Acral lesions of vitiligo: why are they resistant to photochemotherapy?

Transcription

1 DOI: /j x JEADV ORIGINAL ARTICLE Acral lesions of vitiligo: why are they resistant to photochemotherapy? S.M. Esmat, A.M. El-Tawdy, G.A. Hafez, O.A. Zeid,, * D.M. Abdel Halim, M.A. Saleh, T.M. Leheta, M. ElMofty Department of Dermatology, Faculty of Medicine, Cairo University, Cairo, Egypt, and Department of Pathology, Faculty of Medicine, Suez Canal University, Suez, Egypt *Correspondence: O.A. Zeid. agfouad@yahoo.com Abstract Background Acral lesions of vitiligo are usually resistant to conventional lines of treatment as well as surgical interventions. Objective To clarify causes underlying resistance of acral lesions to pigmentation in vitiligo by studying some of the factors associated with mechanisms of repigmentation following photochemotherapy. Methods The study included twenty patients with active vitiligo. Skin biopsies were taken from lesional and perilesional skin of areas expected to respond (trunk and proximal limb) and skin of acral areas, before and after PUVA therapy. Sections were stained with H and E, Melan-A, MHCII, CD1a, SCF and c-kit protein. Results Before treatment acral areas showed significantly lower hair follicle density, melanocyte density, Langerhans cell (LC) density, epidermal MHCII expression, lesional SCF expression and perilesional c-kit expression. Following treatment with PUVA in both non-responsive acral and repigmenting non-acral lesions identical immunohistochemical changes in the form of significant decrease in LC density, epidermal MHC-II and SCF expression were observed. Conclusion The surprisingly similar histochemical changes in response to PUVA in acral and non-acral lesions did not manifest with clinical repigmentation except in non-acral ones. Factors such as inherent lower melanocyte density, lower melanocyte stem cell reservoirs and or lower baseline epidermal stem cell factor may be considered as possible play makers in this respect. Received: 1 March 2011; Accepted: 20 July 2011 Conflicts of interest None. Introduction Ultraviolet radiation is considered the first line of therapy for vitiligo 1 that achieves partial repigmentation in 50 80% of patients. 2 TheresultsofPUVAtherapyvaryconsiderablyaccordingtothe site of vitiligo. The best results are obtained on the face, trunk and proximal parts of extremities. 3 Certain areas are known to resist repigmentation, including hands and feet, 4 lip-tip vitiligo, 5 association with leucotrichia 6 and segmental vitiligo. 7 Few reasons for resistance of the acral areas were suggested. The most popular theory is the relatively lower density or absence of pilosebaceous follicles, the reservoirs from which the melanocytes migrate. 8 Relatively less melanocyte density as well as higher chances of koebnerization over these friction and injury prone anatomical sites were other suggestions. 9 Although the resistance of acral lesions is very disappointing, to our knowledge nobody studied the causes of resistance of such lesions and all studies were directed towards their therapy. Langerhans cells (LC) are antigen presenting cells that express MHC II molecules which bind peptide antigens resulting from exogenous antigens degradation. 10 They are suggested to play a role in the pathogenesis of vitiligo 11 but their number in vitiligenous skin is controversial The proto-oncogene c-kit encodes a transmembrane tyrosine kinase receptor, which binds stem cell factor (SCF). 15 The interaction of SCF with its receptor, c-kit, is critical for the activation 16 and survival of melanocytes. 17 Some authors detected significant lower expression of SCF in vitiligenous skin, 18 whereas others reported its up-regulation. 19 This study was carried out in an attempt to find the differences that make acral vitiligo lesions resist repigmentation compared to other sites of the body that usually respond to PUVA therapy. We tried to achieve this by studying some of the factors that have been reported to change after PUVA therapy and verified the fundamentals for successful repigmentation; namely, LC density and

2 1098 Esmat et al. antigen-presenting capacity 20 and stem cell factor (SCF) expression. 21 Materials and methods A prospective controlled clinical study was done including twenty vitiligo patients who accepted to be volunteers in this study after signing a written consent. An approval from the local ethical committee was obtained. Patients selected were those suffering from generalized active vitiligo with both acral and non-acral lesions and did not receive any topical or systemic therapy for at least 3 months. Patients below 12 years were excluded. Routine laboratory investigations were carried out to exclude patients with any liver diseases or kidney dieases; an ophthalomic examination was also performed. Each patient was subjected to: history taking, thorough clinical examination, 4 mm punch skin biopsy and follow-up. Biopsy taking From each patient biopsy samples were taken from two areas: 1 Type A: Areas highly expected to repigment: upper back, trunk or limbs containing black hairs (Fig. 1). 2 Type B: Areas highly unlikely to repigment: acral lesions; (medial side of ankles or volar aspect of the wrist) were chosen (Fig. 1). Sun-exposed areas were excluded. Biopsies were taken from the lesional (within 5 mm from the edge of the lesion) and perilesional skin (<5 mm away from the lesion) of both type A and B areas before treatment and were repeated after PUVA therapy as soon as repigmentation started in either lesions. Therapy Patients received psoralen tablets 0.7 mg kg 3 times week and were exposed 90 min later to total body broadband UVA (BB- Type A before Type B before (c) (d) Type A after Type B after Figure 1 Type A lesions before and after treatment (c). Type B lesions before and after (d) treatment. Type A lesions are abdominal. Type B lesions are on the lateral maleolus.

3 Why is acral vitiligo resistant? 1099 UVA) phototherapy (UV1000 Waldmann lighting, Villingen Schwenningen, Germany) equipped with 26 lamps emitting a radiation spectrum of nm with a peak of 365 nm. Weekly follow-up Repigmenation was judged by two investigators throughout the study. Clinical development of perifollicular pigmentation, tanning or narrowing of the lesion was considered as evidence of repigmentation and time to take the post-treatment biopsies (Fig. 1). Each biopsy was fixed in formalin, embedded in paraffin and subjected to Haematoxylin and eosin staining Sections (4 lm thick) were cut,stainedbyh&eandwereexaminedblindly,toassessthe presence of inflammatory infiltrate, its distribution, component cells, its type and its magnitude. The infiltrate was assessed semi-quantitatively as mild (+) or moderate (++). The density of follicles and sweat glands were measured by dividing their number by the length of the basement membrane of the epidermis measured in mm using an image analysis software; Analysis (Soft Imaging System GmbH, Germany), installed at a computer attached to the microscope. CD1a, MHC- II, Melan A, SCF & c-kit staining Sections to be stained for Langerhans cells were incubated with the ready to use monoclonal anti-cd1a antibody RTU-CD1a-235 (Clone MTB1) (Novocastra Laboratories Ltd, Newcastle upon Tyne, UK) for 60 min. Sections to be stained for melan-a were stained using anti melan-a (Clone A103), NCL-MelanAmouse monoclonal antibody (Novocastra Laboratories Ltd, Newcastle upon Tyne, UK), at a dilution of 1 : 25 for 60 min. Sections to be stained for HLA-class II were incubated with DakoCytomation monoclonal mouse Anti- Human HLA-DP, DQ, DR Antigen (Clone CR3 43, Code no. M 0775) at a dilution of 1 : 50. Sections to be stained for SCF were incubated with the rabbit polyclonal anti-scf primary antibody (Chemicon, Catalogue number AB1498P) diluted at 1 : 100. Sections to be stained for c-kit were stained using the ready to use mouse monoclonal antic-kit antibody RTU-CD117 (Clone T595, Novocastra Laboratories Ltd, Newcastle upon Tyne, UK) for 30 min at room temperature. Binding of the primary antibodies was detected by the DakoCytomation LSAB2 System- HRP Code K0675 by sequential 10-min incubations with a biotinylated link antibody (containing anti-rabbit and anti-mouse immunoglobulins) and peroxidase-labelled streptavidin. Staining was completed after incubation with the substrate chromogen (AEC) for 10 