EXCEL Data Highlight Bleeding Patterns After CABG, PCI for LM Disease

In patients undergoing revascularization for left main coronary artery disease, CABG is associated with higher rates of all bleeds and in-hospital major bleeding, while PCI is linked to more postdischarge major bleeding through 5 years, according to new data from the EXCEL trial.

The findings can help inform treatment choices, but also point to a path forward for improving outcomes regardless of which revascularization method is used, investigators say.

Over the last decade and a half, researchers have “really identified bleeding to be at least as important a prognostic adverse event as recurrent ischemia after PCI in most settings,” senior author Gregg W. Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), told TCTMD. “We’ve learned a lot about this topic over the years, but nobody’s ever looked at the frequency and impact of major bleeding after left main intervention[s], which of course are some of the highest-risk patients that we do in terms of the amount of myocardium at risk.”

Unsurprisingly, major bleeding in EXCEL was a “very strong independent predictor” of both all-cause and cardiovascular mortality through 5 years, he said, though it’s possible that this link is stronger in this cohort of patients with left main disease than it would be in lower-risk populations.

As to when bleeding occurred in patients receiving surgical versus percutaneous revascularization, it likely comes down to the prolonged use of dual antiplatelet therapy (DAPT) in the latter, according to Stone. “We likely can improve outcomes by shortening the duration of dual antiplatelet therapy, particularly in high-risk bleeding patients but in all patients in whom left main stenting is successful, particularly if guided by intravascular imaging, so you know you have a very good result,” he said.

Commenting on the study for TCTMD, Mario Gaudino, MD, PhD (Weill Cornell Medicine, New York, NY), cautioned that the findings should not be used to pit the different procedures against each other, but rather to confirm “what we know very well: that CABG has a higher procedural risk and lower long-term risk than PCI and vice versa.”

The data are “a snapshot in time for a rapidly evolving field,” he added. “I don’t want to hear PCI versus CABG anymore. I want to hear PCI and CABG, because quite frankly, they’re different interventions, they have different mechanisms, and they have different early versus late risks and benefits. In most patients, one or the other is clearly indicated, and there are a minority of patients where both are indicated and those are the patients generally included in clinical trials.”

That said, Gaudino observed, the study highlights opportunities for improvements for both PCI and CABG in reducing bleeding events.

EXCEL Findings

For the study, published in the December 10, 2024, issue of the Journal of the American College of Cardiology, Gennaro Giustino, MD (Gagnon Cardiovascular Institute, Atlantic Health System, Morristown, NJ), Stone, and colleagues included all 1,905 patients from the original EXCEL trial with unprotected left main CAD who were randomized to undergo PCI (n = 948) or CABG (n = 957) and followed through 5 years.

In total, 11.4% of patients reported at least one major bleeding event, defined as TIMI major or minor bleeding, BARC types 3-5 bleeding, or any overt bleeding requiring a blood transfusion. The median time to the first major bleeding event overall was 7 days—shorter after CABG than PCI (6 vs 11 days; P = 0.048).

I don’t want to hear PCI versus CABG anymore. I want to hear PCI and CABG, because quite frankly, they’re different interventions. Mario Gaudino

Overall, major bleeding events were less common in those who received PCI compared with CABG (7.9% vs 14.8%; OR 0.48; 95% CI 0.36-0.65). When broken down further, in-hospital major bleeds remained less common following PCI (3.8% vs 13.5%; OR 0.25; 95% CI 0.17-0.37) but were more common than CABG postdischarge (4.5% vs 2.0%; OR 2.33; 95% CI 1.33-3.09; P for interaction < 0.0001).

Notably, all 41 PCI patients who reported major bleeding events were on DAPT, while only half of the 18 CABG patients who reported major bleeding were on DAPT.

Major bleeding was associated with an elevated risk of both all-cause (adjusted HR 2.71; 95% CI 1.95-3.77) and cardiovascular mortality (adjusted HR 2.95; 95% CI 1.87-4.63) regardless of whether the bleeds occurred in the hospital or after discharge.

Changing DAPT Perceptions

Patients undergoing PCI with left main disease will need to be properly evaluated for their relative risks of both bleeding and ischemia when individualizing their postprocedural plan, Stone said. While recent studies have indicated that shorter-term DAPT is better for these patients and many clinicians have come to better appreciate the way bleeding can affect prognosis, challenges remain.

“A major issue is that in the United States we tend to start people on dual antiplatelet therapy, and as long as they’re not bleeding, we just keep them on,” Stone said. “But of course, nobody has a major bleed until they have a major bleed, right? So, if you no longer need the medication which is constantly exposing the patient to the risk of bleeding, then you should withdraw it when it is no longer necessary.”

Particularly elderly patients, those with renal insufficiency or anemia, or those taking oral anticoagulation “should be on a real limited duration of DAPT,” he said, specifying 3 months or even 1 month in those at higher bleeding risk. “For most patients, especially if you have a successful left main PCI guided by intravascular imaging, the rate of stent thrombosis around the left main coronary artery is very low.”

Nobody has a major bleed until they have a major bleed, right? Gregg W. Stone

Stone said he’d like to see more registry studies that demonstrate the safety of early discontinuation of DAPT after left main PCI. “I think people are still concerned that it’s not safe to do, even though all of these randomized trials do suggest that it is safe,” he said. “Large, real-world studies in tens of thousands of patients would be confirmatory.”

And even though in-hospital major bleeding is rare post-PCI, at about 4%, this rate still is “not trivial,” according to Stone. Increased use of radial artery interventions can potentially reduce these events following left main PCI, he said, adding that “in EXCEL, most of the patients still had femoral artery intervention.”

CABG Solutions

In the surgical space, Gaudino said the field has placed a heavy emphasis on prevention of bleeding and transfusions in recent years, using technologies like intraoperative autologous blood donation and cell saver devices to reduce the risk of bleeds after CABG. Additionally, he said, the new generation of miniaturized cardiopulmonary bypass circuits will continue to help in this regard.

Ongoing research looking at optimal antiplatelet regimens after CABG, specifically the ODIN trial, should also provide more information to help surgeons tailor these medications in their patients, he said.

Stone suggested that further efforts to prevent postsurgical atrial fibrillation, which often requires chronic oral anticoagulation, would also contribute to lowering the bleeding risk in CABG patients.

“We can do better, and we will do better,” Gaudino observed. “But again, the general concept remains part of the moving the target. There is no way CABG will ever achieve lower periprocedural risk than PCI, and there is no way PCI can achieve the more durable result of CABG in the long term.”