Author + information
- Received February 16, 2017
- Revision received March 8, 2017
- Accepted March 10, 2017
- Published online May 15, 2017.
- Alessandro Sciahbasi, MD, PhDa,
- Enrico Frigoli, MDb,
- Alessandro Sarandrea, Engc,
- Martina Rothenbühler, MScd,
- Paolo Calabrò, MD, PhDe,
- Alessandro Lupi, MDf,
- Francesco Tomassini, MDg,
- Bernardo Cortese, MDh,
- Stefano Rigattieri, MD, PhDa,
- Enrico Cerrato, MDg,
- Dennis Zavalloni, MDi,
- Antonio Zingarelli, MDj,
- Paolo Calabria, MDk,
- Paolo Rubartelli, MDl,
- Gennaro Sardella, MDm,
- Matteo Tebaldi, MDn,
- Stephan Windecker, MDo,
- Peter Jüni, MDd,
- Dik Heg, PD, PhDd and
- Marco Valgimigli, MD, PhDo,∗ ()
- aInterventional Cardiology, Sandro Pertini Hospital, Rome, Italy
- bEustrategy Association, Forlì, Italy
- cDepartment of Radiation Protection, HSE Management, Rome, Italy
- dCTU Bern, and Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland
- eDepartment of Cardio-Thoracic Sciences, Second University of Naples, Naples, Italy
- fCardiology, ASL VCO, Domodossola, Italy
- gDepartment of Cardiology, Infermi Hospital, Rivoli, Italy
- hInterventional Cardiology, Fatebenefratelli Hospital, Milan, Italy
- iHumanitas Research Hospital, IRCCS, Rozzano, Italy
- jInterventional Cardiology Unit, IRCCS AOU San Martino, IST, Genova, Italy
- kCardiology Unit, Misericordia Hospital, Grosseto, Italy
- lVilla Scassi Hospital, Genova, Italy
- mDepartment of Cardiovascular Sciences, Policlinico Umberto I, Rome, Italy
- nCardiology Unit, Azienda Ospedaliera Universitaria di Ferrara, Ferrara, Italy
- oSwiss Cardiovascular Center, Bern University Hospital, Bern, Switzerland
- ↵∗Address for correspondence:
Dr. Marco Valgimigli, Swiss Cardiovascular Center Bern, Bern University Hospital, 3010 Bern, Switzerland.
Background It remains unclear whether radial access increases the risk of operator or patient radiation exposure compared to transfemoral access when performed by expert operators.
Objectives This study sought to determine whether radial access increases radiation exposure.
Methods A total of 8,404 patients, with or without ST-segment elevation acute coronary syndrome, were randomly assigned to radial or femoral access for coronary angiography and percutaneous intervention, and collected fluoroscopy time and dose-area product (DAP). RAD-MATRIX is a radiation sub-study of the MATRIX (Minimizing Adverse Haemorrhagic Events by Transradial Access Site and Systemic Implementation of AngioX) trial. We anticipated that 13 or more operators, each wearing a thorax (primary endpoint), wrist, and head (secondary endpoints) lithium fluoride thermoluminescent dosimeter, and randomizing at least 13 patients per access site, were needed to establish noninferiority of radial versus femoral access.
Results Among 18 operators, performing 777 procedures in 767 patients, the noninferiority primary endpoint was not achieved (p value for noninferiority = 0.843). Operator equivalent dose at the thorax (77 μSv) was significantly higher with radial than femoral access (41 μSv; p = 0.02). After normalization of operator radiation dose by fluoroscopy time or DAP, the difference remained significant. Radiation dose at wrist or head did not differ between radial and femoral access. Thorax operator dose did not differ for right radial (84 μSv) compared to left radial access (52 μSv; p = 0.15). In the overall MATRIX population, fluoroscopy time and DAP were higher with radial compared to femoral access: 10 min versus 9 min (p < 0.0001) and 65 Gy·cm2 versus 59 Gy·cm2 (p = 0.0001), respectively.
Conclusions Compared to femoral access, radial access is associated with greater operator and patient radiation exposure when performed by expert operators in current practice. Radial operators and institutions should be sensitized towards radiation risks and adopt adjunctive radioprotective measures. (Minimizing Adverse Haemorrhagic Events by Transradial Access Site and Systemic Implementation of AngioX; NCT101433627)
Use of radial instead of femoral access for coronary angiography and percutaneous coronary intervention (PCI) has been associated with lower risk of bleeding and vascular complications as well as better survival in patients with acute coronary syndrome (ACS) undergoing invasive management (1,2). European clinical practice guidelines endorsed the use of radial access in patients with non-ST-segment elevation ACS undergoing invasive management with a Class I recommendation over femoral access (3), and the uptake of radial access is increasing worldwide (4).
However, prior studies have raised concerns over the increased risk of radiation exposure for both patients and operators with radial instead of femoral access (5). Only a minority of randomized controlled studies evaluated radiation doses (5), especially in ACS patients (6), and none used dedicated dosimeters to assess operator exposure. As part of the MATRIX (Minimizing Adverse Haemorrhagic Events by Transradial Access Site and Systemic Implementation of AngioX) program (NCT01433627) (7), we collected fluoroscopy time (FT) and dose-area product (DAP), and we equipped radial operators consenting to participate with dedicated dosimeters to assess operator radiation dose with radial or femoral access.
The design of the MATRIX trial and the radiation (RAD-MATRIX) sub-study has been previously reported (7,8). Briefly, all patients with an ACS with or without ST-segment elevation myocardial infarction (STEMI) were randomized to radial or femoral access (Online Appendix). Only expert radial operators (performing more than 100 transradial procedures per year) were involved in RAD-MATRIX.
Before coronary angiography, all patients were centrally randomized (1:1) to radial or femoral access for diagnostic angiography and PCI if clinically indicated. The randomization sequence was computer generated and modified using minimization on intended new or ongoing use of ticagrelor or prasugrel, presence or absence of STEMI, troponin positivity, and anticipated use of immediate PCI in non-STEMI patients. Operators participating in the radiation sub-study were to follow central randomization in regard to radial or femoral access for the primary endpoint comparison (operator radiation exposure at the thorax), and for the patient radiation exposure comparison. A further treatment assignment was performed in patients centrally allocated to radial access based on the patient identification (ID) number with odd ID numbers assigned to right radial and even ID numbers to left radial access. These patient IDs were automatically generated by the centralized web-based randomization and data capture system, so were not under control of the study personnel. This allowed a fairly balanced proportion of right radial access versus left radial access to assess whether the use of left radial as compared to right radial is associated to lower radiation burden (secondary endpoint).
Access site management during and after the diagnostic or therapeutic procedure was left to the discretion of the treating physician and closure devices were allowed as per local practice. Standard operator radioprotection was ensured using a lead apron, a thyroid lead collar, lower body x-ray curtain fixed on the angiographic table, and an upper mobile leaded glass suspended from the ceiling. Staged procedures were allowed, with no restriction with respect to timing, during which the protocol mandated that the access site remain as originally allocated.
Radiation measures collected were FT (expressed in minutes) and the DAP (expressed in Gy·cm2). The DAP is the product of the absorbed dose to air and the cross-sectional area of the x-ray field for all segments of an interventional radiology procedure. This parameter was measured using specially designed ionization chambers mounted at the collimator system and calculated by the software present in each angiographic system. DAP provides a good estimation of the total radiation energy delivered to a patient during a procedure and is correlated with the long-term stochastic risk of cancer (9).
Operator radiation exposure was measured for each participating operator using 3 dedicated lithium fluoride thermoluminescent dosimeters with a range of linearity from 1 μGy to 10 Gy, separate for femoral, left radial, and right radial randomized access site. They were to be worn during each procedure by the participating operator on the left wrist, at mid-thorax level, in the breast pocket outside the lead apron, and at head level (in the middle front to measure the eye dose) (Online Figures 1 and 2). The dosimeters used different detectors according to their location (superficial for the wrist, 3-mm depth for the eye, and 10-mm depth for the thorax). Each dosimeter was distributed to operators in a sealed envelope and was labeled with operator’s code, access site (femoral, right or left radial), and body destination (eye, thorax or wrist: 3 locations × 3 access sites = 9 dosimeters per operator). No protocol violation was declared by participating operators regarding type and position of dosimeters throughout study execution. All dosimeters were collected for central reading at TECNORAD SRL (Verona, Italy) and represent cumulative exposure during all procedures performed by the operator, separated for femoral, left radial, and right radial randomized access site. After central reading and correction for the radiation weighting factor (for x-rays, this factor is 1), the results were expressed as equivalent doses in microSieverts (μSv). The equivalent dose at the thorax was also converted in operator effective dose dividing it by a factor of 33 according to apron thickness, 0.5-mm lead equivalent with a tube voltage under the table (10). Patient effective dose has been calculated using a conversion factor of 0.20 mSv/Gy·cm2, as previously shown (11).
As noted, the primary endpoint of the study was the cumulative operator radiation dose at the thorax. Secondary endpoints included operator radiation dose at left wrist or at head level, patient procedural radiation dose assessed with DAP values, as well as total fluoroscopy time.
The primary noninferiority hypothesis was that radial access was not associated to higher operator radiation dose compared to femoral access (8). Because dosimeters measure the cumulative procedural radiation dose for each operator, the sample size was calculated for the number of operators (i.e., dosimeters) needed rather than for the number of procedures or patients. Using previous information (12), it was estimated that at least 13 operator dosimeters were needed to prove noninferiority with an absolute noninferiority margin of 25 μSv, 1-sided alpha level of 0.05, and 80% power. An arbitrary minimum of 13 procedures per operator and per main access site was mandated to minimize the risks of imbalances due to variation in the complexity of the diagnostic or therapeutic procedures within each operator. The noninferiority test for the primary outcome was performed using a 1-sided unpaired Student t test to estimate the upper bound of the confidence interval of the difference in thorax radiation dosage comparing radial versus femoral on the operator level. Superiority testing for the primary endpoint was performed using 2-sided Wilcoxon rank sum unpaired test. A further secondary analysis using a paired Wilcoxon rank sum test was also performed. Details on the statistical analysis are available in the Online Appendix.
Role of the funding source
The MATRIX program was designed by the last author and approved by the institutional review board at each participating center. The RAD-MATRIX sub-study (8) was pre-specified in the main study protocol and approved by all participating centers as amended number 5 to the original study protocol. MATRIX was sponsored by the Italian Society of Invasive Cardiology, a nonprofit organization, and received grant support from The Medicines Company and TERUMO (Online Appendix).
Between October 2011 and November 2014, 8,404 patients in 78 centers in Italy, the Netherlands, Spain, and Sweden were randomly allocated to radial (n = 4,197) or femoral access (n = 4,207). DAP was collected for 6,902 patients and a total of 7,570 procedures (Online Figure 3). A total of 767 patients undergoing 777 procedures were included in the operator radiation sub-study (RAD-MATRIX) performed by 18 operators (Online Figure 3). Four operators refused to further randomize radial patients to left or right radial access (due to the unwillingness to sustain a prolonged uncomfortable position during left radial access in 3 operators, and in 1 due to perceived lack of clinical equipoise between left and right radial access) and were excluded from this sub-analysis. As a result, 252 radial procedures were performed in 250 patients by 14 operators, which were allocated to left radial (131 procedures in 130 patients) or right radial access site (121 procedures in 120 patients) (Online Figure 3).
Clinical characteristics between radial and femoral groups were similar (Online Table 1). PCI was attempted in more than 80% of the patients in each group (Table 1). Patients allocated to the radial group more frequently received the nonrandomly allocated access than patients in the femoral group (7% vs. 5%: p = 0.0002). In the RAD-MATRIX subsample, crossover rates were balanced in the 2 access groups (Table 1).
In the radial group, median FT was 10.2 min (interquartile range [IQR]: 6 min to 16 min) compared to 9.1 min (IQR: 5.1 min to 15 min) in the femoral group (p < 0.0001) (Table 1). Median DAP values were also higher in the radial group, 64.7 Gy·cm2 (IQR: 28.6 Gy·cm2 to 120.3 Gy·cm2) versus 59.1 Gy·cm2 (IQR: 25.9 Gy·cm2 to 109.5 Gy·cm2) in the femoral group (p = 0.0001) (Table 1). Mean difference of DAP values between radial and femoral access stratified for pre-specified subgroups is shown in Online Figure 4. The results were consistent according to the angiographic system used (Online Table 2). FT and DAP values were consistently correlated in the radial (R = 0.56) as well as in the femoral group (R = 0.56) (Online Figure 5).
The primary noninferiority hypothesis was not reached (mean difference: 34.34 μSv, with an upper 95% confidence limit of 49.57; p value for noninferiority = 0.843). The median operator dose per procedure at the thorax level was higher in the radial compared to femoral access group: 77 μSv (IQR: 39.9 μSv to 112 μSv) versus 41 μSv (IQR: 23.4 μSv to 58.5 μSv), respectively (p value for superiority = 0.019) (Central Illustration, Table 2). A paired analysis yielded identical results. After normalization of the operator dose either for FTs or DAP, the difference between radial and femoral access remained significant (Table 2). Procedural operator doses at the left wrist and head levels did not differ, although both were numerically higher with radial access (Central Illustration, Table 2). The higher radiation dose with radial as compared to femoral access was consistent across individual operators (Online Figure 6).
Baseline and procedural features, including DAP and FT, were similar between left and right radial access groups (Online Table 3). Median procedural operator dose at the thorax did not differ significantly in the right radial compared to the left radial access (84 μSv vs. 52 μSv; p = 0.15) (Table 3, Figure 1). Compared to femoral access, radiation dose did not differ for left radial access but was significantly higher with right radial access (Online Tables 4 and 5). The radiation doses at the wrist and head did not differ with right radial compared to left radial access (Table 3, Figure 1).
Our study is the largest to date evaluating radiation exposure in patients and operators during PCI with radial or femoral access. Our main finding was that in the setting of ACS with or without ST-segment elevation, operator and patient radiation exposure was higher with radial compared to femoral access. The average increase in radiation exposure for patients undergoing radial instead of femoral access was relatively small, in the range of 10%. However, compared to femoral access, radial access was associated with an almost 2-fold increase in operator radiation exposure at the thorax level. Our results confirmed previous observations (13) that DAP is a weak predictor of operator exposure.
In a recent meta-analysis, the difference in patient radiation exposure with radial versus femoral access was shown to narrow over time, suggesting that this difference may not be present in current practice with experienced radial operators (5). Conversely, our findings supported the notion that this difference persists in contemporary practice with experienced operators and it might be much greater than previously anticipated especially for complex multivessel intervention, such as in non-STEMI patients or those with diabetes mellitus.
There are multiple potential explanations for the higher patient and operator radiation exposure associated with radial access. Procedures undertaken via radial access are technically more demanding for operators, especially in the case of tortuosity of the subclavian-aortic axis, which can be observed in up to 30% of patients. More intense catheter manipulation is therefore required to overcome this vascular tortuosity and engage the coronary ostia; whereas the success rate in expert hands is similar to femoral access, these maneuvers increase FT and, consequently, the radiation dose to patients and operators. Our study confirmed previous findings that FT and DAP are correlated and that both are significantly higher in the radial group (5,6).
Other aspects should be considered regarding operator radiation exposure between radial and femoral access. Operator position with respect to the x-ray tube and the patient can affect radiation exposure by a factor of 40 during percutaneous procedures (14). At variance with operator position during femoral access, which is well standardized, operator position during radial access can substantially vary across centers or even within the same center. In many instances, to better manipulate the catheter at the insertion site of the radial artery, operators are closer to the x-ray tube and less shielded by the leaded glass mobile panel. Also, the upper ceiling leaded glass is frequently positioned closer to the patient during radial instead of femoral access, for better direct access to the arterial sheath. Unfortunately, this has translated into less effective shielding of the operator from scatter radiation.