min. Counterstaining was performed by putting the slides in Mayer s haematoxylin for 5 min and mounting was done in glycerol (Sigma). Immunostained sections were examined blindly, for the presence of cytoplasmic positivity of Melan-A, CD1a, and MHC II. Epidermal and dermal MHC II were assessed semi-quantitatively as weak (+) or strong (++). Epidermal CD1a positive cells were counted in all the specimens. Only the cell bodies were counted. Dendrites without cell bodies were ignored. The density of CD1a positive cells were obtained by dividing the number of CD1a by the length of the basement membrane of the epidermis measured in mm using image analysis software called Analysis (Soft Imaging System GmbH, Germany) that was installed in a computer attached to the microscope. SCF staining was described as either strong when it was marked and extensive involving the whole epidermis or weak when it showed weak staining most noticeable in the lower epidermis. Statistical methods The results of the LC density were expressed using the mean and the standard deviation of the density in all the specimens. Comparisons were performed using the t-test for the difference between paired observations. The difference was considered significant if P was <0.05. The results of SCF were analysed using McNemar s test for comparison of paired proportions. Results A-Clinical results Twenty patients were the subjects of this study (12 female patients and 8 male patients). Their ages ranged years. The disease duration ranged years. Patients were Fitzpatrick skin type III and IV. All patients had active progressive vitiligo. Type A lesions responded to treatment after a mean of 12 weeks, whereas type B lesions did not show any form of pigmentation by this time. B-Microscopic examination of H&E sections Inflammatory infiltrate A perivascular lymphocytic infiltrate of variable densities was observed in the upper dermis which was significantly higher in lesional vs. perilesional areas in both type A and B skin. No significant differences were detected between type A and B lesional and perilesional skin before or after therapy. Hair follicle density Was significantly lower in type B skin (lesional and perilesional) (0.01+, )0.03 mm of epidermal basement membrane length) compared to type A skin (lesional and perilesional) (0.37+, )0.21 mm of epidermal basement membrane length). Sweat gland density Was significantly higher in type B skin (lesional and perilesional) (0.54+, )0.24 mm of epidermal basement membrane length) compared to type A skin (lesional and perilesional) (0.24+, )0.22 mm of epidermal basement membrane length). C-Immunohistochemical staining Melan-A staining Melanocytes were absent from all lesional skin before treatment and appeared only in responding lesional skin

4 1100 Esmat et al. after treatment. Perilesional melanocytes in type A were significantly higher than in type B both before and after PUVA. Increased melanocyte density in perilesional skin was noted after PUVA in both types without statistical significance (Table 1). CD1a The CD1a positive Langerhans cells were only observed in the epidermis showing their dendritic nature and occupying their usual suprabasal location (Fig. 2). Before treatment, both type A andbskinlesionsshowedhighercd1a+lccomparedtoperilesional skin. On comparing both types together, type B lesions showed significantly lower density than type A lesional skin. After treatment, both types showed reduction of LC density in lesional and perilesional skin but it was significant only in lesional skin (Table 2). MHC class II The MHC II positive cells were seen in all specimens in the epidermis and dermis (Fig. 3). The positive epidermal staining occurred on Langerhans cells and keratinocytes. Epidermal MHC II showed a statistically significant lower expression in type B than in type A skin in lesional skin before treatment. It was significantly higher in both lesional skin of type A and B compared to perilesional skin (Table 2). Both types showed a significant Table 1 Comparison between the number of perilesional melanocytes per 100 basal keratinocytes (as detected in melan A stained sections) between type A and B lesions (before and after