The value of fractional exhaled nitric oxide in occupational diseases – a systematic review

In the initial stage, 246 articles were identified from the databases. After the automatically filtering for reviews, case studies and age of participants (to exclude studies on children and adolescents), 148 were reviewed. Among these, the title and abstract screening identified 23 studies referring to the general population, 35 analyzing non-occupational diseases, 10 case reports and case series with less than 15 participants, and 3 duplicates. Thirty-six studies referred only to symptoms, without a targeted diagnosis and 11 publications analyzed only the exposure effect of various allergens. This led to 28 articles which were included in the final analysis (Fig. 1).

Fig. 1figure 1

Selection process for the reviewed articles

FeNO utilization for occupational asthma

In total, 16 studies had as primary endpoint the utilization of FeNO for asthma. Of these, 13 addressed the value of FeNO in the diagnostic criteria, and 3 studies included FeNO as a surrogate endpoint for an intervention.

Utilization of FeNO as criterion for diagnosing OA

The identification of a significant threshold value of FeNO - either the baseline value or the variation after workplace exposure or specific inhalation challenge test (SIC) - was the main purpose of these studies. In most cases, they referred to the variation of FeNO from baseline to 24 h after a SIC, but there were also a few studies which considered the variation of FeNO compared to the reference values in the general population [16].

Based on the subjects included in the studies, we identified two different designs (See Supplementary Table 1, Additional File 1): the first included research conducted in specialized centers to confirm the suspicion of OA in patients with very diverse exposure.

The second type of study design focused on a specific exposure (agent or occupation) or a category of OA agents, categorized in high molecular weight (HMW) and low molecular weight (LMW).

Six studies conducted in reference centers used SIC as a confirmation test for OA and one the serial peak expiratory flow (PEF) monitoring. The exposure assessment was comprehensive, as generally requested by the specific compensation rules in each country. They considered possible confounders and used standard methods for the measurement of all variables (See Supplementary Table 2, Additional File 2).

Overall, the reference center studies showed that FeNO after SIC increased significantly compared with before SIC, but reduced sensitivity of the final expert based diagnosis of OA [17, 18]. The relation with SIC depended on the level considered as a significant variation of FeNO before and post challenged [17,18,19], the presence of atopy [18] and on the type of agent (HMW versus LMW) [20]. There was no agreement on the optimal FeNO variation; the cutoff varied from ≥10 ppb to 17.5 ppb, 25 ppb, and even 50 ppb. The sensitivity ranged between 36.8 and 45.3% and the specificity between 81.2 and 100%.

One study used as a reference test the serial PEF measurement analyzed by Oasys computer program [21]. For smokers with higher than 14.7 ppb and non-smokers with higher than 22.1 ppb FeNO values measured within 24 h of work exposure had good correlation with the nonspecific bronchial hyper-responsiveness.

Van Kampen et al. [22] conducted a small study to measure FeNO for 2 weeks of work exposure and 2 weeks without work exposure. A cutoff of 20 ppb was set as a significant work-related increase. Patients underwent a comprehensive evaluation (atopy, agent specific sensitization, lung function and serial FEV1, nonspecific and specific bronchial hyperresponsiveness) and were finally classified as OA or non-OA by a medical expert. Based on the 20 ppb cutoff, nine out of ten finally classified as OA showed an increase of FeNO after work exposure. All positive FeNO cases came from exposure to substances known to induce immunologic OA. Four out of 23 cases which were finally classified as non-OA did not show a 20 ppb increase: one with cobalt exposure, one exposed to formaldehyde and plastic dust, one to lacquers, and one to detergent enzymes. Except for the last one, all the others showed no sensitization to the incriminated agent.

Atopy increased the baseline level of FeNO [19]. Inconsistent results concerning the FeNO variation after SIC were found in different samples [18, 23]. Exposure to HMW agents was the only factor associated with a ≥ 17 ppb variation in FeNO after SIC [23] in a large sample of OA diagnosed by SIC. In another study, the baseline FeNO value was higher in HMW than LMW agent exposure, but a significant increase post SIC was found only for LMW agents [19]. A relatively small study dedicated to LMW agents found no significant differences in FeNO change (increase by 20% or > 6 ppb) 24 h after SIC in 16 positive compared to 16 negative SIC cases [24]. To sum up, it is difficult to draw any conclusion about the significance of FeNO variation related to the molecular weight of the occupational agents, partly because the threshold considered as significant was different; in some studies, the number was too small to reach the statistical significance and, furthermore, the final assessment as OA differs in each country.

The studies dedicated to some specific exposures generally include a low number of cases. When SIC was available, the significance of baseline FeNO or variation of FeNO was compared to this gold standard. For example, in the case of cleaning products, such as sodium hypochlorite [25], a significant increase was found after SIC with bleach, but this was not reflected in all SIC positive cases. In another study which included patients exposed to a mixture of cleaning products [26] the baseline FeNO values were similar to those in the control group.

SIC with isocyanates induced a significant increase in FeNO in two independent studies aiming to identify the mechanism of inflammation in this type of asthma [27, 28]. The first identified the airway wall as the source of the FeNO increase. After 24 h, both FeNO at expiratory flow of 50 ml/s and the bronchial FeNO concentration increased significantly only in the SIC positive patients. The second found a good correlation between the variation of FeNO and sputum eosinophils and provided interesting data on the duration of the increase in FeNO after SIC, which peaked at 24 h only in the SIC positive patients. Levels higher than the initial ones were maintained up to 7 days after SIC, even if not statistically significant.

We found only one study referring to bakers and hairdressers, two occupations exposed to a variety of allergens, most of which are in the HMW category,. Although the number of persons included in the analysis was rather large, the number of cases was low and the imbalance between these two occupations in cases and controls could bias the exposure [29]. As in the studies conducted in reference centers, there was a better specificity than sensitivity of the baseline FeNO values for asthma diagnosis, with a cutoff set to > 25 ppb, as recommended by the American Thoracic Society [15]. When the levels of FeNO were compared to the theoretical reference values for age, gender, and smoking status [16], or when the cutoff was set to lower levels (> 8.5 ppb) together with a positive questionnaire, the specificity decreased while the sensitivity increased [29]. A positive questionnaire was considered if the person reported a diagnosed asthma or at least one of the respiratory symptoms (wheezing, breathlessness, chest tightness, cough and sputum) during the last 12 months; the symptoms have appeared after inception of apprenticeship; and symptoms are present during the working days and improve or disappear during week-ends or holidays [30].

Efficacy of an intervention

FeNO was also used as a surrogate endpoint for the efficacy of an intervention in the prevention of respiratory symptoms and OA (Table 1).

Table 1 Studies in which FeNO was utilized as surrogate endpoint of an intervention in a workplace with potential risk of occupational asthma.

One study [31] compared two interventions (education + better exposure control) with an educational program alone and with no intervention in a randomized group trial. The study concerned 18 supermarket bakeries in one town and the group randomization process referred to a selection of an equal number of units stratified based on the number of employees and production output. The outcomes of the intervention consisted of a reduction of the work-related respiratory symptoms and a more than 10% decrease of the initial FeNO. A year after the intervention, a reduction was observed only in subjects with an initial FeNO > 25 ppb. No other objective measure (e.g. bronchial hyperresponsiveness or lung function) was used to compare the effectiveness of the intervention.

Another project, conducted by Dressel et al. was dedicated to the prevention of respiratory symptoms and allergies in animal farmers [32] by introduction of new educational program. This study included only farmers with diagnosed occupational asthma. FeNO and lung function were measured before and after the program implementation. Particularly in those with high initial values, FeNO was reduced. The achieved low FeNO values were maintained after another year of follow up [33]. Spirometric values did not changed significantly neither in the short term (4–6 weeks) after the interevntion, nor after 1 year. The selection bias and changes in the exposure management during the follow up [31], or the dropout rate [33] were the elements which classified 2 out of the 3 studies in the group with moderate quality (Table 2).

Table 2 FeNO as surrogate endpoint of an intervention in prevention of OA: grading of the studiesOther occupational obstructive lung diseases

Only two studies were specifically dedicated to occupational obstructive pulmonary disease (See Supplementary Table 3, Additional File 3). A large, population study [37] and one from a nanoparticles research team [38].

The first study was conducted on 13,336 subjects from the National Health and Nutrition Examination Survey who underwent FeNO and spirometry measurements. COPD was defined as pre-bronchodilator FEV1/FVC < 70%. Occupational exposure to mineral dusts, organic dusts, exhaust fumes, other fumes, and second-hand smoking was significantly correlated with COPD. Long-term occupational exposure to organic dusts, exhaust fumes, and second-hand smoking in the workplace positively correlated with COPD in subjects with FeNO ≤50 ppb. The probably asthmatic group (defined based on a FeNO > 50 ppb) from workplaces with long-term organic dust and exhaust fumes exposure had lower risk for COPD. This would suggest two conclusions: first, that eosinophilic inflammation is less associated with COPD in workplaces with exposure to inhalants and second, that similar long-term exposure might lead to different types of airway inflammation.

The second study focused on chronic bronchitis. This was a small study investigating the personnel with long time exposed to nanoparticles (average + standard deviation = 18 + 10.3 years), in which post shift spirometry, NO, and tumor necrosis factor (TNF), leukotriene B4 (LTB4), leukotriene E4 (LTE4) in exhaled breath condensate were compared to non-exposed (office workers) controls. Chronic bronchitis was identified from anamnesis. Baseline NO was not significantly different between the exposed and non-exposed groups, but reduction in NO, FEV1, and FEV1/FVC was noticed post shift in the nanoparticles workers. The authors explained the FeNO reduction as an expression of the oxidative mechanisms induced by nanoparticles in the airways which lead to consumption or scavenging of NO [38]. Because there was no chronic bronchitis in the non-exposed group, direct comparisons of FeNO in patients with occupational bronchitis and controls were not performed.

