MDCT: In CT lung cancer screening, don’t forget the heart

Radiologists who screen smokers for lung cancer with CT should remember to look at the heart as an important predictor of mortality in its own right, according to a Tuesday talk at the International Symposium on Multidetector-Row CT (MDCT).

 New studies show that mortality is strongly linked to coronary calcium, even if there are no major adverse cardiac events, and that thoracic aortic calcium can heighten the risk of stroke, said Dr. Mathias Prokop from Radboud University in Nijmegen, the Netherlands. Considering that lung screening scans encompass the heart, radiologists must look at it and for other potential abnormalities in the image.

“One of the big issues is that we are scanning the whole chest, and the question is why should we not look at other areas? [Chronic obstructive pulmonary disease (COPD)], the heart, the bones? Lots can be done there, and you can actually use this information for many reasons” said Prokop, who heads the department of radiology and nuclear medicine in Nijmegen.

Ungated exams?

Some findings are incidental, some are important, and others are less so for getting an accurate picture of the patient’s health and risk, Prokop said. For now, however, the heart is the most useful area to examine. Someday COPD will also rise in importance if reasonable treatments are developed for it.

Of course, one issue with coronary calcium is the need for gated scans to stop cardiac motion; CT lung cancer screens are ungated.

Dr. Mathias Prokop

Dr. Mathias Prokop from Radboud University.

“But we all know that we see calcium on these [ungated] scans, and there is a good correlation between gated and ungated CT,” he said. “There is a slight underestimation with ungated scans, but it doesn’t mean there is always underestimation.”

Some images will show more calcium than exists and some less, but generally the calcium that readers see on ungated scans is really there, Prokop said.

Even an Agatston score of zero on an image doesn’t mean calcium is not present, he said. The literature shows that follow-up scans three months later often show coronary artery calcium, and the variability of the scans (about 50% for Agatston score and 36% for volume score) means that calcification was probably there from the start.

A new study in Radiology (Chiles et al, March 9, 2015) that examined the association between coronary artery calcium and mortality in the National Lung Screening Trial (NLST) found that visual assessment of lung cancer screening scans, even without Agatston scoring, was sufficient to assess cardiac risk.

In that paper, visual assessment of calcifications showed good agreement with the categorized Agatston scores (weighted κ = 0.75), and radiologists using a visual assessment assigned participants to the same risk category as the Agatston score in 73% (1,052/1,442) of CT scans and within one risk category in 99.7% of scans.

“The simplest method, an overall visual assessment of [coronary artery calcium] as none, mild, moderate, or heavy, was able to separate patients into risk categories on the basis of either [coronary disease] death or [acute myocardial infarction],” Chiles and colleagues wrote in Radiology. “All five radiologists in this study preferred the visual assessment, not only because it was faster and simpler than either of the other two methods, but it also eliminated the need for additional software.”

“Visual assessment is good enough,” Prokop said. “Looking just at an overall impression is good enough to give you a good estimate of the hazard of your particular patient.”

Lung scans predict cardiovascular risk

But does the presence of calcification on lung screening exams predict cardiovascular risk? It does, Prokop said. Takx and colleagues just completed an examination of participants in the Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) lung cancer screening trial (Journal of Cardiovascular Computed Tomography, January-February 2015, Vol. 9:1, pp. 50-57). They found that any calcification increases the risk of death among smokers screened for lung cancer, and that Agatston scores 400 and higher increased the risk of death by a factor of 12 — “quite a dismal survival factor,” Prokop said.

In 2012, Jacobs and colleagues found that an Agatston score of 1,000 or more on a low-dose CT scan increased the chance of dying within 2.6 years by a factor of 10, compared with a scan with no detected calcium (American Journal of Roentgenology, March 2012, Vol. 198:3, pp. 505-511).

“Coronary calcium actually measures more: It probably measures your ability to withstand other diseases, so if there’s cardiac comorbidity, your chance of dying from this other disease is higher,” Prokop said. And calcium is a “really massive risk predictor” when it is combined with the classic situation of hypercholesterolemia, he added. Finally, the number of calcifications is a big predictor of elevated cardiovascular risk.

