Author + information
- Received July 20, 2012
- Revision received September 27, 2012
- Accepted October 27, 2012
- Published online March 1, 2013.
- Sanjit S. Jolly, MD, MSc⁎,⁎ (, )
- John Cairns, MD†,
- Kari Niemela, MD, PhD‡,
- Philippe Gabriel Steg, MD§,
- Madhu K. Natarajan, MD, MSc⁎,
- Asim N. Cheema, MD, PhD∥,
- Sunil V. Rao, MD¶,
- Warren J. Cantor, MD#,
- Vladimír Džavík, MD⁎⁎,
- Andrzej Budaj, MD, PhD††,
- Tej Sheth, MD⁎,
- Vicent Valentin, MD‡‡,
- Anthony Fung, MBBS†,
- Petr Widimsky, MD§§,
- Emile Ferrari, MD∥∥,
- Peggy Gao, MSc⁎,
- Barbara Jedrzejowski, MEng⁎,
- Shamir R. Mehta, MD, MSc⁎,
- RIVAL Investigators
- ↵⁎Reprint requests and correspondence:
Dr. Sanjit S. Jolly, Room C3, 118 CVSRI Building, Hamilton General Hospital, 237 Barton Street East, Hamilton, Ontario L8L 2X2, Canada
Objectives The authors sought to compare the radiation dose between radial and femoral access.
Background Small trials have shown an increase in the radiation dose with radial compared with femoral access, but many were performed during the operators' learning curve of radial access.
Methods Patients were randomized to radial or femoral access, as a part of the RIVAL (RadIal Vs. femorAL) trial (N = 7,021). Fluoroscopy time was prospectively collected in 5740 patients and radiation dose quantified as air kerma in 1,445 patients and dose-area product (DAP) in 2,255 patients.
Results Median fluoroscopy time was higher with radial versus femoral access (9.3 vs. 8.0 min, p < 0.001). Median air kerma was nominally higher with radial versus femoral access (1,046 vs. 930 mGy, respectively, p = 0.051). Median DAP was not different between radial and femoral access (52.8 Gy-cm2 vs. 51.2 Gy·cm2, p = 0.83). When results are stratified according to procedural volume, air kerma was increased only in the lowest tertile of radial volume centers (low 1,425 vs. 1,045 mGy, p = 0.002; middle 987 vs. 958 mGy, p = 0.597; high 652 vs. 621 mGy, p = 0.403, interaction p = 0.026). Multivariable regression showed procedural volume was the greatest independent predictor of lower air kerma dose (ratio of geometric means 0.55; 95% confidence interval 0.49 to 0.61 for highest-volume radial centers).
Conclusions Radiation dose as measured by air kerma was nominally higher with radial versus femoral access, but differences were present only in lower-volume centers and operators. High-volume centers have the lowest radiation dose irrespective of which access site approach that they use. (A Trial of Trans-radial Versus Trans-femoral Percutaneous Coronary Intervention (PCI) Access Site Approach in Patients With Unstable Angina or Myocardial Infarction Managed With an Invasive Strategy [RIVAL]; NCT01014273)
Radiation exposure from cardiac procedures is an important risk for both patients and physicians. Radiation has been associated with an increased risk of cancer for patients undergoing cardiac procedures, based upon observational analyses from administrative databases (1). Furthermore, radiation is likely the most important occupational hazard for healthcare providers working in cardiac catheterization laboratories. Small randomized trials and observational studies have suggested that the use of radial access leads to higher radiation doses for both patients and healthcare workers compared with femoral access (2–4). This increase in radiation may be due to the technical challenges of using radial access, including subclavian tortuosity and reduced guide support (5). However, a significant limitation of some of the prior single-center studies is that many may have been performed during the operators' learning curve of radial access.
The RIVAL (RadIal Vs. femorAL) trial was an international multicenter randomized trial of 7,021 patients that compared radial versus femoral access for coronary angiography and intervention in patients with acute coronary syndromes (ACS) (6). The overall trial showed no difference in the primary outcome of death, myocardial infarction, stroke, or non-coronary artery bypass surgery–related major bleeding between groups; however, radial access was associated with a more than 60% reduction in vascular access site complications compared with femoral access. During the course of the trial, detailed data relating to radiation exposure were collected.
The objective of this analysis was to compare the radiation dose with radial versus femoral access as measured by air kerma and dose-area product (DAP) and to determine independent predictors of radiation dose.
The design of the RIVAL trial has been published (7). It was a prospective randomized trial of patients with ACS comparing radial versus femoral access for coronary angiography and same-sitting percutaneous coronary intervention (PCI) if clinically indicated. Between June 6, 2006, and November 3, 2010, 7,021 patients were enrolled from 158 hospitals in 32 countries (6).
