PLATO Subanalysis: Major Bleeding Linked With Increased Short-term Mortality in ACS

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

While major bleeding is consistently associated with subsequent mortality in patients with acute coronary syndromes (ACS), this connection is strongest in the first 30 days and for spontaneous bleeding, according to a substudy of the PLATO trial published online September 26, 2014, ahead of print in EuroIntervention.

Spontaneous Bleeding Significant

After a mean follow-up of 272.2 days, 11.8% (n = 2,189) of patients had at least 1 major bleeding event, per PLATO criteria. Of these, 6.8%, 0.9%, and 0.3% of patients had 2, 3, or at least 4 bleeding events, respectively. The median time from randomization to first major bleed was 12.8 days.

More patients with than without major bleeding died over the follow-up period (10.8% vs 4.1%; P < .001). Unadjusted risks of short-term and long-term mortality increased in conjunction with major bleeding. After accounting for baseline demographics, medical history, medications, laboratory values, and disease characteristics, major bleeding was still associated with more than a 9-fold increased risk of short-term mortality (adjusted HR 9.28; 95% CI 7.50-11.48), but the association with long-term mortality did not persist (adjusted HR 1.28; 95% CI 0.93-1.75).

Among all patients with major bleeding, 22.2% of bleeding events were spontaneous while the rest were procedure-related (1,503 related to CABG, 236 to PCI or angiography, and 86 to noncoronary procedures). Spontaneous bleeding events led to death more often than did procedure-related bleeding over the follow-up period (25.2% vs 7.7%). Additionally, spontaneous bleeding was associated with both short- and long-term mortality, while procedure-related bleeding was only associated with the former (table 1).

Table 1. PLATO Major Bleeding, Mortality

Procedure-related major bleeding deaths were further broken down into those relating to CABG (6.7%), PCI or angiography (10.6%), and noncoronary procedures (25.6%). Only CABG- and noncoronary-procedure-related bleeding were associated with increased short-term mortality (P < .001 for both), but none of these categories was associated with higher long-term mortality.

Analyses using GUSTO and TIMI bleeding criteria for major bleeding produced similar results.

Minor bleeding was observed in 879 patients (4.7%), of whom 158 also reported a major bleed. Minor bleeding also increased the risk of short-term mortality (adjusted HR 2.12; 95% CI 1.37-3.28) but the association was less marked than for major bleeding (P < .001).

Lastly, the link between major bleeding and mortality was similar in patients treated with clopidogrel and ticagrelor for both short- and long-term mortality (P for interaction = .81 and .22, respectively).

Type of Bleeding Deserves Notice

Dr. Steg and colleagues write that though the “detrimental effect of bleeding on short-term outcomes has already been described in patients with ACS,” the effect of major bleeding on long-term mortality is “much more controversial.”

In a telephone interview with TCTMD, Neal S. Kleiman, MD, of Methodist DeBakey Heart and Vascular Center (Houston, TX), commented, “[J]ust about everybody has looked at major bleeding and outcomes and said that 1-year mortality is worse if [patients] have major bleeding. The question is if [the death is] related to the bleeding or whether the bleeding was just an indicator of [a sicker patient].”

Additionally, 9 months is not quite “long-term” with regard to outcomes, but the message would be “likely the same” with longer follow-up, he noted.

Tracy Y. Wang, MD, MHS, MSc, of Duke University Medical Center (Durham, NC), told TCTMD in an email that “bleeding type or location has prognostic significance on top of severity. We did a study that showed that if bleeding was severe, then an association with long-term outcomes is consistently observed. But if bleeding severity was not as high, then type makes a difference.”

She agreed that spontaneous bleeding events “may be associated with worse long-term outcomes than procedural ones in part because these are sicker patients who are prone to these types of bleeds. Their bleeding may also not be as well-controlled, which reduces the likelihood of being resumed on appropriate antiplatelet therapies.”

The authors suggest that bleeding type should be considered in the future when revising bleeding classifications.

“You need to know what the effect of the drug is on the risk of bleeding beyond the site of instrumentation and beyond the CABG period,” Dr. Kleiman concluded. “I think those are very different considerations when you are deciding whether to treat someone and how to manage them.”

Source:
Ducrocq G, Schulte PJ, Becker RC, et al. Association of spontaneous and procedure-related bleeds with short- and long-term mortality after acute coronary syndromes: an analysis from the PLATO trial. EuroIntervention. 2014;Epub ahead of print.

Disclosures:

• Dr. Steg reports receiving research grants from Sanofi and Servier; serving as a speaker or consultant for numerous pharmaceutical companies; and holding stock in Aterovax.
• Dr. Wang reports receiving institutional research support from AstraZeneca, Daiichi Sankyo, Eli Lilly, Gilead, and Glaxo Smith Kline and honorarium from AstraZeneca.
• Dr. Kleiman reports no relevant conflicts of interest.

Related Stories

• Bleeding Common but Often Unreported in ACS Patients on Antiplatelet Therapy
• PLATO Substudy: Ticagrelor Reduces Mortality in NSTE-ACS Patients Regardless of Revascularization
• Predictors of Serious Bleeding in ACS Patients from TRITON-TIMI Identified