treatment) Type A Type B Mc Nemar s test Perilesional before ± ± P < 0.01** Perilesional after ± ± P < 0.01** Mc Nemar s test P > 0.05* P > 0.05* *Not significant, **Highly significant. Lesional skin Perilesional skin Figure 2 Lesional skin showing increased CD1a stained cells, Perilesional skin showing decreased CD1a stained cells. Table 2 Comparison between the immunohistochemical findings in type A and type B lesions (I) before treatment, (II) after treatment (I) Before treatment P-value (II) After treatment P-value Type A Type B Type A Type B CD1a Lesional 7.09 ± ± 1.64 P < 0.005* 2.35 ± ± 1.24 P < 0.005* Perilesional 2.09 ± ± 2.32 P > ± ± 2.4 P < 0.005* P-value P < 0.005* P > 0.05 P > 0.05 P > 0.05 MHCII Lesional Perilesional P < 0.05* P > P-value P < 0.01* P < 0.05* P < 0.05* P < 0.05* SCF Lesional Perilesional P < 0.01* P > P-value P < 0.01* P > 0.05 P > 0.05 P > 0.05 *Signficiant; +, weak; ++, strong P > 0.05 P > 0.05 P > 0.05 P > 0.05

5 Why is acral vitiligo resistant? 1101 Figure 3 Lesional skin showing increased epidermal MHC II stained cells. Perilesional skin showing decreased epidermal MHC II stained cells. Lesional Perilesional Type A lesions before PUVA (strong staining) Type A lesions a er PUVA (c) (d) Figure 4 SCF expression before and after PUVA in both type A and type B lesions. type A lesion showed strong expression of SCF compared to type B (c). A statistically significant reduction was noted between pre- and post-treatment results in both groups A and B skin (b d). Type B lesions before PUVA Type B lesions a er PUVA (weak staining) reduction of MHC II after treatment in both lesional and perilesional skin. Stem cell factor staining Epidermal SCF staining was seen mostly as cytoplasmic staining with occasional nuclear staining (Fig. 4). It was seen in the sweat glands and to a lesser extent in hair follicles of vitiliginous and normally pigmented skin. Before treatment, lesional skin in group A and B showed higher expression of SCF compared to perilesional skin but this difference was significant only in group A. In addition, lesional type B showed lower expression of SCF compared to type A and the difference was statistically significant (Table 2). A statistically significant reduction of SCF expression was noted after treatment in both group A and B skin (Fig. 4). In Type B lesional skin, all cases showing strong expression of SCF before treatment showed weak staining of SCF after treatment. In type A lesional skin, 14 of the 20 cases showing strong expression of SCF before treatment, showed weak staining of SCF after treatment. C-kit staining Perilesional sections stained for c-kit revealed the presence of c-kit positive melanocytes at the lower epidermis above the basement membrane in the form of membranous positivity in all specimens (Fig. 5). C-kit positive cells were absent from all lesional skin before treatment except in the infundibula of hair follicles. Perilesional skin of type B showed significant lower

6 1102 Esmat et al. Perilesional skin Lesional skin Figure 5 Pretreatment perilesional skin showing c-kit positive melanocytes at the lower epidermis above the basement membrane. Weak staining of the adjacent basal keratinocytes is also seen. Pretreatment lesional vitiligo skin showing no c-kit positive cells in the epidermis. Table 3 C-Kit expression in type A and B perilesional skin before and after treatment Type A Type B Paired t-test Pre-treatment ± ± P < 0.01* Post-treatment ± ± P < 0.01* Paired t-test P > 0.05 P > 0.05 *Significant. expression of c-kit compared to perilesional skin of type A before and after treatment (Table 3). Post-treatment positive staining with c-kit occurred in type A lesions only. Discussion Acral lesions of vitiligo are known to resist repigmentation even with surgical treatment. 