Leukomelanoderma following acute cutaneous graft‐ versus‐host disease

ARTICLE

Auteur(s) : Hee Young KANG, Won Hyoung KANG

Department of Dermatology, Ajou University School of Medicine Suwon, Korea, 442-721

Article accepted 23/02/2004

Acute graft versus host disease (GVHD) is characterized by a selective epithelial inflammation that can affect the skin, digestive tract, and liver [1]. Skin involvement is usually characterized by a measles-like or scarlatiniform rash. Development of pigmentary abnormalities can be observed in sites where acute cutaneous GVHD has occurred, and usually consists of hyperpigmented spots. We observed an atypical pigmentary change consisting of a combination of hyper- and hypopigmentation in a young allogeneic bone marrow recipient who suffered repetitive acute GVHD. The histopathological examinations showed features of a post-inflammatory process. Because keratinocytes produce inflammatory cytokines including tumor necrosis factor (TNF)-α and interleukin (IL)-1α, which may be implicated in the inflammatory phenomena seen in acute GVHD, we studied whether these inflammatory cytokines may be implicated in these pigmentary changes [2].

Patient and methods

Patient and skin samples

A 24-year-old man with acute lymphocytic leukemia underwent bone marrow transplantation from his HLA-identical brother. He was skin type IV. On day 25, the patient developed an acute GVHD manifested by a typical rash, abdominal pain, diarrhea, hyperbilirubinemia, and fever. The scarlatiniform rash with confluent sheets of erythema occurred initially on the upper portions of the trunk that gradually spread to include the entire trunk, face, and extremities. There was blistering on parts of the face and upper trunk. Despite treatment with a systemic steroid and cyclosporin A, one relapse of cutaneous lesions of the face and the trunk occurred during the next 3 months. Subsequently, his condition improved slowly and the skin lesions faded leaving residual pigmentary changes. Currently (2 years post-BMT) he is well, but has persistent atypical skin pigmentation. The pigmentation consists of hyperpigmentation with islands of hypopigmentation, so called leukomelanoderma, on his face (Fig. 1A) and upper trunk (Fig. 1B). The other parts of skin, apart from the face and upper trunk, showed diffuse melanoderma. He was sure that the leukomelanoderma on his face and upper trunk might be related to the repetitive attack and blistering during acute cutaneous GVHD. In order to study the pathogenesis of the different pigmentary responses (hyper- or hypopigmentation) in our patient, histological studies were carried out. We also studied whether the inflammatory cytokines produced by keratinocytes during acute GVHD may be implicated in these pigmentary changes. The patient had three skin biopsies of the face, i.e. hyperpigmented, normal appearing, and hypopigmented skin. Three µm paraffin-embedded sections were subjected to routine light microscopic examination. Hematoxylin and eosin and Fontana-Masson (for melanin) staining were performed.

Immunohistochemistry

Four µm paraffin-embedded sections were processed for immunohistochemistry as described earlier [3]. A monoclonal antibody to NKI/beteb (Monosan, Uden, the Netherlands) for melanocytes was diluted 1: 20 and incubated for 2hr at RT [3]. Monoclonal antibodies to IL-1α (R&D systems Inc., MN, U.S.A.), IL-2 (Santa Cruze Biotechnology, CA, U.S.A.), and IL-10 (Santa Cruze) were diluted 1: 50 and incubated overnight at 4°C. A monoclonal antibody to TNF-α (Hycult Biotechnology, Uden, the Netherlands) was diluted 1: 200 and incubated for 18 min at 45°C. Negative controls were performed by using phosphate-buffered saline instead of a primary antibody. These controls always yielded the expected negative results.

Image analysis

The image analysis was evaluated using Image Pro Plus Version 4.5 (Media Cybertics Co., MD, U.S.A.). The stained area per epidermal area (SA/EA) was measured in hyperpigmented, normal appearing and hypopigmented skin.

