Abstract
Background
Adequate intravascular enhancement is essential for accurate coronary computed tomography angiography (CCTA) evaluation; however, conventional empirical contrast protocols often result in significant variability in enhancement because they do not account for individual physiological differences. Test-bolus techniques represent a potential approach for individualized contrast delivery; however, their precise quantitative relationship with final vascular enhancement requires further clarification. This study aimed to investigate the quantitative association between test-bolus peak attenuation and final aortic enhancement (FAE) in CCTA and to describe subgroup variations based on body mass index (BMI) and cardiac output (CO).
Methods
This single-center retrospective study included 50 patients who underwent CCTA. Test-bolus peak attenuation and FAE were systematically recorded. To evaluate the quantitative association between test-bolus peak attenuation and FAE, univariate linear regression analysis was performed, followed by subgroup analyses stratified by BMI (normal, overweight, or obese) and CO (normal or high).
Results
Test-bolus peak attenuation was a significant positive predictor of FAE (β=0.924, P<0.001), accounting for 26.1% of the overall variance in FAE. Exploratory subgroup analyses revealed descriptive differences in the strength of this predictive trend across BMI categories. Notably, a stronger predictive trend was observed in obese patients (R2=0.812, P=0.037) than in normal weight (R2=0.310) and overweight (R2=0.247) patients. Conversely, while predictive efficacy remained significant in patients with normal CO (R2=0.288, P=0.001), it was attenuated and statistically non-significant in those with high CO (R2=0.240, P=0.090).
Conclusions
Test-bolus peak attenuation demonstrated significant potential as a quantitative predictor of aortic enhancement in CCTA, suggesting that test-bolus peak attenuation may be useful in exploring individualized contrast protocols. The descriptive trends observed in the obese subgroup are strictly hypothesis-generating. Exploratory findings suggest that the predictive trend might vary according to BMI and CO. Larger studies are needed to confirm these descriptive observations.
Keywords: Coronary computed tomography angiography (CCTA), test bolus, aortic enhancement, body mass index (BMI), cardiac output (CO)
Introduction
Coronary computed tomography angiography (CCTA) is a first-line non-invasive imaging modality for the evaluation of coronary artery disease (CAD) and has an exceptionally high negative predictive value. It not only enables the simultaneous qualitative and quantitative assessment of plaque characteristics but also provides important information for the risk stratification of cardiovascular events (1). Achieving adequate intravascular enhancement within the coronary lumen—typically ranging from 350 to 500 Hounsfield units (HU)—is an essential prerequisite for accurate evaluation (2). Suboptimal enhancement can limit the assessment of non-calcified plaques and may necessitate repeat scans, whereas excessive enhancement can exacerbate blooming artifacts, frequently leading to the overestimation of luminal stenosis.
Conventional CCTA protocols predominantly rely on fixed-dose or empirically determined weight-based contrast administration. However, such strategies largely overlook important individualized physiological variables, particularly body mass index (BMI) and cardiac output (CO). Physiologically, BMI profoundly influences the volume of contrast distribution, while CO determines the rate of contrast dilution and pulmonary transit. The failure to account for these factors frequently results in substantial interindividual variability in vascular enhancement, directly affecting image quality and diagnosis in some cases (3-5).
van den Boogert et al. (6) demonstrated that individualizing contrast administration protocols based on patient weight and tube voltage not only improves luminal attenuation but also effectively reduces the total contrast dose. Further, they found that compared to traditional bolus-tracking techniques, the test-bolus technique may offer advantages in guiding individualized contrast administration. Nevertheless, existing studies applying this technique often fail to incorporate patients’ real-time cardiac functional status, and standardized criteria for determining the relevant parameter settings remain lacking.
Preliminary evidence (6) suggests a correlation between the peak attenuation of a small-dose test bolus and luminal attenuation in the ascending aorta; however, the precise quantitative relationship between test-bolus parameters and subsequent vascular enhancement in routine clinical practice has not been fully elucidated. Therefore, using real-world clinical data, this study aimed to quantitatively assess the association between test-bolus parameters and final CCTA vascular enhancement. Further, this study sought to elucidate how patient-specific physiological profiles—specifically BMI and cardiac functional status—moderate this predictive relationship. Ultimately, the study findings may provide a robust quantitative rationale for optimizing individualized contrast administration protocols, thereby ensuring consistent and diagnostically adequate image quality while minimizing the issues associated with empirical dosing.
Therefore, this study aimed to address gaps in the current literature by: (I) quantifying the predictive value of test-bolus peak attenuation for final aortic enhancement (FAE) in a Chinese cohort; (II) exploring descriptive trends in this relationship across different BMI and CO subgroups; and (III) presenting exploratory subgroup findings to inform future research on individualized contrast administration. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0982/rc).
