Penicillin allergy

Welcome to the Australian Prescriber Podcast. Australian Prescriber, independent, peer-reviewed, and free.

Anyone who asks patients about their allergies will know that one of the most common replies is penicillin. I think a lot of us know this isn't always right, but does matter and, if it does, what can we do about it?

I'm Dr David Liew, your host for this episode. Today I'm speaking to two people who tackle this on a daily basis and who are working to make it better. They both happen to be from the Centre for Antibiotic Allergy and Research, which is at Austin Health, which is of course where I also work.

Misha Devchand is a antimicrobial stewardship pharmacist at the centre, and Jason Trubiano is an infectious diseases physician who is a director of the centre. Together, they've written a truly excellent practical guide in the December edition of Australian Prescriber on how to approach penicillin allergy. They join us here on the podcast today.

Misha and Jason, welcome to the program.

JT: Thanks for having us.

So, Misha, tell us about this. Why should we care about this as an issue? Does it matter if people get labelled as having a penicillin allergy or is it just a nuisance issue?

MD: Penicillin allergy labels are reported in about 20% of hospitalised patients in America. In Australia, we think that figure is at about 10%. Having a penicillin allergy label will definitely affect the treatment that patients get while they're in hospital. So a penicillin allergy label is associated with an increased length of hospital stay, increased readmission rates. So it's quite a big cost on the healthcare system.

This is something that's been well known. So we've known about penicillin allergies since about the 1940s. I think there is now more of an understanding of what a penicillin allergy label actually means in hospitalised patients. And especially now that there's growing concern about antimicrobial resistance, we really want to be using first-line agents for treating our infections, and most of the time those are penicillin-based antibiotics.

Right. So do you think that this is just something that we've been absorbing for a long time? It isn't a new issue, knowing that people have had penicillin allergies. Why is this something that we're only concentrating on now?

MD: So a majority of the patients that we see are labelled with a penicillin allergy in childhood. And generally those are associated more with either adverse drug reactions or with viral infections that they have at the same time as having the penicillin. So they're not actually true reactions.

We're also starting to see a lot more patients report family history of allergies. So the patients haven't actually tried penicillin themselves.

So there really is an immediate imperative to deal with this. I mean, how do people end up with the wrong label in the first place?

JT: It's the one time in life where mothers are not always right. Mothers have told us that we had an allergy in childhood. The vast majority of those, when we go to test them in adulthood, are not able to be reproduced. And this has been found in Australia, in European studies, in the US. The one consistent thing and theme is that penicillin allergy ain't forever.

So what do you mean by that? I think a lot of people would go around in their mind that once you have an allergy on the list, then it's there for life. Why isn't it the case that we stick with allergies forever?

JT: I think it's been a mixture of poor teaching and new understanding, to be perfectly honest. I think the data that's come through over the last probably 10 and even 20 years from the allergists is that some allergies are not persistent. So there's very nice data that demonstrates that 50% of penicillin allergies are lost at five years and 80% are lost at 10 years. So there is waning immunity. So time certainly is on our side, and then when these patients come to the healthcare network in their 40s, 50s, 60s, 70s or even 80s, reporting allergies from the 40s and 50s, it's just next to impossible for them to be able to retain those allergies.

I guess the other question relating to this is about cross-reactivity. I think we're all aware of the different belief system previously about what cross-reactivity meant between different beta-lactams, between penicillins and cephalosporins and the carbapenems. Where do we stand on that nowadays?

JT: Well, we need to go back and reteach this 10% number. I think if you ask a range of healthcare stakeholders, pharmacists, nurses, clinicians, we were all taught there was 10% cross-reactivity between penicillins and cephalosporins. And certainly that is incorrect. The true rate of cross-reactivity between penicillins and cephalosporins is somewhere in the vicinity of 1 to 2%. So knowing these cross-reactivity structures now have been able for us to predict true cross-reactivity, but also bring down that 10% figure to this 1 or 2% number that we now quote.

And you've got a really nice picture of this in the actual article itself. So let's just go through that clinically relevant cross-reactivity again. So it's all down to the side chains. Which cross-reactivity should clinicians know about?

JT: Yeah, I mean, so the first thing to say is that this is not an absolute, but certainly in people that have genuine reactions, we wouldn't want clinicians using those cross-reactive drugs. And so, the particular example is this one of aminopenicillins, so amoxicillin and ampicillin sharing a side chain with cephalexin and ceflacor. That's the standout one. If you look at the literature around the globe, it varies, but certainly we seem to follow our European friends and that is a predominant thing that's been reported in Australia.

