Transradial Versus Transfemoral Access in Patients Undergoing Rescue Percutaneous Coronary Intervention After Fibrinolytic Therapy

JACC: CARDIOVASCULAR INTERVENTIONS

VOL. 8, NO. 14, 2015

ª 2015 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER INC.

ISSN 1936-8798/$36.00 http://dx.doi.org/10.1016/j.jcin.2015.07.028

Transradial Versus Transfemoral Access in Patients Undergoing Rescue Percutaneous Coronary Intervention After Fibrinolytic Therapy Mitul B. Kadakia, MD,* Sunil V. Rao, MD,y Lisa McCoy, MS,y Paramita S. Choudhuri, PHD,y Matthew W. Sherwood, MD,y Scott Lilly, MD, PHD,z Taisei Kobayashi, MD,* Daniel M. Kolansky, MD,* Robert L. Wilensky, MD,* Robert W. Yeh, MD, MSC,x Jay Giri, MD, MPH*

ABSTRACT OBJECTIVES The purpose of this study was to assess usage patterns of transradial access in rescue percutaneous coronary intervention (PCI) for ST-segment elevation myocardial infarction (STEMI) and associations between vascular access site choice and outcomes. BACKGROUND Transradial access reduces bleeding and mortality in STEMI patients undergoing primary PCI. Little is known about access site choice and outcomes in patients undergoing rescue PCI after receiving full-dose fibrinolytic therapy for STEMI. METHODS Patients in the National Cardiovascular Data Registry’s CathPCI Registry undergoing rescue PCI for STEMI between 2009 and 2013 were studied. Patients were divided on the basis of access site. Patterns of access use and baseline demographics were noted. Unadjusted and propensity-matched analyses were performed comparing in-hospital bleeding, vascular complications, and mortality outcomes among transradial and transfemoral access patients. The falsification endpoint of gastrointestinal bleeding was specified to assess for persistent unmeasured confounding. RESULTS Transradial access was used in 14.2% of cases. In propensity-matched analyses, transradial rescue PCI was associated with significantly less bleeding than transfemoral access (odds ratio [OR]: 0.67; 95% confidence interval [CI]: 0.52 to 0.87; p ¼ 0.003), but not mortality (OR: 0.81; 95% CI: 0.53 to 1.25; p ¼ 0.35). Gastrointestinal bleeding was less frequent in the radial group (OR: 0.23; 95% CI: 0.05 to 0.98; p ¼ 0.05). CONCLUSIONS In a large, “real-world” registry, transradial access was used in a minority of cases and was associated with significantly less bleeding than transfemoral access in patients undergoing rescue PCI. However, given persistent differences in a falsification endpoint, the influence of treatment-selection bias on these results cannot be ruled out. Further studies are needed to determine predictors of bleeding and mortality in this understudied high-risk group. (J Am Coll Cardiol Intv 2015;8:1868–76) © 2015 by the American College of Cardiology Foundation.

From the *Cardiovascular Medicine Division, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania; yDuke Clinical Research Institute, Durham, North Carolina; zDivision of Cardiovascular Medicine, Ohio State University, Columbus, Ohio; and the xDivision of Cardiology, Massachusetts General Hospital, Boston, Massachusetts. Statistical support for this study was funded through a grant from the National Cardiovascular Data Registry. Dr. Rao is a consultant for Terumo Medical, Medtronic, and The Medicines Company (all modest); and has received an honorarium from Terumo Interventional Systems. Dr. Wilensky is a scientific advisory board member for Cardiostem, GenWay, Soteria, and Vascular Magnetics; and has equity interest in Johnson & Johnson. Dr. Yeh is an investigator for the Harvard Clinical Research Institute; has served on the advisory board of Abbott Vascular; and has served as a consultant for Gilead Sciences, Abbott Vascular, and Boston Scientific. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received April 24, 2015; revised manuscript received July 8, 2015, accepted July 30, 2015.

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

F

1869

TR Versus TF Access for Rescue PCI

ibrinolytic therapy is a cornerstone of therapy

Society for Cardiovascular Angiography and

ABBREVIATIONS

in patients with ST-segment elevation myocar-

Interventions. Details of the CathPCI Registry

AND ACRONYMS

dial infarction (STEMI) who present to facil-

have been previously described (15,16). The

ities that do not offer prompt access to primary

registry collects demographic, clinical, pro-

percutaneous coronary intervention (PCI). At least

cedural, and institutional data elements from

12% to 17% of patients in the United States are treated

diagnostic catheterizations and PCI at 1,453

with fibrinolytic therapy (1,2). Up to 40% to 50% of

participating centers. Data is entered via

patients do not achieve restoration of Thrombolysis

a secure web-based platform or software

In Myocardial Infarction flow grade 3 in the infarct-

provided by ACC-approved vendors. There

related artery after administration of full-dose fibri-

is a comprehensive data quality program,

nolytic therapy (3,4) and may undergo urgent cardiac

including specifications for data capture and

catheterization with coronary intervention, so-called

transmission, education and training for site data

“rescue PCI.” In the setting of antithrombotic and

managers, and an auditing program. The data ele-

CI = confidence interval NCDR = National Cardiovascular Data Registry

OR = odds ratio PCI = percutaneous coronary intervention

STEMI = ST-segment elevation myocardial infarction

antiplatelet agents used to treat acute coronary syn-

ments and definitions were created by a specially

dromes, bleeding complications in patients with

assigned ACC committee. The data collection form

STEMI are common and are associated with worse

and a comprehensive description of the data ele-

clinical outcomes and prognosis (5). These risks may

ments are available online (17).

be increased in those receiving fibrinolytic therapy

The study population included adult (age >18

for STEMI prior to cardiac catheterization (6).

years)

Many of these bleeding complications are related

treated with fibrinolytic therapy who subsequently

to the femoral artery access site used for cardiac

underwent rescue PCI via either a transfemoral or

catheterization (7).

transradial approach from July 1, 2009, to June 30, SEE PAGE 1877

patients

presenting with

STEMI

initially

2013. Rescue PCI was defined as PCI for STEMI after failed full-dose fibrinolytic therapy. The following

Several analyses have demonstrated a reduction in

patients were excluded: those who required femoral

bleeding outcomes in STEMI patients undergoing PCI

access for intra-aortic balloon pump or other me-

via the transradial approach (8–14). However, little is

chanical ventricular support; those who underwent

known about differences in bleeding outcomes on

in-hospital coronary artery bypass surgery; those who

the basis of access approach in patients undergoing

underwent >1 PCI during the index hospitalization;

rescue PCI. Given their high-risk clinical presentation

those on hemodialysis; and patients who underwent

and the concurrent use of antiplatelet, anticoagulant,

catheterization >72 h after fibrinolytic administra-

and thrombolytic therapy, these patients may have

tion. Following these exclusions, 9,494 patients from

the most to gain from “bleeding-avoidance strate-

603 sites were included in this analysis (Figure 1).

gies” like radial access. However, patients who fail to reperfuse with thrombolysis represent a high-risk subgroup whose procedural success rates may be worse, and the influence of transradial access upon procedural success and mortality in this group is unknown. The National Cardiovascular Data Registry (NCDR) CathPCI Registry provides a unique opportunity to evaluate access site usage patterns and the differences in bleeding outcomes and mortality in patients undergoing transradial versus transfemoral access following fibrinolytic therapy in STEMI in clinical practice. We hypothesized that transradial access would be associated with improved outcomes in rescue PCI patients.

ENDPOINTS AND DEFINITIONS. The primary endpoint

for this analysis was bleeding complications. Bleeding was defined as the presence of 1 or more of the following within 72 h of PCI: overt access site bleeding, retroperitoneal hemorrhage, intracranial hemorrhage, gastrointestinal or genitourinary bleeding, cardiac tamponade, non–bypass surgery-related blood transfusion in patients with a pre-procedure hemoglobin #8 g/dl, or an absolute decrease in hemoglobin $3 g/dl from pre- to post-PCI in patients with a pre-procedure hemoglobin value of <16 g/dl. Secondary endpoints were in-hospital mortality and other vascular complications requiring treatment. STATISTICAL ANALYSIS. Patient and hospital char-

acteristics were compared by vascular access site used

METHODS

for rescue PCI (radial vs. femoral). These included patient demographics, medical history, risk factors,

DATA SOURCE AND STUDY SAMPLE. The NCDR

intraprocedural and post-procedural events, lesions

CathPCI Registry is an initiative of the American

and devices, discharge, and hospital characteristics.

College of Cardiology (ACC) Foundation and the

Categorical variables are presented as frequencies

1870

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

TR Versus TF Access for Rescue PCI

symptom onset to presentation, and time from fibri-

F I G U R E 1 Patient Selection Process

nolytic therapy to procedure time. Pre– and post– inverse probability of treatment weighting (IPTW) balance of the covariates between radial and femoral patients was also checked. To determine the association between radial versus femoral PCI and in-hospital outcomes, logistic regression with generalized estimating equations to account for within hospital clustering was used. The propensity to undergo radial or femoral access was included in the model by inverse probability of treatment weighting. Weights were stabilized using the marginal probability of radial access to maximize efficiency. For radial patients, the weights were calculated by dividing the marginal probability of radial access by the individual patient’s propensity score. Weights for femoral access patients were calculated as: (1  marginal probability of radial access)/(1  individual patient’s propensity score). Pre- and post-IPTW balance of the covariates between

treatment

groups

was

assessed

using

Cramer’s phi for categorical variables and R 2 for continuous variables. Values closer to 0 indicated better balance. After IPTW adjustment, the Cramer’s phi measure for each categorical variable except for Patient population from the CathPCI Registry selected for the study displayed as a CON-

history of prior coronary bypass surgery was <0.035

SORT diagram. CABG ¼ coronary artery bypass surgery; IABP ¼ intra-aortic balloon pump;

and the R2for each continuous variable was <0.003

PCI ¼ percutaneous coronary intervention; STEMI ¼ ST-segment elevation myocardial infarction.

