CardioExchange Archive

CardioExchange’s Harlan Krumholz interviews Peter Berger, coauthor of an editorial accompanying the results of the HEAT-PPCI study comparing bivalirudin and unfractionated heparin during primary PCI. The study is published in the Lancet; details are available in CardioExchange’s news coverage.

Krumholz: Is this study game-changing enough to shift practice and influence guidelines?

Berger: Usually a single study is not sufficient to change practice. However, the bigger and better the study, the less that rule ought strictly to be adhered to. Also, it must be remembered that there aren’t any prior studies that indicate that bivalirudin is superior to heparin alone (at the dose in this study; let alone with other doses, radial access, and potent P2Y12 inhibition). I am not counting the BAT trial from the late 1980s, which compared different doses of both drugs, included prolonged postprocedural infusions, and involved 8-Fr sheaths; that trial’s findings were reported as negative in the New England Journal of Medicine but were reanalyzed years later and reported to be positive in American Heart Journal— and the BAT trial was not in STEMI patients. So the question is not whether prior studies ought be abandoned because of a single, more-recent study. There aren’t any prior studies to help clinicians decide whether to use bivalirudin or heparin alone in STEMI patients.

When study results favored bivalirudin or suggested equivalence of heparin with a more expensive regimen, proponents of the more-expensive drug could not be heard advocating for additional studies.

It should also be noted that due to a great many price increases, bivalirudin is approximately as expensive (average weighted cost of $875 per vial, and in some cases two vials are required) as the glycoprotein (GP) IIb/IIIa inhibitors abciximab and eptifibatide. (Tirofiban is much less expensive than the other GPIIb/IIIa inhibitors, but has <5% of market share in the U.S. It is more commonly used in Europe.)

So, my answer is that clinicians will have to decide for themselves whether one large, well-done trial — the first examining these two drugs in these doses — should lead them to change practice and use a drug that was shown to be more effective and as safe as another, at <1/400th the price.

Krumholz: You have indicated elsewhere that you were already weighing changes in your practice before these findings were published. What evidence prompted you to consider such changes, and why?

Berger: The cath lab where I practice is almost entirely radial (except in patients with contraindications to its use). The most common site of vascular hemorrhage is the arterial access site, which is reduced 70%–90% by use of the radial artery, compared with the femoral. So we were already examining whether the frequency of bivalirudin use ought be reduced in certain patients and clinical situations.

Krumholz: How will this influence the use of decision aids for determining bleeding risk in patients undergoing PCI?

Berger: We don’t use any formal decision aids for determining bleeding risk. We include arterial access site and other clinical and procedural characteristics known to be associated with reduced bleeding. I readily acknowledge that our ability to predict bleeding in an individual patient is poor. But an important question is, if a patient is at increased risk of bleeding, which antithrombin should be used? Bivalirudin, say many proponents of bivalirudin; the HEAT-PPCI results would suggest heparin (at 70U/kg). And of course, the results of any trial should only be applied to patients similar to those enrolled (STEMI patients, etc., in HEAT-PPCI).

Krumholz: You point out the high percentage of radial access in this study, but subgroup analysis found no significant interaction was found between access site and the primary outcome. Doesn’t this imply that the results apply to non-radial cases as well?

Berger: Subgroup analyses lack the power to reliably detect potential differences that may exist. And statistical tests for interactions are notoriously insensitive. We do know for a fact that bleeding is more common with femoral than radial access. We don’t know if the results of HEAT-PPCI would have differed had femoral access been utilized exclusively. My point was that proponents of bivalirudin believe it to be safer than heparin (albeit at higher doses); if such a safety difference exists and were to favor bivalirudin over even 70 U/kg of heparin, it might be less relevant if radial artery access is used.

I will note that NAPLES III did not favor bivalirudin over 70 U/kg of heparin, and only femoral artery access was permitted in that trial of exclusively patients at increased risk of bleeding. And if efficacy and safety outcomes are indeed similar with heparin and bivalirudin, that is obviously a huge win for heparin.

Krumholz: Is this a win overall for low-cost care?

Berger: Yes. I am not sure what role the cost of drugs or procedures ought to play when the more expensive one is associated with a superior outcome. Which readers of this blog want the cheaper but less effective or less safe treatment for themselves? But when different drugs are comparable in safety and efficacy, price should play a major role in selecting which one to administer. And again, in this case, 70 U/kg of heparin was more effective than bivalirudin, and as safe.