The other standout cross-reactivity or is the absence of cross-reactivity is that with cefazolin. Cefazolin is a first-generation cephalosporin, the most common antibiotic used in surgical prophylaxis or skin and soft tissue infections. And we found people that have an allergic reaction to cefazolin or anaphylaxis are able to tolerate, after we do testing, other penicillins and cephalosporins. So this is one of the unique examples of an absence of cross-reactivity.

The final one is for our friends that look after complex respiratory patients or cystic fibrosis patients and there's a drug called ceftazidime and it clearly shares cross-reactivity with a forgotten antibiotic, called aztreonam. So in primary care, really that's not going to be a scenario that people encounter, but in a complex respiratory patient, well, that would be something to be wary of.

And is this the kind of thing that just applies to hypersensitivity? What about other reactions to antibiotics?

JT: So type A side effects, if we touch on those quickly, I mean, they're non-immune, they're side effects. They don't get predicted at all by cross-reactivity. We should be wiping those of everybody's drug chart, in my opinion.

The other types of hypersensitivity that stand outside the Ig or the immediate, things like anaphylaxis, urticaria, angioedema, well, the data is less well described in these, particularly the delayed reactions. We've recently published some data where cross-reactivity in severe delayed reactions might be a bit more widespread. For example, if you have a reaction to flucloxacillin, well, you would certainly avoid all other penicillins like Amoxil. But even in those groups of patients we found that cephalosporins were safe.

So I think the advice from us would be, certainly most of the data is in IgE or immediate hypersensitivity, and emerging data in delayed or T-cell mediated hypersensitivities.

We've got a bit of a sense of the kind of allergies that we can have and the relevance of the different types of cross-reactivity. But, Misha, how do we apply this in practice?

MD: The key point is to have a good allergy history. So we ask questions on three key topics. So we talk to the patient about the timing of their antibiotic allergy. So when did it happen? How long ago did it happen? We talk to them about the severity of their reaction. So, while we ask them how severe the reaction was, we also ask them other questions. We ask them things like, did you go to hospital for the reaction? Were you given any injections for the reaction? And then we also ask the patient about tolerance. So what antibiotics have you tolerated since the reaction? We say that the point of tolerance is from when they've had the reaction. So any antibiotic that they've tolerated prior to the reaction no longer matters, and it's only since the reaction, what antibiotics have they tolerated?

Okay. So we're getting some really useful information there in terms of being able to make an assessment. I mean, all those things are very relevant, but what does that mean in terms of whether we give a patient a penicillin again or not in the future?

JT: So I think that the new Therapeutic Guidelines has made an attempted at doing this as well, and we've highlighted some of this that immediate and delayed are two important distinctions to make. So was this an immediate onset allergy or was this a delayed onset? And then once you get those, it's important to work out how severe. Is this a severe immediate or a non-severe immediate? And so, classic examples would be a severe immediate reaction, anaphylaxis. Those are patients we shouldn't be taking much risk with. Only using cephalosporins for targeted situations, life-threatening infections in somebody that reports a penicillin allergy.

The patient that has a non-severe immediate like a history of mild urticaria from a few years back, well, then things like cephalosporins in a supervised setting are fine.

The flip side is the delayed, and I think most of our attention should be focused on the non-severe delayed. These are things like the childhood reaction, mild maculopapular rash. They're ones that we certainly should be employing cephalosporin use in. And in the hospital setting we are certainly undertaking oral challenge in those patients, and I think that's the space looking forward where in primary care we need to work out ways in which we can undertake supervised challenge for those patients as well. So I think putting people into boxes into severity and the type really can help in terms of prescribing.

So why don't you tell me about what you do in terms of an oral challenge. Because I think that sounds like a scary thing for a lot of people, taking someone with an antibiotic allergy and then going ahead and giving them an antibiotic.

JT: Well, most of the time, I mean, before we go to oral challenge, the work is done if a clinician or pharmacist is smart. By just pure reconciliation, you can directly de-label a vast majority of patients. For example, they report amoxicillin allergy in hospital, you ring the GP and said, "Oh, by the way, they had Amoxil last week." Well, clearly you can wipe that off the drug chart.

So once you're then left with somebody that has an allergy, if they fit into a low-risk criteria, and certainly in our mind, which we're undertaking in a large health services program here at Peter Mac, we coonsider low-risk things like a childhood rash or a rash where they can't remember it more than 10 years ago, we undertake a low-dose test dose challenge, which is giving the smallest dose possible of an antibiotic in a supervised environment. This is very safe. In our pilot data, it was extremely safe, and certainly in our emerging data it's the same way.

