Dermoscopy for the Family Physician

Noninvasive in vivo imaging techniques have become an important diagnostic aid for skin cancer detection. Dermoscopy, also known as dermatoscopy, epiluminescence microscopy, incident light microscopy, or skin surface microscopy, is performed using a handheld instrument called a dermatoscope or dermoscope, which has a transilluminating light source and standard magnifying optics (10×). The dermatoscope facilitates visualization of subsurface skin structures located within the epidermis, dermoepidermal junction, and papillary dermis, which are otherwise not visible to the unaided eye.^1,2

Traditionally, dermoscopy has been taught to and used by dermatologists; however, it is gaining popularity in other fields of medicine, including primary care.^3–5 Because many patients are seen routinely in the primary care setting, family physicians play an important role in screening for skin cancer and early detection of skin cancer. Dermoscopy has been shown to improve diagnostic accuracy, sensitivity, and specificity for skin cancer diagnosis by dermatologists.⁶ Once thought to be a subspecialist tool, the literature has shown that family physicians can also improve their diagnostic accuracy and triage ability by using dermoscopy.^3–5,7–11

During naked eye examinations of the skin, much of the light transmitted toward the skin is reflected off the stratum corneum, thus precluding the observer from seeing structures located below this layer. Dermatoscopes illuminate the skin via the use of light-emitting diode bulbs, with or without the use of polarizing filters. Dermatoscopes using nonpolarized light require direct contact between the skin and the scope, and require a liquid interface, such as ultrasound gel or alcohol, to be placed between the skin and glass plate of the dermatoscope.¹²

Once the liquid interface has been applied to the skin, the scope can be placed on top of the lesion and gently pressed against the skin; it is imperative that enough pressure be applied to eliminate air bubbles. The liquid interface prevents light from being reflected off the stratum corneum and improves refraction, thereby allowing the visualization of structures below the stratum corneum. Dermatoscopes equipped with cross-polarized filters obviate the need for direct skin contact and a liquid interface.¹³ Cross-polarized filters eliminate the glare off the skin surface and increase light refraction, allowing the observer to visualize deeper skin structures.

Three types of dermoscopic techniques are currently used: (1) classic or standard contact; (2) polarized contact; and (3) polarized noncontact (Table 1^2,13 ). Studies have demonstrated that polarized and nonpolarized dermoscopy provide complementary information.^14,15 Polarized dermoscopy may have a higher sensitivity for detecting skin cancer based on its ability to enhance the visualization of vascular and crystalline structures, both of which are commonly seen in skin cancer^15,16 (Figure 1). Nonpolarized dermoscopy improves specificity because it permits easier visualization of other structures commonly seen in benign lesions, such as milia-like cysts in seborrheic keratosis.

Dermoscopy can enhance a clinician's ability to detect skin cancer. However, improvement in diagnostic accuracy is contingent on acquiring dermoscopy training. Without formal training, the use of dermoscopy may result in poorer performance compared with naked eye examination.¹⁷ Primary care physicians trained in dermoscopy can reduce their referral rate or benign-to-malignant excision ratio (from 9.5 to 3.5).⁵ In addition, primary care physicians trained in dermoscopy improve their ability to identify skin lesions suggestive of skin cancer compared with naked eye examination alone (76% to 79% vs. 54%).^3,4 In other words, in primary care, dermoscopy improves sensitivity for the diagnosis of melanoma and improves the ability to recognize suspicious lesions that require biopsy, without significantly changing specificity (71%).^3–5 Dermoscopy also has been shown to improve the diagnostic accuracy for nonmelanocytic lesions, such as basal cell carcinoma.¹⁸ The indications for dermoscopy are listed in Table 2,^6,16,19,20 and the advantages and limitations are listed in Table 3.^{3,6,14,20–32}

The visualization of structures within the epidermis and papillary dermis using dermoscopy has generated new terminology and clinical criteria for cutaneous lesions.³³ It is imperative that physicians using dermoscopy learn and become proficient at identifying the colors and structures, because they are required in generating a correct diagnosis.

Colors visible under dermoscopy depend on the quantity and location of keratin, blood, collagen, and melanin. Accordingly, colors visualized include yellow, red, and white for keratin, blood, and collagen, respectively. Colors for melanin range from black and brown to gray and blue, depending on the location of melanin, in terms of depth, within the skin layers.³⁴

Structures visible under dermoscopy have histopathologic correlates (see Appendix below). The presence of specific structures permits the classification of a lesion as a melanocytic tumor (i.e., melanocytic nevus or melanoma). Common structures visualized in melanocytic lesions are provided in Table 4^{21,22,35–39} and Figure 2. When a lesion has been determined to be melanocytic, the next step is to determine if it is a nevus or a melanoma. Melanocytic nevi tend to manifest one of the symmetric patterns depicted in Figure 3, and melanomas reveal at least one of the 10 melanoma-specific structures provided in Table 5.

