Prostate cancer risk calculators and the decision to have a prostate biopsy


A new paper in the Journal of Clinical Oncology provides some interesting insights into the relative merits of the two major prostate cancer risk calculators widely used today in North America: the original PCPT-based risk calculator (PRC) and the Sunnybrook risk calculator (SRC). The premise behind the use of both these risk calculators is that they should help a physician and his or her patients to make good decisions about whether to have a prostate biopsy — and ideally better decisions than would be possible based solely on the experience and expertise of the individual physician.

The complete set of easily available information on this new paper includes:

Let’s begin by summarizing the actual paper by Nam et al.

The authors’ primary goal was to test whether either the PRC or the SRC was significantly “better” than the other in predicting the presence of any cancer at all and high-grade cancer in particular in men at potential risk. The study was carried out prospectively (between 2007 and 2009) in a cohort of more than 2,000 patients at five Canadian medical centers. It is important to recognize that the selection of these centers introduced a limitation in how informative this study may really be (see below).

Here is what the study actually found:

  • The study was prospective and recruited 2,130 men (of median age 63 years and median PSA 5.7 ng/ml).
  • 331/2,130 men (15.5 percent) had a positive digital rectal exam.
  • 870/2,130 men (40.7 percent) were found to have cancer of any grade.
  • 403/2,130 men (18.9 percent) were found to have cancer of Gleason score 7 or higher.
  • 1,263/2,130 men (59.3 percent) had no cancer identified on biopsy.
  • The SRC was slightly better overall than the PRC in predicting the presence of cancer.
  • The SRC was also slightly better than the PRC in predicting the presence of cancers with a Gleason score of 7 or higher.
  • Neither the SRC nor the PRC was better than the other in predicting the presence of cancer in men with a relatively low risk threshold (below 30 percent risk) or a relatively high risk threshold (above 55 percent risk).

In interpreting these data, it is important to understand a variety of factors, as follows:

  • As mentioned above, the form of the PRC used in this study was the original form, based on the following input: the patient’s age, PSA level, ethnicity, family history of prostate cancer, digital rectal exam result, and biopsy history. Recent updates to the PRC can take account of data from %free PSA studies and PCA3 studies too, but these were not considered in the current prospective trial.
  • The PRC was created based on data collected from all across America as part of the Prostate Cancer Prevention Trial, so the data were collected from a wide range of men who were not necessarily believed to be at any specific risk for prostate cancer.
  • The SRC is based on slightly different input data from the PRC, including: the patient’s age, urinary voiding symptom score or IPSS, PSA level, %free PSA level, ethnicity, family history of prostate cancer, and digital rectal exam result.
  • The SRC was created based on data collected from a series of men referred for evaluation to the Sunnybrook center in Ontario, and who can therefore reasonably be expected to have been at higher risk for prostate cancer than the men who participated in the Prostate Cancer Prevention Trial.
  • The fact that the Sunnybrook calculator was developed using a cohort of men largely recruited from Ontario in Canada, and that 83 percent of the men participating in the current study were also recruited at institutions in Ontario, may have introduced a regional population bias that favored the Sunnybrook calculator into the current trial.

It is not possible to use the current study to conclude that the SRC is any better or worse than the currently available versions of the PRC. However, this first prospective evaluation of the two major risk calculators is highly informative about the potential for further research into how best to determine the relative merits of differing types of calculator and their use in prostate cancer diagnosis.

According to Vickers, in his editorial comments, what is also clear from the current study is that the practical value of the Sunnybrook calculator may not be that good when risk is low (because it seems to underestimate risk when that risk is low) and  also may not be that good when risk is higher (compared to just biopsying every man with a high PSA or a positive digital rectal exam).

So what’s the bottom line?

The role of prediction tools in the management of prostate cancer is still evolving. The currently available prostate cancer risk calculators do offer guidance as to risk, but they can not tell an individual with great accuracy whether a biopsy is or is not necessary for him as an individual. On the other hand, by using the available calculators, patients and their doctors may be able to have a more reasonable discussion about the necessity for biopsy based on the available information than a discussion based exclusively on an elevated PSA result or a positive biopsy result alone.

Even though this trial may have raised more questions about the value of prostate cancer risk calculators than it has answered, Nam and his colleagues should be congratulated for their willingness to conduct such a prospective trial of the two different risk calculators most widely used in North America today. It is only through conduct of such prospective studies that we will be able to actually improve the practical, day-to-day diagnosis of clinically significant prostate cancer and reduce the risk for over-treatment of potentially indolent, low-risk, and very low-risk disease.

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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: