Diagnostic tests
Scanning for melanoma
- Elisabeth M Wurm, H Peter Soyer
- Aust Prescr 2010;33:150-5
- 1 October 2010
- DOI: 10.18773/austprescr.2010.070
Non-invasive diagnostic tools aim at increasing accuracy of melanoma diagnosis. Clinical naked eye observation in combination with dermoscopy can be regarded as the practical reference standard to identify lesions for histopathological evaluation. Pigmented lesions need to be evaluated in the context of patient history to identify risk factors for melanoma, followed by a dermoscopically-aided entire skin examination. Patients with identified risk factors should be further examined.Total body photography is widely used in the follow-up of high-risk patients (particularly those with numerous and dysplastic naevi) and can be coupled with digital dermoscopy or videodermoscopy. New non-invasive diagnostic aids comprise multispectral image analysis, reflectance confocal microscopy and computer assisted diagnostic systems. Also, molecular profiling of lesions is an emerging technique under investigation for melanoma diagnosis.
Early detection of melanoma remains a significant challenge for clinicians. The critical issue is to remove all lesions that may be malignant while minimising the excision rate of harmless benign lesions. Since naked eye examination has a comparatively low sensitivity in melanoma detection, additional non-invasive diagnostic tools such as dermoscopy are being used in daily practice and have improved the sensitivity of diagnosis when applied by experts.1–3 The current diagnostic gold standard is visual inspection with dermoscopy followed by histopathological examination as required. A high number of unnecessary surgical procedures are still performed. A recent report dealing with primary skin cancer care in Queensland showed that 19.6 pigmented lesions are excised per melanoma.4
Several new non-invasive diagnostic tools aimed at increasing the accuracy of skin cancer diagnosis and thereby minimising unnecessary surgical procedures have emerged in recent years (Table 1). This expanding choice of diagnostic tools may cause confusion among doctors about what they are and how they can be used. Most systems offer a combination of diagnostic methods which may add to the uncertainty.
A patient history to identify risk factors for melanoma as well as a full body examination aided by dermoscopy should be performed for every new patient. Since further evaluation is time-consuming, those individuals at risk, as well as lesions that are considered as atypical or suspicious, should be identified. A detailed history should include:
An inspection of the entire cutaneous surface should include the axillae, groin, the interdigital webs of the hands and feet, the nail apparatus and the scalp.
The ABCD acronym – Asymmetry, Border irregularity, Colour variegation, and large Diameter – supplemented with an E for Evolution, represents the clinical guideline for melanoma diagnosis. In contrast, the EFG acronym – Elevated, Firm and Growing progressively – is more appropriate for nodular melanomas that often have a more subtle clinical appearance.7On physical examination, new and changing naevi should be detected as well as any 'ugly ducklings', that is, lesions that are dissimilar to the rest.
However, any one single visual inspection fails to detect small melanomas and amelanotic melanomas. Thus for high-risk individuals, six-monthly full cutaneous examinations supported by total body photography and dermoscopy as well as patient education for self-examination have been recommended by Australian guidelines.8
Total body photography is widely used in the follow-up of high-risk patients, particularly those with numerous and dysplastic naevi. The technique can be performed with any camera, and a standard digital camera that provides good image quality for digital sectional body images is the most cost-effective option. To document nearly the entire body surface the patient should assume standardised positions under good light conditions. Images should be taken of the face, neck, area behind the ears, scalp (in bald individuals), anterior and posterior trunk, and the extremities (including palms and soles). Subsequent new or changing lesions that may be indicative of melanoma can be recognised in follow-up examinations by comparing the images with the patient's skin. Specific digital skin photography systems are available which facilitate standardisation of imaging and data storage (Table 2).
Total body photography has been reported to enable melanoma detection at an early stage.9 However, small changes in naevi will probably be missed when only applying macroscopic imaging. A combined dermoscopic and total body photography approach is therefore recommended for patients who have atypical moles. Whereas digital dermoscopic images can be obtained with dermoscopic lenses that can be attached easily to most commercially available cameras, various skin imaging devices offer a combination of total body photography and dermoscopy (Table 2). Some devices are also able to automatically compare two overview images and highlight new and changing lesions on the screen, though large-scale clinical studies of a high-risk population are needed to validate these findings.
Dermoscopy, also known as dermatoscopy or epiluminescence microscopy, has been widely adopted into everyday clinical use. It enables visualisation of subsurface anatomic structures of the epidermis and upper dermis. A dermoscope consists of a light source and a magnifying lens. While non-polarised dermoscopes require operation with an immersion medium, such as oil or alcohol, dermoscopes with polarised light do not.
Digital dermoscopy or videodermoscopy is also now widely used. As well as easy storage and retrieval, digital dermoscopic and clinical images can be sent electronically. This is called teledermatology or teledermoscopy (Table 1).
Numerous diagnostic algorithms have been proposed to assess a lesion including pattern analysis, ABCD rule, Menzies method, seven-point checklist and three-point checklist. All these algorithms have been proven to be of high specificity and sensitivity in the diagnosis of melanoma. The choice of which one to use should be made upon personal preferences.*
Dermoscopy, when performed by specialists, increases diagnostic sensitivity without diminishing specificity and has been shown to decrease unnecessary excisions.1-3,10 Two meta-analyses on studies published before 2000 verified that dermoscopy is superior to naked eye examination when used by experts.1,2
Another study reported that even a one-day tutorial on dermoscopy can improve the ability of primary care physicians to correctly refer individuals with suspicious lesions to a skin lesion clinic.11A recent meta-analysis focused exclusively on studies that were performed in a clinical setting and found the relative diagnostic odds ratio (a measure for diagnostic accuracy) was 15.6 for dermoscopy compared to naked eye examination.3This strong scientific evidence indicates that dermoscopy is presently the practical reference standard for non-invasive diagnosis of melanoma.
Due to the impracticability of removing all lesions, follow-up is crucial so that melanomas which lack atypical features at the first visit are not missed. Suspicious lesions can be monitored by serial dermoscopic and macroscopic imaging. Digital dermoscopic (and clinical) images are taken and linked to the body site via a computer. At the follow-up visit, the same lesion is photographed again for comparison. This is especially useful for patients with multiple lesions, and reportedly improves sensitivity in melanoma diagnosis.3,12 Re-examination after three months with subsequent follow-up visits every 6–12 months seems to be a useful strategy. For individuals with familial atypical multiple mole and melanoma syndrome, follow-up every three months is recommended.13
A major disadvantage of the method is, however, that only preselected lesions are monitored, whereas changes in a previously unsuspicious lesion or ade novolesion might be missed. Follow-up should never be performed in nodular lesions, because if they are malignant they tend to grow faster than other melanoma types. Even short delays in treatment might increase the risk of a poor prognosis.
Digital dermoscopic imaging enables primary care physicians to forward dermoscopic images (together with clinical information and macroscopic images) to specialists for a second opinion.14 Studies have shown good agreement between face-to-face diagnosis and diagnosis based on digital images.15,16 This is especially useful in remote areas, where referral is associated with considerable healthcare costs, and time for the patient.
Modern skin imaging devices combine dermoscopy and total body photography with teledermoscopic networks and computer-assisted automated diagnosis (Table 2). The company MoleMap, established by dermatologists, for example, offers a system in which a detailed examination followed by total body photography and comprehensive dermoscopic image capture is obtained by a specifically trained nurse. This information is then sent electronically to a dermatologist for expert analysis.
Multispectral imaging relies on the principle that light of different wavelengths, of the visible and infrared spectrum, penetrates the skin to different depths. When coupled with computer-based analysis, certain features not visible in macroscopic and dermoscopic analysis can be visualised.
High-frequency ultrasound provides a vertical image of the skin based on its different acoustic properties. However, because of the limited resolution, ultrasound alone is not a reliable diagnostic aid. It is more appropriately used for preoperative management in dermatology, for example, in assessing tumour thickness and vascularity.
Optical coherence tomography is comparable to ultrasound, however it uses light instead of sound waves. It has better resolution than ultrasound but only penetrates to a depth of up to 1 mm, which approximately corresponds to the reticular dermis.
The resolution of optical coherence tomography does not reach the capabilities of reflectance confocal microscopy or histopathology, however cellular details can be viewed with the more modern devices. Although there are studies regarding the various features of skin cancer, reports of the diagnostic accuracy of optical coherence tomography are lacking. It seems that this technique might also play a role in other skin diseases in the future, such as contact dermatitis, psoriasis and bullous diseases, as well as monitoring of topical treatment.17
Confocal laser scanning microscopy can be operated in fluorescence or reflectance mode, but reflectance confocal microscopy is more suitable for clinical applications. Reflectance confocal microscopy allows visualisation of the epidermis and papillary dermis at a quasi-histological resolution. Horizontal sections of a lesion can be scanned and viewed using a near-infrared laser. This method is ideally suited for melanoma diagnosis as melanin provides strong contrast and is easily visualised. Diagnostic algorithms for melanoma detection have been proposed and show improved diagnostic specificity and sensitivity.18–20 Furthermore, a glossary of terms commonly used in reflectance confocal microscopy has been published.21 This type of microscopy has also been used in non-melanoma skin cancer, Mohs surgery,in vivosurgical margin assessment and in follow-up of response to topical treatment. However, large-scale clinical studies are needed to assess the method's full clinical potential.
