JAC

Journal of Antimicrobial Chemotherapy (2004) 54, 11551157

DOI: 10.1093/jac/dkh454
Advance Access publication 14 October 2004

Is it safe to use carbapenems in patients with a history

of allergy to penicillin?
Manica Sodhi1, Sandra S. Axtell2*, Joyce Callahan1 and Raja Shekar1
1

Infectious Diseases, Huron Hospital Cleveland Clinic Health System, East Cleveland, OH; 2Pharmacy,
Hillcrest Hospital Cleveland Clinic Health System, Mayfield Heights, OH, USA
Received 21 June 2004; returned 24 July 2004; revised 30 August 2004; accepted 11 September 2004

Methods: A retrospective review was conducted in a total of 266 patients who were administered either
imipenem/cilastatin or meropenem. The patients were admitted to the Cleveland Clinic Health SystemEastern Region Hospitals during the years 2001 and 2002.
Results: Fifteen of the 163 patients (9.2%) with reported penicillin allergy developed a hypersensitivity
reaction to meropenem or imipenem/cilastatin whereas 3.9% of the 103 patients without penicillin
allergy developed a hypersensitivity reaction to meropenem or imipenem/cilastatin. These results are
not statistically significant.
Conclusions: Based on this study and other similar studies, the true incidence of cross-hypersensitivity reactions between penicillin and carbapenems may be lower than previously reported. Carbapenem
use may be reasonable for penicillin allergic patients if caution is exercised.
Keywords: imipenem/cilastatin, meropenem, hypersensitivity

Introduction
Carbapenems are similar in chemical structure to penicillins and
therefore are associated with a risk for allergic cross-hypersensitivity. The package inserts for both Merrem (meropenem) and
Primaxin (imipenem/cilastatin) list a warning stating that
serious and occasionally fatal hypersensitivity (anaphylactic)
reactions have been reported in patients receiving therapy with
beta-lactams. These reactions are more likely to occur in individuals with a history of sensitivity to multiple allergens. There
have been reports of individuals with a history of penicillin
hypersensitivity that have experienced severe hypersensitivity
reactions when treated with another beta-lactam. Before initiating therapy careful inquiry should be made concerning previous
hypersensitivity reactions to penicillins, cephalosporins, other
beta-lactams, and other allergens.1,2 Selection of antibiotic

therapy can therefore be severely limited by the medication

allergies reported by the patient. Penicillin is commonly cited as
a drug allergy. Patients mistakenly report reactions such as
experiencing gastrointestinal upset as an allergy.3 6 Patients may
report intolerance to penicillin from several decades ago when
penicillin formulations were less pure and had more of an allergenic potential.6 Also, patients may mistakenly report symptoms
of the illness they were being treated for as an allergy such as
fever and yellow spots on the tonsils or a rash due to a viral
infection.5,6 Allergies may also fade over time. An adult whose
reaction may have occurred many years ago may no longer be
allergic, although in many cases the patients self-reported history may be unreliable.6 An accurate drug allergy history is one
method to limit unnecessary avoidance of certain antibiotics
including carbapenems especially when they are necessary to
treat serious, life-threatening infections.

..........................................................................................................................................................................................................................................................................................................................................................................................................................

*Corresponding author. Tel: +1-440-312-4396; Fax: +1-440-312-7104; E-mail: saxtell@cchseast.org

..........................................................................................................................................................................................................................................................................................................................................................................................................................

1155
JAC vol.54 no.6 q The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.

Downloaded from http://jac.oxfordjournals.org/ by guest on February 18, 2013

Objectives: The purpose of this retrospective study was to ascertain the clinical safety of administering
carbapenems, namely imipenem/cilastatin and meropenem, in patients with a history of penicillin
allergy compared with administering carbapenems in patients with no reported penicillin allergy. Carbapenems are similar in chemical structure to the penicillins and therefore are associated with a risk
for allergic cross-hypersensitivity. Carbapenems are commonly avoided in patients with a reported
penicillin allergy on the basis of a potential cross-hypersensitivity with penicillin, however, very few
studies have been conducted describing the incidence of cross-hypersensitivity between penicillin and
carbapenems.