(Figure 2). IPTW-adjusted subgroup analysis was performed in several key subgroups, including age >75 years, male sex, BMI $30 kg/m 2, diabetic patients, creatinine clearance <45 ml/min, GP IIb/IIIa use,

(percentages), and differences were assessed using

bivalirudin use, and cardiogenic shock. Given the

the chi-square test. Continuous variables are pre-

known issue of residual confounding in observational

sented as medians with interquartile ranges and were

analyses of radial versus femoral access (19), we tested

compared using the Wilcoxon rank-sum test.

a falsification endpoint of gastrointestinal bleeding on

We accounted for potential selection bias in those

both groups. This endpoint served as a “negative

who underwent transradial versus transfemoral pro-

control” in our analysis, as rates of gastrointestinal

model

bleeding were unlikely to be influenced by the choice

for vascular access using logistic regression. Variables

of access site. To further examine the role of baseline

were selected a priori in a nonparsimonious fashion

differences in bleeding risk between the 2 cohorts, we

from the NCDR CathPCI version 4 pre-catheterization

performed an analysis to ascertain the pre-procedure

in-hospital mortality and bleeding models (8,18).

predicted bleeding risk of the 2 groups on the basis

Specifically, we adjusted for sex, age, race, body mass

of a validated NCDR CathPCI bleeding risk model (20).

cedures

by

constructing

a

propensity

index (BMI), hypertension, hyperlipidemia, family

All analyses were performed by the Duke Clinical

history of premature coronary artery disease, prior

Research Institute using SAS version 9.3 (SAS Insti-

myocardial infarction, heart failure history, cerebro-

tute, Cary, North Carolina). This study was approved

vascular disease, peripheral vascular disease, prior

by the Duke University Medical Center institutional

PCI, prior coronary artery bypass graft, diabetes,

review board and was determined to meet the defi-

tobacco use, history of dialysis, glomerular filtration

nition of research not requiring informed consent.

rate, heart failure on presentation (New York Heart Association functional class), cardiac arrest, aspirin

RESULTS

use, thienopyridine use, anticoagulant, glycoprotein (GP) IIb/IIIa use, pre-procedure hemoglobin (as a

BASELINE

continuous linear spline with 1 knot at 13), time from

Among

CHARACTERISTICS

9,494

patients

AND

undergoing

TREATMENT.

rescue

PCI

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

following fibrinolytic therapy for STEMI, 8,146 (85.8%) had femoral access and 1,348 (14.2%) had radial

TR Versus TF Access for Rescue PCI

F I G U R E 2 Covariate Balance Before and After Propensity Matching

access. Those undergoing radial access were more often younger and male, and more often had higher BMI, family history of coronary artery disease, and history of congestive heart failure. They less often had a history of hypertension, dyslipidemia, previous coronary artery bypass graft surgery, cerebrovascular disease, or chronic lung disease (Table 1). On presentation, patients undergoing radial access less often had heart failure within the 2 weeks prior to presentation, cardiogenic shock, and cardiac arrest. Mean glomerular filtration rate was higher in the radial access group. There was no statistically significant difference in baseline hemoglobin levels. The median time from fibrinolytic administration to cardiac catheterization was 189 min (interquartile range: 108 to 918 min). There were no differences in ejection fraction between the 2 groups. Radial access patients more often received aspirin and unfractionated heparin. There were no other statistically significant differences in medical therapies between the 2 groups, including usage of thienopyridines and GP IIb/IIIa inhibitors (Table 2). Patients treated with transradial access had lower pre-procedure predicted bleeding risk scores (10.2  8.2%) than those treated with transfemoral access (12.02  8.5%) on the basis of a validated NCDR CathPCI bleeding risk model (p < 0.0001) (20). ACCESS APPROACH AND CLINICAL OUTCOMES.

Fluoroscopy time was longer in the radial access group (14.4  10.4 min vs. 12.1  9.2 min; p < 0.0001). Contrast volume was greater in femoral access patients (197.1  84.2 ml vs. 180.9  78.4 ml; p < 0.0001). There was no statistically significant

Pre– and post–inverse probability of treatment weighting (IPTW) balance of the covariates between radial and femoral patients for categorical variables (A) and continuous variables (B).

difference in successful dilatation of the culprit lesion

GFR ¼ glomerular filtration rate; HF ¼ heart failure; NYHA ¼ New York Heart Association;

(96.1% in radial access patients vs. 94.7% in femoral

PCI ¼ percutaneous coronary intervention.

patients; p ¼ 0.06) (Table 2). Unadjusted analysis demonstrated a significantly lower incidence of the primary bleeding endpoint in

IPTW-adjusted analysis demonstrated that radial

the radial access group compared with the femoral

access was associated with significantly less bleeding

access group (6.9% vs. 12.0%; p < 0.0001). The

than femoral access (odds ratio [OR]: 0.67; 95% con-

number needed to treat to prevent 1 bleeding

fidence interval [CI]: 0.52 to 0.87; p ¼ 0.003).

complication with the radial approach was 20.

In-hospital mortality was not significantly different

There were fewer blood transfusions (1.1% vs. 2.8%;

between the groups (OR: 0.81; 95% CI: 0.53 to 1.25;

p ¼ 0.0003), vascular complications (0% vs. 0.4%;

p ¼ 0.35) (Figure 3). IPTW-adjusted subgroup analyses

p ¼ 0.02), access site bleeding events (0.2% vs. 1.1%;

in several key subgroups were consistent with the

p ¼ 0.001), access site hematomas (0.22% vs. 1%;

overall results (Figure 4).

p ¼ 0.004), and retroperitoneal bleeding events (0%

The

falsification

endpoint

of

gastrointestinal

vs. 0.4%; p ¼ 0.02) in the radial access group. There

bleeding demonstrated less gastrointestinal bleeding

was also a borderline associated mortality benefit

in the radial group after IPTW adjustment (OR: 0.23;

in the radial group (1.7% vs. 2.6%; p ¼ 0.05) (Table 3).

95% CI: 0.05 to 0.98; p ¼ 0.05).

1871

1872

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

TR Versus TF Access for Rescue PCI

and this may be an underutilized bleeding avoidance

T A B L E 1 Demographics and Prior History of Study Patients

strategy in this understudied patient population. Radial (n ¼ 1,348)

Femoral (n ¼ 8,146)

p Value

Age

57.4  11.2

58.9  11.8

<0.0001

Sex

1,069 (79)

6,160 (76)

0.0033

with this approach had significantly lower pre-

BMI, kg/m2

30.3  6.2

29.7  6.0

0.0004

procedure predicted bleeding risk scores. Despite

Previous MI

207 (15)

1,406 (17)

0.08

the utilization of a nonparsimonious propensity-

28 (2)

270 (3)

0.02

matched model, persistent differences in the falsifi-

307 (23)

1,825 (22)

0.75

Cerebrovascular disease

54 (4)

442 (5)

0.03

cation endpoint of gastrointestinal bleeding between

Peripheral vascular disease

63 (5)

363 (4)

0.72

Previous CHF Diabetes

We observed a “risk-treatment paradox” with the utilization of transradial access, as patients treated

the groups point to the presence of unmeasured

Chronic lung disease

114 (8)

842 (10)

0.03

confounders influencing our adjusted outcomes.

Hypertension

817 (61)

5,188 (64)

0.03

Hence, our results must be interpreted as hypothesis-

Current/recent smoker

664 (49)

3,987 (49)

0.8

generating.

Dyslipidemia

750 (56)

4,922 (60)

0.001

Family history of CAD at age <55 yrs

324 (24)

1,743 (21)

0.03

outcomes

Previous PCI

220 (16)

1,430 (18)

0.27

although nearly all of these studies have either

Previous CABG GFR

23 (2)

450 (6)

<0.0001

82.8  25.4

79.7  30.3

<0.0001

Several prior studies have examined bleeding with

transradial

access

for

STEMI,

excluded or not specifically analyzed rescue PCI patients. Two prior randomized trials and a large

Values are mean  SD or n (%).

registry analysis have demonstrated improvements

BMI ¼ body mass index; CABG ¼ coronary artery bypass graft surgery; CAD ¼ coronary artery disease; CHF ¼ congestive heart failure; GFR ¼ glomerular filtration rate (using the Modification of Diet in Renal Disease study equation); MI ¼ myocardial infarction; PCI ¼ percutaneous coronary intervention.

in outcomes with utilization of transradial access in primary PCI, but rescue PCI patients were excluded from these studies (11,12,14). The RIFLE-STEACS (Radial Versus Femoral Randomized Investigation in ST-Elevation Acute Coronary Syndrome) trial

DISCUSSION

randomized

1,001

STEMI

patients

to

radial

or

femoral access and found lower mortality, bleeding, We found that the majority of patients in the United

and length of hospital stay in the radial access

States undergoing rescue PCI following fibrinolytic

group. Similarly, a prior NCDR CathPCI registry–

therapy for STEMI have the procedure performed via

based analysis of 91,000 United States patients

the transfemoral approach. In patients undergoing

found lower adjusted risk of bleeding and in-

radial catheterization in this setting, procedural suc-

hospital mortality in radial access patients com-

cess rates were high. Our primary analyses demon-

pared with femoral access patients (8). Although

strated lower associated bleeding risk in patients

rescue PCI patients were included in these 2 studies,

undergoing rescue PCI via the transradial approach,

outcomes specific to this important group were not examined.

T A B L E 2 Clinical Presentation and Concomitant Medical Treatments of

Study Patients

Only 2 prior analyses have attempted to address differences in clinical outcomes between rescue PCI patients treated with varying access strategies. A

Heart failure within 2 weeks

Radial (n ¼ 1,348)

Femoral (n ¼ 8,146)

p Value

58 (4)

502 (6)

0.007

2004 study retrospectively examined transradial versus transfemoral access in 111 patients undergoing

Cardiogenic shock

41 (3)

456 (6)

<0.0001

rescue PCI with adjuvant GP IIb/IIIa use (21). Addi-

Cardiac arrest

74 (5)

665 (8)

0.0007

tionally, a 2007 study examined 287 patients under-

Fluoroscopy time

14.4  10.4

12.1  9.2

<0.0001

going rescue PCI, 44 of whom underwent transradial

Contrast volume

180.9  78.4

197.13  84.2

<0.0001

access (22). Both studies suggested improvements

Ejection fraction

47.0  11.7

46.4  11.3

0.17

Aspirin

1,257 (94)

7,349 (91)

0.0007

Clopidogrel

1,002 (74)

6,013 (74)

0.79

213 (16)

1,143 (14)

0.34

1,150 (85)

6,347 (78)

<0.0001

Prasugrel Unfractionated heparin

in bleeding outcomes with use of transradial access. However, these studies were unadjusted analyses that were modest in size. To our knowledge, there has been no large trial

Bivalirudin

450 (33)

2,625 (32)

0.38

or observational analysis with statistical adjustment

Low-molecular-weight heparin

230 (17)

1,475 (18)

0.36

that has analyzed the association of access approach

GP IIb/IIIa inhibitor

424 (32)

2,753 (34)

0.10

with bleeding complications and mortality specif-

Values are n (%) or mean  SD. GP ¼ glycoprotein.

ically in STEMI patients treated with fibrinolytic therapy who have undergone rescue PCI. The current analysis also represents by far the largest population

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

1873

TR Versus TF Access for Rescue PCI

of rescue PCI patients ever studied. Moreover, many prior analyses of access approach are based on clinical

T A B L E 3 Clinical and Bleeding Outcomes of Study Participants

Radial (n ¼ 1,348)

Femoral (n ¼ 8,146)

Lesion successfully dilated

1,295 (96.1)

7,710 (94.7)

0.06

Any vascular complication

0 (0)

33 (0.4)

0.02

trial populations, which exclude the elderly and patients with significant comorbidities who have increased risk of bleeding and in whom pharmaco-

p Value

logical reperfusion is less commonly employed

RBC/whole blood transfusion

15 (1.1)

227 (2.8)

0.0003

(23–25). Our analysis avoids these pitfalls through

Bleeding event within 72 h

18 (1.3)

304 (3.7)

<0.0001

examination of a contemporary patient population

Bleeding at percutaneous entry site

2 (0.2)

86 (1.1)

0.001

across a nationally representative group of hospitals.

Hematoma at access site

3 (0.22)

83 (1)

0.004

Patients undergoing primary PCI for STEMI are at

Retroperitoneal bleeding

0 (0)

33 (0.4)

increased risk of bleeding (20), and this risk may be

Gastrointestinal bleeding

2 (0.2)

69 (0.9)

0.006

Overall bleeding

93 (7)

967 (12)

<0.0001

23 (1.7)

212 (2.6)

0.05

increased in those who receive fibrinolytic therapy prior to PCI (6). Bleeding has been associated with increased mortality in patients presenting with acute coronary syndromes (26). Thus, the reduction of

Death

0.02

Values are n (%). RBC ¼ red blood cell.

bleeding in this group of patients is of particular interest. Despite possible reductions in bleeding via the transradial approach, our data indicates that 85% of patients undergoing rescue PCI are approached via the transfemoral access route. Additionally, patients chosen for treatment with transradial access were actually at lower predicted bleeding risk than those chosen for transfemoral access. This “risk-treatment paradox” in the use of transradial access for PCI has been noted in other studies (27). Our findings may be due to a lack of operator comfort with the transradial approach, lack of awareness of the benefits of transradial approach, and facility/catheterization laboratory staff unfamiliarity with radial setup leading to delays in timesensitive

situations

such

as

acute

myocardial

infarction. Multiple societies aim to educate and increase U.S. operator utilization of the radial approach, particularly

for

higher-risk

patient

populations

(28,29). Of note, there were no associated improvements in in-hospital mortality observed in the current study.

populations that have lower PCI-related rates of mortality. In these groups, despite benefit in terms of bleeding and vascular complications, the mortality benefit of transradial access has been less clear (30). Another possibility is that the post-thrombolytic milieu may create a more favorable environment that leads to less distal embolization and no-reflow phenomenon, improving mortality to such a degree that it eliminates the propensity-matched mortality benefit in the radial group. Alternatively, given the low in-hospital mortality rates seen in the present analysis, it is possible that we did not have the power to accurately assess mortality in this patient group. Our findings raise several interesting questions for future research and development in the understudied

rescue

PCI

population.

Other

bleeding

avoidance strategies, such as the use of bivalirudin, have been shown to reduce bleeding in STEMI, but have not been studied specifically in the rescue PCI population (31). Importantly, in our analysis, even in

This is in contrast to prior findings from the NCDR that examined all patients presenting with STEMI

F I G U R E 3 Adjusted Outcomes of Transradial Versus Transfemoral Access

treated with radial access (8), although it is consistent with the more recent STEMI-RADIAL (ST Elevation Myocardial Infarction treated by RADIAL or femoral approach) trial (30). It is likely that the currently examined rescue PCI population represents a group at intermediate risk of mortality as compared with the STEMI population as a whole. Our patient cohort had lower in-hospital mortality rates (1% to 2%) than allcomers with STEMI (5.5%) (1). In particular, patients requiring intra-aortic balloon pump or those requiring multiple PCIs were excluded from our analysis, because these procedures require additional access

Association between transradial access and in-hospital outcomes, including mortality and

sites that could confound bleeding outcomes. As

the composite bleeding endpoint. Odds ratios and 95% confidence intervals listed on the

such, the current population may have more in

right side of the figure.

common

with

non-STEMI

and

elective

PCI

1874

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

TR Versus TF Access for Rescue PCI

of radial versus femoral access that have been shown

F I G U R E 4 Adjusted Bleeding Outcomes in Subgroups of Interest

to experience possible residual confounding when assessed by falsification endpoints (19). It has proven difficult to obviate this residual confounding despite varying methods of statistical adjustment. The results of our falsification endpoint analysis demonstrate the necessity of employing more stringent methodologies to cardiovascular observational analyses to produce valid data that is unlikely to be contradicted by subsequent randomized studies. In addition to possible residual confounding, our study has limitations inherent to registry analysis. The examined registry only contained data regarding in-hospital mortality. It is possible that radial access could confer a mortality benefit over the longer term (30 days or 2 years), as has been suggested by prior studies of radial access in STEMI (11,13,14). Additionally, it is possible that all bleeding outcomes were not captured, as the data is based on self-reporting by participating hospitals. Although the NCDR conducts

Association between transradial access and bleeding in key subgroups. Odds Ratios and

a robust auditing program, all outcomes are not

95% confidence intervals listed on the right side of the figure. BMI ¼ body mass index;

independently adjudicated as is the case in many

CrCl ¼ creatinine clearance; GP ¼ glycoprotein.

randomized trials. Access site was recorded as the primary access site, and data on access-site crossover was not available. The dataset is limited to those that report to the CathPCI registry. Although this

the subgroup of patients receiving bivalirudin, there

includes the largest contemporary group of in-

was a sustained reduction in bleeding outcomes

stitutions in the world performing PCI and does

suggesting an added benefit of transradial access over

reflect a variety of hospitals by size, location, private/

bivalirudin alone. Additionally, we found increases in

public, and academic/nonacademic, it does not reflect

fluoroscopy time with radial access among rescue PCI

all hospitals in the United States. Finally, analyzed

patients, consistent with prior analyses of both PCI

data are only from the United States, and global

and diagnostic catheterization via the transradial

practice patterns likely vary.

approach (32,33). Although not associated with differences in overall procedural success, this consistent

CONCLUSIONS

finding across studies does underscore the need for continuing technological development that facilitates

In a large, “real-world” registry, transradial access

transradial procedures as well as continuing educa-

was used in 15% of patients undergoing rescue PCI

tion to increase operator experience and skill from

with high procedural success rates. Propensity-

the transradial approach.

matched

analyses

suggested

significantly

less

bleeding associated with a transradial approach in STUDY LIMITATIONS. We specified a falsification

this population. Given persistent differences in a

endpoint of gastrointestinal bleeding to formally

falsification endpoint, the influence of treatment-

assess for possible residual confounding in this

selection bias on these results cannot be ruled out.

observational analysis. Despite robust statistical

However, given the lack of research regarding

adjustment and the use of a nonparsimonious pro-

bleeding avoidance in rescue PCI, the present study

pensity model, there may still be unmeasured resid-

is likely to represent the best available data in this

ual confounders in the association among patient

area for the foreseeable future.

characteristics,

access

approach,

and

outcomes.

