Dermoscopic Examination of Nail Pigmentation

Background  Diagnosis of longitudinal melanonychia is usually difficult, and neither a single clinical criterion nor a combination of symptoms currently can be used to clearly distinguish malignant from benign bandlike pigmented nail lesions. Biopsy is painful and often leaves definitive dystrophic scars.

Objectives  To describe and evaluate dermoscopic patterns associated with longitudinal nail pigmentation.

Patients and Methods  A total of 148 unselected consecutive cases of longitudinal melanonychia were included over a period of 4 years (20 melanoma, 37 nevi, 16 drug-induced nail pigmentation, 45 nail apparatus lentigo of various types, 8 ethnic-type nail pigmentation, and 22 subungual hemorrhages). All patients were recruited from the dermatology unit outpatient clinic of the Hôtel Dieu de Lyon. All cases were photographed in vivo under oil immersion (dermoscopy). Patterns were recorded prior to final pathologic diagnosis. An independent biostatistics unit performed statistical evaluation using 7 semiologic patterns.

Results  Melanoma cases were significantly associated with a brown coloration of the background and the presence of irregular longitudinal lines (P = .001). Blood spots were mostly observed in subungual hemorrhages (P = .001); however, their presence could not rule out melanoma. Micro–Hutchinson sign was observed only in melanoma, but its rare occurrence did not allow any statistical evaluation of its specificity. Nail apparatus nevi were significantly associated with a brown coloration of the background and the presence of regular lines (P = .001). Nail apparatus lentigo, ethnic-type pigmentation, and drug-induced pigmentation were significantly associated with homogeneous longitudinal thin gray lines and gray coloration of the background (P = .001). Microscopic longitudinal grooves were unspecific, occurred in several conditions, and were associated with any type of ungual discoloration.

Conclusions  We believe that dermoscopic examination of the nail plate in cases of longitudinal melanonychia provides useful information that could help clinicians to more accurately decide if a nail apparatus biopsy should be performed; however, histopathologic diagnosis remains the gold standard in doubtful cases.

EARLY DIAGNOSIS of nail apparatus melanoma in its pigmented form is difficult because several benign conditions share the same clinical features (ie, longitudinal melanonychia).^1-14 Several clinical criteria have been proposed to help clinicians more accurately distinguish suggestive cases,¹⁵ but neither a single clinical criterion nor a combination of symptoms currently can be used to avoid numerous useless painful nail matrix biopsies that may leave dystrophic scars.^16-19 Longitudinal melanonychia is suggestive when it is unique, polychrome, increasing in width, or accompanied by the Hutchinson sign (pigmentation of the cuticle). Also suggestive are onset during adulthood, absence of history of pigmentation-inducing treatment, and ethnic origin not usually associated with nail pigmentation. Indications of probable benign multiple lesions are phototype V or VI skin or lesions present since childhood.¹⁵

The literature on dermoscopic examination of the nails is limited to very few published observations,^20-23 in contrast to the large body of information available on dermoscopy of pigmented lesions elsewhere on the skin.^24-27 The goals of our study were to (1) systematically examine under oil immersion (dermoscopy) a large consecutive cohort of patients referred to our center for pigmented band(s) on nails; (2) try to individualize relevant dermoscopic patterns; and (3) evaluate pattern correlation with 6 different etiologic diagnoses (melanoma, nevus, drug-induced nail pigmentation, nail apparatus lentigo, ethnic-type nail pigmentation, and subungual hemorrhages).

A total of 148 unselected consecutive cases of longitudinal melanonychia were included in this study over a period of 4 years (20 melanoma, 37 nevi, 16 drug-induced nail pigmentation, 45 nail apparatus lentigo of various types, 8 ethnic-type nail pigmentation, and 22 subungual hemorrhages). All patients were recruited from the dermatology unit outpatient clinic of the Hôtel Dieu de Lyon; informed consent was obtained. All cases were photographed under oil immersion with a dermoscopic camera (Heine Dermaphot, Herrshing, Germany). All pictures were taken prior to biopsy, and observed features were recorded without any knowledge of the final pathologic diagnosis. The final diagnosis was based on pathologic examination of a representative nail matrix biopsy specimen as well as records of the patient's medical history (especially pigmentation-inducing drug treatments and repetitive trauma or inflammation of the nail region), ethnic origin, total skin and mucous membrane examination (especially centered on the detection of the various types of lentiginoses such as Laugier-Hunziker disease), and postoperative evolution of the lesion with a minimum follow-up of 6 months. Twenty cases of melanoma were included (all acral lentiginous melanoma), with a mean Breslow thickness of 1.07 mm (range, 0.2-2.1 mm); 2 cases were Clark level I melanomas, all other cases were invasive. Thirty-seven cases were nail apparatus melanocytic nevi; 45 corresponded to nail apparatus lentigo (among which 12 cases were part of a Laugier-Hunziker syndrome and 1 corresponded to Peutz-Jeghers-Touraine syndrome).

Previous observations allowed us to individualize 7 distinct dermoscopic features of longitudinal melanonychia (Figure 1). These features were recorded for all cases prior to biopsy: (1) Blood spots were characterized by proximally well-circumscribed dots or blotches. Their coloration varied from purple-blue in recent lesions to brown in older lesions, and the distal edge was in some cases distorted with a somewhat linear pattern, especially in older lesions (Figure 2). (2) Brown coloration of the background in the area of the clinically observed dark band varied from light to dark with sharply delimited lateral borders. The proximal border corresponded in most cases to the proximal nail plate immediately adjacent to the cuticle; the distal border corresponded to the distal edge of the nail plate (Figure 3, Figure 4, Figure 5, and Figure 6). In most cases this background was associated with either regular or irregular lines. (3) Regular lines, usually associated with the brown background, displayed regularity in coloration, spacing, and thickness; no disruption in their parallelism was observed (Figure 3). (4) Irregular lines were also usually associated with a brown background and showed variations in coloration, spacing, and thickness. Some areas of disruption in parallelism were also observed (Figure 4 and Figure 5). (5) Grayish background in the area of the clinically observed band varied from light to dark (Figure 7). Thin gray lines regular in coloration, thickness, and spacing were usually associated with this gray background (Figure 8). (6) Micro–Hutchinson sign was defined by the visibility on dermoscopy of a pigmentation of the periungual tissues that could not be seen with the naked eye (Figure 6). (7) Microscopic grooves were observed in many cases, not always superimposed on the pigmented area. These grooves were either whitish or grayish and had a linear longitudinal disposition (Figure 9).

Observed features and final diagnoses were cross tablulated at the end of the study, and statistical analysis was performed by an independent biostatistics unit using the χ² test or, in case of a small number of recorded events, Fisher exact test (SAS software; SAS Institute, Cary, NC). Results gave significance scores for all differential diagnoses of melanoma. P values of .05 or lower were considered significant.

The results of all analyses are summarized in Table 1. Twenty cases of nail apparatus melanoma were included in the study. Prominent dermoscopic features observed in melanoma were the association of brown pigmentation of the background (19/20; 95%) with longitudinal brown to black lines irregular in their coloration, spacing, thickness, and parallelism (19/20; 95%) (Table 1, Figure 1B, Figure 1D, Figure 4, and Figure 5). The irregular pattern of the lines was significantly associated with melanoma when compared with all other diagnoses (P = .001) taken either individually or as a group (P = .001 in all 5 differential diagnoses). Melanoma shared with melanocytic nevus the brown coloration of the background. The micro–Hutchinson sign as defined by the visibility of a pigmentation of the cuticle only on dermoscopy (Figure 1F and Figure 6) was observed in only 3 cases, but all were melanoma. This rare feature was also significantly associated with melanoma when compared with all other diagnoses (P = .001). Interestingly blood spots were observed in 1 case of melanoma; their presence, therefore, cannot rule out this diagnosis.

Thirty-seven cases of melanocytic nevus of the nail were included in the study. Prominent observed features were the brown background (37/37; 100%) and the regular pattern of the longitudinal lines (35/37; 95%) (Figure 3). The presence of these lines, regular in their thickness, spacing, coloration, and parallelism, was found statistically sufficient to distinguish nevus from melanoma (P = .001).