Methods
Study population
This retrospective cohort study included consecutive patients with suspected CAD who underwent CCTA at Minhang Hospital, Fudan University, from September 2025 to February 2026. The inclusion criteria for this study were as follows: (I) age ≥18 years; (II) presence of normal sinus rhythm; (III) absence of hyperthyroidism; (IV) no history of allergy to iodinated contrast media; and (V) absence of severe hepatic or renal dysfunction. The exclusion criteria were as follows: (I) severe coronary artery calcification; (II) incomplete clinical or imaging data; and/or (III) failure to comply with the prescribed contrast medium injection protocol. Ultimately, 50 patients were included in the final analysis. The detailed patient selection process is illustrated in Figure 1.
Figure 1.
Flowchart of patient selection. CAD, coronary artery disease; CCTA, coronary computed tomography angiography.
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Minhang Hospital. Informed consent was waived due to the retrospective nature of the study.
CCTA examination protocol
All CCTA examinations were performed using a third-generation dual-source computed tomography (CT) scanner (SOMATOM Definition Force, Siemens Healthineers, Germany).
Patient preparation
Resting heart rate was routinely assessed before scanning. Patients with a resting heart rate >70 beats per minute (bpm) were administered oral beta-blockers (metoprolol, 25–50 mg) 30–60 minutes before scanning (n=34, 68%) to achieve a target heart rate ≤65 bpm. Contraindications to beta-blocker administration included resting bradycardia (<60 bpm), second- or third-degree atrioventricular block, and known beta-blocker allergy. Sublingual nitroglycerin was not administered before CCTA in any patient, as this is not part of Minhang Hospital’s routine protocol for CCTA. Patients were then positioned supine, and the scan range extended from 1 cm below the tracheal bifurcation (carina) to 2 cm below the diaphragm.
Image acquisition parameters
A retrospective electrocardiogram (ECG)-gated axial scanning mode was employed. Tube voltage was automatically modulated using CARE kV, and tube current was dynamically adjusted in real-time using an automated tube current modulation technique (CARE Dose4D; Siemens Healthineers, Erlangen, Germany). Specific scan parameters included a detector collimation of 2 mm × 96 mm × 0.6 mm, a gantry rotation time of 250 ms, and a temporal resolution of 66 ms. Full cardiac cycle data were acquired using automated ECG-triggered acquisition.
Contrast injection protocol
A nonionic contrast medium (iomeprol, 400 mg I/mL) and a saline chaser were administered via the right antecubital vein using a dual-syringe power injector (MEDRAD Stellant; Bayer, Germany).
Test bolus
A dynamic single-slice scan was performed at the level of the main pulmonary artery (with a 1-s scan interval over 20–25 phases). An initial 20-mL bolus of contrast medium was injected at a fixed flow rate of 3.5 mL/s, immediately followed by a 20-mL saline chaser at the same rate. A region of interest (ROI) was placed in the ascending aorta to generate a time-density curve, from which the peak attenuation (HU) and time to peak attenuation were recorded. In routine clinical practice at Minhang Hospital, the test bolus is primarily used to qualitatively assess circulatory transit time and to generate the time-density curve for documentation. The main scan was triggered using bolus tracking (with a 60-HU threshold in the descending aorta) rather than the test-bolus timing, as this is the standard protocol at Minhang Hospital. The test-bolus peak attenuation values were extracted retrospectively from the stored time-density curves for research purposes. A circular ROI (200 mm2) was placed in the ascending aorta at the level of the main pulmonary artery, carefully avoiding vessel wall calcifications and plaques. Attenuation values were recorded as absolute HU without baseline subtraction. In cases of respiratory motion or beam-hardening artifacts, the ROI position was manually adjusted by the operator to ensure measurement accuracy.
Main scan
The total contrast volume was calculated using the following formula: Main scan contrast volume (mL) = [7 + estimated scan time (s)] × injection flow rate (mL/s). The injection flow rate was individualized based on the tube voltage automatically selected by CARE kV, with flow rates of 3.5 mL/s at 70 kV and 4.0 mL/s at 80 kV, followed by an increase of 0.5 mL/s for every subsequent 10-kV increment. Bolus tracking was performed with the monitoring ROI placed in the descending aorta and a trigger threshold of 60 HU. The main scan was automatically initiated once the attenuation threshold was reached. Immediately following the contrast injection, a 30-mL saline chaser was administered at the same flow rate. This contrast administration protocol was based on Minhang Hospital’s protocol, which was adapted from previous literature and Siemens vendor recommendations for CCTA contrast delivery (7). The constant “7” in the formula represents the baseline injection duration (in seconds), ensuring sustained vascular enhancement throughout the entire acquisition period.