22 Comparison between lesional and perilesional skin in areas responding to PUVA (type A) and acral areas (type B) resistant to PUVA revealed interesting findings. A lymphocytic perivascular upper dermal infiltrate was found in all sections. There was no significant difference in the extent of this infiltrate between type A and type B lesions indicating that both acral, poorly responsive, and non-acral, responding, vitiligo skins were similar with regard to the presence of inflammation. The density of the infiltrate was significantly higher in lesional compared to perilesional skin in contrast to studies that showed that the infiltrate is more marked in perilesional skin. 11,23 This is probably because all our cases were active and progressive. This confirmed the suggestion that vitiligo is an inflammatory disease where the dermal lymphocytic infiltrate is the most likely first immunological effect. 24 Staining with melan A revealed that the melanocyte density in perilesional skin in type B lesions was significantly lower compared to type A perilesional skin, which might suggest an anatomical variation in the number of melanocytes. AfterPUVAtherapy,bothtypeAandtypeBperilesionalskin responded similarly by increasing perilesional melanocyte density, but only type A showed clinical repigmentation of the lesions. Failure of repigmentation in type B may be explained by the lower original melanocyte count in these areas. However, a defect in motility of melanocytes or a defect in the transfer of melanosomes could be suggested, as migration of melanocytes upwards along the outer root sheath and radially in the vitiligenous skin have been considered essential for repigmentation. 25 Regarding Langerhans cells, CD1a positive cells were found to be significantly higher in lesional than perilesional skin in both groups indicating the important role of LC in the pathogenesis of active vitiligo lesions. When both type A and type B skins were compared before treatment, type B lesions showed significantly lower LC density compared to type A and this difference remained significant after treatment. The lower density of LC in type B lesions could be explained in different ways. Migration to regional lymph nodes to present antigens released from destroyed melanocytes during the course of active vitiligo may be considered. In addition, an anatomical variation in LC density in acral areas supported by lower perilesional LC density in type B than type A, though not statistically significant, is yet another consideration. Furthermore, destruction by cytotoxic factors released during the course of melanocyte damage has been suggested by Kao and Yu 26 as a possible cause of lowered LC density in active vitiligo. Thus, the lower number of LC in acral lesions might suggest a more aggressive lesion. After PUVA therapy there was a significant reduction of LC in both type A and type B lesions. This shows that type B lesions responded to PUVA the same way as type A lesions by reducing LC density, which is one of the mechanisms suggested for PUVA induced repigmentation. 20 However, this decrease was not associated with clinical repigmentation. Acral lesions showed lower expression of epidermal MHCII indicating lower density of antigen-presenting cells in the resistant sites. After PUVA both types responded by significant reduction in epidermal MHCII expression but only type A was associated with

7 Why is acral vitiligo resistant? 1103 clinical repigmentation. The previous findings might indicate that the reduction in LC density and MHCII are not the only events required to induce repigmentation by PUVA, and that acral skin lacks one or more of the other factors needed to complete the process of repigmentation. The pretreatment SCF epidermal expression in type A and B skin was significantly higher in lesional than perilesional skin, this was in agreement with Kitamura et al. 21 who found an accentuated expression of SCF and endothelin-1 in lesional vitiligo skin vs. non-lesional skin. However, Moretti et al. 18 demonstrated a lower expression of SCF, GM-CSF and b-fgf in lesional vitiligo. The controversy in these findings might indicate that epidermal SCF level changes at different stages of the disease. Increased expression of SCF was reported in hyperpigmentary disordersasdermatofibroma, 27 mastocytosis 28 and lentigo senilis 29 indicating the importance of stem cell factor c-kit pathway in proliferation of melanocytes and melanogenesis. It could be suggested that excessive production of SCF by keratinocytes in lesional vitiligo skin could be a trial of keratinocytes to stimulate melanogenesis in response to the existing melanocyte damage. Significant lower expression of SCF was noticed in type B lesions compared to type A. This could reflect a contributing factor such as keratinocyte dysfunction in acral areas that may clarify the difference in clinical response to therapy. Apoptotic keratinocytes may synthesize less SCF than non-apoptotic cells do, with a consequent influence on neighbouring melanocyte survival in the lesions of vitiligo. 