Sixteen cases of drug-induced nail pigmentation were included in this study. In most cases the pigmentation was observed on several fingernails or toenails. The 3 most frequently recorded responsible drugs were zidovudine (AZT), hydroxyurea, and minocycline. The main dermoscopic features in these cases were a grayish coloration of the background (15/16; 94%) and the presence of thin longitudinal gray lines with regular thickness, spacing, coloration, and absence of parallelism disruption (Figure 7). These dermoscopic findings were not different from the ones observed in ungual lentigo or ethnic-type pigmentation but significantly differed from melanoma (P = .001).

Forty-five cases of nail apparatus lentigo of various types were included in the study. Most cases were isolated, but 12 cases were part of a Laugier-Hunziker syndrome,²⁸ and 1 corresponded to Peutz-Jeghers-Touraine syndrome. Dermoscopic findings were similar to those observed in drug-induced pigmentation or in ethnic-type melanonychia. Prominent features were the grayish coloration of the background (44/45; 98%) and the presence of thin longitudinal gray lines regular in their coloration, thickness, and spacing (42/45; 93%). The presence of these 2 criteria significantly differentiated nail lentigo from melanoma (P = .001).

Eight cases of ethnic-type pigmentation of the nail were included in the study. In most cases the pigmented longitudinal bands were multiple on fingernails and/or toenails. The patterns in these cases were similar to those previously described in ungual lentigo and drug-induced nail pigmentation but significantly different from those of melanoma (P = .001). The 2 characteristic dermoscopic features of ethnic-type nail pigmentation were the grayish background (7/8; 87.5%) and the thin, regular gray lines (7/8; 87.5%).

In the present study, 22 cases of subungual hemorrhage were included. Only cases with a longitudinal bandlike disposition of the pigmentation were included to evaluate situations in which melanoma could be included in the clinical differential diagnosis of the lesion. Blood spots characterized by a well-limited, rounded proximal edge and a purple to brown coloration were observed in all cases of subungual hemorrhages (22/22; 100%). The distal edge of older lesions showed a somewhat linear distortion that could rarely result in the formation of either regular of irregular lines. It is important to note that, even though the blood spot pattern was significantly associated with subungual hemorrhages when compared with melanoma (P = .001), this easily recognizable pattern cannot be considered to be exclusive, and its presence is not sufficient to rule out melanoma.

Early diagnosis and treatment of melanoma hold out the only possibility of curative treatment.^29-32 Nail apparatus melanoma in its pigmented form is usually revealed by a longitudinal pigmentation of the nail plate or melanonychia striations.^9-14 This clinical presentation is common to several other diagnoses,^33,34 and positive diagnosis is based on histopathologic examination of a biopsy specimen of the nail matrix.³⁵ This procedure is usually painful and often followed by the creation of definitive nail dystrophy due to the surgical injury to the nail matrix.^16-19 Different opinions have been expressed about the management of longitudinal melanonychia.^36-38 Some authors recommend systematic biopsy³⁹; others propose histopathologic diagnosis in suggestive cases only.¹⁵ From the clinical point of view, suggestive lesions have their onset during adulthood, are localized on only the nail, range in color from light brown to black, enlarge progressively, and are associated with a pigmentation of the eponychium called Hutchinson sign.^40-44 In contrast, lesions are more likely to be benign if there is a history of inflammatory disease of the nail or pigmentation-inducing drug intake, if similar lesions are found on several nails, if the onset of the lesion was in childhood,^45-47 if the pigmentation is regular, and if Hutchinson sign is absent.⁴⁸

Dermoscopy is thought to be a useful tool in the diagnosis of skin pigmented lesions. Several algorithms have been proposed to accurately diagnose melanoma, nevi, seborrheic keratoses, pigmented basal cell carcinomas, and hemangiomas on skin.^24-27 Some dermoscopy studies of pigmentation of the nail have been published, but no clear attempt to establish a semiologic classification of such cases has been produced. In their textbook, Stolz et al²² proposed 5 cases of dermoscopic examination of nail pigmentation. Two cases concerned subungual hemorrhages and showed features very similar to those classified as "blood spots" in the present study: the proximal edges of the spots were rounded, whereas the distal edges had a somewhat linear pattern. A third case was classified as mixed bacterial and fungal infection. Our experience in such cases is similar, but we did not include this diagnosis in our study. The fourth case was an invasive acral lentiginous melanoma, and the observed pattern consisted of irregular lines with a brown background. The last case showed vascular changes in scleroderma, which was not the object of our work.