Image reconstruction and post-processing
Images were reconstructed using Advanced Modeled Iterative Reconstruction (ADMIRE; strength level 4; Siemens Healthineers) with a slice thickness of 0.75 mm, an increment of 0.5 mm, and a Bv40 vascular kernel. The reconstructed datasets were transferred to a dedicated workstation (syngo.via; Siemens Healthineers) for post-processing, including volume rendering, maximum intensity projection, and curved planar reformation.
Measurement of aortic enhancement
Objective image measurements were independently performed by a radiologist with 10 years of experience in cardiovascular imaging who was blinded to the contrast injection protocols. The CT attenuation of the aortic root was measured on the axial images at the level of the left main coronary artery ostium. These images were derived from standard ECG-gated reconstructions used for routine coronary evaluation; depending on the patient’s heart rate and rhythm, the reconstruction phase was selected as the best diastolic phase (typically 70–80% of the R-R interval) for patients with regular and low heart rates, or the best systolic phase (typically 35–45% of the R-R interval) for those with higher or irregular heart rates.
A fixed ROI area of 200 mm2 was used, carefully avoiding vessel wall calcifications and non-calcified plaques. Measurements were repeated twice for each patient, and the average value was recorded as the FAE intensity. To assess the reproducibility of the objective image parameters, interobserver reliability for the FAE measurements was evaluated using the intraclass correlation coefficient (ICC). The ICC was calculated based on a two-way mixed-effects model with absolute agreement. An ICC value >0.75 indicated excellent reproducibility.
A representative case illustrating the comprehensive imaging and measurement workflow is presented in Figure 2. This figure includes the test-bolus time-density curve, the standardized objective measurement of FAE, the corresponding high-quality coronary angiograms, and the automated assessment of cardiac function (as detailed in Section “Measurement of cardiac functional parameters”).
Figure 2.
Representative images illustrating the measurement of patient-specific physiological parameters, the test-bolus time-density curve, and FAE. (A) Automated assessment of cardiac functional parameters using a dedicated post-processing workstation, indicating a normal CO of 7.81 L/min. (B) Time-density curve generated from the test-bolus scan. The optimal peak attenuation was 230.6 HU with a time to peak attenuation of 17.0 seconds. (C) Standardized measurement of FAE. An axial image at the level of the left main coronary artery ostium demonstrates a fixed ROI (2.00 cm2) positioned within the aortic root, yielding an excellent mean attenuation of 411 HU. (D-F) Curved planar reformations of the RCA (D), LAD (E), and LCX (F) in the same patient, demonstrating homogeneous and highly diagnostic intraluminal contrast enhancement. CO, cardiac output; ED, end-diastolic; ES, end-systolic; FAE, final aortic enhancement; HU, Hounsfield units; LAD, left anterior descending artery; LCX, left circumflex artery; LV, left ventricle; RCA, right coronary artery; ROI, region of interest; SD, standard deviation.
Measurement of cardiac functional parameters
The reconstructed multiphase CT datasets were transferred to the post-processing workstation (syngo.via), which automatically calculated and recorded the left ventricular end-diastolic volume (LVEDV) and left ventricular end-systolic volume (LVESV). Cardiac functional parameters were calculated using the following formulas: stroke volume (SV) = LVEDV – LVESV; ejection fraction (EF) = (SV/LVEDV) ×100%. CO was calculated as CO = SV × heart rate, using the average heart rate recorded by ECG during the scan (8).
Study variables and grouping primary variables
The primary study variables were defined as follows—
Test-bolus peak attenuation: the maximum CT attenuation value (HU) derived from the test-bolus time-density curve;
FAE: the average CT attenuation value (HU) of the aorta measured according to the criteria in Section “Measurement of aortic enhancement”.
Time to peak attenuation was also recorded from the test-bolus time-density curve; however, preliminary correlation analysis revealed a weaker association with FAE (r=−0.284, P=0.046). Given its collinearity with CO and the focus of the study on peak attenuation as the primary predictor, time to peak attenuation was not included in the main regression models.
Based on established clinical reference ranges for CO (males: 2.82–8.82 L/min; females: 2.65–5.98 L/min), no patient in the cohort had low CO. Patients were thus dichotomized into normal CO (n=37) and high CO (n=13) groups.
Observation metric
Based on the optimal enhancement range required for routine clinical diagnosis (350–500 HU), the number of cases with aortic enhancement <350 HU was recorded to evaluate the clinical suitability of the routine contrast injection protocol.
