TY - JOUR
T1 - Prostate-specific antigen levels are higher in African-American than in white patients in a multicenter registration study
T2 - Results of RTOG 94-12
AU - Vijayakumar, Srinivasan
AU - Winter, Kathryn
AU - Sause, William
AU - Gallagher, Michael J.
AU - Michalski, Jeff
AU - Roach, Mack
AU - Porter, Arthur
AU - Bondy, Melissa
PY - 1998/1/1
Y1 - 1998/1/1
N2 - Purpose: To compare serum prostate-specific antigen (PSA) levels in a national sample of African-American and white men with prostate cancer, and to attempt to explain any differences by using self-reported individual- level socioeconomic status adjustments. Methods and Materials: During 4 1/4 months in 1994-95, 709 patients with nonmetastatic prostate cancer were enrolled in this prospective study 17.5% were African-American and 82.5% were white. Information about clinical stage, tumor grade, pretreatment PSA, type of insurance and educational and income status was obtained. Serum PSA levels were measured and racial differences were found; how the differences were influenced by other patient- or tumor-related factors and if the differences could be explained by socioeconomic status disparities were determined. In univariate analyses, factors associated with the mean PSA levels were studied; log-converted values were used to yield a normal distribution. Multivariate analyses were done on log-linear models for description of association patterns among various categorical variables; a perfectly fitted model should have a correlation value (CV) of 1.0. Results: The mean PSA level was higher in African-Americans (14.68 ng/ml) than in whites (9.82 ng/ml) (p = 0.001). Clinical stage (p = 0.001), Gleason sum tumor grade (p = 0.0001), educational level (p = 0.001), and household income (p = 0.03) were also associated with mean PSA levels: age, type of biopsy, and insurance status were not. Disease stage (p = 0.0001), grade (p = 0.0001), education (p = 0.07), and income (p = 0.02) were all associated with PSA levels for whites, but none of these factors were important for African-Americans (all p values > 0.1). The best fitted log-linear model (CV = 0.99) contained PSA (< 10, 10-20, and > 20), Gleason sum grade (2-5, 6-7, and 8-10), race, and two interactions: PSA by race (p = 0.0012) and PSA by Gleason sum (p = 0.0001). Models replacing race for either income (CV = 0.82) or education (CV = 0.82) or both (CV = 0.78) did not fit as well. Conclusions: African-Americans with nonmetastatic prostate cancer have higher serum PSA levels at diagnosis than whites, implying a higher tumor cell burden. Individual-level household income, education, or insurance status alone or in combination account for racial differences, but only partially.
AB - Purpose: To compare serum prostate-specific antigen (PSA) levels in a national sample of African-American and white men with prostate cancer, and to attempt to explain any differences by using self-reported individual- level socioeconomic status adjustments. Methods and Materials: During 4 1/4 months in 1994-95, 709 patients with nonmetastatic prostate cancer were enrolled in this prospective study 17.5% were African-American and 82.5% were white. Information about clinical stage, tumor grade, pretreatment PSA, type of insurance and educational and income status was obtained. Serum PSA levels were measured and racial differences were found; how the differences were influenced by other patient- or tumor-related factors and if the differences could be explained by socioeconomic status disparities were determined. In univariate analyses, factors associated with the mean PSA levels were studied; log-converted values were used to yield a normal distribution. Multivariate analyses were done on log-linear models for description of association patterns among various categorical variables; a perfectly fitted model should have a correlation value (CV) of 1.0. Results: The mean PSA level was higher in African-Americans (14.68 ng/ml) than in whites (9.82 ng/ml) (p = 0.001). Clinical stage (p = 0.001), Gleason sum tumor grade (p = 0.0001), educational level (p = 0.001), and household income (p = 0.03) were also associated with mean PSA levels: age, type of biopsy, and insurance status were not. Disease stage (p = 0.0001), grade (p = 0.0001), education (p = 0.07), and income (p = 0.02) were all associated with PSA levels for whites, but none of these factors were important for African-Americans (all p values > 0.1). The best fitted log-linear model (CV = 0.99) contained PSA (< 10, 10-20, and > 20), Gleason sum grade (2-5, 6-7, and 8-10), race, and two interactions: PSA by race (p = 0.0012) and PSA by Gleason sum (p = 0.0001). Models replacing race for either income (CV = 0.82) or education (CV = 0.82) or both (CV = 0.78) did not fit as well. Conclusions: African-Americans with nonmetastatic prostate cancer have higher serum PSA levels at diagnosis than whites, implying a higher tumor cell burden. Individual-level household income, education, or insurance status alone or in combination account for racial differences, but only partially.
KW - African-Americans
KW - Prostate cancer
KW - Prostate-specific antigen
KW - Race
UR - http://www.scopus.com/inward/record.url?scp=0031984026&partnerID=8YFLogxK
U2 - 10.1016/S0360-3016(97)00834-1
DO - 10.1016/S0360-3016(97)00834-1
M3 - Article
C2 - 9422553
AN - SCOPUS:0031984026
SN - 0360-3016
VL - 40
SP - 17
EP - 25
JO - International Journal of Radiation Oncology Biology Physics
JF - International Journal of Radiation Oncology Biology Physics
IS - 1
ER -