The sensitivity of the PSA test as an indicator of the presence of prostate cancer is known to be about 80 to 90 percent. In lay terms, what this means is that 80 to 90 percent of men who actually have prostate cancer cells in their prostate will have an elevated PSA level compared to men who show no sign of prostate cancer after having at least one biopsy. (Why 80 to 90 percent? Because it depends where you draw the cut-off line for an “elevated” PSA level at 4 ng/ml or at 3 ng/ml, or at 2.5 ng/ml. The lower the cut-off line, the higher the sensitivity.)
The selectivity of the PSA test is about 50 percent. In lay terms again, this means that an elevated PSA level (again compared to PSA levels of men who show no sign of having prostate cancer after at least one biopsy) is a complete toss-up. You could have some prostate cancer cells … but you could just as easily have none.
According to an article on ther Medscape web site , Pincus et al. have now reported (in a presentation at the annual meeting of the American Society for Clinical Pathology) that the sensitivity of a first prostate biopsy is about 80 percent. In other words, 80 percent of men with prostate cancer cells in their prostate will have a positive biopsy at their first attempt (and 20 percent won’t). They also report that 100 percent of men in their cohort who had a PSA level of 40 ng/ml or higher had a first biopsy that was positive for prostate cancer.
Now let us be very clear indeed. Having some prostate cancer cells in your prostate and having a clinical condition called prostate cancer are not the same thing. To have the clinical condition, you must have prostate cancer cells in your prostate. But having prostate cancer cells in your prostate may not have any clinical significance at all. Why is this important?
It is important because Dr. Pincus and his colleagues have used a circular argument to justify their opinion that “PSA is an excellent marker for prostate cancer, since it has a similar sensitivity to a first prostate biopsy and is less invasive with less risk of morbidity.” What? Say that again!
OK. Let’s say that again this time in English … Dr. Pincus is arguing that because a prostate biopsy has an 80 percent specificity for finding prostate cancer cells in the prostate, and 80 to 90 percent of men who actually have prostate cancer also have an elevated PSA, then PSA is “an excellent marker for prostate cancer.” The problem, of course, is threefold:
- 50 percent of men with an elevated PSA level still don’t have prostate cancer.
- Of the 50 percent of men who do have an elevated PSA level, about 30 percent who have a subsequent positive biopsy will have clinically insignificant disease which may now get treated unnecessarily.
- And the suggestion that the PSA test is “an excellent marker ” is in comparison to a first biopsy. Is anyone seriously suggesting that we should just biopsy every man in America when he hits 40 (0r 50 or whatever) years of age?
Only a pathologist could think like this. It ignores the entire reality of whether the man considering his potential risk for prostate cancer has a right to an opinion on the subject.
The research team carefully points out that 96 percent of the 1,665 patients at the Veterans Administration New York Harbor Health Care System on whom their data are based and who were found to have prostate cancer (cells in their prostates) on biopsy underwent biopsy only because of an elevated PSA level! That means that (presumably) about half these men had a first biopsy because of an elevated PSA level, and that their first biopsies were all negative (even though 20 percent of them were false negatives). Of course there are no data provided about whether even one of those patients who had a negative biopsy got a serious systemic infection or worse still died as a consequence of such an infection but that is well within the limits of probability.
If there was ever a study that clearly demonstrated the need for a much better test to assess the risk of clinically significant prostate cancer, this was that study.
The “New” Prostate Cancer InfoLink does not believe that the PSA test is a “bad” test or that it shouldn’t be used. We also do not think that people shouldn’t get biopsies. But we do believe that it’s high time to stop pretending that the PSA test is the test we need. It is simply the test we have. It needs to be used with caution. Subsequent biopsies should also be used with caution. We need to find a much better test just as soon as we possibly can! And we need to be a great deal clearer in our heads that having some prostate cancer cells in our prostates is not necessarily the same as having the clinical condition formerly (and formally) known as prostate cancer.
The sensitivity of the PSA test as an indicator of the presence of prostate cancer is known to be about 80 to 90 percent. In lay terms, what this means is that 80 to 90 percent of men who actually have prostate cancer cells in their prostate will have an elevated PSA level compared to men who show no sign of prostate cancer after having at least one biopsy. (Why 80 to 90 percent? Because it depends where you draw the cut-off line for an “elevated” PSA level at 4 ng/ml or at 3 ng/ml, or at 2.5 ng/ml. The lower the cut-off line, the higher the sensitivity.)
The selectivity of the PSA test is about 50 percent. In lay terms again, this means that an elevated PSA level (again compared to PSA levels of men who show no sign of having prostate cancer after at least one biopsy) is a complete toss-up. You could have some prostate cancer cells … but you could just as easily have none.
According to an article on ther Medscape web site , Pincus et al. have now reported (in a presentation at the annual meeting of the American Society for Clinical Pathology) that the sensitivity of a first prostate biopsy is about 80 percent. In other words, 80 percent of men with prostate cancer cells in their prostate will have a positive biopsy at their first attempt (and 20 percent won’t). They also report that 100 percent of men in their cohort who had a PSA level of 40 ng/ml or higher had a first biopsy that was positive for prostate cancer.
Now let us be very clear indeed. Having some prostate cancer cells in your prostate and having a clinical condition called prostate cancer are not the same thing. To have the clinical condition, you must have prostate cancer cells in your prostate. But having prostate cancer cells in your prostate may not have any clinical significance at all. Why is this important?
It is important because Dr. Pincus and his colleagues have used a circular argument to justify their opinion that “PSA is an excellent marker for prostate cancer, since it has a similar sensitivity to a first prostate biopsy and is less invasive with less risk of morbidity.” What? Say that again!
OK. Let’s say that again this time in English … Dr. Pincus is arguing that because a prostate biopsy has an 80 percent specificity for finding prostate cancer cells in the prostate, and 80 to 90 percent of men who actually have prostate cancer also have an elevated PSA, then PSA is “an excellent marker for prostate cancer.” The problem, of course, is threefold:
Only a pathologist could think like this. It ignores the entire reality of whether the man considering his potential risk for prostate cancer has a right to an opinion on the subject.
The research team carefully points out that 96 percent of the 1,665 patients at the Veterans Administration New York Harbor Health Care System on whom their data are based and who were found to have prostate cancer (cells in their prostates) on biopsy underwent biopsy only because of an elevated PSA level! That means that (presumably) about half these men had a first biopsy because of an elevated PSA level, and that their first biopsies were all negative (even though 20 percent of them were false negatives). Of course there are no data provided about whether even one of those patients who had a negative biopsy got a serious systemic infection or worse still died as a consequence of such an infection but that is well within the limits of probability.
If there was ever a study that clearly demonstrated the need for a much better test to assess the risk of clinically significant prostate cancer, this was that study.
The “New” Prostate Cancer InfoLink does not believe that the PSA test is a “bad” test or that it shouldn’t be used. We also do not think that people shouldn’t get biopsies. But we do believe that it’s high time to stop pretending that the PSA test is the test we need. It is simply the test we have. It needs to be used with caution. Subsequent biopsies should also be used with caution. We need to find a much better test just as soon as we possibly can! And we need to be a great deal clearer in our heads that having some prostate cancer cells in our prostates is not necessarily the same as having the clinical condition formerly (and formally) known as prostate cancer.