The Effect of Functional and Intra-Coronary Imaging Techniques on Fluoroscopy Time, Radiation Dose and Contrast Volume during Coronary Angiography

The aim was to analyze the effect of fractional flow reserve (FFR), intravascular ultrasound (IVUS) and optical coherence tomography (OCT) on fluoroscopy time (FT), radiation dose (RD) and contrast volume (CV) in patients undergoing coronary angiography. This case-control study included consecutive patients above the age of 18, who underwent coronary angiography. FT, RD, and CV after each procedure were retrospectively recorded. Multivariate models were used to demonstrate the effect of these complementary studies and other factors, on radiation and contrast exposure. A total of 1047 patients were included, 74.5% were men and the mean (SD) age was 62.4 (12.1) years. Complementary studies performed were: IVUS (n = 237), FFR (n = 56) and OCT (n = 37). FFR and IVUS had a small effect on FT (η = 0.008 B = 2.2, p < 0.001; η = 0.009, B = 2.5, p < 0.001), while OCT had no effect (η = 0.002 B = 2.9, p < 0.183). IVUS, FFR and OCT had no effect on the RD. IVUS did not affect contrast volume (η = 0.002 B = 9.4, p < 0.163) while OCT and FFR had a small effect on CV (η = 0.006 B = 39, p < 0.01; η = 0.008 B = 37, p < 0.003). The number of placed stents had a significant effect on FT (η = 0.192, Β = 4.2, p < 0.001), RD (η = 0.129, Β = 511.8, p < 0.001) and CV (η = 0.177, Β = 40.5, p < 0.001). The use of complementary studies in hemodynamics did not modify the received RD and had a minor effect on FT and the CV used.

Coronary angiography is the gold standard for the diagnosis of coronary artery disease 1 . Over the past decade, functional and intra-coronary imaging techniques have emerged to overcome the limitations of coronary angiography. These new techniques are Fractional Flow Reserve (FFR), Intravascular Ultrasonography (IVUS) and Optical Coherence Tomography (OCT). FFR measures pressure differences across coronary artery stenosis, using a standard guide catheter with a pressure tip. It is defined as the pressure distal to stenosis divided to the pressure before the stenosis. IVUS uses an ultrasound probe and the principle of pulse-echo ultrasonography to create a plaque image giving valuable information such as plaque composition, positive remodeling, etc. and OCT creates an image of the plaque from a probe that ejects pulsating near-infrared photons 2 . FFR is of clinical importance because of its association with lower cardiovascular mortality 3 . IVUS and OCT, can aid in decision-making, guide interventions and optimize the results of percutaneous coronary intervention 4 .
Radiation and exposure to contrast medium have been associated with metabolomic changes in cardiomyocytes, endotheliopathy, atherosclerosis and contrast nephropathy [5][6][7] . Many individual factors associated to radiation and contrast exposure have been reported such as vascular access, age, and female sex, but the impact of the use of complementary studies such as FFR, IVUS and/or OCT has been scarcely studied 8,9 .
The aim of this study was to analyze the effect of angiographic complementary studies such as FFR, IVUS and OCT on fluoroscopy time (FT), radiation dose (RD)* and contrast volume (CV) in patients undergoing coronary

Materials and Methods
This study followed STROBE methodology 10 . This was an observational, retrospective case-control study that included consecutive patients undergoing coronary angiography from 2012 to 2016. The study was conducted in the Cardiology and Internal Medicine Departments of Hospital Christus Muguerza in Monterrey, Mexico. We included men and women above the age of 18 years, who underwent simple coronary angiography or in conjunction with one of the following: FFR, OCT, IVUS or a combination of these. Patient characteristics were obtained from medical charts and included family history of coronary artery disease, gender, age, BMI, personal history of dyslipidemia, type 2 diabetes mellitus, chronic kidney disease and arterial hypertension. Also we obtained admission diagnosis (stable and unstable angina, non-ST and ST elevation myocardial infarction, heart failure, positive ischemia test), used complementary studies (FFR, OCT, IVUS), number of placed stents, severity of coronary lesion, vascular approach, vascular approach-related complications, and in-hospital stay (days). RD in milli-gray (mGy) and FT (min) were obtained with a General Electric Innova 3100 fluoroscope; CV (mL) was extracted from medical records. We excluded patients who underwent aortocoronary bypass and those with incomplete anthropometric or procedure information. We eliminated patients with left femoral and radial access because of the small sample size. Statistical analysis. Continuous variables were expressed as means and standard deviations (SD) while categorical variables were expressed as frequencies and percentages. Normality was explored for continuous variables by computing skewness and kurtosis and applying the Shapiro-Wilk test. Log-normalization was used when necessary. We used two sample t-test and chi square for group comparisons. Linear multiple regression models were constructed to predict the effect of multiple variables on FT, RD, and CV in patients undergoing coronary angiography. We use eta-squared (η) as an estimation of variance of the response variable (i.e. RD), explained by the explicative covariable (i.e. FFR). The eta-squared value was computed to calculate the effect size of variables in the models; a value of <0.02 was considered small, 0.02-0.09 medium and >0.09 large. To generalize the models, we used a 10-fold cross-validation. The models were two-sided and a p value < 0.05 was considered significant. There were no missing values. Sample size for a two-sided linear multiple regression model of 15 predictors, effect size f 2 of 0.1, α 0.01, β 0.95 was 182. We used G*Power to calculate sample size and the statistics program R.Studio v 3.4.0. and SPSS version 24.
Ethical approval. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent. Ethics Committee/Institutional Review Board waived the need for informed consent as part of the study approval. This Study has obtained IRB approval from Hospital Christus Muguerza Alta Especialidad and the registration number is CMHAE-047-17.

