In this episode, I’ll discuss penicillin allergy and cross-reactivity.
Subscribe on iTunes, Android, or Stitcher
A big shout-out to those of you that took a moment to leave a review of my book A Pharmacist’s Guide to Inpatient Medical Emergencies on Amazon. The reviews are essential for helping a potential reader decide if the book is for them. One anonymous customer wrote: Very practical. This is a great resource for new hospital pharmacists students and residents. Good advice on how to be most effective as a pharmacist responding to emergencies, especially for newer pharmacists or those new to rapid response teams.
If you have the book, please head over to Amazon to leave a quick review.
The penicillin allergy dance is one that replays itself over and over on a daily basis. With recent warnings on non-beta-lactam antibiotic classes such as fluoroquinolones, the importance of determining what beta-lactam antibiotics a patient with a penicillin allergy can safely tolerate has never been greater.
Most patients who report a penicillin allergy don’t have one. In fact, the number of patients who report a penicillin allergy but can tolerate penicillin may be as high as 90%.
Patients that report penicillin allergy on hospital admission have longer lengths of stay and more complications than patients without a reported penicillin allergy. Specifically, C. difficile, MRSA, and VRE infections are higher in patients who report a penicillin allergy. Presumably, this is because second-line non-beta-lactam antibiotics are used when penicillins are avoided.
Cross-allergenicity rates
Cross-allergenicity rates are much less than what you may have learned in training. Many clinicians quote a cross-allergenicity rate between penicillins and cephalosporins as 10%. The actual rates of cross-allergenicity are far less.
Cephalosporin-penicillin cross-allergenicity rates are somewhere between 0.1% and 2%.
Carbapenem-penicillin cross-allergenicity rates are less than 1%.
Aztreonam-penicillin cross-allergenicity rates are 0%.
The risk of cross-allergenicity may not be clinically significant, depending on the type of reaction. IgG-mediated reactions such as rash are never life-threatening. IgE-mediated reactions such as hives, angioedema, and anaphylaxis are often severe. A patient’s reaction will never switch from an IgG-mediated to an IgE-mediated reaction, but an IgE-mediated reaction may manifest the first time as hives and the next time as anaphylaxis.
Determining the type of allergy
When evaluating a patient with a reported penicillin allergy, I will first try and determine whether the patient has an allergy and if so is it IgG or IgE-mediated? Here are the four questions I will typically ask:
1. How long ago did the reaction occur? B-cell memory fades, so a childhood reaction in an elderly patient is often not significant.
2. What exactly was the reaction? Whole-body rashes imply an IgG-mediated reaction, while hives, anaphylaxis, and angioedema imply IgE-mediated.
3. How was the reaction treated? Reactions that were treated with an ED visit or hospitalization are more worrisome for being IgE-mediated.
4. What other antibiotics can you take without a problem?
Risk assessment
After I determine whether the allergy is true and whether it is likely IgG or IgE mediated, I ask myself a patient-specific question. For example, if I am considering using a carbapenem in a patient with a rash to penicillin I ask “Is it acceptable to take a less than 1% chance of causing a rash in this patient?” Of if I am considering using a cephalosporin in a patient with anaphylaxis to penicillin I ask “Is it acceptable to take a 2% chance of anaphylaxis in this patient?” Framing the question like this makes it easy to decide whether the risk is acceptable or if an alternative antibiotic is preferred.
Skin testing
I can not remember ever using skin testing to evaluate a patient’s penicillin allergy. If this is done major and minor penicillin determinants must be used. Even if a skin test is negative, cross-allergenicity may still occur due to similar side-chains.
Side chains rather than the beta-lactam ring may be responsible for cross-allergenicity. This is why patients allergic to ceftazidime may also be allergic to aztreonam; the side chains are similar.
Graded challenge
A graded challenge does not affect a patient’s immune response to a potential allergen. Instead, it is hoped that any cross-allergenicity will be mild if a small dose of the allergen is given. To perform a graded challenge first give 1/100th or 1/10th of a typical dose, then increase by a factor of 10 until the entire dose is given.
Desensitization
If a penicillin must be given to a patient with an IgE-mediated penicillin allergy, desensitization must occur. The CDC has a thorough protocol on how to perform penicillin desensitization.