This is reflected by the persistent difference in a

REPRINT REQUESTS AND CORRESPONDENCE: Dr.

bleeding outcome that is unlikely to be directly

Jay Giri, Cardiovascular Medicine Division, Hospital

related to access approach (gastrointestinal bleeding)

of the University of Pennsylvania, Gates Pavilion, 9th

between the 2 groups after IPTW adjustment. This is

Floor, 3400 Spruce Street, Philadelphia, Pennsylva-

an issue that may be particularly relevant to studies

nia 19104. E-mail: [email protected].

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

TR Versus TF Access for Rescue PCI

PERSPECTIVES WHAT IS KNOWN? Although radial access has been

no difference in in-hospital mortality (OR: 0.93; 95% CI:

shown to reduce bleeding and mortality in STEMI, little is

0.61 to 1.44; p ¼ 0.75). Gastrointestinal bleeding was

known about the role of radial access in rescue PCI.

also less frequent in the radial group (OR: 0.23; 95% CI:

WHAT IS NEW? Using the NCDR CathPCI Registry, we

the effect of unmeasured confounders on this analysis.

0.05 to 0.98; p ¼ 0.05) making it impossible to rule out examined 9,494 STEMI patients undergoing rescue PCI and found that 14.3% of procedures were performed with

WHAT IS NEXT? Given the lack of research regarding

radial access. IPTW-adjusted analysis demonstrated a

bleeding avoidance in rescue PCI, the present study is

significantly lower risk of bleeding in the radial access

likely to represent the best available data in this area for

group (OR: 0.67; 95% CI: 0.52 to 0.87; p ¼ 0.003), but

the foreseeable future.

REFERENCES 1. Roe MT, Messenger JC, Weintraub WS, et al. Treatments, trends, and outcomes of acute myocardial infarction and percutaneous coronary intervention. J Am Coll Cardiol 2010;56:254–63. 2. Kadakia MB, Desai NR, Alexander KP, et al. Use of anticoagulant agents and risk of bleeding among patients admitted with myocardial infarction: a report from the NCDR ACTION Registry— GWTG (National Cardiovascular Data Registry Acute Coronary Treatment and Intervention Outcomes Network Registry—Get With the Guidelines). J Am Coll Cardiol Intv 2010;3:1166–77. 3. The GUSTO Angiographic Investigators. The effects of tissue plasminogen activator, streptokinase, or both on coronary-artery patency, ventricular function, and survival after acute myocardial infarction. N Engl J Med 1993;329: 1615–22. 4. Burjonroppa SC, Varosy PD, Rao SV, et al. Survival of patients undergoing rescue percutaneous coronary intervention: development and validation of a predictive tool. J Am Coll Cardiol Intv 2011;4:42–50. 5. Manoukian SV, Feit F, Mehran R, et al. Impact of major bleeding on 30-day mortality and clinical outcomes in patients with acute coronary syndromes: an analysis from the ACUITY trial. J Am Coll Cardiol 2007;49:1362–8. 6. Gershlick AH, Stephens-Lloyd A, Hughes S, et al. Rescue angioplasty after failed thrombolytic therapy for acute myocardial infarction. N Engl J Med 2005;353:2758–68. 7. Berkowitz SD, Granger CB, Pieper KS, et al., for the Global Utilization of Streptokinase and Tissue Plasminogen activator for Occluded coronary arteries (GUSTO) I Investigators. Incidence and predictors of bleeding after contemporary thrombolytic therapy for myocardial infarction. Circulation 1997;95:2508–16.

Registry (2007 to 2011). J Am Coll Cardiol 2013; 61:420–6. 9. Cantor WJ, Puley G, Natarajan MK, et al. Radial versus femoral access for emergent percutaneous coronary intervention with adjunct glycoprotein IIb/IIIa inhibition in acute myocardial infarction— the RADIAL-AMI pilot randomized trial. Am Heart J 2005;150:543–9. 10. Hamon M, Rasmussen LH, Manoukian SV, et al. Choice of arterial access site and outcomes in patients with acute coronary syndromes managed with an early invasive strategy: the ACUITY trial. EuroIntervention 2009;5:115–20.

16. Moussa I, Hermann A, Messenger JC, et al. The NCDR CathPCI Registry: a US national perspective on care and outcomes for percutaneous coronary intervention. Heart 2013;99: 297–303. 17. National Cardiovascular Data Registry. Data collection. Available at: http://www.ncdr.com/ webncdr/cathpci/home/datacollection. Accessed October 10, 2015. 18. Brookhart MA, Schneeweiss S, Rothman KJ, Glynn RJ, Avorn J, Sturmer T. Variable selection for propensity score models. Am J Epidemiol 2006;163:1149–56.