Grouping criteria
The grouping criteria were as follows—
CO stratification: patients were categorized into normal CO and high CO groups based on established clinical reference ranges (males: 2.82–8.82 L/min; females: 2.65–5.98 L/min) (9);
BMI stratification: according to the Guidelines for Prevention and Control of Overweight and Obesity in Chinese Adults, patients were classified into three groups: the normal weight group (BMI <24 kg/m2), the overweight group (24≤ BMI <28 kg/m2), and the obese group (BMI ≥28 kg/m2).
Statistical analysis
Statistical analyses were conducted using SPSS 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables are reported as mean ± standard deviation or median (interquartile range) based on the results of the Shapiro-Wilk normality test; categorical variables are presented as frequencies (%). Pearson correlation analysis was initially performed to evaluate associations between FAE and all candidate variables, including age, sex, BMI, CO, tube voltage, flow rate, contrast volume, test-bolus peak, and time to peak attenuation. Univariate linear regression analysis was subsequently performed to quantify the predictive relationship between test-bolus peak attenuation and FAE. Subgroup analyses stratified by BMI and CO were then conducted to evaluate whether this predictive relationship varied across these physiological profiles. Due to the retrospective and exploratory nature of this study and the use of consecutive temporal sampling, no formal a priori sample size or power calculation was performed.
Results
Patient characteristics and correlation analysis
A total of 50 patients (mean age, 58.1±13.3 years; range, 18–88 years) were included in this study. The baseline clinical and procedural characteristics of the study cohort are summarized in Table 1.
Table 1. Baseline clinical and procedural characteristics of the study population (n=50).
| Variable | Value |
|---|---|
| Demographic characteristics | |
| Age (years) | 58.1±13.3 |
| Male sex | 31 (62.0) |
| Height (m) | 1.66±0.09 |
| Weight (kg) | 67.8±12.6 |
| BMI (kg/m2) | 24.1±3.1 |
| BMI classification | |
| Normal (<24 kg/m2) | 24 (48.0) |
| Overweight (24 to <28 kg/m2) | 21 (42.0) |
| Obese (≥28 kg/m2) | 5 (10.0) |
| Cardiac functional parameters | |
| Stroke volume (mL) | 68.5±14.2 |
| Ejection fraction (%) | 64.8±7.5 |
| CO (L/min) | 6.5±1.7 |
| CO classification | |
| Normal CO | 37 (74.0) |
| High CO | 13 (26.0) |
| Heart rate (bpm) | 66.8 (60.0–72.0) |
| Scan and contrast injection parameters | |
| Tube voltage | |
| 70 kV | 46 (92.0) |
| 80 kV | 3 (6.0) |
| 100 kV | 1 (2.0) |
| Injection flow rate (mL/s) | 3.6±0.3 |
| Test-bolus contrast volume (mL) | 20 |
| Test-bolus iodine dose (g) | 8 |
| Main scan contrast volume (mL) | 40.2±3.5 |
| Main scan iodine dose (g) | 16.1±1.4 |
| Total contrast volume (mL) | 60.2±3.5 |
| Total iodine dose (g) | 24.1±1.4 |
| Test-bolus peak attenuation (HU) | 249.8±88.6 |
| Time to test-bolus peak (s) | 19.0±3.3 |
| Scan duration (s) | 3.5 (3.2–3.9) |
| Iodine delivery rate (g I/s) | 1.4 (1.4–1.6) |
| Iodine dose per body weight (g/kg) | 0.88±0.15 |
| Enhancement results | |
| Final aortic attenuation (HU) | 714.3±143.2 |
| Suboptimal enhancement (<350 HU) | 0 (0.0) |
Data are presented as mean ± standard deviation, median (interquartile range), or number of patients (percentage). BMI, body mass index; bpm, beats per minute; CO, cardiac output; HU, Hounsfield units.
Pearson correlation analysis revealed a significant positive correlation between test-bolus peak attenuation and FAE (r=0.511, P<0.001). Further, the quantitative FAE measurements demonstrated excellent intraobserver reproducibility, with an ICC of 0.94 [95% confidence interval (CI): 0.89–0.97]. Conversely, FAE exhibited significant negative correlations with several variables, including CO (r=−0.573, P<0.001), BMI (r=−0.488, P<0.001), and tube voltage (r=−0.422, P=0.002). Additional clinical variables and baseline characteristics are presented in Table S1.
Overall predictive efficacy of the test-bolus peak
A univariate linear regression model demonstrated that the test-bolus peak attenuation was a robust positive predictor of FAE (β=0.924, P<0.001), independently accounting for 26.1% of the overall variance in FAE (Table 2). The overall linear association is illustrated by the scatter plot in Figure 3A.