Nonmelanocytic lesions, including basal cell carcinoma, seborrheic keratosis, hemangioma, and angiokeratoma, also demonstrate specific structures under dermoscopy. Pigmented and nonpigmented basal cell carcinomas can reveal vascular structures such as arborizing blood vessels^36,38 (Table 4^{21,22,35–39} ; Figure 4). In addition, pigmented basal cell carcinoma can demonstrate pigmented structures, including leaf-like structures, spoke wheel–like structures, ovoid nests, and nonaggregated blue-gray globules.

Seborrheic keratosis is one of the most common benign epidermal tumors. Clinically, these lesions are characterized by well-circumscribed plaques or papules with pseudohorn cysts. They range from light brown to dark brown or black. Dermoscopically, seborrheic keratoses display multiple milia-like cysts, comedo-like openings, gyri and sulci, fingerprint-like structures, and moth-eaten borders (Table 4^{21,22,35–39}; Figure 5).

Cherry hemangiomas are bright red to violaceous papules that consist of dilated capillaries. Similarly, angiokeratomas are dark violaceous to black papules that are often keratotic or firm on palpation. Dermoscopically, hemangiomas and angiokeratomas are distinguished by the presence of well-demarcated lacunae that can be red, maroon, blue, or black ³⁶ (Table 4^{21,22,35–39}; Figure 6).

Some lesions, particularly amelanotic or hypomelanotic tumors, cannot be classified based on the aforementioned features. It is important to search for the presence of blood vessels in these lesions. If present, the blood vessel morphology can assist in making the correct diagnosis. In most amelanotic and hypomelanotic lesions, the only clue to the diagnosis depends on evaluating the blood vessel morphology and distribution, whereas in pigmented lesions (as previously mentioned), the analysis of vascular structures provides secondary clues to the diagnosis.^21,40

The easiest way to visualize blood vessels is to use a noncontact polarized dermatoscope. If a contact dermatoscope is used, it is beneficial to use ultrasound gel as the liquid interface, because the gel acts as a cushion that minimizes the amount of pressure applied to the skin, which in turn prevents blanching of the vessels.²¹ Although it is important to analyze blood vessel morphology, especially in amelanotic lesions, it is beyond the scope of this article to review all the vessel morphologies; readers should refer to the referenced articles for more information.^21,22

The two-step dermoscopy algorithm is the foundation for dermoscopic evaluation of skin lesions, and it guides the observer through the diagnosis and management decision-making process³⁶ (Figure 7).

FIRST STEP: MELANOCYTIC VS. NONMELANOCYTIC LESIONS

The first step of the algorithm is based on a seven-level criterion ladder, which is intended to help differentiate melanocytic lesions from the following nonmelanocytic lesions: dermatofibroma, basal cell carcinoma, seborrheic keratosis, and hemangioma³⁶ (Table 4^{21,22,35–39} ; Figure 7). Even if the two-step algorithm cannot reliably differentiate between melanocytic and nonmelanocytic lesions, it has been structured to ensure that cutaneous malignancies will not be missed.^36,39 For example, occasionally basal cell carcinoma and melanoma can manifest overlapping features that may lead to the incorrect classification of some basal cell carcinomas as melanocytic (i.e., melanoma) and some melanomas as nonmelanocytic (i.e., basal cell carcinoma). Despite this potential error, the concern for a cutaneous malignancy will remain high, and the management decision to perform a biopsy will not be altered.

SECOND STEP: BENIGN VS. MALIGNANT MELANOCYTIC LESIONS

The second step of the algorithm is intended only for the evaluation of melanocytic lesions, including common acquired melanocytic nevi, blue nevi, Spitz nevi, atypical nevi, and melanoma. In essence, the second step is intended to help differentiate nevi from melanoma.³⁶ To accomplish this, numerous quantitative and qualitative methods have been created.^34,41–45 In particular, scoring systems have proven to be relatively simple, accurate, and reproducible methods for diagnosing melanoma.^6,19,44

The three-point checklist is considered the simplest method to learn and use, and has the highest sensitivity for identifying melanoma³⁶ (Figure 8). It is intended as a screening algorithm for detecting skin cancer (melanoma and pigmented basal cell carcinoma) and applies only to pigmented skin lesions. One point is assigned to each of the following criteria present in the lesion⁴³:

• Asymmetry in distribution of dermoscopic color and/or structures in one or two perpendicular axes. The contour or silhouette of the lesion does not factor into whether the lesion is symmetric or not.
• Irregular or atypical pigment network consisting of thick lines and irregular holes.
• Blue-white veil and/or white scar-like depigmentation and/or blue pepper-like granules.

