Cardiac CT Angiography and Radiation Dose

From Heartwire — a professional news service of WebMD

February 6, 2009 — A new American Heart Association (AHA) science advisory says there are no data to support the use of cardiac computed tomography angiography (CCTA) in asymptomatic, low-risk patients, so the albeit-small risk of cancer from radiation currently outweighs the lack of any demonstrable benefit in such people [1]. Dr Thomas C Gerber (Mayo Clinic, Jacksonville, FL) and colleagues discuss the subject in a paper published online February 2, 2009, in Circulation.

The advisory comes as a newly published trial in the February 4, 2009, issue of the Journal of the American Medical Association (JAMA), the Prospective Multicenter Study on Radiation Dose Estimates of Cardiac CT Angiography in Daily Practice (PROTECTION 1), shows that there is great variation between centers in the amount of radiation emitted when a CCTA is performed and in the use of dose-saving strategies to reduce radiation exposure [2]. As reported by heartwire, Dr Jörg Hausleiter (German Heart Center, Munich, Germany) first presented the PROTECTION-1 findings at the American College of Cardiology (ACC) meeting in Chicago last year.

Gerber, who is also a coauthor on the JAMA paper, told heartwire: "The main message of our advisory is that it's important to match the right type of test to the right type of patient. CCTA is not the right test for asymptomatic, low-risk patients, because you have the very small risk of potential cancer on the one hand and no outcome data to suggest any benefit on the other. However, if you move up the ladder to intermediate-risk patients, there is no question that CCTA (or nuclear perfusion stress tests) can be very helpful in those who have symptoms or at the very least are at high risk of heart artery disease."

Hausleiter has a slightly different take on the matter, however. He believes that CCTA may turn out to have benefits in asymptomatic patients, "but this needs to be proven." In the meantime, he agrees that "we are lacking data on the benefits of CCTA in asymptomatic patients."

Radiation Dose is Estimate, Link Between Dose and Cancer Unclear

Gerber said that an advisory is the lowest level of document, below guidelines or a statement, and indicates that there is "opinion" on a subject "but not a whole lot of data."

"This advisory is written for doctors, and we wanted to clarify a number of issues. First, doctors don't realize that the dose of radiation cannot really be measured — it's always a fairly crude estimate. What can be measured are certain parameters for CT-scanner radiation output, but this is not necessarily the same as the amount of radiation that the patient absorbs.

The second issue we wanted to clarify is the relationship between radiation dose and risk of cancer, which is the subject of ongoing debate. We're using a fairly conservative model and assuming a linear relationship from very high-dose levels — the kind of doses to which people were exposed in Hiroshima and Nagasaki — to the very low-dose levels we're talking about in medical imaging."

Assuming this linear relationship to be true, Gerber says the median radiation dose that was used until recently in CCTA — around 20 mSv — would potentially mean that one in 2000 exposed patients would develop a fatal cancer. However, he stresses that the median doses of radiation used in medical imaging are constantly being reduced due to new dose-saving technology.

"We are really trying to call on doctors to think about the potential risk vs the expected benefit for that patient. Although this is not what everyone wants to hear, we have no data yet that suggest that using CCTA in patients who have never had symptoms would be helpful in guiding management," he stresses.

Radiation Dose Not Affected by Procedure Volume, Center Expertise

In PROTECTION 1, which was observational and industry independent, Hausleiter and colleagues show a wide variation in radiation emitted by CCTA at the 50 sites examined worldwide. They analyzed 1965 procedures performed during one month, the majority of which were 64-slice. The main outcome was dose-length product (DLP) of CCTA; they also examined the use and efficacy of radiation dose-saving strategies (algorithms).

The DLP varied widely between study sites; it was used to estimate a median radiation dose of 12 mSv, which Hausleiter says is similar to that of an abdominal CT scan "that we order every day." He notes that the median dose given in the paper differs slightly from that reported at the ACC meeting last year — 15.4 mSv — due to a subsequent lowering of the conversion factor used to change the DLP to estimated radiation dose.

Independent factors associated with radiation dose included patient weight, absence of stable sinus rhythm, and varying use of dose-saving strategies.

Predictors for Estimated Radiation Dose in a Multivariate Linear Regression Analysis^a

Predictors	Effects (%)	p
Patient wt, 10-kg increase	5	< 0.001
Indication, noncoronary vs coronary	–1	0.31
Heart rhythm, nonsinus vs sinus	10	0.01
Heart rate, 10-bpm increase	1	0.98
Scan length, 1-cm increase	5	< 0.001
Automated exposure control	0	0.97
ECTCMb	–25	< 0.001
Tube voltage 100 kV vs ≥ 120 kV or greater	–46	< 0.001
Sequential vs spiral scanning	–78	< 0.001
Site experience in CCTA, 12-mo increase	–1	0.03
Performed CCTAs/mo, 10-CCTA increase	0	0.03
64-slice CT system vs Siemens single-source 64c
GE 64	97	< 0.001
Philips 64	11	< 0.001
Siemens dual-source 64	23	< 0.001
Toshiba 64	59	< 0.001

a. Predictors for radiation dose are presented as % change in DLP (mGy x cm).
b. Electrocardiographically controlled tube current modulation.
c. The Siemens single-source 64-slice CT system with the lowest median DLP in this study was used as a reference. The association with DLP is shown for the remaining four 64-slice systems within the linear regression analysis.

Hausleiter told heartwire he was surprised by the variation in radiation dose, both by site and by the machine employed. "We cannot explain the differences by expertise or by volume, so there is still potential to reduce the dose at a lot of sites," he notes.

But Gerber said he was not at all surprised by these findings. "There are a whole lot of confounders in there that are difficult to account for, and we shouldn't use a one-size-fits-all approach to protocols for CTs." For example, he says, for very obese patients, the additional tissue means that operators have to double the output of radiation to achieve the same quality of images, "so a center that images mainly obese patients will come out with a higher average reference level [of radiation] than a center that images mainly patients with normal body weight."

Another example is coronary artery bypass grafting, "where we have to image a much larger portion of the chest than if we are just imaging the heart, so that results in more radiation." And many dose-saving techniques cannot be used in patients with very fast or irregular heart rates, so "depending on the patients, some centers have no opportunity to use these algorithms."

Most Centers Used Dose-Saving Strategies, But Still Room for Improvement

Nevertheless, both Hausleiter and Gerber point out that the majority of the centers in PROTECTION 1 — over 70% — did use strategies for dose-saving. "This is a healthy message," says Gerber. "People are aware of these dose-saving algorithms, and they are being used." And this figure will likely improve, says Hausleiter, as people become more aware of dose-saving strategies and newer machines are employed that make greater use of such technologies.

In an editorial accompanying the JAMA paper [3], Dr Andrew J Einstein (Columbia University Medical Center, New York, NY) also applauds the fact that dose-reduction techniques were used in the majority of patients in PROTECTION 1 and says these results "should serve as a wake-up call to those cardiac CT labs that do not routinely use these methods."

Addressing specific dose-reduction strategies, Einstein says: "Given the strength of evidence supporting it, electrocardiographically controlled tube current modulation (ECTCM) should be widely applied; the evidence for sequential scanning is rapidly accumulating, and it should also be given serious consideration for appropriate patients. Low-voltage scanning should also be considered, perhaps especially for patients who are nonobese and at higher risk of radiation-associated cancer, such as children and young women."

Einstein adds that the variability between sites "that had not previously been appreciated" offers the potential to decrease radiation burden while maintaining image quality "by instituting quality-improvement programs to close the gap." This is something that high-volume centers can also learn from, he notes, pointing out the lack of association between procedure volume and dose.

Hausleiter told heartwire he expects to present further data at the ACC meeting this year, from PROTECTION 2 and 3, which are examining whether image quality is maintained despite the use of lower doses of radiation.

The study was supported solely by an unrestricted research grant from Deutsches Herzzentrum München, Klinik an der Technischen Universität, Munich, Germany. None of the participating physicians received any compensation for study participation. Dr. Hausleiter and coauthors Cynthia McCollough, PhD, and Stephan Achenbach, MD, have received research grants from Siemens Medical Systems unrelated to the current study. Dr. Achenbach was supported by a grant from the German Bundesministerium für Bildung und Forschung. The other study authors have disclosed no relevant financial relationships.

Dr. Einstein has served as a consultant for GE Healthcare, has received travel funding from GE Healthcare, INVIA, Philips Medical Systems, and Toshiba America Medical Systems, and has received support for previous research through a grant funded by Covidien and collaboration with employees of Siemens Medical Solutions. He is supported in part by a National Institutes of Health K12 Institutional Career Development Award.

