Variation of PSA values over time in individual patients


One of the problems with the use of the PSA test as an indicator of risk for prostate cancer is that the PSA values of specific individuals can vary over time … whether cancer is present or not.

A new study reported by Christensson et al. has tracked the intra-personal variation in PSA levels at three time points over about 2 weeks among men who had previously received a biopsy for perceived risk of prostate cancer.

Specifically, the trial enrolled 149 men who were already scheduled for a prostate biopsy. After the initial biopsy, the men were asked to undergo three blood draws, so that their PSA levels and other kallikrein levels could be tracked over time. All samples were stored, processed, and analyzed according to standard protocols. The markers levels to be measured were:

  • Total PSA level (tPSA)
  • Free PSA level (fPSA)
  • Percent fPSA (%fPSA)
  • Kallikrein-related peptidase 2 level (hK2)
  • Intact PSA level (iPSA)

The core results of the study were as follows:

  • The average (median) age of the patients was 65 years.
  • Based on their biopsy results, 52/149 patients (34.9 percent) had prostate cancer and 97/149 patients (65.1 percent) had benign disease.
  • The initial blood draw (BD1) was carried out at least 2 weeks after the scheduled prostate biopsy.
  • The second blood draw (BD2) was carried out an average (median) of 4 hours after BD1.
  • The third blood draw (BD3) was carried out an average (median) of 12 days after BD2.
  • The median tPSA level was 6.8 ng/ml (interquartile range, 4.5-9.6 ng/ml) at baseline.
  • Intra-individual variation was low for all biomarkers, and lowest for tPSA.
  • The ratio between SD1 and SD2 for tPSA was in the range 0.91 to 1.09 for 80 percent of participants.
  • The ratio between SD1 and SD2 for %fPSA was in the range 0.89 to 1.15.
  • 89 percent of men had a variation in tPSA of between 0 and 1 ng/ml from SD1 to SD2.
  • 17 percent of men had a variation in tPSA of > 2 ng/ml from BD1 to BD3.
  • Larger absolute changes were more likely in a man with a high baseline PSA.
  • The tPSA variation between BD1 and BD2 was 4 percent and between BD1 and BD3 was 11 percent.
  • Coefficients of variation between SD1 and SD2 for tPSA, fPSA, %fPSA, hK2, and iPSA were 4.0, 6.6, 6.0, 9.2, and 9.5 percent, respectively.
  • Variation in tPSA levels would have resulted in 4/149 men (2.7 percent) not meeting their institutional cutoff of 3 ng/ml for getting a biopsy.
  • If a cutpoint of 4 ng/ml had been used rather than 3 ng/ml, then 10/149 men (6.7 percent) would not have been given a biopsy.

The authors conclude that “the intra-individual variation for all the kallikrein-like markers studied was relatively small, especially for samples drawn the same day.” They note that few cases were reclassified between the time points and that this indicates the high short-term biological and technical reproducibility of the tests in clinical use.

The problem, of course, is that the variation in PSA results is just enough to present a real problem when it comes to the results in one individual. Furthermore, these data were all collected at specialized prostate cancer centers, and the blood tests were run according to well-controlled, standardized protocols. It is tempting to wonder whether the researchers would have obtained similar results from a similar series of patients  “out in the real world” … or whether a greater degree of variance would have been evident. It would not be difficult to test this. Working with (say) 10-12 community-based practices, one would just need to split each blood sample in two. One part would undergo the standard PSA methodology used by the community practices; the other would go to a centralized facility and be processed according to a standard protocol.

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: