A small number of metastatic melanoma cases (2%–5%) have an unknown primary tumor. In this situation, we should suspect of a previously melanocytic lesion misdiagnosed as benign or of a possible complete regressing melanoma. The problem is bigger when that metastasis is also negative for usual immunohistochemical markers such as S100, HMB45, Melan-A, and tyrosinase but, fortunately, that is a very rare situation.1
Otherwise, melanoma-associated leukoderma or depigmentation is a recognized entity; it is believed to confer a better prognosis to patients in advanced stages of melanoma.2
We present a woman who developed leukoderma vitiligo-like and lymph node metastasis, where BRAF-V600E mutated was demonstrated, in the territory of a lesion diagnosed as melanosis tumoral.
A 82-year-old woman, with a personal history of high blood pressure and chronic renal disease associated, was send to the dermatology department to evaluate a lesion in her left foot that she believes she has had for years.
The patient had been admitted to investigate a constitutional syndrome and a painful and rapidly growing mass in the left groin of a month of evolution (Fig. 1A).
On physical examination, there was an 11-cm erythematous and hard mass in the left groin that was firmly attached to deep tissues. A 1-cm black–bluish irregular tumor and a 3 mm black papule in its proximity were identified in her left foot (Figs. 1C, D). Under nonpolarized dermoscopy, numerous black round and confluent areas were seen with slight desquamation. No ulceration, reticular network, or vascular structures were identified. A total body computed tomography (CT) scan reveled a 12 cm-size adenopathic conglomerate in the inguinal area (Fig. 1B). Under 18F-fluoro-2-deoxy-d-glucose positron emission tomography, this mass showed an increased metabolism of 33.74 Standard uptake value. No other lymphadenopathies or involvement of internal organs were identified, and the brain CT scan was normal. A biopsy of the inguinal mass and excision of cutaneous lesions were performed.
Biopsy of the skin showed an epidermis without remarkable changes. In the dermis, an intense infiltrate of melanophages (CD68+) with some hemosiderophages were found in a background of fibrosis, mild lymphocytic infiltrate, and neovascularization (Figs. 2A–D). No melanoma cells were identified in serial sections by immunohistochemical stains (S100, HMB-45, Melan-A, and tyrosinase; Figs. 2E–H). The proliferative index in the dermis was almost nil (Fig. 2I).
Needle core of the groin mass showed a proliferation of epithelioid-habit cells with eosinophilic cytoplasm, heterogeneous vesicular nuclei, and occasional basophilic nucleoli. Up to 5 mitosis/10 high-power field were counted (Figs. 3A–C). Tumoral cells were vimentin and CD99 positive and negative for S100, CD34, HMB-45, Melan-A, SOX 10, tyrosinase, C-KIT, CD45, and CKAE1/AE3, among others (Figs. 3D–I). Considering the possibility of sampling error and the heterogeneity of tumors, the sample was repeated; however, the same immunophenotype was demonstrated in other viable areas.
Two weeks later, the patient presented bleaching of both eyebrows and achromic spots in the palms and face (Figs. 1E, F).
With all this information, tumoral melanosis was diagnosed, and the possibility of a complete regressing melanoma was suggested, whereas the groin mass was interpreted as a metastatic melanoma with aberrant phenotype. BRAF-V600E mutation was subsequently detected in the lymph node sample. Treatment with vemurafenib or dabrafenib was considered, but because our patient had a poor general baseline situation and a significant deterioration in the kidney function, this possibility was ruled out because both are nephrotoxic drugs. No treatment could be initiated because her kidney function worsened, and she died 1 month later.
Regression of cutaneous melanocytic lesions is a well-recognized phenomenon. Complete regression of melanomas is a very rare event. In the study by Ching et al,3 a frequency of 0.07% was found among 5134 melanomas reviewed.
Clinically, regression of melanocytic lesions can occur spontaneously without any identifiable antecedent event, after trauma, or immunomodulatory therapy.4
This regression may be partial or complete. If the regression is partial, cytology and histology of the remaining lesion should be carefully evaluated.
Small nests made up of bland cytology melanocytes, showing maturation in deep, arranged within, or on the periphery of the regression zone, are frequently indicators of benignity.4
In the regression zones, there is a variable lymphocytic infiltrate in the dermis in actively regressing melanocytic lesions. Angioplasia is usually present.4
Martín et al,5 in a study of 13 patients followed by digital dermoscopy for nevi, in which there was histological evidence of complete regression, described a particular variety of fibrosis, of subepidermal localization, consisting of very fine bundles of collagen.
