Lesions Mimicking Melanoma at Dermoscopy Confirmed Basal Cell Carcinoma: Evaluation with Reflectance Confocal Microscopy

Background: Atypical basal cell carcinoma (BCC), characterized by equivocal dermoscopic features typical of malignant melanoma (MM), can be difficult to diagnose. Reflectance confocal microscopy (RCM) enables in vivo imaging at nearly histological resolution. Objectives: To evaluate with RCM atypical melanocytic lesions identified in dermoscopy, according to common RCM criteria for the differential diagnosis of BCC, and to identify representative RCM parameters for superficial (sBCCs) and nonsuperficial (nsBCCs) basal cell carcinomas (BCCs). Methods: A retrospective analysis of consecutive patients evaluated with RCM, selecting excised lesions classified at dermoscopy with ≥1 score from the re visited 7-point checklist, mimicking melanoma, registered between 2010 and 2016. Cluster analysis identified BCC subclassifications. Results: Of 178 atypical lesions, 34 lesions were diagnosed as BCCs with RCM. Lesions were confirmed BCCs with histopathology. Dermoscopic features included atypical network (55.9%) and regression structures (35.5%) associated with sBCCs, and an atypical vascular pattern (58.8%) and irregular blotches (58.8%) with nsBCCs. Hierarchical cluster analysis identified 2 clusters: cluster 1 (100% sBCCs) was characterized by the presence of cords connected to the epidermis (90%, p < 0.001), tumor islands located in the epidermis (100%, p < 0.001), smaller vascular diameter (100%, p < 0.001) and solar elastosis (90%, p = 0.017), and cluster 2 (nsBCCs 85%) was defined by the dermic location of tumor islands (87.5%, p < 0.001) with branch-like structures (70.8%, p = 0.007) and surrounding collagen (83.3%, p = 0.012), peripheral palisading (83.3%, p = 0.012) and coiled vascular morphology (79.2%, p < 0.001) with a larger vascular diameter (50%, p < 0.001). Conclusions: RCM is able to diagnose BCCs mimicking melanoma at dermoscopy and seems able to identify sBCCs and nsBCCs.


Introduction
Basal cell carcinoma (BCC) is the most frequent type of skin cancer, with a high incidence in the fair skin population and more commonly encountered in photo-exposed body areas. [1]. The first steps in the diagnostic algorithm of cutaneous lesions are clinical and dermoscopic examinations [2]. Dermoscopic features of typical basal cell carcinomas (BCCs) are widely known [3,4]. In some cases, BCCs share clinical and dermoscopic features with neoplastic and inflammatory conditions [5], and differential diagnosis can be particularly challenging. In these cases, BCCs can present a blue-whitish veil, multiple brown dots, pseudopods, atypical pigment network or radial streaks, dermoscopic features usually associated with malignant melanoma (MM) [4,6,7]. Melanocytic and nonmelanocytic lesions with difficult dermoscopic characteristics have been widely documented in the lit-erature [8][9][10][11], whilst BCCs with atypical dermoscopic features have been poorly described.
Reflectance confocal microscopy (RCM) is a noninvasive technology, which enables in vivo skin imaging at almost histological resolution and has been proven effective in increasing the accuracy of melanocytic and nonmelanocytic skin cancer diagnoses and reducing the number of benign excised lesions [12,13]. RCM may be especially helpful in difficult lesions, where clinical and dermoscopic features are not straightforward [14][15][16][17].
RCM parameters of typical BCCs, and the superficial (sBCC) and non-superficial (nsBCC) subtypes, are well known [18][19][20][21][22]. The primary aim of our study was to evaluate atypical melanocytic lesions identified in dermoscopy, with RCM according to common RCM criteria for the differential diagnosis of BCC. A secondary aim of the study included the identification of representative RCM parameters associated with the sBCC and nsBCC subclassifications.

Materials and Methods
For further details, see the supplementary material (for all online suppl. material, see www.karger.com/doi/10.1159/000493727) ( Fig. 1)

Results
Of the 178 consecutive lesions registered at RCM in the study period with at least one of the features of atypical melanocytic lesions evidenced with dermoscopy and without specific features for other lesions, a total of 34 cases were classified at RCM retrospective analysis as BCCs. The remaining 144 cases were classified as other lesions (dermatofibroma, seborrheic keratosis, squamous cell carcinoma or nonspecific). All 34 lesions were confirmed BCC at histopathological diagnosis, confirming a perfect agreement. The diagnostic agreement of RCM for BCC diagnosis for all 1,484 lesions evaluated was 99.46% (8 false negatives).
The 34 patients included in this study were mostly female (68%) with a mean age of 64 years (±17.2). More than half of the lesions were located in photoexposed areas (52.9%) and in patients with phototypes II (44.1%) and III (55.9%). The retrospective analyses highlighted that the most common dermoscopic features observed included the atypical network (55.9%), atypical vascular pattern (58.8%) and irregular blotches (58.8%) ( Table 1, column 1).