11. Jolly SS, Yusuf S, Cairns J, et al. Radial versus femoral access for coronary angiography and intervention in patients with acute coronary syn-

19. Wimmer NJ, Resnic FS, Mauri L, Matheny ME, Yeh RW. Comparison of transradial versus transfemoral percutaneous coronary intervention in

dromes (RIVAL): a randomised, parallel group, multicentre trial. Lancet 2011;377:1409–20.

routine practice: evidence for the importance of “falsification hypotheses” in observational studies of comparative effectiveness. J Am Coll Cardiol 2013;62:2147–8.

12. Mehta SR, Jolly SS, Cairns J, et al. Effects of radial versus femoral artery access in patients with acute coronary syndromes with or without ST-segment elevation. J Am Coll Cardiol 2012;60: 2490–9. 13. 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. 14. Valgimigli M, Saia F, Guastaroba P, et al. Transradial versus transfemoral intervention for acute myocardial infarction: a propensity scoreadjusted and -matched analysis from the REAL (REgistro regionale AngiopLastiche dell’EmiliaRomagna) multicenter registry. J Am Coll Cardiol Intv 2012;5:23–35.

20. Rao SV, McCoy LA, Spertus JA, et al. An updated bleeding model to predict the risk of 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. 21. Kassam S, Cantor WJ, Patel D, et al. Radial versus femoral access for rescue percutaneous coronary intervention with adjuvant glycoprotein IIb/IIIa inhibitor use. Can J Cardiol 2004;20: 1439–42. 22. Cruden NL, Teh CH, Starkey IR, Newby DE. Reduced vascular complications and length of stay with transradial rescue angioplasty for acute myocardial infarction. Catheter Cardiovasc Interv 2007;70:670–5.

8. Baklanov DV, Kaltenbach LA, Marso SP, et al. The prevalence and outcomes of transradial percutaneous coronary intervention for ST-

15. Dehmer GJ, Weaver D, Roe MT, et al. A contemporary view of diagnostic cardiac catheterization and percutaneous coronary intervention in the United States: a report from the CathPCI Registry of the National Cardiovascular Data

23. Alexander KP, Newby LK, Armstrong PW, et al. Acute coronary care in the elderly, part II: ST-segment-elevation myocardial infarction: a scientific statement for healthcare professionals

segment elevation myocardial infarction: analysis from the National Cardiovascular Data

Registry, 2010 through June 2011. J Am Coll Cardiol 2012;60:2017–31.

from the American Heart Association Council on Clinical Cardiology: in collaboration with the

1875

1876

Kadakia et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 8, NO. 14, 2015 DECEMBER 21, 2015:1868–76

TR Versus TF Access for Rescue PCI

Society of Geriatric Cardiology. Circulation 2007; 115:2570–89. 24. de Boer MJ, Ottervanger JP, van ’t Hof AW, et al. Reperfusion therapy in elderly patients with acute myocardial infarction: a randomized comparison of primary angioplasty and thrombolytic therapy. J Am Coll Cardiol 2002;39:1723–8. 25. Gurwitz JH, Col NF, Avorn J. The exclusion of the elderly and women from clinical trials in acute myocardial infarction. JAMA 1992;268: 1417–22. 26. Ndrepepa G, Berger PB, Mehilli J, et al. Periprocedural bleeding and 1-year outcome after percutaneous coronary interventions: appropriateness of including bleeding as a component of a quadruple end point. J Am Coll Cardiol 2008;51: 690–7.

27. Wimmer NJ, Resnic FS, Mauri L, et al. Risktreatment paradox in the selection of transradial access for percutaneous coronary intervention. J Am Heart Assoc 2013;2:e000174. 28. Applegate RJ. Radial access for primary percutaneous coronary intervention for ST-segment elevation myocardial infarction: time for a paradigm shift? J Am Coll Cardiol 2014;63:973–5. 29. Bertrand OF, Patel T. Radial approach for primary percutaneous coronary intervention: ready for prime time? J Am Coll Cardiol 2012;60: 2500–3. 30. Bernat I, Horak D, Stasek J, et al. ST-segment elevation myocardial infarction treated by radial or femoral approach in a multicenter randomized clinical trial: the STEMI-RADIAL Trial. J Am Coll Cardiol 2014;63:964–72.

31. Stone GW, Witzenbichler B, Guagliumi G, et al. Bivalirudin during primary PCI in acute myocardial infarction. N Engl J Med 2008;358:2218–30. 32. 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. 33. Lange HW, von Boetticher H. Randomized comparison of operator radiation exposure during coronary angiography and intervention by radial or femoral approach. Catheter Cardiovasc Interv 2006;67:12–6.

KEY WORDS rescue PCI, STEMI, transradial access