Severe cutaneous adverse drug reactions

About 50% of the implicated drugs in a SCAR are antibiotics. And that's how infectious disease physicians are involved and that's why hospital settings see these reactions because antibiotics make up a big chunk of them.

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

Hi, I'm Jo Cheah, and this is the Australian Prescriber Podcast. Joining me today is Associate Professor Jason Trubiano, an ID physician and Deputy Director of the Centre for Antibiotic Allergy and Research at Austin Health in Melbourne. Jason has co-authored an article with Assistant Professor Ana Copaescu, the assessment of severe cutaneous adverse drug reactions. Welcome to the podcast, Jason.

Thanks for having me, Jo. It's a pleasure.

So to start us off, would you please talk us through the defining features of a severe cutaneous adverse drug reaction or ADR, and what are some examples of these reactions?

Well, SCAR, I think's a great way to describe them because it’s exactly what they do ­– severe cutaneous adverse drug reactions – they're a group of immune-mediated reactions caused by our body's T cells and they're really interesting because they're quite a diverse spectrum. You've got things such as acute generalised exanthematous pustulosis or AGEP, which is like thousands of small, tiny pimples all over the body, generally occurring soon after a drug. And it's the baby of the SCAR because it's not as severe as the others.

Then you move through to drug reaction with eosinophilia and systemic symptoms or DRESS, which really is internal organ involvement resulting from a drug reaction. You get raised eosinophils, you can have liver failure or renal failure and quite extensive rash. And this is really a significant drug reaction.

And then you go to the real baddies, Steven-Johnson syndrome and toxic epidermal necrolysis. Really they're blistering skin reactions. They're where you put a thumb on somebody's skin and it literally peels off. That's what ends people in intensive care and has a mortality rate upwards of 25%.

So you've got this huge spectrum of severe cutaneous adverse drug reactions and outside that we've probably got a range of ones that are not quite them. And we call some of them mini DRESS, where they've got a bad enough reaction that they've ended up in hospital, but doesn't make the diagnostic criteria of those syndromes. They're under reported, probably underappreciated, and we probably need clinicians understanding them a bit more.

And which drugs are most commonly implicated and why?

The drugs implicated really seem to go with the different phenotype or the type of reaction. And so not all drugs will cause certain reactions. So we know, for example, Steven-Johnson syndrome, antiepileptics, allopurinol, co-trimoxazole are probably the most commonly implicated drugs we know. And some of this data is interestingly being extrapolated from our national registry of severe cutaneous reactions, AUS-SCAR. But you flip over to DRESS, and DRESS is commonly caused in an Australian setting particularly by vancomycin and also co-trimoxazole. And then you go to AGEP, and it's commonly caused by simple aminopenicillins or aminocephalosporins. The universal feature though with causality seems to be that about 50% of the implicated drugs in a SCAR are antibiotics. And that's how infectious disease physicians are involved and that's why hospital settings see these reactions because antibiotics make up a big chunk of them.

Interesting and sorry, what was that database you mentioned before? AUS-SCAR?

It's really a registry for collecting patients that have had a SCAR from around sites in Australia. And there are 21 sites participating in this through Australia. And so hospitals can join this study, record patients that have a SCAR reaction. And it's really so we can answer some of these things you've already asked us. What are the commonly implicated drugs in an Australian setting? What are our outcomes? Do we see more SJS or DRESS? And this kind of setup's been done in Europe before, but really not in Australia.

This sort of data collection is on top of or a duplication of what you would be submitting to the TGA anyway?

It's an add-on and it certainly collects a lot more granular data about the phenotype and it also requires patient consent. And we can even get some blood from these patients to try and work out some diagnostics, some blood tests to try and ascertain causality. So it's got some unique features to it, more than the simple TGA report.

Yes, yes, definitely. And I guess since I work at a paediatric hospital, have you considered enrolling paediatric sites yet to AUS-SCAR?

Absolutely. They're not just little adults, are they? And they need to be included separately. Some of the logistics have obviously been blood collecting in a paediatric viewpoint. But no, we need to address this in peds. It happens in peds as well. So we can't forget them.

Very interesting. Can you talk us through the drug causality tool – the Naranjo score – which you've mentioned in the article?

So, there are a range of tools available. Naranjo's just one of them. It's the old school one, it's been used probably the most in most studies. And it was initially developed in Toronto back in the nineties. And it's just a way of ascertaining the likely causality of a drug being involved in a drug reaction. And I really encourage clinicians of all types to do this when they're thinking about a drug involved and the scoring system goes from, I'm pretty sure minus one to nine. And across that spectrum, it gives you an idea of the likelihood of the reaction being probable or definite, nine being definite. And it asks some simple things like, is there previous reports of this reaction? Did the reaction happen after the drug? Did it go away after you stopped the drug? Was there evidence of the drug in the body, if it was one that we could do therapeutic levels?

And so it's a really good way of trying to define – is this drug really implicated, because I think a lot of the time we implicate every single drug on a drug chart, or every single drug in a blister pack, where the patient may have been on them for many years or stopped taking them and they're not involved. So Naranjo is a really good way of doing that. And there are other scores. I would implore everybody to think about looking at a RegiSCAR score, which is the one we use for DRESS. But Naranjo is one we can use for all drug reactions.

And from a prescribing and a medication review point of view, for example if you are a GP or a doctor in a hospital, or community pharmacist or a hospital pharmacist reviewing drug charts, it can get a little bit concerning when you start seeing drugs prescribed where there's a risk of drug allergies and cross sensitivities. And I know that the data in this area has changed, or the guidelines around what's acceptable and what's not. So do you have any advice or practical tips for people who are in those situations?

The key thing is that most of the time, drug cross-reactivity or drug allergy is irrelevant because the vast majority of people aren't allergic. But in this situation, we're talking about SCAR, which are real and true. And we really don't know a lot about cross-reactivity and safe prescribing in this group. Some good tips are about sulfurs and I've talked about sulfurs being an implicated drug in a lot of these. If somebody reports a non-antibiotic sulfur allergy, they're free to take antibiotic sulfurs and vice versa, even in something like SCAR. So that's a clear misnomer. The thing we have to be worried about is prescribing related cross-reactivity beta-lactams, particularly in somebody with a beta-lactam-reported SCAR, or antiepileptics. These really need careful consultation and we should never rush into prescribing these.

I think the thing we can all do is just not ignore the allergy tab in the medical record, or when we're doing the outpatient script in the GP practice. If somebody reports an allergy, it's just worth spending five minutes or even three minutes asking about the phenotype and looking for some red-flag signs. Did they have a blistering rash or mucosal involvement? Did they have a drug reaction that involved their internal organs? Did they have a rash that lasted more than 15 days? If they answer yes to any of those, you may have had a phenotype that is along the SCAR spectrum, and we shouldn't be using drugs in the same class or cross-reactivity in those groups. And I think there're simple things that we can all ask that don't take very long and a patient will definitely remember if they've had a SCAR before. There's no doubt about that.

And in terms of a person's risk to developing SCARs, are there any factors that increase a person's risk? And if that's the case, do we routinely screen for these sorts of factors in Australia?

We can do it for some. And we know that HLA, these are genes that predict the way white bloods cells present antigens to our immune system, certain HLA codes or types predispose to certain reactions, and they particularly relate to ethnicity. And so particularly in Asian populations, allopurinol, dapsone, and a lot of the antiepileptics where we know we can screen for the HLA before they get a drug. And I think we really should be thinking in those ethnicities where there are commonly associated HLAs, that we should be doing more screening. The problem is, you asked me at the beginning, the most commonly implicated drugs, none of them currently have validated screening tools or algorithms we can use. We'll get there. Things like AUS-SCAR will enable us to do that and find causality and relationships, but there are a few in specialist centres, but that's it really. Dapsone, allopurinol, antiepileptics, outside that we're probably still in a research zone.

Very interesting. And generally speaking, how should SCARs or adverse drug reactions be monitored and reported?

You're talking about the ideal hospital, the fairy land hospital?

Ideal. Yeah.

Ideal, every single hospital or health service would have a medication safety review panel or an adverse drug reaction committee. And you'd have a group of clinicians, pharmacists, nurses running through clinical cases, ascertaining a Naranjo score, assigning a phenotype, and then reporting that to the TGA and then reporting that to the patient with a little card documentation, and then sticking it in their My Health Record so that everybody could see it.

And so I think we have to get to a stage where there is no excuse for us not reporting a SCAR and they often get missed. And we've got these terrible stories of people being labelled as a rash and they re-present and get the same drug again. And then actually, when you look through the history of the other hospital, they had a SCAR. So I think it's on us all to report the patient adverse reaction back to the patient and give them confirmed documentation. Where we can, update our own medical records and send it off to community hospitals and linkages that we know are going to also want to know, GPs, community pharmacies, and find better ways of getting this into a universal medical record. And I know we can't do this for all drug reactions, but we've got to be doing it for the tip of the iceberg. And when somebody's flagged as having a SCAR, I think it's on us to do that, moving forward.

Yeah. And you talked a little bit about it already in terms of looking for red flags when you're taking history from a patient. But if we go back to that ideal setting again, what is the ideal history of an allergy that a clinician, pharmacist, nurse, doctor can take when they've met a patient for the first time or on subsequent visits?

Well, we try and teach this ideal world and I think we're going to get to a stage where we have to, now that the National Standards said that we must do an antibiotic allergy assessment on patients admitted to hospital. So we're getting there. It starts with firstly naming the drug. Simply just saying a class is not appropriate. We've got to find the name of the drug, because we know people can tolerate drugs within a different class. So first one is drug.

Second one is time. How long ago? Five, 10, 30 years? These kind of time frames are really important because we know people lose allergy over time. So we know that 50%, for example, of minor drug allergies are lost over time after five years, 80% after 10. So when we're talking about non-SCAR, that 5–10-year frame work's really important. And that's why we use that as a cutoff. But for SCAR, it's irrelevant. If they've had a SCAR five, 10, 30 years ago, you don't re-challenge somebody just because of time passed.

Three is reaction. What was the reaction type? Tell me about the reaction. What do you remember? And four is tolerance. Tell us what you can take safely. Because a lot of the time that gives us really key clues about what they can tolerate in terms of cross-reactive drugs. And it's always important to say what you've tolerated after the adverse drug reaction because whatever has come before that is irrelevant. It's only important what they've tolerated since.

All right. So, that actually brings us to the end of the episode. So, thank you so much for chatting with me today Jason.

It's been a pleasure. Thanks for having me.

[Music]

Jason and Anna's full article is available online. Jason was supported by the Austin Medical Research Foundation and by a National Health and Medical Research Council postgraduate scholarship. The views of the hosts and the guests on the podcast are their own and may not represent Australian Prescriber or NPS MedicineWise. I'm Jo Cheah, and thanks for listening to the Australian Prescriber Podcast.