Hypersensitivity to antibiotics in patients with cystic fibrosis

Journal of Cystic Fibrosis 13 (2014) 205 – 211 www.elsevier.com/locate/jcf

Original Article

Hypersensitivity to antibiotics in patients with cystic fibrosis☆ Jobst Fridolin Roehmel a,b,⁎, Carsten Schwarz a , Anne Mehl a,b , Philippe Stock a,b , Doris Staab a,b a

Division of Cystic Fibrosis, Pediatric Pneumology and Immunology, University Children's Hospital, Charité University Medicine, Augustenburger Platz 1, 13353 – Berlin, Germany b Division of Allergology, Pediatric Pneumology and Immunology, University Children’s Hospital, Charité University Medicine, Augustenburger Platz 1, 13353 – Berlin, Germany Received 27 May 2013; received in revised form 29 September 2013; accepted 1 October 2013 Available online 26 October 2013

Abstract Background: Hypersensitivity reactions to parenterally administered antibiotics (HRPA) are a substantial problem in managing CF. We conducted this observational study to assess their nature and frequency as well as risk factors. Methods: By reviewing medical records and conducting interviews, age, sex, FEV1, ΔF508-genotype, pseudomonal colonisation, allergy history, antibiotic exposure and HRPA were recorded. Results: Of 100 patients included in the study, 60 had ≥ 1 HRPA. Overall, 3205 antibiotic courses with 185 HRPA were ascertained. Changes in therapy followed 65% of HRPA. Eighty-four percent of severe HRPA occurred during days 1–4. Approximately 10% of treatment courses with cefepime and piperacillin/tazobactam caused HRPA. Years of pseudomonal colonisation and cumulative annual exposure were significant risk factors. Conclusions: During days 1–4 of antibiotic treatment patients are at elevated risk for HRPA. HRPA are drug-specific and dependent on cumulative annual exposure. Elucidation of HRPA's immunological mechanisms and development of diagnostic algorithms for clinical use are required. © 2013 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. Keywords: Drug hypersensitivity; Cystic fibrosis; Allergy; Antibiotic treatment; Antibiotics; Hypersensitivity; Aztreonam; Ceftazidime; Piperacillin; Tazobactam; Colistin; Drug Allergy

1. Introduction Due to chronic pulmonary infections patients with cystic fibrosis (CF) frequently have a very high exposure to intravenous antibiotics, most commonly, penicillins, cephalosporines and aminoglycosides. Clinical observations and several cohort studies [1–4] have demonstrated that hypersensitivity reactions to parenterally administered antibiotics (HRPA) are a substantial problem in the management of CF, particularly in patients with

☆ Notation of prior abstract publication/presentation: A poster with preliminary results was presented at the German Cystic Fibrosis Conference 11/2011 in Wuerzburg/Germany — not published. A poster was presented at the North American Cystic Fibrosis Conference 10/2012 in Orlando/USA and the abstract was published as a conference supplementary to Pediatric Pulmonology. A poster was presented at the German Pediatric Pulmonology Conference 3/2013 in Lübeck/Germany. ⁎ Corresponding author. E-mail address: [email protected] (J.F. Roehmel).

advanced lung disease. Clinicians are frequently confronted with HRPA in their day-to-day clinical practice [2,5]. Most common symptoms are pruritus and rash [1–4]. Since the underlying mechanisms remain mostly unclear, antibiotic treatment is frequently terminated, resulting in inconsistent anti-infectious treatment. Due to a substantial increase in the life-expectancy for patients with CF in the last decades [6] and therefore longer treatment periods, most of the published data about HRPA [1,3,4] do not reflect today's patients and treatment regimens. Here we present data of our recent observational study which assessed the nature and frequency of HRPA as well as predisposing risk factors. 2. Methods 2.1. Setting and study population Of 285 patients with CF (185 adults, 100 children) currently treated in our center, 100 of 125 pediatric and adult patients,

1569-1993/$ -see front matter © 2013 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jcf.2013.10.002

