CardioExchange Archive

CardioExchange’s Harlan M. Krumholz and John Ryan interview Sunil V. Rao, lead author of SAFE-PCI for Women, a registry-based randomized trial of a radial versus a femoral approach to PCI. The study is published in JACC: Cardiovascular Interventions.

Krumholz and Ryan: Please describe your study and its major findings.

Rao: We conducted a randomized trial comparing a radial approach with a femoral approach in women undergoing cardiac catheterization (with the possibility of PCI) or scheduled PCI. Our rationale for the study was that women undergoing invasive procedures have a higher risk for bleeding than men, as well as a greater challenge in completing radial procedures, given smaller radial artery diameter and spasm. We had two primary endpoints: an “efficacy” endpoint (a composite of BARC* type 2, 3, or 5 bleeding or vascular complications requiring intervention) and a “feasibility” endpoint (accounting for access-site crossover). This was a “registry-based trial”: The sites were identified through the ACC NCDR CathPCI registry, and a large proportion of the data collection was through the registry.

The trial was stopped early because the event rate was much lower than expected. Our primary-analysis cohort consisted of women who underwent PCI, a subgroup of all women who were randomized. Interestingly, the primary efficacy endpoint showed no significant difference between radial and femoral access among women who underwent PCI; however, among all women who were randomized (including those who underwent just diagnostic coronary angiography), radial access was associated with a significant 60% lower incidence of the primary efficacy endpoint. In both cohorts, the rate of access-site crossover was significantly higher among women assigned to radial than to femoral access.

Krumholz and Ryan: How will this study affect your practice?

Rao: Our practice is already “radial first,” so this study bolsters our confidence in the safety of the radial approach in a high-risk population such as women. One can look at this trial from two perspectives: The purist would say that the trial is negative; in the primary-analysis cohort of women undergoing PCI, radial access did not reduce complications but did require operators to bail out to femoral access. The pragmatist would say that the PCI cohort is a subgroup of the overall randomized population and is underpowered to detect a difference. The directionality of the effect in the PCI cohort favors a radial approach, and in the larger sample size of all women who were randomized, the radial approach showed a significant reduction in bleeding and vascular complications, albeit at the higher rate of femoral crossover. The interaction term for the primary efficacy endpoint between patients who underwent diagnostic catheterization versus PCI was not significant, indicating a consistent effect across both cohorts.

Krumholz and Ryan: It is a “pragmatic trial” — please define that for our readers.

Rao: I think the meaning probably differs for different people. From our perspective, it was pragmatic because we had a system for site identification and data collection that increased the efficiency of the trial. Let me explain: Traditionally, when we identify study sites, a questionnaire is sent to sites’ principal investigators (PIs) asking questions about their practice, the number of cases they perform, the kinds of patients they see, and so on. This is highly subject to recall bias. For the SAFE PCI for Women trial, we needed sites that truly had expertise with a radial approach, so instead of relying on PI recall, we queried the CathPCI registry and identified sites that were performing radial procedures frequently. These sites formed our initial study sites. With respect to data collection, the study sites were already submitting procedure-related data to the CathPCI registry, so a large proportion of the registry data auto-populated the trial case-report form. This significantly reduced the workload for the site’s study coordinators. A clinical-events committee adjudicated all endpoint data.

Krumholz and Ryan: You built the trial to leverage the NCDR database. How did you do that, and how did it affect cost? What was the cost per subject, if you don’t mind sharing?

Rao: The software interface that connected the CathPCI registry with the case-report form, called the National Cardiovascular Research Infrastructure, was funded by the NHLBI. Bob Harrington MD, David Kong MD, and Brian McCourt led the effort at the Duke Clinical Research Institute, collaborating closely with the American College of Cardiology. We don’t have data on cost per subject, but the total budget for this trial — which randomized 1787 women, had 30-day follow-up, and included event adjudication — was $5 million.

JOIN THE DISCUSSION
How do the findings from Dr. Rao’s trial affect your perspective on radial access for PCI in women?

*BARC = Bleeding Academic Research Consortium