Reply: Glucocorticoid use and skin cancers

Sir,

We appreciate Dr Purdue's interest in our report on use of glucocorticoids and skin cancers. As suggested in his letter, we reclassified indication for steroid use into two categories: (1) atopic conditions (i.e. ‘respiratory conditions and asthma’ and ‘allergies’) and (2) non-atopic conditions (i.e. ‘musculoskeletal and connective tissue disease’, ‘neoplasm’, ‘gastrointestinal disease’, and ‘other condition’) (Table 1). In the reanalysis of the data, we noted an error in Table 1 of the original report. There were only two controls that used glucocorticoids for allergies (the seven controls in the original report included users of inhaled steroids). The correct numbers appear in Table 1 below and are restricted to oral users. The odds ratios associated with oral glucocorticoid use, stratified by Dr Purdue's classification, are shown in Tables 2 and 3. The results for ‘any use’ differ very slightly from the original report because we removed individuals with more than one indication or with missing data on the indication for use for this reanalysis.

Table 1 Oral glucocorticoid use according to indication among skin cancer cases and controls

Table 2 Association between oral glucocorticoid use and BCC, with further analysis by indication

Table 3 Association between oral glucocorticoid use and SCC, with further analysis by indication

The data for basal cell carcinomas (BCC) provide some support for his hypothesis, that is, the association may be limited to non-atopic conditions (Table 2). However, lack of precision in our risk estimates prevents us from drawing any firm conclusions. In our original report, the association was stronger for squamous cell carcinomas (SCC), and for this cell type, the risk associated with oral glucocorticoid use does not differ much by type of condition (Table 3). Unfortunately, we were unable to perform the recommended analysis of inhaled steroid use since the indications for use almost exclusively fell into the atopic category (Table 4). The potential effect of immune response and immunosuppression on the incidence of skin cancer is of considerable interest to us. We hope that further studies will address commonly used immune-modulating drugs along with atopic conditions, and that this, in turn, will shed light on this pathway to human malignancies.

Table 4 Indication for glucocorticoid use among those who use inhaled steroids