What patients should be told before agreeing to a blood test that could change their lives

Iwould guess that most asymptomatic patients tend to think of their periodic medical examinations along the lines of routine automobile maintenance—top off lubricant levels, check the brake pads, perhaps even flush the radiator, and all will go well for another year. When it comes to screening for cancer, however, the stakes are higher, more along the lines of preventing disaster—finding the loose lug nuts before the wheels come off on the expressway. Physicians tend to agree. Cancers, even those most relentlessly destructive when metastatic, such as melanoma, pancreatic cancer, or renal cell carcinoma, can often be surgically cured if identified when localized. However, few cancers identified in symptomatic patients are susceptible to such ready control. When the explanation for nagging, progressive local symptoms or vague symptoms of systemic debility is cancer, it will not be containable for long, much less eradicable, despite our best, most technically impeccable efforts. Caring for patients with these cancers provides a disproportionate share of our saddest, most frustrating experiences as physicians. Is it unsurprising, then, that it is hard to hide from our patients our enthusiasm for a new technology that could find tens of thousands of cancers every year before they metastasize? How not, when the glum alternative is a late diagnosis, after which all we will have to tell the patient will, on balance and ultimately, be bad news.

It is probably psychologically inevitable that cancer catastrophes dominate our thinking about cancer screening, but, for patients, the usual result of screening and its consequences is more mundane—no cancer diagnosis. The emotional cost of most cancer screens is the anxiety that precedes comforting reassurance—an uneasy moment or two before a “normal” reading, an anxious week or so awaiting a “benign” biopsy result, or a few months of occasional unbidden and unwelcome thoughts before the “follow-up study suggested” that in the end finds no worrisome change. Although unpleasant, such experiences end in closure, and, especially in retrospect, subside quickly into the untidy flow of human existence. Even in the uncommon and far more momentous situation in which the screen finds cancer, well, we tell ourselves, it could have, would have, been worse. Most screen-detected cancers are not obviously metastatic and incurable, and a consensus approach to the initial treatment is usually available or achievable. When the staging studies are complete, the treating oncology specialist leads an eventful, perhaps emotionally ragged, but usually harmonious discussion, that, under the specialist’s unquestioned authority, leads to a treatment plan. Soon, the surprisingly reassuring routine of treatment eases the way from the diffuse emotional cataclysm of cancer into a focus on the daily details of the new medical agenda. Surgery, radiation, chemotherapy (infrequently), or even a complicated multimodality approach begins. At least for a while, the focus is on execution, and all the bad news, if any, is in the future.

Obtaining informed consent for a procedure is especially difficult when the medical justification is ambiguous. The goal of informed consent is for patients to arrive at a decision they are comfortable with, based on a comprehensible, science-based rationale. The physician’s role is to provide information, respond to the patient’s doubt and uncertainty, and ensure the final decision is a good one that is based on the medical facts, addresses all important patient worries, and fits within the patient’s values and the family’s comfort zone. Compared with the brief recitation of the purpose and likely complications of most diagnostic procedures, the discussion about screening for prostate cancer is a messier process. Prostate-specific antigen (PSA) screening for prostate cancer produces fewer of the good, reassuring, definitive screening results, more of the ambiguous results from which the shadow of cancer is hard to shake, and a sudden growing awareness that no treatment for early prostate cancer bears a consensus stamp of approval, or is even assuredly necessary. We do know that when prostate cancer becomes metastatic, it will within a few years become essentially untreatable, and for the patient, the figurative wheels will fall off. But, we also know that many more patients suffer the complications of treatment than are spared from prostate cancer to die of another cause. Furthermore, through PSA testing in follow-up, we imbue their lives with a cyclical fear of recurrence. In Tannock’s formulation,1 PSA monitoring has nearly eradicated asymptomatic prostate cancer by making it symptomatic: “testing of a man who is symptom free [to] provid[e] him with evidence of our inability to cure his disease.” We do not know, or even know how to decide, whether, why, by how much, at what time, and for which patients, despite all the aggravation we increasingly document, the whole process is worthwhile.