Even more to look for

All-cause mortality also spikes with a rise in thoracic aortic calcification, as Buckens et al reported in European Radiology (January 2015, Vol. 25:1, pp. 132-139), Prokop said.

Buckens and colleagues showed that even though adverse coronary events are mainly determined by coronary artery calcium, if you look at noncoronary vascular events such as stroke, aortic aneurysm, and peripheral artery occlusive disease, then aortic calcium is predictive. The study also showed that osteoporosis affects survival, Prokop said.

“If you have vertebral fractures, you have twice the chance of someone who has the same other risk factors [of dying] within six years,” he said.

Should calcification be reported on lung cancer screening scans? Absolutely, according to Prokop.

“Do we know that it will affect patients?” he said. “That is actually the big question mark, as there are no trials yet that show that treatment will improve survival.”

That missing bit of data is especially hard to fathom considering that coronary calcium research has been around for close to 30 years, Prokop added.


The Importance of Lung Cancer Screening With Low-Dose Computed Tomography for Medicare BeneficiariesLung Cancer Screening for Medicare PatientsLung Cancer Screening for Medicare Patients

The National Lung Screening Trial has provided convincing evidence of a substantial mortality benefit of lung cancer screening with low-dose computed tomography (CT) for current and former smokers at high risk. The United States Preventive Services Task Force has recommended screening, triggering coverage of low-dose CT by private health insurers under provisions of the Affordable Care Act. The Centers for Medicare & Medicaid Services (CMS) are currently evaluating coverage of lung cancer screening for Medicare beneficiaries. Since 70% of lung cancer occurs in patients 65 years or older, CMS should cover low-dose CT, thus avoiding the situation of at-risk patients being screened up to age 64 through private insurers and then abruptly ceasing screening at exactly the ages when their risk for developing lung cancer is increasing. Legitimate concerns include false-positive findings that lead to further testing and invasive procedures, overdiagnosis (detection of clinically unimportant cancers), the morbidity and mortality of surgery, and the overall costs of follow-up tests and procedures. These concerns can be mitigated by clear criteria for screening high-risk patients, disciplined management of abnormalities based on algorithms, and high-quality multidisciplinary care. Lung cancer screening with low-dose CT can lead to early diagnosis and cure for thousands of patients each year. Professional societies can help CMS responsibly implement a program that is patient-centered and minimizes unintended harms and costs.

Medicare Advisers Say No to Lung Cancer Screening.

Annual low-dose CT lung cancer screening for high-risk individuals doesn’t have enough evidence for benefit over harms to be covered by Medicare, an advisory panel concluded.

The Centers for Medicare and Medicaid national coverage determination panel voted a mean 2.2 on a 5.0-point scale for confidence in that regard.

Key concerns were the high false-positive rate of CT screening, indication creep outside of the intended screening population, inability to assure quality scans with low radiation dose, and consistent interpretation and diagnostic workup in routine practice.

“If we don’t do it right now, it’s a genie that won’t be able to be stuffed back in the bottle,” Peter Bach, MD, MAPP, of Memorial Sloan-Kettering Cancer Center in New York City, who has been a prominent voice of caution in the national discussion, argued at the panel meeting.

The vote came in again at a low to intermediate level of 2.3 for confidence that harm could be minimized in the Medicare population.

The only thing the panel was confident about was that there are significant gaps in the evidence on how the screening would be used outside a clinical trial, with a mean vote of 4.4 on the 5.0-point scale.

The only reason that score wasn’t higher was that one of the voters flipped the scale and voted 2 when going for 4.


One of the main concerns cited was generalizability of the pivotal National Lung Screening Trial (NLST) upon which the U.S. Preventive Services Task Force (USPSTF) grade B recommendation was based and in turn prompted the national coverage determination.

Only a quarter of the trial population was 65 or older and none were enrolled past age 74.

Harms, particularly to those with chronic obstructive pulmonary disease (COPD) or other comorbidities, could be greater among older adults, Bach pointed out.

Lung cancer disproportionately affects older adults, both in terms of incidence and in deaths.

As of 2011, 14% of Medicare beneficiaries were current smokers and 44% were former smokers, Joseph Chin, MD, noted in introducing the topic on behalf of CMS at the meeting.

