Aspects of skin cancer diagnosis in clinical practice

Abstract: Skin cancer incidence is increasing in fair-skinned populations. The three most common skin cancers are basal cell carcinoma (BCC), squamous cell carcinoma (SCC) and malignant melanoma (MM). A correct diagnosis is crucial for an efficient and tailored treatment for the skin cancer patient. The purpose of this thesis was to evaluate different aspects of the preoperative diagnosis of skin cancer. The studies making up this thesis were based on analysis of data from a register including all skin tumour excisions at the Department of Dermatology in Helsingborg, Sweden, from March 2008 to January 2015. The registered data included e.g. sex and age of the patient, tumour site and size, dermoscopic features of the tumour, the preliminary preoperative and final postoperative (histopathological) diagnosis as well as tumour cells at surgical margins. The preliminary preoperative clinical diagnosis was compared with the final histopathological diagnosis in 2,953 excised tumours, whereof 1,626 (55.1%) were malignant, showing high diagnostic accuracy for the diagnosis of malignant tumour and for the diagnosis of basal cell carcinoma (BCC). A total of 96.0% of all excisions had tumour-free margins. The number needed to excise (NNE) for melanoma (the number of pigmented lesions excised to find one melanoma) was calculated for 1,717 cases of excised skin tumours (252 melanomas, 1,395 naevi and 70 seborrhoeic keratoses (SK)). The overall NNE value during the study period was 6.5 (SKs not included). When SKs were included in the calculations the NNE was 6.8. The NNE value decreased with increasing age of the patient and varied for different body locations, with the highest values found for the trunk and the lowest for the arms. When the ABCD rule of dermoscopy was used preoperatively at the bedside in 309 cases (46 melanomas and 263 naevi), use of the algorithm achieved 83% sensitivity and 45% specificity for melanoma diagnosis. A sensitivity of 74% and specificity of 91% were seen for the clinical diagnosis. A considerable percentage (19.6%) of very early melanomas were preoperatively not expected to be melanomas by the dermatologist. The prediction of histopathological subtype of BCC is important for choosing optimal treatment in BCC patients and was assessed in 1,501 cases with pre- or postoperative diagnosis of BCC. The prediction of superficial BCC (sBCC) significantly improved after an educational update on dermoscopic criteria for sBCC in cases assessed by dermoscopy. In conclusion, these studies have shown high accuracy of the preoperative diagnosis of malignant tumour and BCC. With increasing age of the patient, a higher rate of excised pigmented skin lesions was melanomas. Bedside use of the ABCD rule of dermoscopy achieved high sensitivity but low specificity for melanoma diagnosis; however, clinical information seemed to add to specificity. Prediction of sBCC was enhanced after a dermoscopy training session and when dermoscopy was mandatory.