Epidemiology of Merkel cell carcinoma in New Zealand: a population-based study

Dear Editor, Current evidence suggests that ultraviolet (UV) radiation (UVR) exposure is an important risk factor in the aetiology of Merkel cell carcinoma (MCC).1,2 New Zealand (NZ), with its high UV levels and population of mainly Celtic descent, has some of the highest rates of UV‐related skin cancers in the world, with particularly high rates of melanoma.3 Given the similarities in aetiology and biology between MCC and melanoma it could be anticipated that the incidence of MCC in NZ would also be comparatively high.4 Two recent studies from Queensland and Western Australia (regions with similar UV indices and racial profiles to NZ) have reported the highest rates of MCC in the world.5,6 However, to date, no published reports have evaluated the status of MCC in NZ. In order to determine the incidence (and associated trends) of MCC in NZ, as well as describe patient demographics and clinical characteristics, population data on MCC cases over a 10‐year period from 2002 to 2011 were collected from the New Zealand Cancer Registry (NZCR) by way of an electronic search of International Classification of Diseases‐10 codes. Demographic data (age at diagnosis, site of MCC, year of diagnosis, ethnicity and sex) were extracted. Disease characteristics such as site of occurrence and disease stage at diagnosis were also analysed. Site of occurrence was divided into head and neck, upper limb, lower limb and trunk. MCC stage was reported according to the three‐tiered system proposed by Yiengpruksawan et al. (stage I, localized skin disease; stage II, regional lymph node disease; stage III, metastatic disease).7 Estimated populations were obtained from NZ census data and provided the rate denominators for all analyses.8 Basic descriptive statistics and figures were used to summarize data. The two‐sided Student's t‐test was used to determine whether differences existed between the means of independent samples; the Chi squared test or Fisher's exact test were used to determine differences in categorical variables. A P‐value < 0·05 was considered statistically significant. Age‐adjusted incidence rates [per 100 000 person‐years (PY)] were calculated for sex, year of diagnosis and 5‐year age group using direct standardization to the U.S. standard 2000 population.9 Linear regression was used to investigate trends in incidence by calendar year. From 2002 to 2011 a total of 356 cases of MCC were reported to the NZCR. Demographic data are displayed in Table 1. MCC typically occurred in elderly patients, with an overall mean ± SD age at diagnosis of 77·2 ± 11·2 years. Europeans accounted for the majority, making up > 91% of all cases. Most cases were localized (stage I) at time of diagnosis (47·2%), with stage II and stage III MCCs accounting for a further 20·8% and 10·7% of cases, respectively (21·3% of cases had insufficient histological data to allow for accurate staging). There were no significant differences in tumour stage at diagnosis between men and women. The head and neck region was the most common site of MCC (49·0%), followed by the lower (17·5%) and upper (16·3%) limbs. Interestingly, women were significantly more likely to develop MCC of the lower limb than men (P =0·01). Patient demographics and clinical characteristics Data are given as n (%) unless otherwise indicated. NS, not significant. Patient demographics and clinical characteristics Data are given as n (%) unless otherwise indicated. NS, not significant. During this 10‐year period the overall age‐standardized incidence was 0·88 per 100 000 PY, with men having a higher age‐standardized incidence than women (1·05 per 100 000 PY vs. 0·74 per 100 000 PY; P =0·04). Overall age‐specific incidence increased with age at diagnosis, from 0·03 per 100 000 PY in the 35–39 year age group to 17·56 per 100 000 PY in the 85+ year age group. This trend was also observed when comparing men and women, with the highest incidence of MCC (19·96 per 100 000 PY) occurring in men in the 85+ year age group, followed by women in the 85+ year age group (16·42 per 100 000 PY) (Fig. 1). From 2002 to 2011 there was a statistically significant increase in the overall incidence of MCC (P =0·02), which was more pronounced in men (P =0·04) than in women (P =0·21) (Fig. 2). Age‐standardized incidence rate of Merkel cell carcinoma in New Zealand by age at diagnosis and sex. Age‐standardized incidence rate of Merkel cell carcinoma in New Zealand by year of diagnosis. This is the first population‐based study evaluating MCC in NZ. As expected, given NZ's high incidence of UV‐related skin cancers, MCC rates in NZ are among the highest reported in the literature (second only to the state‐wide incidence rate of 1·6 per 100 000 PY in Queensland, Australia), and represent the highest reported national incidence rate of MCC in the world.5,6,10,11,12,13 Furthermore, the incidence of MCC in NZ is increasing, more so in men than in women. This finding is comparable with trends reported in Australian, U.S. and European populations.5,11,13 Reported patient demographics and clinical characteristics from our population add further support to the importance of UVR in the development of MCC, with higher rates of MCC observed in Europeans and on sun‐exposed sites. Interestingly, women were significantly more likely to develop MCC of the lower limb than men (P =0·01), an anatomical distribution previously documented with malignant melanoma in NZ.3 This is thought to be a result of sex differences in patterns of UV exposure secondary to clothing, sun‐seeking behaviours and hair cover.14 Mean age at diagnosis (77 years) is comparable with the current literature, as is age‐specific incidence, with the highest incidence of MCC in the 85+ years age group and men aged 85+ years having the highest overall incidence.1,5,6,11,12,15 Overall, MCC occurred in men and women equally; however, age‐standardized incidence was significantly higher in men, suggesting that when correcting for the fact that women make up a greater proportion of the elderly NZ population, a male predominance exists – a finding consistent with Australian studies.5,6 In NZ, 10·7% of patients presented with distant metastases at diagnosis. This is in line with the population of Queensland, Australia (15%); however, it is much higher than that of most population‐based studies, which range from 4% to 8%.5,6,10,11,12,15 Tumour stage at diagnosis is an important prognostic factor, with more advanced disease having significantly higher mortality rates.2,5,6 Although the overall incidence of melanoma in NZ is more than 40 times higher than that of MCC, MCC is associated with significantly higher mortality rates,1,3 and its incidence in NZ is increasing. Therefore, as is the case with melanoma, addressing NZ's high rates of MCC should be considered a healthcare priority. Further investigation of aetiological factors and treatment outcomes in NZ are warranted. This will not only aid in providing a better understanding of tumour biology, but will also improve outcomes in the diagnosis, management and prevention of MCC.