THE STARBURST GIANT-CELL IS USEFUL FOR DISTINGUISHING LENTIGO MALIGNAFROM PHOTODAMAGED SKIN

Background: Because lentigo maligna (LM) occurs in areas of the body t hat are subjected to long-term UV radiation (UVR), it may be difficult to distinguish atypical melanocytes in LM from the pleomorphic, atypi cal melanocytes in actinically damaged skin. Objective: The purpose of this study was to determine whether the presence of multinucleated me lanocytes would help to make this distinction. Methods: A total of 89 cases of LM were reviewed for the presence or absence of multinucleate d melanocytes and if present, the maximum number of nuclei was recorde d. As controls, 107 elliptical excisions of basal cell carcinoma or sq uamous cell carcinoma were randomly selected. The tips of the ellipses were reviewed for the presence or absence of multinucleated melanocyt es. Results: Multinucleated melanocytes with a ''starburst'' appearanc e, because of their prominent dendritic processes, were present in 85% of LM cases but in only 21% of sun-damaged control specimens (p < 0.0 0001;odds ratio [OR] = 22.6; 95% confidence interval [CI] = 10.6-47.9) . The sensitivity and specificity of starburst giant cells (SGCs) in t he diagnosis of LM was 85% and 78%, respectively. The maximum number o f nuclei per SGC ranged from 2 to 30 in the LM cases (mean, 6.8 +/- 6. 1) and from 2 to 6 in the controls (mean, 2.7 +/- 1.1)(p < 0.001). If only those SGCs with more than two nuclei are considered, 76% of cases but only 8% of controls contained SGCs (p < 0.00001; OR = 35.3; CI = 15.2-81.7). Similarly, 64% of cases and 3% of controls had SGCs with m ore than three nuclei (p < 0.00001; OR = 61.8; CI = 18.1-210.6). Concl usion: The SGC is a useful indicator for the diagnosis of LM. The diag nosis of LM is also more likely as the number of nuclei in SGC increas es.