Results

General histopathological features

The histopathological examinations showed features of a post-inflammatory process. All specimens revealed some melanophages in the dermis. Fontana-Masson staining showed an increased epidermal melanin pigmentation and more pigmentary incontinence in the dermis of hyperpigmented skin than in normal appearing and hypopigmented skin (Fig. 2A). NKI-beteb detection of epidermal melanocytes showed an increased number and stronger intensity of staining in hyperpigmented lesions compared with normal appearing and hypopigmented skin (Fig. 2B). The number of melanocytes in hypopigmented skin was not clearly different from normal appearing skin.

Immunohistochemistry for cytokines

Staining with antibody against TNF-α revealed a strong immunoreactivity throughout the hyperpigmented skin, whereas there was only a weakly positive staining in the hypopigmented skin (Fig. 2C). Image analysis showed an increase in SA/EA of 0.42 in hyperpigmented skin as compared with hypopigmented skin (0.26). The SA/EA of normal appearing skin was 0.33. In contrast, staining with IL-1α, IL-2 and IL-10 antibodies showed no differences (data not shown).

Discussion

Postinflammatory hyper- and hypopigmentation are frequently encountered problems and represent the sequelae of various inflammatory cutaneous disorders as well as therapeutic interventions. However, the underlying mechanisms and the variability in individuals for developing hypopigmentation or hyperpigmentation are not well understood [4]. Recent evidence has suggested that melanocyte function is regulated by several cytokines that are secreted by surrounding epidermal cells in a paracrine fashion [5, 6]. In particular the chemical mediators released in the inflammatory process have considerable influence on melanogenesis [5]. We studied whether inflammatory cytokines produced keratinocytes during acute cutaneous GVHD which may be implicated in the pigmentary changes of our patient. The cytokines tested were IL-1α, IL-2, TNF-α, (which stimulate the development of acute GVHD) and IL-10 (which exerts opposing effects at various points in the network). While TNF-α was variably distributed in proportion to different degrees of pigmentation, other molecules were detected at minimal levels in all samples. The inflammatory TNF-α is usually present in the dermal and epidermal layers of normal skin with its local concentrations modified by several stimuli. In this case, the staining of TNF-α was higher in the suprabasal layer of the hyperpigmented skin compared to the hypopigmented suprabasal skin. This observation may indicate that the production of TNF-α by the epidermal microenvironment may be involved in postinflammatory pigmentary changes, i.e. the expression level of TNF-α may decide the degree of pigmentation.
Regarding the effect of TNF-α on melanogenesis, conflicting data have been published. TNF-α has been upregulated in the epidermis of lentigo senilis or seborrhic keratosis relative to the perilesional normal epidermis [7-8]. The authors suggested that an accentuation of the epidermal endothelin cascade by TNF-α plays an important role in the mechanism involved in the hyperpigmentation of lentigo senilis or seborrhic keratosis. These findings and our own results open the possibility that high amounts of TNF-α may cause melanocyte activation resulting in hyperpigmentation. On the contrary, it has been published that TNF-α is a paracrine inhibitor of human melanocyte proliferation and melanogenesis in vitro [9]. Furthermore TNF-α was expressed more highly in vitiligo skin compared with perilesional skin, which suggests that TNF-α caused apoptosis of melanocytes which resulted in melanocyte loss [6]. Interestingly, we also found that there were no differences in the expression of TNF-α between the hyperpigmented skin of our patient and the vitiligo skin of another patient, both showing high levels of TNF-α (data not shown).
TNF exerts its biological activity by binding to type 1 and type 2 receptors. Type 1 receptor signals both cell survival and cell death signals, whereas type 2 receptor primarily mediates cell survival signals [10]. Both cell survival and death signals mediated by TNF-α require distinct sets of adapter and other downstream signaling molecules, which can make a decision between survival and death during TNF-α induced signaling. This might offer an explanation as to why the staining of TNF-α is increased in both vitiligo and hyperpigmented skin. Therefore it is conceivable that TNF has bifunctional effects on melanocytes; it may stimulate melanogenesis, and it may apoptose melanocytes.
Several alternative hypotheses may be proposed. Melanocyte-lymphocyte interaction during the early phase of GVHD, in a way similar to that occurring in keratinocytes, may result in this melanocyte dysfunction evolving into leukomelanoderma. Although there was no evidence of melanocyte-lymphocyte interaction by light microscopy in our case, Claudy et al. showed this interaction in the hyper- and hypopigmentation of chronic GVHD by light and electron microscopy [11]. Destruction of melanocytes by alloreactive lymphocytes would be a possible mechanism for vitiligo-like pigmentary changes which have been reported as a manifestation of chronic GVHD [12, 13]. Also, the effect of TNF-α on the pathophysiology of GVHD may have been complicated by the conditioning regimen (irradiation and/or chemotherapy) which may damage and activate the host tissue and induce secretion of this inflammatory cytokine. Further studies should be undertaken with a more detailed focus on the mechanism of the pigmentary changes induced by TNF-α. n

References

1. Aractingi S, Chosidow O. Cutaneous graft-versus-host disease. Arch Dermatol 1998; 134: 602-12.

2. Deeg HJ. Cytokines in graft-versus-host disease and the graft-versus-leukemia reaction. Int J Hematol 2001; 74: 26-32.

3. Im S, Kim J, On WY, Kang WH. Increased expression of α-melanocyte-stimulating hormone in the lesional skin of melasma. Br J Dermatol 2002; 146: 155-74.

4. Ruiz-Maldonado R, Orozco-Covarrubias ML. Postinflammatory hypopigmentation and hyperpigmentation. Seminars Cutaneous Med and Surg 1997; 16: 36-43.

5. Takiwaki H, Shirai S, Kohno H, Soh H, Arase S. The degrees of UVB-induced erythema and pigmentation correlate linearly and are reduced in a parallel manner by topical anti-inflammatory agents. J Invest Dermatol 1994; 103: 642-6.

6. Moretti S, Spallanzani A, Amato L, Hautmann G, Gallerani I, Fabiani M, Fabbri P. New insights into the pathogenesis of vitiligo: imbalance of epidermal cytokines at sites of lesions. Pigment Cell Res 2002; 15: 87-92.

7. Kadono S, Manaka I, Kawashima M, Kobayashi T, Imokawa G. The role of the epidermal endothelin cascade in the hyperpigmentation mechanism of lentigo senilis. J Invest Dermatol 2001; 116: 571-7.

8. Manaka L, Kadono S, Kawashima M, Kobayashi T, Imokawa G. The mechanism of hyperpigmentation in seborrhoeic keratosis involves the high expression of endothelin-converting enzyme-1alpha and TNF-alpha, which stimulate secretion of endothelin 1. Br J Dermatol 2001; 145: 895-903.

9. Kondo S. The role of keratinocyte-derived cytokines in the epidermis and their possible responses to UVA irradiation. J Invest Dermatol 1999; 4: 177-83.

10. Gupta S. A decision between life and death during TNF-α-induced signaling. J Clin Imm 2002; 22: 185-91.

11. Claudy AL, Schmitt D, Freycon F, Boucheron S. Melanocyte-lymphocyte interaction in human graft-versus-host disease. J Cutan Pathol 1983; 10(5): 305-11.

12. Au WY, Yeung CK, Chan HH, Lie AK. Generalized vitiligo after lymphocyte infusion for relapsed leukaemia. Br J Dermatol 2001; 145 (6): 1015-7.

13. Nagler A, Goldenhersh MA, Levi-Schaffer F, Bystryn JC, Klaus SN. Total leucoderma: a rare manifestation of cutaneous chronic graft-versus-host disease. Br J Dermatol 1996; 134 (4): 780-3.