USPSTF recommends CT lung cancer screening

In a landmark decision, the U.S. Preventive Services Task Force (USPSTF) on Monday issued a draft recommendation in favor of CT lung cancer screening for long-term smokers. The decision is a major step toward making the exam an accepted test with widespread application, much like breast or colorectal cancer screening.

The recommendation is subject to a month of public comments and then revision before it can be finalized, but it offers most of what screening advocates had asked for, including screening criteria that are a little broader than those used in the major study driving the recommendations: the National Lung Screening Trial (NLST). The NLST in 2011 found a mortality benefit of at least 20% in long-term smokers who underwent annual low-dose CT screening.

"They recommended what we have long sought, and the impact is going to be monumental in terms of the size of the community affected," Laurie Fenton-Ambrose, the president of screening advocacy organization Lung Cancer Alliance (LCA), told

Decision will affect millions

That community numbers in the millions. Lung cancer is the leading cause of cancer death in the U.S. and is diagnosed in more than 200,000 people each year, said USPSTF chair Dr. Virginia Moyer in a statement accompanying the draft recommendation.

Almost 90% of those who are diagnosed die from the disease because it isn't detected until it has reached an advanced stage. By screening those at high risk, lung cancer will be detected at earlier stages, when it is more likely to be treatable, Moyer said. The risk of lung cancer increases with the duration and extent of smoking.

Rather than limiting its screening recommendation to heavy smokers ages 55 to 74 years as the NLST did in its inclusion criteria, the USPSTF draft recommends screening for ages 55 through 79. Like NLST, the draft requires a smoking history of 30 or more pack-years for inclusion, and if the individual has quit smoking, he or she must have quit within the past 15 years.

But unlike NLST's screening recommendation of three annual low-dose CT scans, the draft does not specify the number of annual screenings, leaving that question, for now, to the discretion of patients and healthcare providers.

USPSTF conferred a grade B recommendation to CT screening, which denotes a high certainty of moderate benefit, or moderate certainty of moderate-to-substantial benefit. At the same time, the draft cautions against the screening of individuals with significant comorbidities, who are at risk of death from other causes, particularly heart disease, before lung cancer could claim their lives.

In practical terms, grade B means that the recommendation, if approved, will be automatically included for Medicare coverage under the terms of the Patient Protection and Affordable Care Act (PPACA), which takes full effect next year, according to Fenton-Ambrose. "Because of the way the Affordable Care Act looks at these ratings, if it's an A or a B it becomes an essential health benefit and is then accorded coverage," she said.

Separately, many private insurers are already stepping up to the plate to offer CT lung cancer screening. The draft recommendation comes a little earlier than expected, and from here things could progress rapidly, with a final decision and Medicare coverage possibly rendered by the end of the year, Fenton-Ambrose believes.

"We're going to do everything possible to keep this process moving forward as rapidly as possible," she said.

Underpinnings in literature

The draft recommendation was based on an extensive review of published lung cancer studies. It was the product of a request made by USPSTF, which sought to update its 2004 report showing insufficient evidence to conclude a significant benefit from CT lung cancer screening.

USPSTF joined with the Agency for Healthcare Research and Quality (AHRQ) to formulate key questions about the harms and benefits of screening, focusing on populations, interventions, outcomes, and harms associated with low-dose CT. The review was published concurrently with the draft recommendation in the Annals of Internal Medicine (July 29, 2013).

The review found "strong evidence" that low-dose screening with CT can reduce both lung cancer mortality and death from all causes. While cautioning that CT screening of long-term heavy smokers must balance the benefits of screening with the harms, the review authors concluded that evidence favored CT screening of high-risk individuals.

The review found its strongest evidence in NLST, the only trial the authors rated as "good quality." They detailed shortcomings in the evidence mostly from three other trials they deemed strong enough for analysis. Even in studies rated only as "fair," the group found clues that it considered helpful in addressing issues ranging from the psychosocial consequences of screening to incidental findings, overdiagnosis, and false positives.

The NLST trial was key in demonstrating a benefit for CT lung screening, wrote Dr. Linda Humphrey and Dr. Mark Deffebach from Oregon Health and Science University and Portland Veterans Affairs Medical Center, who were lead authors of the Annals of Internal Medicine paper.

"Until the last year or so, screening has really not been recommended by any organization, and the reason for that is because there haven't been randomized, controlled trials that show a benefit," Humphrey told

NLST is also the study most generalizable to the U.S. population because it was conducted in the U.S., she said.

Noteworthy trials

The studies covered in the USPSTF review came from a search of the Medline database for publications from 2000 to 2013, as well as several other sources. Reviewers rated each trial as good, fair, or poor.