Patients were eligible for RIVAL if: 1) they presented with non–ST-segment elevation ACS or ST-segment elevation ACS; 2) they were to be managed with an invasive approach; 3) they had intact dual circulation of the hand documented by Allen's test; and 4) the interventional cardiologist was willing to proceed with either the radial or femoral approach (and had expertise for both, including at least 50 radial procedures within the previous year).
Patients were not eligible for RIVAL if they presented with cardiogenic shock, severe peripheral vascular disease precluding a femoral approach, or prior coronary bypass surgery using both internal mammary arteries.
Radiation measurements in the form of air kerma and DAP were collected in 2,569 patients enrolled in the trial from sites that had the facilities to provide these measurements. Fluoroscopy time was also collected for both diagnostic and PCI procedures in the overall population.
Air kerma is defined as the radiation energy adsorbed per unit mass of air (kg) and was reported in milliGrays (mGy). DAP is a product of the air kerma and exposed area, and was reported in Gy·cm2.
Baseline characteristics were recorded for patients who had radiation measurements reported (air kerma and/or DAP). Median air kerma, DAP, and fluoroscopy times for radial versus femoral access were compared using a Wilcoxon rank sum test because the data were not normally distributed. All analyses were performed as intention-to-treat. A significance level of 0.05 with a 2-sided test was used, and SAS version 9.1 (SAS institute, Cary, North Carolina) was used for analyses.
We performed subgroup analysis identical to those performed in the main RIVAL study, including tertiles of radial PCI volume by center, tertiles of individual radial operator volume, PCI versus no PCI, age <75 years versus ≥75 years, sex, ST-segment elevation myocardial infarction (STEMI) versus non–ST-segment elevation ACS, and body mass index (BMI) <25, 25 to 35, or >35 kg/m2. The tertiles of radial PCI volume by center (expressed as median operator volume at each center) were: 1) low (≤60 radial PCIs per year per operator); 2) intermediate (61 to 146 radial PCIs per year per operator); and 3) high (>146 radial PCIs per year per operator) (6). The tertiles of individual operator volume were: 1) low (≤70 radial PCIs per year per operator); 2) intermediate (71 to 142 radial PCIs per year per operator); and 3) high (>142 radial PCIs per year per operator) (6). Because fluoroscopy time, air kerma, and DAP were not normally distributed, we performed logistic regression for above and below the median to derive an interaction p value for subgroups.
The measured air kerma was logarithmically transformed because the distribution was positively skewed. To determine independent predictors of air kerma, multivariable linear regression was performed utilizing known predictors of radiation dose, which include age, sex, radial access, diabetes, prior coronary artery bypass grafting, BMI, prior peripheral arterial disease, PCI, 100% occlusion of culprit lesion, STEMI, and tertiles of center radial PCI volume and then repeated using tertiles of operator radial PCI volume. Ratios of the geometric means were back transformed to provide a clinically relevant estimate of the predictive value of each of the preceding variables. The same approach was followed for DAP, which also had a positively skewed distribution.
Of the 7,021 patients enrolled in the RIVAL trial, 2,569 patients at 42 centers in 16 countries had a radiation measurement reported in the form of air kerma or DAP for their index procedure. Of these 2,569 patients, 1,445 had air kerma reported and 2,255 had DAP reported, whereas 1,131 had both air kerma and DAP reported.
Baseline characteristics for the patients with radiation measurements, shown in Table 1, are well balanced between radial and femoral access and similar to the overall trial participants.
Overall radial versus femoral access
Fluoroscopy time was increased with radial compared with femoral access (radial 9.3 min vs. femoral 8.0 min, p < 0.0001) (Table 2). Median air kerma was increased nominally with radial compared with femoral access (1,046 vs. 930 mGy, respectively, p = 0.051) (Table 3). Median DAP was not different between radial and femoral access (52.8 vs. 51.2 Gy·cm2, p = 0.83) (Table 4).
Procedural volume and radial versus femoral access
When centers were stratified into tertiles of radial PCI volume, fluoroscopy time was increased with radial compared with femoral access in low-volume (10 vs. 8.5 min, p < 0.001) and middle-volume centers (9.5 vs. 7.8 min, p < 0.001). At the individual operator level, fluoroscopy time was increased in all 3 tertiles of operator volume. There were significant interactions with much smaller differences of fluoroscopy time in high-volume centers (8.3 vs. 8.0 min, p = 0.059, interaction p = 0.021) and among high-volume operators (8.7 vs. 8.0 min, p = 0.024, interaction p = 0.002) (Table 2).
Air kerma was increased with radial versus femoral access only in the lowest-volume radial centers (low 1,425 vs. 1,045 mGy, p = 0.002; middle 987 vs. 958 mGy, p = 0.597; high 652 vs. 621 mGy, p = 0.403; interaction p = 0.026) (Fig. 1,Table 3). When results were stratified by tertiles of operator radial PCI volume, air kerma was only increased in low-volume radial operators with radial compared with femoral access (Fig. 2,Table 3), but the interaction term was not significant (p = 0.263).
