Adult male or female participants who were members of the Tokyo General Construction Workers Union and the Tokyo Doken National Health Insurance Association were invited to participate in this research screening, since they are or have been at a risk of asbestos exposure, considering the nature of their occupations, which included construction or plumbing. When the participants underwent annual health check-ups, written informed consent was obtained prior to their participation, regardless of their asbestos exposure history or the duration of their construction work history. This being a large-scale screening, the target number of participants was ~40,000.

This research screening was approved by the Institutional Review Board (IRB) of Juntendo University School of Medicine, its Affiliated Hospital (Tokyo, Japan)and the Immuno-Biological Laboratories (Gunma, Japan). This study was conducted at 85 research sites, which were approved by the IRB. Blood samples were collected on an annual basis. The high-risk population was identified based on the results of an annual N-ERC/mesothelin assessment. The participants who were selected as members of the high-risk population were advised to visit Juntendo University or its Affiliated Hospital for further assessments to achieve early diagnosis. As a control population for data analysis, a low-risk population was also selected for comparison with the high-risk population.

As previously reported (12), we initially defined the participants as high-risk if they exhibited abnormal levels (>8.0 ng/ml) of N-ERC/mesothelin. However, it was considered that this criterion was not sufficient for a screening and that more optimal criteria were required. We therefore set the new criteria for the high-risk population as follows: i) human anti-mouse antibody (HAMA) not detected; ii) absence of any evidence of renal dysfunction based on medical history and a laboratory test. This criterion was included to consider the effect of renal failure on this marker (13); iii) age ≥35 years; iv) detection of abnormal values of N-ERC/mesothelin on more than two occasions during the annual assessments.

The final confirmation of the high-risk population was performed by a case review committee, which was held on a monthly basis.

Data on a history of asbestos exposure were not collected unless the participants developed mesothelioma, since the scope of our research was a mass scale screening of subjects with a risk factor.

This committee was comprised of at least one physician, one data manager, one technician familiar with N-ERC/mesothelin and one administrator from the Tokyo General Construction Workers Union. The role of the committee was to ultimately identify the high-risk population.

In order to characterize the high-risk population, a low-risk population was also selected as a control prior to the initiation of data analysis. In total, 7,850 participants from 17 approved research sites were selected as the low-risk population, who underwent N-ERC/mesothelin measurements at least twice and exhibited no abnormal values.

The confirmatory diagnosis of mesothelioma was determined via tumor biopsy in the context of appropriate clinical and radiological findings.

The anti-N-ERC/mesothelin monoclonal antibody (mAb) clones 7E7 and 16K16 were described previously (10,11). N-ERC/mesothelin, expressed in Escherichia coli as a glutathione-S-transferase and histidine-tagged fusion protein, was purified and used as an immunogen. Splenocytes from immunized mice were fused with the X-73-Ag8.653 mouse myeloma cells. Using ELISA, the supernatants of the hybridoma cells were screened by their reactivity to the immunogen and several positive clones were selected with the limiting dilution method. The 7E7 and 16K16 clones were selected for ELISA.

Blood N-ERC/mesothelin levels were determined with the sandwich ELISA system using 7E7 and 16K16 mAbs. The details of this analysis were previously described (10,11). Briefly, microtiter plates (96 wells) were coated with 100 μl/well of 100 mmol/l carbonate buffer (pH 9.5) containing 7E7 mAb. The plates were washed with phosphate-buffered saline with 0.1% Tween-20 (PBS-T) and blocked with 200 μl/well 1% (w/v) bovine serum albumin (BSA) in PBS containing 0.05% NaN₃. After washing with PBS-T, 100-μl aliquots of test samples or recombinant N-ERC/mesothelin as a standard, serially diluted in PBS-T with 1% BSA, were added in duplicate to the wells, followed by incubation at 37°C for 1 h. After washing with PBS-T, 100 μl horseradish peroxidase (HRP)-conjugated 16K16 mAb was added to each well, followed by incubation for 30 min at 4°C. The wells were washed with PBS-T and 100 μl of freshly prepared tetramethyl benzidine solution was added as a substrate. The resulting mixture was incubated in the dark for 30 min at room temperature. The reaction was terminated by the addition of 100 μl of 1 N H₂SO₄. The absorbance of the solution was measured at 450 nm in an ELISA reader (EMax; Molecular Devices Co. Ltd., Sunnyvale, CA, USA). In the case of samples with extra value due to HAMA, they were re-measured with sample diluents containing normal mouse IgG in PBS-T with 1% BSA as substitute for PBS-T with 1% BSA.

A longitudinal data analysis model was used to estimate the mean change from Time 1 (the first assessment) over time. The model included factors for group, time and group-by-time interaction as fixed effects. An unstructured covariance matrix was used to model the correlation among repeated assessments. Ninety-five percent confidence intervals (95% CIs) were provided for the difference in change from Time 1 values over time between the high-risk and low-risk groups, using an appropriate contrast of least squares means and referencing a t-distribution. P values were also computed based on the contrast statistics. A step-down closed testing procedure was employed to test the differences in change from the Time 1 value. The procedure started at Time 5 and continued in a descending order in time until the lower limit of the 95% CI for difference in change from Time 1 at a particular time point failed to exceed 0 ng/ml. P<0.05 was considered to indicate a statistically significant difference.

As of September 26, 2010, ~40,000 participants were enrolled in this research screening and a total of 124,288 blood samples were collected and analyzed for N-ERC/mesothelin. Among these, 1,603 samples exhibited abnormal values, defined as >8.00 ng/ml of N-ERC/mesothelin, which was rarely detected among healthy volunteers (10,11). These samples were re-analyzed, in order to exclude HAMA detection. Following re-analysis, 714 samples still exhibited high values. Based on the decision of the case review committee, 62 participants were finally identified as the high-risk population. According to X-ray findings, there were no reports indicating suspected mesothelioma among the high-risk population in this interim assessment. The mean age of the high-risk subjects was 56 years (range, 35–76 years).

Mean N-ERC/mesothelin value by risk group, gender and assessment time are presented in Table I. The mean N-ERC/mesothelin level in the low-risk population were ~2.20–4.33 ng/ml, similar to that in healthy volunteers and the mean N-ERC/mesothelin level in the high-risk population was 9.83–17.28 ng/ml, similar to that in patients with mesothelioma, as we previously reported (11). No significant differences in the mean N-ERC/mesothelin level by age (data not shown in Table I) or gender were identified between the two populations, although the number of data points in the high-risk population was limited.

We also calculated the mean ± SE of N-ERC/mesothelin value and the mean change in N-ERC/mesothelin value from the first assessment in the two populations (Table II). In the high-risk population, the level increased by ~2.0 points annually, whereas it was increasing by only 0.4 points annually in the low-risk population. Therefore, there was a tendency for N-ERC/mesothelin levels to increase significantly during this study period in the high-risk population, compared to those in the low-risk population (Fig. 1). In addition, two participants in the high-risk population ultimately developed mesothelioma. The background information and outcome of the participants who developed mesothelioma is summarized below:

We previously reported a summary of this participant in the form of a case report (13).

The participant was a 72-year-old male, with a 50-year history of asbestos exposure. This participant submitted his blood samples for N-ERC/mesothelin analysis annually during the first 3 years and the values were abnormally high on all three measurements (9.3, 8.8 and 11.1 ng/ml). However, since the N-ERC/mesothelin values were stable, the participant did not receive a N-ERC/mesothelin test in the 4th year, although the site staff advised him to receive a test several times. When the patient visited the research site in the 5th year, the N-ERC/mesothelin levels had suddenly escalated to 78.6 ng/ml and he was diagnosed with epithelioid-type mesothelioma.

Notably, two other participants from the high-risk population developed other types of cancer (lung and appendiceal cancer). Since other cancer types were beyond the scope of this research screening, we were not able to compare the incidence of other types of cancer between the high-risk and the entire populations.