Limitations in dermoscopic assessment of pigmented auricular lesions: a single-center retrospective study

Introduction

Dermoscopy is an important tool for diagnosis pigmented skin lesions, however, it has some limitations, especially in the assessment of special site locations, such as the face, acral regions or mucosa. Cutaneous microanatomy varies depending on the anatomic region, which translates into different histopathological and dermoscopic presentations according to the lesion’s location. This structural diversity obliges clinicians to acquire specific anatomical knowledge for early and accurate diagnosis of site-specific melanomas [1]. However, some researchers challenge the concept of site specificity in skin lesions, highlighting the fact of significant overlap in the histopathological appearance of melanocytic lesions regardless of their location [2]. Nevertheless, histopathological examination of special sites’ lesions possesses a significant challenge and requires particular attention from pathologists. A retrospective study of 21 auricular benign melanocytic lesions conducted by Saad et al. revealed pagetoid spread and cytologic atypia in approximately half of the analyzed cases [3]. Moreover, melanomas in special site locations are rare and therefore, uncommonly observed in daily practice [4]. However, several algorithms dedicated to these special regions have been developed in the recent years, i.e., the BRAAFF checklist for diagnosis of acral melanoma or dermoscopic inverse approach to examine flat pigmented facial lesions [5, 6].
The auricle is a relatively poorly studied anatomical area in terms of dermoscopic spectrum. Few articles that focused on this topic show that melanocytic lesions in this region may show a combination of features observed in facial and extra-facial melanomas, however no dedicated algorithm has been created so far [7–9].
In contrast to scant literature data, a recent prospective study by Rishpon et al. showed that almost 30% of patients attending a pigmented lesion clinic had at least one pigmented auricular lesion [9].

Objective

The aim of this study was to identify dermoscopic features of pigmented lesions located on the external ear, which were initially considered suspicious based on known dermoscopic screening algorithms.

Material and methods

A retrospective study was conducted on patients admitted to the Department of Dermatology, Venereology and Allergology between January 1, 2017 and December 1, 2023 due to pigmented lesions located on the external ear. The MedStream Designer™ software was used to search the database and extract the clinical data.
Videodermoscopic examinations were performed using a Medicam 800 HD camera attached to a FotoFinder Vexia videodermoscope (20× magnification; non-polarized light). Two clinicians (J.Ż and M.Sł.) independently reviewed dermoscopic images using 5 dermoscopic algorithms: ‘Chaos and clues’, 7-point checklist, CASH algorithm, Menzies method, and 3-point checklist [10–14]. Additionally, the lesions were assessed for the presence of facial melanoma criteria [15, 16]. The dermoscopists were blinded to the final diagnosis, and any discrepancies were resolved by a third evaluator (M. So.).
Pigmented auricular lesions fulfilling at least one of the preselected dermoscopic screening algorithms were included in the study. Lesions located outside the external ear and lesions exceeding the size of the imaging area of the videodermoscopic camera were excluded from the analysis. Histopathological examination was conducted to rule out melanoma and confirm benignity.

Results

Six patients (3 females and 3 males) with histopathologically confirmed benign melanoma mimickers located on the external ear were evaluated. The mean age of analyzed patients was 50.5 years (range: 22–68 years). The Fitzpatrick skin phototype of the included patients was as follows: II – 4 patients; III – 2 patients. Two patients had a history of skin cancer (2/6; 33.3%) and one subject had a positive family history of malignant skin tumors. Two lesions were located on the helix and one each was located on the crus of the helix, scaphoid fossa, lobule, and the posterior aspect of the earlobe. Clinically, most of the lesions presented as pigmented macules (4/6; 66.7%) and two as pigmented papules (2/6; 33.3%). In all cases, lesions were excised shortly after clinical and dermoscopic evaluation, without prior videodermoscopic follow-up. Four out of six lesions (4/6; 66.7%) were histopathologically classified as nevi (3 compound nevi and one dermal nevus) and the remaining two as seborrheic keratosis and melanosis (each 1/6; 16.7%).
Lesions were marked as suspicious in 5 out of 6 cases according to 7-point checklist and ‘Chaos and clues’ algorithms (both 5/6; 83.3%). Menzies method indicated 4 out of 6 lesions as potentially malignant (4/6; 66.7%), whereas one third of lesions fulfilled dermoscopic criteria of CASH algorithm and 3-point checklist (2/6; 33.3%). All the lesions presented structural and color asymmetry, which in association with clinical background and the presence of melanoma specific structures led to the diagnostic excision. The most commonly observed general melanoma criteria were irregular globules, small hyperpigmented areas (both features 3/6; 50%) and polymorphic vessels (2/6; 33.3%). Among the most frequently noted dermoscopic clues of facial melanoma were asymmetric pigmented follicular openings (4/6; 66.7%), hyperpigmented follicular openings (3/6; 50%) and angulated lines/rhomboids (3/6; 50%). The pseudonetwork was noted in one-third of the analyzed cases (2/6; 33.3%). Two out of six lesions (2/6; 33.3%) were devoid of facial-specific melanoma dermoscopic features. More detailed patients’ records, i.e., demographic and clinical data, dermoscopic findings and histopathological reports are presented in the table 1.

