Biden Has Prostate Cancer. Should You Get Tested?

One of the trickiest topics in epidemiology.

Pictured: Not a prostate screening centre. I think. Photo by Graham Ruttan on Unsplash

Joe Biden has just been diagnosed with prostate cancer. Many people are now wondering if they should also get tested for the disease, which is very treatable if you start early. You can also test for prostate cancer with a very simple blood test, so it seems like a no brainer to go and get some pathology done.

Here’s why the decision is much more complicated than it might first appear.

Screening Is Hard

In the world of epidemiology, there are a few major fights over evidence. There’s the question of whether countries should mandate bicycle helmets, or the discussion about the benefits of breastfeeding, or the endless battle over whether vaping is the newest way to get kids hooked on cigarettes or a fantastic strategy to help adults quit.

One of the longest-running and hardest to answer questions is pretty simple: who should be screened for cancer? It seems like a very easy thing to answer, and most people would automatically just say “everyone at risk”. If we catch cancer early it is usually far more treatable, and screening is often quite low-cost, so it feels like a win-win to just test every person who could possibly have a case of cancer for the disease.

There are free stock photos of mammography from the National Cancer Institute, which is pretty cool. Photo by National Cancer Institute on Unsplash

The problem is that cancer screening is not entirely benign. There are risks - however small - to things like x-rays, and many cancers require a biopsy to properly diagnose the disease. Biopsies are pretty safe, but they do involve cutting out small pieces of tissue, which is never risk-free. There’s also the psychological impact of having a positive test for cancer, even if it later turns out to be a false alarm. For the people you catch with early cancer, the screening can be amazing, but it can cause a lot of harm for the people who have a false positive test.

There’s a really famous example used to illustrate the issue here. Women aged 40-59 have a rate of breast cancer of about 1 per 1,000. Let’s say you have a test that correctly identifies 99% of women with breast cancer as having the disease - i.e. almost no false negatives - and has a 95% success rate at correctly showing that a person doesn’t have cancer.

If an average 48 year old woman tests positive for breast cancer using this test, what is the likelihood that she really has the disease?

Most people answer either 99% or 95%. These are the intuitive responses, because they seem to show how effective the test is at finding cancer. When you tell them that those two are wrong, you often get guesses of 90%, or perhaps 80%. People tend to be shocked that the correct answer is just 2%. In a famous experiment, most doctors asked this question got it wrong too!

The math here is quite simple. Let’s say you test 1,000 women. Of those, you correctly find the 1 person with cancer. But you also test 999 women who don’t have cancer, and your test incorrectly shows that they have cancer 5% of the time. That means that there are 49 or 50 false positives in your group of women. So the chances that any single person has cancer in the pool of people who tested positive is 1/50 = 0.02 or 2%.

If we were talking about false positive diabetes tests, this wouldn’t really matter. Diabetes is diagnosed exclusively through blood tests, and the main harm is having to go in and do another test. It’s unpleasant and a bit stressful, but not a huge problem.

But for cancer, you need further scans, and sometimes biopsies. These additional investigations carry a low risk, but it’s easy to see that if 98% of the people who go on to have extra tests don’t have cancer you’re going to create quite a bit of harm for very little benefit.

And that’s why population mammography isn’t recommended for breast cancer screening in younger women. We do test people who are high risk - you might’ve been told to get a mammogram at age 30 if you have a family history of breast cancer, for example - but it’s generally inadvisable to test everyone just to check.

What About Prostates?

This brings us to the question of prostate cancer. As someone with a prostate, it’s definitely of great interest to me.

I asked for a picture of a prostate, and Unsplash delivered. Hurray? Photo by Europeana on Unsplash

There are a few relatively unique things about prostate cancer. Firstly, it’s almost always very slow to progress. The average person diagnosed with prostate cancer is only about 5% more likely to die over 15 years than someone without the disease. There’s some data showing that, even without treatment, very few people with prostate cancer die because of it unless they have very aggressive forms of the disease.

Prostate cancer treatment also sucks. Aside from chemotherapy, the main treatment for is to remove the prostate. This has consequences for continence, sexual function, and more.

On the other hand, it’s very easy to screen for prostate cancer. You can do it with a simple blood test that looks at something called Prostate Specific Antigen (PSA). While the test isn’t perfect, it’s quite good at picking up people who likely have cancer, and it can easily be rolled out to massive populations.

All of this leads to a difficult question - should we screen for prostate cancer in men? Fortunately, we’ve been asking this question for a long time, and there are a number of huge clinical trials looking at it. We have gold-standard evidence which shows definitively that the answer to the prostate cancer screening question is: maybe.

What the data shows is that prostate cancer screening across large populations of men probably reduces their risk of dying from prostate cancer by a bit, and may reduce their risk of dying overall by a very small fraction. But the benefit reduces as men get older, because other things start to kill them more quickly than prostate cancer. In addition, for every death from prostate cancer that’s prevented, 1 person will go to hospital for sepsis, 3 will become incontinent, and 25 will experience some measure of sexual dysfunction.

This is all made more complex by the fact that the findings are very uncertain. There is some debate in the epidemiological community about whether prostate cancer screening is useful at all, because some of the randomized clinical trials do not show a benefit for sending men off to be screened at any age.

All of this is why the CDC only recommends PSA screening for men aged 55-69. The UK National Health Service, however, does not recommend regular PSA screening for people of any age. The only reason that the UK tests people is if they specifically ask for a test or if they have symptoms that are in line with prostate cancer. If you want to read about the extensive controversy the recent Australian guidelines on PSA testing are a great introduction into the topic - they have dozens of pages looking at the different clinical trials and eventually conclude that if people ask for it then PSA testing is fine every 2 years between 55 and 69 after lengthy discussion with their clinician about the risks and benefits.

Joe Biden is a case in point of all this complexity. He has an aggressive form of cancer that is amenable to treatment with hormone therapy. On average, fewer than 10% of men diagnosed with this sort of disease die from prostate cancer over the ensuing 15 years. 80% of them die in that period, but mostly from things that are not cancer. However, Biden has secondary metastases - specifically, cancer in his bones - which makes the prognosis quite a lot worse. The average man of his age would probably die from something other than cancer, but for Biden the outcomes are not as good.

If you have a prostate and are aged 55-69, you might want to talk to your doctor about whether you should be getting regular PSA testing. The evidence is hotly-debated, but there’s likely some benefits to getting the tests*. If you don’t fit into that age group, and have no other factors placing you at high risk, then there’s probably no reason to go and get tested.

*Note: I have barely touched on the controversy, and this sentence will undoubtedly get me angry emails from PSA testing opponents. I think the argument in the Australian guidelines is reasonable, and that while it’s possible that some screening programs are ineffective, it’s quite likely that testing men in this age group does reduce the death rate from prostate cancer and the overall mortality by a small but not insignificant amount. I have also simplified the endless argument about mammography down, because again the debate is…lengthy.