publication date: Feb. 1, 2019
The Prostate Cancer Prevention Trial and nine lessons in prostate cancer medicine
Otis W. Brawley
Bloomberg Distinguished Professor of Oncology and Epidemiology
Johns Hopkins University
The publication by Ian M. Thompson and colleagues in last week’s New England Journal of Medicine regarding long term follow-up of patients in the NCI-sponsored Prostate Cancer Prevention Trial (PCPT) marks a good opportunity to review and reflect on the history of the trial and the past 30 years of prostate cancer medicine.
George Santayana is oft quoted as saying that “those who do not remember the past are condemned to repeat it.” There are many lessons here that should influence our approach to modern health care issues in general and prostate cancer specifically, such that we do not repeat the harm that had been caused by closed-mindedness and ignorance around prostate cancer.In the 1990’s, prostate cancer was (and indeed still is) a significant cause of human suffering. The nineties was a time of prostate cancer hysteria. Society was just starting to talk openly about cancer. Studies indicated prostate specific antigen (PSA) screening could find prostate cancer early. Many public figures were willing to publicly say that they had prostate cancer and PSA screening saved their lives. Many felt an obligation to be an example to save other men. They preached decades-old messages of cancer early detection.
By 1992, patient advocacy was becoming increasingly prominent in AIDS and women’s health. Advocacy moved into the prostate cancer arena with formation of a group called Us TOO. The group advocated for prostate awareness, screening and treatment. Their prejudice toward screening and treatment is understandable due to fear of cancer and the years of messaging regarding early detection.
In the US, hospitals and clinics responded to this hysteria by putting prostate cancer screening and treatment into their business plans. Mass screening in malls and community centers was being advertised. While most thought that they were doing a public service, the fact that screening and treatment were lucrative helped keep healthcare providers from questioning their health benefits. It is fascinating that the introduction and use of PSA was more rational in Europe.
The importance of clinical trials
In early 1990’s prostate cancer medicine, few prostate clinical studies had been attempted. There were studies showing that PSA could find localized cancer. No study had been conducted to show that PSA screening saved lives. Indeed, no study had been conducted to show that treatment of localized disease saved lives.
PSA was, and to this day is, FDA-approved for diagnosis of disease in men for whom there is a clinical suspicion of prostate cancer and for following progression of diagnosed disease. It was not—and still is not—FDA-approved for screening.
Prospective randomized trials take a long time. Some felt a sense of urgency, arguing against doing the appropriate clinical trials and staying the course, i.e. continuing to screen and treat, because people are dying.
Truth is, bad or inappropriate medicine can also kill. Medicine’s inability to predict effectiveness is fraught with interventions that were thought beneficial, but really caused harm. Fortunately, over the past two decades, well-designed trials definitively demonstrated that:
Treatment of prostate cancer can prolong survival in 1997. The first study to show a treatment benefit demonstrated that radiation therapy and hormones was better than radiation alone for locally advanced disease. (1)
Radical prostatectomy for localized disease prevents death in 2002. (2)
PSA screening combined with treatment saves lives in 2009. (3)
The majority of men with screened detected prostate cancer do not need treatment in 2016. (4) One screening study, the NCI The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial, was crippled by the American screening frenzy. (5)
It was hard to convince a man randomized to the control group to not get screening when all of his friends were doing it. Many Americans felt that it was unethical to do a study which required a group randomized to a non-screening arm.
I was involved in the launch of PCPT in 1992. The study was severely criticized by some in the medical community. While some criticism or skepticism is healthy and always legitimate in medicine, the loudest critics of PCPT spoke without decency and often misrepresented the science while denigrating me and the trial. They indicated ignorance of the science. For some, the concept of prevention was neither understood nor valued. They attacked the study because it took away from the “screen, screen, treat!!!” message.
The tendency of “experts” to be extremely vocal often demonstrates ignorance of the science. I have seen it in many cancer screening arguments. As I recall the often impolite, sometimes ad hominem comments from PCPT critics, I am reminded of the Abraham Lincoln quote: “Better to remain silent and be thought a fool than to speak and to remove all doubt.”
