Conversations in Cardiology: Plavix and PPI

Dr. Morton Kern has engaged his colleagues in 'conversations in cardiology' brief, informal dialogue on relevant topics in interventional cardiology.  With permission from Dr. Kern and his colleagues, we present their conversations for the benefit of the interventional cardiology community.  Feel free to comment at the bottom of the page.

 

Our Colleague, Alice A. Perlowski, University of Chicago Medical Center, asked Ted Feldman, Evanston Hospital,  this question. . .

I was hoping you could shed some light on something that I am puzzled by.

Despite the fact that the Consensus statement in 2010 stated that there was an unknown and unproven clinical ramification for the interaction between Plavix and PPIs, why does the most recent FDA Safety comment on this in 2012 continue to state avoid concomitant use of the two meds?

I ask because at Walgreens and CVS, a "major Interaction warning" continues to pop up in their system when patients go to fill their medications .  The pharmacist is physically unable to dispense the Plavix when this warning comes up until a physician overrides this. They are supposed to contact the ordering physician when this happens.

Now the problem. . . .

I stented a proximal LAD on Tuesday. The patient was discharged on Thursday. When he went to fill his hospital Plavix Rx on Friday, the huge warning popped up on the screen about the interaction (patient already on Nexium I believe filled there at some other time).

Apparently no attempt was made to contact the ordering physician("since it was after hours") and the med was not dispensed. The guy came in today with horrendous stent thrombosis; had to angiojet the heck out of it. Pretty certain the stent was mechanically fine because I IVUSed it after placement on Tuesday.

Obviously pharmacy not calling the MD was mistake #1, but its really concerning that this warning prohibits pharmacists to dispense the med. In addition, it seems like there needs to be another warning box that comes up that says if they plan to withhold a thienopyridine that it should be an emergency phone call to a physician...

Thanks for your thoughts, Alice.

Synoposizing the question:

1.  Is there sufficient data to justify FDA Black Box.  If not how should we approach this?

2.  Should we address PPI for patients to receive Plavix each time we prescribe?

3.  Who should contact the pharmacy re this adverse interaction issue?

To answer the questions from my view. . .

1 and 2. I thought this was a dead issue and do not concern myself any longer.  I was unaware of the Warning to pharmacies and their failure to contact physicians over the critical issue above.

3.  I think an ambitious fellow could make this a focused review and with SCAI help contact FDA for a ruling.

Mort Kern
University of California, Irvine

 


Here are my thoughts in response to Mort's questions:

1.  Is there sufficient data to justify FDA Black Box.  If not how should we approach this?  The interaction with prilosec and nexium  is NOT in the black box, but is a "warning" in the label. There is incontrovertible evidence from several randomized pharmacokinetic and pharmacodynamic studies that these 2 PPIs reduce Plavix active metabolite (AM) levels  [by as much as 40% in some studies) and reduce the antiplatelet effect of Plavix. Therefore, patients getting these drugs may not get the intended effect, which is why the FDA put it in the label. They do this in the label for all sorts of medications with particular drug-drug interactions.

Whether you CARE that it interferes with plavix AM levels is another question entirely! Event rates are so low after PCI (esp. with stable CAD) that for most patients it probably doesn't matter. But for some it very well could. And Alice is correctly worried that her patient wasn't taking plavix…in patients with borderline response to begin with, this will essentially be what you are making happen if you add prilosec or nexium!   While the COGENT trial was negative (no worse CV events with combo plavix/omeprazole vs plavix alone), the study was underpowered, the endpoint was not only thrombotic events, and only a portion were recent stent or ACS patients; I have attached the FDA's response to the trial that was published in NEJM.

2.  Should we address PPI for patients to receive Plavix each time we prescribe? I try to avoid prilosec/nexium, and give one of the others instead. 

3.  Who should contact the pharmacy re this adverse interaction issue? I would be really surprised if this pharmacist followed a Walgreen's policy – this should be checked. Since the interfering agent is omeprazole, that was the one that should have been stopped. I was called recently by a pharmacy for just this – warning me that there could be an effect, and whether they could substitute omeprazole for another agent, eg., pantoprazole (Protonix).

