Dermoscopy, with and without visual inspection, for diagnosing melanoma in adults

Background: Melanomahas one of the fastest rising incidence rates of any cancer. It accounts for asmall percentage of skin cancer cases but is responsible for the majority ofskin cancer deaths. Although history-taking and visual inspection of asuspicious lesion by a clinician are usually the first in a series of ‘tests’to diagnose skin cancer, dermoscopy has become an important tool to assistdiagnosis by specialist clinicians and is increasingly used in primary caresettings. Dermoscopy is a magnification technique using visible light thatallows more detailed examination of the skin compared to examination by thenaked eye alone. Establishing the additive value of dermoscopy over and above visualinspection alone across a range of observers and settings is critical tounderstanding its contribution for the diagnosis of melanoma and to futureunderstanding of the potential role of the growing number of other high-resolutionimage analysis techniques.

Objectives: To determine the diagnostic accuracy of dermoscopy for thedetection of cutaneous invasive melanoma and atypical intraepidermalmelanocytic variants in adults, and to compare its accuracy with that of visualinspection alone. Studies were separated according to whether the diagnosis wasrecorded face-to-face (in-person) or based on remote (image-based) assessment.

Search methods: We undertook a comprehensive search of the followingdatabases from inception up to August 2016: Cochrane Central Register ofControlled Trials; MEDLINE; Embase; CINAHL; CPCI; Zetoc; Science CitationIndex; US National Institutes of Health Ongoing Trials Register; NIHR ClinicalResearch Network Portfolio Database; and the World Health OrganizationInternational Clinical Trials Registry Platform. We studied reference lists andpublished systematic review articles.

Selection criteria: Studies of any design that evaluated dermoscopy inadults with lesions suspicious for melanoma, compared with a reference standardof either histological confirmation or clinical follow-up. Data on the accuracyof visual inspection, to allow comparisons of tests, was included only ifreported in the included studies of dermoscopy.

Data collection and analysis: Two review authors independently extractedall data using a standardised data extraction and quality assessment form(based on QUADAS-2). We contacted authors of included studies where informationrelated to the target condition or diagnostic threshold were missing. Weestimated accuracy using hierarchical summary ROC methods. Analysis of studiesallowing direct comparison between tests was undertaken. To facilitateinterpretation of results, we computed values of sensitivity at the point onthe SROC curve with 80% fixed specificity and values of specificity with 80%fixed sensitivity. We investigated the impact of in-person test interpretation;use of a purposely developed algorithm to assist diagnosis; observer expertise;and dermoscopy training.

Main results: A total of 104 study publications reporting on 103 studycohorts with 42,788 lesions (including 5700 cases) were included, providing 354datasets for dermoscopy. The risk of bias was mainly low for the index test andreference standard domains and mainly high or unclear for participant selectionand participant flow. Concerns regarding the applicability of study findingswere largely scored as ‘High’ concern in three of four domains assessed.Selective participant recruitment, lack of reproducibility of diagnosticthresholds and lack of detail on observer expertise were particularlyproblematic.

The accuracy of dermoscopy for the detection of invasive melanoma or atypicalintraepidermal melanocytic variants was reported in 86 datasets; 26 forevaluations conducted in-person (dermoscopy added to visual inspection) and 60for image-based evaluations (diagnosis based on interpretation of dermoscopicimages). Analyses of studies by prior testing revealed no obvious effect onaccuracy; analyses were hampered by the lack of studies in primary care, lackof relevant information and the restricted inclusion of lesions selected forbiopsy or excision. Accuracy was higher for in-person diagnosis compared toimage-based evaluations (relative diagnostic odds ratio (RDOR) of 4.6; 95% CI2.4, 9.0, P<0.001). Accuracy was compared for (a) in-person evaluations of dermoscopy (26 evaluations; 23,169 lesions and 1664 melanomas) versus visual inspection alone (13 evaluations; 6740 lesions and 459 melanomas) and for (b) image-based evaluations of dermoscopy (60 evaluations; 13,475 lesions and 2851 melanomas) versus image-based visual inspection (11 7/13/2018 #164 Dermoscopy, with and without visual inspection, for the diagnosis of melanoma in adults https://archie.cochrane.org/popups/view.jsp?url=%2Fsections%2Fdocuments%2Fview%3Fdocument%3Dz1501211604482829565727921312… 2/255 evaluations; 1740 lesions and 305 melanomas). For both comparisons, meta-analysis found dermoscopy to be more accurate than visual inspection alone, with RDORs of (a) 4.7 (95% CI: 3.0 to 7.5; P ><0.001).