Use of complementary studies and stent implantation. Complementary studies (FFR, IVUS
or OCT) were used in 293 patients (27.3%) and in 3.3% (n = 35) of cases, at least two were necessary. Table 2 describes the complementary studies required in our population. Stents were implanted in 202 (68.9%) patients who underwent complementary studies. Stent implantation was performed in 60.8% (n = 474) of patients in whom complementary studies were not necessary (n = 780). The mean number of stent implants in both groups was 1.25 (SD 1.22) and 1.1 (SD 1.19), respectively (p = 0.07).
Factors modifying fluoroscopy time. Fluoroscopy mean (SD) time was 13.8 (11.9) min. We computed two linear multiple-regression models in order to evaluate the factors that affected FT ( Table 3). The first model (A) included the number of complementary studies, adjusted by multiple covariates, and the second model (B) evaluated the effect of each complementary study adjusted by multiple covariates: Where FT = Fluoroscopy time, ß0 = intercept, ß1-5 = Covariates estimates, DM2 = Type 2 diabetes (present/absent, ADA criteria), CKD = Chronic kidney disease (present/absent if Glomerular Filtration Rate < 60 ml/min/1.73m 2 ), Complementary Studies = number of complementary studies (1-3), Stent = Number of stents placed. Obtaining:    Factors modifying contrast volume. The mean (SD) CV used for each procedure was 199.6 (111.1) ml.
We computed two linear multiple-regression models (

Discussion
Our study demonstrated that OCT, FFR and IVUS had a small effect on FT. OCT and FFR had a small effect on CV and none of them had an effect on RD.
The number of FFR, IVUS or OCT in patients who underwent coronary angiography did not affect the RD and had a small effect by increasing FT and CV.
For instance, the best predictors of FT (see Table 3, model A and B) were the number of stent placed and the use of complementary studies; nonetheless these models only predicts 22.8% of the variance of FT; this means that most of the FT (77.2%) is caused by other factors. This contradicts the tendency to attribute complementary studies for excessive FT exposure. The same principle applies to RD (model C and D) and CV (model E and F).
There is increasing evidence supporting the use of complementary studies. In the last decade, the use of IVUS increased six-fold and this tendency continues to rise. In the United States, only 6% of coronary angiographies are guided by IVUS or OCT, while in our population, complementary studies were used in 27.2% 11,12 . The availability of these techniques in our center and actual worldwide trends probably explain our results.
Some authors have reported that complementary studies demand operator expertise, they increase coronary angiography time and, hypothetically, RD and CV 1 . Ionizing radiation is associated with an indirect stress response of the heart, endotheliopathy, atherosclerosis, cancer and in vitro apoptosis [5][6][7] . The risk of nephropathy increases when using contrast media. Because of the frequent use of these techniques and the deleterious events associated with radiation and contrast media, it is important to determine the effect of complementary studies on these variables. Our study found that these complementary studies have a minimal effect on radiation and contrast exposure.
A retrospective study by Ntalianis et al. quantified extra angiography time, RD and CV when using FFR in patients who underwent diagnostic angiography 13 . Their results, expressed as a percentage of the entire procedure, showed that FFR used in one artery added an extra 13-26% of FT, 16-30% of RD, and 16-31% of CV, respectively. This study did not include patients undergoing IVUS or OCT, and the interaction between FFR and other radiation and contrast predictors was not assessed (i.e. vascular site access). We included more predictors of radiation and contrast exposure (i.e. lesion severity, number of affected arteries or stents placed) and created multivariate models to observe the effect of IVUS, OCT and FFR on these variables.
Our multivariate models could only predict 22-36% of the variance of FT, RD and CV; this means that other predictors must be measured in the future (Table 3).
This study demonstrated that complementary studies explain only a small percentage of the radiation and contrast media variations when used in coronary angiography. There are many advantages to complementary studies such as decreasing mortality, guidance in stent placement, and better angiography outcomes 3,14 . Patients can avoid unnecessary stent placement.
The number of stents placed was the only variable that showed a strong association with RD, CV and FT, so complementary studies are safe in daily practice.
Male sex was associated with greater radiation and contrast medium exposure even when men were younger than women and had less type 2 diabetes mellitus and arterial hypertension. Arterial hypertension, type 2 diabetes mellitus, BMI and severity of the coronary lesion were associated with a greater CV and RD. This may be due to the fact that these comorbidities are associated with diffuse atherosclerosis, complex plaque characteristics (i.e. diffuse lesions), increased angiography complexity and complication development [15][16][17] . Although CKD was present in a small sample (n = 49), this was the only factor that showed a negative association with FT, RD and CV, but more studies are needed to reach a definite conclusion.
A radial vascular approach has been previously associated with greater RD and FT than the femoral approach 9 . In our models, a femoral approach did not influence the radiation dose or fluoroscopy time but exposed patients to greater CV. The use of the radial artery approach is increasingly more common because of the low complication rate (i.e. hematoma and bleeding) which represents an additional advantage 18 . The femoral vascular approach was associated with a greater number of complications and a prolonged in-hospital stay, as reported in previous studies 19 .
Study limitations. Although this study represents a real-life study, the lack of randomization can lead to selection bias. Another important limitation is that inter-operator variability in coronary angiography (an operator-dependent procedure) was not measured, and one can assume that operator expertise is a factor associated with radiation and contrast exposure.

Conclusion
Complementary studies did not increase RD and had a small effect on FT and CV. Stent placement is the factor that most increased radiation and contrast medium exposure.

Data availability
Data base are available upon request