When centers were stratified into tertiles of radial PCI volume, DAP was not different between radial and femoral in all 3 tertiles (Table 4). When results were stratified by tertile of individual operator volume, DAP was increased in low-volume radial operators with radial access (low 60 Gy·cm2 vs. 50 Gy·cm2, p = 0.046; middle 55 Gy·cm2 vs. 55 Gy·cm2, p = 0.643; high 42 Gy·cm2 vs. 44 Gy·cm2, p = 0.605), but the interaction term was not significant (p = 0.072) (Table 4).
For fluoroscopy time, there were consistent findings of increased fluoroscopy time for radial versus femoral access in all subgroups as shown in Table 2. Air kerma levels were significantly higher with radial compared with femoral access in only those patients with STEMI (1,272 vs. 1,072 mGy, p = 0.039, interaction p = 0.763) and those who underwent PCI (1,317 vs. 1,230 mGy, p = 0.040, interaction p = 0.085) (Table 3). For DAP, the results were consistent with no differences between radial and femoral access in the additional subgroups (Table 4).
Independent predictors of air kerma were radial center volume, PCI for index procedure, 100% occlusion of culprit lesion, radial access group, female, BMI, and diabetes (Fig. 3). On the basis of the ratios of the geometric means, the 2 most important predictors were PCI (ratio 2.29; 95% confidence interval [CI]: 2.13 to 2.47) and high radial center volume (ratio 0.55; 95% CI: 0.49 to 0.61) (Fig. 3). When the analysis was repeated utilizing radial operator instead of center volume, the overall results were similar, and radial operator volume was a significant predictor of air kerma (ratio 0.74; 95% CI: 0.68 to 0.82) (Online Fig. 1).
Independent predictors of DAP were radial center volume, PCI for index procedure, 100% occlusion of culprit lesion, female, diabetes, prior CABG, BMI, and peripheral vascular disease (Fig. 4). Radial access was not a significant independent predictor of DAP (ratio 1.04; 95% CI: 0.97 to 1.11). When the analysis was repeated using radial operator instead of center volume, the overall results were similar, and radial operator volume was a significant independent predictor of DAP (ratio 0.82; 95% CI: 0.74 to 0.90) (Online Fig. 2).
The RIVAL trial is the first large, international multicenter randomized trial to compare the radiation dose of radial versus femoral access for coronary interventions. Although fluoroscopy time was increased with radial compared with femoral access, radiation dose as measured by air kerma was only nominally increased with radial access, with this difference only occurring in low-volume radial centers. By contrast, radiation dose as measured by DAP was not different between radial and femoral access. Finally, procedural volume appears to be a more important predictor of radiation dose than either radial or femoral access, thereby underscoring the importance of operator and center experience regardless of access site used.
In the overall RIVAL trial, a significant interaction for the primary outcome was observed for radial center volume with a benefit observed for radial access in the highest tertile of center volume. The results of the present analysis reinforce the concept that expertise and training are important for radial access. It was previously unknown whether the observed increase in radiation dose related to radial compared with femoral access could be overcome with increased training. It is clear from this analysis from the RIVAL trial that high-volume centers and operators are able to achieve similar radiation doses as measured by air kerma with either access site approach.
The magnitude of differences observed between low- and high-volume centers was much greater than that between access site approaches. This suggests that operator training and education are very important to reduce the radiation exposure to patients, operators, and healthcare workers working in cardiac catheterization laboratories.
Why a difference in fluoroscopy time and not other measures, such as DAP?
Most of the radiation for PCI is from cine-angiography with only a small minority from fluoroscopy. As a result, fluoroscopy time is a poor surrogate measure of radiation dose for the patient. Small increases in fluoroscopy time during crossing of the subclavian artery or traversing the brachial artery with radial access do not appear to lead to significant increases in total radiation dose compared with femoral access.
Clinicians may ask why a difference was observed for fluoroscopy time and a trend for air kerma but not DAP. A potential reason is that DAP is dependent on the size of the radiation field so that imaging practices such as differential coning by operators could have led to similar DAP between the groups despite nominal differences in air kerma. By contrast, differential coning would not have affected air kerma values.
Previous observational analyses have suggested that radial access may be associated with an increased radiation dose, but this may have been confounded by selection bias (8). The 3 prior randomized trials of radial versus femoral access examining radiation dose were limited by the fact that they were: 1) single-center trials with a small number of operators; 2) did not report air kerma values; and 3) did not report outcomes by procedural volume (2–4). In these trials, 2 demonstrated increased DAP and fluoroscopy time with radial compared with femoral access, and 1 trial showed no difference.