A project conducted in a cohort of diesel engine testers explored the occupational exposure impact on the respiratory system. This project found an obstructive lung pattern to be representative for the long-term effects of these hazards, after adjustment for age, weight, height, smoking, and drinking habit [39]. Moreover, when smokers and non-smokers were compared inside the diesel exhaust group, the lung function was similar, suggesting that occupational exposure had greater effect on lung function than smoking. In the same sample, FeNO was also measured, with no difference between the exposed and the non-exposed subjects [40]. Unfortunately, the relation between FeNO and the diagnose of COPD was not presented. However, there was a significant reduction in FEV1 and FEV1/FCV, compatible to the ventilatory pattern of occupational bronchitis in the exposed workers; the FeNO was similar to the control group.

Two other studies refer, although not as a main scope, to the relation between FeNO and occupational obstructive lung disease. The first explored several inflammatory markers, such as FeNO, interleukin 8, and nitrite in the exhaled breath condensate in non-smoking employees working in a repeated water damaged building, with improper ventilation and a high level of mould [41]. The exposure started approximatively 5 years before. No relation between current symptoms and FeNO was found. FeNO was significantly lower only in the physician diagnosed chronic bronchitis group. The second study found a significant decrease in FEV1 and FEV1/FCV in workers exposed for a short time to petrochemical hydrocarbons from oil refineries as compared to a matched group of white collar workers. The samples of this study did not include COPD or asthma patients. FeNO was measured once, during the working hours, and there was no specification of the relation of this measurement to the recent exposure or duration of exposure. Compared to controls, the FeNO mean was lower, but not statistically significant [42], although this study had some uncertainties on the selection procedures (See Supplementary Table 4, Additional File 4). In patients with distal airways obstructive syndrome previously exposed to fiber glass dust, FeNO was not correlated to the cumulative exposure [34] but the alveolar component of the FeNO was not measured. The study was retrospective and the selection of the patients is not clearly stated.

Interstitial lung disease

FeNO was also a subject of research in interstitial lung diseases (See Supplementary Table 5, Additional File 5 [43,44,45]). Initial findings in HP showed that the alveolar flux of FeNO was higher than in asthma and in healthy controls [35]. These results were confirmed by other research which highlighted FeNO as a distinctive feature of HP [36]. For this purpose, even a cut-off value of 41 ppb, as optimal sensitivity (76.9%) and specificity (85.4%) to diagnose HP was defined. Unfortunately, none of these studies mentions any data about the occupational exposure.

The two studies on HP which covered also the occupational exposure revealed no signal of FeNO levels in occupational HP. The first [46] compared 11 cases of confirmed HP to 14 cases of suspected HP, which did not meet all major criteria: identification of the exposure and appropriate medical history and/or detection of precipitins in serum or broncho-alveolar lavage, histologic pattern of HP, and SIC positive. FeNO was measured prior and 24 h after SIC. The study showed no difference in the baseline FeNO between the two groups and no difference in the FeNO level before and after SIC in cases confirmed with HP, with positive SIC.

The second study was specifically designed for the investigation of small airway disease in HP [47]. FeNO was measured at baseline and after 4 weeks of treatment. Despite the functional and clinical improvement (reduced symptoms, better 6-minute walk test), FeNO did not change. Data collection and confounders are properly addressed, but there is still some bias in the selection process; they are both cases from reference centers, a bias which can be difficult to surmount for a relatively rare disease, which needs extensive and sophisticated tests for diagnose (See Supplementary Table 6, Additional File 6 [43,44,45]).

In pneumoconiosis, the utilization of the FeNO was evaluated in a smaller sample of retired coal miners [48] with no clear data on the representativeness. FeNO was significantly lower in current smokers and in those with lower FEV1. No differences were noted between patients with small or large opacities and controls (formerly exposed workers without silicosis). Exposure to carbon nanotubes was also related to lower FeNO [49] after adjustment for doctor diagnosed cardiovascular, inflammatory or metabolic disease, educational level, recent infection, white blood count and previous exposure to chemicals. The relation was more robust in nonsmokers and became statistically no significant when corrected for previous exposure to nanoparticles.

On the contrary, in workers with asbestosis and asbestos plaques, FeNO was significantly higher than in controls [50]. However, patients with asbestos-related diffuse pleural thickening had similar FeNO as the controls. An inverse relation between FeNO and total lung capacity was found in cases with asbestosis. The authors suggested a continuation of the inflammation even in quiescent lesions (plaques; in diffuse pleural thickening. In their interpretation, either the process of fibrosis is completed and local inflammation was minimal, or the fusion of visceral and parietal pleural layers altered the NO production in these areas.

FeNO had the tendency to decrease with higher cumulative exposure to beryllium [51], but the differences among high exposed, low exposed and controls were not statistically significant. The number of patients with diffuse interstitial fibrosis was not mentioned in this study, but adjustment to this variable did not influence the FeNO.

留言 (0)

沒有登入
gif