No abstract

Using individual subject data from 10 case-control studies, comprising over 3000 cases and almost 4000 controls, we have estimated the relative risk of melanoma associated with aspects of complexion, namely, hair, eye and skin colour and freckling in adulthood, and have examined the relationships between these factors and naevus count in terms of melanoma risk. Compared with individuals with black or dark brown hair, the relative risks for developing melanoma in those with light brown, blonde and red hair were 1.49 (95% CI 1.31, 1.70), 1.84 (95% CI 1.54, 2.21) and 2.38 (95% CI 1.90, 2.97), respectively. Individuals with blue eyes had a risk 1.55 (95% CI 1.35, 1.78) times that for those with brown eyes, or 1.15 (95% CI 0.94, 1.40) after adjusting for hair colour and freckling in adulthood. The relative risks associated with hair and eye colour were independent of those for naevus count and skin colour. Light skin colour and high freckle density were also highly significant risk factors, independent of each other and of naevus count and hair and eye colour. The risks associated with these factors, while individually modest, are largely independent, and thus pigmentation characteristics and freckling tendency should be useful in identifying high risk groups to be targeted for prevention.

Summary A study of 953 invasive cutaneous malignant melanomas of the head and neck was performed to determine differences between lentigo maligna melanoma and other histogenetic types with regard to patients and sites affected; prognosis was analysed in 595 of these cases. The cases studied comprised all head and neck melanomas registered with the Scottish Melanoma Group between 1979 and 1992, apart from the 3% of cases that were unclassifiable or rare histogenetic types. The histogenetic types of melanoma were 498 (52%) lentigo maligna melanoma (LMM), 237 (25%) superficial spreading melanoma (SSM) and 218 (23%) nodular melanoma (NM). All types increased in incidence throughout the study period. Patients with LMM (mean age 73 years) and NM (mean 68 years) were significantly older than those with SSM (mean 57 years). There were significant anatomical subsite differences related to sex of patients and histogenetic type of melanoma; melanomas on the face were more frequent in females and 90% of LMM occurred at this site, whereas melanomas on the scalp, neck and ears were more frequent in men. Kaplan-Meier estimates of the probability of survival were produced for the 595 of these 953 patients with 5 year follow-up details. In this group of patients the prognostic significance of tumour thickness, Clark level of invasion, ulceration, histogenetic type of melanoma and number of mitoses were studied using stepwise variable selection of procedures. Each of these possible prognostic factors attained individual significance but the tumour thickness was the dominant risk factor in the proportional hazards analysis. When patients were divided into four sex/ulceration subgroups (male/ulcerated, female/ulcerated, male/non-ulcerated, female/non-ulcerated) and analysed by proportional hazards analysis, no variable other than the tumour thickness had any further prognostic effect. Histogenetic type did not remain an independent prognostic variable at this stage. Despite sex and subsite differences, the prognosis for invasive lentigo maligna melanoma does not differ from that for other histogenetic types after controlling for tumour thickness.

There is a need to increase the awareness of skin cancer and safe sun practices among the European population.

Of the 125 patients questioned, only 20 (16%) had received appropriate surgical treatment withini three months of becoming aware of a suspicious cutaneous pigmented lesion; 63 (50%) had received surgical treatment three to 12 months after first noticing such a change, and in 42 cases (34%) this interval was over one year. In 102 of 105 patients in whom the interval exceeded three months the patient alone was responsible for the delay; in only three cases was the family doctor partially at fault. No evidence of delay by the hospital service was identified. Because of these findings a public education campaign was launched in the west of Scotland in June 1985 with the aim of improving recognition of early malignant melanoma. In the next six months the proportion of patients in the west of Scotland with primary melanomas categorised as "thin, good prognosis" had-risen from 38% to 62%, and the proportion with tumours categorised as "thick, poor prognosis" had fallen from 34% to 15%. IntroductionIn a recent extensive survey of incidence and survival statistics in patients with cutaneous malignant melanoma from centres in all parts of the world the overall five year survival rate for over 8000 patients presenting with stage I malignant melanoma (mainly