Influence of Body Mass Index on Radiation Exposure Across Imaging Modalities in the Evaluation of Chest Pain

Abstract

Background

Essential to a patient‐centered approach to imaging individuals with chest pain is knowledge of differences in radiation effective dose across imaging modalities. Body mass index (BMI) is an important and underappreciated predictor of effective dose. This study evaluated the impact of BMI on estimated radiation exposure across imaging modalities.

Methods and Results

This was a retrospective analysis of patients with concern for cardiac ischemia undergoing positron emission tomography (PET)/computed tomography (CT), cadmium zinc telluride single‐photon emission CT (SPECT) myocardial perfusion imaging, or coronary CT angiography (CCTA) using state‐of‐the‐art imaging modalities and optimal radiation‐sparing protocols. Radiation exposure was calculated across BMI categories based on established cardiac imaging–specific conversion factors. Among 9046 patients (mean±SD age, 64.3±13.1 years; 55% men; mean±SD BMI, 30.6±6.9 kg/m²), 4787 were imaged with PET/CT, 3092 were imaged with SPECT/CT, and 1167 were imaged with CCTA. Median (interquartile range) radiation effective doses were 4.4 (3.9–4.9) mSv for PET/CT, 4.9 (4.0–6.3) mSv for SPECT/CT, and 6.9 (4.0–11.2) mSv for CCTA. Patients at a BMI <20 kg/m² had similar radiation effective dose with all 3 imaging modalities, whereas those with BMI ≥20 kg/m² had the lowest effective dose with PET/CT. Radiation effective dose and variability increased dramatically with CCTA as BMI increased, and was 10 times higher in patients with BMI >45 kg/m² compared with <20 kg/m² (median, 26.9 versus 2.6 mSv). After multivariable adjustment, PET/CT offered the lowest effective dose, followed by SPECT/CT, and then CCTA (P<0.001).

Conclusions

Although median radiation exposure is modest across state‐of‐the‐art PET/CT, SPECT/CT, and CCTA systems using optimal radiation‐sparing protocols, there are significant variations across modalities based on BMI. These data are important for making patient‐centered decisions for ischemic testing.

Nonstandard Abbreviations and Acronyms

CACS

coronary artery calcium score

ICRP

International Commission on Radiological Protection

Clinical Perspective.

What Is New?

• This study compared radiation exposure across body mass index (BMI) categories in patients undergoing state‐of‐the‐art positron emission tomography/computed tomography (CT), cadmium zinc telluride single‐photon emission CT/CT, and coronary CT angiography with radiation‐sparing protocols.

• Overall, radiation exposure was modest, but there was significant variation according to BMI, particularly in those undergoing coronary CT angiography.

• Patients with BMI of <20 kg/m² had similar effective doses with all 3 imaging modalities, whereas those with larger BMIs had the lowest effective dose with positron emission tomography/CT.

What Are the Clinical Implications?

• Essential to a patient‐centered approach to imaging individuals with chest pain is knowledge of differences in radiation effective dose across imaging modalities.

• Tailoring testing strategy based on patient BMI may lead to lower radiation effective dose when imaging for chest pain.

Evaluation of chest pain in patients with known or suspected coronary artery disease is frequently accomplished with either coronary computed tomography angiography (CCTA) or nuclear stress testing, using either single‐photon emission computed tomography (SPECT) or positron emission tomography (PET). The 2021 multisocietal chest pain guidelines ¹ and 2023 appropriate use criteria ² recommend either CCTA or nuclear stress testing as first‐line imaging modalities in the evaluation of chest pain and advocate for a patient‐centered imaging strategy. Knowledge of radiation exposure on the patient level is essential to guiding a patient‐first strategy, as is mentioned in the guidelines. ¹

Larger patients present challenges for all 3 of these imaging modalities, as larger size is associated with greater attenuation and hence lower image quality, unless image protocol adjustments are made. The magnitude of differences in radiation exposure between modalities in relation to body size in daily practice has received little attention. In this study, we compared total body radiation effective doses across patient sizes, as measured by body mass index (BMI), using state‐of‐the‐art instrumentation and aggressive radiation‐sparing protocols for all 3 modalities.

METHODS

The data underlying this article will be shared on reasonable request to the corresponding author.

