Complications of Peripheral Vascular Intervention: Time to Stop the Bleeding^∗Free Access

Introduction

Bleeding complications occur in 2% to 3% of patients following percutaneous coronary intervention (PCI) (1). Significant PCI-related bleeding is associated with increased mortality, morbidity, and health care costs (2,3). This has stimulated significant scrutiny from cardiovascular clinicians and other stakeholders, and as a result, deliberate refinements in interventional technique and periprocedural management have culminated in reduced bleeding and improved outcomes in recent years (4).

Peripheral vascular intervention (PVI) has been used with increasing frequency over the past decade to ameliorate symptoms of life-style-limiting claudication and preserve limb viability in the growing population of patients with peripheral artery disease (PAD). PVI has surpassed bypass surgery as the dominant mode of revascularization for lower extremity PAD, and contemporary practice has evolved such that PVI is increasingly being performed in nonhospital, outpatient-based laboratory settings (5,6). To keep step with the changing paradigms for PVI, it is important to identify factors contributing to patient- and procedure-specific risk, in order to tailor therapy and optimize outcomes. In this light, comprehensive understanding of PVI-related bleeding complications is imperative to assess procedural safety and to foster initiatives to improve quality of care. Until recently, however, sparse published data have been available to inform clinicians regarding bleeding risk associated with PVI and to direct best practices.

In this issue of JACC: Cardiovascular Interventions, Bhardwaj et al. (7) report the findings of a large observational analysis examining the prevalence and predictors of bleeding complications in patients undergoing PVI for lower extremity PAD. This analysis included 18,289 procedures from 2014 to 2016 at 76 facilities participating in the NCDR PVI (National Cardiovascular Data Registry Peripheral Vascular Intervention) registry. Bleeding was defined as a hemoglobin decrease of >3 g/dl with overt clinical bleeding, any hemoglobin decrease >4 g/dl, or transfusion for those with baseline hemoglobin >8 g/dl within 72 h of the procedure. Overall, bleeding occurred in 4.1% of the procedural cohort. There was nearly 5-fold variation in bleeding rates across participating institutions (1.2% to 5.9%). The majority of overt bleeding occurred at the site of vascular access (40.4%). Retroperitoneal bleeding, a potentially lethal complication, accounted for 22.6% of the overt bleeding events, representing an overall complication rate of 0.4%. Several characteristics predictive of bleeding were identified, including age, female sex, history of heart failure, critical limb ischemia procedures, nonfemoral access, and lack of vascular closure device use. Periprocedural bleeding was associated with increased in-hospital mortality after adjustment (odds ratio: 10.87; 95% confidence interval: 6.95 to 17.02), and the magnitude of the association correlated with the type of bleeding event, such that patients having clinically overt bleeding or requiring transfusions had the greatest risk.

Bhardwaj et al. (7) are to be commended for completing this thoughtful and timely analysis. Although other studies have assessed complication rates following PVI, these analyses have been limited primarily to access site–related bleeding (8). As the investigators suggest, the findings of this analysis represent a starting point from which modifiable bleeding risk factors during PVI may be targeted to reduce complication rates. Moreover, as is often the case with studies examining procedural outcomes, the investigators’ discovery that there is substantial variation in bleeding outcomes across participating NCDR PVI registry institutions indicates that there may be significant opportunity to establish and disseminate best treatment practices, in order to improve performance at those institutions found to have higher bleeding rates. By way of comparison, similar initiatives in the coronary interventional arena have been shown to reduce PCI-related bleeding complications significantly (9).

In this study, the reported bleeding rate of 4.1% following PVI is high relative to that of contemporary PCI practice for stable coronary artery disease (1) and may yet constitute an underestimate of the true bleeding risk because of inherent limitations of the dataset that was analyzed. Many PVI procedures are performed on an outpatient basis; as a result, any bleeding that does not occur during the immediate post-procedural period may not be identified and captured as a complication in this registry. Moreover, the definition of bleeding used in this analysis was quite stringent: a hemoglobin decrement >3 g/dl or the need for transfusion following PVI typically represents a very severe clinical bleeding event. Less severe bleeding complications such as the formation of an access site hematoma may result in a change in post-procedural management (e.g., the need for overnight hospital observation in a patient who would otherwise have been discharged) or perhaps the modification of periprocedural medical therapy (such as delaying the resumption of long-term systemic anticoagulant therapy or dual-antiplatelet therapy).

Such “minor” bleeding events, although not captured as bleeding complications in this analysis, may have an impact on long-term prognosis in a population of patients with a high prevalence of coronary and other cardiovascular disease.

Although the stringent methodology used to determine the rate of bleeding may spur some debate, it is evident that future strategies must be developed to reduce bleeding and improve outcomes of patients undergoing PVI. Because PVI procedures are often nonurgent, there may be time and opportunity to engage techniques, established as beneficial in other interventional therapies, to improve bleeding-related outcomes. For example, adopting the best transfemoral access technique, including the use of ultrasound guidance, may potentially reduce vascular access site–related bleeding (10) and may reduce the reported retroperitoneal hemorrhage rate of 0.4%, which is significant for a generally elective procedure. Tailoring pharmacology during PVI, such as the use of bivalirudin instead of heparin, may offer promise during PVI. Although not demonstrated to be of significant benefit in this study, likely in part because of to its overall low use (7), bivalirudin has previously been shown to reduce hemorrhagic complications in carotid stenting (11) and in lower extremity PVI (12) and may warrant evaluation in prospective, randomized controlled fashion. The location of vascular access may also provide opportunity to reduce peri-PVI bleeding. In the past 2 decades, the adoption of transradial access in PCI has dramatically reduced periprocedural bleeding rates and has improved outcomes (4). The development of dedicated peripheral technologies to facilitate transradial access for PVI, and the increased uptake of retrograde transpedal access, will likely confer similar benefits in reducing bleeding complications in PVI (13), particularly as these approaches may allow the avoidance of access sites such as the brachial and popliteal arteries, which are traditionally associated with higher rates of vascular complications (14). Because PVI is performed by specialists with varying skill sets from a spectrum of clinical disciplines, the development and dissemination of education regarding best treatment practices will have tremendous opportunity to improve outcome for patients undergoing PVI.

The NCDR PVI registry analysis by Bhardwaj et al. (7) demonstrates that bleeding is a significant complication of PVI and that bleeding confers considerable clinical risk. For clinicians engaged in the care of patients with lower extremity PAD, this represents an opportunity to establish best practices and improve patient outcomes: it is time to stop the bleeding.

Footnotes