Screening for Disease Where No Symptoms Exist

Credit: Photograph by Adam Levey

The idea of screening for disease sounds so reasonable that no one could be opposed to it: After all, forewarned is forearmed. "It's a good idea that we don't know how to do very well," Hadler says. "Our screening tests stink, and we're testing for too many things that aren't serious problems or for which we don't have effective treatments. I would be furious if anyone checked my cholesterol or PSA." The premise of looking for disease in people who don't have symptoms began, according to Welch, in the mid-1960s, when the Veterans Administration conducted a small study that treated about 70 men who had severely high blood pressure with antihypertensive drugs, and then compared the outcomes with a similar group of 70 who had received no treatment. During the six-month study period, 27 untreated men suffered some major health event, including four deaths, while only two in the treated group fell ill. This amazing success story was the eureka moment that convinced doctors that if they could identify people with merely moderately elevated blood pressure, then they could treat it and get good results. But, Welch says, sky-high blood pressure turned out to be "the low-hanging fruit." Since then, screening successes have been few and far between. When you narrow the definition of normal – many doctors routinely classify a guy with a 130/85 blood pressure as borderline hypertensive – you expand the pool of people available to be treated. That's good for the drug companies, but you cross a line in which the small benefit that may be derived by a few will be wiped out by side effects needlessly endured by the many. "It's hard to make basically well people better," Welch says, "but it's not that hard to knock them off their game."

Blood Pressure
Blood pressure is the one health measure that virtually all doctors agree should be taken regularly. (It can be done with a home monitor or the cuff at your local pharmacy.) As the veterans study made clear, high blood pressure is a serious threat – think heart attack, stroke, death – and if you've got a systolic BP of 160 or higher, you want to know about it. Even doctors who are pharmaceutical skeptics mostly recommend bringing it down by any means necessary, including drugs.

The problem isn't measuring blood pressure per se; it's interpreting the results when they're not terrible but still less than ideal. If getting consistent readings of systolic pressure in the borderline or prehypertension range of 130 to 139, or even in the mild hypertension range of 140 to 159, pushes you to make changes in diet, exercise, and handling stress, then that's a valuable screening test that will pay key dividends. If it prompts your physician to put you on antihypertensive drugs, then it's a screen of dubious value since the most comprehensive study, a 2012 meta-analysis of previous research, found that treating mild hypertension with drugs in people without heart disease didn't do any good.

The standard lipid panel screen that's done at your checkup gives you a reasonably accurate estimate of the total amount of LDL (low-density lipoprotein, or "bad cholesterol") circulating in your blood. But focusing on LDL fits an outdated model of heart attacks: Plaque inside the coronary arteries, partly composed of LDL, grows so large that it cuts off the blood supply to the heart. We now know that heart attacks result from a complex interaction between plaque and inflammation: A piece of plaque breaks off from the arterial wall, causing a blood clot to form, which blocks blood flow, causing the heart attack. And long-running population studies like the Framingham Heart Study confirm that LDL is only part of the puzzle; by itself, it's not a strong predictor of cardiac mayhem. Newman notes that three-quarters of first heart attacks occur in people with normal LDL. Still, a lot of clinicians put great significance on a high LDL result, and it's a problem for which they have a solution: statin drugs, which indeed drive the number down. Some cardiologists who are willing to admit that LDL has been a flop as a screen for heart disease put their faith in a later generation of advanced lipid tests, which give a more specific picture of LDL cholesterol in action – for instance, the size of the particles or the number that invade the artery walls. But no one has ever done a rigorous study showing that these tests are better at predicting future heart attacks than the plain-vanilla LDL number. "These 'advanced' tests are flaming horseshit," Newman says. Put more generously, the jury is out.

"When it comes to screening," Brawley says, "a doctor who says, 'Let's err on the side of caution,' may actually be erring on the side of grave harm." Nowhere are the stakes higher or the errors more grievous than with prostate cancer. It is, after lung cancer, the cancer that kills the most American men, responsible for 30,000 deaths a year. Logically, early detection should be a blessing, but little about prostate cancer screening is logical.

Until the mid-1990s, screening consisted of the venerable digital rectal exam, the physician's latexed finger traveling up the rectum to feel the gland for hardness and irregular shape. If the cancer was advanced enough to reveal itself by touch, it usually deserved to be treated aggressively. By then, in fact, it had probably gone metastatic, spreading throughout the body. Treatment was damage control and, the urologists hoped, to buy time, not to save lives.

