I'm so sorry to hear that you're in this stress-producing situation. My 86 year-old mother, as she was instructed to do, went faithfully for her annual mammograms. This year, they finally found something in the right breast. Ultrasonically-guided needle biopsy showed a well-differentiated non-invasive cancer. Simple mastectomy was recommended. Well, the bright old girl said, "I'd rather have a flat chest that one breast, and I'll eliminate the chance of this ever happening again." Mammogram and MRI of the left breast was normal. She talked the surgeon into bilateral mastectomy. Histologic exam confirmed the one known cancer in the right breast, and *three others* on the left. The surgery was far more threatening to her survival than the cancers, and *she should never have known about them anyway*. Had she done what the tests indicated were necessary, she would have retained 75% of her cancer load.

If you haven't done so, read the first comment in this thread, especially the links. Your comment "known cancer causes a "high" PSA nearly 80% of the time" has no clinical practicality: if you know you have cancer, you don't need a PSA test, unless you plan some unproven program of following it along until it reaches the "interventional" level. You mention better and better blood tests: like my mother's mammograms, sooner or later some test will confirm your prostate cancer, if you keep at it long enough. As the first thread comment confirms, a 59 (or 54) year-old man with a normal PSA has only 6% less chance of having prostate CA than one with a PSA of 4.1-10.

There is *almost* no evidence that metastatic prostate CA patients have greater survival if their treatment includes removal of the primary (the prostate gland), rather than waiting and treating the metastatic symptoms. *And*, the younger men are more likely to have the aggressive, early-metastasizing cell lines, which are usually metastatic when diagnosed.

I have to hope that you did *not* read the first comment, or that if you did, you have good reason to reject it. The stress to which you are subjecting yourself is *known* to decrease suvival (MI's, etc), while the aggressive treatment of prostate CA is not known to increase survival.

Sadly, you are in the position described by Eliezer Yudkowsky (I've substituted "ignorance" for the original "stupidity"): "The happiness of ignorance is closed to you.  You will never have it short of actual brain damage, and maybe not even then.  You should wonder, I think, whether the happiness of ignorance is optimal - if it is the most happiness that a human can aspire to - but it matters not.  That way is closed to you, if it was ever open." And it was.

Expand full comment

A lot of public-health medicine is based on statistics, but the sample size I care about is n=1.

The fact is, PSA is a test of limited value because it is not very specific; i.e., a "high" number correlates with presence of cancer < 20% of the time. It is, though, fairly sensitive: known cancer causes a "high" PSA nearly 80% of the time.

So here I am at 59, five years after my initial PSA came in at 7.0. Negative biopsy, and it's STILL 7.0. So either I haven't got cancer, or the biopsy missed it and it's not very aggressive. Sure wish I knew which, as I plan to live past 90.

Better tests are available. The PCA-3 is supposed to be 97% specific, so it appeals to people in my situation. Not approved in the US but available at about $450. I just read that Austria started using it regularly this summer.

Another group is using PCA-3 and five other DNA markers; it's supposed to be even better.

And a big study of EPCA-2 is coming up.

I think the focus ought to be on "which test to screen with at what age, and in the presence of what results from earlier in life".

Expand full comment

From the WaPo link:

"It remains unclear whether the screening is worthwhile for younger men, a federal task force concluded yesterday."

And today I saw:

"New research suggests that roughly one in five American men in their 40s has had a blood test to screen for prostate cancer within the last year."


Ok, so my back of the envelope math suggests that last year the US thus wasted at the low end some US$144,486,241 on this one test alone. At the high end, possibly US$825,635,600.

Whoa. Even considering that some estimate the US spent about US$2,214,400,000,000 on health care in 2004 (Can that actually be right? Do we actually now spend more on health care than food?), that's serious cash.

I must be making a mistake, can it really be that much?