30 Thus the defect in SCF production may be one of the contributing factors in the pathogenesis of resistance in these sites. After photochemotherapy, the expression of SCF decreased significantly in both groups. This decrease may be related to consumption of SCF by binding to its receptor c-kit to induce melanogenesis. This decrease was associated with repigmentation in group A only but not in group B which might indicate that the amount of SCF is insufficient in group B to bind to its receptor and produce a clinical effect. Complete absence of c-kit was noted in lesional areas in both groups. Similar findings were obtained by Norris et al. 19 and Kitamura et al. 21 who suggested that decreased expression of c-kit protein and its downstream effector MITF-M by melanocytes may be associated with the dysfunction or loss of melanocytes in vitiligo. The failure of epidermal expression of c-kit receptor was associated with significantly higher expression of SCF by keratinocytes in vitiligo lesions in both types A and B. The lower expression of c-kit receptor might be a factor that causes excessive expression of its ligand SCF as both of them are connected by a negative feedback mechanism where excessive c-kit production decreases SCF expression through CbI protein 31. Our observation of the presence of strong expression of SCF with negative c-kit in the sweat glands can support this suggestion. After PUVA therapy, c-kit receptor was expressed in the lesional skin in responding areas in group A, but still negative in group B lesions. This is probably due to failure of appearance of melanocytes in acral lesions after therapy. Therefore, acral vitiligo skin is significantly different from the clinically responding vitiligo skin. It does not only have lower density of pilosebaceous follicles, but also lower density of perilesional melanocytes and higher density of sweat glands. The lesions also show significantly lower SCF, c-kit, MHCII expression as well as lower density of LC. After exposure to PUVA the acral areas showed the same immunohistochemical changes as other responding areas (a decrease in LC count, MHCII expression and SCF expression and an insignificant rise in perilesional melanocyte density) but they revealed neither clinical repigmentation nor histological appearance of melanocytes in the lesions. We could therefore conclude that acral vitiligo lesions respond to PUVA therapy the same way responding lesions do, but these changes fail to induce repigmentation. This might be due to the originally lower density of melanocytes, melanocyte stem cell reservoir as well as the lower SCF production. The cytotoxic potential of acral skin as well as melanocyte motility and other melanocyte mitogens need to be studied. Acknowledgement The authors thank Dr Wedad Mostafa, Professor of Dermatology, Cairo University, for her help in revising the editing of this manuscript. References 1 Falabella R, Barona MI. Update on skin repigmentation therapies in vitiligo. Pigment Cell Melanoma Res 2008; 22: Vussuki E, Ziv M, Rosenman D, David M. Long-term effects of PUVA therapy on Israeli patients with vitiligo. Harefuah 2006; 145: Hann S. Vitiligo Emedicine. [WWW Document] URL (last accessed: 15 December 2008). 