Table 2. Univariate and subgroup linear regression analyses for predicting FAE using test-bolus peak attenuation.
| Subgroup | Number of patients | Correlation coefficient (r) | R2 | Regression coefficient, β (95% CI) | Regression equation | P value |
|---|---|---|---|---|---|---|
| Overall cohort | 50 | 0.511 | 0.261 | 0.924 (0.473–1.375) | FAE =499.72+0.924 × Peak | <0.001 |
| Stratified by BMI (kg/m2) | ||||||
| Normal (BMI <24) | 24 | 0.557 | 0.310 | 0.977 (0.351–1.604) | FAE =545.90+0.977 × Peak | 0.004 |
| Overweight (24≤ BMI <28) | 21 | 0.497 | 0.247 | 0.697 (0.114–1.280) | FAE =505.34+0.697 × Peak | 0.022 |
| Obese (BMI ≥28) | 5 | 0.901 | 0.812 | 3.300 (0.396–6.204) | FAE =−58.16+3.300 × Peak | 0.037 |
| Stratified by CO | ||||||
| Normal CO | 37 | 0.537 | 0.288 | 1.017 (0.469–1.566) | FAE =499.62+1.017 × Peak | 0.001 |
| High CO | 13 | 0.489 | 0.240 | 0.618 (−0.113 to 1.348) | FAE =507.23+0.618 × Peak | 0.090 |
The coefficient of determination (R2) indicates the proportion of the variance in FAE explained by the regression model based on test-bolus peak attenuation. Regression equations are presented in HU. P<0.05 was considered statistically significant. BMI, body mass index; CI, confidence interval; CO, cardiac output; FAE, final aortic enhancement; HU, Hounsfield units.
Figure 3.
Scatter plots with linear regression lines demonstrating the predictive relationship between test-bolus peak attenuation and FAE. (A) In the overall cohort (n=50), a significant positive linear correlation was observed (β=0.924, R2=0.261, P<0.001). (B) Subgroup analysis stratified by BMI. The predictive performance appeared strongest in the obese group, which demonstrated a markedly steeper regression slope (red line; R2=0.812, P=0.037) compared with the normal-weight (green line) and overweight (blue line) groups. (C) Subgroup analysis stratified by CO. The predictive relationship was significant in the normal CO group (green line; R2=0.288, P=0.001), whereas the regression slope was attenuated and the association was no longer statistically significant in the high CO group (red line; R2=0.240, P=0.090). Shaded areas represent 95% CIs. BMI, body mass index; CI, confidence interval; CO, cardiac output; FAE, final aortic enhancement.
Exploratory subgroup analyses stratified by BMI
Exploratory subgroup regression analyses stratified by BMI demonstrated descriptive variations in the predictive trend across BMI categories (Table 2 and Figure 3B). Moderate correlations were observed in the normal-weight and overweight groups (R2=0.310 and 0.247, respectively), while the predictive efficacy was notably stronger in the obese group. In obese patients, test-bolus peak attenuation accounted for 81.2% of the variance in FAE, with a markedly steeper regression slope (β=3.300, P=0.037) than those observed in the normal-weight and overweight groups. A complete leave-one-out sensitivity analysis, alongside individual influence diagnostics for this subgroup, is provided in Table S2.
Exploratory subgroup analysis stratified by CO
In the exploratory analysis stratified by cardiac functional status, descriptive differences in the regression patterns were observed (Table 2 and Figure 3C). In patients with normal CO, test-bolus peak attenuation maintained significant predictive value (R2=0.288, β=1.017, P=0.001). In the high CO group, the regression slope was flattened (β=0.618) and did not reach statistical significance (P=0.090). As no formal interaction analysis was performed, this finding should be interpreted as an uncertain and potentially underpowered subgroup pattern rather than as definitive evidence of attenuated predictive efficacy.
Discussion
This study quantitatively demonstrated that test-bolus peak attenuation is a strong predictor of FAE in CCTA. Further, the exploratory analyses suggest that this predictive relationship may exhibit descriptive variations according to BMI and cardiac functional status. Rather than a “one-size-fits-all” correlation, the test bolus showed a strong correlation with the FAE in obese patients. Conversely, the findings for patients with high CO represented an uncertain and potentially underpowered subgroup pattern. Collectively, these findings support the potential use of the test bolus as a quantitative parameter rather than a qualitative visual assessment and provide a physiological rationale for further investigation of individualized contrast protocols in a Chinese population.
Consistent with previous studies (6,10), we found a positive correlation between test-bolus peak attenuation and FAE. van den Boogert et al. (6) reported that every 1-HU increase in the test-bolus peak attenuation resulted in an average increase of 0.58 HU in aortic attenuation. In our overall cohort, the regression coefficient was 0.924, which was slightly higher than that reported by van den Boogert et al., but of a similar magnitude. This discrepancy may stem from our use of a higher-concentration contrast medium (400 mg I/mL vs. the conventional 350 mg I/mL) and differences in the injection protocols. The physiological basis for using a test bolus to predict final enhancement is that the small-dose test bolus and the subsequent main bolus traverse the exact same circulatory pathway and undergo similar dilution and transit processes (11). Consequently, the peak attenuation may reflect the patient’s individualized hemodynamic status (e.g., transit time and central blood volume) (11)], thereby carrying important information for predicting the main-bolus enhancement (12,13).