Patients

The cardiovascular imaging database at the Saint Luke's Health System was used to identify patients who underwent clinically indicated SPECT, PET, or CCTA imaging studies to evaluate for known or suspected coronary artery disease. Patients were referred to the specific cardiovascular imaging modality based on ordering provider preference. Consecutive patients were included from January 1, 2019 through December 31, 2021 for PET and SPECT and through August 31, 2023 for CCTA. As our institution provides these services at multiple urban and rural sites with variable cameras and protocols, as previously described, ³ this study focused on only those patients who were imaged using state‐of‐the art instrumentation and radiation‐sparing protocols. Data collected included age, sex, BMI, type of test, injected radionuclide doses in mCi for patients undergoing SPECT and PET, and dose‐length product for patients undergoing CCTA. The study was approved by the Institutional Review Board of Saint Luke's Hospital of Kansas City and Columbia University, and informed consent was waived.

Cameras

All included patients underwent imaging using state‐of‐the‐art cameras. SPECT myocardial perfusion imaging (MPI) studies were performed on 1 of 2 Spectrum Dynamics D‐SPECT cameras, which use cadmium zinc telluride detectors. PET/computed tomography (CT) MPI was performed using Siemens Biograph Vision digital PET/CT systems. CCTA was performed on a Siemens 384 slice dual‐source SOMATOM Force scanner. Attenuation correction CT scans for the SPECT studies were performed on the Siemens Biograph Vision PET/CT system or the Siemens SOMATOM Force scanner.

SPECT/CT Imaging Protocol

SPECT/CT MPI studies were all performed using stress/rest (63% of the cohort) or rest/stress (37% of cohort) protocols. The stress/rest protocol was performed with stress imaging first, with rest imaging omitted if stress images were normal, as has long been recommended for radiation exposure reduction. ⁴ At peak exercise or immediately following vasodilator stress, a protocol‐ and weight‐adjusted dosage (range of ±20%) of either Tc‐99m sestamibi or Tc‐99m tetrofosmin was injected. For the 2‐day protocol, a target activity of 15 mCi was used for the stress and an additional 15 mCi for the rest study, if needed. For the 1‐day protocol, if BMI <30 kg/m², 4 mCi was used for the first injection (rest or stress) followed by 3× the dose for the second injection, if needed. If BMI was 30 to 40 kg/m², 5 mCi for the first injection and 4× the dose for second were used. If BMI was >40 kg/m², 10 mCi for the first injection and 30 mCi for the second were used. Imaging was started after there was adequate hepatobiliary clearance, which was typically 45 to 60 minutes following radiotracer injection; and the low‐dose optimized acquisition protocol, as recommended by Spectrum Dynamics, was used to acquire images in each of the supine and upright positions for ≈1 million myocardial counts. Additionally, all patients underwent either a low‐dose non‐ECG gated CT scan or coronary calcium scoring (if no known coronary artery disease or no prior coronary artery calcium score [CACS] within the past 5 years) for attenuation correction of the emission data.

PET/CT Imaging Protocol

PET/CT MPI was performed as a rest/stress study with dynamic, static, and gated acquisitions. Rubidium‐82–injected activity was 30 mCi (±20%) for all patients regardless of body size for both the rest and stress images. The CardioGen‐82 infusion system (Bracco Diagnostics) was used with a flow rate of 50 mL/min. All patients underwent a low‐dose non‐ECG gated CT scan for attenuation correction of the emission data. CACS was similarly added if there was no known coronary artery disease or no prior CACS within the past 5 years.

CCTA Imaging Protocol

Before the CCTA study, patients were premedicated with β‐blockers and nitrates as per standard protocol. ⁵ A test bolus of 15 mL of contrast was used to determine the timing delay for CCTA, followed by a bolus of 75 mL at 5 mL/s or 100 mL at 7 mL/s contrast (if weight <250 or >250 lbs, respectively) for the study. Automatic tube potential selection was used with tube current modulation using retrospective ECG gating (with MinDose protocol), prospective ECG triggering, or high‐pitch helical protocols and software for radiation reduction. Scan mode was selected by the interpreting physician based on patient BMI, heart rate, and heart rate stability after premedications as well as the indication for the study. Patients with a regular heart rate of <70 beats per minute were eligible for high‐pitch helical or prospective acquisition protocols. Tube potential (kVp) and current (mAs) were selected automatically by the Siemens software using the CARE kV and CARE Dose4D patient‐specific automatic dose adjustment protocol based on the scan mode selected. The best diastolic phase images were used.

