Microphthalmia-Associated Transcription Factor-Positive Pseudonests in Cutaneous Lupus Erythematosus

Abstract

To the Editor: Nicholson and Gerami1 recently reported 2 cases where microphthalmia-associated transcription factor (MITF) aided in distinguishing Mart-1-positive pseudomelanocytic nests from true melanocytic nests in the setting of lichenoid dermatitis. We have been interested in the pseudomelanocytic nest phenomenon since it was first reported by Maize et al2 in 2003. The difficulty of differentiating true Mart-1/Melan-A-positive melanocytic nests from pseudonests is a practical dilemma, as both may occur in the setting of lichenoid tissue reactions, and there is a potential for misdiagnosis if pathologists rely on a single immunostain.1-5 We have found that even combinations of stains like Mart-1 plus MITF have the potential to result in misdiagnosis and wish to alert the pathologists to this potential pitfall. Most of the reports of pseudomelanocytic nests have been described in pigmented actinic keratosis, cutaneous lupus erythematosus (CLE), or lichenoid inflammatory lesions from the head and neck suggesting some relationship to chronically sun-damaged skin.2,4-5 We previously determined that the Mart-1-positive pseudonest phenomenon is uncommon in cases of lichen planus, suggesting that it may be uncommon in sun-protected skin.6 Our current study sought to establish the frequency of pseudonests in the setting of CLE on sun-damaged skin. A search was preformed for cases of CLE from the files of Brooke Army Medical Center and Wilford Hall Medical Center from 1999 to 2009. Each of the cases was initially reviewed by one pathologist to make sure that the biopsies demonstrated marked solar elastosis within the dermis. Cases without significant solar elastosis were excluded from the study. Fresh sections from 53 cases of CLE on sun-damaged skin were stained with hematoxylin and eosin (H&E), Mart-1, and MITF. Three dermatopathologists reviewed the (H&E) slides to determine whether a vacuolar interface or lichenoid interface reaction pattern was present and if collections of cells were present at the dermal-epidermal junction suspicious for pseudonests. Both the Mart-1 and MITF stains were then reviewed in each case. Of the 53 cases of CLE, 30 showed lichenoid interface dermatitis, whereas the other 23 showed a vacuolar interface dermatitis. A total of 6 cases demonstrated areas suspicious with pseudonests on H&E stain (Table 1). Each of these cases demonstrated a lichenoid interface dermatitis. One case (case 30) was remarkable for nested cells that were strongly positive for both Mart-1 and MITF but negative for S100 (Fig. 1A-D). The MITF stain demonstrated at least 2 nonlabeling nuclei (Fig. 1C, inset), which could have alerted the pathologist to the possibility of an aberrant staining pattern. S100 is highly sensitive for melanocytes, and S100-negative melanocytic lesions are rare.7-9 The patient has been followed since 2004 for cutaneous lupus and has developed additional morphologically similar lesions that responded to topical fluocinonide and oral plaquenil but has not demonstrated any evidence of a melanocytic lesion at the biopsy site. In the clinical context of CLE, there is little chance that the solitary aggregate represented an incidental melanocytic nest. Failure to stain with S100 and clinical follow-up support our contention.FIGURE 1: A, Case of CLE with a small nest of cells at the dermal-epidermal junction (H&E: ×20; inset ×60). B, Avid cytoplasmic staining for Mart-1 (×20; inset ×60). C, MITF positive, note that 2 nuclei in the aggregate lack staining (×20; inset ×60). D, Negative staining with S100 (×20; inset ×60).TABLE 1: Cases With Pseudonests Present in H&E or Immunostained SectionsThe pathogenesis of the pseudonest phenomenon remains speculative.2,5 A combination of melanocytes, degenerated keratinocytes, and macrophages has been proposed. Ultimately, studies of the ultrastructural characteristics of pseudonests may help to determine their composition. Our results suggest that MITF may stain pseudonests in a manner similar to Mart-1/Melan-A and cannot be relied upon to distinguish pseudonests from true melanocytic nests. A recent poster at a national dermatopathology conference also described the presence of MITF-positive pseudonests in a case of lichen planus pigmentosus, suggesting our experience is not unique.10 We caution pathologists about overreliance on MITF. We believe that clinical correlation is essential, and when Mart-1/Melan-A-positive and MITF-positive aggregates are noted in the setting of lichenoid dermatitis, S100 staining should be performed. CPT Margaret Abuzeid, MD* LTC Scott R. Dalton, DO† Tammie Ferringer, MD†‡ Richard Bernert, MD§ Dirk M. Elston, MD†‡ Department of *Pathology, Brooke Army Medical Center and Wilford Hall Medical Center, San Antonio, TX Department of †Dermatology; and ‡Pathology, Geisinger Medical Center, Danville, PA §Arizona Dermatopathology, Scottsdale, AZ