206

J.F. Roehmel et al. / Journal of Cystic Fibrosis 13 (2013) 205–211

which were identified to have previously received at least four intravenous antibiotic treatment courses, consented. We restricted our observations to parenterally administrated antibiotics because this route of administration is most commonly associated with hypersensitivity reactions [1]. In contrast to prior publications [2–4] which focused on β-lactam antibiotics only, we also included other classes of intravenous (iv) antibiotics. This study was approved by the ethics committee of Charité — University Medicine Berlin. 2.2. Study design After providing written informed consent, patients were interviewed with a self-designed, semi-open structured questionnaire asking for treatment centers, FEV1 predicted (FEV1), CFTR-status, age when diagnosed CF, history of allergy and atopy, and history of iv antibiotics and HRPA. Then medical records were reviewed to gather and verify information about: HRPA, age, sex, latest FEV1, ΔF508 genotype, allergy history including the latest results of a standard IgE screening test for aeroallergens (an identical assay was used), specific IgE against aspergillus, past allergic bronchopulmonary aspergillosis (ABPA), latest total IgE serum level and years of chronic Pseudomonas aeruginosa (PA)-colonisation (defined as ≥ 2 detections/a). To distinguish patients according to the completeness of documentation in their medical records, we created three groups: (i) the total cohort termed “total” comprised of 100 patients who were currently under treatment at our center, who met the inclusion criteria and had participated in a face-to-face interview; (ii) out of those we defined the group “5-year-doc” (N = 66) that was comprised of patients who had at least 5 years of continuous medical records at our institution; (iii) out of this latter group we formed the group “lifetime-doc” (N = 50) that was comprised of patients whose lifetime medical documentation was available and done at our centre. For the groups “5-year-doc” and “lifetimedoc,” we documented every available parenteral antibiotic treatment course, including the type of antibiotic drug, the length of treatment course, daily dose and the use of pre-emptive measures such as rapid desensitisation [5,7] or administration of anthistamines or corticosteroids. We defined HRPA as any form of hypersensitivity reaction occurring during an iv antibiotic treatment course which required a subsequent medical intervention, such as discontinuation or change of antibiotic treatment or administration of antihistamines and corticosteroids. It was further required that previous CF treatment remained unchanged. No preselection of symptoms in the sense of allergy symptoms was done. Whenever we could not identifiy the HRPA-causing antibiotic due to multiple antibiotic treatment the reaction was excluded from the observations. Criteria for an antibiotic to be accountable were the onset of the reaction in relation to the infusion and tolerance to the second agent after discontinuation of the first (in 69% tobramycine was one of the two agents). HRPA were documented in detail including the date of onset and the day of the antibiotic treatment course, the onset of the reaction in relation to administration of the infusion, type of symptoms, treatment of the reaction and continuation

