A meta-analysis on the relationship between pneumoconiosis and lung cancer

An excess of lung cancer deaths among pneumoconiosis patients has been noted for many years but there has not been agreement on the excess of lung cancer deaths among pneumoconiosis patients in Japan. A meta-analysis on the relationship between silicosis/pneumoconiosis and lung cancer mortality was conducted, by using 36 mortality studies published from 1950 to 1994. MEDICINE and other information sources were used for the information retrieval. After the critical appraisal, 32 eligible studies were identified. Study estimates were then pooled by using both the fixed effect model and the random effect model. An excess of lung cancer mortality among people with silicosis/pneumoconiosis was shown in all 32 studies. Although homogeneity among the studies was rejected (p<0.05) because the shape of the funnel plots assumed a symmetrical pyramidal shape, the summary rate ratio was estimated. The estimated rate ratio was 2.74 (95 percent confidence interval 2.60-2.90) in all 32 studies, and 2.77 (2.61-2.94) in 25 cohort studies. The random effect model results were almost the same as the fixed model results. The estimates in the Japanese studies were a little higher than the overall estimates, which indicated that lung cancer mortality was about three times higher among silicotic patients than among people in the control. This indicated a causal-relationship between silicosis and lung cancer. This means that lung cancer should be regarded as one of the important complications of silicosis/pneumoconiosis. We recommend further research on the relationship between silica exposure and lung cancer.