Edited by: Thomas Bieber and Hans-Uwe Simon.
Clinical and inflammatory features of occupational asthma caused by persulphate salts in comparison with asthma associated with occupational rhinitis
Article first published online: 16 DEC 2009
© 2009 John Wiley & Sons A/S
Volume 65, Issue 6, pages 784–790, June 2010
How to Cite
Moscato, G., Pala, G., Perfetti, L., Frascaroli, M. and Pignatti, P. (2010), Clinical and inflammatory features of occupational asthma caused by persulphate salts in comparison with asthma associated with occupational rhinitis. Allergy, 65: 784–790. doi: 10.1111/j.1398-9995.2009.02288.x
- Issue published online: 4 MAY 2010
- Article first published online: 16 DEC 2009
- Accepted for publication 30 October 2009
- Occupational asthma;
- occupational rhinitis;
- persulphate salts;
- nasal secretions;
- induced sputum
To cite this article: Moscato G, Pala G, Perfetti L, Frascaroli M, Pignatti P. Clinical and inflammatory features of occupational asthma caused by persulphate salts in comparison with asthma associated with occupational rhinitis. Allergy 2010; 65: 784–790.
Background: The relationships between asthma and rhinitis are still a crucial point in respiratory allergy and have scarcely been analysed in occupational setting. We aimed to compare the clinical and inflammatory features of subjects with occupational asthma only (OA) to subjects with OA associated to occupational rhinitis (OAR) caused by persulphate salts.
Methods: The clinical charts of 26 subjects diagnosed in our Unit as respiratory allergy caused by ammonium persulphate (AP), confirmed by specific inhalation challenge (SIC), were reviewed. Twenty-two out of twenty-six patients underwent pre-SIC-induced sputum challenge test (IS) and 24/26 underwent nasal secretion collection and processing.
Results: Twelve out of twenty-six patients received a diagnosis of OA-only and 14/26 of OAR. Duration of exposure before diagnosis, latency period between the beginning of exposure and asthma symptom onset, basal FEV1, airway reactivity to methacholine and asthma severity did not differ in the two groups. Eosinophilic inflammation of upper and lower airways characterized both groups. Eosinophil percentage in IS tended to be higher in OAR [11.9 (5.575–13.925)%] than in OA-only [2.95 (0.225–12.5)%] (P = 0.31). Eosinophilia in nasal secretions was present both in subjects with OAR [55 (46–71)%] and in subjects with OA-only [38 (15–73.5)%], without any significant difference.
Discussion: Our results indicate that OA because of ammonium persulphate coexists with occupational rhinitis in half of the patients. Unexpectedly, rhinitis did not seem to have an impact on the natural history of asthma. The finding of nasal inflammation in subjects with OA-only without clinical manifestations of rhinitis supports the united airway disease concept in occupational respiratory allergy as a result of persulphates.
nonspecific bronchial hyperresponsiveness
occupational asthma associated with occupational rhinitis
provocative dose of methacholine causing a 20% fall in FEV1
specific inhalation challenge
skin prick test
united airway disease
Among the general population (1), there is increasing scientific evidence that asthma and rhinitis might be a unique disease with manifestations in different sites of the respiratory system [‘united airway disease’ (UAD)] (2). Recent comprehensive articles on occupational rhinitis (OR) indicating a high frequency of association between rhinitis and asthma (3–5) and a study by Castano et al. (6) demonstrating a parallel significant physiological reaction of the nose and lungs after challenge with occupational agents seem to support the UAD concept in occupational respiratory allergy as well.
Persulphate salts are low-molecular weight (LMW) compounds used as oxidizing agents in hair bleaches preparations which are considered a relevant cause of occupational asthma (OA) (7–13) and contact dermatitis (14) in hairdressers in several countries. Several articles also reported OR in association with persulphate asthma (3–6).
The relationships between asthma and rhinitis because of persulphates have never been analysed. In this study, we compared the clinical and inflammatory features of subjects with OA-only as a result of ammonium persulphate (AP) with those of subjects with OA associated with OR (OAR) caused by the same compound.
