LAUSR

Keratinocyte cancers are exceedingly common in fair-skinned populations around the world, but accurate measurement of their incidence has proven difficult, principally because the sheer volume of cases notified to cancer registries has exceeded their historical capacity to check, code and annotate the data in a timely manner. Thus, for many registries, keratinocyte cancers either were not recorded at all (e.g. in most Australian and U.S. cancer registries), or else only the first occurrence was ever recorded for any particular person. For cancers that are prone to multiplicity this approach has led to gross under-reporting. With the advent of digital pathology reporting and sophisticated coding algorithms, the resource constraints that inhibited earlier manual coding efforts have largely disappeared. Taking advantage of new data linkage and processing technologies, Venables and colleagues have assembled up-to-date and comprehensive data on the incidence of basal cell carcinoma (BCC) and cutaneous squamous cell carcinoma (cSCC) in the U.K. between 2013 and 2015. They obtained linked data from the National Cancer Registration and Analysis Service, the patient administration system and the cancer outcomes and services dataset. The authors identified new diagnoses of BCC and cSCC using coding algorithms, and importantly tested the accuracy of these methods by manually reviewing the pathology reports of 500 cases of BCC and 500 cases of cSCC. In a novel twist, and in contrast to previous investigations of keratinocyte cancer incidence, the authors used a measure of incidence called ‘PPPA’ – that is, the incidence of first BCC and first cSCC per person per annum. This measure is subtly different from that used in earlier reports, which were mostly based on the first ever BCC or cSCC per patient and ignored any subsequent diagnoses, even if the interval between cancers was many years. Venables and colleagues counted the first new BCC or cSCC arising in each year; thus, in the interval 2013–2015 covered by their analysis, a single patient could have contributed up to six new cancers (i.e. one new BCC and one new cSCC in each calendar year), compared with a maximum of two cancers (i.e. one BCC and one cSCC) under the ‘old’ coding rules. In using this method, the authors observed an increase in KC incidence of 51% compared with estimates obtained using the traditional analysis. These new data estimate the incidences of BCC and cSCC to be 285 and 77 per 100 000 person-years, respectively. In terms of descriptive epidemiological features, the authors noted typical and expected increases in incidence with age, and a marked male predominance for cSCC. Geographically, the authors observed an overall inverse gradient with latitude, with the highest rates of BCC and cSCC in the southwest of England, and the lowest rates in northern Scotland. For both BCC and cSCC, the incidence was lower in Greater London, most likely due to ethnic diversity. Of interest, the authors observed higher rates of keratinocyte cancers in affluent areas, which may reflect either increased exposure to causal factors (presumably higher levels of sun exposure through access to vacations) or greater diagnostic opportunity through superior access to healthcare (i.e. surveillance bias), or both. Continued registration and surveillance of keratinocyte cancers are important not only to monitor trends in incidence across the population, but also to identify potential gaps in the system and to plan future health services. For example, this same group has recently analysed these data to estimate the incidence of metastasis among patients with cSCC. Future analyses might focus on defining the characteristics of those who first come to clinical attention with advanced disease, or understanding the correlates of mortality, in an effort to reduce the toll incurred by these cancers. There would also be practical utility in further exploring the extent of multiplicity. These data suggest that over a 3-year interval, at least 50% of new keratinocyte cancers in the U.K. arise in a person who has previously had such a cancer (which explains the higher incidence generated by the new PPPA method), but it does not quantify the range and distribution of multiple keratinocyte cancers arising in this population. Finally, these new data highlight the magnitude of the skin cancer burden in the U.K., with more than 160 000 people each year being diagnosed with BCC, and nearly 45 000 diagnosed with cSCC. Indeed, these numbers are likely to be underestimates, as the lesions counted in these analyses were all excised and confirmed by histology. It is not known how many keratinocyte cancers during the study interval were treated destructively, without histological diagnosis, and thus not captured. Regardless of possible undercounting, it is clear that BCC and cSCC incur a sizeable health burden in the U.K., and yet both of these cancers of the skin are almost entirely preventable. Addressing this issue must remain a priority.