OUP user menu

Combined detection of coronary artery disease and lung cancer

Harvey S. Hecht, Claudia Henschke, David Yankelevitz, Valentin Fuster, Jagat Narula
DOI: http://dx.doi.org/10.1093/eurheartj/ehu296 2792-2796 First published online: 11 August 2014

Introduction

Coronary artery disease (CAD) and lung cancer have several important features in common. First, their dramatic increases are in large part attributable to societal ills, including worsening dietary patterns, obesity, and tobacco use. Secondly, as these behaviours permeate the world, the diseases are disproportionately increasing in the poorer societies with limited resources for healthcare. Consequently, it is necessary to develop cost-effective strategies. Both disease states may be amenable to early detection by a single low radiation dose CT scan.

Early detection of coronary artery disease: guidelines and general considerations

Gated multidetector computed tomographic imaging of coronary artery calcium (CAC) is a robust strategy for the early detection of CAD in asymptomatic patients that has performed superior to risk factor-based paradigms such as the Framingham Risk Score (FRS) and the European Society of Cardiology Score.1 Notwithstanding these data, the US Preventive Services Task Force (USPSTF) in 2009 concluded that the current evidence was insufficient to assess the balance of benefits and harms of using non-traditional risk factors, including CAC scans, to screen asymptomatic men and women with no history of CAD to prevent coronary events.2 However, in 2010, CAC assessment was incorporated into ACC/AHA Guidelines with a Class IIa status (recommendation in favour of treatment or procedure being useful/effective). Measurement of CAC was considered reasonable for cardiovascular risk assessment in asymptomatic adults at intermediate risk, and all diabetic patients 40 years or older.3 The ACCF/SCCT/ACR/AHA/ASE/ASNC/NASCI/SCAI/SCMR 2010 appropriate use criteria deemed CAC appropriate for intermediate-risk patients as well as for low-risk individuals with a family history of premature disease.4 In 2012, the European Society of Cardiology awarded a similar Class IIa recommendation, and suggested CAC for cardiovascular risk assessment in asymptomatic adults at moderate risk.5

In an attempt to simplify cholesterol treatment, the 2013 ACC/AHA Guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults6 restricted the use of CAC to the patients who did not fall into four designated conventional risk factor-based categories. Coronary artery calcium was downgraded to a Class IIb status (recommendation's usefulness/efficacy less well established),7 ignoring the robust CAC literature that initially engendered the Class lla recommendation, and the more recent data from three prospective, population-based outcome trials that the CAC Net Reclassification Index of the FRS is extremely high, particularly for the intermediate-risk group (52–66%) (Table 1).810 The guidelines are summarized in Table 2.

View this table:
Table 1

Reclassification of Framingham Risk Score risk by coronary artery calcium: primary prevention outcome studies

Study% ReclassifiedNAgeFollow-up (years)
MESA8
 FRS 0–6%11.6587862.25.8
 FRS 6–20%54.4
 FRS > 20%35.8
 NRI25
Heinz Nixdorf9
 FRS < 10%15.0448745–755.0
 FRS 10–20%65.6
 FRS > 20%34.2
 NRI22.4
Rotterdam10
 FRS < 10%12202869.69.2
 FRS 10–20%52
 FRS > 20%34
 NRI19
  • NRI, Net Reclassification Index.

View this table:
Table 2

Coronary artery calcium guidelines and appropriateness criteria

PopulationRecommendation
2009 USPSTF2NAC
2010 ACC/AHA Risk Guidelines310–20% intermediate riskIIa
Diabetics >40 years oldIIa
6–10% low-to-intermediate riskIIb
2010 Appropriate Use Criteria410–20% intermediate riskAppropriate
Low risk with family history of premature coronary diseaseAppropriate
High riskUncertain
Low riskInappropriate
2012 ESC Risk Guideline5Intermediate riskIIa
2013 ACC/AHA Cholesterol and Risk Guidelines6,7NAIIb

In addition to the early detection, CAC has been shown to increase adherence to statin and ASA treatment, to diet and exercise1113 and to improve lipids, BP, and weight.14 With adequate evidence that CAC appropriately detects high-risk patients and that treatment of high-risk patients with statins improves their outcomes,15 it would seem reasonable to connect the two premises and conclude that CAC would improve outcomes. Based on primary prevention trials, the reduction in events would be expected to be ∼30%.15

Critics of CAC argue that the effect of CAC on outcomes in randomized-controlled trials has not been demonstrated. In addition, risk factor-based stratification, such as the FRS, is inexpensive, universally available, and requires no technology other than simple lipid, blood sugar, and blood pressure measurements. While attractive on the surface, unfortunately, the risk factor-based paradigms have also not been validated by randomized-controlled outcome trials, do not improve adherence, and remain inferior to CAC for risk prediction in every study.

