Questionable evidence of increasing incidence of invasive penile cancer in Denmark

Abstract

In agreement with the findings in two prior Danish studies, which together covered the period 1943–2003 [1, 2], Baldur-Felskov et al. [3] provided compelling evidence of a stable incidence of invasive penile cancer in Denmark during 1978–2008, as judged from data in the Danish Cancer Registry (DCR) [4]. However, after adding pathology data that are notoriously incomplete before 1998, the authors reported a statistically significant rising trend in penile cancer incidence over the 31-year period. In other incidence studies of human papillomavirus-associated invasive cancers from the same group [5–7], the authors relied exclusively on DCR data. However, for unstated reasons, Baldur-Felskov et al. chose to search for additional invasive penile cancers in the Danish Pathology Data Bank (DPDB) [8], to ‘increase completeness of data’. Adding penile cancers recorded only in the DPDB is problematic, because this data source is unable to make comparable corrections of potentially missing DCR data for all 31 years under study. The authors may well have identified some missing cases of penile cancer in recent years that could not be identified under penile cancer codes in the DCR (e.g., penile skin cancers inaccurately recorded as skin cancers), but the DPDB data did not permit similar corrections before 1998. Consequently, the apparent longterm trend of increasing penile cancer incidence based on the combined dataset most likely represents an artifact resulting from selective addition in recent years of penile cancer cases identified only in the DPDB. Moreover, Fig. 1 in Baldur-Felskov et al.’s paper provides no evidence to support the claim of increasing penile cancer incidence in the most recent decade, whether based on DCR data alone, DPDB data alone, or the two combined. With respect to precancerous lesions, the authors rightly point out that the observed increase in incidence of highgrade penile intraepithelial neoplasia (PIN 2/3) during 1998–2008 might, at least in part, be ‘explained by increasing knowledge about and greater awareness of HPV-associated diseases’. Also, the reported increase in PIN 2/3 incidence should be viewed in the light of the way the authors treated recurrent episodes in the analysis. Counting recurrences more than 24 months after a prior episode as new incident cases of PIN 2/3 most likely contributed importantly to the observed doubling in incidence. Such analyses may be relevant in attempts to estimate the PIN 2/3-associated burden to the healthcare system. However, to learn if, and to what extent, the annual number of newly affected men with PIN 2/3 is truly increasing, it would be informative if the authors presented a graph of the PIN 2/3 incidence during 1998–2008 in which they counted only the first recorded episode of PIN 2/3 in each patient. As a minor point, it is unclear why the authors deliberately excluded the relatively few cases of penile cancer recorded under ICD-O3 topography code C60.8. Such cancers include lesions that overlap the boundaries of two or more anatomical sublocalizations on the penis (e.g., cancers involving both the glans and the prepuce) and penile cancers whose exact point of origin on the penis cannot be determined. Despite difficulties in determining their exact point of origin on the penis, such cancers are nevertheless penile cancers and should be included in the definition of penile cancer. The authors refer to similar findings in a recent Dutch study [9] to add credence to their claim of increasing invasive penile cancer rates in Denmark. Indeed, findings M. Frisch (&) C. J. Ulff-Moller J. Simonsen Department of Epidemiology Research, Statens Serum Institut, 2300 Copenhagen S, Denmark e-mail: mfr@ssi.dk