(with antiallergic medication), and change or discontinuation of antibiotic treatment. As the risk of hypersensitivity to an antibiotic we regarded the proportion of the recorded antibiotic courses of a given drug which featured an HRPA. Home iv antibiotic treatments were included whenever sufficient documentation was available. Reactions involving the bronchopulmonary (e.g. acute bronchospasm) or the cardiovascular (e.g. hypotension) system were classified as severe, reactions meeting the WAO criteria for anaphylaxis [8] as anphylactic. With regard to their onset, reactions were classified according to the “European Network of Drug Allergy” [9]: those during or within 1 hour of the infusion were classified as “immediate,” all others as “non-immediate.” Following Hjortlund et al. [10,11] we also classified reactions as immediate which occurred within treatment courses (days 1–14). 2.3. Statistical analysis Descriptive analyses in the three (nested) patient populations were conducted based on the extent of data available for each group. Student's t-tests for continuously distributed variables and χ2 tests for proportions were calculated with corresponding 95% confidence intervals and p-values. Univariate analyses were performed to identify risk factors for HRPA. For studying the influence of multiple risk factors on the occurrence of HRPA, a logistic regression model was developed in the “lifetime-doc” cohort with HRPA as the outcome variable and age, gender, years of parenteral antibiotic treatment, FEV1, and years of PA-colonisation as independent variables. Another logistic regression model analysed these risk factors for severe reactions only. Cumulative exposure (gram/year) and lifetime dose to the two most HRPA-causing antibiotics were analysed in the “lifetimedoc” group comparing the subgroups with and without reactions. Due to obvious deviation from the normal distribution, we utilized the nonparametric Mann–Whitney–Wilcoxon test (MWW) for estimating and testing the difference between the two groups (with associated confidence intervals) [12]. For all statistical analysis we used “R” (The R Project for Statistical Computing, www.r-project.org). For figures we utilized “R” package “ggplot2” [13]. 3. Results Of 125 the patients who met the inclusion criteria, 100 (80%) agreed to participate in the study (Table 1). 3.1. Prevalence of reactions (“total,” N = 100) We observed 185 HRPA with 376 symptoms in 60 patients (an average of 3.1 reactions per affected patient). Sixty-five percent of the reactions led to a change or discontinuation of antibiotic treatment. In patients with HRPA 46% had ≥ two and 31% had ≥ three reactions. For pediatric patients (N = 8) we documented four reactions in three patients. Analysing the classes of antibiotics, 79% of all reactions occurred after administration of β-lactam antibiotics. Penicillins accounted

J.F. Roehmel et al. / Journal of Cystic Fibrosis 13 (2013) 205–211

207

Table 1 Patient characteristics in the different groups. Group:

“Total”

“5-Year-doc”

“Lifetime-doc”

“Pediatric”

N patients Age Ø years Sex % male FEV1 Ø % N courses Ø PA-time in years Atopy screen + % ≥ 1 HRPA in %

100 27.9 SD 10.8 38 52.3 SD 25.3 NA 16.1 SD 6.6 29 60

66 28.5 SD 11.6 36 54.4 SD 26.1 NA 17.4 SD 6.0 29 65

50 28.1 SD 13 39 55 SD 27.1 50.4 SD 31.5 16.7 SD 6.3 32 65

8 9 SD 5.1 50 71.7 SD 39.2 21.5 SD 14.4 5.8 SD 7.7 0 38

FEV1 = latest FEV1 in %. PA-time years = Pseudomas aeruginosa (PA) colonisation time in number of years. Atopy screen + = screening tests for a common panel of aeroallergens which were positive. N courses = number of intravenous antibiotic treatment courses.

for 35.1%, cephalosporins for 34.6%, carbapenems for 9.6%, quinolones for 4.3% and polypeptides for 4.3% of the reactions.

highest risk for anaphylactic HRPA had aztreonam (6%) and cefuroxime (6%).

3.2. Risk of antibiotics to cause hypersensitivity reactions (“5-year-doc,” N = 66)

3.3. Characteristics of HRPA (”total,” N = 100) The broadly varying symptoms were grouped into eight entities: 53% were dermal (34% pruritus, 18% erythema and flushing and 17.5% urticaria), 10% drug fever, 9% bronchospasm, 9% neurological symptoms (49% dizziness, 28% paresthesia, 26% headache), 7% gastrointestinal (60% nausea, 28% vomiting), 4% cardiovascular (75% tachycardia, 18% hypotension), and 7% others. Analysed for double combinations of symptom entities which formed HRPA, dermal and bronchopulmonary with a proportion of 9% was most frequent, followed by dermal

In 3205 antibiotic courses analysed, the two antibiotics with the highest absolute number of HRPA were piperacillin/ tazobactam (N = 63) and ceftazidime (N = 50), both also accounted for the most anaphylactic HRPA (piperacillin/ tazobactam N = 9, ceftazidime N = 7). The risk among all evaluated antibiotics was highest for vancomycin (12%). Among antipseudomonal drugs the risks for HRPA were highest for cefepime (11%) and piperacillin/tazobactam (10%) (Fig. 1). The

