Introduction

Previous guidelines to screen for prostate cancer (PCa) using serum PSA levels led to millions of unnecessary prostate biopsies that revealed no PCa or low-risk PCa that was unlikely to affect survival. PSA density has shown promise as a better marker to indicate prostate biopsy and has been well investigated in the past using sextant or octant biopsies. While most prior studies only showed that PSA density could distinguish between BPH and PCa, we compared the sensitivity and specificity of PSA density versus PSA to detect any and significant PCa across different PSA ranges using the current standard of extended template prostate biopsies.

Methods

Participants were a prospective cohort of men referred for prostate biopsy using an extended template biopsy scheme to evaluate PCa at 26 sites throughout the US. We analyzed the area under the receiver operating characteristic curve to assess the predictive accuracy of PSA density versus PSA across three PSA ranges (<4, 4-10, >10 ng/mL) and in men with or without a prior negative biopsy. We assessed the detection of any and significant (Gleason score ≥ 7) PCa.

Results

Among 1,290 patients with available data, 585 (45%) and 284 (22%) men had any PCa and significant PCa, respectively. PSA density was significantly more predictive than PSA for detecting any and significant PCa in the PSA ranges of 4-10 and >10 ng/mL. AUC for significant prostate cancer was 0.72 (0.68, 0.77), p<0.0001 for PSA 4-10 ng/mL and 0.82 (0.75, 0.89), p<0.0001 for PSA >10 ng/mL. PSA density was significantly more predictive than PSA in detecting any and significant PCa in men with and without a prior negative biopsy. However, the incremental AUC value was larger for significant PCA than any PCA in men who had a prior negative prostate biopsy (AUC 0.81 vs 0.70, p = 0.0042), and those who did not (AUC 0.77 vs 0.73, p = 0.0026).

Conclusions

In contrast to previous studies, we found that PSA density outperformed PSA most within the PSA range >10 ng/mL, suggesting that PSA density will save large volume prostate patients from the costs associated with the over-diagnosis of PCa. Additionally, PSA density performed best among men with a prior negative biopsy, saving these men from the burden of repeated biopsies that are likely to be negative in this population. With the current need for better markers to indicate prostate biopsy, PSA density may have significant value as a more sensitive and specific test than PSA to detect PCa when used with an extended template biopsy scheme.

Funding

None