Early detection of lung cancer: guidelines and general considerations

In contrast to CAC, low-dose CT scanning for lung cancer screening has been supported by outcome studies, including the International Early Lung Cancer Action Program (IELCAP)16 and the National Lung Screening Trial (NLST).17 The IELCAP study demonstrated an 80% cure rate for patients diagnosed early by CT screening.16 The NLST enrolled 53 454 high-risk participants to receive three rounds of either CT or chest radiography screening, and 5–7 years of follow-up from the time of random assignment to either screening.17 After achieving its stated goal of a 20% reduction in mortality in the CT compared with the radiographic arm, a different outcome parameter than cure rate, the trial was terminated and scanning was recommended for high-risk patients by the National Comprehensive Cancer Network,18 the American College of Chest Physicians and the American Society for Clinical Oncology,19 the American Cancer Society,20 the American Association for Thoracic Surgery and the Society of Thoracic Surgeons,21 and the American Lung Association22 (Table 3). In recognition of the compelling data, the USPSTF created a Grade B (net benefit is moderate or there is moderate certainty that the net benefit is moderate to substantial) recommendation for annual low-dose screening in persons at high risk for lung cancer based on age and smoking history, including a 30 pack-year or more history of smoking in subjects aged 55–79 who have smoked within the past 15 years.23 The expected mortality reduction is likely to exceed the 20% reported in the NLST, which was terminated as soon as the 20% criterion was met. It was not designed to quantify the maximum mortality reduction and is subject to the downward bias of the true mortality reduction characteristic of such trials.24 Post NLST, there are seven ongoing European-randomized trials with a combined enrolment of 37 000 patients, with projected mortality reductions in the CT arm of 15–45% depending on the particulars of the study.25,26 Screening will very likely extend to lower risk categories as well. In Japan, for example, never-smokers and smokers as young as 40 years of age have been enrolled in screening programmes. Increasing evidence for the pathogenicity of second hand smoke exposure may result in its inclusion for screening eligibility.

View this table:
Table 3

Low-dose lung scan guidelines

AgePack-yearsWithin past (years)
National comprehensiveCancer network1850–74≥3015
≥20 with additional15
risk factor
American College of Chest Physicians and American Society for Clinical Oncology1955–74≥3015
American Cancer Society2055–74≥3015
American Association for55–79≥3015
Thoracic Surgery2150–79≥20 with 5%15
5-year risk
American Lung Association2255–74≥3015
US Preventive55–79≥3015

Combined low-dose heart and lung scanning: a proposal

Since the heart is always visualized on lung scans, it is difficult to argue against performing CAC analysis and ignore the opportunity to identify high-risk patients. Conversely, the lung nodule findings have always been and will continue to be evaluated on CAC scans, despite the exaggerated concern about incidental findings. However, because of the limited field of view, encompassing only 40% of the lungs, CAC scans are insufficient for effective lung cancer screening. The pre-requisite of a combined scan is straightforward: the accuracy of separate scans must be preserved at a low radiation dose.

Patient selection

Patients who meet both CAC and lung scan criteria or lung scan criteria alone should be eligible for a combined full chest low-dose-gated CT scan. In reality, almost all patients who fulfil lung screening criteria by USPSTF, estimated to be 7 000 000 in the USA,27 are also at least intermediate risk for cardiac events because of their advanced age and long-standing smoking history, and will therefore be candidates for both evaluations (Figure 1). Because of the poor penetration of CAC scanning into the diagnostic algorithm to date, lung scan patients will not have the opportunity to benefit from CAC evaluation were it not for the lung scan. In CAC candidates who do not meet lung scanning criteria, acquisition should be limited to the heart. Despite the low radiation dose (to be discussed below), full chest scanning is not appropriate in those who do not fulfil the lung criteria. Coronary artery calcium scans performed as part of coronary computed tomographic angiography should encompass the full chest if lung scan criteria are fulfilled.

Figure 1

USA estimates, and overlap, of coronary artery calcium and lung scan eligible patients. The number of eligible patients in the USA is estimated at 33 million for coronary artery calcium scanning (orange)28 and 7 million for lung scanning (yellow).27 Excluding lung scan eligible patients who have established coronary disease (5.3%, unpublished data from the I-ELCAP database) yields an overlap of 6.6 million lung scan patients who would be expected to benefit from coronary artery calcium scanning.

Scan acquisition

Lung CT scanning is routinely performed without EKG gating, whereas CAC has always been analysed on gated scans to minimize motion artefact; a minimum of eight slice MDCT is recommended for accurate scoring. Coronary artery calcium is apparent on non-gated chest CT screening scans and several analytic approaches have been employed. Coronary artery calcium Agatston scoring of non-gated scans has correlated well with the standard of care scoring on gated scans. However, there is a higher inter-scan variability, 20% underestimation of high-risk patients and 8% false negatives.29 Ordinal scoring based on a semi-quantitative analysis has correlated well with cardiac outcomes30 but is subjective, has no reproducibility data and no correlative database to CAC scoring. Recently, an automatic technique for analysing non-gated scans has been described.31 These alternative methods are not utilized by the cardiology community. Therefore, ideally, combined lung and CAC scans should fulfil minimum CAC criteria, necessitating a shift in lung scan technology by limiting acquisitions to ≥8 slice scanners equipped with EKG gating and CAC analysis software. At least in developed countries, such CT scanners are readily available. Achieving greater penetration in technology poor, less developed countries may require utilization of non-gated scans despite the limitations. An example of gated and ungated combined low-dose lung and CAC scans is displayed in Figure 2.

