Managed Care

A New Cancer Screen Filter

Experts are rethinking the routine cancer screen. In the future, screening recommendations may be tailored, risk-stratified, and based on our genes. (2016)

 

By Charlotte Huff

Cancer screening recommendations, long issued in one-size-fits-all, may be moving toward a bespoke approach, calibrated to more individualized risk factors and, perhaps eventually, a person’s genes.

Currently, many screening recommendations hinge on just a few scraps of information about a person, including age and a family history of the disease. Meanwhile, the U.S. Preventive Services Task Force (USPSTF) and other groups have struggled with striking the right balance between the benefits of earlier detection and the risks of overdiagnosis and overtreatment. In 2012, USPSTF chose to steer clear of the perils of overdiagnosis and overtreatment when it recommended against routine PSA screening for prostate cancer, citing the harms of unnecessary biopsies and surgery and the difficulties of distinguishing slow-growing from aggressive malignancies.

But now the task force is revisiting that advice, and some researchers hope the result will be something of a comeback for PSA screening, but a new-and-improved version that takes into account more risk factors and stratifies screening more carefully by a man’s chances of getting the disease. “My feeling and the feeling of a lot of people is that [the 2012 recommendation] was like throwing the baby out with the bath water,” says Mark Preston, MD, a urologic cancer surgeon at Brigham and Women’s Hospital in Boston.

Preston coauthored a study that found that men, ages 45 to 69, with an elevated PSA were more likely to develop lethal prostate cancer later in life than men whose PSA is not elevated. (It should be noted, though, that the analysis was based on the Physicians’ Health Study, with a study population and thus PSA readings primarily from white men.) In their results, published in August in the Journal of Clinical Oncology, Preston and his colleagues made the suggestion that screening could vary accordingly, so men with low PSA levels compared with the median PSA of their age group wouldn’t need to be screened moving forward as often as those with high levels.

Another study, which began this year at NorthShore University HealthSystem in Evanston, Ill., is looking ahead to when genetic tests might become more routine. Jianfeng Xu, director of personalized medicine at NorthShore, has developed a risk score based on genetic tests for breast, colorectal, and prostate cancer. Now Xu and his colleagues have designed a study to find out whether people’s screening behavior will change depending on their genetic risk score. Genetic tests like this will add some expense upfront but might save money if they winnow out unnecessary screening and the follow-up tests, they say. “We think that it’s not going to necessarily increase health care costs, but rather make our utilization of precious health care resources much more efficient and beneficial,” says Charles Brendler, MD, executive research director of the Program for Personalized Cancer Care at NorthShore.

Homing in on high risk

In its final research plan for the revision of the prostate cancer screening recommendation, the USPSTF has signaled its willingness to consider a risk-stratified approach to screening. One of the questions the task force decided to address is whether the PSA test’s effectiveness varies with a man’s age, race, ethnicity, family history, and other factors.

The prostate cancer screening is not the only cancer screening test getting a second look. Screening for lung cancer is a relatively recent development. In 2013, the USPSTF gave lung cancer screening with CT scans a “B” recommendation for adults ages 55 to 80 who have at least a 30-pack year history, including current smokers and those who have quit within the last 15 years. The task force’s findings are highly influential, as any recommended test given an “A” or “B” rating must be covered by insurers under the Affordable Care Act.

Although it is more targeted than other kinds of cancer screening, lung cancer screening looked to be more expensive because it involves a CT scan, although some research suggested otherwise. For example, an analysis published in 2012 in Health Affairs found that the cost per projected life saved for lung cancer screening was $19,000, which was less than the per-life-saved cost for breast, cervical, or colon cancer screening. The low-dose CT screening test is expensive, and in this analysis the cost was pegged at about $250. But the economics of lung cancer screening are favorable because the screening is targeted, says Bruce Pyenson, the lead author and a principal consulting actuary at Milliman. “You have a concentrated risk population. You also have a cancer where the difference in survival between early and late stage [cancer] is very dramatic.”

But concern about false positives and the risks associated with lung biopsies and other procedures has led to some research of a strategy that might more precisely identify smokers at the highest risk of developing lung cancer, using factors like family history, an emphysema diagnosis, and body-mass index.

Earlier this year, a research team led by NCI researchers reported the results of some sophisticated computer modeling of this approach in JAMA. Their results showed that if lung cancer screening were to zero in on smokers with the highest five-year risk of developing lung cancer, it would be more effective. The number needed to be screened to prevent one lung cancer death would be 194 compared with 162 using the USPSTF recommendations. This research also found that it would be more efficient, with 116 false positives per prevented lung cancers compared with 133 using the USPSTF recommendations.

More late-stage cancers

But any tightening up of the criteria for screening carries with it concern that cancers will go undetected and be diagnosed at a later, less treatable stage. Some experts predicted that would happen after the 2012 USPSTF decision against routine PSA screening. And a study published earlier this year showed a small uptick in late-stage cases—those that have spread to the bone or other organs—in men ages 50 to 69 beginning in 2004 as various groups moved away from advising routine screening PSA tests. It’s a small increase, and it’s not yet known whether it will translate into more men dying from prostate cancer, notes Richard Hoffman, MD, one of the study’s authors and a general internist at the University of Iowa Carver College of Medicine. Even so, Hoffman says. “Going from overscreening to not screening at all, I think is a mistake.”

One middle-ground approach is to add some interpretation to PSA results that would help determine when and if further screening is necessary. In addition to identifying a correlation between above-median PSA results and lethal cancers, Preston’s study in the Journal of Clinical Oncology also found that if a man’s PSA levels stayed below the median level until he was about 60, the chance of him developing life-threatening prostate cancer was low. Further research is needed, but it might turn out that a man with a still-low reading at age 60 won’t need another PSA test, Preston says. “So what we can do is pretty effectively target our screening efforts,” he says.

At NorthShore, the primary care physicians involved in the study of the effect of genetic tests on screening will discuss the cancer risk score results with 500 patients. Hoffman, at the University of Iowa, says that genetic tests could eventually solve one of the biggest challenges in cancer: distinguishing between the slow- or non-growing malignancy that has little effect on a person’s health from the aggressive ones that should be treated. Meanwhile, it’s hard to know whether other risk-stratifying efforts—such as relying on a mid-life PSA—are meaningful, Hoffman says. Screening may catch the less harmful, slow-growing tumors while more lethal malignancies evade early detection regardless of the schedule. “We can’t say that if we looked for it earlier and more aggressively we’re going to make any difference,” says Hoffman, who hopes that the USPSTF will support a more personalized decision-making approach.

If genetic tests were developed to reliably predict risk (a big if, at this point) it might spare patients a lot of trouble—and the American health care system a lot of expense. Brendler at NorthShore points out that he’s gotten an annual PSA test for some 25 years. The total cost of those tests, he figures, adds up to more than $3,000—and his PSA level has remained unchanged. A genetic snapshot at age 40 might have suggested more occasional screening, he says.