Estimation of Radiation Doses

For nuclear imaging, radiation effective doses were calculated by incorporating conversion factors for Tc‐99m–labeled tetrofosmin and sestamibi for SPECT ⁶ and rubidium‐82 for PET ⁷ on both the stress and rest images. An alternative conversion factor for rubidium‐82 was also used for comparison. ⁸ The contribution of the radiation exposure from the low‐dose attenuation correction CT or CACS was calculated using a k‐factor of 0.015 mSv/[mGy×cm]. As there are currently no published estimated k‐factors for converting dose length product to mSv for attenuation correction CT scans, this was estimated using commercially available software (Radimetrics, Bayer Healthcare, Germany) from attenuation correction scans obtained on a Siemens mCT scanner. For this analysis, dose‐length product was obtained on 283 patients who underwent PET/CT MPI, and Radimetrics‐derived effective doses using International Commission on Radiological Protection (ICRP) 103 tissue weighting factors were calculated on the rest and stress scans (total of 566 scans).

For CCTA, radiation effective dose was calculated by the dose‐length product and multiplied by the k‐factor 0.026 mSv/(mGy×cm), which has been validated for coronary CT imaging protocols. ⁹

Statistical Analysis

Radiation exposure by imaging modality was calculated by BMI category: ≤20, >20 to 25, >25 to 30, >30 to 35, >35 to 40, >40 to 45, and >45 kg/m². The radiation exposure to the cohort was analyzed using the as‐imaged protocol, which included CACS if this was performed. Summary statistics were displayed using medians (interquartile ranges [IQRs]) and mean±SD. Medians were compared using the Kruskal‐Wallis test, means with ANOVA, and categorical variables with Pearson χ² test. Multivariable linear regression was used to evaluate predictors of as‐imaged radiation effective dose after adjusting for age, BMI, sex, study type, and whether or not CACS was performed. Box plots were generated by BMI category, and curves were produced to model predicted values using locally weighted scatterplot smoothing. Multivariable regression models were used to assess the effects of BMI and imaging modality on radiation effective dose after controlling for age, sex, and acquisition of a CACS. Statistical analysis was performed using Stata, version 14.2 (Stata LLC, College Station, TX).

RESULTS

Overall, 9046 patients were included in the analysis (4787 PET, 3092 SPECT, and 1167 CCTA). Mean±SD age was 64.3±13.1 years, mean±SD BMI was 30.6±6.9 kg/m², 55% were men, and 53.3% had a concomitant CACS acquired (Table 1). Patients undergoing CCTA were younger and more likely women, and there were small but significant differences in BMI and the rate of CACS when comparing the 3 imaging modalities (Table 1). The distribution of BMIs was similar among all 3 imaging modalities, although SPECT was used less in those with BMI >40 kg/m² (Figure 1). Stress‐only SPECT was done in 39.2% of patients and was more common in patients with lower BMIs (49% if BMI <20 kg/m² and 34% if BMI >45 kg/m²). The 2‐day protocol was performed in 575 (18.6%) patients who underwent SPECT. Median total isotope activity for PET was 60.2 mCi of rubidium‐82 (IQR, 40.2–70.6 mCi), and for SPECT was 18.4 mCi of sestamibi (IQR, 3.1–47.4 mCi) or 18.0 mCi of tetrofosmin (IQR, 6.3–48.6 mCi). CCTA acquisition protocol was high‐pitch helical in 179 patients (15.3%), prospective ECG triggering in 654 patients (56%), and retrospective ECG gating in 334 patients (28.6%), and protocol was associated with radiation effective dose (Table S1). Median dose length product was 270 (IQR, 153–432).

Table 1.