Lesson 1: Some so called “expert opinion leaders” are not very expert.
Interestingly, the lay public was very much interested in prevention and frightened of prostate cancer. The fast accrual to PCPT was evidence of this. The trial participants were overwhelmingly educated and middle class. Even a high proportion of the racial minorities on the trial had a graduate degree.
Lesson 2: Most men will never be diagnosed with prostate cancer.
PCPT was one of the first trials where a special emphasis was placed on accruing minorities and the traditionally underserved. As the trial went on, we learned that poor people are extremely rational. They do not have time to get involved in a study to assess prevention of a disease they are not likely to get.
The prevention findings
After more than a decade, the study found that 7 years of 5-alpha reductase therapy with finasteride lowers the period prevalence of prostate cancer by 25 percent. (6) It found that 24.4 percent of men on the placebo arm were diagnosed with prostate cancer, compared to 18.4 percent on the treatment arm. PCPT also showed that men taking finasteride and diagnosed with prostate cancer had a higher likelihood of having high grade prostate cancer. This led to the concern that finasteride causes high grade prostate cancer.
A second 5-alpha reductase inhibitor dutasteride was marketed. It was tested in a prospective randomized placebo-controlled prevention study, and it too reduced prostate cancer prevalence with an increase in the proportion of high-grade cancers in those diagnosed while on 5-alpha reductase therapy.(7)
A number of studies pointed away from 5-alpha reductase therapy causing high grade prostate cancer and demonstrated that it improves ability of PSA screening to find high-grade cancers. Despite these data, the FDA required a warning label that finasteride and dutasteride might increase the risk of high-grade prostate cancer. It is unfortunate that the FDA action came after an Oncologic Drug Advisory Committee (ODAC) meeting in which some of the PCPT critics would not allow presentation and consideration of all of the evidence.
Lesson 3: It is importance that consensus panels be open-minded and consider all the literature.
The warning led to lawsuits, as money-hungry attorneys sought prostate cancer patients who had taken these drugs. This scared drug companies from interest in prostate cancer risk reduction.
Lesson 4: Long-term use of 5-alpha reductase inhibitors reduces risk of prostate cancer and may reduce risk of prostate death.
Lesson 5: Treatment with 5-alpha reductase inhibitors improves the operating characteristics of PSA screening.
The long-term follow-up of participants in the PCPT is the clearest and best evidence that finasteride therapy reduces risk of prostate cancer and that it’s safe. It suggests that the effects of finasteride go well beyond the seven-year period of therapy and it may even reduce risk of prostate death. The finding is open to some legitimate skepticism, as follow-up through the National Death Index is not as good as clinical follow-up. Unfortunately, long-term follow-up of NCI clinical trial participants is going away due to budget constraints.
Despite this one weakness, I believe the evidence is sufficient to warrant removal of the FDA warning placed on 5-alpha reductase inhibitors, and the evidence now supports an indication for prostate cancer risk reduction. The 5-alpha reductase inhibitors are now generic. No drug company has a financial interest in obtaining an indication for prostate cancer risk reduction using finasteride or dutasteride. This is unfortunate for the public health.Perhaps the National Cancer Institute, which started this body of work nearly 30 years ago, will complete it by filing for the indication.
Prostate cancer screening
The PCPT taught a number of lessons about prostate cancer screening. The placebo arm is one of the best-controlled PSA screened cohorts ever. The study enrolled men with a PSA of 3 ng/ml or less. A reasonable criticism was that the study was trying to prevent prostate cancer in men at low risk of prostate cancer. It was a big surprise that prostate cancer was diagnosed in 24.4 percent of these “low risk” men.
Lesson 6: Prostate cancer is extremely common in healthy men. We have the technology to diagnose it in at least one in four men in their sixties.