Materials and methods
We retrospectively reviewed the clinical charts of all subjects (N 26, 25 hairdressers and 1 employed in a factory producing bleaching products) who received a diagnosis of bronchial and/or nasal occupational allergy because of AP in our Unit in the years 1996–2008. Patients underwent the common diagnostic pathway for OA and OR (4, 5, 15) including the specific inhalation challenge (SIC) with AP (12), skin prick tests (SPT) for AP and patch tests for occupational agents. Prior to the SIC, 22 patients underwent induced sputum (IS) challenge and 24 nasal secretion (NS) collection and processing to assess lower and upper airway inflammation. Subjects reporting respiratory symptoms specifically related to dye exposure also underwent SIC with paraphenylenediamine.
At the time of diagnosis, the patients did not have any respiratory infection symptom. Medications had been stopped according to a previously published protocol (12, 16). Asthma was diagnosed according to GINA Guidelines (17). Occupational asthma and OR were diagnosed on the basis of a suggestive clinical history, diagnosis of asthma or rhinitis, positive bronchial or nasal response to SIC, respectively (4, 5, 12, 15).
This study conformed to the declaration of Helsinki and was approved by the Internal Review Board of our Institute.
We performed SPTs for common allergens and latex with commercial extracts (Lofarma Allergeni, Milan, Italy) and SPTs with freshly prepared AP solutions (1% and 5% weight/volume in saline) (12). Atopy was defined by at least one positive SPT response to common allergens.
Patch tests were performed by the application of occupational agents (FIRMA, Florence, Italy) on the back with readings at 48 and 72 h (18). Contact dermatitis was diagnosed in the presence of dermatitis and positive patch result (18). Total IgE and specific serum IgE (sIgE) for common inhalant allergens and for latex were also measured (ImmunoCAP; Phadia, Uppsala, Sweden).
Lung function tests
Spirometry, bronchial challenge with Mch and bronchial reversibility test were performed as previously described (12). The provocative dose of Mch causing a 20% fall in FEV1 (PD20) was expressed in micrograms and was considered to be positive when the PD20 was <1000 mcg.
The bronchial reversibility test was considered positive in case of increase in FEV1 > 12% (in addition to an absolute increase of 200 ml) (19). Subjects were sub-grouped according to GINA asthma severity classification (17). All patients were currently working when lung function tests were performed.
Specific inhalation challenge
Each patient signed an informed consent for the SIC approved by the Ethical Committee of our Institute. Specific inhalation challenges were performed in an inhalation chamber (7.46 m3) according to previously published protocols (12, 20). In subjects with pollen allergy, SICs were performed out of the pollen season. On the first day, the patient was exposed to ethanol for 30 min (cumulative exposure) as a control. In the absence of any significant change in FEV1 (≥10%) within 7 h, on the following day, the patient was exposed to a solution of AP (8 mg in 3 ml of distilled water) for a 30-min nebulization (cumulative exposure) (12). Ammonium persulphate concentration in the inhalation chamber was evaluated during two different challenges on separate days, by air sampling at 10, 20 and 30 min after starting exposure using an air pump with a constant flow of 3 l/min. The analysis was performed by means of mobile phase ion chromatography. A mean concentration of 1.01 + 0.11 mg/m3 over the 30-min exposure was detected. The patient was asked to remain in the challenge room for progressively longer periods: 30, 90 s and 2, 4, 8 and 14 min up to a cumulative exposure of 30 min.
Specific inhalation challenge with paraphenylenediamine was performed by a 30-min nebulization using the solution of paraphenylenediamine (12.5 mg in 3 ml of ethanol 65% in water). Each patient was consecutively exposed for 2, 4, 8 and 16 min, resulting in a 30-min total exposure time. At each time-point during SIC, both nasal and lung responses were recorded (12).
Evaluation of bronchial response
Spirometry and peak expiratory flow were registered after each exposure during the challenge and at 5, 15, 30 and 60 min after the end of exposure and then every hour for 7 h and after 24 h. The test was considered positive in case of FEV1 fall ≥ 20% compared to baseline. The responses were classified as early if they occurred within 60 min after the end of exposure, late if they occurred after ≥60 min or dual if they are characterized by both an early and a late response (20).
