Objectives: To investigate whether the inclusion of occult cancer in the control group can influence the association of prostate cancer and the polymorphism of manganese superoxide dismutase (MnSOD).
Methods: Prostate specimens and sera were obtained from 194 deceased men who did not have a history of prostate cancer. Eighteen-core biopsy specimens and whole-mount sections were evaluated histologically. The MnSOD genotype of the specimens was determined by polymerase chain reaction restriction fragment length polymorphism analysis.
Results: Tumors were present in 57 of the prostates, and biopsy detected 33 (including 1 false-positive finding). It detected 17 (1 false-positive finding) and missed 14 tumors in the subgroup of 135 specimens with a prostatic-specific antigen <4 ng/mL. The MnSOD AA genotype was associated with prostate cancer found in the step-sectioned specimens vs the control group in whom the absence of occult prostate cancer had been verified. However, no association was found if the control group consisted of subjects with negative biopsy results from the overall group or the subgroup with a prostatic-specific antigen level of <4 ng/mL.
Conclusions: The MnSOD AA genotype was associated with prostate cancer in our study; however, contamination of occult prostate cancer in the control group reduced the power of analysis and might yield seemingly negative results. Epidemiologic studies should strive to include control groups with a verified absence of occult cancer.