Researchers in the US suggest that the prostate specific antigen (PSA) test is now less useful in prostate cancer screening, and may have lost some of
its predictive value, because biopsy techniques have improved in the ten years or so since the research that justified the use of PSA as a diagnostic tool
The study is the work of Drs Douglas Scherr and Michael Schwartz, and colleagues at the New York Presbyterian Hospital of the Weill Medical College of Cornell University, based in New York City, and is published in the 10 March online issue of the journal Cancer.
The correlation between PSA levels and positive prostate biopsy results was established at a time when biopsy patterns were not as they are today. Much of the data to justify the PSA test as a useful diagnostic tool was generated in the 1990s. Today, more biopsies are taken, and more samples are taken from each patient.
Scherr, Schwartz and colleagues examined data on 1,607 patients whose prostate biopsies were carried out at their hospital between 1993 and 2005 and grouped them according to when they received their biopsies: 1993-1997, 1998-2001, and 2002-2005. They then looked for how patterns of practice, and correlation between PSA and biopsy results shifted with time.
Using the number of biopsies performed, the number that were positive, the number of biopsy samples taken, the age of the patient, the most recent PSA before the biopsy and the volume of the prostate, the researchers assessed the potential links between PSA levels and the rate of positive biopsies.
They found that:
- The number of biopsies performed, the percentage of positive biopsies, and patient age did not change significantly over time.
- But there was more frequent use of PSA level 2.5 to 3.99 ng/ml as a biopsy indication (ie tendency toward lower PSA threshold).
- And there was an increase in the median number of core samples taken at the time of the biopsy.
- In men with a normal digital rectal exam, there was a moderate correlation between PSA level and positive biopsy rate between 1993 and 1997, but not after that.
- Fewer biopsies were performed for the indication of a suspicious digital rectal exam.
- The positive biopsy rate in men with PSA levels between 2.0 and 3.99 ng/ml was the same or higher than that for men with higher PSA levels.
- Multivariate analysis showed that PSA was not a significant predictor of biopsy result in men with a normal digital rectal exam.
"Early in the PSA era, the predictive power of PSA depended on multiple factors: high prevalence of disease, higher prevalence of high-grade disease, and low likelihood of prostate cancer diagnosis in men with low PSA."
But today, because of the "culling effect" of more frequent biopsies and shifts in biopsy patterns, they said that:
"The correlation between PSA and biopsy result is lost in men with normal DRE [digital rectal exam]," and that "diagnosing a higher proportion of tumors in men with a PSA between 2.0 and 4.0 ng/mL has negatively influenced the predictive value of PSA for cancer detection".
According to the American Cancer Society, the PSA level normally goes above 4 ng/ml when cancer is present, but around 15 per cent of men with a PSA under 4 will have prostate cancer.
Scherr told WebMD that whether the PSA level was as low as 2.5 or as high as 10, the chance of finding cancer was about the same in the men who had biopsies, so the usefulness of using PSA to find the men who need biopsies has been lost, he said, especially given that the prostate cancers they are finding are also smaller.
The researchers said there was an urgent need for new blood or urinary markers to predict who should have a prostate biopsy, because apart from family history or signs from previous biopsy results, there is little help available to give doctors a clue as to who should be referred for a prostate biopsy.
Dr Robert Smith, Director of Screening at the American Cancer Society told WebMD that while PSA is an imperfect test, it "happens to be the best one we have at the moment", and that:
"Because of the uncertainties, leading cancer organizations, including ACS, recommend that men discuss the pros and cons of this test with their doctors."
Dr Andrew Vickers a PSA researcher and statistician at the Memorial Sloan-Kettering Cancer Center in New York, also spoke to WebMD and said that PSA testing was a useful way to predict decades in advance which men will develop prostate cancer. He co-authored a paper recently that showed taking a single PSA reading by the age of 50 was a very strong predictor of cancer up to 25 years later.
"Negative Influence of Changing Biopsy Practice Patterns on the Predictive Value of Prostate-specific Antigen for Cancer Detection on Prostate Biopsy."
Michael J. Schwartz, David H. Hwang, Andrew J. Hung, Jullet Han, Justin W. McClain, M. Mendel Shemtov, Alexis E. Te, R. Ernest Sosa, E. Darracott Vaughan, Jr., and Douglas S. Scherr.
Cancer Published Online (early view): March 10, 2008.
Click here for Abstract.
Sources: Cancer press release, journal abstract, WebMD.
Written by: Catharine Paddock, PhD