Skip to main content

22-08-2011 | Endocrinology | Article

Novel PSA test could improve diagnostics


Free abstract

MedWire News: Researchers report the development of a novel prostate screening test that is more sensitive than current diagnostic methods at indicating the presence of prostate cancer.

The test involves taking a urine sample and using it to separate prostate specific antigen (PSA) protein structures into 'cancer' and 'non-cancer' pathologies, a process known as solvent interaction analysis (SIA).

The current prostate screening method involves taking a blood test to calculate a total PSA cut-off level that indicates the presence or absence of prostate cancer. However, the test is not completely reliable as it cannot accurately distinguish between malignant and benign disease, and PSA levels can be raised for reasons other than prostate cancer.

Regarding the novel PSA/SIA test, lead author Mark Stovsky (Case Western Reserve University, Boston, Massachusetts, USA) said: "This test provides a new way to look at prostate cancer diagnosis utilizing a novel biological assay which differentiates PSA molecular structures arising from cancer versus non-cancerous glands."

Stovsky and colleagues used data for 222 patients with an indication for prostate biopsy, 122 of whom had benign results, and 100 of whom had prostate cancer. This gave an overall false-positive rate of 55% for the standard diagnostic models used to select patients for prostate biopsy.

SIA involves the partitioning of proteins in aqueous-based 2-phase systems to enable sensitive detection of diverse structural changes, explain the researchers. After partitioning PSA proteins in urine, the two aqueous phases contain different amounts of PSA isoforms, at which point conventional immunosorbent assay can be used to create a ratio between the two PSA levels. This numerical ratio is called K, say Stovsky et al, and they used a cut-off value of 1.73 and higher to indicate prostate cancer.

As reported in the journal Urology, this K cut-off resulted in a sensitivity of 100% and a specificity of 80% for the PSA/SIA test.

Furthermore, even though the cut-off for K was determined retrospectively, the raw data showed an area under the receiver operating characteristic curve score (AUC) of 0.90 (where 1 = perfect discrimination) for detecting prostate cancer using the PSA/SIA test, compared with an AUC of 0.58 for the total PSA test.

"This has the potential to be a major advance in the development of more accurate tests for prostate cancer diagnosis," said Stovsky.

He and his team believe that another use of the PSA/SIA test could be its "influence on the recommended PSA cut-off screening value." The high sensitivity and specificity of their novel test may enable the total PSA test threshold to be lowered, without increasing the number of biopsies taken.

Future studies may also reveal a use for the PSA/SIA in correlation with other prostate cancer clinical variables such as Gleason score, PSA velocity, age, race, and assessment of biochemical failure after treatment, suggests the research team.

"It is encouraging to note that differential partitioning of the structurally heterogenous PSA isoform population predicted biopsy results better than current selection criteria," conclude Stovsky et al.

By Sarah Guy

Related topics