M. Sodhi et al.

Materials and methods

A retrospective chart review was conducted in a total of 266 patients
who were administered either imipenem/cilastatin or meropenem.
The patients were admitted to the Cleveland Clinic Health SystemEastern Region Hospitals during the years 2001 and 2002.
The review consisted of 163 patients with reported penicillin allergy
as well as 103 patients without reported penicillin allergy. The
details of penicillin allergy were collected. Allergy to imipenem/

cilastatin and meropenem was based on the appearance of a

hypersensitivity reaction, which resolved upon discontinuation of
the carbapenem. A hypersensitivity reaction was defined as developing a rash, hives, fever, angio-oedema, bronchospasm or pruritis that
led to the discontinuation of the carbapenem. Other antibiotic allergies were noted and also the number of days the carbapenem was
administered before the development of the reaction.
Charts of 163 patients with no reported penicillin allergy who
received either imipenem/cilastatin or meropenem were identified to
serve as controls. Sixty of the patients identified without penicillin
allergies were not included due to the unavailability of the charts or
not meeting inclusion criteria. These patients were matched in time
(admission to the hospital 3 days of the case patient) and age ( 5
years of the case patient) to patients with reported penicillin allergy.
Patients needed to receive at least 1 day of treatment with a carbapenem to be included.
Analysis of the data was accomplished using the Students t-test
to compare mean ages and the z statistic to compare proportions.
Results were deemed statistically significant if P < 0.05 and b = 0.2.
Statistical analysis was completed using Primer of Biostatistics.

Results
Fifteen of the 163 patients (9.2%) with reported penicillin allergies developed a hypersensitivity reaction to a meropenem or
imipenem/cilastatin whereas 3.9% of the 103 patients without
penicillin allergy developed a hypersensitivity reaction to
meropenem or imipenem/cilastatin. The z score was 1.390 and
P = 0.164, thus there was no statistically significant difference
between the two groups. Results are listed in Table 1.
Among the 15 penicillin allergic patients that developed a
hypersensitivity reaction, 11 manifested a maculopapular rash as
the reaction to the carbapenem. Only one of the patients developed facial oedema and lip swelling in addition to the maculopapular rash. One patient developed pruritis and another patient
developed drug fever in addition to a rash. The drug fever was
viewed as an adverse drug reaction rather than an allergic

Table 1. Comparisons of patients with reported penicillin allergy and without penicillin allergy
Patients with penicillin
allergy (n = 163)

Patients without penicillin

allergy (n = 103)

z score
T = 2.014

0.045

3.179

0.001

Mean age, years (range)

70.6 (32 91)

74 (39 91)

Sex, M/F (% male)

69/94 (42%)

65/38 (63.1%)

P value

Classification of the reported reaction to penicillin

rash
facial swelling and/or anaphylaxis
unknown
other

34.9%
6.1%
54%
4.9%

0%
0%
0%
0%

6.609
2.223
8.984
1.912

<0.001
0.026
<0.001
0.056

Other antibiotic allergies

cephalosporins
sulfa drugs
fluoroquinolones
macrolides

4.3%
22%
10.4%
5.5%

4.9%
10.7%
1.9%
1.9%

0.073
2.191
2.383
1.124

0.941
0.028
0.017
0.261

9.2%
6.7%
0.6%
0.6%
0.6%

3.9%
2.9%
1%
0%
0%

1.390
1.073
0.360
0.252
0.252

0.164
0.283
0.719
0.801
0.801

Allergic reaction to carbapenem

maculopapular rash
drug fever and rash
facial oedema
pruritis

1156

Downloaded from http://jac.oxfordjournals.org/ by guest on February 18, 2013

Carbapenems are commonly avoided in patients with a

reported penicillin allergy on the basis of a potential crosshypersensitivity with penicillin, however, very few studies have
been conducted describing the incidence of cross-reactivity
between penicillin and carbapenems. One study reported the
incidence of cross-reactivity between penicillin and carbapenems
as being 47%.7 The cross-reactivity was determined by the reactivity to both penicillin and imipenem/cilastin skin tests. The
clinical relevance of this cross-reactivity, however, is not
known. A retrospective review in neutropenic bone marrow
transplant patients with reported penicillin allergy that received
imipenem/cilastatin demonstrated an overall incidence of imipenem/cilastatin allergy in 9.5% of these patients. The authors
concluded that the incidence of cross-hypersensitivity between
penicillin and imipenem/cilastatin might be lower than previously reported.8 A recent retrospective review determined an
incidence of carbapenem allergy in 11% of the penicillin allergic
patients.9 Reported incidence of carbapenem-associated hypersensitivity in the general population is estimated to be 13%.1,2
The objective of this retrospective study was to ascertain the
clinical safety of administering carbapenems, namely imipenem/cilastatin and meropenem, in patients with a history of penicillin allergy compared with administering carbapenems in
patients with no reported penicillin allergy to help clinicians
make better informed decisions regarding the choice of antibiotics for patients with a history of penicillin allergy.

Carbapenem use in penicillin allergic patients

practice where listed allergies are usually treated as fact
regardless of their actual nature.
Retrospective studies inherently have their limitations due to
the dependence on the quality of documentation in the medical
records. Penicillin allergy histories and descriptions of hypersensitivity reactions relied on the details documented in the medical
record. A prospective study would allow for more accurate and
complete assessment of the allergy history and allergic reaction
description. In this retrospective review, typically only one antibiotic was discontinued when a reaction was noted which
increased the likelihood of correctly identifying the causative
agent.