Evaluation of nasal response
Rhinoscopy was performed before and at 5, 15, 30 and 60 min after the end of exposure and then every hour for 7 h and at 24 h. Mucosal oedema was scored from 0 (no oedema) to 3 (severe oedema) (rhinoscopy score). At the same time-points, nasal symptoms (itching, obstruction, sneezing, rhinorrhoea) were assessed by the same operator and scored from 0 to 3 to define the nasal symptom score. The test on the control day was considered negative, and the subject was considered suitable to undergo SIC with the active substance if the challenge score (symptom score plus rhinoscopy score) was ≤3 (16). The test was regarded as positive if the recorded challenge score was ≥6 at any time-point and OR caused by this agent was diagnosed (4, 5). The threshold of 6 was selected from the analysis of challenge score variability for all the studied subjects during their control day.
Airway and nasal inflammation assessment
Sputum was induced, and nasal blown secretions (NS) were collected and processed as previously described (12, 16, 21, 22). Significant eosinophilic airway inflammation was defined when sputum eosinophils were >3% (23).
A threshold of 1.5% that represents the mean +2SD of eosinophils in healthy subjects was considered a cut-off for nasal eosinophilic inflammation (16).
Continuous variables were expressed as median (1st–3rd quartiles). Data were analysed using Mann–Whitney U-test, chi-square test, Fisher exact test, Spearman rank test and Kruskal–Wallis test (when a significant difference was found, multiple Mann–Whitney U-tests were performed between the subgroups applying the Bonferroni correction). Analysis was performed using the statistical software (statistica for Windows, release 4.5; StatSoft, Tulsa, OK, USA).
Twelve subjects (46.2%) had been diagnosed with OA-only, and 14 (53.8%) OAR because of AP (Table 1). One OA-only and two OAR as a result of AP were also diagnosed with OR because of paraphenylenediamine. No significant difference was found between OA-only and OAR in age, sex, personal or family history of atopy and smoking habits.
|Patients||OA + OAR (n = 26)||OA-only (n = 12)||OAR (n = 14)||P-value*|
|Age (years)||22 (21–30)||22 (20–27)||23,5 (21–34)||0.32|
|OR to other occupational agents n (%)||3 (11.5)||1 (8%)||2 (14,2%)||–|
|Contact dermatitis to AP n (%)||8 (38.4)||4 (33.3%)||4 (28.5%)||–|
|Total IgE (kU/l)||211 (127–359)||209 (145–382)||215 (93–339)||0.79|
|Atopy n (%)||14 (53.8)||7 (58.3)||7 (50)||–|
|Family history of atopy/allergy (yes/no)||19/7||8/4||11/3||–|
|Basal FEV1 (L − percentage of predicted)||3.17 (2.9–3.3) − 102.5 (96.25–109.75)||3 (2.9–3.2) − 102 (95.5–113.5)||3.2 (3–3.4) − 101.9 (96–109)||0.21–0.89|
|Basal Mch challenge (PD20FEV1) μg||908.5 (516.75–1920)||834.5 (318–1722)||908.5 (702–2178)||0.53|
|Bronchodilator test (positive/negative)||5/21||2/10||3/11||–|
|SPTs to AP (±)||0/26||0/12||0/14||–|
|Duration of exposure before diagnosis (months)||84 (48–156)||84 (42–132)||81 (48–240)||0.66|
|Latency OA (months)||48 (24–120)||68 (21–108)||45.5 (24–132)||1|
|Latency OR (months)||45.5 (24–78)||–||45.5 (24–78)||–|
|Lapse OR–OA (months)||0 (0–7)||–||0 (0–7)||–|
|Lapse onset–diagnosis OA (months)||18 (12–24)||24 (12–24)||15 (8–34)||0.51|
|Lapse onset–diagnosis OR (months)||21 (12–36)||–||21 (12–36)||–|
None of the patients reported previous occupations with possible risk factors for asthma or rhinitis. The total duration of exposure before diagnosis and the time elapsed between the beginning of exposure and asthma symptom onset did not differ in the two groups (Table 1).
