LAUSR

In this issue, Nakashima, et al. reported the first Japanese case of malignant pleural mesothelioma (MPM) that developed 25 years after thoracic radiotherapy for Hodgkin’s lymphoma without a history of asbestos exposure (1). A table of 29 similar cases in the literature (Table 1) (1-19) shows distinguishing features, which are summarized as follows: 1) women accounted for 55% of cases, which is much higher than the proportion of women who develop asbestosrelated MPM (<20%); 2) the median age at the diagnosis of MPM was 44 years old, whereas asbestos-related MPM is usually diagnosed in the late 50s or older (20); 3) 66% of patients had received radiotherapy for their first malignant tumor before 30 years old; and 4) the median period of latency was 16 years, which was shorter than the median period from the initial asbestos exposure to death of MPM (estimated to be 32 years) (21, 22). These unique characteristics, which differ from the clinical profile of conventional asbestos-related MPM, underscore the association between radiotherapy and MPM. However, the causal relationship between radiotherapy and MPM, the main focus of the current case, has been difficult to establish in humans, although a mouse model successfully showed that radiation exposure induced the development of malignant mesothelioma in the early 1970s (23). Bradford Hill proposed criteria to consider for connecting the association and causation, as follows: 1) strength of the association between a cause and a disease, 2) consistency in different circumstances, 3) specificity in the cause, 4) temporality of the cause that precedes the disease, 5) doseresponse relationship between risk factors and the disease variables, 6) biological plausibility, 7) coherence with generally known facts concerning the natural history and biology of the disease, 8) experimental evidence, and 9) analogy with commonly accepted phenomena (24). The association between radiotherapy and MPM satisfies some of these criteria obviously, while further discussions are needed for others (Table 2). The strength of the association can be measured by the size of the effect, such as hazard ratio (HR) and relative risk (RR) in epidemiological studies. A longitudinal analysis of the US Surveillance, Epidemiology, and End Results (SEER) database showed that external beam radiotherapy actually increased the risk of MPM, but its HR (95% confidence interval) was only 1.34 (1.01-1.77). This value is similarly high to the RR of 1.27 for the association between passive smoking and lung cancer in never smokers (25) but lower than the RRs of 4-6 for other subsequent malignant tumors in childhood cancer survivors (26). Furthermore, this study has an absolute limitation, as the authors did not directly obtain the history of asbestos exposure but instead derived a proxy measure of the exposure by modeling the RR of primary mesothelioma among men in the county of residence (27). The specificity is not met strictly, as it is always difficult to exclude completely non-occupational asbestos exposure, such as that from the neighborhood, domestic, and household (28). In addition, many of these patients received cytotoxic chemotherapy as well, which is generally considered to have carcinogenic effects. The dose-response relationship could not be demonstrated in the SEER study of radiation-induced MPM (27), nor was any association or tendency observed between the dose of radiotherapy and latent period in this case series summary (Table 1). The current case reminds us strongly of the unmet need concerning effective strategies for the prevention and early detection of subsequent MPM as well as other malignancies in cancer survivors who received radiotherapy for their primary tumor.