Figure 2

Combined gated heart and lung scan. A 65-year-old asymptomatic male smoker with 40-pack-year history and hyperlipidaemia underwent combined scanning. Images were prospectively acquired in a step and shoot mode on a 256 slice scanner at 120 kV and 25 mAs, with 3 mm slice thickness and radiation exposure of 0.95 mSv. Left: Calcium scan demonstrating extensive calcified coronary plaque in the left coronary artery (pink). The total Agatston calcium score was 1467. Right: Lung window reconstruction reveals a 3 mm left lower lobe nodule (green arrow). Bottom: EKG gating signal (yellow dot on R wave).

Radiation

The literature and lay press are replete with warnings of cancers attributable to routine cardiac and lung CT scanning, all of which are based on hypothetical extrapolation, predominantly from atomic explosions, and not based on actual data from the scanning itself. In response, the American Association of Physicists in Medicine has affirmed that the risks of medical imaging at effective doses <50 mSv for single procedures or 100 mSv for multiple procedures over short-time periods are too low to be detectable and may be non-existent.32 Nonetheless, we are in full accord with the as low as reasonably achievable mandate, and with the ESC principle that ‘each patient should get the right imaging exam, at the right time, with the right radiation dose’.33

Low-dose lung scans in the average NLST patient were 1.4 + 0.5 mSv34 and CAC scans have progressively decreased to the 1 mSv level.35 However, recently developed iterative reconstruction algorithms available on all the latest generation scanners will facilitate the acquisition of sub 0.5 mSv lung scans and CAC studies without loss of accuracy. Model-based reconstruction algorithms under development are expected to further decrease radiation. Extrapolating the CAC scan radiation to full chest coverage will yield <1 mSv combined scans, in the same ballpark as screening mammography.36 Less developed countries may not have access to the latest dose reducing algorithms in which case the combined scan radiation would be higher but not likely to exceed 2 mSv.

Cost effectiveness

Analyses of cost effectiveness suffer from a lack of uniformity of evaluation and varying assumptions of costs and benefits that are constantly changing as potentially more expensive new technologies and treatments are introduced and older, still applicable ones become less expensive. Perhaps in recognition of this dilemma, cost-effective analyses are not part of the guidelines and screening justifications. Their complexity is undoubtedly greatly magnified in the less developed countries. Using a single low-dose CT scan for two major diseases should be more cost effective than using two separate CT scans, but needs to be demonstrated.

Conclusions

The imperative to contain the growing global burden of CAD and lung cancer is clear, as is the complexity of the problem. Identification of at risk patients is not sufficient but would eventually form a pre-requisite for implementation of appropriate intervention. To accomplish this goal, low-dose lung screening has received the imprimatur of the USPSTF and will hopefully be adopted globally in similar if not identical form. Analysis of lung scans for CAC seems appropriate27 and has the potential to be the standard of care, ideally by gated techniques or non-gated if there is no alternative. However, the global burden of CAD far exceeds that of lung cancer in the number of patients and costs to society, and CAC analysis of lung scans, while an important first step, would address only a small part of the CAD burden.

Whether it is reasonable to directly target the entire appropriate use population by widespread CAC scanning without waiting for an official ‘screening’ label remains to be debated. Such a recommendation would be antithetical to many, whereas others emphasize that to optimize patient care in light of the available evidence, screening guidelines should not be confined to the necessity of the randomized control trial.

Numerous questions remain to be answered. A detailed analysis of available CT scanners and the costs and logistics of screening, particularly in less privileged countries, are required, as are the programmes for implementing intervention and adherence. Of equal importance is a fundamental change in mind-set in the prevention community to accept the use of the best available tools as the first step in easing the global burden of disease. Waiting for outcome trials for CAC which are unlikely to be performed, or for replication of the NLST results, especially in developing countries, would incur an immeasurable delay. At the present time, it is logical and reasonable that gated CAC scanning be performed on lung scans and that the entire thorax be imaged during CAC scans in those who meet recommendations for both evaluations.

Funding

Flight Attendant Medical Research Institute.

Conflicts of interest: Harvey Hecht: Philips Medical Systems Consultant; Claudia Henschke: None declared; David Yankelewitz: patents for measurement of chest nodules; Valentin Fuster: None declared; Jagat Narula: Philips Medical Systems and General Electric research grants.

Footnotes

  • The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.

References

View Abstract