- Email: [email protected]
Clinical and histopathologic characteristics of trichrome vitiligo
Clinical and histopathologic characteristics of trichrome vitiligo Seung-Kyung Hann, MD,a Yi-Sun Kim, MD,a Jung Hoan Yoo, MD,a and Yoon-Sun Chun, MDb Seoul, Korea Background: The term trichrome vitiligo describes lesions that have a tan zone of varying width between normal and totally depigmented skin, which exhibits an intermediate hue. However, the pathogenesis and the histopathologic characteristics of trichrome vitiligo are unknown. Objective: Our purpose was to investigate the clinical and histopathologic characteristics and the pathogenesis of trichrome vitiligo. Methods: Four punch biopsy specimens were taken from 21 patients with trichrome vitiligo; they were from vitiliginous skin, light brown skin, perilesional normal skin, and normal skin as far as 5 cm from the nearest vitiligo spot. The sections were stained with hematoxylin-eosin; in selected cases, we performed immunohistochemical staining with S-100 protein and CD1a. Results: Trichrome vitiligo occurred most frequently on the trunk in active vitiligo vulgaris. Focal vacuolar degeneration of the basal cell layer and mild inflammatory cell infiltration of the epidermis and dermis were more prominent in light brown skin and perilesional normal skin than in vitiliginous skin and normal skin. The number of melanocytes was decreased in light brown skin compared with perilesional normal skin (P < .05) and in vitiliginous skin compared with light brown skin (P < .05); a few melanocytes were observed even in skin affected by trichrome vitiligo. The number of Langerhans cells was increased in the epidermis of light brown skin and perilesional normal skin compared with vitiliginous and normal skin (P < .05). PUVA therapy yielded excellent repigmentation. Conclusion: Trichrome vitiligo is a variant of active vitiligo. The changes of melanocytes, keratinocytes, and Langerhans cells may be involved in the pathogenesis of depigmentation in trichrome vitiligo. (J Am Acad Dermatol 2000;42:589-96.)
T
he term trichrome vitiligo was first suggested in 1964 by Fitzpatrick.1 The lesions have an intermediate zone of hypochromia located between the achromic center and the peripheral unaffected skin. This results in 3 shades of color—brown, tan, and white—in the same patient1,2 (Fig 1). The trichrome lesion naturally evolves to a typical vitiligo macule.3 The significance of trichrome is unknown, but it is clearly a metastable or transitional pigmentary state, though it may persist for months to years with little change.4 Fitzpatrick1 and Pincus5 interpreted From the Department of Dermatology, Yonsei University College of Medicine,a and the Pochon CHA Medical College, Pundang CHA Hospital.b Accepted for publication Dec 9, 1999. Reprint requests: Seung-Kyung Hann, MD, Department of Dermatology, Yonsei University College of Medicine, CPO Box 8044, Seoul, Korea. E-mail: [email protected]. Copyright © 2000 by the American Academy of Dermatology, Inc. 0190-9622/2000/$12.00 + 0 16/1/104896 doi:10.1067/mjd.2000.104896
Fig 1. This patient shows the characteristic light brown skin between normal and vitiliginous skin.
trichrome as suggestive of a gradual centrifugal spread of hypomelanosis or a stepwise depigmentation. However, other reports pointed out that the sharp demarcation between the 3 areas in their cases, 589
590 Hann et al
J AM ACAD DERMATOL APRIL 2000
Table I. Clinical characteristics of trichrome vitiligo Patient No.
Sex
Age (y)
Duration (y)
Type
Lesion site
Activity
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
M F F F M F M F M F M M M M M F F M M F F
60 18 11 32 45 51 32 30 19 62 34 29 40 33 26 35 20 25 34 56 9
4 4 3 14 7 15 1 13 2 3 8 10 25 0.6 14 20 3 10 5 3 1
Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Focal
Back Back Abdomen Arm Buttock Back Buttock Back Abdomen Abdomen Back Back Back Abdomen Back Back Back Back Chest Back Leg
Stable Spreading Spreading Spreading Spreading Spreading Spreading Spreading Spreading Spreading Spreading Stable Spreading Spreading Stable Spreading Spreading Spreading Spreading Spreading Spreading
MATERIAL AND METHODS
Table II. Location of trichrome vitiligo Location
No. of patients (%)
Back Abdomen Buttock Chest Arm Leg Total
12 (57) 4 (19) 2 (9.5) 1 (4.8) 1 (4.8) 1 (4.8) 21 (100)
as well as the lack of gradual changes of color and the stability of the lesion, is inconsistent with the interpretation of trichrome vitiligo as an active centrifugal spreading lesion.4 Therefore, whether trichrome vitiligo is a temporary phenomenon of active spreading vitiligo or a hypomelanosis showing an unusual progression pattern, remains to be defined. The present study was designed to observe detailed changes of the epidermis and dermis in vitiliginous skin, light brown skin, perilesional normal skin, and normal skin as far as 5 cm from the nearest vitiligo spot in trichrome vitiligo. The sections were stained with hematoxylin-eosin, and immunohistochemical staining was performed in selected cases. The clinical and histopathologic features of trichrome vitiligo were studied to clarify its characteristics and pathogenesis.