Baseline Characteristics

Variable	PET/CT (n=4787)	SPECT/CT (n=3092)	CCTA (n=1167)	P value
Age, y	67.3±12.0	64.1±10.9	52.2±15.9	<0.001
Male sex	2729 (57.0)	1741 (56.3)	504 (43.3)	<0.001
Height, in	67.4±4.2	67.6±4.0	67.2±4.1	0.002
Weight, lb	203.0±52.4	195.3±44.7	188.3±47.3	<0.001
BMI, kg/m2	31.4±7.4	29.9±5.9	29.2±6.6	<0.001
Known coronary artery disease	2838 (59.3)	1597 (51.6)	106 (9.1)	<0.001
CACS acquired	2462 (51.4)	1708 (55.2)	655 (56.1)	<0.001
Radiation effective dose, mSv	4.4 (3.9–4.9)	4.9 (4.0–6.3)	6.9 (4–11.2)	<0.001

Data are given as mean±SD, number (percentage), or median (interquartile range). BMI indicates body mass index; CACS, coronary artery calcium scoring; CCTA, coronary CT angiography; CT, computed tomography; PET, positron emission tomography; and SPECT, single‐photon emission CT.

Figure 1. Proportion of patients by imaging modality and body mass index (BMI) class.

Generally similar distributions across BMI were noted in patients undergoing positron emission tomography (PET), single‐photon emission computed tomography (SPECT), and coronary computed tomography angiography (CCTA), although CCTA tended to be performed more often in patients at lower BMIs, and PET in those at higher BMIs.

Median (IQR) radiation effective doses in the cohort were 4.4 (3.9–4.9) mSv for PET, 4.9 (4.0–6.3) mSv for SPECT, and 6.9 (4.0–11.2) mSv for CCTA. Measurements of the effective dose of radiation by imaging modality (Table 1) and by BMI (Table 2; Figure 2A and 2B) illustrate that PET had the most consistent radiation exposure with the smallest deviation from the median, and CCTA had the largest variation, particularly at higher BMIs. After multivariable adjustment, PET had the lowest radiation effective dose, followed by SPECT (P<0.001) and CCTA (P<0.001 versus both PET and SPECT; Table S2). Interaction testing also confirmed that the effect of BMI on radiation effective dose varied depending on which imaging modality was used (P<0.001). If an alternative conversion factor for rubidium‐82 was used, effective dose for PET would have been even lower (median, 3.38 [IQR, 2.93–3.86] mSv). ⁸

Table 2.

Radiation Effective Dose by BMI Category

Variable	PET/CT (n=4787)	SPECT/CT (n=3092)	CCTA (n=1167)	P value
BMI <20 kg/m2	n=143	n=79	n=53
Effective dose, mSv	4.1 (3.6–4.1)	4.3 (1.6–5.1)	2.6 (0.9–4.0)	<0.001
CACS acquired	77 (53.8)	53 (67.1)	24 (45.3)	0.036
BMI 20–25 kg/m2	n=796	n=499	n=258
Effective dose, mSv	4.1 (3.6–4.2)	4.4 (3.8–5.6)	3.5 (1.8–5.4)	<0.001
CACS acquired	386 (48.5)	278 (55.7)	138 (53.5)	0.033
BMI 25–30 kg/m2	n=1394	n=1118	n=403
Effective dose, mSv	4.1 (3.7–4.4)	4.5 (3.1–5.6)	6.5 (4.3–9.3)	<0.001
CACS acquired	658 (47.2)	589 (52.7)	229 (56.8)	<0.001
BMI 30–35 kg/m2	n=1144	n=889	n=246
Effective dose, mSv	4.4 (3.9–4.8)	5.7 (4.3–6.5)	9.2 (6.6–13.4)	<0.001
CACS acquired	565 (49.4)	476 (53.5)	145 (58.9)	0.013
BMI 35–40 kg/m2	n=675	n=383	n=129
Effective dose, mSv	5.4 (4.3–5.4)	6.1 (5.1–7.5)	11.6 (8.1–17.2)	<0.001
CACS acquired	365 (54.1)	222 (58.0)	72 (55.8)	0.47
BMI 40–45 kg/m2	n=414	n=41	n=45
Effective dose, mSv	6.3 (5.0–6.3)	8.2 (5.0–8.9)	16.0 (11.6–23.0)	<0.001
CACS acquired	251 (60.6)	22 (53.7)	20 (44.4)	0.089
BMI >45 kg/m2	n=221	n=83	n=33
Effective dose, mSv	6.7 (5.1–6.7)	8.9 (5.7–9.4)	26.9 (20.1–30.2)	<0.001
CACS acquired	160 (72.4)	68 (81.9)	27 (81.8)	0.17

Data are given as median (interquartile range) or number (percentage). Effective dose estimates include attenuation correction CT and CACS (if performed). BMI indicates body mass index; CACS, coronary artery calcium score; CCTA, coronary CT angiography; CT, computed tomography; PET, positron emission tomography; and SPECT, single‐photon emission CT.