Johr and Izakovic²¹ recently published a report of 4 cases of nail pigmentation observed by dermoscopy. In the first case, the longitudinal melanonychia was associated with an epithelial inclusion cyst. No melanocytic hyperplasia was found, and the dermoscopic pattern was homogeneous and light grayish-brown, somewhat similar to the gray homogeneous pattern we observed in subungual lentigo. Their second case was an invasive acral lentiginous melanoma with a dermoscopic pattern of irregularly thick, spaced, and colored lines with a brown background. The third case was a histopathologically proven nail apparatus nevus, and the authors described a symmetrical disposition of parallel brown lines on dermoscopy. The last case corresponded to a benign possibly ethnic-type nail pigmentation with a grayish homogeneous pattern on dermoscopy. No biopsy was performed. These authors concluded that asymmetry in color and in disposition of the pigmentation suggested high-risk lesions. They proposed that dermoscopic examination be included in the clinical evaluation of longitudinal melanonychia prior to biopsy and suggested that biopsy be performed only in high-risk lesions. They also suggested that dermoscopic follow-up could help in the proper management of such cases.

Kawabata and coworkers²⁰ examined by dermoscopy 6 cases of nail apparatus melanoma and 18 cases of ungual nevi. Melanoma cases were characterized by an irregular pattern of the pigmented lines, whereas nevi showed a more regular pattern of pigmented lines. Our findings are therefore very similar to theirs regarding the pigmentation of the nail plate. These authors emphasize the semiologic value of the irregular pigmentation of the periungual tissues (Hutchinson sign) for the diagnosis of melanoma; they found this pigmentation in all 6 of their cases. However, in our cases, we could not calculate the significance score of this pigmentation because of its relative rarity. Members of our research group²³ recently described a distinct dermoscopic pattern with a rounded blue homogeneous coloration in a subungual blue nevus. However, this observation cannot be compared with the cases included in the present study because the lesion was not clinically characterized by a longitudinal melanonychia.

In the present study, our data indicate that brown coloration of the background is associated with prominent melanocytic hyperplasia either in subungual nevi or in melanoma. In contrast, lesions associated with mild hyperplasia of melanocytes (lentigo of various types) or hyperpigmentation of the epithelium without melanocytic hyperplasia (drug-induced or ethnic-type pigmentation) are characterized by a grayish coloration of the background. The most powerful criterion suggestive of melanoma was the irregularly colored, spaced, and thick pattern of the lines with areas of disruption of their parallelism. We confirmed that subungual hemorrhages are easily recognizable by their blue to black structureless blood spots with a rounded proximal edge. We would like to stress that the presence of such blood spots alone cannot rule out melanoma; the absence of other features, especially lines, must be confirmed to do that.

Our data confirm the previously expressed opinion that dermoscopy can help clinicians accurately decide if a biopsy of the nail apparatus is necessary in cases of longitudinal melanonychia. We did not demonstrate that dermoscopic follow-up of patients with nail pigmentation would be of any value, but we believe that further work should be done to clearly define criteria for high-risk lesions in cases of change in the dermoscopic pattern during follow-up.

In conclusion, our findings indicate that dermoscopy should be included in the clinical evaluation of longitudinal nail pigmentation. We believe that irregularly thick, spaced, parallel, and colored lines associated with a brown background are strongly suggestive of melanoma. We believe that careful examination under oil immersion of the nail in some cases of longitudinal melanonychia could avoid unnecessary painful surgery. Dermoscopy, of course, adds new criteria for diagnosis of ungual pigmentation, but it does not replace histopathologic diagnosis, and biopsy should be performed in cases of doubtful lesions. Further studies are needed to evaluate the role of dermoscopic examination of the nails in the clinical follow-up of patients with nail pigmentation.

Accepted for publication April 16, 2002.

This work has been supported in part by the Insitut National pour la Recherche Médicale, INSERM U346, and by research grants (appel d'offres 2000) from the Hospices Civils de Lyon (Dr Thomas), Lyon, France.

This study was presented in part at the First International Dermoscopy Meeting, Rome, Italy, February 24, 2001.

Corresponding author and reprints: Luc Thomas, MD, PhD, Dermatology Unit, Hôtel Dieu de Lyon, 69288 Lyon CEDEX 02, France (e-mail: luc.thomas@chu-lyon.fr).