Rorshach and biomarkers

To read this blog on my blog site and sign up to be notified of future posts (as well as being able to search for previous blogs) click here.

Psychology, or for that matter being able to read others’ personalities, has never been a strong suit for me. Neither was art – I still am at the stick figure stage when drawing. It turns out that Hermann Rorshach was probably good at both. The question of what you see when looking at an ink blot seems relevant to the current status of biomarkers in prostate (and others) cancer. On the one hand, some biomarkers are fabulous – for example the Philadelphia chromosome, described in 1959, was the first unique cancer marker that ultimately resulted in a specific targeted treatment, imatinib (Gleevec), dramatically improving survival for patients with chronic myelogenous leukemia. PSA, on the other hand (our “favorite”) is not so great, and as I previously noted, may give rise to the “PSA Clock” effect in which patients ruin their lives by clock watching. But, as we know, it is remarkably useful as a weather vane. When a prostate cancer patient is being followed on any sort of therapy, going down is good and going up is bad.

Thus, there have been thousands of articles attempting to either make PSA interpretation  better, or to replace it with more sensitive or more accurate predictors of prostate cancer behavior. I reviewed some of these, and the challenges here. Today, yet another article on a rather “simple” biomarker, PTEN loss, showed up among the >20 prostate related emails I receive each day. Writing in European Urology, a group of well-known prostate cancer investigators looked at immunohistochemistry (using special stains to highlight a protein in cells under a microscope slide) to evaluate loss of PTEN, a tumor suppressor gene, in prostatectomy specimens. This simple test (in this particular experiment) was as good as the commercial Prolaris test that evaluates a panel of genes related to how fast cells are dividing in predicting biochemical recurrence (PSA relapse) or prostate cancer specific mortality. With PTEN loss, the chances of having a biochemical relapse (rising PSA) or developing metastases or dying in a 10 year followup period were significantly greater than if you did not have PTEN loss. A simple, inexpensive test might replace a more complicated one.

Here’s where Dr. Rorshach’s psychological construct comes into play at so many levels. If you are the scientist looking through the microscope, do you score a loss when there is only faint staining? Are you sure you are looking at a cancer cell and not a normal cell or  a stromal cell, or maybe even an immune cell? If you decide on giving a score to each cell, say “1+, 2+, or 3+” staining, how do you add all those up?  How many cells should you examine? All parts of the tumor, or only the most aggressive (Gleason pattern ≥ 4) And if you can figure all that out, can you teach your colleagues to look at the same specimen(s) and come up with the same answer? These are the challenges we face when we move a lab experiment into the clinic (and they are well recognized by the authors).

But…there is more! Look at the graphs. Obviously you would rather be on the upper curve with PTEN present, but how bad is it really? At 10 years, only ~10% of the men had developed metastases or died in this study. Recognize that these men were a cross section of patients, median age 59, median PSA 5.9, 64% Gleason 3+3 and another 23% Gleason 3+4 with pretty low a priori risk (did we need the PTEN test to tell us?). So the real issue is whether you would want anything different done to you if you were one of the few patients with Gleason 3+3 and PTEN loss, just because you have this new information? And what would that be?? Radiation? Hormone therapy? How much and how long? -all in the psychology of looking at those curves. Some men might want nothing more done, while others would want “the kitchen sink” thrown at them, even if they had relatively little (and unproven) to gain.

So, medicine remains as much an art as it is a science (with no offense to my mathematical statistical colleagues). As the father of American Internal Medicine, William Osler, told his students, “Common sense in matters medical is rare, and is usually in inverse ratio to the degree of education.”