Figure 2. Effect of body mass index (BMI) on radiation dose.

A, Locally weighted smoothing curves show that at lower BMI (<20 kg/m²) median radiation effective dose for coronary computed tomography angiography (CCTA) is lower than for both positron emission tomography (PET) and single‐photon emission computed tomography (SPECT). As BMI increases to >20 kg/m², the effective dose increases more dramatically for CCTA compared with PET and SPECT. Box plots by BMI class. B, PET offered relatively low radiation effective dose, with increases by BMI attributable to the contribution of attenuation correction CT and CACS, if performed.

In patients with BMIs <20 kg/m² (n=275, 3% of cohort), all 3 imaging modalities had similar effective doses after adjustment (CCTA versus PET: P=0.511; CCTA versus SPECT: P=0.861; PET versus SPECT: P=0.536). Under 1% of the cohort received <1 mSv of radiation, and all were in the CCTA cohort (n=71, 6.1% of CCTA cohort), and all but 1 had no CACS performed. If BMI was 20 to 25 kg/m², PET had the lowest radiation effective dose (P<0.001 versus SPECT and CCTA), and SPECT and CCTA had comparable effective doses (P=0.567). In BMI classes >25 kg/m², PET had the lowest radiation effective dose (P<0.001 for all comparisons), followed by SPECT (P<0.001 versus CCTA). The increase in radiation exposure was particularly evident in patients with class 3 obesity or higher (BMI >40 kg/m², n=837, 9.3% of the cohort), where the median radiation exposure from CCTA was 3.3× as high as PET (20.9 versus 6.3 mSv). Radiation effective dose with CCTA was 10× higher in patients with BMI >45 kg/m² compared with BMI <20 kg/m² (26.9 versus 2.6 mSv).

Effective dose was also evaluated on the basis of height and weight. Radiation dose increased for all modalities at higher weights (Figure 3A), and results were more dramatic with CCTA than PET or SPECT. At weights <148 lbs (14.3% of cohort), CCTA offered lower effective dose than PET. At weights >148 lbs, PET offered lower effective doses than either SPECT or CCTA (P<0.001 for all comparisons), and SPECT offered lower effective doses than CCTA (P<0.001 for all weight categories). Effective dose also increased with increasing heights in those who underwent CCTA, although increases in the PET and SPECT cohorts were not as dramatic (Figure 3B).

Figure 3. Radiation effective dose by weight and height.

A, Radiation effective dose increased for all 3 imaging modalities across septiles of weight, with the largest increase noted with coronary computed tomography angiography (CCTA). B, There was not a predictable trend in radiation effective dose across septiles of height in single‐photon emission computed tomography (SPECT) or positron emission tomography (PET), but there was an increase with CCTA.

With respect to image quality, 58 of 4787 (1.2%) PETs and 73 of 3092 (2.4%) SPECT MPIs were interpreted as equivocal or nondiagnostic. Of 1167 CCTAs, 43 (3.7%) were either repeated or interpreted as nondiagnostic because of image quality. There were no significant differences in quality by BMI (Table S3).

DISCUSSION

This study analyzed radiation effective dose among 3 radiation‐exposing imaging modalities using state‐of‐the‐art imaging hardware and software and aggressive radiation dose‐reducing techniques. The effect of BMI on radiation exposure varied significantly among the 3 imaging modalities. PET offered a predictable and modest radiation exposure, with radiation exposure increasing slightly in the higher BMI categories because of the contribution of the attenuation correction CT and CACS (when performed). For both CCTA and SPECT, BMI had a substantial impact on radiation exposure. All 3 imaging modalities generally offered low radiation exposure in small patients (BMI <20 kg/m²), whereas PET offered the lowest exposure in those with BMI >20 kg/m², followed by SPECT and then CCTA. The largest variability in radiation effective dose was seen in patients receiving CCTA, where median effective dose was >10× higher in the highest BMI category compared with the lowest.

