A Registry-Based Randomized Trial Comparing Radial and Femoral Approaches in Women Undergoing Percutaneous Coronary Intervention

JACC: CARDIOVASCULAR INTERVENTIONS

VOL. 7, NO. 8, 2014

ª 2014 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER INC.

ISSN 1936-8798/$36.00 http://dx.doi.org/10.1016/j.jcin.2014.04.007

A Registry-Based Randomized Trial Comparing Radial and Femoral Approaches in Women Undergoing Percutaneous Coronary Intervention The SAFE-PCI for Women (Study of Access Site for Enhancement of PCI for Women) Trial Sunil V. Rao, MD,* Connie N. Hess, MD, MHS,* Britt Barham, BA,* Laura H. Aberle, BSPH,* Kevin J. Anstrom, PHD,* Tejan B. Patel, MD,y Jesse P. Jorgensen, MD,z Ernest L. Mazzaferri Jr., MD,x Sanjit S. Jolly, MD,k Alice Jacobs, MD,{ L. Kristin Newby, MD,* C. Michael Gibson, MD,# David F. Kong, MD,* Roxana Mehran, MD,** Ron Waksman, MD,yy Ian C. Gilchrist, MD,zz Brian J. McCourt,* John C. Messenger, MD,xx Eric D. Peterson, MD, MPH,* Robert A. Harrington, MD,kk Mitchell W. Krucoff, MD*

ABSTRACT OBJECTIVES This study sought to determine the effect of radial access on outcomes in women undergoing percutaneous coronary intervention (PCI) using a registry-based randomized trial. BACKGROUND Women are at increased risk of bleeding and vascular complications after PCI. The role of radial access in women is unclear. METHODS Women undergoing cardiac catheterization or PCI were randomized to radial or femoral arterial access. Data from the CathPCI Registry and trial-specific data were merged into a final study database. The primary efficacy endpoint was Bleeding Academic Research Consortium type 2, 3, or 5 bleeding or vascular complications requiring intervention. The primary feasibility endpoint was access site crossover. The primary analysis cohort was the subgroup undergoing PCI; sensitivity analyses were conducted in the total randomized population. RESULTS The trial was stopped early for a lower than expected event rate. A total of 1,787 women (691 undergoing PCI) were randomized at 60 sites. There was no significant difference in the primary efficacy endpoint between radial or femoral access among women undergoing PCI (radial 1.2% vs. 2.9% femoral, odds ratio [OR]: 0.39; 95% confidence interval [CI]: 0.12 to 1.27); among women undergoing cardiac catheterization or PCI, radial access significantly reduced bleeding and vascular complications (0.6% vs. 1.7%; OR: 0.32; 95% CI: 0.12 to 0.90). Access site crossover was significantly higher among women assigned to radial access (PCI cohort: 6.1% vs. 1.7%; OR: 3.65; 95% CI: 1.45 to 9.17); total randomized cohort: (6.7% vs. 1.9%; OR: 3.70; 95% CI: 2.14 to 6.40). More women preferred radial access. CONCLUSIONS In this pragmatic trial, which was terminated early, the radial approach did not significantly reduce bleeding or vascular complications in women undergoing PCI. Access site crossover occurred more often in women assigned to radial access. (SAFE-PCI for Women; NCT01406236) (J Am Coll Cardiol Intv 2014;7:857–67) © 2014 by the American College of Cardiology Foundation.

From the *The Duke Clinical Research Institute, Durham, North Carolina; yUnity Health System, Rochester, New York; zUniversity of South Carolina School of Medicine-Greenville, Greenville, South Carolina; xThe Ohio State University Medical Center, Columbus, Ohio; kMcMaster University, Hamilton, Ontario, Canada; {Boston University School of Medicine, Boston, Massachusetts; #Beth Israel Deaconess Medical Center, Boston, Massachusetts; **Mount Sinai Hospital, New York, New York; yyWashington Hospital Center, Washington, DC; zzPenn State Hershey Medical Center, Hershey, Pennsylvania; xxThe American College of Cardiology, Washington, DC; and the kkStanford University Medical Center, Palo Alto, California. Abbott Vascular, Medtronic Vascular (grant no. A 1054367), Terumo Medical, The Medicines Company, Daiichi Sankyo/Eli Lilly and Company (grant no. H7TUS-X014), ACIST Medical, Guerbet, FDA Office of Women’s Health (grant no. HHSF223201111381P), and The Duke Clinical Research Institute funded this study. None of the funding sources had a role in design and conduct of the study; collection, management,

858

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

I

ABBREVIATIONS AND ACRONYMS ACS = acute coronary syndrome

BARC = Bleeding Academic

schemic heart disease is a leading cause

increase the efficiency of trial operations, we used the

of

National

death

among

women.

Although

Infrastructure

tegies such as antithrombotic therapy and

randomization into an ongoing PCI registry and

revascularization improve outcomes in pa-

serves as a platform for streamlined data collection.

tients with unstable angina and myocardial

CI = confidence interval

infarction (1), they have been studied in predominantly male populations. Compared

Monitoring Board

with men, women are at increased risk of

NCRI = National Cardiovascular

adverse

Research Infrastructure

syndrome (2) and invasive procedures such

OR = odds ratio

as percutaneous coronary intervention (PCI)

PCI = percutaneous coronary

(3). They are also at increased risk of

intervention

bleeding complications of both medical ther-

STEMI = ST-segment elevation

apies for acute coronary syndrome (ACS)

myocardial infarction

Research

(NCRI), a clinical trial infrastructure that incorporates

Research Consortium

DSMB = Data Safety

Cardiovascular

currently recommended treatment stra-

outcomes

after

acute

coronary

(4) and femoral artery access for PCI (5). Bleeding is the most common complication of PCI and is associated with increased short- and longterm morbidity, mortality, and costs (6,7). Using the radial artery as the access site for PCI significantly decreases bleeding and access site complications compared with the traditional femoral artery approach (8), but in contemporary clinical practice, women are much less likely to undergo transradial PCI (9). Women presenting for cardiac catheterization present a unique challenge because, although they are at higher risk of femoral arterial access site bleeding, radial artery access may not be feasible.

METHODS STUDY DESIGN. The SAFE-PCI for Women trial was a

multicenter, prospective, open-label, randomized, controlled clinical trial that used the NCRI as the platform for randomization and data collection. This was an investigator-initiated trial, the design of which has been reported previously (12). All trial management activities including data management and statistical analyses were performed at the Duke Clinical Research Institute. Duke University Medical Center’s Institutional Review Board approved the study, as did the review Boards of each participating center. All subjects provided written informed consent before randomization. None of the trial’s funding sources had a role in the design or implementation of the study or in the reporting of the data. The authors wrote all drafts of the manuscript and vouch for the integrity of and completeness of the data and analyses. Members of the trial committees are listed and the trial protocol is given in the Online Appendix. NATIONAL CARDIOVASCULAR RESEARCH INFRA-

Compared with males, females have smaller radial

STRUCTURE. The NCRI is an investigator network

arteries (10) that may be more prone to spasm, which

created through collaboration between the Duke

is a major cause of radial procedure failure (11). In

Clinical Research Institute and the American College

addition, women have been significantly underrep-

of Cardiology and funded by the National Heart,

resented in previous studies comparing radial with

Lung, and Blood Institute (grant 1RC2HL101512-01).

femoral access. Thus, the role of radial artery access

The network includes sites participating in the Na-

in women undergoing PCI remains unclear. Accord-

tional Cardiovascular Data Registry’s CathPCI Regis-

ingly, we performed a large simple multicenter,

try, an ongoing PCI registry co-sponsored red by the

prospective, randomized trial to determine the effi-

American College of Cardiology and the Society for

cacy and feasibility of transradial PCI in women. To

Cardiovascular Angiography and Intervention. Under

analysis, and interpretation of the data; preparation or approval of the manuscript; and decision to submit the manuscript for publication. Some funding sources reviewed the manuscript before submission. Dr. Rao is a consultant for The Medicines Company and Terumo Medical. Dr. Jolly has received research funding from Medtronic. Dr. Waksman has received consulting fees and honoraria from Biotronik, Medtronic, Boston Scientific, Abbott Vascular, and AstraZeneca; is on Speakers’ Bureau of AstraZeneca; and is a shareholder in Endothelix, Inc. Dr. Mehran has received research grant support from The Medicines Company, Bristol-Myers Squibb, Sanofi-Aventis, Eli Lilly & Co./Daiichi Sankyo, Regado Biosciences, and STENTYS and consulting fees from Abbott Vascular, AstraZeneca, Boston Scientific, Covidien, CSL Behring, Janssen (JNJ), Maya Medical, and Merck and is on the Advisory Board of Covidien, Janssen Pharmaceuticals, and Sanofi-Aventis. Dr. Harrington has received research grants from Merck, The Medicines Company, Sanofi-Aventis, Bristol-Myers Squibb, and Portola Pharma and consulting fees from Merck, Bristol-Myers Squibb, AstraZeneca, Gilead, Jansen/J&J, Merck, WebMD, CSL Behring, MyoKardia, The Medicines Company, and Amgen. Dr. Krucoff has received consulting fees and research support from Medtronic, Abbott Vascular, Terumo, and Acist. Dr. Anstrom has received research support from AstraZeneca, Lilly, Medtronic and consulting fees from Abbott Vascular, AstraZeneca, BMS, Pfizer, GSK, and Ikaria. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received April 3, 2014; accepted April 7, 2014.

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

the construct of the NCRI, the data stream of registry-

and included patients who underwent PCI as defined

participating sites is accessed electronically to popu-

in the following. The primary analysis of efficacy and

late a clinical trial database for patients consented

feasibility was performed in the PCI cohort, and a

and enrolled in the SAFE-PCI for Women trial. Patient

sensitivity analysis was performed in the total ran-

demographics, medical history, concomitant medi-

domized cohort.

cations, procedure details, and in-hospital clinical outcomes routinely entered in the registry’s data collection system using standardized data elements are transferred to an electronic trial case report form. Additional trial-specific information including procedure information related to vascular access site and

RANDOMIZATION. After providing written informed

consent, patients were randomized 1:1 to either radial or femoral arterial access. Randomization was performed via an online randomization module incorporated into the registry trial database.

outcomes of interest not obtained as part of the reg-

STUDY PROCEDURES. Patients assigned to radial

istry are collected using additional electronic case

access underwent radial artery puncture by either

report form pages per usual clinical trial standards.

the counterpuncture technique or the anterior wall

The NCRI computer systems, including randomiza-

technique (14) on the basis of operator preference.

tion components, are formally validated, and the in-

For femoral arterial access, it was recommended that

tegrated registry and trial-specific data conform to

operators use either fluoroscopic or ultrasound guid-

the same quality requirements. Because radial access

ance. Cardiac catheterization and PCI technique was

accounts for a minority of procedures in the United

per operator preference. Adjunctive antithrombotic

States (9), initial study sites were identified through

therapy was recommended for all transradial pro-

the CathPCI Registry on the basis of their actual

cedures at a minimal dose of 40 IU/kg unfractionated

transradial PCI volume as described previously (12).

heparin bolus to minimize the risk of radial artery

Participating sites and principal investigators are lis-

occlusion (15). For transfemoral diagnostic pro-

ted in the Online Appendix.

cedures, antithrombotic therapy was at the discretion

STUDY POPULATION. Women undergoing urgent or

elective PCI or diagnostic cardiac catheterization with the possibility of PCI were considered eligible for the trial if they were older than 18 years of age, able to provide informed consent, and undergoing evaluation or treatment for ischemic heart disease. Exclusion criteria included conditions precluding arterial access in either the femoral or radial artery (peripheral arterial disease severe enough to preclude arterial access, absence of collateral flow in both hands assessed by use of the Allen or the Barbeau (13) test, active hemodialysis fistula or graft in an arm to be used in case of assignment to radial access, inter-

of the operator. Bivalirudin was recommended for PCI in all patients, at a dose per the package insert, in conjunction with aspirin and an oral P2Y 12 inhibitor. Use of glycoprotein IIb/IIIa inhibitors was also by operator preference. An immediate post-procedure “patent” hemostasis technique (16) was recommended for all patients after radial access; femoral arterial sheaths were removed $2 h after the discontinuation of bivalirudin or when the activated clotting time was <150 s if unfractionated heparin was used for the PCI. Manual compression or closure devices were permitted at the operator’s discretion.

national normalized ratio $1.5 in the presence

ENDPOINTS AND DEFINITIONS. There were 2 pri-

of ongoing treatment with vitamin K antagonists,

mary endpoints: the primary efficacy endpoint was

receipt of an oral factor IIa or Xa inhibitor in the 24 h

bleeding or vascular complications requiring inter-

before the PCI procedure), known valvular heart dis-

vention occurring within 72 h of the procedure or

ease requiring valve surgery, planned right heart

by hospital discharge, whichever came first; the pri-

catheterization, primary PCI for ST-segment elevation

mary feasibility endpoint was access site crossover.

myocardial infarction (STEMI), known presence of

Bleeding was defined according to the Bleeding Aca-

bilateral internal mammary coronary bypass grafts,

demic Research Consortium (BARC) definition (17),

participation in an investigational drug or device

and the endpoint of interest for the study included

study within 30 days before enrollment, and planned

a composite of BARC type 2, 3, or 5 bleeding

staged PCI within 30 days after index PCI.

events (definitions given in the Online Appendix).

Two patient cohorts were pre-specified: the total

Vascular complications included a composite of

randomized cohort and the PCI cohort. The total

arteriovenous fistula, arterial pseudoaneurysm, or

randomized cohort included all patients randomized

arterial occlusion requiring intervention. Access site

regardless of whether they underwent PCI. The PCI

crossover was defined as the inability to complete

cohort was a subgroup of the total randomized cohort

the procedure from the assigned arterial access site,

859

860

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014

Radial vs. Femoral Access in Women

AUGUST 2014:857–67

requiring conversion from radial to femoral access or

convened due to a lower than expected rate of the

vice versa for procedure completion. Conversions to

primary efficacy endpoint. The DSMB recommended

the contralateral radial artery in case of assignment

terminating the trial because it was unlikely that

to the radial approach (or contralateral femoral

the trial would show a difference between the access

artery in case of assigned femoral approach) were not

site strategies on the basis of the planned sample size.

considered access site crossovers. Secondary end-

There was no evidence of harm in either arm. The

points included procedure duration, total procedure

steering committee met to discuss the recommenda-

radiation dose to the patient measured as air kerma in

tion and, because there were no safety issues,

milligrays (mGy), total contrast volume used during

decided to continue enrollment until enrollment in a

the procedure measured in milliliters, and the 30-

quality-of-life substudy was completed (300 patients

day occurrence of death, vascular complications (as

undergoing coronary angiography or PCI). The results

defined previously), or unplanned revascularization.

of this substudy will be reported separately.

At 30 days, we also assessed the patient’s access

Descriptive summaries of the distribution of

site preference for their next procedure should the

continuous variables are presented as mean  SD, or

patient require one. Secondary and 30-day outcomes

median, 25th, 75th percentiles, and subject counts.

were assessed only in patients who underwent PCI. A

Categorical variables are summarized in terms of

clinical events committee adjudicated all suspected

frequencies and percentages. The primary analysis of

primary efficacy endpoint events. Because the risk of the primary efficacy endpoint

efficacy and feasibility was performed by modified intention-to-treat in the PCI cohort, and sensitivity

(bleeding or vascular complications) is dependent

analyses of efficacy and feasibility were performed in

on the combination of intensive anticoagulation

the total randomized cohort. Odds ratios for the pri-

and arteriotomy, a patient was grouped in the PCI

mary efficacy and feasibility endpoints were gener-

cohort if a coronary guidewire exited the coronary

ated using logistic regression with indicator variables

guide catheter for the purpose of diagnosis or treat-

for randomized assignment, planned use of glyco-

ment, and systemic antithrombotics were given to

protein IIb/IIIa inhibitors during PCI, and elective

achieve therapeutic levels of anticoagulation. This

PCI for stable angina versus ACS. Analyses of sec-

included diagnostic procedures such as fractional

ondary endpoints also followed modified intention-

flow reserve, intravascular ultrasound, and optical

to-treat principles and were conducted within the

coherence tomography as well as coronary angio-

PCI cohort. Secondary analysis of 30-day death, vas-

plasty and/or stenting.

cular complications, or unplanned revascularization used the logistic regression and included indicator

STATISTICAL ANALYSIS. On the basis of the actual

variables for randomized assignment, planned use

rate of bleeding events using a definition approxi-

of glycoprotein IIb/IIIa inhibitors during PCI, and

mating BARC type 2, 3, or 5 bleeding among women

elective PCI for stable angina versus ACS. Analyses

without STEMI undergoing PCI in the CathPCI Reg-

of secondary continuous outcome measures of pro-

istry, we assumed that the rate of bleeding or vascular

cedure duration, total radiation dose, total contrast

complications in the femoral access arm would be 8%

volume were conducted using linear models with a

(18). Using an estimate of bleeding reduction associ-

binary indicator variable for randomized assignment.

ated with radial access from observational and

Two pre-specified subgroup analyses for the pri-

previous clinical trial data (8), we assumed a 50%

mary efficacy endpoint were performed: ACS versus

decrease in the rate of the primary efficacy endpoint

non-ACS in the PCI cohort and quartiles of site radial

in the radial access arm. A sample size of 1,576 women

volume in the total randomized cohort. A post-hoc

undergoing PCI provided 90% power to detect sta-

interaction for the primary efficacy endpoint was

tistically significant differences at a 2-sided alpha

examined in the total randomized cohort between

of 0.05. The sample size was set at 1,800 patients

patients who underwent PCI and those who did not.

(900 patients per arm) due to uncertainty in event

We also performed an analysis according to the access

rates. Because of the common practice of ad-hoc PCI,

site used to complete the procedure (as treated

we planned to randomize 3,000 women undergoing

analysis) in both the randomized and PCI cohorts.

cardiac catheterization to obtain 1,800 undergoing

Exploratory comparisons between the radial and

PCI. No formal power calculation was used for the

femoral groups were conducted for 72 h or hospital

primary feasibility endpoint.

discharge bleeding events using the Thrombolysis In

After 1,120 patients had been randomized (446

Myocardial Infarction (19) and ACUITY (Acute Cathe-

of whom had undergone PCI), an unplanned meeting

terization and Urgent Intervention Triage Strategy)

of the Data Safety Monitoring Board (DSMB) was

(20) trial definitions, as well as bleeding and vascular

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

complications that were identified by sites and re-

PCI, at 60 sites in the United States; 44 of 62 (70.9%)

ported in the CathPCI Registry. Except for Throm-

activated sites enrolled at least 10 patients. Of the

bolysis In Myocardial Infarction major, the definitions

total randomized cohort, 893 were assigned to radial

were slightly modified on the basis of data availability

access and 894 were assigned to femoral access; of

(Online Appendix). These exploratory endpoints were

the PCI cohort, 345 and 346 were assigned to radial

reconstructed from events reported by sites and were

and femoral access, respectively (Fig. 1). Follow-up

not adjudicated. A 2-sided p value # 0.05 was used

for the primary endpoints was available for 99.9% of

for statistical significance. All analyses were per-

the PCI cohort and 99.3% of the total randomized

formed at the Duke Clinical Research Institute using

cohort. Baseline characteristics of the total random-

SAS version 9.2 (SAS Institute, Cary, North Carolina).

ized and PCI cohorts are shown in Tables 1 and 2. There were no significant differences between the

RESULTS

randomized groups in either cohort. The majority of procedures were performed for non–ST-segment

Between September 2011 and July 2013, 1,787 women

elevation ACS. Unfractionated heparin was given

were randomized, 691 (38.7%) of whom underwent

during the diagnostic catheterization to 65.9% of the

F I G U R E 1 Randomization and Follow-up

Women undergoing urgent or elective PCI or coronary angiography with the possibility of PCI were randomized to either radial or femoral arterial access. Shown are the number of patients in each group, the number of patients in each group who underwent PCI, and the number of patients assessed at 30 days for the secondary endpoint. FFR ¼ fractional flow reserve; IVUS ¼ intravascular ultrasound; OCT ¼ optical coherence tomography; PCI ¼ percutaneous coronary intervention.

861

Rao et al.

862

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

femoral groups, respectively. Vascular closure de-

T A B L E 1 Baseline Characteristics of the Total Randomized Cohort

Age, yrs 2

BMI, kg/m

vices were used for hemostasis in 65.5% of the pa-

Radial (n ¼ 893)

Femoral (n ¼ 894)

p Value

63.3 (55.1, 72.2)

63.9 (55.7, 72.0)

0.77

30.5 (26.1, 35.1)

30.8 (26.5, 35.8)

0.29

tients assigned to femoral access. The primary efficacy and feasibility outcomes and secondary outcomes for the PCI and total ran-

243 (27.2)

216 (24.2)

0.14

domized cohorts are shown in Table 3. The rate of

Hypertension

710 (79.5)

714 (79.9)

0.85

BARC type 2, 3, or 5 bleeding or vascular complica-

Previous MI

160 (17.9)

175 (19.6)

0.37

tions requiring intervention in the femoral arm of the

40 (4.5)

57 (6.4)

0.08

PCI cohort, which was the primary analysis cohort,

3 (0.3)

3 (0.3)

0.99

was lower than predicted (2.9%). Compared with

51 (5.7)

54 (6.0)

0.77

314 (35.2)

313 (35.0)

0.95

femoral access, radial access reduced the primary

Current or recent smoking

Previous CABG Hemodialysis Peripheral arterial disease Diabetes mellitus Clinical presentation

0.66

Non-ACS

418 (46.8)

389 (43.5)

NSTEACS

471 (52.7)

503 (56.3)

STEMI

4 (0.4)

efficacy endpoint in this cohort by 60% in the PCI cohort, which did not reach statistical significance. There were no significant interactions noted in the pre-specified subgroups of ACS versus non-ACS in

2 (0.2)

the PCI cohort (Fig. 2). In the total randomized Values shown are number (%) or median (25th, 75th percentiles). BMI ¼ body mass index; MI ¼ myocardial infarction; CABG ¼ coronary artery bypass grafting; ACS ¼ acute coronary syndrome; NSTEACS ¼ non–ST-segment elevation acute coronary syndrome; STEMI ¼ ST-segment elevation myocardial infarction.

cohort, radial access significantly reduced bleeding and vascular complications by 70% compared with femoral access. There was no significant interaction by volume of radial procedures performed at participating sites; similarly, the post-hoc analysis of the

patients assigned to radial access and 25.7% of the patients assigned to femoral access in the PCI cohort. Bivalirudin was used as the anticoagulant during PCI for 59.1% and 65.8% of procedures in the radial and

interaction between patients who underwent PCI and those who did not was also not significant (p ¼ 0.58) (Fig. 2). With respect to the primary feasibility endpoint, the rate of conversion to femoral access from radial access was significantly higher than the rate of conversion to radial access from femoral access in both cohorts. Only 1 patient, who was assigned to femoral access, did not have her diagnostic cathe-

T A B L E 2 Baseline Clinical and Procedure Characteristics of the PCI Cohort

terization procedure successfully completed from

Age, yrs

Radial (n ¼ 345)

Femoral (n ¼ 346)

p Value

either access site. The major reason for conversion

65.1 (56.5, 73.7)

63.9 (56.5, 72.9)

0.32

from radial to femoral access was radial artery spasm

30.1 (25.9, 34.5)

30.5 (26.9, 35.4)

0.14

occurring in 42.9% of crossover patients in the PCI

Current or recent smoking

106 (30.7)

102 (29.5)

0.72

cohort and 43.6% of crossover patients in the total

Hypertension

296 (85.8)

294 (85.0)

0.76

randomized cohort. Other reasons for access site

Previous MI

82 (23.8)

96 (27.7)

0.23

crossover are listed in the Online Appendix, as are

Previous CABG

25 (7.2)

34 (9.9)

0.22

2 (0.6)

2 (0.6)

>0.99

BMI, kg/m

2

Hemodialysis Peripheral arterial disease

the primary efficacy results in the as-treated analysis in both cohorts, which were similar to the intention-

23 (6.7)

29 (8.4)

0.39

144 (41.7)

154 (44.5)

0.46

Non-ACS

98 (28.4)

94 (27.2)

total radiation dose were not statistically different

NSTEACS

244 (70.7)

251 (72.5)

between the radial and femoral access groups

Diabetes mellitus Clinical presentation

STEMI

0.50

3 (0.9)

1 (0.3)

PCI status* 133 (46.5)

123 (43.6)

Urgent

149 (52.1)

157 (55.7)

4 (1.4)

2 (0.7)

166 (59.1)

181 (65.8)

0.10

Glycoprotein IIb/IIIa inhibitor*

32 (11.4)

32 (11.6)

0.93

Vascular closure device*

15 (5.1)†

Bivalirudin*

194 (65.5)

icantly higher among patients assigned to femoral access. Thirty-day follow-up was available for 582 patients in the PCI cohort (84.2%). The rate of 30-day

< 0.001

Values shown are number (%) or median (25th, 75th percentiles). *Excludes patients who underwent fractional flow reserve, intravascular ultrasound, or optical coherence tomography without coronary angioplasty or stenting. †Patients who had any femoral arterial access. PCI ¼ percutaneous coronary intervention; other abbreviations as in Table 1.

Secondary outcomes of procedure duration and

(Table 4); the mean total contrast volume was signif0.49

Elective Emergent

to-treat results.

death, vascular complications, or unplanned revascularization was not significantly different between the 2 arms. Exploratory analyses using different bleeding definitions also showed no significant differences in either the total randomized or PCI cohort (Table 5). At 30-day follow-up, 71.9% of patients assigned to the radial approach preferred radial access for their next procedure; 23.5% of patients

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

863

Radial vs. Femoral Access in Women

assigned to the femoral approach preferred femoral access for their next procedure.

T A B L E 3 Primary Efficacy and Feasibility Endpoints in the Total Randomized

and PCI Cohorts PCI Cohort

DISCUSSION The SAFE-PCI for Women trial represents several “firsts” for clinical trials. It is the first U.S.-based

Radial (n ¼ 345)

Femoral (n ¼ 345)

OR (95% CI)

4 (1.2)

10 (2.9)

0.39 (0.12–1.27)

0.12

3.65 (1.45–9.17)

<0.01

Femoral (n ¼ 884)

OR (95% CI)

p Value

5 (0.6)

15 (1.7)

0.32 (0.12–0.90)

0.03

Type 2

4 (0.4)

10 (1.1)

Type 3

1 (0.1)

3.70 (2.14–6.40)

<0.01

BARC type 2, 3, or 5 bleeding or vascular complications

multicenter, prospective, randomized trial comparing

Type 2

3 (0.9)

6 (1.7)

radial and femoral approaches to cardiac catheteriza-

Type 3

1 (0.3)

4 (1.2)

tion or PCI and is the first randomized trial of PCI

Type 5

0

0

0

0

1 (0.3)

0

strategies performed solely in women. In addition, it is the first registry-based trial in the United States with patient-level randomization. Importantly,

Arteriovenous fistula Arterial pseudoaneurysm Arterial occlusion Access site crossover

this registry-trial infrastructure provided several

This study was terminated early at the recommendation of the DSMB on the basis of a lower than expected rate of bleeding or vascular complications. Although there was a trend toward benefit, radial

6 (1.7)

Radial (n ¼ 891)

workload. Thus, the results and methodology of the model for future clinical research.

0

21 (6.1)

Total Randomized Cohort

efficiencies in site selection, data collection, and site trial inform clinical practice and provide a unique

0

p Value

BARC type 2, 3, or 5 bleeding or vascular complications

Type 5

0

Arteriovenous fistula

0

Arterial pseudoaneurysm

access did not significantly reduce bleeding or

Arterial occlusion

vascular complications among women undergoing

Access site crossover

1 (0.1)

4 (0.5) 0 0 1 (0.1)

0

0

60 (6.7)

17 (1.9)

PCI. The PCI cohort was essentially a subgroup of the total randomized cohort in which radial access did

BARC ¼ Bleeding Academic Research Consortium; CI ¼ confidence interval; OR ¼ odds ratio; PCI ¼ percutaneous coronary intervention.

significantly reduce bleeding and vascular complications. Thus, this lack of a significant difference may be due to the limited sample size of women undergoing PCI. There was no significant interaction be-

proven to be as safe as the radial approach. Taken

tween patients undergoing PCI and those undergoing

together with previous studies, the results of the

diagnostic catheterization for the primary efficacy

SAFE-PCI for Women trial suggest that an initial

endpoint, suggesting that the benefit of radial access

radial access strategy is reasonable in women under-

is consistent across both the PCI and total randomized

going cardiac catheterization or PCI with the recog-

cohorts. Although the reduced sample size limits the

nition that a proportion of patients may require

statistical certainty of this interpretation, it is

conversion to femoral access.

encouraging that the findings in the total cohort of

Although ischemic heart disease is a leading cause

women undergoing diagnostic catheterization or PCI

of mortality in women, they are significantly under-

are consistent with the benefits of the radial approach

represented in clinical trials. For example, women

to bleeding reduction in mixed-sex populations (8).

have often comprised less than a third of the pop-

Also consistent with previous studies was the need to

ulations of clinical trials of recommended secondary

convert to femoral access from radial, which was

prevention strategies (23). In addition, trials that have

significantly higher than the need to convert to radial

defined the role of revascularization strategies for

access from femoral access. The major reason was

high-risk ACS patients or failure of medical therapy

arterial spasm, and the rate of conversion was similar

in stable angina have included mostly men (24,25).

to that seen in the RIVAL (Radial Versus Femoral

There is reason to believe that women may respond to

Access

treatment strategies differently from men, and it is

for

Coronary

Intervention)

trial,

which

included both men and women (21). Future iterations

clear that women are at significantly higher risk of

in PCI catheters and devices, such as improved hy-

bleeding with antithrombotic therapy for ACS and

drophilic

extremely

after PCI (18). Such bleeding complications often

slender 3- and 4-French systems (22), may permit the

coatings

for

catheters

and

occur at the vascular access site and are associated

application of the radial approach to more women

with subsequent adverse events such as myocardial

and reduce access site crossover rates. In addition,

infarction, stroke, stent thrombosis, and even death

ulnar artery access may also be an alternative if

(26). The majority of PCI procedures in the United

Rao et al.

864

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

F I G U R E 2 Odds Ratios for the Primary Efficacy Endpoint in Subgroups of Patients

Odds ratios for the incidence of Bleeding Academic Research Consortium type 2, 3, or 5 bleeding events or vascular complications requiring intervention occurring within 72 h of the index PCI or at hospital discharge, whichever occurred first. Analysis of the interaction between ACS and non-ACS patients was performed in the PCI cohort; analysis of the interaction by site radial volume and between patients who underwent PCI versus those who underwent diagnostic catheterization was performed in the total randomized cohort. Analysis of the interaction between patients who underwent PCI versus those who underwent diagnostic catheterization was performed post-hoc. ACS ¼ acute coronary syndrome; PCI ¼ percutaneous coronary intervention.

States are performed via femoral artery access (9).

necessitating prolonged post-PCI bed rest. In addi-

The femoral artery is large and readily accommodates

tion, the femoral artery is deeply situated, particu-

PCI equipment, and spasm is rare. On the other

larly in overweight or obese patients, can have

hand, hemostasis after femoral access requires the

atherosclerotic narrowing, needs relatively intense

effect of procedural anticoagulation to wane, thus

manual compression to achieve hemostasis after PCI, and is difficult to monitor for occult bleeding such as retroperitoneal hematomas, which are associated with increased morbidity and mortality (26).

T A B L E 4 Secondary Endpoints

Procedure duration, min Total radiation dose, air kerma, mGy Total contrast volume, ml 30-day death, vascular complications, or unplanned revascularization*

Several strategies have emerged to reduce the risk

Radial (n ¼ 290)

Femoral (n ¼ 292)

p Value

of post-PCI bleeding and are termed “bleeding

51.6  32.3

49.9  30.5

0.46

avoidance strategies,” which include the direct

1,604.0  1,394.0

1,472.3  1,274.1

0.26

thrombin inhibitor bivalirudin, potentially the use of

152.7  76.9

165.6  82.7

0.03

vascular closure devices, and radial access (27). In

15 (5.2)

10 (3.4)

0.26

observational studies, radial artery access is associated with a large reduction in access site bleeding and

Patient preference for next procedure† Radial

230 (71.9)

Femoral

26 (8.1)

Either

64 (20.0)

82 (25.7)

major vascular complications compared with femoral

75 (23.5)

access (8). The radial artery is superficial and readily

162 (50.8)

compressible and rarely has atherosclerosis. In patients at high risk of bleeding, such as those with

Values shown are mean  SD or number (%). *Odds ratio: 1.58; 95% confidence interval: 0.71 to 3.48. †Available in 320 patients assigned to radial access and 319 patients assigned to femoral access.

STEMI, radial access for primary PCI may also reduce mortality by reducing major bleeding events (28).

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

T A B L E 5 Bleeding Outcomes at 72 H or Hospital Discharge Using

Different Bleeding Definitions in the Total Randomized Cohort Radial (n ¼ 891)

Femoral (n ¼ 884)

and a database that is compliant with the U.S. Food and Drug Administration guidelines on electronic records (30), was w$5 million, which is far less than a comparably sized trial would have cost without the

TIMI major

3 (0.3)

5 (0.6)

NCRI model. This model is an ideal platform for

TIMI minor

3 (0.3)

5 (0.6)

more efficient and productive investigational drug or

ACUITY major

10 (1.1)

10 (1.1)

device studies. Conducting clinical trials for innova-

CathPCI Registry bleeding

10 (1.1)

12 (1.4)

tive new therapies is increasingly becoming cost

1 (0.1)

1 (0.1)

prohibitive; in this context, the NCRI model appears

CathPCI Registry vascular complications

Events are reconstructed from site-reported events and not adjudicated; see Online Appendix for definitions.

to be a promising platform for future clinical investigations (31).

TIMI ¼ Thrombolysis In Myocardial Infarction.

STUDY LIMITATIONS. First, the trial was terminated

early, limiting the statistical power to detect the However, its small caliber can exclude the use of

effect of radial access on outcomes in women,

large-bore catheters in some patients, and arterial

particularly in the subgroup of women undergoing

spasm is a major cause of radial access site crossover

PCI. However, the benefit of the radial access on the

necessitating conversion to the femoral approach (11).

primary efficacy endpoint was also larger than ex-

Women, in particular, are noted to have smaller

pected, thus permitting detection of a statistically

caliber radial arteries than men (10), and thus spasm

significant benefit in the total randomized cohort.

may be a more prevalent technical issue. Hence, our

Although we used the CathPCI Registry to estimate

rationale for performing the SAFE-PCI for Women

the rate of bleeding or vascular complications, the

trial was rooted in the lack of data on radial access

rate in the trial was much lower than expected. This

in women and the equipoise that although the

low underlying rate of the primary endpoint could

radial approach may reduce post-procedure bleeding,

be attributed to vascular access expertise at the

it may require conversion to femoral access in some

sites that participated in the trial compared with the

patients.

other sites in the registry, the Hawthorne effect

Performing a study to address this equipoise in

(32), the preponderance of bivalirudin use, or a

an underrepresented patient subgroup like women

combination

posed several challenges including identifying in-

vestigators who participated in the trial were all

vestigators with requisite transradial procedure pro-

experienced in using the radial approach. It is likely

ficiency and the costs of conducting a prospective,

that these results may not translate to novice radial

randomized trial (12,29). An important aspect of the

approach operators who may initially experience

SAFE-PCI for Women trial was the embedding of

higher rates of conversion to femoral approach.

the randomization into an ongoing PCI registry. At

Studies indicate that although the learning curve for

the

access

transradial procedures is not steep (33), operators

accounted for <10% of PCI procedures performed in

interested in adopting radial access may have to

the United States (9). Therefore, it was fundamentally

contend with a transient increase in the conversion

important to include sites with proficiency in the

to femoral access during the learning phase. Finally,

radial approach. Rather than rely on the traditional

we excluded patients undergoing primary PCI in our

method of using site surveys to determine eligibility,

trial due to the relatively lower overall proficiency

which is hampered by recall bias, we leveraged the

with radial approach in the United States at the

CathPCI Registry to identify sites using actual trans-

time that the trial began, which could have led to

radial procedure volumes. This strategy achieved trial

prolonged times to reperfusion. Data that emerged

metrics for enrollment that far exceed traditional

during the conduct of the SAFE-PCI for Women trial

clinical trial operations. More than 96% of sites

suggests that transradial primary PCI may lower

enrolled at least 1 patient and nearly 71% of sites

mortality compared with the femoral approach (34).

enrolled at least 10 patients. In addition, a large

On the basis of the site activity and experience in

proportion of the trial case report form was autopo-

the SAFE-PCI for Women trial and the demonstrated

pulated with data from the registry, thus reducing

efficiency of the NCRI model, a randomized trial

site-level workload. These efficiencies in workflow

comparing both clinical outcomes and door-to-

translated into faster enrollment and decreased trial

balloon time metrics between radial and femoral

costs. The overall budget for the SAFE-PCI for

access in STEMI patients seems both warranted and

Women trial, which included endpoint adjudication

feasible.

time

that

enrollment

began,

radial

of

these

factors.

Second,

the

in-

865

866

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

Radial vs. Femoral Access in Women

CONCLUSIONS

from radial access to femoral access in 6.7% of pa-

In the SAFE-PCI for Women trial that was terminated

performed in the United States, this study demon-

early, radial access did not significantly lower the

strates a new paradigm for conducting efficient

incidence of bleeding or vascular complications in

pragmatic clinical trials.

tients. As the first registry-based, randomized trial

women undergoing PCI; however, in the larger sample size of women undergoing cardiac catheterization

REPRINT REQUESTS AND CORRESPONDENCE: Dr.

or PCI, radial access significantly reduced bleeding or

Sunil V. Rao, 508 Fulton Street (111A), Durham, North

vascular complications. There was a need to convert

Carolina 27705. E-mail: [email protected].

REFERENCES 1. Anderson JL, Adams CD, Antman EM, et al. ACC/ AHA 2007 guidelines for the management of

8. Bertrand OF, Belisle P, Joyal D, et al. Comparison of transradial and femoral approaches for

Academic Research Consortium. Circulation 2011; 123:2736–47.

patients with unstable angina/non ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines for the Management of Patients With Unstable Angina/

percutaneous coronary interventions: a systematic review and hierarchical Bayesian meta-analysis. Am Heart J 2012;163:632–48.

18. Rao SV, McCoy LA, Spertus JA, et al. An updated bleeding model to predict the risk of

Non ST-Elevation Myocardial Infarction): developed in collaboration with the American College of Emergency Physicians, the Society for Cardiovascular Angiography and Interventions, and the Society of Thoracic Surgeons: endorsed by the American Association of Cardiovascular and Pulmonary Rehabilitation and the Society for Academic Emergency Medicine. J Am Coll Cardiol 2007;50:652–726. 2. Hochman JS, Tamis JE, Thompson TD, et al. Sex, clinical presentation, and outcome in patients with acute coronary syndromes. Global Use of Strategies to Open Occluded Coronary Arteries in Acute Coronary Syndromes IIb Investigators. N Engl J Med 1999;341:226–32. 3. Lansky AJ, Hochman JS, Ward PA, et al. Percutaneous coronary intervention and adjunctive pharmacotherapy in women: a statement for healthcare professionals from the American Heart Association. Circulation 2005; 111:940–53. 4. Alexander KP, Chen AY, Newby LK, et al. Sex differences in major bleeding with glycoprotein IIb/IIIa inhibitors: results from the CRUSADE (Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the ACC/AHA guidelines) initiative. Circulation 2006;114:1380–7. 5. Doyle BJ, Ting HH, Bell MR, et al. Major femoral bleeding complications after percutaneous coronary intervention: incidence, predictors, and impact on long-term survival among 17,901 patients treated at the Mayo Clinic from 1994 to 2005. J Am Coll Cardiol Intv 2008;1:202–9. 6. Verheugt FW, Steinhubl SR, Hamon M, et al. Incidence, prognostic impact, and influence of antithrombotic therapy on access and nonaccess site bleeding in percutaneous coronary intervention. J Am Coll Cardiol Intv 2011;4:191–7. 7. Amin AP, Marso SP, Rao SV, et al. Costeffectiveness of targeting patients undergoing percutaneous coronary intervention for therapy with bivalirudin versus heparin monotherapy according to predicted risk of bleeding. Circ Cardiovasc Qual Outcomes 2010;3:358–65.

