Among the several hundred workplace agents implicated as causes of occupational asthma (OA), several are encountered in the pharmaceutical manufacturing industry [
The first case of occupational antibiotic allergy was described in 1953, but a relatively small number of individual case reports have been published subsequently. Very large quantities of antibiotics are produced in almost every country but the frequency of OA in those who manufacture them remains unclear. While prevalence estimates of around 10% have been reported [
New cases were seen and diagnosed at Royal Brompton Hospital in London, UK between 1995 and 2009. Specific IgE measurements, where attempted, were carried out by either the radioallergosorbent test (RAST) method or the commercial ImmunoCAP assay. Each patient underwent controlled, single-blind, specific provocation testing using a dust-tipping method, with small quantities of the relevant antibiotic powder mixed with a larger amount of dried lactose; active and control exposures were carried out on sequential days with the patients having stopped any asthma medications prior to the tests. Responses were assessed using serial FEV1 measurements following challenge and by changes in bronchial responsiveness to inhaled histamine using the Yan technique [
We were not able to identify any previous systematic review of this subject. We searched the published literature using the Medline database between 1953 and February 2010. Both key word- and text word-based searches were performed with combinations of the terms “occupational asthma”, “asthma”, “respiratory sensitisation”, and “antibiotics”. In addition, we examined the reference lists of relevant articles. Both case reports (
We used a quantitative structure-activity relationship (qSAR) model to examine the potential for each published antibiotic to act as a respiratory sensitising agent. “Hazard indices” for each were calculated, where possible, using the Chemical Asthma Hazard Assessment Program [
Cases of occupational asthma from antibiotics identified at Royal Brompton Hospital in the period 1995–2009.
case | Year of diagnosis | Workplace exposure | Allergic symptoms | Latency | Specific IgE | Bronchial provocation test | ||
Agent | FEV1 response | Increase in histamine reactivity | ||||||
a | 1995 | penicillin | wheeze | 19 years | not done | penicillin | late | Yes |
b | 1996 | amoxicillin | wheeze, cough | 27 years | penicilloyl G (+) | amoxicillin | late | yes |
c | 2000 | amoxicillin | wheeze, cough | 27 years | amoxicilloyl (+) | amoxicillin | late | yes |
d | 2009 | erythromycin | wheeze, rhinitis | 2 years | erythromycin ethylsuccinate (−) | erythromycin ethylsuccinate | late | Yes |
Changes in FEV1 and histamine reactivity following bronchial provocation testing in four antibiotic manufacturing workers. Serial FEV1 measurements (
Each of the four cases was involved in the primary manufacture or formulation of antibiotics in the United Kingdom and had presented with new onset, work-related asthmatic symptoms. Three worked with penicillins and their derivatives, the fourth with a variety of medicines including erythromycin. This last, a woman of 52, developed asthma 22 years previously, two years after starting work on the packaging lines. She reported that her symptoms worsened with exposure to granulated erythromycin and on two occasions in the year prior to referral had required treatment with oral corticosteroids. At referral, she had normal spirometry, was demonstrated to be atopic on skin prick testing, and completed a series of peak flow measurements which showed significant variability on both work and rest days, with no clear work-related pattern.
All four patients underwent single-blind, controlled specific provocation testing using a dust-tipping method. Each of the four developed a (predominantly) late asthmatic reaction (Figure
Published cases of occupational allergy to antibiotics 1953–2009.
Penicillins | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
Reference | Year | Country | No. | Exposure | Latency | Respiratory symptoms | Skin test | Specific IgE | Bronchial provocation test | Oral challenge |
[ | 1953 | USA | 2 | penicillin | #1: “weeks” | #1: cough, rhinitis | Procaine penicillin (+) | not reported | not reported | not reported |
[ | 1957 | France | 4 | penicillin | #1: “years” | #1: cough, rhinitis, dyspnoea | not reported | not reported | not reported | not reported |
[ | 1960 | France | 2 | penicillin | #1: 1 year | #1: asthma, urticaria | #1: PMP* (+) | not reported | not reported | not reported |
[ | 1974 | UK | 3 | ampicillin | #1: 2 years | asthma ( | ampicillin (−) | not reported | #1: ampicillin (+LR*)6APA* (+/−) | #1: ampicillin (−) |
[ | 1980 | Spain | 2 | amoxicillin | #1: 1 year | #1: rhinitis, dyspnoea, wheeze | not reported | negative | not reported | not reported |
[ | 1982 | Germany | 1 | ampicillin | NS* | cough, rhinitis, dyspnoea, fever | ampicillin (−) | BPP* (+) | antibiotic mix (+LR*) | not reported |
[ | 1997 | Belgium | 1 | amoxicillin | 6 months | cough, wheeze, rhinitis | not reported | not reported | amoxicillin (+ER* LR*) | not reported |
[ | 1998 | Spain | 1 | amoxicillin | 27 years | cough, rhinitis, wheeze, dyspnoea | amoxicillin (−) | amoxicillin (+) | amoxicillin: (+ER*) | amoxicillin (+LR*) |
Cephalosporins | ||||||||||
Reference | Year | Country | No. | Exposure | Latency | Allergic symptoms | Skin test | Specific IgE | Bronchial provocation test | Oral challenge |
[ | 1980 | Spain | 1 | cephalexin, | 3 months | cough, wheeze | PP* (−) | negative | not reported | cephalosporin (NS) (+) and rhinitis, urticaria |
[ | 1981 | UK | 2 | 7ACA* | #1: NS* | #1: cough, rhinitis, chest tightness | #1: 7ACA (+) | not reported | #1: 7ACA (+ER*) | not reported |
[ | 1995 | UK | 1 | ceftazidime | 1 year | rhinitis, dyspnoea | not reported | not reported | ceftazidime (+ER* LR*) | not reported |
[ | 1996 | Italy | 1 | cefmetazole | 1 year | cough, rhinitis, bronchospasm | cefmetazole (−) | penicillin G (−) | cefmetazole (+NS*) | not reported |
[ | 1999 | Spain | 1 | cefadroxil | 9 months | cough, rhinitis, dyspnoea, chest tightness | PP* (−) | penicillin G (−) | cefadroxil (+ER*) | amoxicillin (−) |
[ | 2003 | Korea | 2 | cefteram | NS* | NS* | #1: cefteram (+) | #1: cefteram-HSA* (+) | #1: cefteram (+ER*) | not reported |
[ | 2004 | Korea | 2 | 7-ACA* | 2 years | #1: rhinitis, respiratory symptoms | #1: 7-ACA* (+) | #1: 7-ACA-HSA* (+) | #1: 7-ACA* (+ER*) | not reported |
[ | 2009 | Italy | 1 | 7-TACA* | 8 months | cough, rhinitis, dyspnoea | not reported | not reported | 7-TACA*: (+ER*) | not reported |
Miscellaneous | ||||||||||
Reference | Year | Country | No. | Exposure | Latency | Allergic symptoms | Skin test | Specific IgE | Bronchial provocation test | Oral challenge |
[ | 1977 | India | 1 | tetracycline | 1 year | cough, wheeze, dyspnoea | not reported | not reported | tetracycline (+ER*) | tetracycline (+ER*) and urticaria |
[ | 1975 | UK | 1 | spiramycin | 1 year | cough, rhinitis, dyspnoea, dermatitis | spiramycin (+) | not reported | spiramycin (+LR*) | not reported |
[ | 1979 | Italy | 1 | spiramycin | 1 year | cough, asthma, dermatitis | spiramycin (+) | not reported | chick feed with spiramycin (+LR*) | not reported |
[ | 1984 | Italy | 2 | spiramycin | #1:14 years | #1: dyspnoea | not reported | not reported | #1: spiramycin adipate (+ER* LR*) | not reported |
[ | 1995 | Italy | 1 | piperacillin | 22 months | rhinitis, dyspnoea, wheeze, rash | piperacillin (+) | not reported | piperacillin (+ER*) | not reported |
[ | 2006 | Korea | 2 | thiamphenicol | NS* | #1: rhinitis, asthma | #1: thiamphenicol (+) | #1: thiamphenicol (+) | #1: thiamphenicol (+ER*) | not reported |
[ | 2009 | Korea | 1 | vancomycin | 5 months | rhinitis, chest tightness | vancomycin (−) | vancomycin-HSA* (−) | not reported | not reported |
[ | 2010 | Spain | 1 | colomycin | 3 months | rhinitis, cough, wheeze, dyspnoea | not reported | negative | colomycin (+ER*) | not reported |
*NS: not specified, *ER: early (asthmatic) response, *LR: late (asthmatic) response.
*HSA: human serum albumin, *MDM: minor determinant (penicillin) mix, *BP: benzylpenicillin.
*(B)PP: (benzyl)penicilloyl polylysine, *PMP: phenoxymethyl penicillin, *6APA: 6 amino penicillanic acid, *7-ACA: 7aminocephalosporanic acid, *7CTD: tosylate dihydrate derivative of 7ACA.
*7-TACA: 7-amino-3thiometihyl-3-cephalosporanic acid.