Pack-year smoking histories, upon which screening hinges, aren’t known.

But the best estimate is that about 9% of Medicare beneficiaries would be eligible for screening under the criteria recommended by the USPSTF, Bruce Pyenson, an actuary with the consulting firm Milliman, told the panel.

Subgroup analysis of the 65 and older population within the NLST suggested a lower and nonsignificant 13% reduction in lung cancer mortality, although without a significant interaction by age on the impact of screening.

Panel member Harry Burke, MD, PhD, an internal medicine clinician at the Walter Reed National Military Medical Center in Bethesda, Md., called this the biggest reason he voted no confidence in the evidence for the Medicare population.

Another issue was that the majority of patients in the NLST were screened at large academic medical centers by experienced radiologists adhering to protocols for performance.

The same 20% lung cancer mortality reduction is unlikely to be replicated in the community setting, Doug Campos-Outcalt, MD, MPA, chair of family, community, and preventive medicine at the University of Arizona in Phoenix, told the panel.

He represented the American Academy of Family Physicians, which disagreed with the USPSTF on the B recommendation and called for designation as I for insufficient evidence.

Community centers have increasingly been rolling out programs, and several spoke about their experience at the meeting, citing low rates of biopsy rivaling the NLST and coordination with primary care.

“Community based screening can be performed responsibly,” said Vickie Beckler, RN, lung screening coordinator at the Wellstar community hospital system in Marietta, Ga. “Please do not impose barriers to access.”

One awkward consequence of no Medicare coverage would be that coverage will end right as people are entering the period of highest lung cancer incidence, commented James Jett, MD, a pulmonologist in the oncology division at National Jewish Health in Denver.

Under the Affordable Care Act, private insurers will be required to cover the USPSTF-recommended screening without cost-sharing beginning Jan. 1, 2015.

“Older Americans 65 and older will not have that opportunity,” he told MedPage Today. “It’s a tremendously missed opportunity by Medicare.”

Further data for the 65 and older population could come from a registry, uncontrolled observational findings, and ongoing randomized but smaller European trials.

There is not likely to be another trial like the NLST, noted guest panelist Michael Gould, MD, director of health services research and implementation science at Kaiser Permanente Southern California.

“We’re accustomed to making recommendations, thumbs up or thumbs down, everyone should get it,” he said, but the lesson has been learned from prostate specific antigen screening for prostate cancer. Lung cancer is even more “a poster child for the situation where every individual has to weigh benefits and harms.”

Quality Assurance

The 96% false positive rate among the 27% of positive scans found in the NLST drew plenty of attention.

The LUNGRADS structured reporting and management system criteria should help fix that, argued Ella Kazerooni, MD, director of cardiothoracic radiology at the University of Michigan in Ann Arbor and a representative for the American College of Radiology.

She said the criteria would bring the false-positive rate down to one in 10, although the panel seemed unconvinced without seeing data to back it up.

Another concern is “extreme variability” in the false-positive rates between radiologists in the NLST, noted Paul Pinsky, MD, of the National Cancer Institute Division of Cancer Prevention and one of the researchers in the NLST.

It ranged from 10% or lower for some but up around 50% or higher for others even among the experienced radiologists in the trial, he pointed out.

There was no standard algorithm to say what to recommend for follow-up in the trial, he added.

That level of variation in interpretation in a relatively controlled environment raises serious concerns as it rolls out in the community, said panelist Jo Carol Hiatt, MD, MBA, of Kaiser Permanente.

Inappropriate Use

However, with such nuanced screening recommendations there’s no reason to believe a coverage decision would be implemented that way in the community, argued panelist Allan M. Fendrick, MD, of the University of Michigan School of Public Health in Ann Arbor.

CMS is still spending a billion dollars on screening 80-year-olds for colon cancer despite lack of evidence for benefit in that group, he noted. “I do not want this to be PSA,” he said. “I have reservations.”

There is a tendency for criteria to slip to extending the criteria, which means the risk-benefit ratio of the NLST would no longer apply, agreed guest panelist Steven H. Woolf, MD, MPH, director of the Center on Society and Health at Virginia Commonwealth University in Richmond.

