Bivalirudin vs Heparin Patient-Level Meta-analysis Stirs an Old Debate

Taken together, could a round-up of the eight major trials carve out a role for bivalirudin in STEMI patients at this stage?

Bivalirudin vs Heparin Patient-Level Meta-analysis Stirs an Old Debate

A pooled analysis of individual patient data from the major bivalirudin versus unfractionated heparin trials concludes that preprocedural bivalirudin could be considered the agent of choice in the setting of STEMI, a finding that flies in the face of several studies that convinced operators to all but abandon—and guidelines to downgrade—its use over the last decade.

“We have spent the better part of the last 4 years collecting, cleaning, and analyzing the data,” for this meta-analysis, said Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), who presented the results during a late-breaking clinical science session at TCT Connect 2020 today and discussed them with the press yesterday. “There was less thrombocytopenia, there was less need for blood transfusions, there were substantial benefits in the STEMI patients with bivalirudin, and now that it's generic the cost is not very different [from heparin], so I don't see why it wouldn't be used.”

“It's a full cycle with bivalirudin,” said Ajay Kirtane, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), who chaired yesterday’s press conference. “It was an orphan drug, then taken up by a company, trials came out, then people sort of cast it away. Now it's generic, and here we have this [new] data. So I think it's a pretty interesting synthesis, certainly because we saw so many of these presentations of bivalirudin data at TCTs in the past.”

But not everyone is persuaded. Adnan Kastrati, MD (Deutsches Herzzentrum München, Munich, Germany), noted during the press conference that the pendulum has swung so far in favor of heparin, it’s hard to imagine these data could change current practice. “It might be too late for Europe,” he said. “I don't know the situation in the US, but to lead to a revival of bivalirudin in patients with STEMI, it might be too late.”

Bivalirudin initially gained popularity after early acute MI studies showed it reduced bleeding with no price tag in ischemic events when compared with heparin plus a glycoprotein IIb/IIIa inhibitor (GPI). It was knocked from that perch following trials like HEAT-PPCI and VALIDATE-SWEDEHEART, which showed that when GPI use was not mandated, and particularly when radial access was used, the agents are comparable; heparin won out as the cheaper strategy.

Speaking with TCTMD, Rod Stables, MD (Liverpool Heart and Chest Hospital, England), the principal investigator for HEAT-PPCI, said there is nothing in this new report to convince him to switch back to bivalirudin. “This meta-analysis represents a phenomenal effort, years in the making, hours and hours and hours of painstaking work, and the detail with which it has been performed is meticulous. I have nothing but admiration for the dedication and scientific rigor with which it’s been undertaken,” Stables said.

Nevertheless, the inclusion of “flawed” trials, which mandated GPI usage, are utterly out of step with the times, he continued. “The reason why I find it not a practice-changing meta-analysis is because it continues to be confounded by the same principal problem that has dogged this field from very early on, and that is that you cannot make a correct comparison of the relative safety and efficacy of heparin and bivalirudin if there is disproportionate GPI use in one arm.”

Patient-Level Data

Stone and colleagues collected patient-level data from all RCTs that enrolled at least 1,000 patients, ultimately including eight trials—ACUITY, HORIZONS-AMI, ISAR-REACT 4, EUROMAX, BRIGHT, HEAT-PPCI, MATRIX, and VALIDATE-SWEDEHEART. Patients were excluded if assigned to bivalirudin plus a GPI, did not have STEMI or NSTEMI, or were not treated with PCI, leaving a total of 27,409 patients. In all, roughly 55% of the cohort were STEMI patients and 45% NSTEMI, evenly matched in terms of study drug allocation.

As Stone showed in his presentation, overall outcomes for the entire cohort were no different for patients treated with bivalirudin versus heparin for either the primary efficacy endpoint, death at 30 days, while the primary safety endpoint of serious bleeding was reduced with bivalirudin (3.4% vs 5.7%; adjusted HR 0.60, 95% CI 0.52-0.68).

Broken down by ACS type, there was no difference in the rates of overall mortality or cardiac mortality by anticoagulant type for the NSTEMI patients, but the difference for both endpoints was significant among STEMI patients. In NSTEMI, the efficacy endpoints did not significantly differ, but the safety endpoint did.

Major Endpoints: STEMI and NSTEMI Patients




HR (95% CI)


All-Cause Death



0.80 (0.64-1.01)

Cardiac Death



0.72 (0.57-0.91)

Serious Bleeds



0.57 (0.47-0.68)


All-Cause Death



1.21 (0.84-1.73)

Cardiac Death



1.27 (0.87-1.87)

Serious Bleeds



0.63 (0.52-0.76)

Additional analyses make it clear that bivalirudin followed by a postprocedure infusion (that Stone noted lasted, on average, 2 hours) was associated with the best outcomes in terms of mortality reduction, with a high-dose infusion mitigating the risk of MI and stent thrombosis that tend to tick higher with bivalirudin versus heparin. Only 50% of the included trials studied post-procedure infusions.

Other analyses attempted to tease out the relative safety and efficacy of both agents with versus without postprocedure infusions with bivalirudin as well as the use of planned GPIs in the heparin-treated patients. Nearly half of the studies involved a planned GPI infusion in the heparin arm.

No Going Back?

For Stables, the planned versus non-planned GPI analyses fall short.