8

*der 100% *der 43%, car 24%, fev 18%

95 390

*der 45%, pul 14%, fev 13%

24

*der 29%, car 29%, gas 29%

36 94 17 18 770 94

*der 34%, pul 17%, neu 17% *der 40%, pul 30%, gas 10%

*der 100% *der 67%, gas 33% *der 48%, neu 18%, pul 13% *der 20%, pul 20%, gas 20%

342 *der 22%, gas 22%, fev 17% 35

*der 33%, pul 33%, oth 33%

41 *der 83%, neu 17% 1110 *der 38%, gas 25%, fev 13%

% Proportion of treatment courses which lead to a reaction Fig. 1. Risk of HRPA and severe HRPA according to antibiotic treatment courses. Group “5-year-doc” (N = 66). Proportion of antibiotic treatment courses which led to a reaction (black bar = severe reactions) in percent. Severe reactions were defined as bronchopulmonary or cardiovascular reactions. Numbers inside of the bars stand for total number of recorded single antibiotic courses. The information right to the bars (with*) demonstrates the three most commonly observed symptom entities for the specific drugs in percent (der = dermal, pul = bronchopulmonal, neu = neurologic, gas = gastrointestinal, car = cardiovascular,fev = drug fever, oth = others).

208

J.F. Roehmel et al. / Journal of Cystic Fibrosis 13 (2013) 205–211

and neurologic with 6.4% and dermal and fever with 4.8%. We classified 24% of the HRPA as severe and 15% as anaphylactic. No HRPA required ICU care. 3.4. Timing of reactions (”total,” N = 100) In relation to the infusion 53% of HRPA occurred while administering the antibiotic, 35% within 1 hour and 12% more than 1 hour after infusion regardless of the day of the treatment course. In relation to the day of the treatment course 81% of all, 84% of severe and 80% of anaphylactic HRPA occurred between days 1 and 4 of a treatment course (for combined analysis see Fig. 2). 3.5. Risk factors for HRPA (“lifetime-doc,” N = 50) In univariate analyses we identified the following significant risk factors for an individual to develop HRPA: reduced FEV1, duration of PA-colonisation and the number of years of iv antibiotic treatment. In contrast, ΔF508-genotype, sex and markers for atopy such as screening tests for a common panel of aeroallergens and total IgE concentration were not significant risk factors. The total number of antibiotic courses and dichotomous variables like “b 20 treatment courses” were not significantly different between patients with and without HRPA. While for any type of “reaction” the multivariable

regression model did not result in significant factors except for the variable years of PA-colonisation, the results were more pronounced for “severe reactions” (Table 2). The following factors were significant predictors of severe HRPA: female sex, reduced FEV1, and years of PA-colonisation. The addition of the variable “number of antibiotic courses” did not improve the estimation in the regression analyses. 3.6. Cumulative dose as a risk factor (“lifetime-doc,” N = 50) Patients who experienced an HRPA did not differ significantly in the lifetime dose of the investigated HRPA-causing drugs from those who did not experience an HRPA. Yet the yearly cumulative dose of this drug was significantly higher in patients with reactions than in those without reactions (Table 3). 3.7. Onset of HRPA with respect to initiation of antibiotic treatment (”lifetime-doc,” N = 50) Mean age at the beginning of iv treatment was 12.9 (SD 11.6) years. For patients with at least one HRPA (N = 31) mean age at the beginning of iv treatment was 12.9 (SD 10.3) years and mean time from the first exposure to iv antibiotics to the first reaction was 8.4 (SD 5.6) years. For patients with at least two HRPA (N = 20) mean age was 13.7 (SD 8.18) years at first iv treatment and the mean time from first exposure to iv antibiotics to the first

Fig. 2. Proportion of HRPA related to the day of the treatment course. Divided into immediate (grey) and non-immediate (black) reactions. Analysed were 185 HRPA (group “total,” N = 100) for all HRPA, severe HRPA and anaphylactic HRPA.