A total score of 2 or 3 is considered positive, and the lesion should be biopsied or the patient referred for further evaluation.⁴³ Given its simplicity and high sensitivity for detecting pigmented skin cancer, the three-point checklist may be ideally suited for clinicians with little experience in dermoscopy and for use as a screening tool in the primary care setting.¹⁹ Atypical pigment network and blue-white structures have previously been defined (Figure 2; Table 5; Appendix). Dermoscopic asymmetry requires further clarification, because it differs from the conventional way that clinicians define asymmetry. The conventional view of asymmetry, as in the well-known ABCD mnemonic for melanoma (asymmetry, border irregularities, color variation, diameter [greater than 6 mm]), factors in the contour or shape of the lesion. In contrast, symmetry or asymmetry in dermoscopy does not factor in the contour or shape, but rather the distribution of colors and structures within the lesion (Figure 8). For example, using the conventional definition of symmetry, a heart-shaped lesion that reveals only globules would be considered asymmetric in one axis; however, under dermoscopy, this lesion would be considered completely symmetric in both axes.

Pattern analysis is a qualitative method that evaluates dermoscopic structures and their distribution. It requires the ability to recognize, and experience in recognizing, benign patterns^46–48 (Figure 3). It also requires the ability to recognize the 10 melanoma-specific structures listed in Table 5. Any melanocytic lesion that deviates from one of these benign patterns, while also revealing at least one of the 10 melanoma-specific structures, should be biopsied to rule out melanoma.

From a management perspective, the two-step algorithm (Figure 7) is intended to guide the decision-making process on whether to perform a biopsy, or to refer or reassure the patient.⁴⁹

If the lesion is considered benign, the patient can be reassured, provided education on skin self-examination, and instructed to return if any changes or new lesions develop.^25,50 If the lesion is suspicious for melanoma or basal cell carcinoma, it should be biopsied.⁵¹ If the lesion is considered suspicious (i.e., the clinician is uncertain of the diagnosis), it can be biopsied or the patient can be referred for another opinion.

All of the criteria discussed in this article can be applied during the evaluation of any lesion on any area of the skin. However, there are additional criteria that are specific to mucosal, volar, and facial skin lesions. For discussion of the specific features of lesions on these areas, refer to the Atlas of Dermoscopy⁵² or other online tutorials on dermoscopy, such as http://www.dermoscopy-ids.org/index.php/education/podcasts and http://www.genomel.org/dermoscopy.

Dermoscopy is a useful, easy-to-use, and non–time-consuming technique that increases diagnostic accuracy for pigmented and nonpigmented skin lesions⁵³; in particular, it increases the sensitivity for detecting melanoma and basal cell carcinoma. In addition, dermoscopy has been shown to improve specificity by decreasing the number of biopsies of benign lesions, such as nevi, seborrheic keratosis, and hemangioma, that may clinically mimic malignancy.²⁸

Data Sources: A PubMed search was completed in Clinical Queries using the key terms dermoscopy, dermatoscopy, epiluminescence microscopy, skin surface microscopy, and incident light microscopy. The search included meta-analyses, randomized controlled trials, clinical trials, reviews, and case series.

EDITOR'S NOTE: At the time of initial submission, Dr. Marghoob did not list any disclosures on AFP's Conflict of Interest form, which asks for financial relationships with commercial entities that might have an interest in the topic. During the final stages of production, we discovered that Dr. Marghoob had the following relationships with makers of dermoscopes, which we agreed should be disclosed: (1) receiving dermoscope prototypes for testing from the four major manufacturers, and providing unpaid feedback and advice about these devices; (2) receiving honoraria for speaking on the topic of dermoscopy at meetings funded in part by makers of dermoscopes (however, Dr. Marghoob is not on any speakers' bureaus); and (3) being an investigator in Institutional Review Board–approved research projects funded by the National Institutes of Health and Melanoma Research Alliance, some of which partnered (at least to some extent) with companies that produce dermoscopes (however, Dr. Marghoob does not receive any compensation from this grant funding). Given the stage at which these conflicts came to our attention, we performed an internal review of the manuscript and disclosures, and ultimately decided that the manuscript provided an unbiased and nonpromotional description of this technique. In addition, Dr. Marghoob agreed to not enter into any relationships with a maker of dermoscopes that used his AFP article for any presentations on dermoscopy for at least 12 months after the article's publication. For these reasons, we decided to continue with publication. However, to be clear, relationships like these would generally be disqualifying according to our conflict of interest policy (https://www.aafp.org/journals/afp/authors/guide/coi.html).