For now, no one can declare with conviction and defend for long the proposition that screening and aggressive treatment of prostate cancer prolongs life on average, despite the many studies published that purport to weigh in on the value of PSA screening for prostate cancer,2, 3, 4 although this is subject to revision by ongoing European5 and U.S.6 randomized trials. Still, researchers seem unable to resist publishing data indirectly addressing the efficacy of screening. PSA screening diagnoses many more clinically organ-confined cancers than earlier methods,7 but such results would be expected whether or not early diagnosis leads to more successful treatment. Mortality from prostate cancer in Tyrol, Austria, declined faster after widespread screening and aggressive treatment than in other Austrian provinces that did not adopt the approach.8 However, prostate cancer mortality rates in Seattle, one of the most aggressive screen-and-treat areas in the United States, were no different than in Connecticut, one of the most conservative areas, despite prostatectomy rates more than fivefold higher in 1989.9, 10 Great Britain, where PSA screening was sparse and aggressive treatment uncommon, has had declines in prostate cancer mortality comparable to those in the United States, where an aggressive approach was the norm, beginning a year or two later.11 The international correlation between prostate cancer mortality rates and the use of PSA screening is inconsistent.12 Given plausible estimates of the natural history of prostate cancer and the temporal relationship between widespread PSA testing and the recent decline in U.S. prostate cancer mortality, a causal relationship is unlikely.13

Still the enthusiasm to bring in an early verdict in favor of PSA screening has led to the publication of deeply flawed, but superficially compelling, results. LaBrie and colleagues14 concluded that PSA screening saved lives in their randomized trial by lumping those who chose not to accept screening with those not offered it, rather than the correct “intention to screen” approach comparing those invited to screen with those not invited. As a result, the crucial role of randomization in eliminating bias was obliterated.14

We know that the use of PSA screening sharply increases the rates of diagnosing prostate cancer,15, 16, 17 and screening disproportionately diagnoses both earlier and slower growing cancers. Nevertheless, Dugan and colleagues,18 extrapolating on a tumor volume prognostic model with a scanty empirical basis, concluded that the proportion of screen-detected cancers at the Mayo Clinic that were “clinically insignificant” ranged between 0.3% and 14.5% of tumors, depending on their assumed exponential growth rates. In other words, with approximately 200,000 cancers diagnosed annually, about fivefold more than the greatest annual death toll in periods when treatment was largely ineffectual and infrequently used, the investigators concluded that nearly all cancers were “clinically significant.” Etzioni and colleagues19 estimated that 29% of screen-detected cancers in whites and 33% of those in blacks would otherwise not have been diagnosed during life, assuming the average lead time is 5 years, or 39% and 41%, respectively, assuming 7 years. In addition to these otherwise clinically invisible cancers, screening diagnoses cancers that would never metastasize or cause death, and it diagnoses them years earlier. Using population-based data, Hugosson and colleagues20 found that 1 in 4 Swedish men eventually diagnosed with prostate cancer in the PSA-screening era would be dead of the disease within 15 years, and McGregor and colleagues21 estimated that only 16% of screen-detected cancers in Quebec would result in death before age 85. The screening data cited are based on the most common approach to PSA screening, with 4.0 ng/mL the threshold for biopsy and sextant biopsy sampling. By lowering the threshold to 2.5 ng/mL22 or doubling or more the number of biopsy cores23, 24, 25 would further amplify overdiagnosis—ever more cancers, ever less clinically significant. Of course, the longer men are screened, the likelier they are to have a PSA level greater than the 4.0 ng/mL threshold for biopsy sampling. The lifetime mortality rate for prostate cancer is 3% for white men, 4% for black men, and the lifetime risk of a diagnosis of prostate cancer has risen from 10% to 16%. Is our goal 20%? 25%?

So what can we tell our patients before screening? We cannot say “If you are screened, you will on average live longer,” although some day we may be able to. We cannot say “if we find a cancer, you will be better off if we treat it.” We should tell them that, regardless of whether trials find PSA screening efficacious, many more men will have long-term urinary, bowel, and sexual dysfunction than will avoid death from prostate cancer. We certainly cannot say “information cannot hurt you”: anxiety from false-positive tests, the discomfort and inconvenience of follow-up procedures, and the residual uncertainty arising from a “positive” PSA test but negative biopsy findings can certainly harm some men. However, the pain of a process that results in a negative result is small, of course, compared with dying of metastatic prostate cancer, and anyone whose screening results in avoiding death from prostate cancer benefits tremendously. So what we can say is this: “There is a lottery available, the cost is affordable for most people, although it may be higher than they initially think, and there may be some grand prizes awarded. Do you want to buy a ticket?”