Conclusions
Very limited data are available concerning the incidence of
cross-hypersensitivity of carbapenems in patients with penicillin
allergies. A prescriber may commonly avoid prescribing a carbapenem and then may be unnecessarily left with very few alternatives. Selecting antibiotics in severely ill patients with a history
of penicillin allergy can be a very difficult task. Based on this
and other similar studies,8,9 the true incidence of cross-hypersensitivity between penicillin and the carbapenems may be lower
than that previously reported and less concerning than previously
thought. Carbapenem use may be reasonable for patients with
penicillin allergies if caution is exercised.

Discussion
Patients with a history of penicillin allergy had a 9.2% incidence
of a carbapenem allergic reaction compared with 3.9% of the
patients with no penicillin allergy. These results are not statistically significant. Only one of the patients with a history of anaphylaxis to penicillin had an allergic reaction to the carbapenem
and this patient developed a maculopapular rash. These results
are similar to two previously published retrospective reviews
that demonstrated an overall incidence of imipenem/cilastatin
allergy between 9.5% and 11%.8,9 The true incidence of crosshypersensitivity between penicillin and the carbapenems may
therefore be lower than what is reported in the product package
inserts which list 47%.1,2,7
Carbapenem use seems reasonable in penicillin allergic
patients if there is a valid indication for the antibiotic. A possible
trend towards a higher incidence rate of a carbapenem reaction
was observed in penicillin allergic patients with multiple antibiotic allergies most notably to cephalosporins. It may be prudent to use caution when prescribing carbapenems in patients
with multiple antibiotic allergies. Fifty-four percent of our penicillin allergic patients had unknown types of reactions due to
incomplete histories documented in the chart. There was a
potential that some of the reported penicillin allergic patients
actually were not truly allergic. This reflects actual clinical

References
1. Primaxin package insert. (2001). Merck & Co., Inc., Whitehouse
Station, NJ, USA.
2. Merrem package insert. (2002). AstraZeneca Pharmaceuticals,
Wilmington, DE, USA.
3. Salkind, A. R., Cuddy, P. G. & Foxworth, J. W. (2001). Is this
patient allergic to penicillin? Journal of the American Medical
Association 285, 2498505.
4. Kerr, J. R. (1994). Penicillin allergy: a study of the incidence as
reported by patients. British Journal of Clinical Practice 48, 5 7.
5. Oswald, N. T. A. (1983). Penicillin allergy: a suspect label. British
Medical Journal 287, 265 6.
6. Surtees, S. J., Stockton, M. G. & Gietzen, T. W. (1991). Allergy to
penicillin: fable or fact. British Medical Journal 302, 1051 2.
7. Saxon, A., Adelman, D. C., Patel, A. et al. (1988). Imipenem
cross-reactivity with penicillin in humans. Journal of Allergy and Clinical
Immunology 82, 2137.
8. McConnell, S. A., Penzak, S. R., Warmack, T. S. et al. (2000).
Incidence of imipenem hypersensitivity reactions in febrile neutropenic
bone marrow transplant patients with history of penicillin allergy.
Clinical Infectious Diseases 31, 15124.
9. Prescott, W. A., DePestel, D. D., Ellis, J. J. et al. (2004).
Incidence of carbapenem-associated allergic-type reactions among
patients with versus patients without a reported penicillin allergy.
Clinical Infectious Diseases 38, 11027.

1157

Downloaded from http://jac.oxfordjournals.org/ by guest on February 18, 2013

reaction. Among the patients with no reported penicillin allergy,

three developed a maculopapular rash and one developed a drug
fever in addition to a rash. Development of a hypersensitivity
reaction invariably led to the discontinuation of the carbapenem.
The reactions resolved upon discontinuation of the carbapenem.
Reactions occurred on an average of 3.6 days into treatment for
the penicillin allergic patients and 5.5 days for patients without
penicillin allergies.
Patients with multiple drug allergies including penicillin are
inherently more apt to have an allergic reaction to the carbapenem. This tendency was most notable in patients with both penicillin and cephalosporin allergies. Fifty-seven percent (4/7) of
the patients with both allergies also had an allergic reaction to
the carbapenem. Statistical significance could not be determined.
The nature of the previous penicillin allergy was reviewed to
compare differences or similarities with the carbapenem hypersensitivity reaction. Seven (12.3%) out of 57 patients who
reported rash as a reaction to penicillin also developed a reaction
to the carbapenem. Five of these patients developed a rash as a
reaction to the carbapenem, whereas one of them developed
facial oedema and lip swelling. Interestingly, only one out of the
10 patients who reported anaphylaxis with penicillin administration developed a reaction to the carbapenem. This patient
developed a maculopapular rash as a reaction to the
carbapenem.