Gary Strauss

MD MPH

tufts medical center

Professor of Medicine;Director of Lung Cancer Program; Asso. Director of Hem/onc Fellowship Program

PLCO (Learn more about this study)

PLCO-231

Sep 30, 2016

re-analyze PSA screening data in PLCO trial

We are requesting data from the prostate-cancer screening component of the PLCO (Prostate-Lung-Colon-Ovary) trial. In May 2012, the US Preventive Services Task Force (USPSTF) recommended against prostate-specific-antigen (PSA)–based screening for prostate-cancer among all men. The recommendation was largely based upon the results of PLCO, which was published in 2009 and failed to demonstrate a significant reduction in prostate-cancer mortality among men aged 55-74. Indeed, the mortality rate ratio among those randomized to PSA-screening was 13% higher in the experimental group.

There has been intense controversy since publication of the results of PLCO in 2009. Results of PLCO contrasted with the results of the European-Randomized-Study-of-Screening-for-Prostate-Cancer (ERSPC), which did report a significant mortality reduction in the experimental group randomized to PSA-screening.

One of the major concerns with regard to the validity of PLCO results relates to the level of contamination in the control arm. Recently, an analysis was published demonstrating that the proportion of control participants who reported having undergone at least one PSA test before or during the trial was close to 90%, higher than the originally reported contamination rate of 50%, and even higher than in the experimental group (85%).
Indeed, the PLCO trial was initiated the year after the American-Cancer-Society recommended that informed men over the age of 50 years should be offered annual PSA screening in 1992. Evidence supports that this recommendation was widely accepted by the public at the time. All 76,693 men participating in PLCO were recruited between 1993 and 2001.

A fundamental hypothesis of our project is that analysis PLCO data would support the conclusion that PSA-screening does save lives, if adjustments were made for the extent of PSA-screening in the control group.
Prostate cancer mortality was considered to represent the primary endpoint in PLCO. However, under specific circumstances, survival can provide an accurate measure of outcome in randomized population trials. This depends upon the ability to exclude or adjust for the magnitude of confounding due to lead-time bias, length bias, and overdiagnosis. A prostate-cancer specific survival analysis was published demonstrating no significant difference in prostate-specific survival rates between the experimental and control groups, with 10-year survival 94.7% versus 93.5% respectively. However, this analysis made no attempt to adjust for contamination in the control arm.

The occurrence of metastatic disease is a very important contributor to the suffering caused by prostate cancer. In the ERSPC study, PSA screening significantly reduced the risk of developing metastatic prostate-cancer. There was a relative reduction of 30% in the intention-to-screen analysis and a 42% (p = 0.0001) reduction for men who were actually screened.

We propose to conduct an analysis of PSA screening efficacy while adjusting for PSA screening in control participants prior to entering the trial and during active screening period. We will assess prostate cancer-specific mortality. Moreover, we will evaluate prostate cancer survival, progression-free of metastasis and prostate cancer mortality. We also want to evaluate the extent of overdiagnosis and reevaluate the extent of contamination in the control arm.

1. To compare prostate cancer mortality, prostate cancer-specific survival, and survival free of metastatic disease in experimental and control arms in the PLCO trial, while adjusting for prior PSA screening both before and during the PLCO experimental period. Prostate cancer cases will be stratified into low-risk (Gleason 2-6, clinical stage T1c (non-palpable) and PSA<10 ng/ml), intermediate risk (Clinical tumor stage T2b or T2c, Gleason score of 7 or PSA level of 10–20 ng/mL), and high-risk (Gleason 8-10, PSA>20 ng/ml or clinical stage T3). The specific hypotheses that will be investigated regarding this specific aim include the following:
a) PSA screening improves outcomes for men who develop high-risk prostate cancer, defined as Gleason 8-10; or PSA>20 ng/ml, or clinical stage T3 disease (palpable and bulky).
b) PSA screening may not improve outcome among men who develop low-risk prostate cancer, which will be defined for purposes of this analysis of Gleason 2-6; or clinical stage T1c (non-palpable), or PSA<10 ng/ml
c) Very high level of contamination in the PLCO control group largely accounts for the negative published results with regard to both disease-specific mortality and disease-specific survival.

2. We will explore the relative ability of mortality and survival endpoints to accurate measure PSA screening efficacy, after adjusting for level of contamination in the control.

3. Assess the extent of overdiagnosis, stratified by risk group. The methods used for this analysis will be to correlate and define the extent to which prostate cancer treatment effects disease-specific mortality, survival, and freedom from metastatic disease among those diagnosed with prostate cancer in both the experimental and control arms of the study. The specific hypotheses that will be investigated regarding this specific aim include the following:
a) There is no substantial overdiagnosis among men with high-risk disease.
b) There is substantial overdiagnosis among men with low-risk prostate cancer.

4. Re-evaluate the contamination rate in the control arm, differentiating screening PSA versus early symptom-driven PSA testing.