PSA and Its Misunderstanding

We have examined the use of PSA over the past four plus years in some detail including it extensively in our Draft book on Prostate Cancer Genomics. Simply our conclusions are:

Past analyses of PSA effectiveness fail for several reasons.

First the ask the wrong question. The real question is what temporal PSA data and what threshold provide what level of mortality reduction from PCa.

Second the tested the wrong thing. For example the now classic European study used 4.0 but sampled every few years and not annually. 

Third, temporal data of PSA, %Free PSA, and PSA velocity should be used with data taken annually over a minimum of ten years.

Fourth, there are great differences in PSA assays. I have seen variations as great as 25-30% from assay to assay at the same time. It is argued that this is not true but anecdotally I have seen this too many times to believe the research.

Now the Mayo has published a piece decrying the use of PSA. It seems to be in line with the Task Force Recommendations, which we vehemently disagree with. The Mayo paper states:

The contrasting view, perhaps espoused most visibly by the US Preventive Services Task Force (USPSTF) is as follows: 

There is no randomized clinical trial that has reported overall lives being saved by PSA or equivalent screening programs.

Some of the published randomized trials have actually reported a deficit in survival among some of the screened populations (eg, older-aged cohorts) 

Most published data have not found a real population benefit from current prostate screening algorithms. 

As a result, the USPSTF has recommended against routine prostate screening strategies that are based on the currently available tools.

As we have demonstrated several times before, the USPSTF results are in error. The reasons we have summarized above. They conclude:

The important message of this study is that we should be focusing on the genesis of this cancer because it has become an important demographic and epidemiologic challenge. Toward this goal, these types of large and expensive surveys should be used more efficiently to cover a broad range of targets. 

We also need to apply PSA in its full context, moving rhetoric to the side, and attempt to
implement thoughtfully designed and well structured, hypothesis-driven studies to reveal ways to diagnose the more dangerous prostate cancer variants earlier, or prevent them from occurring, and to develop better management paradigms for those who present with advanced disease. 

International data suggest that there is a new epidemic of prostate cancer among urban Chinese, Koreans, and Japanese, domiciled in their homelands or in Western nations,16 and this may present an opportunity for further field testing of some of the strategies enumerated in this editorial at a sufficiently early time to have a real population-based effect.

The author bases the result on such comments as:

Prostate cancer is a remarkably heterogeneous disease, and there clearly is a subtype that will coexist in elderly men for many years, which poses no threat to their longevity or lifestyles unless disrupted by the consequences of aggressive post screening treatment algorithms. 

We need to educate men about the existence of prostate cancer, its symptoms and presentations, the availability of treatment, and key facts relating to the debate about screening. 

Yes, PCa is complex, and as we have demonstrated in our Draft book the genetic pathways are one of those complexities.  However there are both indolent and aggressive forms and at present we do not understand the difference. Stopping the gathering of data will not assist in that differentiation, gathering more data will. To the extent possible, we should also be genetically profiling every tumor, and with as many cells as is possible. Denigrating the few tests we have really does little.