Cancer incidence rates among Lawrence Livermore National Laboratory (LLNL) employees: 1974-1997.

BACKGROUND: In the mid-1970's an excess of malignant melanoma of the skin was noted among employees at the Lawrence Livermore National Laboratory (LLNL). A 1984 cancer incidence study showed a non-significant excess of total cancers among female employees with significant excesses for melanoma, rectum and anus, and salivary gland cancers. For male employees, there was a non-significant deficit of total cancer with significant excesses in melanoma and non-brain nervous system cancers. This paper reports the results of a surveillance effort to update our understanding of the patterns of cancer incidence in this population.

METHODS: We used California Cancer Registry (CCR) data to ascertain employees who had worked for six or more consecutive months at LLNL during the 24-year period of 1974 through 1997 who were diagnosed weith cancer during that time frame. We used the Standardized Incidence Ratio (SIR) in our analyses.

RESULTS: There were 17,785 employees who provided 186,558 person-years of observation: 145,203 were from males and 41,355 were from females. The CCR, through its linkage techniques, identified 541 individuals with invasive cancer and 96 with in situ cancer. A total of 404 males had invasive cancer and 33 had in situ cancer whereas there were 137 females with invasive cancer and 63 with in situ cancer. The SIR for invasive cancer in males was 69 (95% CI 62-76). The overall cancer SIR for males was unaffected by calendar time. There were only two invasive cancer sites with significant excess: melanoma and cancer of the testes. For eight categories or cancer sites, we found a statistical deficit in cancer incidence. The most striking deficit occurred in cancer of the lungs and bronchus with a SIR of 36 (95% CI 26-50). The SIR for invasive cancer in females was 80 (95% CI 67-94). The overall cancer SIR for females decreased over calendar time. There was a statistically significant deficit for cancers of the female genital organs. There were 84 cases of invasive and in situ melanoma in both genders. Time-trend analyses for melanoma showed a significant excess during the years 1974-1985 but a reduction to community rates from 1986 through 1997. There were 21 individuals with testicular cancer with a SIR of 207 (95% CI 129-317). There were no differences in age at diagnosis or cell type with the comparison population. We analyzed the data using the same radiosensitive cancer categories used in the 1984 study. There were no increases in SIRs in any of these categories.

CONCLUSIONS: We found that the LLNL employees had less cancer than expected with males having relatively fewer cancers than females. The lung cancer rate for males was remarkably low. Since 1986 the melanoma rates resemble the community rates. Testicular cancer rates are modestly elevated and appear to have been so for the past 20 years. Lifestyle patterns, including smoking, and cancer screening activities are probably important contributors to the observed low cancer rates.