Matthew Price
Scripps Clinic

 


I agree with all of Matt’s points. What this demonstrates is that patients with stable CAD have a very low rate of stent thrombosis and other thrombotic events after PCI. TRIGGER-PCI tested this exact situation and reported 1 event through the first 6 months in more than 200 patients who had high platelet reactivity on clopidogrel after a successful DES. Of note, however, decreased metabolism of clopidogrel in patients taking certain PPIs is not the same as taking no clopidogrel at all – so the drug should still be prescribed, although with best-in-class stents (especially fluoropolymer-based everolimus-eluting stents, which time and again have been demonstrated to have the lowest rates of definite stent thrombosis), and especially if IVUS was used, thrombotic complications should rarely occur. Nonetheless, this issue can be avoided by prescribing a PPI that doesn’t interfere with clopidogrel metabolism, such as pantoprazole. This is unlikely to solve the pharmacist problem, however. I doubt their warnings are that sophisticated – or are they?

In patients with acute coronary syndromes, the story is very different – the stent thrombosis rates are much higher, as is platelet reactivity (even in clopidogrel responders). These patients should in general all be treated with ticagrelor or prasugrel, absent major bleeding predisposition. And Champion Phoenix has undeniably demonstrated that effective ADP inhibition prior to PCI with IV cangrelor prevents intra-procedural stent thrombosis and other thrombotic events, as well as peri-procedural MI (across the spectrum of CAD). The implications for the oral agents are that in NSTEACS, ticagrelor (rarely prasugrel) should be given prior to cath, ideally 2-4 hours ahead of time given the potential for delayed absorption.

Alice, can you tell us whether your patient presented with an ACS or stable coronary syndrome? What type of stent was used? Was clopidogrel pre-loaded? If so, how long before the procedure? Were there any procedural complications? It would also be interesting to see the films and IVUS to exclude under-expansion or edge issues, or other rare causes of stent thrombosis such as longitudinal stent deformation, fracture, etc. Finally, in some of these patients we find an anti-phospholipid antibody, protein C or S deficiency, etc., although when you hear hoof beats…

Gregg Stone
NewYork Presbyterian
Columbia University Medical Center
Cardiovascular Research Foundation

 


This really is an unbelievable case and one that requires investigation. Clearly the pharmacist was wrong, and obviously the fact that we have these "warnings" that are now computerized and prevent drug-drug interactions is a two edged sword.

The workaround obviously is either changing the PPI as Gregg suggests -- but that may not really solve the computerized warning -- or changing the antiplatelet agent -- prescribe effient or ticagrelor rather than plavix. Also totally agree that even with all of that said, a stent thrombosis that quickly does suggest the possibility that in this case there was residual thrombus or malapposition or edge dissection or some other contributing factor.

Finally, did he take the aspirin? I have seen stent thrombosis with omitting one or the other agents, but its more common in my anecdotal experience when both agents arent being taken.

Lloyd W. Klein
Rush University Medical Center

 


I agree with what’s been said so far.  It’s worth keeping in mind that clopidogrel is a less desirable product for higher-risk patients who are a appropriate for prasugrel or ticagrelor, so the conversation should focus on changing the anti-platelet therapy rather than the PPI. 

The communication lapse with the pharmacy was the essential missed opportunity; if pharmacists are going to be prohibited by FDA from dispensing clopidogrel to patients taking PPIs, they should be empowered to provide a limited amount of alternative anti-platelet drug rather than send the patient home empty handed – this case demonstrates the risk, and a simple algorithm to assure appropriate patients get alternative drugs for a few days would be easy to construct.  That would allow time to assure communication.

How did the patient do?

Kirk N. Garratt
Lenox Hill Hospital
 


I also agree with Matthew’s comments. Attached is the manuscript with the PK and PD data that were requested by the FDA to better define the drug interaction which had been previously identified by other investigators – OCLA study. This prompted specifically designed studies to be performed which included not only PD but also PK assessments in the setting of 4 dedicated PK/PD studies - attached). The observation of an ~40% reduction in generation of the active metabolite due to the interaction (this was a real interaction at the hepatic levels as determined by assessment of the active metabolite of clopidogrel and not an absorption problem secondary to changes in gastric pH given that circulating clopidogrel pro-drug was also determined and not affected). This interaction (assessed by PK) resulted in an ~20% absolute change in PD. The obvious question is whether this is clinically meaningful or not. As highlighted by Matthew and Gregg, the potential for an adverse clinical event  depends on the patient profile (with more concerns in an ACS patient because more effective platelet inhibition is required). A nice paper in JACC from Montalescot’s group identified a profile of these patients at higher risk.

