Of risk and reality — the role of prostate cancer testing in young, high-risk patients

None of the trials conducted to date have provided sound information as to the risk and benefits associated with PSA testing among younger men (aged, say, between 35 and 50 years of age) who have well-known risk factors for prostate cancer. Such risk factors, of course, include:

  • A meaningful family history of prostate cancer
  • African American race, or other forms of Black African ethnicity (e.g., Afro-Caribbean and, potentially, Afro-European)
  • A history of exposure to certain types of environmental risk (e.g., Agent Orange)

In a new paper in the Journal of Urology, Vertosick et al. set out to see if they could offer insight into two questions:

  • “Do race and family history increase risk sufficiently to justify differential screening?”
  • “Might there be other risk factors that lead to superior risk stratification?” (compared to race and family history).

We should be clear that this is a complex paper involving a significant number of estimates and assumptions. It is not a prospective analysis of data from a large clinical trial; it is a theoretical and hypothesis-generating analysis based on extant data. It therefore has to be read with care and interpreted with caution. However, having said that, here is a summary of what Vertosick and her colleagues argue that they have been able to demonstrate from the data that are available:

  • Screening based on family history as a risk criterion involves about 10 percent of men and accounts for about 10 percent of prostate cancer-specific deaths.
  • Screening based on Black race as a risk criterion involves about 13 percent of men and accounts for about 28 percent of prostate cancer-specific deaths.
  • 44 percent of prostate cancer-specific deaths occur in the 10 percent of men who have the highest PSA levels at age 45 years.
  • Under no identifiable scenario that assessed risk for prostate cancer-specific mortality based on family history or Black race did the ratio  of the risk group size to the number of prostate cancer-specific deaths in that risk group approach that of PSA level.

Given the fact that we have no evidence that either race or family history have any impact on the relationship between PSA level and risk for prostate cancer-specific mortality, Vertosick and her colleagues draw three core conclusions from their data:

  • PSA levels at age 45 offered the best risk-related ratio between men screened and potential prostate cancer-specific deaths avoided.
  • PSA-based risk stratification would probably include men of Black race and men with a family history of prostate cancer who are at significant risk for aggressive forms of prostate cancer themselves.
  • Differential screening based on risk should be informed by a baseline PSA level.

The term “informed by” immediately above has been deliberately emphasized by us and not by the authors.

The key point that the authors are trying to make is that if early testing of high-risk subsets of patients is to be recommended by guideline-writing groups such as the American Urological Association or the National Comprehensive Cancer Network (and other guideline writing groups around the world), then baseline PSA testing should be a key factor in how those patients are being assessed. We would note that Vertosick and her colleagues neither recommend nor do they argue against the value of early testing of high-risk patients.

In an editorial comment on this paper, Hoffman implicitly commends Verstosick  et al. for not recommending baseline PSA testing in all men before the age of 50 years. However, the authors responded to that editorial comment by observing that:

There are no good reasons to believe that screening a man at age 45 years would result in greater harms or fewer benefits than screening a man at age 50 years. Rather, it is likely that the opposite is true since the risk of over diagnosis is low in a man with an anticipated life expectancy of greater than 30 years.

In the opinion of The “New” Prostate Cancer InfoLink, the realities of the value of PSA testing are, to a very large extent, still being masked by the behaviors of doctors and patients (as opposed to the actual risks). If an elevated PSA level of > 0.7 ng/ml led only to careful monitoring of that patient over time until a biopsy was actually justifiable, and if active surveillance was the actual standard of care until treatment was actually necessary, then it would become a great deal easier to appreciate and act on the concept of a baseline PSA level for most men at age 45 or thereabouts. The problem, of course, is that any elevated PSA level still tends to lead to a biopsy and any amount of cancer being found in the prostate still tends to lead to early, if not immediate, treatment. Until we can resolve that problem, we need to continue to be cautious about the over-use of baseline PSA testing in men who are under (say) 50 years of age.

As Hoffman writes in his editorial comment,

… clinicians should inform younger men at high risk about the uncertain value of early screening.

