Consistency and synergy

To the editor:

We have read with interest the study by De Bruin et al¹  reporting an increased mesothelioma risk among irradiated Hodgkin lymphoma (HL) survivors in The Netherlands. Their almost 30-fold increased risk is similar to that reported by Hodgson et al²  (relative risk = 20), who used 9 NCI SEER registries and other registries in Europe. The association seen in these studies is consistent with our work using the SEER database only, although the magnitude of our increase for males treated with radiation was much smaller (standardized incidence ratio [SIR] = 6.59).³ 

There are several possible explanations for this difference. Both of the European studies restricted their cohorts to those who had survived HL for more than 5 years, and De Bruin et al restricted HL cases to those diagnosed before age 51. We had only a 2-month survivor restriction and no age restrictions, yet all of our male cases of mesothelioma treated with radiation survived more than 5 years, and all but one were less than 51 years of age at HL diagnosis. If we had applied the same exclusion criteria as De Bruin et al, our person-years of observation would have been fewer, while only one case of mesothelioma would have been excluded. The rate of mesothelioma among those treated with radiation (and, therefore, our SIR) would have been greater. Furthermore, we may have missed mesothelioma cases diagnosed in geographic regions not covered by SEER, whereas the study by De Bruin et al actively followed patients for second cancers nationwide,⁴  and nearly half the cohort in Hodgson et al was from national cancer registries, where loss to follow-up is less of a problem.

We disagree, however, that the data “suggest potential synergy” between radiotherapy and asbestos. Asbestos exposure is presented only for those diagnosed with mesothelioma. It is, therefore, impossible to assess separate or joint effects. Rather, the authors compare mesothelioma cases with past asbestos exposure (54%) to an expected prevalence of less than 50% in the Dutch population. The reference cited to support a 50% prevalence in the general Dutch population is inappropriate; the estimate was taken from a review paper by Peterson et al,⁵  who cite this percentage from a publication of Canadian and North American male mesothelioma cases.⁶  Even if this were a valid comparison, the limited data suggest an independent effect of asbestos, not synergy. All 7 of the mesothelioma cases with reported asbestos exposure are men. The cases of mesothelioma in females, none of whom had reported asbestos exposure, have a SIR of 85.2, much higher than that for males (17.9). The authors cite 2 experimental studies as providing support for this hypothesis. The in vitro study did not address mesothelioma, but the effect on biomarkers of oncogenic transformation in a cell model, and the in vivo study reported the frequency of mesothelioma in mice after asbestos exposure only.

Other recently published studies have reported elevated mesothelioma rates associated with radiotherapy.^7,–9  This study adds support to the conclusion that high-dose radiotherapy causes mesothelioma, although it does not provide evidence to evaluate whether there is synergy between asbestos and radiotherapy.