What criteria reliably distinguish melanoma from benign melanocytic lesions?

The differential diagnosis of melanocytic lesions is fraught with difficulty and a common source of litigation either if a lesion misreported as 'benign' recurs locally or re-presents with nodal metastases or if an atypical naevus is called 'malignant' leading to a cosmetically unsatisfactory wider resection, unwarranted anxiety about prognosis and adverse life insurance prospects. Several authors have claimed that there are valid morphological criteria which, alone or in combination, enable reliable distinction between benign and malignant melanocytic lesions. Others question these criteria and, doubting the extent to which unequivocal diagnoses can be rendered in all cases, believe that the diagnosis is purely subjective and that most diagnostic errors are non-negligent. To address these issues, expert opinions were commissioned from three sets of authors. Okun, Edelstein & Kasznica emphasize that a significant minority of melanocytic lesions are so borderline morphologically that diagnostic uncertainty is allowable and that such uncertainty can be handled responsibly. Kirkham, in favouring the methodical use of criteria, concedes that they are 'largely opinion-based rather than evidence-based, but do go beyond mere subjective pattern analysis'. In agreement with Okun and his colleagues. Slater emphasises that no single feature is reliable by itself and that all aspects, including clinical details, should be interpreted together; he has no hesitation in reporting the diagnosis as 'uncertain' in doubtful cases. In the absence of a specific marker pathognomonic of melanocytic malignancy, the diagnosis will continue to rely on the judicious application of morphological criteria with a small proportion of elusive cases in which diagnostic uncertainty should not be concealed.