Although variable degrees of melanophagy were evident, just in 2 of 13 regressed nevi, it was intense.5 The mean size of these nevi was 0.5 cm, and their mean time for complete regression was 6.4 months (between 3 and 11 months).5
In the case of the halo nevi (HN), nests of melanocytes are destroyed by an inflammatory infiltrate, predominantly CD8, which is initially very prominent and gradually decreases.6
Sometimes, HN melanocytes may demonstrate significant reactive cytological atypia that could lead to suspicion that it is actually regressed melanoma (RM).6 The absence of fibrosis in the final phase has been reported in HN.5
Also, a previous incomplete shave nevus (recurrent nevus) may mimic RM, but in this case, fibrosis will be orderly and parallel to the epidermal surface.4
RM will show an irregular silhouette, areas of pagetoid spread, cytological atypia, and lack of maturation at the edge or within an area of regression.4 Epidermal atrophy was rare in nevi, whereas it was a usual finding in RM.5
The inflammatory infiltrate in RM is usually asymmetric and patchy, predominantly CD4,7,8 whereas fibrosis is compact.5 The latter is due to the fact that fibrogenic cytokines, interleukin-6, platelet-derived growth factor, and transforming growth factor-beta were more frequently expressed in melanoma than in the halo nevus, irrespective of regression.8
Kamino et al studied 33 primary invasive malignant melanomas with evidence of late stage regression and 10 malignant melanomas with a recent scar secondary to a previous biopsy, and they found that areas of regression had a well‐defined compressed layer of thin elastic fibers pushed down from the papillary dermis to the base of the fibrosis. This was the opposite to an abrupt transition to the thick elastic fibers of the spared reticular dermis in the base of scars.9 In our case, the staining of elastic fibers showed contradictory results. In some areas, elastic fibers seemed to be pushing down from the fibrotic regression areas, but in others, they showed an abrupt transition to the normal dermis.
It has been stated that the regression of a melanoma area less than 50%–75% does not affect its prognosis, unlike what happens when the regression area is more extensive.10
When we are faced with the diagnosis of fully regressed melanocytic lesions the challenge is even greater, being very difficult to determine its lineage or nature, because regression is not specific of melanocytic lesions and may be found in solar lentigo and pigmented basal cell carcinomas.5,11,12
Completely regressed melanocytic lesions clinically simulate blue nevi and deep penetrating nevi.12 A similar morphology has also been reported in patients with remote melanomas treated with dabrafenib–trametinib and pembrolizumab.13,14
Histologically, these lesions are often diagnosed as tumoral melanosis, in which, melanophages (CD68+) are numerous forming nodular masses within the dermis,11 whereas the melanocytes are absent (Melan-A, S100, HMB45, and SOX 10 are negative). Careful evaluation must be performed to detect any remaining tumor cells masked by the dense inflammatory infiltrate or confused as melanophages.4
Complete RM is difficult to document. In most cases, it is the development of metastases of unknown primary tumor in which the detailed physical examination of the patient allowed to subsequently discover these types of lesions.5,15 In 2%–5% of metastatic melanoma cases, the primary tumor is unknown, and the diagnosis of a regressing primary melanoma is made following the criteria of Smith and Stehlin modified by Ronan.15–18
Other possibilities that we should explore when facing a metastatic melanoma from unknown primary tumor, would be the malignization of subcapsular nevi of lymph nodes, melanomas of the gastrointestinal tract, genital tract, the eye, or another part of the body. However, many authors state that most of the time it is a prior excision of a cutaneous lesion clinically or histologically misdiagnosed as benign, which was actually a primary melanoma, usually thin or “nevoid” type.19
The diagnosis of metastatic melanoma in lymph nodes is based on cellular morphology, immunohistochemistry, molecular pathology, or even ultrastructural microscopy.20 The morphology of a melanoma is variable and can recall epithelial, hematologic, mesenchymal, and neural tumors. For this reason, it is important to confirm the suspicion by immunohistochemistry.
Zubovits et al21 studied 126 cases of metastatic melanoma, of which 98% were positive for S-100, 93% NK1/C3, 82% Melan-A, and 76% for HMB-45. Note that S-100 and NK1/C3 also stained histiocytes present in lymph nodes, whereas Melan-A and HMB-45 stained only melanocytes, for this reason, if it is true that S-100 was the most sensitive marker for melanocytic lesions, HMB-45 and, especially, Melan-A are more specific markers.21,22In the series of Zubovits, 1 of 126 cases exhibited absence of staining for any of these markers.21 Negative cases for melanocytic markers are usually diagnosed by the presence of a known primary in the territory drained by the affected node.1
A study indicated a possible link between tumor metastasis and progression or atypical morphology with loss of immunoreactivity for S100.23
Seventeen cases S100-negative were found between 1553 metastatic melanoma studied by Aisner.24 Of these, 10 (59%) were immunoreactive for both HMB-45 and Melan-A4, 5 had a documented primary cutaneous melanoma, and the other 2 had a prior metastatic lesion S100+. In 11 cases of S-100 negative metastasis, where were possible evaluate the primary tumor, 9 were S100-immunoreactive (82%).24
Nonaka et al evaluated 35 cases of malignant melanomas and compared their immunohistochemical profile in necrotic and viable tumoral areas. They found that necrotic areas had nonspecific staining for S-100, whereas HMB-45, Melan-A, and tyrosinase focally stained necrotic areas where the outline of the tumor cells could still be recognized as ghost cells but not in areas of complete coagulative necrosis.25
In addition, the sensitivity of SOX-10 has been evaluated in metastatic melanomas in lymph nodes that do not express melanocytic markers by immunohistochemistry, being especially useful because its nuclear staining and absence of expression in dendritic cells.26,27
Moreover, it should be noted that complete histological regression of melanoma metastases has also been reported, although it is considered very rare and is very likely to be under reported.3,28
In this case, we found the BRAF-V600 mutated. BRAF is a serine/threonine protein kinase involved in the signaling cascade (MAPK pathway) that drives cell proliferation, differentiation, migration, survival, and angiogenesis. It is well known that BRAF is mutated in 50%–60% of melanomas, but it is also mutated in other tumors such as gliomas, colorectal carcinoma, thyroid carcinoma, nonsmall cell lung cancers, and hairy cell leukemia. However, none of them was suspected in our patient after imaging and positron emission tomography-CT.29
Achromic patches reminiscent of vitiligo, somewhat symmetrical but not typical, were discovered in our patient after more intensive examination. Several report of cases and even studies have associated hypomelanosis/despigmentation vitiligo-like, HN, and melanoma.13,30 This is understandable given that the regression of melanomas would be mediated by the immune rejection of malignant and also benign melanocytes.4
The relative rate of this association ranges from 1.4% to 20%, being vitiligo 7–10 times more prevalent in patient with melanoma; in majority, their onset is after the diagnosis of melanoma and often in metastatic cases.30–32 Careful evaluation with Wood's lamp could detect many cases scarcely evident to the naked eye.33
Furthermore, in an analysis of 135 studies on melanoma immunotherapy, the overall cumulative incidence of vitiligo was 3.4% and found that these patients had 2 to 4 times lower risk of disease progression and death, respectively, compared with patients without vitiligo development.34,35
In conclusion, our patient presented clinically a typical skin melanoma lesion with a homolateral inguinal mass, pointing to a melanoma with inguinal metastases. The skin sample showed complete histological regression, and the lymph node biopsy showed a tumoral proliferation morphologically compatible with metastasis of melanoma but negative for all the immunohistochemical melanocytic markers. However, a mutated BRAF-V600E was discovered. These findings together with the concomitant development of vitiligo-like lesions favor the diagnosis of metastatic melanoma from a complete regressive primary cutaneous melanoma, with an aberrant immunohistochemical profile.
The authors thank Dr Carlos Santonja, Pathology Department of the Fundación Jimenez Díaz, for collaborating in carrying out the Sox-10 staining of the samples.
1. Gao Z, Stanek A, Chen S. A metastatic melanoma
with an unusual immunophenotypic profile. Am J Dermatopathol. 2007;29:169–171.
2. Teulings HE, Lommerts JE, Wolkerstorfer A, et al. Vitiligo-like depigmentations as the first sign of melanoma: a retrospective case series from a tertiary vitiligo centre. Br J Dermatol. 2017;176:503–506.
3. Ching D, Amini E, Harvey NT, et al. Cutaneous tumoural melanosis: a presentation of complete regression of cutaneous melanoma. Pathology. 2019;51:399–404.
4. McCardle T, Messina J, Sondak V. Completely regressed cutaneous melanocytic lesion revisited. Semin Oncol. 2009;36:498–503.
5. Martín JM, Rubio M, Bella R, et al. Regresión completa de nevos melanocíticos: correlación clínica, dermatoscópica e histológica de una serie de 13 casos. Actas Dermosifiliogr. 2012;103:401–410.
6. Mooney MA, Barr RJ, Buxton MG. Halo nevus or halo phenomenon? A study of 142 cases. J Cutan Pathol. 1995;22:342–348.
7. Tefani FL, Barneston RS, Halliday GM, et al. Immunohistochemical analysis of the cellular infiltrate in primary regressing and non-regressing malignant melanoma. J Invest Dermatol. 1991;97:197–202.
8. Moretti S, Spallanzani A, Pinzi C, et al. Fibrosis in regressing melanoma versus nonfibrosis in halo nevus upon melanocyte disappearance: could it be related to a different cytokine microenvironment? J Cutan Pathol. 2007;34:301–308.
9. Kamino H, Tam S, Roses D, et al. Elastic fiber pattern in regressing melanoma: a histochemical and immunohistochemical study. J Cutan Pathol. 2010;37:723–729.
10. Kaur C, Thomas R, Desai N, et al. The correlation of regression in primary melanoma with sentinel lymph node status. J Clin Pathol. 2008;61:297–300.
11. Flax SH, Skelton HG, Smith KJ, et al. Nodular melanosis due to epithelial neoplasms. A finding not restricted to regressed melanomas. Am J Dermatopathol. 1998;20:118–122.
12. Grohs RL, Mesbah Ardakani N. Melanoma manifesting as tumoral melanosis; now you see it, now you don't. Am J Dermatopathol. 2018;40:462–465.
13. Laino A, Shepherd B, Atkinson V, et al. Tumoral melanosis associated with combined BRAF/MEK inhibition (dabrafenib/trametinib) in metastatic melanoma
. JAAD Case Rep. 2018;4:921–923.
14. Woodbeck R, Metelitsa AI, Naert KA. Granulomatous tumoral melanosis associated with pembrolizumab therapy: a mimicker of disease progression in metastatic melanoma
. Am J Dermatopathol. 2018;40:523–526.
15. High WA, Stewart D, Wilbers CR, et al. Completely regressed primary cutaneous malignant melanoma with nodal and/or visceral metastases: a report of 5 cases and assessment of the literature and diagnostic criteria. J Am Acad Dermatol. 2005;53:89–100.
16. Smith JL, Stehlin JS. Spontaneous regression of primary malignant melanoma with regional metastases. Cancer. 1964;18:1399–1415.
17. Ronan SG, Han MC, Das Gupta TK. Histologic prognostic indicators in cutaneous malignant melanoma. Semin Oncol. 1988;15:558–565.
18. Vijuk G, Coates AS. Survival of patients with visceral metastatic melanoma
from an occult primary lesion: a retrospective matched cohort study. Ann Oncol. 1998;9:419–422.
19. Giuliano AE, Moseley HS, Morton DL. Clinical aspects of unknown primary melanoma. Ann Surg. 1980;191:98–104.
20. Karimipour DJ, Lowe L, Su L, et al. Standard immunostains for melanoma in sentinel lymph node specimens: which ones are most useful. J Am Acad Dermatol. 2004;50:759–764.
21. Zubovits J, Buzney E, Yu L, et al. HMB-45, S-100, NK1/C3, and MART-1 in metastatic melanoma
. Hum Pathol. 2004;35:217–223.
22. Ohsie SJ, Sarantopoulos GP, Cochran AJ. Immunohistochemical characteristics of melanoma. J Cutan Pathol. 2008;35:433–444.
23. Bishop PW, Menasce LP, Yates AJ, et al. An immunophenotypic survey of malignant melanomas. Histopathology. 1993;23:159–66.
24. Aisner DL, Maker A, Rosenberg SA, et al. Loss of S100 antigenicity in metastatic melanoma
. Hum Pathol. 2005;36:1016–1019.
25. Nonaka D, Laser J, Tucker R, et al. Immunohistochemical evaluation of necrotic malignant melanomas. Am J Clin Pathol. 2007;127:787–791.
26. Willis BC, Johnson G, Wang J, et al. SOX10: a useful marker for identifying metastatic melanoma
in sentinel lymph nodes. Appl Immunohistochem Mol Morphol. 2015;23:109–112.
27. Jennings C, Kim J. Identification of nodal metastases in melanoma using sox-10. Am J Dermatopathol. 2011;33:474–482.
28. King M, Spooner D, Rowlands DC. Spontaneous regression of metastatic malignant melanoma of the parotid gland and neck lymph nodes: a case report and a review of literature. Clin Oncol. 2001;13:466–469.
29. Dankner M, Rose AAN, Rajkumar S, et al. Classifying BRAF alterations in cancer: new rational therapeutic strategies for actionable mutations. Oncogene. 2018;37:3183–3199.
30. Arpaia N, Cassano N, Vena GA. Regressing cutaneous malignant melanoma and vitiligo-like depigmentation. Int J Dermatol. 2006;45:952–956.
31. Milton GW, McCarthy WH, Carlon A. Malignant melanoma and vitiligo. Australas J Dermatol. 1971;12:131–142.
32. Schallreuter KU, Levenig C, Berger J. Vitiligo and cutaneous melanoma. A case study. Dermatologica. 1991;183:239–245.
33. Barriere H, Litoux P, Le Lay M, et al. Cutaneous achromia and malignant melanoma. Ann Dermatol Venereol. 1984;111:991–996.
34. Teulings HE, Limpens J, Jansen SN, et al. Vitiligo-like depigmentation in patients with stage III-IV melanoma receiving immunotherapy and its association with survival: a systematic review and meta-analysis. J Clin Oncol. 2015;33:773–781.
35. Cui J, Bystryn JC. Melanoma and vitiligo are associated with antibody responses to similar antigens on pigment cells. Arch Dermatol. 1995;13:314–318.