His statement is accurate. Young men at high risk who want to get a baseline PSA test need to appreciate that such a test is not definitive of anything. Although 44 percent of prostate cancer-specific deaths occur in the 10 percent of men who have the highest PSA levels at age 45 years, this also means that 56 percent of men who go on to die of prostate cancer do not have a PSA level in the top decile at age 45.

There are no easy answers … which is why Verstosick et al. are so careful to say only that “Differential screening based on risk should be informed by” (as opposed to defined by) a baseline PSA level.

6 Responses

  1. Sitemaster:

    Thanks for your post.

    It is disturbing that experts use flawed arguments to discourage PSA tests. Their basic assumption is that clinicians and/or patients are too stupid to understand PSA scores and make good decisions. Maybe we shouldn’t check our car’s tire and oil pressure since bad readings would force us to buy new cars?

    As you identified, the biggest problem is not the PSA test but what happens after the PSA test. In my opinion, standards of care now erroneously encourage blind biopsies, which in turn often lead to high-risk over-treatment for lower-risk patients … or, in some cases, to understating the risk and need for treatment.

    Rather, following confirmed excessive PSA scores, protocols should suggest low-cost, noninvasive tests … possibly Free PSA, PCA3, etc. Depending on results, a 3 T mpMRI would be recommended in order to identify areas of interest and potential spread outside the capsule. Depending on MRI results, there would be an MR-guided biopsy. Of course, advanced scans and genomic tests would be appropriate depending on facts and circumstances.

    Bottom line: There is an urgent need for the urological profession to take a fresh look at its standards to reflect changes in technology that will transform prostate cancer diagnosis and treatment. In addition, there is a great need for clinicians and patients to be better educated as to PSA scores, mpMRIs, and alternative treatments.

    Be well :-)


  2. As far as risks associated with Agent Orange exposure, that possibility is next to zero in men aged 35-50. Agent Orange was terminated in 1978. And it was rarely used in the US. This age group was too young for Viet Nam service. The number of women exposed to Agent Orange in Viet Nam was also a very small number so breast milk contamination would also be very rare.

    I think eventually, and possibly soon, biomarkers will help us decide when to biopsy (i.e., phi), or when to treat (quite a few available but really need trial data). Until then, the “baseline at age 40” will lead to more diagnosis and more treatment. Since almost no physicians would risk telling a 44-year-old man that he does not need treatment for any risk level of prostate cancer due to possible litigation, we have a conundrum that cannot be solved yet today.

  3. Tony:

    I have to say that I have very real doubts about the value of the phi test. We have no significant prospective data at all.

  4. I have two questions. First, what PSA level is the bottom of the top decile, at age 45?

    Second, you write: “Although 44 percent of prostate cancer-specific deaths occur in the 10 percent of men who have the highest PSA levels at age 45 years, this also means that 56 percent of men in the 10 percent who have the highest PSA levels at age 45 do not go on to die of prostate cancer (or necessarily to have prostate cancer at all).” Although you are much more learned than I am, something about this does not make sense to me. If 44% of prostate cancer-specific deaths occur in men who have the PSAs in the top decile at age 45, that does not mean 56% of men in that decile do not die of (or even have) prostate cancer. You are not saying 44% of the “top decile men” die of prostate cancer (meaning 56% don’t), you are saying 44% of the prostate cancer-specific deaths occur in the top-decile men (meaning 56% occur in guys who were not in the top decile at age 45). In reality, even among the top decile men, much fewer than 56% die of prostate cancer. Right?

  5. Barry:

    You are correct. My error. I have modified the text accordingly.

    With regard to the lowest PSA level in the top decile, I shall have to go back and look at the details in the paper later today, and I shall post the answer then.

  6. Barry:

    I apologize for the delay in getting back to you about the first part of your question above. As I thought (but wanted to check), the authors write that, “In the Malmö study 44% of prostate cancer deaths (95% CI 34–53) by age 75 years occurred in men in the top 10% of PSA at ages 45 to 49 years, equivalent to PSA 1.60 ng/ml or greater.” So the answer is 1.6 ng/ml.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s