PSA-Density: easily obtainable clinical parameter to help excluding significant prostate cancer.

J Urol. 2017 Mar 31. pii: S0022-5347(17)43846-8. doi: 10.1016/j.juro.2017.03.130. [Epub ahead of print] PubMed PMID: 28373135.

Short summary 

PSA-Density was included into the PIRADS-algorithm to rule out clinically significant prostate cancer. A combination of PSA-Density and PIRADS-Score yielded a higher diagnostic accuracy as compared to each of these parameters alone.   

Conclusion

• The inclusion of PSA-Density into the PIRADS-algorithm improved the NPV in men with suspected prostate cancer (n=1040) and in a subpopulation with previous negative TRUS-biopsy (n=443).
• The improvement in NPV was about 10% in both of the aforementioned groups.
• The addition of information on previous negative biopsy did not add further value to the combination of PIRADS-score and PSA-Density alone.   
• By using the combination of PIRADS-score and PSA-Density to rule out significant prostate cancer, 19% of biopsies could have been avoided.

Patient cohort

Prospective assessment of 1040 consecutive men with suspicion of prostate cancer.

The New and the Good

• Incorporation of an easily obtainable clinical parameter was shown to increase the NPV of mpMRI alone.
• Good Standard of Reference (MR-/TRUS fusion biopsy PLUS transperineal systematic biopsy).
• Prospective analysis of a large patient cohort.

Limitations

• Single reader assigning PIRADS-scores.
• Size threshold could have been included into a further definition of clinical significance.
• No information on the PSA-Density-Group was provided in patients with PIRADS-3 lesions which harboured significant cancer.

Possible consequences for clinical practice

• As PSA-Density is easily obtained and in many institutions routinely included in the report, it may be easily integrated into mpMRI reporting in daily clinical practice.
• A PSA-Density of <0.15 may render the radiologist more confident when reporting a PIRADS-1 or PIRADS-2 score. Consequently the urologist may more confidently refrain from a biopsy in his patient.

Possible consequences for PIRADS v3

• It would have been interesting to see if PSA-Density may guide further assessment in equivocal lesions (e.g. PIRADS-3 lesions):
• If a high percentage of the 177/252 (69.4%) PIRADS-3 lesions which did not harbour significant cancer would have shown a PSA-Density of <0.15, these lesions could be treated as “lower suspicion” PIRADS-3 lesions and could be followed up using MRI and PSA instead of immediate biopsy:

• Based on the findings of a previous study (click here), it could be discussed whether easily obtained clinical data such as PSA-Density and information on results of previous TRUS-biopsy should be included together with quantitative image information (ADC) into a updated PIRADS-algorithm.

Future study ideas derived from this paper

• The findings could/should be validated in a multicenter-study with multiple readers.
• Findings could be extended regarding the purpose of further triaging PIRADS-3 lesions.

Personal Comment

The PSA density is routinely included in the prostate-MRI reports in our institution. My personal cutoff for feeling more confident with an equivocal MRI-finding was always a PSA-Density of 0.15. Hence, I am glad, that this personal cutoff proved to be a scientifically valid cutoff as well and I may continue using it knowing to have a bit more evidence backing me up…