In 5/14 (35.7%) OAR subjects, work-related nasal symptoms had preceded bronchial symptoms, whereas in 8/14 (57.1%), nasal and bronchial symptoms appeared simultaneously. In one subject, work-related asthma symptoms had preceded nasal symptoms of 60 months.
All patients had normal spirometry findings (Table 1). Basal Mch test was positive in 6/12 (50%) OA-only and in 8/14 (57%) OAR workers. In the overall group, patients with more severe asthma (step 3) had a significantly longer total duration of exposure before diagnosis [144 (111–240) months] than patients of step 1 [42 (36–66) months] and step 2 [60 (30–60) months] (Kruskal–Wallis test P = 0.02 and Mann–Whitney test with Bonferroni correction, respectively, P = 0.021 and 0.033) (Fig. 1). No significant difference was found in basal FEV1, PD20FEV1Mch and asthma severity between OA-only and OAR (Tables 1 and 2).
|Patients||OA-only + OAR (N = 26) (%)||OA-only (N = 12) (%)||OAR (N = 14) (%)|
|Step 1||6 (23)||5 (41.6)*||1 (7.2)*|
|Step 2||9 (36.6)||2 (16.6)*||7 (50)*|
|Step 3||11 (42.4)||5 (41.6)*||6 (42.8)*|
Skin prick tests with AP were negative in all cases. Patch tests with AP resulted positive in four OA-only and in four OAR workers. Skin symptoms had preceded respiratory symptoms of 42 (27–54) months in OA-only group and of 30 (9–63) months in OAR. Two of the four OA-only patients and two of the four OAR with positive patch test to AP also proved positive to paraphenylenediamine, one OAR had a positive patch test exclusively to paraphenylenediamine. Nine workers with a positive patch test were diagnosed with occupational contact dermatitis.
Twenty patients (76.9%) had high total IgE levels (≥120 kU/l); sIgE for common allergens were found in 13 atopic patients.
Four out of twelve (33.4%) OA-only subjects showed an early bronchial response to SIC with AP, and 8/12 (66.6%) a late response. In OAR, bronchial response was early in 6/14 (42.8%), late in 6/14 (42.8%) and dual in 2/14 patients (14.4%). The maximum FEV1 fall during SIC was 27.5 (24–41)% in OA-only and 32 (27.25–34)% in OAR subjects. Nasal response was early in 9/14 (64.3%) and late in 5/14 (35.7%).
Eleven out of twelve (91.6%) OA-only and 13/14 (92.8%) OAR patients produced interpretable NS samples. In 10/11 (90.9%) OA-only and in all OAR, nasal eosinophilic inflammation was found. NS eosinophil percentage did not differ between OA-only and OAR [38 (15–73.5)% and 55 (46–71)%, respectively, P = 0.31]. Excluding the 12 subjects sensitized to perennial allergens and the two with OR because of paraphenylenediamine, nasal eosinophilia was still detectable in six of seven (85.7%) OA-only and in all OAR (N = 7) patients, without any significant difference between the two groups (P = 0.14).
Eight out of twelve (66.6%) OA-only and 10/14 (71.4%) OAR subjects produced interpretable IS specimen (Table 3). Four out of eight (50%) OA-only patients and 9/10 (90%) OAR (P = 0.28, ns) patients had eosinophilic airway inflammation. IS eosinophil percentage did not correlate with last work exposure, PD20FEV1Mch, disease duration before diagnosis, FEV1 maximal fall after SIC in any group.