Discussion
Dermoscopic features of typical BCCs have been widely discussed in the literature [3,34], but reporting of atypical BCCs has so far been restricted to short reports or case series evaluated with dermoscopy. These studies have highlighted some common dermoscopy features associated with MM (from the 7-point checklist) [23], including irregular streaks, identified by Bakos et al. [7], and atypical network, blue-white veil and irregular dots/ globules, identified by Ferrari et al. [6]. However, this study confirms that all features of the 7-point checklist were identified in this equivocal set of atypical BCCs. The most frequently identified features in this cohort included atypical network (55.9%), atypical vascular pattern (58.8%) and irregular blotches (58.8%), whereas the features identified in the previously mentioned case reports [6,7] were identified in less than half of the lesions in the current series.
Given the presence of atypical features at dermoscopy in all lesions, the usual clinical pathway for lesions with suspicious MM characteristics includes lesion excision for a definitive histopathological diagnosis [37], postponing accurate diagnoses and consequential therapeutic and patient management [4,16,35]. RCM is a noninvasive, in vivo imaging tool, which assists in timely differential diagnoses and optimizes therapeutic management [12], with reported pooled sensitivity and specificity of BCC detection of 91.7% (95% CI: 0.87-0.95) and 91.3% (95% CI: 0.94-0.96), respectively [37]. Further, RCM features of typical BCC have also been applied to, and have successfully defined, the BCC subtypes [20][21][22]29]. How-ever, to date, RCM features of atypical BCCs mimicking MM in dermoscopy are missing.
In the current study, commonly accepted RCM features associated with typical BCCs were applied to this cohort of atypical BCCs. The main RCM criteria for typical BCC diagnosis include the presence of tumor islands (nests of basaloid cells), peripheral palisading and peritumoral clefts [20,21], and in the current study, these RCM features proved useful in identifying all the BCCs in this cohort of lesions with dermoscopic suspicion of melanoma. This finding has a significant impact for therapeutic patient management, as BCC patients can be offered differential diagnosis from MM, at the patients' bedside, with a noninvasive procedure, saving the patient the concern of an uncertain MM diagnosis and eventually an unnecessary surgical excision, whilst enabling the physician to immediately commence correct treatment.
There are many factors influencing the prognosis of BCCs, including tumor size, site, clinical margins, histopathological features of aggression, failure of previous treatment, immunosuppression, and histopathological subtypes [38,39]. Specific dermoscopy features have been proven potent predictors of typical sBCC and nsBCC [3]. In terms of BCC subtypes, sBCCs are generally associated with a lower risk of recurrence and can be treated with destructive surgical techniques, including curettage and cautery, cryosurgery and carbon dioxide laser and nonsurgical techniques, including topical immunotherapy with imiquimod application and photodynamic therapies [20,38,39]. Conversely, nsBCCs are generally considered at higher risk of recurrence and are treated with more invasive surgical therapies, such as excision with predetermined margins, Mohs micrographic surgery or, in selected cases, invasive nonsurgical radiotherapy [39]. Together with other prognostic factors, the BCC subtype is therefore crucial to optimal patient management.
The current study's secondary aim was to identify whether RCM was able to assist in the identification of the BCC subtypes, sBCC and nsBCC, and the current results suggest that in over 90% of cases, the distribution of RCM features was indicative of the relevant subtypes, as illustrated by the 2 clusters identified at cluster analysis.
The parameters identified in the current study as indicative of sBCC included cords connected to the epidermis, a small tumor island located in the epidermis, absent vascular morphology and solar elastosis. Other studies have identified these RCM features associated with typical sBCCs, underlining the variations in cell size and shape of keratinocytes and architectural disorder of the overlying epidermis as visualized at RCM, usually ascribed to associated actinic injury [29].
The RCM features associated with nsBCCs identified in the current study included the dermic location of large tumor islands with branch-like structures, peripheral palisading, peritumoral clefts, coiled vascular morphology and increased vascular diameter, collagen surrounding the tumor island and the absence of cords connected to the epidermis [20], many of which were underlined in a recent review [21]. The confounding feature in cluster analysis for differential diagnoses of sBCC or nsBCC was the presence of coiled vascular morphology in 3 of the 13 sBCCs. As has been proven in a study of BCC subtypes with optical coherence tomography, 19% of sBCCs were associated with slight or highly present microvascular coil features [40]. Vascular morphology and diameters are complex to evaluate with RCM, and the integration of optical coherence tomography, another noninvasive tool, could assist in more precise in vivo diagnoses [40].
This study is principally limited by the absence in literature of a dermoscopy classification of atypical BCCs, but typical BCC RCM criteria proved useful in identifying this equivocal subset of atypical BCCs. Further, the limited ability of RCM in vascular morphology and vessel diameter evaluation made this feature difficult to assess, and the integration of optical coherence tomography analysis should be considered for future research dedicated to in vivo diagnosis of atypical BCC.
In conclusion, RCM has been proven to be an optimal, noninvasive examination for early differential diagnosis, and this study underlines its application to BCCs with atypical dermoscopy presentations mimicking MM. Further, RCM seems able to assist in differential diagnoses of sBCCs or nsBCCs in this group of atypical BCCs, offering