Sources

1. Ionizing radiation in cardiac imaging. A science advisory from the American Heart Association Committee on Cardiac Imaging of the Council on Clinical Cardiology and Committee on Cardiovascular Imaging and Intervention of the Council on Cardiovascular Radiology and Intervention. Circulation. 2009; DOI: 10.1161/CIRCULATIONAHA.108.191650. Available at: http://circ.ahajournals.org.
2. Hausleiter J, Meyer T, Hermann F, et al. Estimated radiation dose associated with cardiac CT angiography. JAMA. 2009;301:500-507.
3. Einstein AJ. Radiation protection of patients undergoing cardiac computed tomographic angiography. JAMA. 2009;301:545-547.

The complete contents of Heartwire, a professional news service of WebMD, can be found at www.theheart.org, a Web site for cardiovascular healthcare professionals.

Clinical Context

Although CCTA is a useful diagnostic imaging modality to evaluate coronary artery disease, there is concern regarding potential risks associated with exposure to ionizing radiation. Selected indications include low-to-intermediate pretest probability for obstructive coronary artery disease and rapid evaluation of chest pain in patients seen in the emergency department.

Many clinicians may still be unaware of the magnitude of radiation exposure associated with CCTA and with the factors that independently contribute to radiation dose. This information is clearly needed to plan and implement strategies to minimize patient exposure to ionizing radiation without sacrificing image quality.

Study Highlights

• The goals of this study were to estimate the radiation dose of CCTA in routine clinical practice, to evaluate the association of currently available strategies with dose reduction, and to identify the independent factors contributing to radiation dose.
• This cross-sectional, international, multicenter, observational study took place at 50 study sites, including 21 university hospitals and 29 community hospitals.
• Radiation dose was estimated for 1965 patients who underwent CCTA between February and December 2007.
• Independent predictors associated with dose were determined by linear regression analysis.
• The primary study endpoint was the DLP of CCTA.
• Median DLP was 885 mGy x cm (interquartile range, 568 - 1259 mGy x cm). This corresponds to an estimated radiation dose of 12 mSv, which is comparable with 1.2 times the dose of an abdominal CT scan, more than twice that of invasive coronary angiography, or 600 chest radiographs.
• There was a high variability in DLP between study sites (range of median DLPs per site, 331 - 2146 mGy x cm).
• This 6-fold difference in DLP reflects the large variability in CCTA protocols, differences in CT system characteristics, and use of dose-reduction algorithms among individual study sites.
• Independent clinical factors associated with radiation dose were patient weight (relative effect on DLP, 5%; 95% confidence interval [CI], 4% - 6%) and absence of stable sinus rhythm (relative effect on DLP, 10%; 95% CI, 2% - 19%).
• Technical factors associated with radiation dose were scan length (relative effect on DLP, 5%; 95% CI, 4% - 6%), electrocardiographically controlled tube current modulation (relative effect on DLP, −25%; 95% CI, −23% to −28%; applied in 73% of patients), 100-kV tube voltage (relative effect on DLP, −46%; 95% CI, −42% to −51%; applied in 5% of patients), and sequential scanning (relative effect on DLP, −78%; 95% CI, −77% to −79%; applied in 6% of patients).
• Site factors associated with radiation dose were experience in cardiac CT (relative effect on DLP, −1%; 95% CI, −1% to 0%), number of CCTAs per month (relative effect on DLP, 0%; 95% CI, 0% - 1%), and type of 64-slice CT system (for highest vs lowest dose system, relative effect on DLP, 97%; 95% CI, 88% - 106%).
• In this study, algorithms for dose reduction were not associated with deteriorated diagnostic image quality, which may support their use in adequately selected patients.
• The investigators concluded that median doses of CCTA varied significantly among study sites and CT systems. Despite the availability of effective strategies to reduce radiation dose, some strategies are infrequently used.
• Limitations of the study include lack of universal standardization of the definition of tube current, and radiation dose associated with coronary calcium scoring not evaluated.

Pearls for Practice

• Median estimated radiation dose for CCTA was 12 mSv, which is comparable with 1.2 times the dose of an abdominal CT, more than twice that of invasive coronary angiography, or 600 chest radiographs. There was a high variability in DLP among study sites, reflecting the large variability in CCTA protocols, differences in CT system characteristics, and use of dose-reduction algorithms among individual study sites.
• Despite the availability of effective strategies to reduce radiation dose, some strategies are infrequently used. Independent factors associated with radiation dose were patient weight, absence of stable sinus rhythm, modifiable technical factors related to the scanning procedure, and site-specific factors.