The question of what amount of radiation was a patient exposed to during a nuclear or CT imaging study seems simple, but the answer is complex and has changed over time as the understanding of radiopharmaceutical and CT dosimetry has evolved. The ICRP has developed tissue weighting factors relating to tissue blood flow and sensitivity to radiation, which are needed to calculate tissue and organ absorbed doses and effective dose for nuclear and CT imaging. ¹⁰ For example, updated tissue weighting factors for CT scans, reflecting more current epidemiologic data and understanding, have been included in ICRP publication 103¹⁰ and give greater weight to breast tissue compared with the older publication 60 from 1991 (breast weighting factor of 0.12 versus 0.05, respectively). This is 1 reason that the commonly cited conversion factor from dose‐length product (expressed in units of mGy×cm) to effective dose (expressed in units of mSv) of 0.014 is considered outdated and the 0.026 conversion factor is included in multisocietal guidelines. ⁹ , ¹¹ There have been variations in the conversion factors from MBq to mSv with rubidium‐82, and the highly cited data from Senthamizhchelvan et al were used for this analysis. ⁷ More recent numbers for rubidium‐82 dosimetry from Hunter et al, which are the first to be measured directly with PET/CT in human cardiac patients, would have yielded even lower radiation exposure for PET. ⁸ Updates from the ICRP are forthcoming with respect to cardiac radiopharmaceutical dosimetry, but are not expected to substantially change the message of the current analysis.

Although our study describes estimated radiation effective dose as calculated using validated conversion factors, one must keep in mind that effective dose is a concept designed for protection where there are stochastic risks, wherein the concern with higher radiation effective doses is increased cancer risk. A conversion factor for a particular imaging study is defined on the basis of a population average, and not intended for patient‐specific cancer risk assessment. With respect to patient size, larger patients have higher dose‐length products on CT studies and thus higher radiation effective dose, although it is unclear if this truly reflects higher organ doses and thus increased radiation‐attributable cancer risk compared with that in smaller patients. Ultimately, how estimated radiation effective dose translates to cancer risk in patients with larger BMIs is unclear, although ICRP estimates of radiation effective dose are the most effective way to optimize protection against stochastic effects of radiation ¹² and compare exposure across imaging modalities. Furthermore, population‐based tissue weighting factors are used by the ICRP and do not distinguish between men and women, despite the fact that women have increased breast and lung radiation sensitivity and greater risk of cancer when compared with men. ¹³ Future efforts from the ICRP to estimate individual and sex‐specific effective dose estimates are underway ¹⁴ ; hopefully such enhanced methods will enable more precise estimation of patient‐specific risks from cardiac imaging.

Multisocietal guidelines and statements call for a focus on patient‐centered imaging. ¹ , ¹⁵ , ¹⁶ Central to this philosophy is accurate estimation of risks to the patient, including radiation exposure. Although CCTA can offer the lowest effective radiation dose among all 3 imaging modalities, this was only achieved in a small portion of our patient population with lower BMIs and lower heart rates. This heterogeneity continues to exist despite dramatic radiation dose reductions in CT using current hardware and software. ¹⁷ The 2021 multisocietal guidelines do not comment on patient factors that can affect radiation dose like BMI, although average effective dose is described as 3 to 5 mSv for CCTA. This is lower than what was noted in our study using state‐of‐the‐art CCTA imaging technology (6.9 mSv) and lower than in a multicenter European registry (9.6 mSv), ¹⁸ which is likely attributable to the guidelines’ estimate using the outdated conversion factor and not incorporating CACS, which is commonly used in clinical practice. Furthermore, the iodinated contrast used in CCTA increases total radiation effective dose by increasing organ doses, a factor that is not included in standard analyses of radiation exposure and could lead to underestimation of cellular DNA damage with CCTA. ¹⁹ As higher radiation effective doses are often needed to obtain diagnostic quality images in patients with higher BMIs, it should be noted that CT fractional flow reserve, which is used in intermediate coronary lesions to assess for ischemia, requires suitable quality images and 33% may have inadequate quality images. ²⁰

Some of the variability in radiation effective dose with CCTA is related to the protocol chosen, as retrospective spiral protocols use higher radiation than prospective ECG gating and high‐pitch helical scans. Over time, radiation reduction protocols have yielded lower effective doses for CCTA along with a reduction in retrospectively ECG‐gated scans (94% to 11% in PROTECTION VI [Prospective Randomized Trial on Radiation Dose Estimates of Cardiac CT Angiography in Daily Practice]). ²¹ Although the 28% rate of retrospective scans in the current study is above that of PROTECTION VI, we also used more low‐radiation high‐pitch helical scan protocols (15% versus 11%). Median dose length product in our study was 270 mGy×cm, which is similar to that in PROTECTION VI (246 mGy×cm), despite our patients having substantially higher BMI (median, 30.6 versus 26.8 kg/m² in PROTECTION VI).

