Pleural plaques and asbestosis: dose– and time–response relationships based on HRCT data

Study design

A large-scale screening programme for asbestos-related diseases, namely the Asbestos Post EXposure Survey (APEXS), was organised at the request of the French Ministry of Labour between October 2003 and December 2005 in four regions of France (Aquitaine, Rhône-Alpes, Lower Normandy and Upper Normandy). The study was initiated following a national consensus statement published in 1999 that recommended the use of chest CT scan for the surveillance of workers with previous moderate-to-high occupational exposure to asbestos 11. It is noteworthy that, in France, workers are entitled to financial compensation and early retirement in the case of asbestos-related disease, including pleural plaques. The target group for this screening programme comprised unemployed or retired asbestos-exposed workers covered by French National Health Insurance. Various recruitment procedures (letters, newspapers, radio, etc.) were used in each region.

All subjects with supposed occupational asbestos exposure were invited to undergo a free screening programme. The screening programme included physical examination, pulmonary function tests performed by the respiratory physician of the subject’s choice and spiral HRCT performed by a programme-approved radiologist. Medical data and a CD-ROM copy of the HRCT were sent directly to the investigators by the patient’s practitioner or radiologist. All data were then rated and recorded in a coordination centre for each region.

Exposure assessment procedures

Volunteers for the screening programme had to fill out a standardised questionnaire describing all jobs held throughout the subject’s working life as well as specific asbestos-exposing tasks. The questionnaires were analysed, blinded to radiological data, by industrial hygienists in Aquitaine and Normandy and by trained national health insurance agents in Rhône-Alpes. Asbestos exposure was assessed using an a priori simplified job-exposure matrix according to occupation and industrial activities. The level of exposure was defined for each subject’s entire career and classified into four classes: high (defined as continuous exposure for at least 1 yr or discontinuous exposure for at least 10 yrs), low (passive exposure), moderate (all other occupational exposure) and nil (no exposure). A cumulative exposure index (CEI) was then calculated by multiplying this level (3, 2, 1 and 0, respectively) by the total exposure duration (in years).

Radiological procedures

Recommended HRCT acquisition parameters were defined in accordance with the French Thoracic Imaging Society guidelines. The main recommendations were slices with a thickness of 1.5–5 mm, a pitch of 1.5–2.0, 120 kV, 60–150 mA maximum, window width of 1600 HU and window level of -600 HU for parenchymal slices, window width of 400 HU and window level of 50 HU for mediastinal slices, with differences depending on mono- or multi-detector equipment. Acquisition in the prone position was only required in the case of gravitational abnormalities. All CT scans were performed without contrast material.

Radiologists who participated in the programme (nearly 300) had to sign an agreement statement. They received guidelines to perform HRCT for the diagnosis of asbestos-related benign diseases. Pleural plaques were defined as focal areas of pleural thickening with soft tissue attenuation, which could contain calcification, typically located on the posterior wall of the lower half of the pleural spaces, often in the paravertebral regions, on the anterior wall between the third and the fifth costal interspaces, and/or on the diaphragmatic pleura. Their parietal nature was defined by their sharply defined edge, well demarcated from adjacent lung, usually with a typical tableland shape, easily recognised regardless of their thickness, or sometimes with less typical tapered margins. In the case of atypical pleural features, only bilateral pleural changes, located in typical areas, were taken into account. They had to be differentiated from the transversus thoracis muscle in anterior parasternal areas, and from subcostalis muscles along the posterior chest wall.

Diffuse pleural thickening was not defined by dimensional criteria but by the presence of parenchymal bands and/or rounded atelectasis, showing the visceral nature of the pleural changes.

Asbestosis was defined by the following pulmonary interstitial abnormalities initially affecting subpleural posterobasal pulmonary areas: early subpleural dot-like opacities, intralobular reticulation, subpleural curvilinear line, interlobular septal thickening and ground-glass opacity (the most common findings); and honeycombing (in the case of advanced asbestosis). These features had to affect both lungs and had to remain present on prone views to establish the diagnosis of asbestosis.

Radiologists also received specific training in the interpretation of HRCT from experienced radiologists and occupational physicians.

HRCT reports and films were both sent to coordination centres. For the purposes of the present study, the presence or absence of radiological abnormalities was rated by only using the HRCT reports according to the above definitions by physicians blinded to clinical data.

