This article is more than 5 years old. Information may no longer be current.
Schrödinger’s cat: PSA screening is alive and dead
Click Here to Manage Email Alerts
The recent paper from Martin and colleagues in JAMA — which reported the ProtecT trial from the United Kingdom, an important study adding another nail into the coffin of PSA screening — reminds me of Schrödinger’s cat.
To remind you, the Nobel prize-winning Austrian physicist Erwin Schrödinger posed an existential conundrum to challenge the early theoretical view of quantum mechanics advanced by Niels Bohr and Werner Heisenberg.
These equally famed European physicists, in defining the initial theoretical concept of quantum mechanics, postulated that matter could simultaneously exist in multiple forms. Specifically, they suggested that multiple constructs of an electron orbiting around an atom could exist synchronously, but could not easily be measured without the measurement itself disrupting the electron’s orbit and/or velocity, thus vitiating the experiment.
Thus, this theorem did not allow for simultaneous definition of where the electron actually was at any time and how fast it was moving, and it is hardly surprising that it was termed “the uncertainty principle.”
This concept was challenged by Schrödinger in the now-famous mind-game allegory of a cat confined in an enclosed container that included a slowly decaying radioactive source, and a Geiger counter with a trigger linked to a bottle of poison.
Depending on the rate of radioactive production, the Geiger counter might trigger at any time, causing the bottle to shatter, releasing the poison and, thus, killing the cat.
In this mind game, Schrödinger created the parody of quantum mechanics in which the cat could synchronously be viewed as alive or dead — which was illogical — until someone opened the container to assess the situation, while also disrupting the experiment. In his view, the prevailing view of quantum mechanics was equally nonsensical.
This creates an interesting parallel with the impact of the important observation from Martin and colleagues, in contrast to the many external interpretations of data from the previously published randomized trials of PSA screening.
We should not forget the confusion added by the ever-changing recommendations of the U.S. Preventive Services Task Force and its political masters.
Just to remind you, I have addressed the previously published studies from Europe, Scandinavia and the United States, which have shown variously that PSA screening may reduce deaths from prostate cancer, but after decades of follow-up, these studies still have not shown an OS benefit. In men aged older than 70 years, the European study showed a decrease in OS associated with PSA screening.
The key issue of effective screening programs — such as shown for low-dose CT scanning for lung cancer, mammography for breast cancer, and colonoscopy for colorectal cancer — is to produce both cancer-specific survival and OS increases. Although the PSA screening protagonists continue to twist and turn and change the rules of play, they remain unable to show an OS benefit despite all their rhetoric.
That said, they also seem to ignore the fiscal and physical expense of subjecting older-aged men to the rigors of prostate screening and the downstream sequelae of multiple testing and the associated psychological trauma of testing and treatment. This is particularly egregious as so many of the discovered cases represent low-risk disease, and the protagonists add insult to injury by convincing their hapless victims to undergo “surveillance” for an unending period, purporting to be ready to pounce on the cancer when it becomes “active.”
All of this nonsense remains unproven, and the hype and rhetoric certainly exceed fact and/or economic or physical benefit for most of the victims.
Where does the work of Martin and colleagues fit into this scene? These intrepid investigators studied 189,386 men aged 50 to 69 years who underwent a single PSA screen, and they compared the outcomes to those of 219,439 control subjects randomly assigned to standard (non-screened) practice. These subjects were drawn from British general practices that were randomly allocated to screening or nonscreening protocols.
The proportion of men diagnosed with prostate cancer was higher in the screened cohort (4.3%) than nonscreened group (3.6%), but there was no significant difference in prostate cancer mortality (0.29% in each cohort) nor overall mortality (13% in each group) after median follow-up of 10 years.
The screened population had a significant increase in detection of low-risk cancers — which now seem to find their way into active surveillance programs in North America, albeit of uncertain benefit. Interestingly, the patients offered a PSA screen who chose not to undergo testing had a lower detection rate of prostate cancers than either those who were screened or those in the control group.
There are several important points from this study, which should not be ignored.
Although the follow-up is shorter than the European and PLCO studies, the recruitment from 2001-2009 — being more contemporary — provides access to the more recent advances in prostate cancer detection and treatment methods. The randomization of general practices may have reduced the potential for volunteer bias, although this may have been offset by the impact of rejection of screening by patients within practices in which screening was randomly offered.
Of interest, the distribution of intercurrent diseases — such as diabetes, obesity and cardiovascular disease — was similar within the two general practice groups, thus reducing the risk of inadvertent case selection bias.
The bottom line of this important study is that early, one-time screening did not improve outcomes. It should also be noted that the more intensive prostate screening programs previously described had most of their impact in the first episode of screening, with a greater evolution of overdiagnosis with subsequent screening episodes without concomitant survival increment. Thus, the fact that the British used only a one-time approach is not necessarily a deficit.
Why do I draw a comparison with Schrödinger’s cat?
Schrödinger saw the incongruity of a cat simultaneously being viewed as alive and dead in his model of quantum physics. I view this as absolutely the same as the likely impact of this trial being added to the panoply of extant evidence that really does not support PSA screening.
Nonetheless, these negative data continue to be interpreted by some urologists and public health mavens as supporting the ongoing use of routine prostate screening. It makes no sense to expend large sums of money on a technology which, after decades of use, still does not reliably save lives, but which can cause morbidity, fiscal expense and sometimes mortality in the proposed beneficiaries. The continued support of community-based prostate screening may be an example of the use of real data to create fake news!
Reference:
Martin RM et al. JAMA. 2018;doi:10.1001/jama.2018.0154.
For more information:
Derek Raghavan, MD, PhD, FACP, FRACP, FASCO, is HemOnc Today’s Chief Medical Editor for Oncology. He also is president of Levine Cancer Institute at Carolinas HealthCare System. He can be reached at derek.raghavan@carolinashealthcare.org.
Disclosure: Raghavan reports no relevant financial disclosures.