Patient 1

A 66-year-old woman (phototype III) was consulted due to a pigmented macule (6.6 × 7.7 mm) of the helix. The history of the lesion was unknown. Dermoscopy showed asymmetry of colors and structures, pseudonetwork, rhomboids, dark brown blotches, and asymmetric pigmented follicular openings (fig. 1 A). Histopathological examination gave the diagnosis of the compound nevus (fig. 2 A).

Patient 2

A 55-year-old male (phototype II) was consulted due to a pigmented macule (2 × 1.7 mm) of the helix. Dermoscopy showed asymmetry of colors and structures, brown-grey structureless areas, irregular dark brown blotches, polymorphic vessels (of diffuse distribution), and brown globules. In addition, in the lower part of the lesion, asymmetric pigmented follicular openings with obliteration of follicular ostia were observed (fig. 1 B). Based on histopathological evaluation, the compound melanocytic nevus was diagnosed (fig. 2 B).

Patient 3

During a total body skin examination of a 50-year-old male (phototype II) diagnosed with epidermodysplasia verruciformis, a dark brown papule with black spots (5.2 × 3.9 mm) was identified on the crus of the right helix. Dermoscopy showed asymmetry of colors and structures with the presence of dark brown/globules and irregular blotches (fig. 1 C). Histopathological examination gave the diagnosis of a pigmented seborrheic keratosis (fig. 2 C).

Patient 4

A 42-year-old otherwise healthy female (phototype II) presented with the ‘ugly duckling’ lesion (4.8 × 3.7 mm) located in the left scaphoid fossa. Dermoscopy showed a structureless pattern with multiple colors (light brown, dark brown, blue); in addition, within a light brown structureless area polymorphic vessels were observed (fig. 1 D). Histopathological examination gave the diagnosis of the dermal nevus (fig. 2 D).

Patient 5

A 22-year-old male (phototype III) was consulted due to a pigmented papule located at the border between the left lobule and helix (4.5 × 3.1 mm). According to the patient, the lesion has appeared approximately 4 years earlier but enlarged within the previous few months. Dermoscopy showed monoaxial asymmetry of structures and colors. Gradual accentuation of pseudonetwork along with asymmetric pigmented follicular openings and polygonal structures were observed (fig. 1 E). The histopathological image corresponded with a compound melanocytic nevus with features of partial regression (fig. 2 E).

Patient 6

A 68-year-old female (phototype II) with multiple comorbidities referred with an annual dermoscopic skin examination. A dark pigmented macule (4.3 × 3.5 mm) on the posterior aspect of the left auricle was noticed. Dermoscopy showed grey color polygonal structures associated with asymmetrically hyperpigmented follicular openings. Moreover, scattered, irregular dark brown globules were observed (fig. 1 F). Based on the histopathology melanosis was diagnosed (fig. 2 F).