The arrival of the prostate-specific antigen (PSA) screening test changed the equation. It measured the amount of an enzyme, made by the prostate, that leaked into the bloodstream in greater quantity when the gland was enlarged, possibly because of cancer. For the first time doctors had a respectable chance of catching the cancer before it spread, when it was still encased in the gland. In increasing numbers, the urological surgeons cut it out and the radiologists burned it out, using ever more elaborate technologies.

There were drawbacks from the get-go. The screen wasn't very specific. A suspicious PSA reading of two or four or higher might be caused by a normal age-related enlargement of the gland or an infection, prostatitis. According to the research literature, about three-quarters of the men who got alarming PSA scores and were subjected to the discomfort and infection risk of a prostate biopsy turned out to be false-positives – no cancer found. And the men whose biopsies came back with cancer, for whom PSA had done its job of early detection, were in a quandary. "Roughly half the elderly men in America have evidence of cancer in their prostate, yet only 3 percent will die from it," Welch says. In other words, however much a urologist hopes he is doing the right thing by finding and treating prostate cancer in an individual patient, he knows, or should know, that the majority of patients with detectable amounts of cancer would be better off left alone because the treatment is more dangerous than the disease.

The medical system is left with a kind of Sophie's choice: How many men are you willing to treat, at the cost of severe collateral damage (some amount of erectile dysfunction and incontinence is inevitable), to try to cure them of a cancer that in most men will turn out to be more or less harmless (i.e., they'll eventually die of something else)? How many do you harm in the quest to save a relative few? To answer that, you must have some idea of how many lives are being saved by aggressive treatment of localized prostate cancer found through PSA screening. To that end, nearly a quarter of a million men, here and in Europe, have been enrolled in clinical trials to compare the outcomes of men who had been screened with PSA (which means if they had prostate cancer, it was likely found and treated earlier) to men who had not (which means if they had the disease, it was almost certainly found and treated at a later, more lethal stage). The numbers vary some. What they have in common is that they are lousy. Welch puts it broadly: For every man who "avoids a prostate cancer death, roughly 50 are treated needlessly." Both Welch and Brawley, a prostate cancer expert, refuse to get their own PSA tested.

The sobering research results published in the past several years have shaken the urological establishment's faith in screening and early detection. The American Urological Association, which, as recently as 2009, recommended that men at the age of 40 get an initial PSA reading, now acknowledges that the risks of false-positives and overtreatment are so great, only men in the 55-to-69 age-group should even consider it. The U.S. Preventive Service Task Force recommends that no one get the screen; which goes beyond what Brawley and Welch believe, which is that men should have the option. Some may come to the rational conclusion that they're willing to endure most anything if it lessens the risk of dying young. What Brawley and Welch and virtually all epidemiologists hate is the idea of mass screening – the hospital van pulling up to the mall to offer free or discounted PSA tests, ensuring down the line that a lot of scared and ill-prepared men will flock to the hospital for more testing and, in some cases, treatment. You can't tell the average American male he's got a cancer growing in his body without his wanting to get it yanked or burned out, and the hell with a nuanced discussion of the numbers. "The time to talk to your doctor about PSA is before you take the test," Welch says.

Dr. Robert Mordkin, chief of urology at Virginia Hospital Center, notes that there has been a significant shift in attitude among his peers about how to use the PSA test to find the best candidates for prostate surgery. "Narrowing the funnel" he calls it. In other words, recommending treatment only for men who fall at the extremes of our conventional diagnostic measures, with very high scores on PSA and Gleason (a measure of the cancer cells' potential lethality). As well, a new generation of genetic tests should be used to add predictive power, as the eminent prostate cancer pathologist Dr. David Bostwick puts it, "to separate the pussycats from the tigers." The hoped-for net result is fewer men with elevated PSA getting biopsied and fewer men with biopsies that reveal cancer getting treatment. More would fall into the "watch and wait" camp: do nothing besides regular testing and if the numbers start to go south, then consider aggressive treatment.

PART 1: Screening for Disease Where No Symptoms Exist

PART 2: Drugs: Effective for the Few, Prescribed to the Many

PART 3: Surgeries You May Be Better Off Without