Lucas Engelage, Niklas Behnel, Agron Lumiani, Oleksii Bashkanov, Leo Federico Stadelmeier, Alexander Tamalunas, Alexander Buchner, Ronald Sroka
A universal 0.15 ng/ml/cm³ PSAD threshold fails to account for prostate volume variability. Potential volume-adapted cut-offs (0.25 / 0.15 / 0.12 ng/ml/cm³ for < 30 / 30-50 / >50 cm³) better approximate stratum-specific optimal values, although continuous multivariable risk prediction outperformed any fixed-threshold strategy and PSAD is interpreted alongside MRI in practice. Alternative biomarkers should be considered for very large prostates where PSAD accuracy deteriorates. These findings are hypothesis-generating and require external validation with PI-RADS integration before potential guideline implementation.
PURPOSE: PSA density (PSAD) may improve detection of clinically significant prostate cancer (csPCa) by normalizing PSA for prostate volume. Current practice applies a uniform PSAD threshold of 0.15 ng/ml/cm³; however, whether this approach performs adequately across prostate volumes remains unclear. This study evaluated the diagnostic performance of PSAD for csPCa and investigated the influence of age and prostate volume on its accuracy.
METHODS: A total of 3,923 MRI-guided biopsy sessions from 3,087 patients were retrospectively analyzed. PSAD was calculated from total PSA (ng/ml) and MRI-derived prostate volume (cm³). ROC analysis determined AUC and Youden-optimal cut-offs, stratified by age (< 60, 60-69, ≥ 70 years) and prostate volume (< 30, 30-50, > 50 cm³), with bootstrap CIs.
RESULTS: Overall AUC was 0.758 (95% CI 0.743-0.773). Optimal cut-offs varied substantially by volume: 0.246 ng/ml/cm³ (95% CI 0.156-0.292) for small (< 30 cm³, AUC 0.717), 0.144 ng/ml/cm³ (0.124-0.178) for moderate (30-50 cm³, AUC 0.734), and 0.124 ng/ml/cm³ (0.102-0.162) for large prostates (> 50 cm³, AUC 0.702). Accuracy deteriorated progressively in very large prostates (AUC 0.658 for > 80 cm³, 0.609 for > 120 cm³). At volumes > 80 cm³, the 0.15 ng/ml/cm³ threshold achieved only 33.1% sensitivity. Age showed minimal clinical impact (AUC improvement 0.019).
CONCLUSION: A universal 0.15 ng/ml/cm³ PSAD threshold fails to account for prostate volume variability. Potential volume-adapted cut-offs (0.25 / 0.15 / 0.12 ng/ml/cm³ for < 30 / 30-50 / >50 cm³) better approximate stratum-specific optimal values, although continuous multivariable risk prediction outperformed any fixed-threshold strategy and PSAD is interpreted alongside MRI in practice. Alternative biomarkers should be considered for very large prostates where PSAD accuracy deteriorates. These findings are hypothesis-generating and require external validation with PI-RADS integration before potential guideline implementation.