Notably, an exploratory analysis revealed a trend toward a stronger correlation between test-bolus peak attenuation and FAE in the obese subgroup. However, given the very small sample size of the obese subgroup (n=5), these results should be interpreted as hypothesis-generating rather than definitive. In the obese subgroup, the regression slope was 3.300, which was significantly higher than those of the normal-weight and overweight groups.
The “volume of distribution hypothesis” in contrast pharmacokinetics represents a possible explanation for this observation. Obese patients typically possess a larger central blood volume and contrast distribution volume, often accompanied by relatively reduced CO or altered circulatory patterns (14). In such a physiological state—characterized by a larger compartment and slower circulation—the intravascular dilution and transit of the contrast medium may remain relatively stable. The iodine bolus formed by the test injection may thus better preserve its signal characteristics, with less interference from rapid blood flow washout (15). Consequently, test-bolus peak attenuation may more precisely mirror the final enhancement of the main bolus.
This finding is consistent with the conclusions drawn from machine learning models by Lopes et al. (12), which identified test-bolus peak attenuation as the most critical feature for predicting suboptimal enhancement. Our study further quantified this relationship using traditional statistical methods. Because obese patients frequently require high tube voltage scanning, they may be particularly susceptible to marked iodine signal attenuation and thus face a high risk of suboptimal enhancement (16,17), making them the primary target population for individualized contrast protocols. Given the extremely small sample size of our study, the trends observed in the obese cohort are strictly hypothesis-generating. Future large-scale prospective studies are needed to rigorously evaluate whether these physiological observations are reproducible and whether they can ultimately inform individualized contrast dosing or algorithmic injection calculators.
Notably, the mean aortic attenuation in our cohort was 714 HU, substantially higher than the upper limit of the reported optimal range (500 HU), suggesting possible contrast overuse. No patient had enhancement below 350 HU; however, approximately 46 patients (92%) had enhancement exceeding 500 HU, which may increase blooming artifacts and contribute to the overestimation of stenosis. The widespread over-enhancement in our cohort may have been compounded by the combined use of a high-concentration contrast medium (400 mg I/mL) and a low tube voltage (70 kV in 92% of patients). Future individualized protocols should aim to minimize both under- and over-enhancement. Our subgroup-specific regression equations may provide a quantitative framework to achieve this balance.
Further, the descriptive variations observed across cardiac functional statuses warrant further investigation (4,10,18). In the normal CO group, the test-bolus peak demonstrated significant predictive value. Conversely, in the high CO group, the regression slope decreased to 0.618 and did not reach statistical significance (P=0.090). As no formal interaction analysis was performed to statistically compare the two strata, this result should be strictly interpreted as an uncertain and underpowered subgroup pattern. Although it is physiologically plausible that high CO states may accelerate contrast washout, the small sample size of this subgroup (n=13) severely limited the statistical power of the study. Consequently, any definitive mechanistic interpretation regarding the attenuation of the predictive signal remains speculative, and this descriptive pattern requires verification in larger, adequately powered cohorts (19).
This study had several limitations. First, as a single-center retrospective study without an a priori power calculation, the overall sample size was relatively small. Most notably, the obese group comprised only 5 patients. Linear regression based on such a limited sample heavily compromises the stability of the regression estimates. Although statistical significance was reached, the model is highly susceptible to overfitting and the influence of individual outliers, as evidenced by the extremely wide 95% CI (0.396–6.204) for the regression coefficient in this cohort. Consequently, while the predictive trend in obese patients is physiologically plausible, the current quantitative estimates should be considered statistically unstable and strictly hypothesis-generating, and require rigorous validation in future prospective, adequately powered multicenter studies. Further, we acknowledge the inherent statistical limitations of the leave-one-out sensitivity analysis presented in our supplementary data (Table S2). With only four observations remaining in each iteration (n=4), the regression estimates are highly sensitive to minor data variations, and standard influence diagnostics have limited interpretability in such a small sample. These sensitivity analyses are thus provided solely for transparency regarding data leverage, rather than as evidence of model stability. Additionally, the absence of formal interaction testing (e.g., test-bolus peak × BMI group and test-bolus peak × CO group) means that the observed differences in predictive efficacy across subgroups should be interpreted as descriptive subgroup findings rather than evidence of statistically significant interactions.