Of 323 patients with vitiligo seen at the vitiligo clinic in Severance Hospital of Yonsei Medical Center from January through October 1996, evaluation of 21 patients with trichrome vitiligo was performed. Responses to questionnaires regarding type, activity, sex, age at onset, age at initial visit, involved sites, family history, and associated disease were recorded. Four-millimeter punch biopsy specimens of the skin were obtained from 21 patients with trichrome vitiligo. Each patient underwent 4 biopsies: vitiliginous skin, light brown skin, perilesional normal skin, and normal skin as far as 5 cm from the nearest vitiligo spot. The sections were stained with hematoxylin-eosin to study the inflammatory cell infiltration in the epidermis and dermis, vacuolar degeneration of basal cell layer, and number of melanophages. The number of melanophages was counted in two separate sections of each 4-mm punch biopsy specimen, and the mean value was obtained. Immunohistochemical staining with S-100 protein and CD1a was performed in 10 patients showing inflammatory cell infiltration in hematoxylin-eosin staining. Positive cells with S-100 and CD1a were evaluated on 2 sections for each antibody with 3 fields per each section (×400). To count the accurate number of melanocytes in the basal cell layer, we stained the paired slide with CD1a and compared
Hann et al 591
J AM ACAD DERMATOL VOLUME 42, NUMBER 4
Table III. Histologic findings of trichrome vitiligo Inflammatory cell infiltration
Vitiliginous skin Light brown skin Perilesional normal skin Normal skin
Epidermis
Dermis
Vacuolar degeneration of basal cells
1 (4.8)* 13 (61.9) 16 (76.2)
6 (28.6) 12 (57.1) 16 (76.2)
5 (23.8) 13 (61.9) 19 (90.5)
2 (9.5)
3 (14.3)
7 (33.3)
*Number (percentage in parentheses) of positive cases.
the two paired slides for exclusion of CD1a+ Langerhans cells of the basal cell layer. The statistical significance of the results was calculated by the paired t test. Patients with trichrome vitiligo were treated with systemic PUVA, topical PUVA, systemic steroid therapy, or topical steroid therapy for at least 2 months depending on each patient’s situation.
RESULTS Clinical characteristics of trichrome vitiligo Table I summarizes the clinical characteristics of trichrome vitiligo. Of the 21 patients with trichrome vitiligo, 11 were male and 10 were female; their ages ranged from 9 to 62 years. The mean age at disease onset was 33.4 ± 14.8 years, and the mean duration was 7.9 ± 6.8 years. There were 20 cases of generalized type and 1 localized type disease. All the 20 patients with generalized type disease showed features of trichrome vitiligo restricted to a localized area, whereas the remaining areas displayed white spots with well-demarcated borders of typical vitiligo. At the time of the visit, 18 cases were active vitiligo (85.7%) and 3 cases were stable vitiligo (14.3%). The most commonly involved site was the back (57%) followed by abdomen (19%), buttock (9.5%), chest (4.8%), arm (4.8%), and leg (4.8%) (Table II). Histopathologic findings Light microscopic findings of the 4 discrete sites of biopsies revealed inflammatory cell infiltration in the epidermis occurring in 4.8% of vitiliginous skin, 61.9% of light brown skin, 76.2% of perilesional normal skin, and 9.5% of normal skin. Inflammatory cell infiltration in the dermis was 28.6% in vitiliginous skin, 57.1% in light brown skin, 76.2% in perilesional normal skin, and 14.3% in normal. Inflammatory cell infiltration was more prominent in light brown skin and perilesional normal skin than in vitiliginous skin and normal skin (Fig 2). Vacuolar degeneration of the basal cell layer was more prominent in light
Table IV. Number of melanophages in skin from patients with trichrome vitiligo Patient No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Mean ± SD†
VS
LBS
PLNS
NS
1.5* 4 1 0 0.5 0 1 0 2 0 0 2 0 0.5 0 3 7 1 0 0 3
5 16.5 14 3.5 6 18 10 25 23 9 7 6.5 0 7 8 4 13 5.5 15 2 17
3 13 13 8.5 8 17 16 25 5 19 4 3.5 0 8 24 12 3 0.5 5 5 15
0 0.5 0.5 0 0.5 0 0 8 3 2 2 0 0 0 3.5 0 10 2 7 1.5 0
9.9 ± 8.0
2.8 ± 4.0
1.3 ± 1.8 10.2 ± 6.8
LBS, Light brown skin; NS, normal skin; PLNS, perilesional normal skin; VS, vitiliginous skin. *Number of melanophages in two sections of 4-mm punch biopsy specimen. †VS and LBS: P < .05; PLNS and NS: P < .05.