The nuclear imaging modalities also differed in their radiation effective doses. SPECT yielded higher effective doses than PET regardless of BMI, and higher than CCTA if BMI was <25 kg/m². The injected activity of SPECT radiotracers (tetrofosmin and sestamibi) increases with BMI, which contributes to increased radiation effective dose. The use of stress‐first and stress‐only‐if‐normal imaging along with BMI‐adjusted protocols are 2 approaches recommended by guidelines and used by our imaging laboratory to reduce radiation exposure for SPECT studies. ³ However, larger patients are less likely to undergo a stress‐only study, contributing to a higher effective dose. State‐of‐the‐art cadmium zinc telluride crystal cameras allow for injection of a lower radiotracer activity to achieve similar diagnostic images at a lower radiation effective dose than conventional Anger cameras. With respect to PET, rubidium‐82 radiotracer has a short half‐life, and the radiation exposure is extremely consistent within a narrow range across BMI categories. Some centers use a more weight‐based protocol for rubidium‐82 dosing, which may yield even lower radiation effective dose for PET. The higher radiation effective doses with PET noted at higher BMIs are attributable to the contribution from the attenuation correction CT and the CACS, if performed. Finally, the PET findings do not apply for other MPI radiotracers, such as N13‐ammonia or F18‐flurpiridaz.

Finally, effective radiation dose in cardiac imaging studies has continued to decrease over time with the evolution of hardware and software technologies. Although hardware can be considered state of the art in the era in which this study took place, as novel hardware (particularly photon‐counting CT) is adopted, radiation effective dose is expected to continue decreasing. Additionally, interpretation at lower effective doses using deep learning algorithms may be possible. This underscores the importance of modernizing cardiovascular imaging laboratories and generating updated real‐world estimates of radiation effective dose as systems evolve.

Limitations

This study should be interpreted in the context of the following limitations. More patients were imaged using PET and SPECT than CCTA in this cohort, although the number of studies was large for each imaging modality. Legacy nondigital PET, CCTA, and Anger SPECT systems were not included in this analysis, as we attempted to evaluate state‐of‐the‐art hardware and software technologies and protocols. Our nuclear laboratory protocols for radiotracer dosing, attenuation correction CT, and calcium scoring may be different than other local protocols, changing estimated radiation exposure. Radiation dose reduction should be incorporated while maintaining adequate image quality; image quality across these imaging modalities and BMI categories is an area for further research. Finally, stress electrocardiography, echocardiography, and magnetic resonance imaging are additional stress testing modalities that do not expose patients to ionizing radiation.

CONCLUSIONS

Although effective radiation doses using state‐of‐the‐art PET, SPECT, and CCTA systems with optimal radiation‐sparing protocols are modest across modalities, there are significant differences in radiation exposure based on BMI. As BMI increases, so too does radiation exposure, although PET offers the least variability, followed by SPECT, and then CCTA. These data may be helpful to generate more accurate estimations of radiation effective dose by BMI for each imaging modality and are important for making patient‐centered decisions for ischemic testing.

Sources of Funding

None.

Disclosures

Dr Bateman has received research grant support from Bracco, GE Healthcare, Jubilant Drax Image, and Spectrum Dynamics; has served as a consultant to GE Healthcare and Synektik; and has an equity interest in Cardiovascular Imaging Technologies. Dr Einstein reports receiving a speaker's fee from Ionetix, consulting fees from W. L. Gore & Associates, and authorship fees from Wolters Kluwer Healthcare–UpToDate; and serving on a scientific advisory board for Canon Medical Systems USA; his institution has grants/grants pending from Attralus, Bruker, Canon Medical Systems USA, Eidos Therapeutics, GE HealthCare, Intellia Therapeutics, Ionis Pharmaceuticals, Neovasc, Pfizer, Roche Medical Systems, and W. L. Gore & Associates. The remaining authors have no disclosures to report.

Supporting information

Tables S1–S3.