Second, patients with severe left ventricular dysfunction (e.g., reduced EF) were not specifically included in this cohort; all patients had CO values within the normal or high range. Thus, the applicability of our findings to patients with low CO states remains unknown. Further, the absence of suboptimal enhancement (<350 HU) in our cohort limited our ability to evaluate the clinical utility of test-bolus peak attenuation for predicting binary endpoints (e.g., inadequate vs. adequate enhancement). The present findings thus primarily support its use for predicting continuous enhancement values.
Third, several technical factors may have influenced our measurements. The use of a fixed ROI area (200 mm2) for aortic attenuation measurements may not fully account for individual variations in aortic diameter. Beta-blockers were administered to a subset of patients (68%) for heart rate control; given their negative inotropic effects, they may have transiently influenced baseline hemodynamics and CO, and therefore represent a potential unmeasured confounder in the CO subgroup analyses. Additionally, nitroglycerin was not administered in this cohort, which may have affected coronary vasodilation and luminal diameter assessment, although this is unlikely to have influenced the primary outcome of aortic root attenuation.
Fourth, while estimated scan time was incorporated into the total contrast volume calculation, the potential interaction between scan time and final enhancement efficacy was not extensively analyzed. Although tube voltage showed a significant correlation with FAE (r=−0.422, P=0.002), it was not included in the main regression models due to its close association with BMI and the study’s focus on test-bolus peak attenuation as the primary predictive variable. Future studies with larger sample sizes may further elucidate the independent contribution of tube voltage.
Finally, this study did not evaluate diagnostic image quality endpoints such as signal-to-noise ratio, contrast-to-noise ratio, or the proportion of non-diagnostic segments, which are essential for establishing the clinical utility of test-bolus-based dose adjustment. Future prospective studies should incorporate these endpoints to validate the clinical applicability of our findings.
Conclusions
Test-bolus peak attenuation may serve as a useful predictor of FAE in CCTA, although its predictive performance may vary according to BMI and CO. The descriptive trends observed in the obese subgroup should be considered hypothesis-generating, and larger prospective cohorts are required to investigate their potential clinical utility.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Minhang Hospital. Informed consent was waived due to its retrospective nature.
Footnotes
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0982/rc
Funding: This work was supported by the Science and Technology Commission of Minhang District, Shanghai (No. 2024MHZ077).
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0982/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0982/dss
References
- 1. Sandoval Y, Leipsic J, Collet C, Ali ZA, Azzalini L, Barbato E, Cavalcante JL, Costa RA, Garcia-Garcia HM, Jones DA, Khoo JK, Maran A, Nieman K, Pinilla-Echeverri N, Seto AH, Shlofmitz E, Brilakis ES. Coronary computed tomography angiography to guide percutaneous coronary intervention: Expert opinion from a SCAI/SCCT roundtable. J Cardiovasc Comput Tomogr 2025;19:277-90.. 10.1016/j.jcct.2025.04.015 [DOI] [PubMed] [Google Scholar]
- 2. Oda S, Utsunomiya D, Nakaura T, Kidoh M, Funama Y, Tsujita K, Yamashita Y. Basic Concepts of Contrast Injection Protocols for Coronary Computed Tomography Angiography. Curr Cardiol Rev 2019;15:24-9.. 10.2174/1573403X14666180918102031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Fei X, Du X, Yang Q, Shen Y, Li P, Liao J, Li K. 64-MDCT coronary angiography: phantom study of effects of vascular attenuation on detection of coronary stenosis. AJR Am J Roentgenol 2008;191:43-9.. 10.2214/AJR.07.2653 [DOI] [PubMed] [Google Scholar]
- 4. Konno M, Hosokai Y, Usui A, Abe M, Tateishi T, Kawasumi Y, Tsuda M, Ota H, Takase K, Saito H. Cardiac output obtained from test bolus injections as a factor in contrast injection rate revision of following coronary CT angiography. Acta Radiol 2012;53:1107-11.. 10.1258/ar.2012.120276 [DOI] [PubMed] [Google Scholar]
- 5. Tatsugami F, Matsuki M, Nakai G, Inada Y, Kanazawa S, Takeda Y, Morita H, Takada H, Yoshikawa S, Fukumura K, Narumi Y. Body size-adapted dose of contrast material and scanning protocol in 320-detector row CT coronary angiography. J Comput Assist Tomogr 2011;35:475-9.. 10.1097/RCT.0b013e31821f4e00 [DOI] [PubMed] [Google Scholar]
- 6. van den Boogert TPW, Lopes RR, Lobe NHJ, Verwest TA, Stoker J, Henriques JP, Marquering HA, Planken RN. Patient-tailored Contrast Delivery Protocols for Computed Tomography Coronary Angiography: Lower Contrast Dose and Better Image Quality. J Thorac Imaging 2021;36:353-9.. 10.1097/RTI.0000000000000593 [DOI] [PubMed] [Google Scholar]
- 7. Yamamuro M, Tadamura E, Kanao S, Wu YW, Tambara K, Komeda M, Toma M, Kimura T, Kita T, Togashi K. Coronary angiography by 64-detector row computed tomography using low dose of contrast material with saline chaser: influence of total injection volume on vessel attenuation. J Comput Assist Tomogr 2007;31:272-80.. 10.1097/01.rct.0000236422.35761.a1 [DOI] [PubMed] [Google Scholar]
- 8. Bastarrika G, Arraiza M, Pueyo JC, Herraiz MJ, Zudaire B, Villanueva A. Quantification of left ventricular function and mass in cardiac Dual-Source CT (DSCT) exams: comparison of manual and semiautomatic segmentation algorithms. Eur Radiol 2008;18:939-46.. 10.1007/s00330-007-0849-x [DOI] [PubMed] [Google Scholar]
- 9. Rusinaru D, Bohbot Y, Djelaili F, Delpierre Q, Altes A, Serbout S, Kubala M, Maréchaux S, Tribouilloy C. Normative Reference Values of Cardiac Output by Pulsed-Wave Doppler Echocardiography in Adults. Am J Cardiol 2021;140:128-33.. 10.1016/j.amjcard.2020.10.046 [DOI] [PubMed] [Google Scholar]
- 10. Dwivedi AND, Varshney A, Jain D, Singh G. CT coronary angiography as an alternative imaging method to ascertain cardiac output and its correlation with echocardiography. Clin Radiol 2023;78:e831-8.. 10.1016/j.crad.2023.07.019 [DOI] [PubMed] [Google Scholar]
- 11. Masuda T, Nakaura T, Funama Y, Sato T, Higaki T, Kiguchi M, Matsumoto Y, Yamashita Y, Imada N, Awai K. Development and Validation of Generalized Linear Regression Models to Predict Vessel Enhancement on Coronary CT Angiography. Korean J Radiol 2018;19:1021-30.. 10.3348/kjr.2018.19.6.1021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Lopes RR, van den Boogert TPW, Lobe NHJ, Verwest TA, Henriques JPS, Marquering HA, Planken RN. Machine learning-based prediction of insufficient contrast enhancement in coronary computed tomography angiography. Eur Radiol 2022;32:7136-45.. 10.1007/s00330-022-08901-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Masuda T, Funama Y, Imada N, Sato T, Yamagami T, Tatsugami F, Awai K. Prediction of aortic enhancement on coronary CTA images using a test bolus of diluted contrast material. Acad Radiol 2014;21:1542-6.. 10.1016/j.acra.2014.07.002 [DOI] [PubMed] [Google Scholar]
- 14. Bae KT, Seeck BA, Hildebolt CF, Tao C, Zhu F, Kanematsu M, Woodard PK. Contrast enhancement in cardiovascular MDCT: effect of body weight, height, body surface area, body mass index, and obesity. AJR Am J Roentgenol 2008;190:777-84.. 10.2214/AJR.07.2765 [DOI] [PubMed] [Google Scholar]
- 15. Masuda T, Higaki T, Nakaura T, Funama Y, Matsumoto Y, Sato T, Okimoto T, Gotanda R, Arao K, Imaizumi H, Arao S, Hiratsuka J, Awai K. Usefulness of the patient-specific contrast enhancement optimizer simulation software during the whole-body computed tomography angiography. Heart Vessels 2022;37:1446-52.. 10.1007/s00380-022-02024-z [DOI] [PubMed] [Google Scholar]
- 16. Caruso D. Innovation and Optimization of Contrast Media Administration in Computed Tomography. Korean J Radiol 2025;26:210-2.. 10.3348/kjr.2024.1159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Fursevich DM, LiMarzi GM, O'Dell MC, Hernandez MA, Sensakovic WF. Bariatric CT Imaging: Challenges and Solutions. Radiographics 2016;36:1076-86.. 10.1148/rg.2016150198 [DOI] [PubMed] [Google Scholar]
- 18. Singhal M, Subramanian P, Sharma A. Contrast Kinetics in CT Coronary Angiography. Indian J Radiol Imaging 2025;35:234-41.. 10.1055/s-0044-1792038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Qi K, Li L, Yuan D, Zhang Y, Zhang M, Zhang W, Gu Y, Wang Z, Lyu P, Gao J, Liu J. Optimized contrast enhancement and homogeneity in aortic CT angiography: bolus tracking with personalized post-trigger delay. Quant Imaging Med Surg 2025;15:709-20.. 10.21037/qims-24-624 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
The article’s supplementary files as
Data Availability Statement
Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0982/dss