So..is the warning justified? Yes. The Code of Federal Regulations (21CFR 201.57 [e]) requires that “labeling shall be revised to include a warning as soon as there is reasonable evidence of an association of a serious hazard with a drug; a causal relationship need not have been proved.” Therefore, physicians and patients need to be informed on this. Fortunately, there are multiple ways around the interaction: a) consider newer agents (prasugrel, ticagrelor) if clinically indicated; b) if patient needs to be on clopidogrel, consider a non 2C19 interfering PPI (pantoprazole, dexlansoprazole) and avoid omeprazole and esomeprazole.


Dominick J. Angiolillo
University of Florida College of Medicine
Jacksonville, Florida

 


Here is the PPD data from TRILOGY-ACS poster at the ACC.  Unfortunately, it further highlights the challenges in this field.  The PRU data does not support too much of a differential effect by the phenotype, but the clinical data is quite striking.

It may be smarter to avoid clopidogrel if PPIs are used.

Magnus Ohman
Duke University Medical Center

 


Thank you so much for all of your comments, and for including me in this forum; this case has been very educational to say the least. There is no doubt that my patient's ST was multifactorial. He is very high risk for multiple reasons, originally presented with a NSTEMI, and required bare metal stent placement due to concerns for noncompliance and need for bilateral hip replacements in the next few months. No doubt that the omission of Plavix may have been just the tip of the iceberg. His case highlights many of the challenges of my practice environment on the South Side of Chicago, where the majority of patients are uneducated and underserved, as well as noncompliant (thus BMS are used in many cases). I completely agree that prasugrel or ticagrelor would have been a better option for him for many reasons (particularly since he needs to be on Nexium due to h/o bleeding gastric ulcers), yet in our patient population who are largelyuninsured, we depend completely on free samples and free coupons to obtain these meds for them- and these are not always readily available (or are they, and we have missed the boat on obtaining them for all of our patients with no major contraindications?). Therefore, Plavix is still used here in many cases.

I was quite surprised to learn the reason why my patient did not receive his Plavix, and was not aware that these warning screens were still present. In surveying pharmacists from my own hospital, CVS and Walgreens, I was told that the computer screens are reflective of the package inserts of the medications. Interaction warnings are classified as Major, Moderate, or Mild. When a major interaction warning pops up on the computer screen, the pharmacist does have the capacity to override this warning if he or she feels comfortable. Many pharmacists told me that they overridethis warning routinely, or will call the MD to obtain permission to switch the esomeprazole or omeprazole to pantoprazole. Conversely, one pharmacy manager told me that some of his more junior employees would not feel comfortable making an override of a Major Warning independently, and thus rely on physician overrides, which they must phone a physician to obtain.  This is what happened in my patient's case- and the pharmacist was incorrect for not phoning a physician even though it was after hours. When I asked how we in the cardiology community could prevent this type of event from happening again, I was told that when writing the prescription, we should make some acknowledgment of our awareness of the drug interaction and give permission for override of the major warning.

Interestingly, many of us use computer systems like Epic to write our prescriptions. When I prescribe a PPI and Plavix together in Epic (at least in my system at U of C) no interaction warning comes up and I am able to freely sign the prescriptions with no issue. There is no specific place in Epic (at least which produces a statement visible to thepharmacist) where I can check a box that says I override any med interactions, and thus I suppose I will start doing this manually or by telephone call, and will instruct my housestaff at U of C to do the same.

My patient is doing quite well. Although he did take a hit to his anterior wall, he will likely go home today or tomorrow. I switched him to prasugrel post thrombectomy and I am in the process of trying to obtain some samples for him to take home.

Alice A. Perlowski
University of Chicago Medical Center
 


With regard to the free samples issue, I have been led to believe that any patient can get a free first month of ticagrelor. After that, the deal depends on income, insurance coverage, etc.  But the first month should be available free of charge.  You may want to confirm with your local reps in the Chicago area.


David Cohen
Saint Luke's Mid America Heart Institute
 


Its not free unless you have a prescription that is filled at the same time so if you dont have a way to pay for the rest of it, you cant get the "free" sample.

Same with effient, by the way, and also with (another class of drugs, aI know) eliquis. 

Lloyd W. Klein
Rush University Medical Center
 


Recognizing this issue, we have taken a different approach. Our pharmacy committee met with representatives of the companies and we have been able to develop a “bulk” program that enables us to provide a 30 day supply to every patient, brought to the bedside and verified by nurses, at the time that they leave  the hospital. This is independent of whether or not they have insurance and has been an assurance to patient, family, doctors and nurses that we are not dependent upon patients having to procure this themselves.


Steven R. Bailey
University of Texas Health Science Center at San Antonio
 


Comments