We did not observe a clear difference in terms of operator radiation dose between right and left radial access. Several studies have compared operator radiation dose between radial left or right access with inconsistent results (15–19); some showed less operator radiation dose with left instead of right radial access, whereas others reported similar radiation dose or higher operator dose with left radial access. Various operator positions with respect to x-ray tube and inconsistent locations of mobile shielding devices across operators during right or left radial access might account for such heterogeneous observations.
The absolute increase in DAP values for patients receiving radial instead of femoral access was 5.6 Gy·cm2. This difference was small and, when expressed in terms of patient effective dose, was approximately 1.12 mSv. Considering an additional lifetime cancer risk of 2.5%/Sv (1:40,000) between the ages of 40 years and 60 years (20), radial access would be associated with an increased lifetime cancer risk of 1:35,714 (0.0028%). One could consider this an acceptable risk considering that radial instead of femoral access may avoid 6 deaths for every 1,000 patients treated (1).
At variance from patients, interventional cardiologists perform thousands of procedures during their lifetime, with the potential for a cumulative effect. Operator exposure was almost 2 times higher with a radial than a femoral approach. Most of the operator body is covered with dedicated shields, such as lead apron and thyroid collar but some operator body regions, such as the head or arms, remain unprotected and directly exposed to radiation. Since a direct correlation between dose and cancer risk, even for very low dose radiation exposure, has been suggested (11), then also taking the deterministic risk of radiation into account (i.e., the cumulative risk of cataracts) (21), our findings should raise caution within the medical community. The incremental operator effective dose for a single procedure undertaken with radial instead of femoral access was in the range of 1.1 μSv, corresponding to an additive 330 μSv every 300 procedures. This is similar to an additive radiation exposure of 17 chest x-rays.
Some studies suggested significant reductions in operator radiation doses using adjunctive protective drapes placed on patients (22–24) during radial access. Adjunctive personal protections such as nonlead protective caps that reduce head radiation doses also should be considered (25).
The use of thermoluminescent dosimeters allowed only a cumulative analysis of the operator radiation dose. Hence, further analysis of the radiation dose regarding the complexity of each single procedure performed was not possible to, for example, target procedural improvements to reduce radiation exposure. The use of electronic dosimeters that show radiation dose at the end of each procedure would have allowed a better understanding of which factors might ameliorate, or even negate, the differences in radiation exposure observed between radial and femoral access. However, thermoluminescent dosimeters allowed operators to remain blinded to study results. As per study protocol, we did not standardize patient preparation and set-up for radial access but asked each operator to follow his or her routine practice. The inclusion of 18 experienced operators from different centers likely provided a representative sample of current practice with radial access, but cannot be translated to less-experienced operators or operators with limited training in the radial access site. The consistency of higher operator radiation exposure across participating operators with radial instead of femoral access suggested that the greater radiation dose is a common issue in current practice. The null finding of right versus left-right radial comparison in terms of operator exposure might reflect a power issue and requires further investigation.
Our study showed that radial access was associated with higher operator and patient radiation exposure compared to femoral access. Radial operators and institutions should be sensitized towards radiation risks and adopt adjunctive radioprotective measures.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Radial artery access reduces bleeding and mortality in patients undergoing invasive management of acute coronary syndromes compared to femoral access but entails greater radiation exposure for both the operator and patient.
TRANSLATIONAL OUTLOOK: More work is needed to develop systems that better shield both patients and operators against radiation exposure during interventional cardiovascular procedures without compromising angiographic outcomes.
For a list of investigators, program exclusion and inclusion criteria, additional details on statistical analyses, and supplemental tables and figures, please see the online version of this article.