9. Feldman DN, Swaminathan RV, Kaltenbach LA, et al. Adoption of radial access and comparison of outcomes to femoral access in percutaneous coronary intervention: an updated report from the national cardiovascular data registry (2007-2012). Circulation 2013;127:2295–306. 10. Saito S, Ikei H, Hosokawa G, Tanaka S. Influence of the ratio between radial artery inner diameter and sheath outer diameter on radial artery flow after transradial coronary intervention. Catheter Cardiovasc Interv 1999;46:173–8. 11. Dehghani P, Mohammad A, Bajaj R, et al. Mechanism and predictors of failed transradial approach for percutaneous coronary interventions. J Am Coll Cardiol Intv 2009;2:1057–64. 12. Hess CN, Rao SV, Kong DF, et al. Embedding a randomized clinical trial into an ongoing registry infrastructure: unique opportunities for efficiency in design of the Study of Access site For Enhancement of Percutaneous Coronary Intervention for Women (SAFE-PCI for Women). Am Heart J 2013;166:421–428.e421. 13. Barbeau GR, Arsenault F, Dugas L, Simard S, Lariviere MM. Evaluation of the ulnopalmar arterial arches with pulse oximetry and plethysmography: comparison with the Allen’s test in 1010 patients. Am Heart J 2004;147:489–93. 14. Pancholy SB, Sanghvi KA, Patel TM. Radial artery access technique evaluation trial: randomized comparison of Seldinger versus modified Seldinger technique for arterial access for transradial catheterization. Catheter Cardiovasc Interv 2012;80:288–91. 15. Spaulding C, Lefevre T, Funck F, et al. Left radial approach for coronary angiography: results of a prospective study. Catheter Cardiovasc Diagn 1996;39:365–70. 16. Pancholy S, Coppola J, Patel T, RokeThomas M. Prevention of radial artery occlusionpatent hemostasis evaluation trial (PROPHET study): a randomized comparison of traditional versus patency documented hemostasis after transradial catheterization. Catheter Cardiovasc Interv 2008;72:335–40. 17. Mehran R, Rao SV, Bhatt DL, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding

post-procedure bleeding among patients undergoing percutaneous coronary intervention: a report using an expanded bleeding definition from the National Cardiovascular Data Registry CathPCI Registry. J Am Coll Cardiol Intv 2013;6: 897–904. 19. Chesebro JH, Knatterud G, Roberts R, et al. Thrombolysis in Myocardial Infarction (TIMI) Trial, phase I: a comparison between intravenous tissue plasminogen activator and intravenous streptokinase. Clinical findings through hospital discharge. Circulation 1987;76:142–54. 20. Stone GW, McLaurin BT, Cox DA, et al. Bivalirudin for Patients with Acute Coronary Syndromes. N Engl J Med 2006;355:2203–16. 21. Jolly SS, Yusuf S, Cairns J, et al. Radial versus femoral access for coronary angiography and intervention in patients with acute coronary syndromes (RIVAL): a randomised, parallel group, multicentre trial. Lancet 2011;377:1409–20. 22. Matsukage T, Yoshimachi F, Masutani M, et al. A new 0.010-inch guidewire and compatible balloon catheter system: the IKATEN registry. Catheter Cardiovasc Interv 2009;73:605–10. 23. Melloni C, Berger JS, Wang TY, et al. Representation of women in randomized clinical trials of cardiovascular disease prevention. Circ Cardiovasc Qual Outcomes 2010;3:135–42. 24. Clayton TC, Pocock SJ, Henderson RA, et al. Do men benefit more than women from an interventional strategy in patients with unstable angina or non-ST-elevation myocardial infarction? The impact of gender in the RITA 3 trial. Eur Heart J 2004;25:1641–50. 25. Boden WE, O’Rourke RA, Teo KK, et al. Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med 2007;356: 1503–16. 26. Doyle BJ, Rihal CS, Gastineau DA, Holmes DR Jr. Bleeding, blood transfusion, and increased mortality after percutaneous coronary intervention: implications for contemporary practice. J Am Coll Cardiol 2009;53:2019–27. 27. Dauerman HL, Rao SV, Resnic FS, Applegate RJ. Bleeding avoidance strategies. Consensus and controversy. J Am Coll Cardiol 2011;58:1–10.

Rao et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 7, NO. 8, 2014 AUGUST 2014:857–67

28. Romagnoli E, Biondi-Zoccai G, Sciahbasi A, et al. Radial versus femoral randomized investigation in ST-segment elevation acute coronary syndrome: the RIFLE-STEACS (Radial Versus Femoral Randomized Investigation in ST-Elevation Acute Coronary Syndrome) study. J Am Coll Cardiol 2012;60:2481–9. 29. Lauer MS, D’Agostino RB Sr. The randomized registry trial–the next disruptive technology in clinical research? N Engl J Med 2013;369:1579–81. 30. Electronic Records; Electronic Signatures. 21 C.F.R. Part 11 (2013).

Radial vs. Femoral Access in Women

31. Califf RM, Sanderson I, Miranda ML. The future of cardiovascular clinical research: informatics, clinical investigators, and community engagement. JAMA 2012;308:1747–8. 32. McCarney R, Warner J, Iliffe S, van Haselen R,

34. Joyal D, Bertrand OF, Rinfret S, Shimony A, Eisenberg MJ. Meta-analysis of ten trials on the effectiveness of the radial versus the femoral approach in primary percutaneous coronary intervention. Am J Cardiol 2012;109:813–8.

Griffin M, Fisher P. The Hawthorne Effect: a randomised, controlled trial. BMC Med Res Methodol 2007;7:30.

KEY WORDS PCI, radial approach, women and heart disease

33. Ball WT, Sharieff W, Jolly SS, et al. Characterization of operator learning curve for transradial coronary interventions. Circ Cardiovasc Interv 2011;4:336–41.

A PP END IX For the list of trial committee members and the trial protocol, please see the online version of this article.

867