Previous investigations have studied the mortality of workers employed at a chemicals production factory in north Wales, which manufactures and uses vulcanisation inhibitors and accelerators, antioxidants, and many other proprietary products for the rubber industry.1^–3 The overall study cohort comprised 2160 male production workers, and mortality and morbidity for bladder cancer have been investigated to the end of 2005.3 It was concluded that cases of occupational bladder cancer had occurred at the plant, that exposure to ortho-toluidine was responsible for part of the bladder excess, and that the manufacture of phenyl-B-naphthylamine (PBN) and/or exposure to 2-mercaptobenzothiazole (MBT) may also be involved.3 The purpose of this report is to present site-specific findings for cancers in the sub-cohort exposed to MBT to determine whether MBT may be involved in cancers other than bladder cancer. MBT is used as a vulcanising agent in the rubber manufacturing industry, a corrosion inhibitor in auto radiator and metalworking fluids and a stabiliser in the manufacture of plastics. Investigators from the US National Toxicology Program found some evidence of carcinogenic activity from MBT exposure in male and female rats, no evidence of carcinogenic activity in male mice, and equivocal evidence of carcinogenic activity from MBT exposure in female mice.4 MBT and its derivatives were manufactured at the plant from 1932 to 2001.

MATERIALS AND METHODS

Earlier reports1^–3 had identified all 2160 male production employees (hourly paid personnel) with at least 6 months’ employment at the factory and some employment in the period 1955–1984. This cohort had been assembled with the agreement of company management and workforce representatives. The Office for National Statistics (ONS) has, with the approval of the British Medical Association Ethics Committee, continued to supply copies of death certificates with the underlying cause of death coded to contemporaneous revisions of the International Classification of Diseases (ICD), together with details of cancer registrations as recorded in regional cancer registries. Job histories for the period 1930–1988 were available in terms of some 300 unique job-department titles, and a total of 363 employees had ever been employed in one of the 17 job/department titles judged by a former company occupational hygienist (DJW) to involve exposure to MBT. Three of these 17 titles also involved exposure to aniline; many of the MBT workers also had periods of working in other departments involving other chemical exposures. At the closing date of the survey (31 December 2005), 222 members of the MBT cohort had died, 136 were traced alive, and the remaining five subjects were untraced. Many MBT-exposed workers had been first exposed decades earlier such that long latency periods were available for study (decade of first exposure to MBT: 1930s, n = 6; 1940s, n = 36; 1950s, n = 111; 1960s, n = 69; 1970s, n = 109; 1980s, n = 32).

In 1988 DJW had provided assessments of 8 h time-weighted average exposures to both MBT and MBT derivatives for each job/department title and for six calendar periods (1955–1960, 1961–1967, 1968–1977, 1978–1980, 1981, 1982–1985). Jobs attracted either zero exposure, very low exposure (0.1–1 mg.m⁻³), low exposure (1–2.5 mg.m⁻³), medium exposure (2.5–6 mg.m⁻³), or high exposure (6–20 mg.m⁻³). These estimates had been based on limited monitoring data for the period 1977 onwards, a review of process manuals and other company records for earlier years, and discussions with long-serving employees. Some exposure estimates had to be adjusted by a “year fraction” factor to take into account the fact that not all jobs were associated with MBT throughout the whole of a working year (eg, product campaigns). Mid-class values of the adjusted exposure ranges were used as point estimates of exposure levels.1 2 Exposure estimates for 1955–1960 were applied to the earlier period of 1932–1954 and exposure estimates for 1982–1985 were applied to the later period of 1986–1988. Individual cumulative exposures (mg.m⁻³.y) were then estimated as the sum of the products of employment durations and the relevant exposure concentration.

External standard: standardised mortality ratio (SMR) approach

Expected numbers of deaths were calculated from serial mortality rates for England and Wales applied to similarly defined arrays of person-years at risk (pyr) generated by the data. Workers entered the pyr at the end of the 6-month minimum period of employment, 1 January 1955, or the date first exposed to MBT, whichever was the later date. They left the pyr on the closing date of the study (31 December 2005), the date of death, the date of emigration, or date last known alive, whichever was the earliest date. SMRs were calculated as the ratio of observed deaths to expected deaths, expressed as a percentage (both numerator and denominator refer to underlying causes of death). These procedures were accomplished with the PERSONYEARS software.5 Significance tests were two-tailed and no contributions were made to observed or expected numbers past the age of 85. This censoring at 85 was applied for three reasons. First, published mortality rates are only available for the open-ended age group ⩾85 and the distribution of the cohort pyr by single years of age might be very different from that of the general population; second, the reliability of cause of death particulars is probably poorer at later ages; and third, any study subjects incorrectly classified as traced alive at the end of the study would have a disproportionate effect on the expected numbers for the open-ended age group.

Equivalent analyses were also carried out for cancer registration (incidence) data. National cancer registration data are not available before 1971, and the calculation of standardised registration ratios (SRRs) was limited to the period 1971–2005. Cancer sites were selected for further investigation if an SMR or SRR for that site of cancer was significantly elevated (or close to levels of formal statistical significance).

Internal standard: Poisson regression

The SMR analysis had to be limited to an analysis of underlying causes of death. For the internal analysis greater flexibility was available and cases were selected as those deaths for which a primary cancer of interest was mentioned on any part of the death certificate or in the cancer registration particulars recorded in the national cancer registration scheme. For subjects with more than one identification of a cancer of interest (death certificate and cancer registration, or multiple cancer registrations), the earliest date of identification was taken as “date of diagnosis”. Follow-up particulars for the remaining 1797 members of the overall cohort not exposed to MBT were included to provide a suitable “internal” comparison group.

Variables considered to have the potential for influencing cancer risks within the overall cohort were age, calendar period, cumulative exposure to MBT, duration of employment in the aniline-exposed departments, duration of employment in the PBN-exposed department, and duration of employment in the ortho-toluidine-exposed department. Variables were not treated as continuous variables, but rather each variable was categorised into several levels. In constructing the models, it was necessary to ensure that there was at least one death observed at each level of each variable. The analysis allowed subjects to contribute pyr to contemporaneous categories of these six time-dependent variables.

The EPICURE computer program was used to provide pyr and numbers of deaths for all combinations of all levels of the selected variables.6 The EPICURE program was also used to carry out statistical modelling by Poisson regression using maximum likelihood techniques.7 The purpose of the modelling was to calculate point estimates of relative risk (rate ratios) for each of the four levels of the cumulative MBT exposure variable (none, 0.01–21.24 mg.m⁻³.y, 21.25–63.74 mg.m⁻³.y and ⩾63.75 mg.m⁻³.y), with and without adjustment for other variables. The “cut-off” values for these exposure groups were based on exposures estimated to be received from 5 and 15 years of “medium” exposure to MBT, respectively. These exposure categories have been used in previous analyses.1^–3 The significance of any trend in risk across the cumulative MBT exposure categories was assessed by repeating the analysis while treating cumulative exposure as a continuous variable, coded 1, 2, 3 or 4 for the four levels of exposure.

RESULTS

Mortality from all causes combined in the MBT-exposed cohort was close to expectation (observed; Obs 201, SMR 106, 95% CI 92 to 122). Observed and expected numbers of deaths are shown by site of cancer in table 1. Significant excess mortality was found for all neoplasms (Obs 76, SMR 141, 95% CI 111 to 177), cancer of the large intestine (Obs 8, SMR 232, 95% CI 100 to 457) and bladder cancer (Obs 8, SMR 374, 95% CI 162 to 737). Non-significant excesses were shown for lung cancer (Obs 27, SMR 138, 95% CI 91 to 201) and multiple myeloma (Obs 3, SMR 440, 95% CI 91 to 1287). Non-significant excesses were also shown for cancer of the kidney and other cancers of the urinary tract (renal pelvis, urethra and ureter) such that the SMR for all cancers of the urinary tract (ICD-9 188–189) was markedly elevated (Obs 11, SMR 345, 95% CI 172 to 617). Mortality from all neoplasms excluding multiple myeloma and cancers of the large intestine, lung and urinary tract was unexceptional (Obs 27, SMR 100, 95% CI 60 to 145).

Observed and expected numbers of cancer registrations are also shown in table 1. Significant excess morbidity was found for all malignant neoplasms (Obs 97, SRR 148, 95% CI 120 to 181), cancer of the bladder (Obs 12, SRR 253, 95% CI 131 to 441) and multiple myeloma (Obs 4, SRR 465, 95% CI 127 to 1191). Non-significant excesses were shown for cancers of the large intestine (Obs 9, SRR 181, 95% CI 83 to 344) and lung (Obs 26, SRR 152, 95% CI 99 to 223). Non-significant excesses were also shown for cancer of the kidney and for other cancers of the urinary tract such that the SRR for all cancers of the urinary tract (ICD-9 188–189) was markedly elevated (Obs 17, SMR 268, 95% CI 156 to 429). Morbidity from all malignant neoplasms excluding multiple myeloma and cancers of the large intestine, lung and urinary tract (and non-melatonous skin cancer) was slightly elevated (Obs 31, SRR 124, 95% CI 84 to 176).

Cancer of the large intestine, lung cancer and multiple myeloma were selected for further investigation; bladder cancer findings have been reported in detail elsewhere.3 In the total cohort, subjects with cancer of the large intestine (before age 85 years) were identified by death certificate only (12 cases), by cancer registration particulars only (13 cases), or by both sources (14 cases); case subjects leave the pyr on the date of the first notification. Subjects with cancer of the lung were identified by death certificate only (36 cases), by cancer registration particulars only (14 cases), or by both sources (86 cases). Subjects with multiple myeloma were identified by cancer registration particulars only (one case) or by both sources (five cases).

Table 2 shows rate ratios (relative risks) for lung cancer, for cancer of the large intestine and for multiple myeloma by categories of estimated cumulative exposure to MBT and its derivatives. The first column shows rate ratios with adjustment for age and calendar period; the second columns of rate ratios include additional adjustment for duration of employment in the PBN department, duration of employment in the ortho-toluidine-exposed department and duration of employment in the aniline-exposed departments (see table footnotes for variations). There were no cases of multiple myeloma in workers exposed to PNB or ortho-toluidine, and no cases of cancer of the large intestine in workers exposed to ortho-toluidine. Risks are all shown relative to a baseline risk of unity for workers (person-years) in the non-exposed category. Significant positive trends are shown for cancer of the large intestine and for multiple myeloma in relation to MBT exposure; these trends could not be explained by exposure to aniline, PBN or ortho-toluidine. Clearly, the significant trend shown for multiple myeloma is based on small numbers of deaths but the application of exact methods more suited to small samples8 also indicated a significant trend (p value for trend revised from 0.013 to 0.04).

DISCUSSION

On the basis of national mortality and cancer incidence rates, this cohort of MBT-exposed workers experienced unusually elevated risks of multiple myeloma and cancers of the large intestine, bladder and lung (albeit the latter was only of borderline significance). More detailed analysis of bladder cancer risks is available in another report,3 but the more detailed analysis carried out in this report for the remaining three sites of cancer showed statistically significant positive trends both for the risks of cancer of the large intestine (p<0.001) and for multiple myeloma (p = 0.013) with estimated cumulative exposure to MBT.

Confident evaluation of these findings is not possible because they relate to novel findings found in a single study, particularly given the small size of the MBT cohort and the fairly crude exposure assessments available for analysis. A large number of assumptions had to be made in the course of producing the exposure assessments and there must be considerable scope for individual misclassifications to have been made, particularly if the available work histories do not capture all department changes at the plant. The lung cancer findings are further compromised by the absence of cigarette smoking histories. There was, however, no suggestion of elevated risks (on the basis of national rates) for cancer of the large intestine or multiple myeloma (or lung cancer) in workers at the plant not exposed to MBT, so there is limited scope for the positive findings to be due to unadjusted occupational confounding.

Two published reports are available for a “sister” study of MBT workers carried out at the Nitro plant, USA.9 10 The US cohort comprised 1059 production workers of whom 600 were exposed to MBT or its derivatives. The more recent US report found no excess of lung cancer in MBT-exposed workers (Obs 27, SMR 103); findings were not presented for multiple myeloma or cancer of the large intestine.10 Confident interpretation of the UK findings would be assisted by a similarly detailed updated analysis being carried out on the US cohort. Studies of other workers exposed to MBT in its manufacture or use would also be most helpful to test two new hypotheses: MBT exposure in humans increases the risk of (1) cancer of the large intestine, and (2) multiple myeloma. Such studies will take time to implement. In the meantime, perhaps MBT should be handled with increased care as it may be a human carcinogen.