In Reply.—Increasing the PSA threshold to 5 or 6 ng/mL, as Dr Parker suggests, misses 30% to 50% of prostate cancers and decreases the proportion of curable cancers detected, thus defeating the purpose of screening. The cost of free PSA testing in only 9% of screened men (with PSA level of 4-10 ng/mL and with a benign prostate examination) is low compared with the 20% biopsy savings.
Dr Peven comments that the free PSA test "only increases the yield of carcinoma by 3%." Free PSA does not increase sensitivity. Rather, as shown in his Table 1, it increases specificity (avoids 20% of unnecessary biopsies) with minimal loss of sensitivity (detects 95% of cancers). In his theoretical population of 400 men, 60 would be spared unnecessary biopsies and 5 patients with cancer would be missed. Peven questions using free PSA testing to avoid biopsies, since 8% of patients whose free PSA test results are negative have cancer. This is only slightly higher than the prevalence in the general population (4%). Although all men could undergo a biopsy, this approach would be an inefficient use of resources because 92% of the biopsies in men with more than 25% free PSA would be negative. The cancers that would be detected would be primarily in older men with less aggressive tumors.