The project was approved by the Cochin Hospital ethics committee in Paris. All patients received information on the study and gave their written informed consent to the radiologist for the increased radiation dose delivered by HRCT.

Statistical analysis

The variables used to characterise asbestos exposure were duration of exposure, TSFE, level of exposure and CEI. All variables apart from level of exposure were ranked in classes following the quartile distribution of values. The level of exposure was ranked in four classes (nil, low, moderate and high).

Statistical associations between determinants of asbestos exposure and the presence of pleural plaques or asbestosis on HRCT were studied using stepwise unconditional logistic regression. As CEI was a combination of level and duration of exposure, two distinct models were used with either CEI or the other variables (i.e. level and duration of exposure).

In view of the low prevalence of diffuse pleural thickening, this disease was not specifically analysed, but these patients were included in all analyses concerning pleural plaques and asbestosis.

Crude odds ratios (ORs) and adjusted ORs for confounding factors (body mass index (BMI) and smoking status) were obtained separately for pleural plaques and asbestosis. Trend tests for logistic regressions were obtained using categorical variables as quantitative variables and by replacing ordinal variables by median values.

A cross-validation of a random sample of occupational questionnaires (n = 400) and HRCT reports (n = 200) was organised in the four regions, yielding agreements between hygienists or physicians rated as good to excellent by κ statistics.

All statistical analyses were performed with SAS Software version 9.1 (SAS Institute, Inc., Cary, NC, USA).

Cohort selection

A total of 20,157 subjects volunteered to participate in the APEXS and 16,885 subjects returned the completed occupational questionnaire; 13,859 (82.1%) of these subjects were eligible for the screening programme in terms of their National Health Insurance cover and more than half of them (7,275; 52.5%) underwent thoracic HRCT. Subjects whose HRCT reports were not sent to the coordination centre (n = 734) or who presented incomplete data (n = 709), as well as females (n = 255) and nonexposed subjects (n =  32), were excluded from analysis.

Descriptive analysis

Global characteristics of the APEXS population (n = 5,545) are given in table 1⇓. The mean±sd age was 63.5±5.7 yrs (range 39–85 yrs). Among the subjects in whom smoking status was known, 55% (2,066) were former smokers and only 9% (327) were current smokers. Only 26.2% (956) of the population had a normal BMI. The main jobs and industrial activities were metal working, construction and repairs.

Table 2⇓ describes the exposure characteristics. Overall, the mean±sd TSFE was 41.9±8.1 yrs with an average duration of exposure of 27.4±11.1 yrs. 80% (4,443) of the study population was estimated to have a high level of exposure throughout the entire career.

Based on HRCT data, the overall prevalence was 15.9% (882 out of 5,545) for pleural plaques, 6.8% (375) for asbestosis and 0.7% (37) for diffuse pleural thickening. A significant relationship was observed for both pleural plaques and asbestosis with age and a borderline significant relationship was observed for smoking status (table 3⇓). Higher prevalences of pleural plaques and asbestosis were observed in insulation workers, handlers or shipyard repair workers. Strong relationships were noted for pleural plaques and, to a lesser extent, for asbestosis with the four parameters of asbestos exposure (table 4⇓).

Multivariate analyses: pleural plaques

On multivariate analyses, all exposure parameters except for duration of exposure showed positive relationships with the prevalence of pleural plaques (table 5⇓). Significant and independent associations were observed between pleural plaques and TSFE and CEI (model I) and between pleural plaques and TSFE and level of exposure (model II). Trend tests were also highly significant for TSFE (p<0.0001), CEI (p<0.0003) and level of asbestos exposure (p<0.0001). Taking into account the adjustment variables did not modify any of these relationships and had only a minimal effect on ORs.

Asbestosis

Crude analyses of asbestosis (table 6⇓) showed a positive association with CEI (model I) and TSFE and level of exposure (model II). Only CEI (model I) and level of exposure (model II) remained significant after adjustment for smoking status. However, dose–response curves appeared to be weaker than for pleural plaques.

Conclusion

To date, this study is the largest programme based on HRCT for the screening of asbestos-exposed subjects. We report, for the first time, strong and independent correlations between time parameters of asbestos exposure (as expressed by TSFE) and pleural plaques, and between dose parameters (namely cumulative exposure or level of exposure) and both pleural plaques and asbestosis. These findings indicate that these time and dose parameters must be included in the definition of high-risk populations if further screening programmes are considered.