Characteristics of prostate cancers detected in a multimodality early detection program

Abstract

Background. Few data are available to describe the clinical and pathologic characteristics of prostate cancers detected through early detection programs. The American Cancer Society National Prostate Cancer Detection Project (ACS‐NPCDP) is a multimodality, multicenter study of the feasibility of early prostate cancer detection using digital rectal examination (DRE), transrectal ultrasound (TRUS), and prostate specific antigen (PSA). One hundred fifty‐six prostate cancers are available from this project for analysis. Methods. The ACS‐NPCDP is a prospective, comparative study of a cohort of 2,999 men between 55 and 70 years of age not suspected of having prostate cancer. DRE, TRUS, and PSA are performed for each subject on an annual basis for as long as 5 years. Biopsies are performed on the basis of recommendations from DRE or TRUS results. Although elevated PSA alone was not typically a basis for biopsy, in some instances biopsies were recommended because of the degree of elevation in PSA. Diagnoses are confirmed by participating pathologists and by pathologic analysis. Results. A small proportion of cancers detected were advanced in terms of the clinical stage at time of diagnosis. A total of only six cancers were stage C1 to D1, and five of these were preexisting cancers detected at the first examination. Cancers detected by DRE tended to be more advanced than those found on the basis of only TRUS or PSA. A large proportion of patients received curative therapy, involving radical prostatectomy for 67.9% and radiation therapy for 17.9%. Of 100 men presumed to have organ confined disease and treated by prostatectomy, 64 actually proved to have localized cancer, a rate of upstaging of 36.0%. PSA level and PSA density were associated with the detection of organ confined cancer, but several advanced cancers had PSA levels in the normal range, limiting the usefulness of these measures for staging. Conclusions. The cancers resulting from this multimodality detection effort represented a spectrum of pathologic findings. These data, however, suggest that early detection interventions in men not suspected to have prostate cancer will yield tumors with a favorable stage distribution that are likely to benefit from treatment. Further follow‐up evaluation is needed to determine whether these benefits are reflected in long‐term morality and survival experince.