|Variables||OA-only (N = 8)||OAR (N = 10)||P-value*|
|Viability (%)||80.9 (73–84.6)||78.4 (58.6–88.2)||0.75|
|Macrophages (%)||69.1 (58.4–78.8)||48.3 (34.8–69.6)||0.10|
|Neutrophils (%)||13.7 (10–21.2)||22.4 (13.9–34.4)||0.24|
|Eosinophils (%)||2.95 (0.2–12.5)||11.9 (5.5–13.9)||0.28|
|Lymphocytes (%)||1.3 (0.5–2.2)||1.4 (1–1.5)||0.89|
|Epithelial cells (%)||5 (3.3–9.9)||3.6 (1.2–7)||0.65|
|Total cells (cells/mg)||1140 (905–2150)||1180 (1032.5–2430)||0.70|
|Macrophages (cells/mg)||781.1 (625.8–1703.4)||887 (428.8–1348)||0.56|
|Neutrophils (cells/mg)||210.9 (121.8–293.3)||384 (124.4–626.3)||0.48|
|Eosinophils (cells/mg)||45 (6.8–171.3)||60.1 (31.7–234.9)||0.70|
|Lymphocytes (cells/mg)||12.24 (7.1–23.7)||18.2 (14.7–46)||0.27|
|Epithelial cells (cells/mg)||54.7 (36.9–265.1)||44.1 (15.3–63.4)||0.48|
We found that workers with OA-only because of persulphates and those with OA associated with OR had baseline nasal eosinophilic inflammation, irrespective of the presence of rhinitis, suggesting a UAD related to persulphate exposure at the workplace. No significant difference was found between the two groups in personal and occupational risk factors, asthma severity, nonspecific bronchial hyperresponsiveness (NSBH) to Mch and baseline airway inflammation at the time of diagnosis.
Occupational asthma was associated with OR in about half of the patients, in agreement with our previous findings (12). Other authors have reported a even higher prevalence (75%) of rhinitis symptoms in persulphate asthma (10). The association between upper and lower airway symptoms in occupational setting is commonly found (3–5, 23). Malo et al. (24) first reported that rhinitis symptoms are present in up to 92% patients with OA induced by both LMW and high-molecular weight (HMW) agents. Most of our patients had work-related nasal symptom onset simultaneous to asthma, in accordance with what is more frequently found in LMW-induced asthma (4, 24) and in contrast with Munoz et al. (11) who reported that most patients with persulphate asthma had rhinitis preceding the onset of asthma.
The duration of occupational exposure before diagnosis was long (12) and similar for both groups. However, as we have no data on workplace exposure levels, particularly regarding exposure to bleaching materials, we cannot assume that similar duration of exposure means similar total dose of exposure.
According to our previous findings (12), in both groups, the latency period between the beginning of exposure and asthma symptom onset was longer than that reported in LMW-agent sensitization, which usually requires a shorter interval of time than sensitization to HMW agents (25). This finding may be related to the fact that hairdressers usually perform several tasks, resulting in intermittent exposure to persulphates, which may have required a longer time for sensitization.
The time elapsed between asthma symptom onset and diagnosis was relatively short in both subjects with OA-only and in those with OAR. Therefore, patients were seen in an early phase of their asthma, and this finding may probably explain the absence of impairment in respiratory function (15). It is noteworthy that in subjects with rhinitis symptom preceding asthma, the presence of rhinitis did not influence the referral to a physician, who was consulted only when asthma symptoms appeared. The role of OR as a risk factor for the development of OA has recently been emphasized (4, 5), nevertheless work-related rhinitis symptoms still remain underestimated (23). For the reason that the prevention of OR may also provide an excellent opportunity to prevent OA, our findings indicate a greater attention should be paid to the onset of nasal symptoms in subjects exposed to occupational asthmogens (4, 5, 23).
Atopy was found in half of the patients, without any significant difference in the two groups. This figure is similar to previous findings in persulphate asthma (10–12), but it is slightly higher than that reported for other LMW-induced occupational allergy (26) for which atopy is generally considered not to be an important risk factor (15). The high percentage of atopy is consistent with some reports suggesting that the mechanism of AP sensitization may be IgE mediated (10, 11), nevertheless, according to our previous findings (12) and to Diab et al. (27), the negative results of SPTs with AP in all our cases do not support an IgE-mediated mechanism. An involvement of T-lymphocytes in persulphate allergy has been suggested in humans (27, 28) and in mice (29). In preliminary data, we also found that AP can stimulate human basophils and mastocytes through its oxidizing activity (30); therefore, the mechanism of action of persulphates still remains to be clarified and needs further studies.