We're not asking people to go out and randomly challenge people at a low risk. You certainly need governance and protocols around it, but there's certainly a huge group of patients that are going to benefit from this test dose challenge, which is both safe and also ensures patients get the right antibiotics post the test. So there's certainly going to be a big scope for this moving forward.

I mean, it sounds like, from what we've talked about, that there's a massive burden out there of patients with inappropriate labels as far as allergy are concerned, and trying to tackle this on a broader scale is something which might seem intimidating for some people.

JT: Yeah, I mean, there's probably two million Australians walking around with a penicillin allergy, and there's lots of association-level data now that that just means bad health outcomes for that patient. When you're able to remove that allergy, we've seen huge jumps in appropriate prescribing. So almost a 10-fold increase in the studies we've done here, and that's been sort of reverberated around the globe. So you can make a big difference to patient's healthcare.

Okay. So what happens during an oral rechallenge when there is actually someone who does have a true allergy? I mean, have you had problems with that or is that fairly easy to control?

In the pilot data, we had zero in 50 patients, and in our interim data, which we are yet to publish, we're talking in the vicinity of 1 or 2% of patients that have a reaction. And that reaction is a mild rash that doesn't require treatment. So if you pick the right patients, it can be done properly. It doesn't mean that somebody can't have a severe reaction and you shouldn't have safety and all the resources available to do that, but if you pick the right patients, it's very low-risk. And that data's been shown in a lot of Australian studies and overseas studies as well.

The other thing, we've talked a lot about the low risk, we're not trying to dismiss that there are patients out there that are going to have severe reactions and true allergy. There's probably a small proportion, maybe 1% of all patients that report a true allergy. And still, these frameworks of assessment and testing are going to be able to accurately identify those patients and put a nice safety valve around those patients and protect them as well.

So we're not trying to dismiss all people's allergies, we're just trying to make sure that we put people in the right boxes we mentioned before.

So how do we scale this in terms of trying to deal with this population across the community? I mean, it sounds like experts like yourself can really risk-stratify well, but there are obviously a lot of people out there who can potentially be de-labelled. Are there ways in which we can try and tackle that as a community?

We’re just talking amongst ourselves here about this question, to be perfectly honest. Because you mentioned the word community as a community approach, which we desperately need services. We need investment and funding to roll out these assessment criteria and protocols in primary care. It’s certainly feasible. It just requires resources. And that’s going to be going back to medical school and reteaching some of these cross-reactivity facts and also going out there and implementing these health services protocols.

Misha probably can talk about the hospital, but certainly at a pharmacist level, this is scalable in terms of assessment and challenge, do you think?

MD: Yeah, and definitely we've been... So here at the Austin, we've implemented a program where nursing staff and pharmacists do take thorough antibiotic allergy histories and we are being able to identify patients with low-risk penicillin allergies, and we are able to orally challenge those patients while they're in hospital.

JT: Do you think there's an appetite in pharmacists for this? Is this something that is on their radar?

MD: Pharmacists, like doctors, were taught back in the day when there was 10% cross-reactivity, and a lot of that stuff is only… their thought process around that is only changing now. But pharmacists are definitely eager to be involved in antibiotic allergy assessment and they're very interested in understanding this topic and understanding more about how to take a good antibiotic allergy history and what that means in terms of prescribing antibiotics.

So what I'm hearing is that it's about trying to improve broader prescriber education, but it's also about trying to find and stratify the right types of patients, even in hospitals. So we're targeting the high-risk patients. So broadly, what kind of patient groups do you think are going to be the first ones at the front of the line in terms of who we should be looking at?

JT: Well, this to me is simple. Anybody that's immunocompromised should not be allowed to leave a hospital with a penicillin allergy if it's of the crappy variety. So I think immunocompromised patients, they're at the front of the queue.

Just behind them in the line would certainly be people that are frequently engaged in health care – chronic respiratory patients, general medical patients that are frequent flyers to hospital. And I think if you target that group, you're going to have the biggest bang for buck in terms of an intervention. Would it be great to go and nip these in the bud? You know, the 18-year-old that develops a penicillin allergy or reports one from childhood, and then prevent a life of using the wrong antibiotics? Yep, that would be fantastic. But in terms of resource allocation at the moment, that's probably going to be somewhere down the line.

But in the meantime, there's an excellent article in the Australian Prescriber to give you some guidance, at least in terms of trying to give us a start.

Misha and Jason, it's been great having you on the program. Thanks for joining us today.

JT: Thanks a lot.

MD: Thanks for having us.

The views of the guests or the host of this program are their own and may not represent Australian Prescriber or NPS MedicineWise. I'm David Liew, and thanks for joining us once again on the Australian Prescriber Podcast.