4 Taieb A, Picardo M. The definition and assessment of vitiligo: a consensus report of the vitiligo European Task Force. Pigment Cell Research Journal 2007; 20: Panvelkar V, Talwar S, Tiwari VD. Koebner phenomenon. A clinical test to determine therapy and prognosis in vitiligo. Indian J Dermatol Venereol Leprol 1992; 58: Gupta S, Narang T, Olsson MJ, Ortonne JP. Surgical management of vitiligo and other leucodermas: evidence-based practice guideline. In Gupta S, Olsson MJ, Kanwar AJ, Ortonne JP, eds. Surgical Management of Vitiligo, 1st edn. Blackwell publishing Ltd, Oxford, 2007: Mulekar S. Long term follow up study of segmental and focal vitiligo treated by autologous Non cultured melanocyte-kertainocyte cell transplantation. Arc Dermatol 2004; 140: Roelandts Rik. Photochemotherapy for vitiligo. Photodermatology, Photoimmunology and Photomedicine Journal 2003; 19: Mutalik S. Surgical management of acral vitiligo. In Somesh G, Mats O, Amrinder K, Jean-Paul O, eds. Surgical Management of Vitiligo, Blackwell publishing Ltd, Oxford, UK, 2007; Mommas A, Koning F. Immunoelectron microscopic analysis of major histocompitability class II expression on human epidermal Langerhan s cells. In Heidrun Moll, ed. The immune function of epidermal Langerhans cells. Medical intelligence unit. RG Landes Springerverlag, Austin, 1995:

8 1104 Esmat et al. 11 Wijingaard Van den R, Wankowicz-Kalinska A, Le Poole C et al. Local Immune response in skin of Generalized vitiligo patients. Laboratory invest Journal 2000; 80: Claudy A, Rouchouse B. Langerhan s cell and vitiligo: quantitative study of T6 and HLA-DR antigen-expressing cells. Acta Derm Venereol 1984; 64: Caravelho E, Guedes A, Castro L. Quantitative study of Langerhans cells in vitiligo. Dermatology journal 2002; 77: Yagi H, Tokura Y, Furukawa F. Vitiligo with raised inflammatory borders: involvement of T cell immunity and keratinocytes expressing MHC class II and ICAM I molecules. Eur J Dermatol 1997; 7: Anderson DM, Lyman SD, Baird A et al. Molecular cloning of mast cell growth factor, a hematopoietin that is active in both membrane bound and soluble forms. Cell 1990; 63: Grichnik JM, Burch JA, Burchette J, Shea CR. The SCF KIT pathway plays a critical role in the control of normal human melanocyte homeostasis. J Invest Dermatol 1988; 111: Murphy M, Reid K, Williams DE et al. Steel factor is required for maintenance, but not differentiation, of melanocyte precursors in the neural crest. Dev Biol 1992; 153: Moretti S, Spallanzani A, Amato L et al. New insights into the pathogenesis of vitiligo: imbalance of epidermal cytokines at sites of lesions pigment. Cell Res 2002; 15: Norris A, Todd C, Graham A et al. The expression of the c-kit receptor by epidermal melanocytes may be reduced in vitiligo. Br J Dermatol 1996; 134: Schneider L, Hinrichs R, Scharffetter-kochanek K. Phototherapy and photochemotherapy. Clinics in dermatology journal 2008; 26: Kitamura R, Tsukamoto K, Harada K et al. Mechanisms underlying the dysfunction of melanocytes in vitiligo epidermis: role of SCF KIT protein interactions and the downstream effector, MITF-M. J Pathol 2004; 202: Falabella R. Surgical approaches for stable vitiligo. Dermatol Surg 2005; 31: Al Badri AM, Foulis AK, Todd P et al. Abnormal expression of MHC class II snd ICAM-1 by melanocytes in vitilgo. J pathol 1993; 169: Sharquie KE, Mehenna SH, Naji AA, Hani A. Inflammatory changes in vitiligo stage I and II depigementation. Am J Dermatopathol 2004; 26: Jian Cui, Li-Yu Shen, Guang-chao Wang. Role of hair follicles in repigmentation of vitiligo. Journal of investigative dermatology 1991; 97: Kao CH, Yu HS. Comparison of the effect of 8-methoxypsoralen (8-MOP) plus UVA (PUVA) on human melanocytes in vitiligo vulgaris and in vitro. J Invest Dermatol 1992; 98: Imokawa G. Autocrine and paracrine regulation of melanocytes in human skin and in pigmentary disorders. Pigment Cell Res 2004; 17: Longley BJ, Morganroth GS, Tyrrell L et al. Altered metabolism of mast-cell growth factor (c-kit ligand) in cutaneous mastocytosis. N Engl J Med 1993; 328: Hattori H, Kawashima M, Ichikawa Y, Imokawa G. The epidermal stem cell factor is over-expressed in lentigo senilis: implication for the mechanism of hyperpigmentation. J Invest Dermatol 2004; 122: Lee AY, Kim NH, Choi WI, Youm YH. Less keratinocyte-derived factors related to more keratinocyte apoptosis in depigmented than in normally pigmented suction-blistered epidermis may cause passive melanocyte death in vitiligo. J Invest Dermatol 2005; 124: Zeng S, Xu Z, Lipkowitz S, Longley JB. Regulation of stem cell factor receptor signaling by Cbl family proteins (Cbl-b c-cbl). Blood 2005; 105: