Introduction: Coronary CT angiography (CCTA) can safely disposition low to intermediate risk chest pain (CP); however, there is conflicting data with respect to cumulative radiation exposure when compared with usual care over short follow-up intervals.

Objectives: We report the effective radiation dose from index and downstream testing in low to intermediate risk symptomatic patients evaluated for chest pain in the ED with either CCTA or usual care to define various sources of patient radiation dose and quantify effective dose over a year and a half of follow-up.

Patients and Methods: We evaluated radiation exposure from initial and downstream testing in a prospectively collected, matched cohort evaluated for CP in the emergency department (ED) with either CCTA compared with usual care over a median follow-up of 19.6 months. Effective radiation dose was calculated using published conversion factors.

Results: Prospective, ECG-triggered acquisition using a 128-slice dual-source multidetector computed tomography (DSCT) scanner was performed in 92.9% of scans with a median effective dose from CCTA of 6.8 mSv (IQR 5.2, 9.1 mSv). CCTA cohort patients were more likely to undergo cardiac testing with exposure to radiation (P < 0.001); however, the median effective dose in patients exposed to radiation from cardiac testing was significantly lower in the CCTA cohort (7.1 mSv vs. 11.8 mSv, P < 0.001). Fewer patients in the CCTA cohort had additional non-cardiac thoracic imaging radiation exposure (40.8%) compared with usual care (92.8%). Total radiation exposure from any source was similar between the CCTA and usual care groups (100% vs 98.4%, P = 0.087), as was median total effective radiation dose (P = 0.105). Upfront CCTA was not associated with higher rates of incidental non-cardiac findings.

Conclusions: Initial evaluation of acute chest pain in the ED with CCTA was not associated with an increase in total radiation exposure over a follow-up period of 19 months. CCTA offers a more comprehensive evaluation of multiple thoracic organ systems leading to reduced radiation exposure from non-cardiac thoracic testing and no increase in incidental imaging findings. This may represent an added benefit in this population of patients presenting acutely.

Coronary computed tomography angiography (CCTA) is a well validated modality to rapidly define the presence and severity of coronary artery disease (CAD) in symptomatic patients (1). Specifically, there is an ever-growing body of data supporting implementation of CCTA to evaluate symptomatic patients presenting to the emergency department (ED) due to its very high negative predictive value (NPV) (2-5). Patients with coronary stenosis < 50% can be safely discharged home with near zero major adverse cardiovascular event (MACE) rates (2-5). Additionally, upfront CCTA in this population demonstrated a decrease in hospital admissions and a decrease in short-term healthcare costs (6). Despite the strength of this data, concern still exists that exposing patients with inherently lower risk chest pain symptoms to ionizing radiation may be unnecessary. Furthermore, many centers lack ready access to CCTA and thus rely more heavily on functional stress testing that does not involve radiation. With increasing emphasis on patient-centered imaging where patients are empowered to understand the risks and benefits of diagnostic medical imaging involving ionizing radiation, longer follow-up periods where total radiation exposure is reported is imperative (7). Most major professional societies support a linear-no-threshold model with respect to the risk of developing malignancy secondary to radiation exposure (8). Additionally, the most recent Biological Effects of Ionizing Radiation (BEIR VII) report describes a lifetime attributable risk (LAR) based on Japanese and survivors of the atomic bomb who were exposed at age 30 and survived to age 60 (9). Einstein et al. estimated lifetime attributable risk from 64-slice CCTA based on the BEIR VII model and determined the lifetime cancer risk for a 20-year old woman is 1:143 with lower attributable risk with increasing age (10) utilizing CCTA protocols with effective radiation doses between 9 - 21 mSv. This effective dose is at least 3 times greater than the average background radiation exposure from natural sources in the United States. Young women were at highest risk for breast cancer while men were more likely to develop lung cancer as a result of cardiac imaging (10). With ever improving CT scanner technologies and expertise, the effective radiation dose from CCTA continues to fall with data of sub-millisievert scan acquisitions (11, 12).

We report the effective radiation dose from index and downstream testing in low to intermediate risk symptomatic patients evaluated for chest pain in the ED with either CCTA or usual care to define various sources of patient radiation dose and quantify effective dose over a year and a half of follow-up.

3.1. Patient Selection

This is a retrospective analysis of a prospectively identified cohort of patients who presented to the emergency department (ED) with acute chest pain. CCTAs were acquired using a 128-slice dual-source multidetector computed tomography (DSCT) scanner between January 2013 through December 2013 at a single center, tertiary referral hospital (San Antonio military medical center, Joint Base San Antonio-Fort Sam Houston, Texas).

A total of 366 patients presenting to the ED with acute chest pain were analyzed with 182 patients evaluated using CCTA in the ED and 184 patients evaluated in a usual fashion. The median follow-up for the entire cohort was 19.6 months. Table 1 summarizes the baseline demographic data for the entire population, as well as the CCTA and usual care cohorts. The median age at the time of evaluation was 48 years (IQR 42, 57 years) and there were no significant differences between the groups with respect to age, gender, body mass index (BMI), CAD risk factors, nor TIMI risk scores.

Table 1. Baseline Demographic Data and Effective Radiation Dosea

Specifics of the CCTA protocols used in the CCTA cohort are outlined in Table 2. All patients were imaged using prospective, ECG-triggered axial imaging (92.8%) or prospective, ECG-triggered high-pitched helical (8.2%) acquisitions. Both CCTA and CAC scores were obtained in 91.2% of patients. A median patient BMI in the CCTA cohort of 30.1 lb/in² (IQR 26.0, 33.9 lb/in²) resulted in 96.7% of patients being scanned utilizing a tube voltage of 120 kVp. Median tube current was 786 mA (IQR 630, 929 mA) and median padding was 615 ms (IQR 503, 686 ms). The vast majority of CCTAs resulted in diagnostic image quality with a modest 3.6% of the segments being determined not interpretable.

Table 2 . CCTA Cohort Detailsa

Obstructive CAD (≥ 50% stenosis) was present in 6%, nonobstructive CAD (1% - 49% stenosis) was present in 35.7%, and CAD was absent (no stenosis, CAC of 0) in 58.2%. In the CCTA cohort, obstructive CAD was confirmed in 3 patients (75% true positive) by ICA. All three of these patients proceeded to revascularization (2 via percutaneous coronary intervention (PCI) and 1 via coronary artery bypass grafting (CABG). Revascularization was deferred in the fourth patient following a non-significant fractional flow reserve (FFR) of 0.86. Nine ICAs were performed in 8 patients in the usual care cohort. Of the 9 ICAs, 5 were found to have no significant CAD, 2 patients had non-obstructive CAD (stenosis < 70%), and 2 patients underwent PCI for single vessel obstructive CAD (22% true positives).

Median effective dose attributable to CCTA was 6.0 mSv (IQR 4.4, 8.0 mSv), CAC was 0.85 mSv (IQR 0.69, 1.1 mSv), and total scan was 6.8 mSv (IQR 5.2, 9.1 mSv). There was no difference in effective radiation dose from cardiac imaging (Figure 1 A) between patients with obstructive CAD (median 8.4 mSv (IQR 4.9, 40.0 mSv) compared with nonobstructive CAD and no CAD (median 7.1 mSv (IQR 6.0, 9.4 mSv) and 6.7 mSv (IQR 4.6, 8.5 mSv), P = 0.156). Additionally, median total radiation dose (Figure 1B) did not differ significantly based on CAD burden (P = 0.336).

Figure 1. Box Plot Depicting Median Effective Radiation Dose Attributable to A, Cardiac Imaging and B, Total Medical Imaging in the CCTA Cohort as a Function of Coronary Artery Disease (CAD) Burden

Total median effective radiation dose (Table 1) over the 19.6 month follow-up was not significantly different in the CCTA cohort (9.1 mSv (IQR 6.2, 17.0) compared with the usual care cohort (13.1 mSv (IQR 4.7, 29.2), P = 0.105). Likewise, there was no difference between the groups with respect to the total percentage of patients who underwent any diagnostic testing involving radiation (CCTA 100% vs usual care 98.4%, P = 0.087). More patients in the CCTA cohort (Table 1) underwent cardiac testing (55.6% vs 47.2%, P = 0.025). Patients in the CCTA cohort were also more likely to undergo cardiac testing involving ionizing radiation (100% vs 55.2%, P < 0.001). However, of the patients in each group receiving cardiac testing involving radiation, the median effective dose was lower in patients in the CCTA arm (median: 7.1 mSv (IQR 5.3, 9.3) vs median 11.8 mSv (6.0, 13.7), P < 0.001). By comparison, patients in the usual care cohort were far more likely to undergo non-cardiac thoracic imaging (Table 1) than patients initially evaluated with CCTA (77.8% vs 22.2%, P < 0.001). A significant increase was also observed in the percentage of patients receiving ionizing radiation from a non-cardiac thoracic test in the usual care cohort (92.8% vs 40.8%, P < 0.001), though the median effective radiation dose was modest in both groups (usual care 0.18 (IQR 0.06, 11.2) vs CCTA 0.06 (IQR 0.06, 0.18), P < 0.001). Despite modest median radiation dose, 25.1% of the total effective radiation dose in the usual care cohort was attributable to non-cardiac thoracic imaging compared with 5.6% in the CCTA cohort (P < 0.001).

Table 3 summarizes effective radiation dose grouped by age. There was no significant difference between the 2 groups in the percentage of patients in their 30s, 40s, 50s, 60s, or 70s that were exposed to any amount of radiation due to medical imaging. Excluding patients in their 70s, CCTA cohort patients in all of these age groups were more likely to receive radiation from a cardiac imaging source (P < 0.001 in all age groups). Patients in their 40s and 60s in the CCTA cohort received a lower median dose from cardiac testing than usual care (P = 0.003 and P = 0.001, respectively). No significant difference in total effective radiation dose was observed between the 2 groups.

Table 3. Effective Radiation Dose and Radiation Exposure by Decade of Agea

Incidental findings (Table 4) associated with all index and downstream testing were not significantly different between the groups with the exception of incident bony lesions. Bone lesions were more commonly observed incidentally in patients in the usual care cohort (2.7% vs 0%, P = 0.022). Additionally, 3 patients in the CCTA cohort were diagnosed with alternate, non-cardiac etiologies of chest pain due to CCTA (2 patients with pneumonia and 1 patient with a pulmonary embolus).

Table 4. Incidence of Incidental Imaging Findingsa

A strategy of CCTA using DSCT in the ED to evaluate low to intermediate risk acute chest pain patients was not associated with an increase in the incidence of radiation exposure nor total median radiation dose when compared with a propensity matched, usual care cohort over a follow-up period of 19 months. Additionally, the median effective dose secondary to cardiac imaging was lower in the CCTA cohort despite this cohort being comprised primarily of overweight patients (median BMI 30.1 lb/in²) and predominantly utilizing 120 kVp imaging protocols (96.7%). Additional thoracic imaging was more commonly obtained in the usual care cohort accounting for the vast majority of effective radiation dose received in this cohort. These findings, coupled with a previously published analysis from this same population demonstrating that use of CCTA in the ED was associated with more rapid ED disposition times (median 5.9 hours vs. 25.0 hours, P < 0.001), lower hospital admission rates (9.3% vs 98.9%, P < 0.001), and lower total payer cost ($ 182,064.55 vs $ 685,190.77, P < 0.001) with no increased risk to patients, strengthens the argument for the routine use of CCTA in the ED (5). This finding may be explained by the fact that CCTA is a more comprehensive testing modality for the evaluation of acute chest pain patients in the ED. In addition to being the most sensitive modality for the detection of CAD, the scan volumes routinely include the vast majority of the lung fields and routinely allow for visualization of the proximal pulmonary arteries and a large portion of the thoracic aorta.

Large randomized trials utilizing a CCTA strategy for the evaluation of acute CP report mixed findings with respect to radiation exposure and downstream testing over short term follow-up. The rule out myocardial infarction using computer assisted tomography (ROMICAT) II randomized 1000 patients with possible acute coronary syndrome (ACS), a non-ischemic ECG, and negative cardiac biomarkers to CCTA or standard evaluation in the ED (2). Over a relatively short follow-up period of 28 days, CCTA reduced ED length of stay and increased ED patient discharge with no difference in adverse cardiovascular outcomes. Similar to our findings, these authors reported an increase in radiation exposure (97% vs 33% with standard evaluation), but lower mean effective dose in the CCTA group (11.3 mSv for CCTA vs 14.1 mSv for SPECT, P < 0.001). In ROMICAT II, over 70 patients were imaged using 128-slice DSCT technology resulting in half of the effective radiation dose when compared to those who were imaged using older technology (6.2 ± 3.8 mSv vs 12.3 ± 5 mSv). Thus, as CT scanner technology continues to improve, the effective radiation dose gap between SPECT and cardiac CT is likely to grow.

Observational data from the same group in ROMICAT I found that over 40% of patients with no CAD and over 51% of patients with non-obstructive CAD by CCTA (performed in a blinded fashion prior to admission) underwent evaluation with either SPECT imaging or ICA with the associated procedural risk and significantly higher effective radiation dose (13, 19). Thus, in centers where CCTA is not available, a large portion of patients without significant CAD are undergoing unnecessary invasive evaluation or work-up with a less sensitive modality with a higher effective radiation dose than that offered by CCTA. Recently published data from our group suggests the absence of obstructive CAD on CCTA (defined as stenosis ≥ 50%) resulted in lower downstream cost and utilization of additional ischemic testing during subsequent chest pain evaluations (20). This is likely explained by the recent findings from SCOT-HEART where CCTA increased diagnostic certainty when compared to clinical evaluation in combination with exercise stress testing (21). Thus, upfront radiation exposure from CCTA may be offset by reduced downstream radiation exposure over a longer term follow-up period, particularly in patients with no and non-obstructive CAD.

Similar to previously published data, our CCTA cohort was more likely to receive radiation secondary to cardiac imaging with lower median effective radiation dose attributable to cardiac imaging. Some data suggest increased rates of incidental, non-cardiac findings that do not assist in diagnosis but add to downstream testing and radiation exposure. However, we observed no difference in the incidence of these findings between the cohorts (22). Finally, the vast majority of radiation exposure from cardiac imaging occurred in patients age 40 and 60 years, with patients in the youngest decade of this range having a significant reduction in median effective radiation dose. While reduction in radiation dose and exposure may not affect downstream cancer risk as dramatically in this group when compared to patients in their 3rd decade of life, current guidelines recommend minimizing dose and exposure regardless of age. Given the very good prognosis in CCTA patients with no or minimal CAD over 5-6 years of follow-up, perhaps longer term follow-up is needed before the full benefit of an index CCTA is realized (23-25).

Study Limitations: This is a single center analysis involving a relatively small sample size enrolled in a closed-referral healthcare system, thus the findings may not be generalizable to an open referral health system. The use of an organ weighting factor of 0.014 mSv mGy^-1 cm^-1, though used as a standard clinical conversion factor, is reported to underestimate effective dose of CCTA by up to 50%, thus actual CCTA per-patient doses may be higher than calculated (26-28). Effective radiation doses from procedures involving fluoroscopy (either interventional radiology or at the time of a surgical operation) were not included in this analysis as the air kerma levels were not available. The use of DSCT imaging is not readily available in all ED hospitals, thus the reported effective radiation doses may not be generalizable. Given the fact that the patients in the CCTA and usual care cohorts were evaluated during the same time period, it is likely that the majority of patients in the usual care cohort presented on weekends and after normal business hours when CCTA was not available.