Ammonium persulphate is a well-known cause of occupational skin diseases in hairdressers (7, 9–12, 14, 31, 32), and its dermal sensitizing capacity has recently been demonstrated in mice (29). Despite that, only a third of our subjects developed contact dermatitis to AP. We suppose that this observation could be accounted for the fact that most of our workers used gloves in their daily activity, whereas no one reported using any respiratory protection while handling AP. In both groups of our patients, skin symptoms had preceded respiratory symptoms of several years. The importance of skin exposure in occupational respiratory diseases has been suggested especially for isocyanates (33), but the relationships between skin and respiratory diseases induced by AP have never specifically been addressed.
The majority of our patients had intermittent or mild asthma, and 42% had moderate persistent asthma. Few articles have specifically investigated the severity of OA at diagnosis, and the present results in persulphate asthma agree with previous findings in OA as a result of both HMW and LMW agents (34, 35). In contrast to our previous study (34), in the present cases, the degree of severity was positively associated with total duration of exposure before diagnosis. Surprisingly, asthma severity was similar in the two groups of subjects with or without rhinitis (Table 2), whereas one would have expected a higher degree of severity in OAR subjects, as is found in the general population (1). Moreover, the presence of rhinitis did not influence the baseline respiratory function, nor the level of NSBH to Mch. Therefore, in our data, the presence of rhinitis did not seem to have any impact on the natural history of asthma. We are unaware of other studies evaluating the impact of OR on the severity of OA.
The bronchial response to SIC was mostly late in both groups, according to previous articles (7, 10, 31), and in two cases of OAR was associated with an early component (dual response). The maximum FEV1 fall was moderate in both groups, suggesting that the presence of rhinitis does not aggravate the bronchial reaction elicited by the offending agent. The nasal response in OAR group was early in most cases (64.3%), according to Yawalkar et al. (28), whereas Blainey et al. (7) had reported a prevalent late nasal response occurring at about the same time of the bronchial response.
Irrespective of the presence of clinical rhinitis, a high nasal eosinophilic inflammation was detected in almost the totality of patients (96%). Nasal eosinophilia in most OAR subjects was expected, because mucosal inflammation characterizes OR (4, 5), whereas the finding of nasal inflammation in OA-only workers was more surprising. The observation that nasal eosinophilia was still detectable after excluding the subjects, in which exposure to perennial allergens or OR to paraphenylenediamine might have contributed to inflammation strongly suggests a relationship between nasal eosinophilia and AP exposure at the workplace. This finding supports the UAD concept (1, 2, 6) in occupational allergy because of persulphates. The reason why some subjects with nasal inflammation developed OA-only and others developed OAR is not clear.
Seventy-two per cent of workers who underwent sputum induction were found with bronchial eosinophilia. The frequency tended to be higher in OAR (90%) than in OA-only group (50%), but the difference did not reach a statistical significance, probably because of the small number of subjects. Analogously, the degree of eosinophilia and total cell count were similar in the two groups. The presence of sputum eosinophilia in subjects with persulphate asthma had previously been reported by our group (12) and by Macchioni et al. (31). The present data suggest a scarce influence of rhinitis on bronchial inflammation, but further studies are required on a greater number of subjects.
In conclusion, the results of our study show that OA because of persulphates is associated with rhinitis in half of the patients. No personal or occupational factor may predict the development of asthma only or of asthma in conjunction with rhinitis. The presence of rhinitis did not seem to influence the natural history of asthma, the level of NSBH to methacholine or bronchial inflammation. Irrespective of the presence of rhinitis, eosinophilic nasal inflammation was detectable in all subjects and may be considered the expression of the UAD concept in persulphate occupational allergy.
- 17Global Strategy for Asthma Management and Prevention, Global Initiative for Asthma (GINA) 2007. Available from: http://www.ginasthma.org.
- 18Hand dermatitis due to contactants: special considerations. In: RietschelRL, FowlerJF, editors. Fischer’s Contact Dermatitis, 5th edn. Philadelphia, PA: Lippincott, Williams & Wilkins, 2001:261–278., .
- 19Airway responsiveness: standardized challenge testing with pharmacological, physical and sensitizing stimuli in adults; official statement of the European Respiratory Society. Eur Respir J 1993;6(Suppl):S53–83., , , , , .