HORIZONS-AMI Substudy: Anticoagulation After Primary PCI Yields Worse Outcomes

Download this article's Factoid (PDF & PPT for Gold Subscribers)

Use of anticoagulants for specific indications such as acute atrial fibrillation (A-fib) and left ventricular thrombus following primary percutaneous coronary intervention (PCI) increases both ischemic and bleeding events over the first 30 days and longer term. The findings, from a substudy of the HORIZONS-AMI trial, were published online August 13, 2014, ahead of print in the American Journal of Cardiology

However, these outcomes may stem as much from the patients’ high-risk clinical profiles as from the anticoagulation, suggested Christopher Granger, MD, of Duke University (Durham, NC), in a telephone interview with TCTMD. 

In the main trial, more than 3,000 STEMI patients undergoing primary PCI were randomized first to unfractionated heparin plus a GPI or bivalirudin alone then to Taxus paclitaxel-eluting stents (Boston Scientific; Natick, MA) or BMS with the same platform.

Methods
For the current analysis, investigators led by Gregg W. Stone, MD, of Columbia University Medical Center (New York, NY), compared outcomes between patients who did (n = 410; 16.6%) vs did not (n = 2,063; 83.4%) receive anticoagulants for specific indications (as opposed to routine prophylaxis), including:
  • Suboptimal PCI (3.2%)
  • Intra-aortic balloon pump (2.8%)
  • A-fib (2.7%)
  • Post-index procedure revascularization (2.4%)
  • Cardiomyopathy (1.8%)
  • LV, mural, or giant coronary thrombus (1.3%)
Prescribed anticoagulants were unfractionated heparin (n = 181), low-molecular weight heparin (n = 192, of whom 162 received enoxaparin), bivalirudin (n = 138), and fondaparinux (n = 2). Some patients received more than 1 of these drugs at different times.
Patients who received anticoagulants were more likely to have had intervention attempted on the LAD and more often had baseline and post-PCI TIMI 0/1 flow. The number of stents implanted and rates of multivessel treatment were also higher in the anticoagulation group. Importantly, these patients were also less likely to receive dual antiplatelet therapy but more likely to have been given warfarin at discharge (both P < .0001).

 

After propensity-adjusted multivariable analysis, the anticoagulated group showed higher rates of cardiac death, reinfarction, ARC-defined definite/probable stent thrombosis, and MACE (death, reinfarction, ischemia-driven TVR, or stroke) compared with the nonanticoagulated group. They also experienced more non-CABG major bleeding, leading to a higher incidence of net adverse clinical events (NACE, composite of MACE and non-CABG major bleeding; table 1). Similar findings were observed from 30 days to 3 years.

Table 1. Propensity-Adjusted 30-Day Outcomes

 

Anticoagulant

No Anticoagulant

P Value

Cardiac Death

6.1%

1.9%

.04

Reinfarction

4.7%

1.6%

.001

Definite/Probable Stent Thrombosis

6.1%

1.9%

.0002

Non-CABG Major Bleeding

17.3%

6.0%

< .0001

MACE

13.9%

4.4%

< .0001

NACE

26.1%

9.2%

< .0001

Abbreviation: NACE, net adverse clinical events.

In addition, in-hospital acquired thrombocytopenia developed more frequently in patients who received post-PCI anticoagulants than those who did not (19.3% vs 10.8%; P < .0001).

Moreover, 30-day outcomes were worse for anticoagulated patients regardless of stent assignment but similar between those treated for guideline-recommended indications (A-fib; intra-aortic balloon pump; or LV, mural, and giant coronary thrombus; n = 164) vs non-guideline-recommended indications (eg, failed PCI, ventricular aneurysm, and cardiomyopathy; n = 246).

At 30 days patients who received unfractionated heparin (n = 181) had higher rates of adverse events than those given other agents, although the differences disappeared over the longer term. In addition, in the anticoagulated cohort, women fared worse than men as evidenced by higher rates of MACE (P = .045) and NACE (P = .001) at 30 days and increased non-CABG major bleeding (P = .01) and stroke (P = .047) between 30 days and 3 years.

Stopping Antiplatelet Therapy Increases Risk

Overall, patients not on dual antiplatelet therapy at discharge had more adverse events within the first 30 days than did those who remained on the therapy; higher rates of all-cause death and cardiac death (both P = .02) persisted between 30 days and 3 years.

Within the first 30 days after PCI, 10 patients (0.4%) developed thromboembolic events (4 cases of DVT and 6 of pulmonary embolism). Four of these patients had received anticoagulation for specific indications, although the therapy had been stopped in 2 prior to their events. 

“These data question whether postprocedural [anticoagulation] in this setting is beneficial, or in fact, may be harmful (eg, by causing bleeding or thrombocytopenia),” the authors write. “In this regard, [anticoagulation]-induced major bleeding (the majority of which occurred within the first 5 days post-procedure) may have led to early discontinuation of antiplatelet therapy, which … has been strongly linked to ischemic events and death after primary PCI.”

Noting the large discrepancy in outcomes between men and women, Dr. Stone and colleagues suggest, “Further studies are needed to identify the specific patient subgroups in whom postprocedural [anticoagulation] provides clinical benefits warranting the potential risks.” 

Confounding Clouds Findings

“The biggest challenge with this [study] is the inability to fully adjust for confoundingfactors,” Dr. Granger commented. “The reason these patients were anticoagulated was because they had features that placed them at higher risk. Ideally, one would be able to adjust for these features, but we know that the adjustment models are often imperfect and incomplete.”

He noted that unfractionated heparin—widely used in these settings due mainly to lower cost and familiarity—is problematic because it is implicated in causing thrombocytopenia more than other anticoagulants. And thrombocytopenia is associated with increased risk of both bleeding and thrombosis.

Noting that achievement of optimal coagulation with unfractionated heparin is not always possible, Eugenia Nikolsky, MD, PhD, of Rambam Medical Center (Haifa, Israel), said that theoretically bivalirudin may have an advantage, although the issue should be addressed in a prospective trial.

There may be a selective role for the novel oral anticoagulants, such as in patients with acute DVT or A-fib, Dr. Granger said. Both he and Dr. Nikolsky pointed to ongoing randomized trials that may provide further guidance on the best combination of oral anticoagulants and antiplatelet agents for A-fib patients following PCI. Dr. Nikolsky added that one option might be omitting aspirin from a triple-therapy regimen, but the safety of this approach is unclear.

Based on these data, Dr. Granger found no reason to question the current guidelines accepting anticoagulant use in certain circumstances after primary PCI. Instead, he said, they “highlight that this population is at high risk for bleeding and thrombosis and we should be paying more attention to them, both to ensure that we’re very careful in how we use anticoagulants and to generate more evidence about how to best treat them.

Note: Dr. Stone and several coauthors are faculty members of the Cardiovascular Research Foundation, which owns and operates TCTMD.

  


Source: 
Madhavan MV, Géneréux P, Kirtane AJ, et al. Post-procedural anticoagulation for specific therapeutic indications after revascularization for ST-segment elevation myocardial infarction (from the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction [HORIZONS-AMI] trial). Am J Cardiol. 2014;Epub ahead of print.

Related Stories:

HORIZONS-AMI Substudy: Anticoagulation After Primary PCI Yields Worse Outcomes

Disclosures
  • Drs. Stone and Nikolsky report no relevant conflicts of interest.
  • Dr. Granger reports receiving research funding and consulting fees from The Medicines Company.

Comments