Discussion

This study presents six patients with dermoscopically suspicious but histopathologically benign skin lesions located on the auricle. Four out of six lesions were histopathologically classified as nevi and the remaining two were classified as seborrheic keratosis and melanosis.
Most data on dermoscopic features of melanosis come from evaluation of mucosal lesions [17]. The literature data on dermoscopic presentation of extra-mucosal melanosis are scarce. One case of melanosis of the areola was reported depicting brown parallel lines and a cobblestone pattern [18]. Melanosis identified in our series provides a peculiar dermoscopic presentation of melanosis in extra-mucosal location featuring the presence of violaceous-grey polygonal structures.
While most seborrheic keratoses (SKs) are easily recognized based on clinical and dermoscopic assessment, a subset of SKs may mimic melanocytic lesions, including melanoma, irrespective of the anatomical location. SKs presenting with a multicomponent pattern or regression features are among those difficult to diagnose, as confirmed in our study.
In the retrospective analysis of 161 SKs, the most misleading dermoscopic feature of SK was the multicomponent pattern (19.9% of cases), comprised of clusters of irregular dots/globules, atypical pigment network and/or areas of hyper- or hypopigmentation [19].
On the other hand, a study by Carrera et al. showed that hairpin vessels and fingerprint-like areas were dermoscopic features significantly associated with a risk of incorrect diagnosis of melanoma as SK [20]. Application of the convolutional neural networks and identification of RCM patterns are promising strategies to enhance diagnostic efficacy in discriminating SK from SK-like melanoma [21, 22].
Similarly, compound nevi (CN) are known to act as melanoma mimickers. Irregular pigmentation, shape asymmetry and irregular globules seem to be the most common false-positive melanoma criteria [23]. The first two aforementioned criteria have been observed in 3 cases of CN in the studied material and irregular globules in 1/3 cases.
Few studies on dermoscopic features of external ear melanocytic nevi provide a straightforward approach to evaluate pigmented lesions in this anatomical region.
A recently published study by Deinlein et al. showed that face-specific dermoscopic criteria, such as concentric circles, pseudonetwork, or pigmented follicular openings, may be useful in assessing of auricular pigmented lesions [7]. In this study, asymmetric pigmented follicles and annular-granular structures were most strongly associated with malignancy diagnosis and were found in 40% of external ear melanoma (EEM) cases and only 11% and 7% of auricular nevi.
Rishpon et al. showed that EEM tends to be larger in size and dermoscopically more complex compared to benign lesions presenting with a structureless or reticular pattern [9]. In this study follicular openings were observed with a similar prevalence in benign lesions and EEM (69.6% and 79.3% of cases, respectively). However, in EEM they were mostly focally obliterated with irregular pigmentation between the follicles.
In contrast to the studies showing the possible utility of facial melanoma criteria in the diagnostics of auricular lesions, a recent case series by Kaminska-Winciorek et al. demonstrated that EEM frequently exhibits extra-facial melanoma features [8]. Of note, 3/6 cases of EEM in this case series did not present any facial-melanoma criteria. When it comes to our report, the majority of lesions demonstrated at least one dermoscopic feature characteristic for facial melanoma. However, in more than 30% of cases, no facial-specific dermoscopic features were observed, implying that melanocytic lesions located on the external ear may share both facial and extra-facial dermoscopic traits.
Moreover, presence of a specific color does not seem to help to discriminate benign from malignant ear lesions. The case series of 4 blue lesions on the external ear published by Bombonato et al. revealed that the presence of white color in dermoscopy could be a clue for EEM recognition [24]. However, in the same paper, white scar lines resembling blue-whitish veil have been also described in the dermoscopic presentation of apocrine hidrocystoma.
Additionally, a blue-whitish veil is a hallmark of melanoma invasiveness as reported in the case series by Bombonato et al. [24]. Therefore, the presence of white structures is unreliable for early EEM recognition and the predominance of homogenous blue color in the dermoscopy does not allow to rule out EEM.
The current study shows limitations associated with applying existing dermoscopic algorithms for special anatomical locations, such as external ear. False positive dermoscopy results increase the number of surgical procedures, factors responsible for overdiagnosis and overtreatment [25].

Conclusions

The aim of this report was to present diagnostic difficulties in assessing pigmented auricular lesions. We showed that neither dermoscopic algorithms nor the presence of specific colors or patterns in dermoscopy allow for clear determination of the malignancy of lesions located on the external ear. This report underlines that auricular pigmented lesions have to be approached with integration of lesion’s history and clinical appearance along with dermoscopic presentation. However, melanoma on the ear is still a challenge. Clinical and dermoscopic uncertainty, particularly regarding new-onset lesions appearing in adulthood, should raise clinical concern and justifies histopathological confirmation. Pigmented lesions on the external ear are dermoscopically heterogeneous and may display both facial and extra-facial dermoscopic features. 7-point checklist and ‘Chaos and clues’ algorithms were the least specific in the dermoscopic evaluation of auricular lesions. Scarce literature data on dermoscopic features of the lesions located on the ear highlight the need for further studies in this field.

Funding

No external funding.

Ethical approval

The study was conducted in compliance with the Declaration of Helsinki and was approved by the Ethics Committee of the Medical University of Gdansk, Poland (No. NKBBN/248/2023). Informed consent was obtained from all participants included in the study.

Conflict of interest

The authors declare no conflict of interest.

References