Multiphoton laser scanning microscopy works with a near-infrared laser beam which excites endogenous fluorophores. Nicotinamide adenosine dinucleotide phosphate (NADPH) is the primary source of autofluorescence. Like reflectance confocal microscopy, the multiphoton laser scanning microscopy provides horizontal sections of the skin allowing visualisation of cellular and subcellular structures. To date, it is mainly used as a research tool, rather than clinically.
Automated diagnostic systems extract and analyse features of skin lesions and give a diagnosis. They have been shown to reach comparable levels of diagnostic specificity and sensitivity to that of expert dermatologists.22 To date, a few fully automated systems are available, some of which are integrated in the software of videodermoscopy devices (Table 2). MelaFind uses multispectral imaging information from dermoscopic images. The MelaFind system is currently in the final stages of being granted US Food and Drug Administration approval and is anticipated to be available in the not too distant future. There is a tendency of these tools to over-diagnose melanoma. Further studies are required to assess the impact of automated instruments against human performance in the clinical field.
Molecular profiling is an emerging technique in melanoma diagnosis. A method that analyses RNA acquired from tape stripping of a suspicious melanocytic lesion is currently under investigation.
Although newer imaging techniques hold great promise, they cannot replace visual inspection and patient examination. Clinical naked eye observation in combination with dermoscopy can be regarded as the practical reference standard to identify lesions for excision. Histopathological analysis of lesions remains the gold standard in skin cancer diagnosis.
It is important to emphasise that pigmented lesions need to be evaluated in the context of a patient's entire skin examination. Although a general practitioner may easily make the decision to excise a suspicious lesion, there are a few clinical situations where a dermatologist's advice should be sought and further evaluation be performed. These include high-risk patients with multiple (atypical) naevi or naevi on specific anatomical locations such as palms, soles of the feet and under the nails, or on the genitals.
New non-invasive imaging techniques have great potential for monitoring lesion growth and response to treatment, as well as true margin assessment before surgery.
*A detailed description of these algorithms can be found at www.dermoscopy.org/consensus/tutorial.asp
Dermoscopy books, recommended for beginners:
Menzies SW, Crotty KA, Ingvar C, McCarthy WH. An atlas of surface microscopy of pigmented skin lesions: dermoscopy. 2nd ed. Roseville: McGraw-Hill Australia; 2003.
Stolz W, Braun-Falco O, Bilek P, Landthaler M, Burgdorf WHC, Cognetta AB. Color atlas of dermatoscopy. 2nd ed. Berlin: Blackwell Publishing; 2002.
Johr R, Soyer HP, Argenziano G, Hofmann-Wellenhof R, Scalvenzi M. Dermoscopy: the essentials. Edinburgh: Mosby; 2004.
Malvehy J, Puig S, editors. Principles of dermoscopy. Barcelona: Cege Editors; 2002.
Dermoscopy books, further reading:
Marghoob AA, Braun R, Kopf AW, editors. Atlas of dermoscopy. London: Parthenon Publishing; 2004.
Soyer HP, Argenziano G, Hofmann-Wellenhof R, Johr R. Color atlas of melanocytic lesions of the skin. Berlin: Springer; 2007.
Micali G, Lacarrubba F. Dermatoscopy in clinical practice. London: Informa Healthcare; 2009.
Tosti A. Dermoscopy of hair and scalp disorders with clinical and pathological correlations. Bologna: Informe Healthcare; 2008.
Malvehy J, Puig S, Braun, RP, Marghoob AA, Kopf AW. Handbook of dermoscopy. London: Taylor & Francis; 2006.
Online resources:
International Dermoscopy Society
www.dermoscopy-ids.org
Educational website on dermoscopy
www.dermoscopy.org
Dermoscopy atlas by the Skin Cancer Society of Australia and New Zealand
www.dermoscopyatlas.com
Blog dedicated to dermoscopy
www.dermoscopic.blogspot.com
Educational web site for the Australian Dermoscopy Diploma
www.dermoscopyconsult.com
International Dermoscopy Diploma by the University of Graz, Austria
www.meduni-graz.at/IDD
University of Queensland School of Medicine. Master of Medicine (Primary Care Skin Cancer Medicine)
www.skincancermasters.com
The International Community for Teledermatology, Dermatology
www.telederm.org
Professor Soyer is co-founder and shareholder of e-dermconsult GmbH, a spin-off company of the Medical University of Graz, Austria, with emphasis on holistic solutions for teledermatology. He is also shareholder and consultant for MoleMap Australia by Dermatologists Pty Ltd.
Research Assistant, Dermatology Research Centre, The University of Queensland School of Medicine, Princess Alexandra Hospital, Brisbane
Chair, Dermatology Research Centre, The University of Queensland, School of Medicine, Princess Alexandra Hospital, Brisbane