Summary
To summarize, if I encounter a patient with a reported penicillin allergy, I make an attempt to verify the allergy and determine if it is IgG or IgE-mediated. I then make a patient-specific risk:benefit assessment of whether the risk of cross-allergenicity is acceptable. Usually, I will use a cephalosporin if the penicillin reaction was IgG-mediated or a mild IgE-mediated one. If the reaction is a severe IgE-mediated one, I’ll lean toward a carbapenem, aztreonam, or non-beta-lactam antibiotic.
If you like this post, check out my book – A Pharmacist’s Guide to Inpatient Medical Emergencies: How to respond to code blue, rapid response calls, and other medical emergencies.
Most of us were taught in school that the cross reaction between ceph and PCN was ~10-15%. Then data came out stating 3-5% and now its reported to be more like <1% cross reaction. The latest theory that I find most interesting is that it might not as much as a cross reaction phenomenon as much as general immunogenicity. Some evidence suggests that pts with 'PCN Allergy" are basically just a likely to get rash with Septra as they are with Keflex making me believe that this might simply be general immune reactivity versus true cross reaction. Maybe cross reaction can occur but I think its FAR FAR less that what we think. Now that's with 1st/2nd generation.
I have been giving and monitoring Rocephin on an inpt basis to PCN allergic patients for over 10 years without any significant issues. I can only recall one patient having a anaphyl reaction and a few other typical ADRs but I after closer review I doubt these were even true cross reactions. I suspect they would have happened regardless as the ADR pts had extensive allergy lists suggesting immunogenicity (or psych mediated). As you know the cross reaction with 3rd generation is questionable at best.
Just my two cents worth…. I am sure more info will become available and in another 10 years we will be talking about the time we thought the cross reactivity was 1% 🙂
Thanks again Joe for bringing this issue up for discussion. It is something every pharmacist will face multiple times in their career.
Thanks Ash – I can’t wait to say “I’m old enough to remember when cross-reaction was thought to be 1%” 🙂
I appreciate the blog post above. However as a pharmacist who has some experience in allergy assessment and PCN allergy testing in general, I believe a few clarifications are needed.
1- I agree that cross-reactivity between cephalosporin use in a patient with a history of Type 1 reactions to PCN is very low. Romano, et al found that prospective skin testing in a cohort of confirmed PCN allergic (skin test and history of PCN reaction) patients first generation cephalosporins had about a 10% cross reactivity rate while 3rd and 4th generation cephs had essentially no cross-reactivity. This is in line with what we know about cross-reactivity potential of these drugs which is based almost entirely on their side chain structure. If side chains are similar (as is true with for example penicillin and cephalexin) than cross reactivity rates rise.
2-Romano’s data and data we’ve published from our own health-system show NO cross-reactivity between carbapenems and patients who claim a PCN allergy. In fact our study actually found that patients who did NOT claim such an allergy were more likely to have a reaction to a carbapenem than those who did claim such an allergy. Our health-system uses this class regularly in patients (as well as later generation cephalosporins) with no problems
3-Although major and minor determinants are optimal for PCN allergy skin testing that are NOT mandatory–which is a good thing as the minor determinant does not exist in the U.S. The AAAAI parameter on drug allergy clearly states that dilute PCN G (10000 U/mL) can be used as an appropriate substitute for the minor determinant. It also clearly states that if appropriate scratch testing followed by ID testing is performed the test is completely safe even in highly allergy patients. Finally there is little to no “cross-reactivity” between this test and other beta-lactams. In fact the NPV of this test is over 98% and those who test negative can receive the PCN class with no greater risk of a Type 1 reaction than the general population.
Pharmacists CAN perform this test (subject to their state’s rules). We have had an allergy assessment program a component of which includes PCN skin testing that is done by pharmacists for over 10 years now. We are collecting data on this patients for review, but our first experience with this service we published over 10 years ago.
REF:
1-Romano A, et al. Ann Intern Med. 2004 Jul 6;141(1):16-22.
2-Romano A, et al. Ann Intern Med. 2007 Feb 20;146(4):266-9.
3-Romano A, et al, N Engl J Med. 2006 Jun 29;354(26):2835-7
4-Campagna JD, J Emerg Med. 2012 May;42(5):612-20
5-Wall GC, J Chemother. 2014 Jun;26(3):150-3
6-Solinsky R et al. Annals of Asthma Allergy and Immunology 2010;105:e1-77
Excellent points & thank you for the references Geoff!
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6789778/
Making it a bit more interesting.