“What you need to understand is that the phrase ‘unplanned or non-planned GPI’ means nothing. It simply means that the protocol did not demand you gave GPI with your heparin,” Stables explained. “Over the various trials the tendency to use GPI with heparin varied enormously but in many trials it was more commonly used with heparin. By the time we got to SWEDEHEART, people had twigged that GPI use was unhelpful and the rates in the heparin and bivalirudin arms were equal.” That echoes what was done in HEAT-PPCI, he added. “The only way to make a fair comparison is to look exclusively at the trials which had, by design and—more importantly—in terms of treatment delivered, and equal approach to the use of GPI.”

This was the case for both VALIDATE-SWEDEHEART and HEAT-PPCI, and both trials showed no difference between the two anticoagulant regiments.

Asked by session chair Sunil Rao, MD (Duke Clinical Research Institute, Durham, NC), about the differential use of GPIs following his presentation at TCT Connect today, Stone acknowledged that this variability “is a limitation of all of these randomized trials.”

But, Stone continued, “when you pool all the data, when you consider ‘study’ as a random effect and then when you adjust for the differences in baseline characteristics, it shows you that compared to heparin and no routine use of GPI that bivalirudin with a post-PCI infusion substantially is associated with a reduction in mortality. And then if you use the [high dose] post-PCI that mitigates that increased risk of MI and stent thrombosis, while still preserving about a 40% reduction in serious bleeding.

Dominick Angiolillo, MD, PhD (University of Florida College of Medicine, Jacksonville), a panelist in the TCT Connect session, pointed out that it took years to do this meta-analysis “I was using the bivalirudin strategy with a full-dose infusion for a few hours afterwards—just like when you open a good bottle of wine, you gotta use it all,” he said. “But now I’ve switched to using unfractionated heparin and cangrelor [Kengreal; Chiesi], so what do I do?”

Stone replied that this is a valid alternative, but stressed that cangrelor wasn’t a part of any of the trials included in this analysis. “From this data, I can’t respond to that, but I can tell you that the effects we showed in terms of reducing mortality were independent of the potency of the P2Y12 inhibitor that was used,” he said, adding that the decision to use cangrelor would be independent of the decision to use bivalirudin or not.

Kirtane, speaking with TCTMD after the press conference, said he doesn’t think it’s too late for guidelines to change once again. “The trials do generally show a reduction in bleeding and in my opinion, if you can show a reduction in bleeding and there are no deleterious effects and the drug is relatively cheap, I would think about using it especially in patients at high risk for bleeding. . . . There’s no formula that every patient with STEMI should get X, Y, or Z. If I had a patient that was high risk for bleeding, I would probably start using [bivalirudin] again.” 

But Stables, asked if a high-bleeding-risk patient might be best managed with bivalirudin, pointed out that the current meta-analysis does not allow for a comparison of agents restricted to such patients. That said, he added, there are other “genuinely interesting” takeaways: “If for any reason you are using bivalirudin, for example in a patient who can’t receive heparin, these data provide a very persuasive argument for a post procedural infusion. I also think that this continues to provide evidence that GPI use needs to be highly restricted to certain selected cases, only because it certainly would appear that more widespread GPI use creates less favorable outcomes.”

Rao told TCTMD he believes this analysis suggests that bivalirudin in certain patients is a “viable option.” As radial access is more widely adopted, though, it will be harder to demonstrate a reduction in bleeding, he observed. “Interventionalists tend to be committed to their antithrombin choice so while these data are robust, it may not result in a real change in practice.”

Note: Stone is a faculty member of the Cardiovascular Research Foundation, the publisher of TCTMD

Shelley Wood is Managing Editor of TCTMD and the Editorial Director at CRF. She did her undergraduate degree at McGill…

Read Full Bio
  • Stone reports speaker honoraria from Cook; consulting to Valfix, TherOx, Robocath, HeartFlow, Gore, Ablative Solutions, Miracor, Neovasc, Abiomed, Ancora, Vectorious, Cardiomech; and equity/options from Ancora, Qool Therapeutics, Cagent, Applied Therapeutics, Biostar family of funds, SpectraWave, Orchestra Biomed, Aria, Cardiac Success, and Valfix.
  • Kirtane reports institutional grant support/research contracts from Abbott Vascular, Boston Scientific, CSI, Medtronic; Philips, ReCor Medical, and Siemens.
  • Stables has previously reported grants from The Medicines Company and AstraZeneca during the conduct of HEAT-PPCI.
  • Kastrati and Rao report having no relevant disclosures.
  • Angiolillo reports consulting fee/honoraria/speaker's bureau from Amgen, Arena, Aralez, AstraZeneca, Bayer, Biosensors, Boehringer Ingelheim, Bristol-Myers Squibb, Chiesi, Deerfield, Daiichi-Sankyo, Eli Lilly, Haemonetics, Idorsia, Janssen, Merck, PhaseBio, PLx Pharma, Pfizer, Sanofi, and The Medicines Company; and grant support/research contracts from Amgen, AstraZeneca, Bayer, Biosensors, CeloNova, CSL Behring, Daiichi-Sankyo, Eisai, Eli Lilly, Gilead, Idorsia, Janssen, Matsutani Chemical Industry Co., Merck, Novartis, Osprey Medical, and Renal Guard Solutions.