Of men diagnosed with prostate cancer on the placebo arm, half were diagnosed through PSA screening and half were diagnosed through a biopsy that was called for in all men who had normal PSA screening over the seven years of the trial.
Lesson 7: PSA screening of “low-risk” men can find a lot of prostate cancer and miss a lot of cancer.
Indeed, PSA screening missed as much prostate cancer as it found. Prostate cancer can be diagnosed in men with very low PSA level.
Lesson 8: Most people think every man diagnosed with prostate cancer will die of it if not treated. Fact is, most men who are diagnosed with prostate cancer do not die of it and do not need treatment.
PCPT showed that 25 percent of men can be diagnosed (half through PSA screening in a seven-year period). The NCI SEER studies suggest that less than 3 percent of men die of prostate cancer. The Schroeder screening study suggests that lifetime risk can be reduced from about 3 percent to about 2.4 percent.
Lesson 9: Half of all men diagnosed with prostate cancer through screening have a Gleason 6 prostate cancer. A third of those who died of prostate cancer in PCPT initially had a Gleason 6 cancer.
This demonstrates the need for careful observation of those undergoing surveillance and the need for better laboratory tests to predict significant disease.
We have come a long way in our understanding of prostate cancer since the beginning of PCPT and the publication of the other large prostate cancer trials. We have increased appreciation for the need for rigorous assessment of the literature. We also have increased appreciation of the limits of our knowledge.Most screening recommendations involve informed decision-making. Perhaps the American Society of Clinical Oncology (ASCO) decision aid best explains what is known about prostate cancer screening and treatment.(9)Out of 1,000 men age 55 who choose to be screened to age 70, about 96 will be diagnosed with prostate cancer and 4 will ultimately die of the disease.
Out of 1,000 men age 55 who choose not to be screened to age 70, about 60 will be diagnosed with prostate cancer and 5 will ultimately die of the disease.
The 4 deaths per 1,000 vs. 5 deaths per 1000 represents the 20 percent reduction in relative risk of death.
The high number of men who are diagnosed and do not die of prostate cancer indicates that a substantial number of men do not benefit from treatment and again show that the need to be able to identify the men for whom observation is best and the 1 in 1,000 men who will benefit from treatment of localized disease.
In April 2017, the American Urologic Association (AUA), the American Society of Therapeutic Radiation Oncology (ASTRO) and the Society of Urologic Oncology (SUO) released a joint evidence-based practice guideline on clinically localized prostate cancer. It was later endorsed by the American Society of Clinical Oncology. The guideline stresses the importance of involving the patient in shared decision-making regarding prostate cancer screening and treatment. (9)
It also stressed the importance of evaluating a cancer for risk of clinical significance and says that active surveillance is the best initial care option for most low-risk localized prostate cancer.
Pilepich MV, Caplan R, Byhardt RW, et al. Phase III trial of androgen suppression using goserelin in unfavorable prognosis carcinoma of the prostate treated with definitive radiotherapy: report of Radiation Therapy Oncology Group Protocol 85-31. J Clin Oncol. 1997; 15(3) 1013-21
Holmberg L, Bill-Axelson A, Helgesen F, et al. A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer. NEJM 2002 347(11): 781-189.
Schroder FJ, Hugosson J, Roobol MJ, et al. Screening and prostate cancer mortality in a randomized European study. NEJM 2009; 360: 1320-1328.
Hamdy FC, Donovan JL, Lane JA, et al. 10 Year outcomes after monitoring, surgery, or radiotherapy for localized prostate cancer. NEJM 2016; 375: 1415-1424.
Andriole GL, Crawford ED, Grubb RL, et al. Mortality results from a randomized prostate cancer screening trial. NEJM 2009; 360:1310-1319.
Thompson IM, Goodman PJ, Tangen CM, et al. The Influence of Finasteride on the Developmet of Prostate Cancer. NEJM 2003; 349: 215-224.
Andriole GL, Bostwick DG, Brawley OW, et al. Effect of Dutasteride on the risk of prostate of cancer. NEJM 2010; 362: 1192-1202.