Among the 8,215 abstracts studied, 67 full-text articles met the inclusion criteria. Of these, seven randomized, controlled trials reported the results of low-dose CT (LDCT) screening in intervention and control groups. Results were analyzed further in four trials:

  1. NLST, which compared CT scans with radiography exams in more than 50,000 individuals ages 55 to 74 years at 33 U.S. sites
  2. DANTE (Detection and Screening of Early Lung Cancer by Novel Imaging Technology and Molecular Essays), an Italian trial comparing LDCT plus sputum cytology to usual care in about 2,400 male smokers with 20 or more pack-years of smoking history, ages 60 to 74, who had quit within 15 years
  3. DLCST (Danish Lung Cancer Screening Trial), which evaluated low-dose CT in more than 4,000 healthy men and women ages 50 to 70; the study was planned to last five years, and participants had to have quit smoking after age 50 and less than 10 years before enrollment
  4. MILD (Multicentric Italian Lung Detection), a single-center trial comparing annual or biennial LDCT with no screening in 2,400 men and women 49 years or older who were current (20 or more pack-years) or former smokers (quit less than 10 years ago) with no recent history of cancer

DLCST and MILD participants were younger than those in NLST and DANTE, and correspondingly, lung cancer and all-cause mortality were both lower in DLCST and MILD, the group wrote.

From a public health perspective, NLST trial data indicated that just 320 individuals needed to be screened to prevent a single death from lung cancer. In comparison, the number needed to screen for mammography is 1,339 for women ages 50 to 59 years after 11 to 20 years of follow-up, the reviewers noted.

NLST found that lung cancer mortality was cut by 20% and all-cause mortality by 7% compared to chest radiography in the course of the three annual scans, Humphrey and colleagues wrote. One-quarter of all deaths were from lung cancer.

Yet NLST's mortality gains have not yet extended to other trials. Another trial that was reviewed, the ITALUNG trial, suggested a mortality benefit that was judged to be insignificant. Two poor-quality European trials showed no benefit to CT screening in a younger, lower-risk cohort.

Several factors may explain the differences, most likely stemming from participant ages and risk levels, different follow-up periods, and cohorts too small to show a benefit from screening, the group wrote. Also, no studies showed a benefit from chest radiography; however, another study that was reviewed, the Prostate, Lung, Colorectal, and Ovarian (PLCO) trial, suggested a benefit among high-risk individuals and possibly among women.

Beyond NLST, the DLCST study with its younger population and far lower prevalence of lung cancer may suggest that lower-risk individuals would not receive much benefit from screening, though the study was small and underpowered.

"One of the things in the Danish study was that there was even a suggestion of harm, so I think that's very important to keep in mind -- that there may be harm in screening people at lower risk based on their age and smoking exposure," Humphrey said.

A few complications, some overdiagnosis

Indeed, no analysis of screening is complete without a review of the potential harms of the test, which, for CT lung cancer screening, can include follow-up complications. In NLST there were complications associated with 245 CT exams and 85 chest radiography exams; however, major complications were uncommon (occurring following intervention: CT group = 73, radiography group = 23). Cumulative radiation exposure from the CT scans should also be considered a harm, according to Humphrey.

Another issue is overdiagnosis, or the detection of cancer that may never be a threat to the patient. Overdiagnosis was not directly reported, though NLST results suggested its existence. More advanced cancers were reported in the radiography group versus the CT group, and in the other trials more early-stage cancers were found in the CT groups.

"We don't know the extent of overdiagnosis; the best we could tell, it looked to be on the order of 10%," Humphrey said. That overdiagnosis risk is in addition to other risks that come with diagnosing other diseases with CT that could lead to other harms in the course of care, she added.

"One of the most common findings in doing these CT scans was emphysema," she said. "I don't know that there's data to support a benefit of being diagnosed earlier versus later with emphysema." Similarly, coronary calcification was also prevalent, and the benefit of knowing that these individuals are at higher risk for vascular disease isn't surprising, though the benefit of knowing versus not knowing is actively being studied.

Fear of screening

Seven studies evaluated psychosocial consequences of screening. The authors didn't find that screening had a clinically meaningful impact on distress, Humphrey said. There was a slight increase in distress among people testing positive, but it resolved for most patients as follow-up made them negative again. Nor did screening have much effect on smoking behavior.

"One of the concerns has been whether people will say, 'Oh, I can get screened for lung cancer. I don't have to worry about quitting smoking,' " Humphrey said. "We didn't find that in the evidence, and I think a key message for any article about this topic is that not smoking is the key to preventing many things, including lung cancer, and it's such an important part of any screening program or any clinical care we provide."

Are there ways to further stratify the target patient population to reduce false positives? Humphrey considers it unlikely that any new biomarkers on the horizon could enable the definitive discrimination of patients with or without lung cancer.

"I think the best right now is stratifying people by risk based on family history, age, smoking exposure, and lung disease, and using those common things that we identify in clinical practice," Humphrey said. "They've certainly been using some markers for diagnosis, and that might be where there's benefit: in reducing follow-up in people with false positives, and reducing the number of people getting biopsies."

Future research will focus on methods for identifying those at highest risk for cancer, they wrote. In addition, research "to improve discrimination between benign and malignant pulmonary nodules, and to find early indicators of aggressive disease is warranted," they wrote. Specimens have been collected but studies have yet to be reported.

As the USPSTF decision percolates, the radiology community will have plenty to debate in terms of how CT lung cancer screening programs are set up. But for the present, it should take credit for years of hard work in demonstrating the value of CT lung cancer screening, said Fenton-Ambrose.

"The radiology research community should take a bow," she said. "It's the radiology community that helped achieve one of the largest lifesaving benefits that we're going to see in the cancer community for a long time."

Page 1 of 654
Next Page