Do No Harm: Post-PCI Bleeds
Overall, complication rates related to PCI procedures have been drastically reduced, but bleeding complications remain the last barrier to safety. However, ongoing research has shed light on how some techniques and drugs may or may not be effective in reducing bleeding.

Big improvements—far enough?

“In the last 10 years, the safety of PCI has vastly improved, and in particular, the urgent revascularization rate has lowered to 0.1 percent—even less in many registries—due to improved strategies and technologies,” says Steven P. Marso, MD, an interventional cardiologist at Mid America Heart Institute at St. Luke’s Hospital in Kansas City, Mo. Likewise, the current U.S. mortality risk following PCI is 0.4 to 0.5 percent, and the restenosis rate is approximately 6 percent at six to nine months. Therefore, bleeding, with rates as high as 9 percent, is the last common and costly morbid complication that needs to be addressed in the PCI population.

One problem with assessing post-PCI bleeds is the lack of a unified definition of bleeding, according to Sunil V. Rao, MD, director of the cardiac cath lab at the Durham VA Medical Center in Durham, N.C. “The definition that a provider utilizes will clearly influence reported rates,” he adds. This was confirmed by Steinhubl et al in a meta-analysis of 13 trials evaluating antithrombotic drugs in more than 178,000 patients with acute coronary syndrome (ACS). The researchers noted that “it is undoubtedly true” that variations in major bleeding definitions have led to differences in reported rates (Am Heart J 2007;154:3-11).  
“While some studies have used either the GUSTO or TIMI definition, others have used both, and yet others have combined selected elements of both scales,” wrote Vavelle and Rao (Interv Cardiol 2009;1[1]:51-62).

However, there are certain elements of consensus for defining a bleed, including: overt bleeding, bleeding that requires a transfusion and bleeding severe enough to cause the hemoglobin to decrease by 3 grams. “If you combine these three data elements and examine the large registry data, the bleeding risk associated with PCI—whether for STEMI, non-STEMI or elective PCI—tends to be in the 7 to 9 percent range,” Rao says. Also, bleeding rates tend to be higher in women.

While bleeding rates have gradually decreased after PCI due to smaller catheters and targeted anticoagulant therapies, higher risk patients, such as those with STEMI or non-ST-elevation ACS, have experienced no change in bleeding complication rates—maintaining at about 10 percent, Rao says.

However, preventive measures can and should be employed whenever possible. “Even though experienced operators typically are better at preventing and managing complications, practices can always reduce their complication rates further,” says Gregory J. Mishkel, MD, an interventional cardiologist at Prairie Cardiovascular in Springfield, Ill. “Prevention is always preferable to treatment. Before PCI is undertaken, it needs to be assessed whether the risk of the procedure is worth the potential outcome.”

What’s in a bleed?

Among the more than one million PCIs performed annually in U.S. hospitals, bleeding occurs with wide variability and has wide-ranging effects on costs and healthcare resources. Major bleeding events result in an average four- to six-day increase in length of stay, and an average increase of hospital costs by $6,000 to $8,000 (Am J Cardiol 2003;92[8]:930–935). However, certain facilities incur more expenses. For instance, every bleeding event costs the St. Luke’s Health System between $11,000 and $13,000.

Rao points to specific increased resource utilization with bleeds, including consultations, vascular surgery and increased imaging. “It’s a stepwise association, so the worse the bleeding, the more resources are utilized,” he says.

Also, research, such as the ACUITY trial, has shown that major bleeding complications are associated with short-term and long-term morbidity and mortality, which includes recurrent MI, stent thrombosis, stroke and death (J Am Coll Cardiol 2007;49:1362-1368). However, the majority of bleeds have not been found to be causal.  

Despite a heightened focus on post-PCI bleeding, very few U.S. practices assess an individual’s risk for bleeding pre-procedurally. At St. Luke’s, however, caregivers recently began to systematically quantify every individual’s risk for bleeding. The software-driven calculated data are presented to the interventional cardiologist and incorporated into the informed consent process, so the patient also is aware of the risks (see page 27). “Equipped with this information, providers are empowered to offer therapies that might mitigate the patient’s risk of bleeding, even though the process doesn’t instruct physicians on which drug or device to use,” Marso says.

Treatment strategies

“Of late, reducing bleeding has become a clinical priority because we now have multiple therapies available,” Rao says. However, practitioners have not found a unified strategy to mitigate bleeding risk, so ongoing studies are attempting to evaluate the various methods.

For example, Marso et el analyzed data from 1.52 million patients undergoing PCI procedures performed at 955 U.S. hospitals participating in the American College of Cardiology’s (ACC) National Cardiovascular Data Registry (NCDR) CathPCI Registry from Jan. 1, 2004, to Sept. 30, 2008. They examined the use of the antithrombin drug bivalirudin (Angiomax, The Medicines Company) and vascular closure devices (VCDs) to assess the association with lower rates of peri-procedural PCI bleeds. Researchers found that bleeding occurred in 2 percent of the overall patient population.

“Bivalirudin was associated with less bleeding across all patients,” Marso explains, but it was primarily effective in those patients with a bleeding risk of greater than 1 percent—an intermediate risk population.   

The use of bivalirudin could even be cost-savings. “When bivalirudin is administered to high-risk patients, we save the health system approximately $1,500 per patient. Even though drug acquisition cost is $600, the bleeding frequency rate is so high in these patients that overall costs are reduced,” Marso says.

Other new pharmacological approaches also have emerged. For example, in the OASIS-5 follow-up trial evaluating patients with ACS, fondaparinux (Arixta, GlaxoSmithKline) was associated with a 40 percent reduction in major bleeding risk, compared with enoxaparin (Lovenox, Sanofi-Aventis) in those treated with GP IIb/IIIa inhibitors (J Am Coll Cardiol 2009;54:468-476).

While data surrounding VCDs and bleeding risk have not always been clear, Marso et al found that these devices were associated with a lower risk of bleeding, particularly in high-risk patients. Not everyone is so convinced that the VCD question is entirely answered.

“All the observational data have shown that VCDs are associated with a lower bleeding risk, but almost every randomized trial has shown that they are associated with worse bleeding risk. The issue is selection bias inherent in registry data; therefore, we may need to take observational data with a grain of salt,” says Rao, who was part of the Marso et al study. “Also, it’s an expensive strategy to use both a VCD and bivalirudin, as it adds between $200 and $400 to the cost of every procedure. It ultimately becomes a question of how much the facility wants to sacrifice—both financially and with clinical outcomes—and how much benefit is derived.”

The latest guidelines do not recognize VCDs as a strategy that reduces bleeding. However, Marso suggests that the “time has come” for a randomized trial to establish whether their use reduces bleeding and if so, give them a Class 1 indication.

Conversely, Rao points to the transradial access site as an alternative method to reduce bleeding, which “doesn’t add any cost to the procedure.”  

Therefore, it is evident that more research still needs to be undertaken to better define bleeding to result in better reported outcomes, as well as to more fully understand the factors that increase a patient’s risk for bleeding and how to mitigate that.