- Top of page
Background: False-negative responses to specific inhalation challenge (SIC) with occupational agents may occur. We explored whether assessing changes in sputum cell counts would help improve the identification of bronchial reactivity to occupational agents during SICs.
Methods: The predictive value of the changes in sputum cell counts after a negative FEV1 response to a first challenge exposure to an occupational agent was determined using the changes in airway calibre observed during repeated challenges as the ‘gold standard’. The study included 68 subjects investigated for work-related asthma in a tertiary centre. After a control day, the subjects were challenged with the suspected occupational agent(s) for up to 2 h. All subjects who did not show an asthmatic reaction were re-challenged on the following day. Additional challenges were proposed to those who demonstrated a ≥ 2% increase in sputum eosinophils or an increase in nonspecific bronchial hyperresponsiveness to histamine after the second challenge day.
Results: Six of the 35 subjects without changes in FEV1 on the first challenge developed an asthmatic reaction on subsequent challenges. ROC analysis revealed that a >3% increase in sputum eosinophils at the end of the first challenge day was the most accurate parameter for predicting the development of an asthmatic response on subsequent challenges with a sensitivity of 67% and a specificity of 97%.
Conclusions: An increase in sputum eosinophils is an early marker of specific bronchial reactivity to occupational agents, which may help to identify subjects who will develop an asthmatic reaction only after repeated exposure.
There is accumulating evidence that work-related asthma is a prevalent disorder (1), although it is often difficult to differentiate immunologically-mediated occupational asthma (OA) from work-exacerbated asthma in a clinical practice (2). Under- and over-diagnosis of these conditions can both lead to substantial and avoidable adverse consequences for workers, employers, health-care insurance organizations and society (3). As a result, the clinical diagnosis of OA should be established with the highest level of confidence. The clinical history, immunological tests and serial assessments of peak expiratory flow and nonspecific bronchial hyperresponsiveness (NSBR) at and off work can assist clinicians in a stepwise diagnostic approach, although specific inhalation challenges (SIC) with the suspected occupational agent is still acknowledged as the most reliable method for establishing the diagnosis of OA (4, 5). However, even SICs may result in false-negative responses, especially when specific bronchial reactivity to the sensitizing agent has declined after cessation of workplace exposure (6, 7). A postchallenge increase in the level of NSBR has been proposed as an earlier and more sensitive marker of bronchial responsiveness to occupational agents than changes in airway calibre (8, 9). In recent years, airway inflammation, a key feature of asthma and OA, has been increasingly investigated through induced sputum examination, because the technique is noninvasive and can be performed serially over time (10). Sputum studies have shown that exposure to occupational agents at work (11, 12) and in the laboratory (13–16) can induce an eosinophilic and less consistently, a neutrophilic inflammation of the airways in sensitized subjects. It has also been shown that an increase in sputum eosinophils may precede functional changes after repeated exposures to low doses of occupational agents in sensitized subjects (17).
The aim of this study was to examine whether assessment of the changes in sputum cell counts might be useful for minimizing false-negative results of SIC with occupational agents.
- Top of page
This study showed that an increase in sputum eosinophils may precede functional (FEV1 and NSBR) changes after exposure to occupational sensitizers. These results further support the concept that sputum eosinophilia is an earlier marker of specific bronchial reactivity to occupational agents than changes in airway calibre. The findings confirm and expand those reported by Lemiere and co-workers who demonstrated that an increase in sputum eosinophils and interleukin-5 preceded the occurrence of changes in FEV1 and methacholine PC20 after repeated challenges with low doses of occupational agents in patients with ascertained OA (17).
We investigated the usefulness of adding sputum cell counts to functional parameters during SICs in a ‘real-life’ diagnostic process. Hence, the findings may have direct implications in the clinical investigation of OA through inhalation challenges in the laboratory. Our data further outlines that an SIC involving a short period of exposure to the suspected occupational agent (usually 30–60 min) is not sensitive enough for accurately excluding the diagnosis of OA. A substantial proportion [6 of 35, 17% (95% CI: 7–34%)] of subjects with work-related asthma required repeated challenge exposures before demonstrating a positive asthmatic response to occupational agents. The ROC analysis revealed that an increase in sputum eosinophils greater than 3% after a two-hour challenge exposure to occupational agents achieved the optimum trade-off between sensitivity (67%, 95% CI: 33%–95%) and specificity (97%, 95% CI: 82%–99%) for predicting the development of an asthmatic response on subsequent challenge(s) (Fig. 2). These results should, however, be prospectively validated in a larger sample of patients undergoing SICs. Determining cut-off values from ROC curves implies that false-positive and false-negative results are regarded as having similar clinical impacts. In clinical practice, the cut-off value for the postchallenge increase in sputum eosinophils should be lowered if the test is aimed at improving the sensitivity of SICs in order to prevent false negative results, which seems most desirable in the field of OA.
