Statistics from Altmetric.com
As its authors intended—by limiting their analysis of the association to instances in which recognition of malignancy succeeded a diagnosis of sarcoidosis by more than one year—the study by Seersholm et al 1 rigorously tested the hypothesis proposed by ‖Brincker and Wilbek2that persons with respiratory sarcoidosis are predisposed to develop malignancies. Several authors have hypothesised the reverse, that malignancies give rise to sarcoidosis. For example, Suenet al 3 reported six patients (four with lymphomas) in whom sarcoidosis followed chemotherapy with a median interval of nine months, and Pandha et al 4 found 48 cases reported in the literature in which sarcoidosis presented concomitantly or following a diagnosis of testicular cancer.
The subject is of considerable conceptual interest. Several authorities have suggested that the production of systemic non-caseating granuloma might reflect a peculiar reactivity of the host to a variety of causative agents. Epidemiological corroboration of the observation that patients with malignancies are prone to develop sarcoidosis would suggest aetiological heterogeneity. One hopes that the authors, who have at their disposal a large database, will test this hypothesis.
authors’ reply Dr Reich raises an interesting hypothesis that previous malignancy and, in particular, treatment with chemotherapy may cause sarcoidosis. In our study, however, we found no cases of malignancy prior to the diagnosis of sarcoidosis, which would have been the case if the hypothesis is correct. A larger study of the risk of sarcoidosis following chemotherapy may reveal an association but requires national sets of sarcoidosis rates in order to calculate the expected number of cases. Unfortunately a national registry of the incidence of sarcoidosis is not available.