Airway Inflammation Among Workers in the Silicon Carbide Processing Industry

Does exposure to aerosols induce airway inflammation and/or a systemic inflammatory response in workers in the silicon carbide (SiC) processing industry?

As yet, no human studies describing the health effects from exposure during recycling of slurry containing SiC and/or PEG have been reported. However, there are several studies that may be relevant which assess the health effects of other particles. It has been demonstrated that small particles penetrate further into the lungs than larger particles and thereby have stronger irritating properties [2]. Results from 2 experimental studies where the effect variables were lung function, organ weights and characteristics of broncho-alveolar lavage (BAL) [3, 4] indicate that SiC dust and PEG itself are probably relatively inert. However, in a third study it was demonstrated that SiC particles have the potential to stimulate test cells (murine macrophage cell line and human osteoblast-like cell line) to produce inflammatory signal substances (tumour necrosis factor (TNF)-alfa), indicating possible inflammatory responses in the lungs [5]. Furthermore, it is thought that the total amount of inhaled dust is important ("overload") even if the substance itself is relatively inert [6]. Repeated high exposures (peaks) may act as a reinforcing factor.

Because there are few previous studies, and the available data are conflicting, there is a need for more knowledge about the association between health effects and exposure to particles in general and to exposure in the SiC processing industry in particular.

Our hypothesis is that airway inflammation may be induced by exposure to SiC particles and would probably be detectable as increased levels of inflammatory cells and soluble inflammatory markers in the sputum and blood samples from non-smoking, asymptomatic workers.

The primary aim of the present study of workers exposed to SiC dust in the work-place is to determine whether this exposure is associated with inflammatory changes in the lungs.

The secondary aims are:

• to examine possible changes in systemic inflammatory markers, and
• to examine changes in lung function among SiC-processing workers over a period of 3 years.

Possible changes in lung function will be monitored by spirometry measurements (study A). Airway inflammatory responses and changes in circulating inflammatory markers will be assessed by examination of the induced sputum samples and blood samples in a subgroup of workers (study B).

General value An important goal of the study is to provide knowledge that can be important for workers who are exposed to particles and aerosols in the SiC-processing industry or other industries processing SiC and/or PEG. The study may also be relevant to workers in other industries with dusty environments given the possible "overload" mechanisms. The study could possibly also have the potential to provide additional knowledge about effects of exposure to similar particles e.g from air pollution. Because few former studies have used induced sputum in a prospective design, such studies are of particular interest.

If early signs of lung inflammation or a decrease in lung function are demonstrated in the present study, it will be necessary to implement additional measures to reduce the dust levels in this industry; e.g., through enclosure of processes. The investigators believe that the results from this study together with the existing knowledge of inflammatory effects associated with occupational exposure to particles may be relevant for the prevention of chronic respiratory disease.

III - Study design and methods Study population In the present study, the effects of exposure to aerosols on airway inflammation and systemic inflammation will be investigated by means of spirometry, FeNO, induced sputum and blood sampling over a three-year period.

Study A: The whole group of workers will complete a questionnaire on respiratory symptoms and undergo spirometry at baseline, and again after 3 years. All former workers will be contacted and asked to complete a questionnaire to assess the proportion of workers resigning because of (or partly because of) health effects. These workers will also be invited to undergo an examination of lung function (spirometry). This is to ensure that all workers ever employed are included in the study population.

Study B: A subgroup comprising non-smokers will in addition be followed with blood samples, induced sputum tests and exposure measurements every 6 months over the same time period.

Controls: Exposed subjects will be compared with non- or low-exposed subjects (administrative/office workers). In addition to the prospective study, the investigators will compare the baseline findings from the sputum and blood samples with those from already established external control groups from previous studies on healthy non-exposed workers and from cement production workers from the same geographic area [7].

Exposure-response analyses: Workers will be divided into groups with different exposure levels in order to study exposure-response trends of the measured agents in the exposed population (this design is less vulnerable to selection effects).

Baseline examinations will be repeated for all subjects and for the sputum subgroup. Detailed, validated questionnaires will record the medical history, relevant illnesses, use of medications, occupational history as well as present exposures.

Spirometry will allow registration of lung function and FeNO measurements the monitoring of possible airway inflammation.

A group-based strategy will be applied for the collection of exposure measurements at baseline and at follow-up (3 years) for the subjects not included in the sputum part of the study.

Induced sputum collection will be applied to detect possible changes in inflammatory cells and soluble markers from the airways.

Blood analysis will be collected to examine possible changes in circulating inflammatory markers and their association to exposure.

All participants in the sputum subgroup will carry individual exposure measurement.

The employees who leave the cohort will contribute with data while they are employed (defined as time-at-risk) and a drop-out test will be performed to determine whether they experienced work-related respiratory symptoms or diseases. New employees will be included during the study period and assigned a time-at-risk equal to their time of exposure.

Exposure measurement and assessment The dust/aerosol samples will be collected with the use of thoracic and inhaled respirable cyclones to estimate the doses for the deposition in the bronchi and in the alveoli.

Validity Factors that may affect lung function and inflammatory markers in sputum Age, height, BMI and sex will be recorded for the study. Height in meters will be entered into the computer program for calculating spirometry values. Weight (body mass index) affects lung function and will be measured before the examination. The investigators will also record possible acute respiratory infections during the preceding 12 months. An upper respiratory infection during the preceding three weeks may affect the level of inflammatory cells and markers in sputum.

There is a possibility that very early inflammatory effects among these workers could be succeeded by adaptation and thereby reduction of effects. In this case, the inflammatory response at some time after employment could be smaller than observed initially. This may lead to underestimation of effects. The investigators consider that a dynamic cohort design including only subjects exposed for the first time in this industry is not feasible. However, a follow-up of three years is considered sufficient to detect clinically relevant changes. The cohort in this study will also have a relatively short prior employment in the industry because the plants were opened in 2006 and 2009. Spirometry screening of all workers, performed during the first days of employment at the local health services, is available. These will be compared with the baseline recordings to evaluate possible effects prior to inclusion in the present study.

Statistical analysis and study power

The associations to be studied are individual changes in the effect variables:

• the trend over the study period (study A: baseline and 36 months for the whole group of workers; study B for the sputum subgroup: baseline, 6, 12, 24 and 36 months), and
• between exposed and controls (exposed and unexposed/administrative staff/office workers) at baseline and differences in trend during the study period.

T tests will be used to compare cases and controls at baseline for normally distributed variables, and a Mann-Whitney test for non-normally distributed variables. Where multiple tests are to be conducted, ANOVA and Kruskal-Wallis tests will be applied prior to the pairwise comparisons to determine the overall significance level.

Multiple regression models with adjustment for age, sex, smoking habits and other relevant factors for each analysis will be applied when adjusting for other factors at baseline.

To analyse the repeated measures, general linear models (GLM) and mixed models will be used.

The power calculations are based on clinically interesting differences in percentage of neutrophils and spirometry values. The study is designed to include at least 20 participants in each group based on a power estimate of 90%, which assumes a 95% significance level, to detect a true difference in the percentage neutrophils between two tests of at least 16%. Power calculation for FEV1 and FVC% of predicted: if the goal is to detect a difference of 1.5% in FEV1 and FVC% predicted, then given a standard deviation of 5% the sample must consist of 117 persons in each group. This will ensure 90% power to detect a difference at the 5% significance level.

It is expected that the response rate among non-smokers will allow approximately 40 employees and 20 unexposed workers to be included in the sputum part of the study (study B). Based on the power calculations, prior experience and published studies, this will be enough to detect a difference in development between the two groups of the expected magnitude. In total, 150 workers are employed at SiC I and II (to be included in study A). It is planned to include the employees at Glomfjord (Nordland) SiC processing if the sample size from Porsgrunn is not sufficient. To ensure sufficient unexposed workers, the neighbouring Renewable Energy Corporation (REC) plant will be used to recruit controls if needed. The REC plant started production at about the same time and has a similar work-force to the study group.

VI - Ethical aspects It is considered that the purpose of the study and the methods being used do not violate generally accepted ethical values. Participation in the project is voluntary. Collection of biological material and information will take place with the participants' consent. All participants will sign consent forms after receiving written information about the survey. Participants will at any time have the opportunity to withdraw from the study without giving a reason.

According to an agreement with the SIC- processing, the research fellow and project group can freely publish the results of the survey. The project is not financially dependent on grants from the industry, and the industry (SIC) financing of the project is limited to the exposure monitoring.