From a patient perspective, the small overall differences in air kerma (approximately 100 mGy or 10% increase) between radial and femoral access for a single procedure is likely to have minimal risk for subsequent cancer based on the available literature. The impact for the operator over a 30-year career is likely to be more significant, with a 10% increase representing an additional 3 years of radiation equivalent. However, this increase associated with radial access can likely be avoided with increased training and procedural volume.
Measures to reduce operator radiation dose should be used irrespective of the access site and include transparent, ceiling-mounted shielding, below-table shielding, and use of stored fluoroscopy when possible (9). Disposable radiation blocking drapes and lead shields draped over the patient may further lower operator radiation dose (10,11).
Personalized dosimeter measurements for healthcare workers performing the procedures were not collected. Therefore, the measurements reported are of the radiation doses received by the patients. The type of x-ray system was not collected as part of the study. As a result, it is possible that centers with more modern systems participated in the radiation substudy because these systems routinely report air kerma and DAP. As a result, these results may be more applicable to centers with modern x-ray systems. Finally, the participation of trainees was not recorded on case report forms, and participation of trainees can increase radiation dose. Despite these limitations, this is the largest randomized trial to assess the radiation dose between radial and femoral access and the effect of procedural volume.
Radiation dose as measured by air kerma is nominally higher with radial versus femoral access, but differences are present only in lower-volume centers and operators. Procedural volume was a more important predictor of radiation dose than choice of access site. Experience is the most important factor in reducing radiation exposure from coronary procedures regardless of whether radial or femoral access is performed.
For supplementary figures, please see the online version of this paper.
The RIVAL trial was funded by a grant from the Canadian Network and Centre for Trials Internationally (CANNECTIN), an initiative of the Canadian Institutes of Health Research. RIVAL began as an investigator-initiated substudy of the CURRENT OASIS 7 trial, which was funded by a grant to the Population Health Research Institute (PHRI) from Sanofi-Aventis and Bristol-Myers Squibb. Dr. Jolly has received an institutional research grant (to PHRI) from Sanofi-Aventis, Bristol-Myers Squibb, and Medtronic; and consulting fees (modest) from Sanofi-Aventis, GlaxoSmithKline, and AstraZeneca. Dr. Cairns has recently chaired or been a member of the DSMBs of the following industry-sponsored trials: PALLAS (Sanofi-Aventis), ACTIVE (Sanofi-Aventis), and AVERROES (Bristol-Myers Squibb); is a consultant for Boehringer Ingelheim, Canada; has received speaker honoraria from Boehringer Ingelheim and Bayer; serves on the advisory boards of Boehringer Ingelheim and Bristol-Myers Squibb/Pfizer; has received research grants from Medtronic and AstraZeneca; and is a steering committee member of the TOTAL trial, which receives funding from Medtronic, Canada. Dr. Steg has received institutional research grants from Servier; consulting or honoraria (modest) from Astellas, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo/Lilly, GlaxoSmithKline, Merck, Otsuka, Roche, Sanofi-Aventis, Servier, and The Medicines Company; and stock options in Aterovax. Dr. Rao has received consultancy honoraria (modest) for Terumo Medical, The Medicines Company, Eli Lilly, and Zoll. Dr. Budaj has received consulting fees (modest) from Sanofi-Aventis, Eli Lilly, Novartis, AstraZeneca, and Merck; and grants from Sanofi-Aventis, Boehringer Ingelheim, GlaxoSmithKline, Bristol-Myers Squibb, and AstraZeneca. Dr. Mehta has received an institutional research grant (to PHRI) from Sanofi-Aventis and Bristol-Myers Squibb; and consulting fees/honoraria (modest) from Abbott Vascular, Sanofi-Aventis, Eli Lilly, and AstraZeneca. Dr. Džavík has received unrestricted research grants and travel grants from Abbott Vascular; and speaker's honoraria from AstraZeneca. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute coronary syndromes
- body mass index
- confidence interval
- dose-area product
- percutaneous coronary intervention
- ST-segment elevated myocardial infarction
- Received July 20, 2012.
- Revision received September 27, 2012.
- Accepted October 27, 2012.
- American College of Cardiology Foundation
- Eisenberg M.J.,
- Afilalo J.,
- Lawler P.R.,
- Abrahamowicz M.,
- Richard H.,
- Pilote L.
- Brueck M.,
- Bandorski D.,
- Kramer W.,
- Wieczorek M.,
- Höltgen R.,
- Tillmanns H.
- Achenbach S.,
- Roper D.,
- Kallert L.,
- et al.
- Jolly S.S.,
- Niemelä K.,
- Xavier D.,
- et al.
- Mercuri M.,
- Mehta S.,
- Xie C.,
- Valettas N.,
- Velianou J.L.,
- Natarajan M.K.
- Lange H.W.,
- von Boetticher H.