In an excellent extensive article, Graham and Helwig1 presented the concept that Bowen's disease of the skin is associated with internal malignancies in a high proportion of cases. Therefore, the diagnosis of Bowen's disease on a cutaneous lesion should inspire a search for a deeper, more serious lesion. These authors found that in 35 patients with this disease who were known to have died, 20 (57%) had an associated internal cancer. If one were to add the four instances of reticuloendothelial diseases, this would rise to 68.7%, and the further addition of four instances of metastasizing cutaneous cancer would increase the percentage to 80%. These figures are overwhelming and convincing.
However, statistics are difficult to evaluate. Two factors tend to cast doubt on theories based on figures. First, the numbers must be sufficiently large to be significant. While the statistics quoted above are very suggestive, it must be recognized