy would detect greater numbers of clinically significant prostate cancers among AAM. The greater incidence of prostate cancer in AAM would thus translate into lower relative screening costs, and the greater biologic potential from tumor grade differences offers even greater potential mortality reductions from early detection. Lower screening costs with greater potential mortality reductions among AAM thus appear far more relevant and timely issues for prostate cancer early detection than focusing on potentially confusing statistical variations in PSA to derive age-specific, reference value differences. If high-risk groups (i.e., AAM and men with a family history of PCa) are advised to begin early detection at earlier ages , perhaps more work is needed in defining lower PSA decision levels for AAM based on attempts to reduce their disproportionate mortality rate. The focus of early detection should be the greatest potential mortality reduction for all men at the lowest human and economic costs. Conclusive mass screening trial results might not be accomplished in time for practical application or before their technological obsolescence. We believed that PSA decision levels of 2 ng/ml for high-risk men and 4 ng/ml for the general population are much simpler and are only the first step in tailoring individual risk assessment . Alternatively, we could continue calling multiple PSA ranges age-specific for CM  and change the term to age-sensitive PSA reference ranges for AAM . Peter J. Littrup, MD Departments of Radiology, Urology, and Radiation Oncology Wayne State University School of Medicine, Harper Hospital, Detroit, MI 48201 141 REFERENCES 1. Morgan TO, Jacobsen SJ, McCarthy WF, Jacobson DJ, McLeod DG, Moul JW: Age-specific reference ranges for serum prostatespecific antigen in black men. N Engl J Med 335:304–310, 1996. 2. Morton RA: Racial differences in adenocarcinoma of the prostate in North American men. Urology 44:637–645, 1994. 3. Mettlin CJ, Jones G, Averette H, Gusberg SB, Murphy G: Defining and updating the American Cancer Society guidelines for the cancer-related check-up: Prostate and endometrial cancers. Cancer 43:42–46, 1993. 4. Moul JW, Sesterhenn IA, Connelly RR, Douglas T, Srivastava S, Mostofi FK, McLeod DG: Prostate-specific antigen values at the time of prostate cancer diagnosis in African-American men. JAMA 274:1277–1281, 1995. 5. Oesterling JE, Jacobsen SJ, Chute CG, Guess HA, Girman CG, Panser LA, Lieber MM: Serum prostate-specific antigen in a community-based population of healthy men. Establishment of age-specific reference ranges. JAMA 270:860–864, 1993. 6. Mettlin CJ, Lee F, Drago J, Murphy GP: Investigators of the American Cancer Society National Prostate Detection Project: The American Cancer Society National Prostate Cancer Detection Project—Findings on the detection of early prostate cancer in 2,425 men. Cancer 67:2949–2958, 1991. 7. Pienta KT, Demers R, Hoff M, Kau TY, Montie JE, Seversen RK: Effect of age and race on the survival of men with prostate cancer in the metropolitan Detroit tri-county area, 1973–1987. Urology 45:93–102, 1995. 8. Littrup PJ, Sparschu RA, Gross ML, Klein RM, Segel MC, Zingas AP: Cost effective prostate cancer detection: Differences in screening in clinical populations [Abstract]. Radiology 197(P): 436, 1995. 9. Littrup PJ, Sparschu RA: Transrectal ultrasound and prostate cancer risks: The ‘‘tailored’’ prostate biopsy. Cancer 75(Suppl): 1805–1813, 1995. 10. Littrup PJ, Klein RM, Gross ML, Sparschu RA, Segel MC, Zingas AP: Color doppler of the prostate: Histologic and racial correlations [Abstract]. Radiology 197(P):365, 1995. 11. McNeal JE, Villers AA, Red Wine EA, Freiha FS, Stamey TA: Histologic differentiation, cancer volume, and pelvic lymph node metastasis in adenocarcinoma of the prostate. Cancer 66: 1225–1233, 1990.