In a post from December 2015 (“I’m skeptical about … PSA screening”), I borrowed an analogy from Dr. Gilbert Welch, a Dartmouth epidemiologist, who noted that cancers are like animals in a barnyard enclosed by a fence. The goal is to keep them contained. Unfortunately, some cancers will have already metastasized when diagnosed, no matter how early the screening. In this analogy, such cancers are represented by birds, wherein a fence doesn’t contain them, and early screening fails to improve outcomes. At the other end of the spectrum are slow-growing, indolent cancers that are best left alone. These are the docile turtles of the barnyard, for which screening leads to over-diagnosis of cancers never destined to cause harm. That leaves the cancers that are not yet symptomatic but have the potential to become invasive and metastasize. These are rabbits capable of hopping the fence. The goal of screening is to find them. Unfortunately, when viewed through the lens of a screening blood test, rabbits and turtles appear identical. This is the situation we find ourselves when it comes to screening for prostate cancer using a blood test known as PSA (prostate specific antigen). The test diagnoses far more turtles than rabbits because there are far more turtles living in the barnyard. And since the PSA test can’t distinguish between the two, both will be treated. In patients with non-aggressive cancers this results in harm as the side-effects of treatment include urinary incontinence, impotency, GI symptoms, and more. Worse, there are no studies demonstrating an all-cause mortality benefit to PSA screening. In large trials, PSA testing hasn’t even reduced the risk of dying from prostate cancer. Based on this, in 2012 the US Preventative Task Force recommended against PSA screening for prostate cancer in all age groups. There remains a lot of controversy, however, and many primary care physicians and urologists still order the test in their at-risk patients (e.g. African-Americans and/or those with a family history of prostate cancer).
So what should you do if your PSA is elevated and a subsequent biopsy demonstrates localized cancer? Common treatments include radical prostatectomy (surgical removal of the prostate), brachytherapy (radioactive “seeds” implanted into the prostate gland), cryotherapy (freezing), and radiation directed at the prostate from an external source. Adjunctive therapies include androgen deprivation via hormonal manipulation or surgical castration. None of these treatments sound particularly appealing and none of them are. But since we know that most prostate cancers remain localized and ultimately cause no harm, how about withholding aggressive treatment and simply keeping close tabs on the PSA?
In my previous post on this issue, I noted that a trial comparing serial PSA-monitoring versus prostatectomy versus radiotherapy was under review. The results are now available on the New England Journal of Medicine’s website. The ProtecT (Prostate Testing for Cancer and Treatment) trial randomized men with localized prostate cancer into 3 groups: 545 men were monitored, 553 underwent surgery, and 545 received radiotherapy. They were followed for a median duration of 10 years. The primary result was prostate cancer mortality. Secondary outcomes included all-cause mortality, development of metastases, and the rate of disease progression (defined as death due to prostate cancer; metastatic disease; long-term androgen-deprivation therapy; clinical tumor enlargement; rectal fistula, urinary retention due to obstruction, or the need for a permanent catheter). The monitoring process required repeat PSA testing every 3 months for the first year, and every 6 to 12 months thereafter. Monitored patients who demonstrated significant changes in their PSA levels were counseled and given the option to enter active treatment.
The good news is that the death rate from prostate cancer across the board was extremely low, the 10-year cancer survival rate was close to 99%. The results of the trial showed that if you take 3,000 men with non-metastatic prostate cancer, divide them into 3 equal groups, and follow them for 10 years, you can expect 15 prostate cancer deaths in the PSA-monitored group, 9 in the surgery group, and 7 in the radiation group. The small mortality differences noted in the study did not achieve statistical significance, meaning that they could have occurred simply due to chance. In essence, there was no difference in mortality.
Although these results are certainly encouraging, there is more nuance to the question than simply recommending no PSA testing, or only active PSA monitoring when localized cancer is found. Slightly more than half of the men in the monitored group experienced a significant rise in their PSA level at some point during follow up, and nearly all of these went on to receive some form of aggressive treatment. The unanswered question is what would have happened to them if no subsequent therapy had been instituted, since it’s not clear that treatment confers a mortality benefit (i.e. treated patients don’t necessarily live longer).
With regard to the study’s secondary outcomes, it is not surprising that there was more disease progression in the monitored group than in the surgery or radiation groups. After all, no treatment was undertaken against the cancer in that group. For the record, 20% of those patients experienced cancer progression, while just 8.6% of the surgical patients and 8.4% of the radiation patients progressed. The likelihood of metastases in the monitored group was higher as well; 6% in the monitored group, 2.3% in the surgical group, and 2.9% in the radiation group. Although this sounds like aggressive treatment might be the way to go, I’m not so sure given that the treatments didn’t impact mortality.
Don’t want to take chances? Okay, but before deciding, you need to consider the downside to treatment. In a separate paper from the same study group, the authors monitored patient symptoms over time. Here, the patients treated with surgery or radiation fared much worse than the monitored group. After 6 months, the percent of men experiencing urinary incontinence requiring the use of pads was 46% in the surgery group, but only 5% in the radiation group, and 4% in the monitored group. And while the surgery group saw some improvement over time, they never returned to baseline and continued to fare far worse than the others with respect to continence.
As to erectile function, both surgery and radiation groups fared poorly. At baseline, two-thirds of the men were able to generate erections adequate for intercourse, but by 6 months that number had fallen to just 12% in the surgery group, and 22% in the radiation group. The number also fell to 52% in the monitored group, suggesting that the diagnosis of prostate cancer, in-and-of-itself, makes it difficult for men to maintain an erection. Over time, performance improved slightly in both the surgery and radiation groups while it continued to decline in the monitored group. By 6 years out, fewer than a third of the men in any group were able to generate an erection sufficient for intercourse. Prostate cancer is bad news for sex.
Overall, surgical prostatectomy patients had more incontinence and impotency, while radiation patients experienced more bowel problems, urinary voiding issues, and nighttime urination. Monitored patients had more progression to metastatic disease but with a negligible effect on mortality. Lastly, the authors estimate that 27 men would need to be treated with prostatectomy and 33 with radiation versus an active monitoring strategy to avoid 1 patient developing metastatic disease. This means that most men undergoing aggressive treatment don’t benefit from it. A total of 9 men would need to be treated with either prostatectomy or radiation to avoid 1 patient having clinical progression of his cancer. Meanwhile, with respect to harm, the authors conclude that after 2 years, treating just 4 men with prostatectomy or 8 men with radiotherapy would cause 1 additional case of erectile dysfunction. Treating 5 with prostatectomy would add an additional patient with urinary incontinence. To put it another way, prostatectomy is more than 6-times as likely to cause incontinence and more than 5-times as likely to cause erectile dysfunction than it is to prevent metastases. The results of this trial demonstrate that neither surgery nor radiation extends the quantity of life compared to active monitoring, while both forms of treatment negatively impacted the quality of life. Given this, the choice seems clear.
- C. Handy et al., “10-Year Outcomes after Monitoring, Surgery, or Radiotherapy for Localized Prostate Cancer,” NEJM 2016; DOI: 10.1056/NEJMoa1606220.
- L. Donovan et al., “Patient-Reported Outcomes after Monitoring, Surgery, or Radiotherapy for Prostate Cancer,” NEJM 2016; DOI: 10.1056/NEJMoa1606221.