He cited the sentiment to move to screening a younger age group, to not stop screening at 80, and to expand the pack-year and thresholds expressed by representatives of various organizations at the meeting.

Also, it would be pragmatically difficult to hold to such thresholds given self-reported smoking history, he added. “The feasibility of expecting that to be done nationwide is quite challenging.”

“I have serious concerns that coverage will lead to an explosion of inappropriate activities driven by a mix of good intentions and unrestrained entrepreneurialism,” warned Bach, although he did support coverage with requirement for a registry and qualifications for screening sites.

He criticized the Lung Cancer Alliance list of “trusted” sites, noting that 68% of that group’s Centers of excellence he reviewed followed neither the USPSTF nor the professional societies’ recommended criteria for screening.

Gould expressed doubt that any principles issued for screening could be assured to be followed going forward and that it shouldn’t be up to a group like the Lung Cancer Alliance to set those standards.

“Have confidence in the professional societies,” that group’s president and CEO Laurie Fenton Ambrose responded.

Despite the uncertainty that gave screening such low scores, panel vice chair Art Sedrakyan, MD, PhD, called himself convinced that there is a large subgroup of patients that would gain substantial benefit from low-dose CT lung screening.

“We just need to find it,” said Sedrakyan, director of the Patient Centered Comparative Outcomes Research Program at Weill Cornell Medical School in New York City.

The proposed decision is due in mid-November, followed by a 30-day comment period and a final determination 90 days after that.


Selection criteria for lung-cancer screening.

The National Lung Screening Trial (NLST) used risk factors for lung cancer (e.g., >/=30 pack-years of smoking and <15 years since quitting) as selection criteria for lung-cancer screening. Use of an accurate model that incorporates additional risk factors to select persons for screening may identify more persons who have lung cancer or in whom lung cancer will develop.
METHODS: We modified the 2011 lung-cancer risk-prediction model from our Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial to ensure applicability to NLST data; risk was the probability of a diagnosis of lung cancer during the 6-year study period. We developed and validated the model (PLCO(M2012)) with data from the 80,375 persons in the PLCO control and intervention groups who had ever smoked. Discrimination (area under the receiver-operating-characteristic curve [AUC]) and calibration were assessed. In the validation data set, 14,144 of 37,332 persons (37.9%) met NLST criteria. For comparison, 14,144 highest-risk persons were considered positive (eligible for screening) according to PLCO(M2012) criteria. We compared the accuracy of PLCO(M2012) criteria with NLST criteria to detect lung cancer. Cox models were used to evaluate whether the reduction in mortality among 53,202 persons undergoing low-dose computed tomographic screening in the NLST differed according to risk.
RESULTS: The AUC was 0.803 in the development data set and 0.797 in the validation data set. As compared with NLST criteria, PLCO(M2012) criteria had improved sensitivity (83.0% vs. 71.1%, P<0.001) and positive predictive value (4.0% vs. 3.4%, P=0.01), without loss of specificity (62.9% and. 62.7%, respectively; P=0.54); 41.3% fewer lung cancers were missed. The NLST screening effect did not vary according to PLCO(M2012) risk (P=0.61 for interaction).
CONCLUSIONS: The use of the PLCO(M2012) model was more sensitive than the NLST criteria for lung-cancer detection.


Annual Low-Dose CT Screening Better Than Chest X-Ray for Reducing Lung Cancer Mortality

Annual low-dose computed tomography cuts more lung cancer deaths than chest radiography among high-risk patients, according to early, as-yet unpublished findings from the National Lung Screening Trial released by the National Cancer Institute.

More than 53,000 current or former heavy smokers (aged 55 to 74) without signs or symptoms of lung cancer were randomized to undergo low-dose CT or chest radiography at baseline and then annually for 2 years. During follow-up, lung cancer mortality was 20% lower with CT than with radiography. (Based on this finding, the trial’s data and safety monitoring board recommended stopping the study.)

The NCI’s Dr. Christine Berg said: “This is the first time that we have seen clear evidence of a significant reduction in lung cancer mortality with a screening test in a randomized controlled trial. The fact that low-dose helical CT provides a decided benefit is a result that will have implications for the screening and management of lung cancer for many years to come.”