The MATRIX program is conducted with support from The Medicines Company and Terumo. The sponsor and funders had no role in study design, data collection, data monitoring, analysis, interpretation, or writing of the report. Dr. Rigattieri has received fees from AstraZeneca. Dr. Cortese has received fees from The Medicines Company and AstraZeneca; grants and fees from Abbott Vascular; grants, fees, and nonfinancial support from AB Medica; grants and nonfinancial support from Innova HTS, Kardia, and Stentys; and grants from Hexacath and Amgen. Dr. Windecker has received fees from AstraZeneca; grants from Biotronik, Edwards Lifesciences, and Medtronic; grants and fees from Boston Scientific; fees from Daiichi-Sankyo; and has research contracts with Boston Scientific, Biotronik, Abbot, and Medtronic. Dr. Jüni has received support from Abbott Vascular, Biosensors, Medtronic, Johnson & Johnson, Ablynx, Amgen, AstraZeneca, Boehringer Ingelheim, Eisai, Eli Lilly, Exelixis, Geron, Gilead Sciences, Nestlé, Novartis, Novo Nordisc, Padma, Roche, Schering-Plough, St. Jude Medical, and Swiss Cardio Technologies; is a member of the steering committee or statistical executive committee member for trials funded by Abbott Vascular, Biosensors, Medtronic, and Johnson & Johnson; is on the steering committee for AstraZeneca, Biotronik, Biosensors, St. Jude Medical, and The Medicines Company; and has received grants from AstraZeneca, Biotronik, Biosensors International, Eli Lilly, and The Medicines Company. Dr. Valgimigli has received grants from Terumo and The Medicines Company; has received grants and fees from AstraZeneca; and has received fees from Terumo, Bayer, and Biosensors. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute coronary syndrome
- dose-area product
- percutaneous coronary intervention
- ST-segment elevation myocardial infarction
- Received February 16, 2017.
- Revision received March 8, 2017.
- Accepted March 10, 2017.
- 2017 American College of Cardiology Foundation
- Mehta S.R.,
- Jolly S.S.,
- Cairns J.,
- et al.
- Roffi M.,
- Patrono C.,
- Collet J.P.,
- et al.
- Adamo M.,
- Byrne R.A.,
- Baumbach A.,
- Haude M.,
- Windecker S.,
- Valgimigli M.
- Jolly S.S.,
- Cairns J.,
- Niemela K.,
- et al.
- Sciahbasi A.,
- Calabro P.,
- Sarandrea A.,
- et al.
- Hirshfeld J.W. Jr..,
- Balter S.,
- Brinker J.A.,
- et al.
- ↵NCRP Report N° 122. Use of Personal Monitors to Estimate Effective Dose Equivalent and Effective Dose to Workers For External Exposure to Low-LET Radiation. NCRP Report 1995;1–42.
- Picano E.,
- Vano E.,
- Rehani M.M.,
- et al.
- Casey M.,
- Malone D.
- Theocharopoulos N.,
- Damilakis J.,
- Perisinakis K.,
- Manios E.,
- Vardas P.,
- Gourtsoyiannis N.
- Kado H.,
- Patel A.M.,
- Suryadevara S.,
- et al.
- Sciahbasi A.,
- Romagnoli E.,
- Trani C.,
- et al.
- Kallinikou Z.,
- Puricel S.G.,
- Ryckx N.,
- et al.
- Dominici M.,
- Diletti R.,
- Milici C.,
- et al.
- Pancholy S.B.,
- Joshi P.,
- Shah S.,
- Rao S.V.,
- Bertrand O.F.,
- Patel T.M.
- ↵1990 Recommendations of the International Commission on Radiological Protection. Ann ICRP 1991;21:1–201.
- Authors on behalf of ICRP,
- Stewart F.A.,
- Akleyev A.V.,
- Hauer-Jensen M.,
- et al.
- Lange H.W.,
- von Boetticher H.
- Sciahbasi A.,
- Rigattieri S.,
- Sarandrea A.,
- et al.
- Alazzoni A.,
- Gordon C.L.,
- Syed J.,
- et al.