Accuracy was compared for (a) in-person evaluations of dermoscopy (26evaluations; 23,169 lesions and 1664 melanomas) versus visual inspection alone(13 evaluations; 6740 lesions and 459 melanomas) and for (b) image-basedevaluations of dermoscopy (60 evaluations; 13,475 lesions and 2851 melanomas)versus image-based visual inspection (11 evaluations; 1740 lesions and 305melanomas). For both comparisons, meta-analysis found dermoscopy to be moreaccurate than visual inspection alone, with RDORs of (a) 4.7 (95% CI: 3.0 to7.5; P < 0.001) and (b) 5.6 (95% CI: 3.7 to 8.5; P < 0.001). Theseeffects correspond to predicted differences in sensitivity of (a) 16% (95% CI:8%, 23%) (92% for dermoscopy+visual inspection vs 76% for visual inspection)and (b) 35% (95% CI 24% to 46%) (81% for dermoscopy vs 47% for visualinspection) at a fixed specificity of 80%; and to predicted differences inspecificity of (a) 20% (95% CI 7%, 33) (95% for dermoscopy plus visualinspection vs 75% for visual inspection) and (b) 40% (95% CI 27, 57) (82% fordermoscopy vs 42% for visual inspection) at a fixed sensitivity of 80%.

Using the median prevalence of disease in each set of studies ((a) 12% forin-person and (b) 24% for image-based) for a hypothetical population of 1000lesions, an increase in sensitivity of (a) 16% (in-person) and (b) 35%(image-based) from using dermoscopy at a fixed specificity of 80% equates to areduction in the number of melanomas missed of (a) 19 and (b) 81 with (a) 176and (b) 152 false positive results. An increase in specificity of (a) 20%(in-person) and (b) 40% (image-based) at a fixed sensitivity of 80% equates toa reduction in the number of unnecessary excisions from using dermoscopy of (a)176 and (b) 304 with (a) 24 and (b) 48 melanomas missed.

The use of a named or published algorithm to assist dermoscopy interpretation(as opposed to no reported algorithm or reported use of pattern analysis) hadno significant impact on accuracy either for in-person (RDOR 1.4, 95% CI 0.34,5.6; P=0.17) or image-based (RDOR 1.4, 95% CI 0.60, 3.3; P=0.22) evaluations.This result was supported by subgroup analysis according to algorithm used.Higher accuracy for observers reported as having high experience and for thoseclassed as ‘expert consultants’ in comparison to those considered to have lessexperience in dermoscopy was observed, particularly for image-basedevaluations. Evidence for the effect of dermoscopy training on test accuracywas very limited but suggested associated improvements in sensitivity.

Authors' conclusions: Despite the observedlimitations in the evidence base, dermoscopy is a valuable tool to support thevisual inspection of a suspicious skin lesion for the detection of melanoma andatypical intraepidermal melanocytic variants, particularly in referredpopulations and in the hands of experienced users. Data to support its use inprimary care is limited however it may assist in triaging suspicious lesionsfor urgent referral when employed by suitably trained clinicians. Formalalgorithms may be of most use for dermoscopy training purposes and for lessexpert observers, however reliable data comparing approaches using dermoscopyin-person are lacking.