Expert Sheds Light on Prostate Cancer Screening Trial

E. David Crawford, MD

The European Randomized Study of Screening for Prostate Cancer (ERSPC) and United States Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) showed varying results in mortality when investigating screening for men with prostate cancer.

These randomized, controlled trials evaluated the effects of screening versus no screening on mortality. ERSPC found that screening reduced prostate cancer mortality, whereas the PLCO trial found no reduction.

However, a recent analysis published in the Annals of Internal Medicine concluded through statistical models accounting for differences in implementation and settings that ERSPC and PLCO offer compatible evidence that screening lowers prostate cancer mortality.

OncLive: Can you please provide some background of the PLCO study?

In an interview with OncLive, E. David Crawford, MD, a professor of radiation oncology at the University of Colorado, discusses the controversy surrounding the PLCO trial, as well as the role of prostate-specific antigen (PSA) testing for men with prostate cancer.Crawford: We initiated the 2 big screening trials with Europe and the United States. There were 10 centers involved in the United States screening trial. It started out looking at screening for prostate cancer but got held up by the National Institutes of Health because there was an interest in adding other cancers to this screening trial. It took 3 years from the time we planned on starting this in prostate cancer for it to be launched as a PLCO trial in prostate, lung, colorectal, and ovarian cancer.

Unfortunately, in the interim, men were getting screened with the PSA test in the United States. The population of unscreened men did not exist like it did a few years prior. Also, during the trial, when people were randomized to the usual care, which did not involve screening, they would go out and get screened. The contamination became an issue in the United States trial. We recognized that and accounted for it in the statistical analysis—accepting a 25% to 30% contamination.

The contamination got worse, so the control arm contamination was around 50% to 90% over time. It looked at the binary statistics of doing a study like this. It was screen versus no screening, but since there were contaminations in the no-screen arm, many felt that this trial lost its power. What this study did was look at it in a different way.

The European trial was not faced with that dilemma of screening since it was not accepted at that time. They had a population that was more urgent and did not enter screening off the protocol. They essentially had a purer no-screening arm. When these studies were originally reported at 8 years, there was no difference. However, with 10 to 15 years, the European trial showed a difference of 25% to 30% improvement in mortality.

Our trial was sort of trashed in a way by many people because of the contamination. It’s a negative trial and did not seem to be a benefit for people. Many write it off as a waste of money, time, and effort, which is both true and not true. Myself and Dr Anthony D’Amico tried to salvage this study 7 to 8 years ago. We looked at the PLCO trial and the performance status for patients who entered the trial quite healthy and we saw a benefit.

A group got together—including myself and some of the others who were on the original trial— to look at a promising new way to analyze the data. We investigated the intensity of screening, not just an occasional PSA as an endpoint. We tried to look at the PLCO trial and reconcile it to the point where it was compared with the European trial—showing that there is more of a benefit than baseline intensity. It is important to do longer follow-up to see a benefit for prostate cancer screening. Now, there is a benefit of 10, 15, and 20 years and a 25% to 30% improvement in survival rates.

Some screening is good. We see it with the mortality rates when we look at the number of metastatic disease [cases] that occurred back in 1989 when we started screening. The incidence of metastatic disease dropped precipitously; whereas, in breast cancer over the years, there wasn’t any change in metastatic disease. In prostate cancer, we saw the drop in metastatic disease and, therefore, over enough time, we saw an improvement in survival rate. Data in the US Preventive Services Task Force (USPSTF) are focusing on the downside, which was right. There was overdiagnosis and overtreatment. However, right now, these data couldn’t come at a better time for prostate cancer.

What we need to do today is to screen smarter. We need to evaluate people who have a risk of aggressive cancer and we have the tools to do that right now. Currently, 90% of PSAs are ordered by family practitioners. We believe the PSA cutoff is 1.5 and there are data to support that. A PSA above 1.5 is the danger zone, although it doesn’t necessarily mean you have prostate cancer—but you’re in the risk category.

What do you believe the next steps are for screening of prostate cancer?

At that point, another test is needed to determine who has prostate cancer. This resonates nicely with family practitioners. For example, if a patient has an abnormal cardiogram, they do not get rushed off to receive a bypass, they get another test. This is a newer statistical way of looking at things. It is no longer “black or white.” This trial is attempting to add further detail.There are currently questions of which treatment is best and what is the age limit to screen. There will always be criticism for this study, but an area to focus on is that early detection helps.

We looked at ovarian cancer in the PLCO trial and early detection did not show an improvement. However, in prostate cancer, we believe that early detection does matter. We have good tools to screen smarter. Just because a patient has a low-grade cancer, that does not mean that the patient needs a radical prostatectomy or radiation therapy. We want to find cures.

What are the main takeaways from this study?

There is a new urine test called SelectMDx that looks for 2 genes that are overexpressed with aggressive prostate cancer. Regarding PSA, the USPSTF needs to reassess this.We can say that this study supports the concept of smart screening and early detection. This should be offered to men and not denied. As we go forward with this data, we need to have a simple message to family practice doctors. We have given them so many mixed messages over the years regarding screening, such as the cutoff stage, PSA density, PSA velocity, and so forth. The way forward is a PSA cutoff of 1.5, and above 1.5 is a gray zone that needs to be further evaluated. With this information and newer statistical analysis, we are on the path to make a lot of progress in prostate cancer and early detection.

Tsodikov A, Gulati R, Heijnsdijk EAM, et al. Reconciling the effects of screening on prostate cancer mortality in the ERSPC and PLCO trials. Ann Inter Med. 2017;167(7):449-455. doi: 10.7326/M16-2586.