Despite limited scientific support, a repeat prostate-specific antigen (PSA) test before prostate biopsy decisions is common. We analyzed biopsy outcomes in 1686 men from the STHLM3 study with PSA 3-10 ng/mL and two PSA tests taken within eight weeks and before prostate biopsy using percentages and multinomial logistic regression. We found that omitting prostate biopsy for men with PSA values decreasing to PSAs of 3 ng/mL or less would save 16.8% of biopsy procedures, while missing 5.4% of the cancers with Gleason scores (GSs) of 7 or higher. The proportion of cancers with GSs of 6 or lower was independent of the first PSA value, as well as of PSA change. Also, the risk of tumors with GSs of 7 or higher decreased with both decreasing and increasing PSA levels: It was 18.6% (95% confidence interval [CI] = 16.3% to 20.9%) for men with PSA changes of less than 20%, 12.1% (95% CI = 8.0% to 16.2%) for men with PSA levels increasing at least 20%, and 6.6% (95% CI = 3.8% to 9.3%) for men with PSA levels decreasing at least 20%.
Journal of the National Cancer Institute. 2016 Jul 14*** epublish ***
Tobias Nordström, Jan Adolfsson, Henrik Grönberg, Martin Eklund
Department of Medical Epidemiology and Biostatistics (TN, HG, ME), Department of Clinical Sciences at Danderyd Hospital (TN), and Department of Clinical Science, Intervention and Technology (JA), Karolinska Institutet, Stockholm, Sweden; Swedish Agency for Health Technology Assessment and Assessment of Social Services, Stockholm, Sweden (JA)., Department of Medical Epidemiology and Biostatistics (TN, HG, ME), Department of Clinical Sciences at Danderyd Hospital (TN), and Department of Clinical Science, Intervention and Technology (JA), Karolinska Institutet, Stockholm, Sweden; Swedish Agency for Health Technology Assessment and Assessment of Social Services, Stockholm, Sweden (JA)., Department of Medical Epidemiology and Biostatistics (TN, HG, ME), Department of Clinical Sciences at Danderyd Hospital (TN), and Department of Clinical Science, Intervention and Technology (JA), Karolinska Institutet, Stockholm, Sweden; Swedish Agency for Health Technology Assessment and Assessment of Social Services, Stockholm, Sweden (JA)., Department of Medical Epidemiology and Biostatistics (TN, HG, ME), Department of Clinical Sciences at Danderyd Hospital (TN), and Department of Clinical Science, Intervention and Technology (JA), Karolinska Institutet, Stockholm, Sweden; Swedish Agency for Health Technology Assessment and Assessment of Social Services, Stockholm, Sweden (JA) .