In this study, an increase in sputum eosinophils appeared to be a more accurate predictor of bronchial responses to occupational agents than changes in histamine PC20 values, although changes in postchallenge NSBR have been previously proposed as early indices of bronchial reactivity to these agents (8, 9). Such discordances between changes in sputum eosinophil counts and NSBR are consistent with previous studies showing that airway eosinophilia can be dissociated from airway responsiveness and airflow obstruction in asthma (23, 24). Although an increase in sputum neutrophils has been described after SICs, especially with low-molecular-weight chemicals (16), we found that only decreases in % neutrophils after a first challenge were predictive for the development of an asthmatic reaction on repeated challenge(s), which is likely to result from a mirror effect of the changes in eosinophils. Noticeably, we observed an increase in the absolute number of sputum neutrophils after both negative and positive SICs (Table 3), similar to what has been reported in a previous study comparing sputum cells at work and away from work (12).
Several potential limitations of this study should be carefully considered. Subjects with a negative FEV1 response after two active challenge days were not systematically rechallenged on subsequent days. Accordingly, we cannot formally exclude that some subjects might have reacted if all of them, including those without sputum eosinophilia, had been further challenged. Because of practical constraints, we proposed additional challenges only to subjects who showed an increase in NSBR (8) or an increase in sputum eosinophils at the end of the second challenge. An increase in sputum eosinophils ≥ 2% from baseline was selected a priori, because this ‘cut-off’ value has been considered as being clinically significant in previous studies (12, 15). The occurrence of false-negative results seems, however, unlikely since all subjects were challenged with the suspected agent(s) for prolonged periods (at least 4 h). It is also unlikely that asthmatic responses and sputum eosinophilia might have been inhibited by inhaled corticosteroids (12, 25), since the proportion of subjects treated with corticosteroids was similar in subjects with negative and positive SICs. Finally, some subjects may not have been challenged with the agent that actually caused asthma at work. SICs were carried out in a realistic way, aimed at mimicking workplace exposure as close as possible. Sixty-nine percentage of subjects with a negative SIC were challenged with multiple agents to which they were exposed at work. The possibility of OA could be further ruled out by the assessment of peak expiratory flows, NSBR and/or sputum cytology after resuming work in 17 subjects. These procedures could not be systematically executed after negative SICs since most of the subjects had resigned their jobs and were not allowed to be reexposed at work or they were not interested in performing further tests. The major practical limitation of sputum cell assessment results from the failure to obtain suitable samples in a substantial proportion of the subjects (23%), which is similar to figures reported in previous studies (26, 27). Therefore, there is a need for further evaluating the diagnostic value of changes in the levels of exhaled nitric oxide during SICs, since the technique is less time-consuming and would be more easily available (28).
The persistent increase of sputum eosinophils in the four subjects with work-related respiratory symptoms, but with negative SIC response may have been related to eosinophilic bronchitis and/or occupational rhinitis (Table 2). Eosinophilic bronchitis is characterized by asthma-like symptoms, mainly chronic cough, and sputum eosinophilia (>2.5%) in the absence of variable airflow obstruction and NSBR. Three of these four subjects showed a slight, but persistent increase in sputum eosinophils after repeated challenge exposures to chloramine-T (Subject no. 7) or latex (Subjects nos. 8 and 10) without significant changes neither in FEV1 nor in histamine PC20. Subjects nos. 8 and 10 demonstrated IgE-mediated sensitization to latex, while specific IgE antibodies to chloramine-T could not be detected in Subject no. 7. Noticeably, these two agents have been previously documented as causing occupational eosinophilic bronchitis (29, 30). Three of the four subjects (Subjects nos. 7, 8 and 10) also reported work-related symptoms of rhinitis. An increase in sputum eosinophils has been documented after exposure to common inhalant allergens in nonasthmatic patients with allergic rhinitis (31). These findings underline the relevance of assessing airway inflammation by means of induced sputum in the investigation of work-related respiratory symptoms.
Despite its inherent limitations, the present study shows that an increase in sputum eosinophils is useful in identifying subjects in whom specific bronchial reactivity to occupational agents will become demonstrable only after repeated challenges. The addition of an induced sputum examination to the functional tests might therefore improve the diagnostic sensitivity of SICs. In the absence of changes in functional parameters, a significant increase in postchallenge sputum eosinophilia suggests that further challenge exposures in the laboratory and/or at the workplace are needed before excluding the diagnosis of OA. This approach should enhance the degree of confidence with which the diagnosis OA can be excluded after SICs. Further research is needed to determine whether a postchallenge increase in sputum eosinophils could be considered as reflecting a significant bronchial response when changes in spirometry do not fulfil the recommended criteria, which would allow reducing for the number of challenges and the cost of the procedure (32).