Previous published case reports, including a total of 37 patients, are summarised in Table
Three reports describe patients sensitised to penicillins. In the first, published in 1953, Eaton Roberts described two employees in a US factory with clinical evidence of OA ascribed to penicillin; apart from skin prick testing with procaine penicillin (negative), he did not perform any objective investigations [
Synthetic penicillins have also been reported as a cause of work-related asthma. Davies et al. [
There are several reports of cephalosporins and associated precursors and derivatives as causative agents for OA; 11 cases are summarised in Table
The remaining eight reported cases developed OA during the manufacture of antibiotics other than penicillins or cephalosporins. Most have been in workers exposed to the macrolide spiramycin, an antibiotic widely used in the livestock industry and particularly with poultry so that traces may be found in some chicken eggs. Davies and Pepys [
Menon and Das [
We identified seven reports of epidemiological studies carried out in antibiotic manufacturing sites; all but one was of cross-sectional design. In several cases, the absence of detailed information on the size of the exposed populations precludes any estimate of disease prevalence.
Briatico-Vangosta et al. [
Chida and Uehata [
Phenylglycine acid chloride (PG-AC) is a highly reactive compound used chiefly in the manufacture of ampicillin and other antibiotic side chains. A survey of 24 workers involved in the production of PG-AC [
Carnevale et al. [
Two surveys of workers involved in spiramycin manufacture have been reported. Malo and Cartier [
Our cases add three to the previously reported eight cases of OA attributed to inhalation of synthetic penicillins during their manufacture, and the first case of disease arising from workplace exposure to erythromycin. An additional seven cases of penicillin OA were reported to a UK national surveillance scheme between 1989 and 2009 (THOR personal communication).
An examination of published findings from surveillance schemes in other parts of the world found specific reference to antibiotics in none although the Propulse scheme in Quebec collated seven cases of OA (2.4% of the total) attributed to “medical drugs” [
The clinical features of the published cases are similar to those found in other examples of allergic OA and are broadly indicative of an immediate-type respiratory hypersensitivity. A latent period of asymptomatic exposure was reported for all, usually of fewer than 24 months although occasionally far longer, and rhinitis was a frequent accompaniment to asthma symptoms. In those cases where specific inhalation testing was used in diagnosis and the findings reported in full, a late or dual asthmatic response was reported in 43%. In most cases of isolated early asthmatic responses to specific provocation, there was evidence of immunological sensitisation on skin or serum testing. Each of the implicated antibiotics had a very high hazard index derived from a quantitative structure-activity relationship analysis [
Nonetheless, as with many other low-molecular-weight causes of the disease, the immunological details of antibiotic-related OA remain unclear. While an IgE-associated mechanism is likely in many cases, the possibility of sensitisation arising from an alternative mechanism, perhaps through cross-linking of cell surface receptors, cannot be ruled out. Even when, as is not always the case, their techniques are described, the variety of skin test methods used in the published reports makes their interpretation and diagnostic significance uncertain. In any case, the results were often negative and where they were not there is no systematic information on the findings among exposed but nonasthmatic employees. Similar comments apply to the use of tests for serum-specific IgE antibodies which were less often performed and more often negative although some success has been recorded for in-house assays using cephalosporin conjugates [
Penicillins and cephalosporins are common causes of oral drug allergy which in most cases is attributed to an IgE-associated immune response to one or more hapten-protein conjugates. Penicillins are composed of betalactam and thiazolidine rings with one or more differentiating side chains; the instability of the betalactam causes its carbonyl group to form amide-bonds with the amino groups of lysine residues from nearby proteins. “Minor” antigenic determinants, some formed from side chain-protein conjugates, may also induce IgE immune responses. 7 aminocephalosporanic acid, the active nucleus of cephalosporins, is structurally similar to the active nucleus of the penicillins, consisting of betalactam and 6 dyhydrothiazolidine rings. Hapten-protein conjugates of the cephalosporin betalactam are relatively unstable. IgE antibodies that react to cephalosporins detect a large number of specific antigens derived from protein conjugates formed from the side chain(s), the side chain plus a portion of the betalactam ring, or the complete cephalosporin. On the basis of ELISA inhibition assays in two patients with OA from cefteram, Suh et al. [
Seven of the identified case reports describe oral challenges to the causative antibiotic (in six cases a penicillin or cephalosporin) in patients with OA. All but one produced an allergic response. The three cases reported by Davies et al. [
We have been unable to find valid information on the numbers of exposed employees in any parts of the world but we expect these to number many thousands. Much manufacture is carried out in conditions where exposures to employees are very low but this is not always the case, and we note the general movement of pharmaceutical manufacturing away from its traditional base in western Europe and North America. We note also that most published evidence relates to older types of antibiotic. This is likely to reflect the number and exposures of those involved in their manufacture rather than any intrinsic hazard; the hazard indices for newer antibiotics such as flucloxacillin (0.909), clarithromycin (0.997), tobramycin (1.0), and azithromycin (1.0) are no lower. While cases of OA arising from antibiotic manufacture are rarely published, the available epidemiological evidence, admittedly scanty, sometimes deficient and all of it dated, suggests that the risks may be higher than many appreciate. We suggest that further workplace-based surveys, of careful design and supported by improvements in immunological diagnosis, are required.