Expand full comment

It might helpful to read the actual guidelines -- which actually isn't a major change from the old ones. Patients with a life expectancy of less than 10 years probably don't benefit from prostate cancer screening because it is unlikely that prostate ca. is what will kill them. What is new in the guideline is that "less than 10 year life expectancy is hard to figure out", so AHRQ made it simpler, and just said no point in screening men over 75.

Hal, if you are young (i.e. not a senior citizen) -- I don't think its prudent to ignore screening (esp. if you have symptoms). _Most_ cancers are curable if caught early -- with prostate cancer being the most controversial in terms of treatment efficacy because so many men will develop it when they get older and it will not be what kills them. A helpful way to tease out how beneficial an intervention is going to be is to ask a doctor in a straight-forward manner, "how likely is this going to be what kills me?"

Expand full comment

I suspect this is just an underpowered study. Here's a different study from Austria showing not only did prostate cancer deaths plummet with the introduction of aggressive prostate screening, they fell much faster than in regions of Austria which didn't have aggressive screening. http://www.sciencedaily.com... All-cause mortality is generally almost useless unless you're dealing with heart disease, because the effect you're looking for (reduced prostate mortality here) disappears in the noise of all the other causes of death. In addition, there's a major confounding factor in that patients who get treated are more likely to have dangerous cancers - they're a higher risk group than watchful waiting. So the Swedish study is saying a higher-risk group that received treatment doesn't do measurably better than a low risk group, using a poor measurement system. Pretty flimsy stuff to stake thousands of lives on.

A 7% annual drop in death rates is far too much to just be happening. A plausible alternative hypothesis for reduced prostate cancer deaths has to include a model of why this is happening, and it has to be a testable model.

Expand full comment

I wonder when the experts are going to really figure this issue out. Test, don't test, test, don't test. I'm still younger, so I'll get screened in a few years. I'd rather know if I have a choice.

We wrote a post trying to explain this today at http://current.pic.tv/2008/...

Expand full comment

I had a run-in with a high PSA count a few years ago - had to have several prostate biopsies, quite painful and unpleasant. Luckily it came to nothing, apparently just a case of local inflammation, and my PSA is back down now. This new information will certainly make me less likely to accept medical treatment if it goes back up.

Expand full comment

retired, thanks for the expert commentary.

Expand full comment

The statement that prostate cancer death rates have plummeted in many countries following the institution of widespread PSA screening looks susceptible to the post hoc fallacy.

Expand full comment

Screening for prostate cancer is a consumer-driven activity, rather than a science-driven one. When PSA testing was developed, both doctors and the public believed that low levels meant absence of CA, and high levels meant presence of CA. In reality, when sextant biopsy is performed on men with either PSA of 0-4, or PSA 4.1-10, the incidence of cancer is only 6% less in the first group (which is supposed to be normal) than in the second (which is supposed to have CA). In addition, in men in their fifties with 0-4 PSA and a positive biopsy, the cellular morphology tends to be worse, and many already have metastasized (cannot be cured). In autopsy studies of men in their 80's and older, having died of something other than prostate CA, over 80% were found to have microscopic evidence of prostate CA. In other words, if you live long enough, you'll have it. Yes, there are 25,000+ deaths annually in the USA from prostate CA, but the chance of dying WITH prostate CA rather than FROM it is orders of magnitude greater. By definition, those patients who are candidates for radical prostatectomy have no symptoms, while approximately 100% of those who undergo the surgery have loss of some desired function. If the surgery (or radiation) saved a life, there could be an argument for it. As it is, the survivals are almost the same, with or without treatment (an exception is a younger man with severe cellular morphologic changes and no metastases; a quite unusual situation). Since virtually every man on Medicare has a reasonable chance of having prostate CA, routine PSA and biopsy tests are contributing to the impending bankruptcy of the system, without adding to survival.

PSA specificityprostate CA screeningcriticism and reply

Expand full comment

Isn't "the experts are wrong, wrong!" a standard news bias? I'm not sure we can conclude much of anything from just one article.

Expand full comment