J.F. Roehmel et al. / Journal of Cystic Fibrosis 13 (2013) 205–211 Table 2 Results of the logistic regression analyses identifying potential risk factors for HRPA and severe HRPA (group “lifetime-doc,” N = 50). Variable

p-Value reactions

p-Value severe reactions

Years of PA colonisation Reduced FEV1 Female gender Increasing age (continuous) Years of antibiotic treatment Number of courses Atopy screen +

0.0399 0.5947 0.1023 0.2848 0.9130 0.4621 0.6791

0.0324 0.0259 0.0375 0.3376 0.7301 0.4801 0.7501

inclusion criteria and definitions of HRPA were different among these reported studies. Even in a direct comparison with Koch's study [4], our patients experienced nearly twice as many reactions (3.1 vs. 1.68) per capita. This underlines that HRPA are an increasing problem in the management of advanced CF. 4.1. Timing of reactions

Atopy screen + = screening tests for a common panel of aeroallergens which were positive. Significant results are displayed in bold.

reaction was 7.1 (SD 5.08) years. The mean time between the first and second reaction was 2.2 (SD 2.68) years. The time from first antibiotic treatment to the first reaction was significantly longer (p b 0.001) than the time from the first to the second reaction. Having experienced an HRPA to one drug is a risk factor for developing HRPA against a second drug. 3.8. Therapeutic strategies Out of the recorded 3205 single drug courses 34 were done with pre-emptive administration of oral antihistamines and/or corticosteroids (N = 20) or as a rapid desensitisation [5] (N = 14) to deal with prior HRPA. Out of those 9/34 (4/20; 5/14) courses led to another reaction. 4. Discussion HRPA in CF patients are highly relevant from a clinical perspective. Patients with advanced CF and an urgent need for repeated intravenous antibiotic treatment suffer the most from HRPA, since antibiotic discontinuations and changes of treatment the anti-infective therapy might be insufficient. HRPA in CF patients appears to be more frequent than in the general population [14]. In our study population the prevalence of at least one HRPA was 60% and therefore higher than in most previous CF studies analysing antibiotic drug hypersensitivity: 29% in Pleasants et al. [3], 33% in Wills et al. [1], and 36% in Burrows et al. [2]. Koch et al. [4] showed a similar prevalence to our study with 61%. The Table 3 Median cumulative dose per year and life time dose of specific antibiotics as risk factors. Exposed patients a

Patients

N

Median lifetime dose (g)

p

Median cumulative dose (g)/year

p

Ceftazidime N = 48 Piperacillin/taz N = 43

w/ HRPA w/o HRPA w/ HRPA w/o HRPA

18 30 22 21

876 860 366 702

0.92

106 62 235 107

0.012

0.063

209

In our study, 81% of all, 84% of severe and 80% of anaphylactic reactions occurred during the first 4 days of a treatment course. These results are of importance for home parenteral antibiotic treatment. Hospitalisation at start of antibiotic treatment in patients at risk might be considered to improve safety since reactions frequently required immediate treatment. The high magnitude of HRPA on the first day during and within 1 hour after antibiotic administration suggests a significant proportion of IgE-mediated reactions. As published before [1–4] typical IgE mediated skin symptoms such as urticaria and pruritus were most common. Several other immunological or other mechanisms may also cause the spectrum of symptoms we observed with varying times of onset in the treatment course. A non-immunological mechanism may be a direct histamine liberation, called pseudoallergy [15], which has been described for vancomycin (“red-man-syndrome”) and ciprofloxacin. Another symptom was drug fever with a proportion of 10% and a trend to occur later in the course, as previosly published [1–4]. Later in the treatment course we observed fewer reactions with a larger variety of symptoms in the sense of cellular delayed hypersensitivity reactions [9,16,17]. Possibly a proportion of the reactions is due to unspecific adverse effects rather than to specific immunonological mechanisms with the risk for reoccurrence. Our finding that the time from the first to the second reaction was significantly decreased has also been described for penicillin reactions in the general population [18]. 4.2. Risk of antibiotics to cause hypersensitivity reactions As demonstrated, we observed a moderate risk of HRPA for ceftazidime and higher risks for piperacillin/tazobactam and cefepime among the antipseudomonal antibiotics. In our study, the risk of any reaction was almost as high for aztreonam as that for piperacillin/tazobactam. Aztreonam additionally had the highest risk for development of severe HRPA with approximately 6% of all treatment courses leading to a severe reaction. This was unexpected and is in contrast to previous observations [2,4,19]. Remarkably all patients with HRPA to aztreonam had previously experienced HRPA to ceftazidime [20]. 4.3. Risk factors for HRPA

0.011

Piperacillin/taz = piperacillin/tazobactam; w/ = with, w/o = without. Median cumulative dose/year = median cumulative dose until the first HRPA or until the end of the observation. a Analysed for the two most frequently HRPA-causing antipseudomonal antibiotics.

Our study confirmed in part previous findings [2,21] that ΔF508 genotype and markers for atopy had no significant influence on the development of HRPA. Contrary to prior publications [2,4], neither age nor the total number of antibiotic treatment courses were significant risk factors. We found the period of time with PA-colonisation to be a significant predictor for the development of HRPA. The period of time with PA-

210

J.F. Roehmel et al. / Journal of Cystic Fibrosis 13 (2013) 205–211

colonisation, female sex and low FEV1 were risk factors for severe HRPA. Reduced FEV1 and years of PA-colonisation both might reflect the load of antibiotic treatment per time. Female sex as a risk factor could be due to the “gender gap” phenomenon [22,23]. Since in women the disease progresses more rapidly, they probably have more exposure to antibiotics. In contrast to a previous publication [2], the total number of treatment courses and the total lifetime dose of specific antibiotics had no significant impact on the risk of HRPA. However, for all evaluated antibiotics, the cumulative dose of a given antibiotic per time prior to the reaction was significantly higher among patients with compared to those without reactions. We focused on the cumulative dose rather than on the number of treatment courses because antibiotic courses vary broadly in length and dose. Our observation may be in line with a previously made suggestion [24] indicating that repeated specific exposure may increase the risk of HRPA. Our results suggest that HRPA may occur due to a specific, cumulative-dose dependent sensitisation. Unfortunately, we have yet to define the underlying immune reactions leading to different types and time points of the reaction. Some strengths and limitations of our study need to be considered. We were able to record complete data about lifetime antibiotic exposure and HRPA in 50 patients who had been treated at our center only, amounting to a follow-up period of up to 25 years in individual patients. It would not be feasible to ascertain antibiotic exposure during such a long follow-up period in a prospective clinical study. Additionally all recalled HRPA were cross-checked in the medical records. In patients without lifetime documentation at our center, we had to rely on interview data in order to ascertain previous (not documented) antibiotic exposure and HRPA. In these patients, recall bias might be of concern due to the retrospective nature of our study. We therefore classified our study population according to the completeness of the information in the medical records, analysing lifetime exposure only in the group of patients for whom full documentation was available. For the analysis of lifetime antibiotic exposure, this reduced our sample size by half. Therefore for the analysis of cumulative exposure for specific antibiotics as a risk factor only ceftazidime and piperacillin/ tazobactam reached evaluable numbers. Just eight pediatric patients were included due to less iv-antibiotic treatments in this group nowadays. HRPA is a diffuse clinical phenomenon with a broad spectrum of symptoms and a range of possible mechanisms. Facing the difficulty to distinguish between hypersensitivity reactions and unspecific side effects, we included all symptoms, which led to patients being labeled as hypersensitive to an antibiotic. The majority of the reported symptoms are listed in: “WAO guidelines for anaphylaxis.” [8]. We took into account that side effects might be included. This is a limitation since we may have overestimated HRPA's frequency. Otherwise It may also be a strength to gather data without fixed definitions. Retrospectively it was not possible to define HRPA by specific symptoms, since these may not be fully documented. We therefore defined it by specific changes to treatment further supporting the diagnosis of HRPA.

In summary, the high frequency of HRPA in our study is of considerable concern, given the consecutive limitations of antibiotic treatment options for future therapy of exacerbations caused by multiresistant gram-negative bacteria [25]. A possible strategy for prevention of HRPA, which beforehand definitely requires further confirmation of our findings, could be to rotate the antipseudomonal agents rather than to repeat the same regimen which has been well tolerated in past treatments, of course while keeping the microbiological requirements in mind, in order to reduce the specific cumulative dose. As clinical solution several desensitisation protocols have been shown to be effective [5,7,26], in our observations only 65% were successful. In order to avoid over-diagnosis of HRPA, diagnostic algorithms for daily clinical use should be developed. Besides recent publications [27–30] about drug specific lymphocytes in CF patients, we believe that further elucidation of HRPA's immunological mechanisms, also concerning immediate reactions and diagnostics, is needed. The specific cumulative dose per year which we observed as a strong predictor for the development of HRPA requires confirmation in further studies. Competing interest statement The authors declare to hold no conflict of interest with the publication of the results included in this article. Funding This study received no funding. Acknowledgements We thank study subjects and their families for their participation. We also thank both the nurses team on the ward and in the outpatient clinic at our center for their patience and help. References [1] Wills R, Henry RL, Francis JL. Antibiotic hypersensitivity reactions in cystic fibrosis. J Paediatr Child Health Aug 1998;34(4):325–9. [2] Burrows JA, Nissen LM, Kirkpatrick CMJ, Bell SC. Beta-lactam allergy in adults with cystic fibrosis. J Cyst Fibros Jul 2007;6(4):297–303. [3] Pleasants RA, Walker TR, Samuelson WM. Allergic reactions to parenteral beta-lactam antibiotics in patients with cystic fibrosis. Chest Oct 1 1994;106(4):1124–8. [4] Koch C, Hjelt K, Pedersen SS, Jensen ET, Jensen T, Lanng S, et al. Retrospective clinical study of hypersensitivity reactions to aztreonam and six other beta-lactam antibiotics in cystic fibrosis patients receiving multiple treatment courses. Rev Infect Dis Apr 1991;13(Suppl. 7):S608–11. [5] Legere III HJ, Palis RI, Bouza TR, Uluer AZ, Castells MC. A safe protocol for rapid desensitization in patients with cystic fibrosis and antibiotic hypersensitivity. J. Cyst. Fibros., 8(6). Elsevier B.V.; Dec 1 2009. p. 418–24. [6] Parkins MD, Parkins VM, Rendall JC, Elborn S. Changing epidemiology and clinical issues arising in an ageing cystic fibrosis population. Ther Adv Respir Dis Apr 2011;5(2):105–19. [7] Whitaker P, Shaw N, Gooi J, Etherington C, Conway S, Peckham D. Rapid desensitization for non-immediate reactions in patients with cystic fibrosis. J Cyst Fibros Jul 2011;10(4):282–5.

J.F. Roehmel et al. / Journal of Cystic Fibrosis 13 (2013) 205–211 [8] Simons FER, Ardusso LRF, Bilò MB, El-Gamal YM, Ledford DK, Ring J, et al. World Allergy Organization anaphylaxis guidelines: summary. J Allergy Clin Immunol Mar 2011;127(3):587–93 [e1–22]. [9] Romano A, Blanca M, Torres MJ, Bircher A, Aberer W, Brockow K, et al. Diagnosis of nonimmediate reactions to beta-lactam antibiotics. Allergy Nov 2004;59(11):1153–60. [10] Hjortlund J, Mortz CG, Skov PS, Eller E, Poulsen JM, Borch JE, et al. One-week oral challenge with penicillin in diagnosis of penicillin allergy. Acta Derm Venereol May 2012;92(3):307–12. [11] Hjortlund J, Mortz CG, Skov PS, Bindslev-Jensen C. Diagnosis of penicillin allergy revisited: the value of case history, skin testing, specific IgE and prolonged challenge. Allergy Aug 2013;68(8):1057–64. [12] Hollander M, Wolfe DA. Nonparametric Statistical Methods. 2nd ed. Wiley-Interscience; 1999. [13] Wickham H. ggplot2. Springer; 2009. [14] Torres MJ, Blanca M. The complex clinical picture of beta-lactam hypersensitivity: penicillins, cephalosporins, monobactams, carbapenems, and clavams. Med Clin North Am Jul 2010;94(4):805–20 [xii]. [15] Bircher AJ. Drug-induced urticaria and angioedema caused by non-IgE mediated pathomechanisms. Eur J Dermatol Dec 1999;9(8):657–63 [quiz663]. [16] Pichler W. Delayed drug hypersensitivity reactions. Ann Intern Med Sep 24 2003;139:683–93. [17] Whitaker P, Naisbitt D, Peckham D. Nonimmediate β-lactam reactions in patients with cystic fibrosis. Curr Opin Allergy Clin Immunol Aug 2012;12(4):369–75. [18] Sogn DD. Prevention of allergic reactions to penicillin. J Allergy Clin Immunol Nov 1986;78(5 Pt 2):1051–2. [19] Jensen T, Koch C, Pedersen SS, Høiby N. Aztreonam for cystic fibrosis patients who are hypersensitive to other beta-lactams. Lancet Jun 6 1987;1(8545):1319–20. [20] Pérez Pimiento A, Gómez Martínez M, Mínguez Mena A, Trampal González A, de Paz Arranz S, Rodríguez Mosquera M. Aztreonam

[21] [22] [23]

[24] [25]

[26] [27]

[28]

[29]

[30]

211

and ceftazidime: evidence of in vivo cross allergenicity. Allergy Jun 1998;53(6):624–5. Parmar JS. Antibiotic allergy in cystic fibrosis. Thorax Jun 1 2005;60(6): 517–20. Davis PB. The gender gap in cystic fibrosis survival. J Gend Specif Med Mar 1999;2(2):47–51. Chotirmall SH, Greene CM, Oglesby IK, Thomas W, O'Neill SJ, Harvey BJ, et al. 17Beta-estradiol inhibits IL-8 in cystic fibrosis by up-regulating secretory leucoprotease inhibitor. Am J Respir Crit Care Med Jul 1 2010;182(1):62–72. Ramesh S. Antibiotic hypersensitivity in patients with CF. Clin Rev Allergy Immunol Aug 2002;23(1):123–41. Ren CL, Konstan MW, Yegin A, Rasouliyan L, Trzaskoma B, Morgan WJ, et al. Multiple antibiotic-resistant Pseudomonas aeruginosa and lung function decline in patients with cystic fibrosis. J Cyst Fibros Jul 2012;11(4):293–9. Burrows JA, Toon M, Bell SC. Antibiotic desensitization in adults with cystic fibrosis. Respirology Sep 2003;8(3):359–64. Jenkins RE, Yaseen FS, Monshi MM, Whitaker P, Meng X, Farrell J, et al. β-Lactam antibiotics form distinct haptenic structures on albumin and activate drug-specific T-lymphocyte responses in multiallergic patients with cystic fibrosis. Chem Res Toxicol Jun 17 2013;26(6):963–75. El-Ghaiesh S, Monshi MM, Whitaker P, Jenkins R, Meng X, Farrell J, et al. Characterization of the antigen specificity of T-cell clones from piperacillin-hypersensitive patients with cystic fibrosis. J Pharmacol Exp Ther Jun 2012;341(3):597–610. Whitaker P, Meng X, Lavergne SN, El-Ghaiesh S, Monshi M, Earnshaw C, et al. Mass spectrometric characterization of circulating and functional antigens derived from piperacillin in patients with cystic fibrosis. J Immunol Jul 1 2011;187(1):200–11. Elsheikh A, Castrejon L, Lavergne SN, Whitaker P, Monshi M, Callan H, et al. Enhanced antigenicity leads to altered immunogenicity in sulfamethoxazole-hypersensitive patients with cystic fibrosis. J Allergy Clin Immunol Jun 2011;127(6):1543–51.