brown skin (61.9%) and perilesional normal skin (90.5%) than in vitiliginous skin (23.8%) and normal skin (33.3%) (Table III). The mean number of melanophages in the 4-mm punch biopsy specimens was 1.3 ± 1.8 in vitiliginous skin, 10.2 ± 6.8 in light brown skin, 9.9 ± 8.0 in perilesional normal skin, and 2.8 ± 4.0 in normal skin. There was a significant increase in melanophages in light brown skin and perilesional normal skin compared with vitiliginous and normal skin (P < .05) (Table IV). Immunohistochemical findings The number of S-100+ cells in the basal cell layer of the epidermis was counted per 6 high-power fields (×400) and is summarized in Table V. The mean number of melanocytes was 4.8 ± 2.2 in vitiliginous skin, 11.9 ± 3.3 in light brown skin, 16.8 ± 3.0 in perilesional normal skin, and 16.3 ± 3.5 in normal skin. There was no statistically significant difference between perilesional normal skin and normal skin (P > .05). The number of melanocytes was decreased in light brown skin compared with perilesional nor-
592 Hann et al
J AM ACAD DERMATOL APRIL 2000
A
B
C
D Fig 2. Hematoxylin-eosin staining of trichrome vitiligo. Vacuolar degeneration of the basal cell layer and mild inflammatory cell infiltration in epidermis and dermis is more prominent in light brown skin (B) and perilesional normal skin (C) than vitiliginous skin (A) and normal skin (D) of trichrome vitiligo. (Original magnification ×200.)
mal skin (P < .05) and in vitiliginous skin compared with light brown skin (P < .05). A few melanocytes were observed even in vitiliginous skin of trichrome vitiligo (Fig 3). The number of CD1a+ Langerhans cells in the epidermis was counted per 6 high-power fields (×400); these data are summarized in Table VI and representative findings are illustrated in Fig 4. The mean number of CD1a+ Langerhans cells was 29.5 ± 9.8 in vitiliginous skin, 55.7 ± 11.1 in light brown skin, 54.3 ± 10.4 in perilesional normal skin, and 36.7 ± 8.6 in normal skin. There was no statistically significant difference between light brown skin and perilesional normal skin (P > .05). The number of Langerhans cells was increased in light brown skin and perilesional normal skin compared with vitiliginous and normal skin (P < .05). Treatment and response In trichrome vitiligo, treatment results were excellent with systemic PUVA therapy. The 21 patients
enrolled in this study received systemic PUVA, topical PUVA, systemic steroid therapy, or topical steroid therapy for at least 2 months based on their clinical manifestations, biopsy findings, and patient compliance. Among the 21 patients, 6 patients dropped out of the study and the remaining 15 patients were followed up for periodic observation of treatment results. Three patients (20.0%) experienced complete resolution of the lesions or marked improvement (excellent), 2 patients (13.3%) showed improvement of a moderate degree (good), 2 patients (13.3%) had some improvement (fair), 2 patients (13.3%) had slight or no improvement (poor), and 6 patients (40.0%) showed aggravation of the lesions (very poor). All 3 patients who showed excellent treatment results had been treated with systemic PUVA therapy. Eight patients were treated with systemic PUVA therapy and 3 (37.5%) of them showed an excellent response. Among the 5 patients treated with systemic steroid therapy, 4 patients (80%) responded very poorly.
Hann et al 593
J AM ACAD DERMATOL VOLUME 42, NUMBER 4
Table V. Numbers of S-100+ melanocytes in patients with trichrome vitiligo
Table VI. Number of CD1a+ Langerhans cells in patients with trichrome vitiligo
Patient No.
VS
LBS
PLNS
NS
Patient No.
1 2 3 4 5 6 7 8 9 10
— 5 5 2 4 5 8 8 4 2
— 14 15 14 8 15 13 8 13 7
23* 19 18 13 14 16 15 17 16 11
23 — 20 15 16 17 16 12 16 14
1 2 3 4 5 6 7 8 9
VS
LBS
PLNS
NS
— — 39 36 21 33 34 — 14
62* 52 67 28 50 65 — 52 34
73 58 54 44 50 47 60 44 36
36 47 37 — 45 24 31 33 28
Mean ± SD† 29.5 ± 9.8 55.7 ± 11.1 54.3 ± 10.4 36.7 ± 8.6
Mean ± SD† 4.8 ± 2.2 11.9 ± 3.3 16.8 ± 3.0 16.3 ± 3.5 For abbreviations see legend to Table IV. *Number of melanocytes per 6 high-power fields (×400). †PLNS and LBS: P < .05; LBS and VS: P < .05.
DISCUSSION In this study, we investigated whether the tan lesions of trichrome vitiligo and its surrounding perilesional normal skin show features of active vitiligo, both clinically and histologically. Among the 21 vitiligo patients showing trichrome lesions, 95.2% were classified as having vitiligo vulgaris and 85.7% had spreading lesions clinically. Histopathologic findings also showed the characteristics of active spreading vitiligo. Therefore trichrome vitiligo was regarded as a phenomenon occurring in restricted areas of active vitiligo. Of the trichrome lesions, 85.5% were localized to the trunk region, including the abdomen, back, and buttock, leading to the assumption that trichrome vitiligo predominates in unexposed skin. According to previous reports, sun-exposed areas are the predilection sites of vitiligo and lesions in these areas commonly show rapid progression. In contrast, the fact that trichrome vitiligo lesions of our patients predominated in unexposed skin could be one of the reasons the characteristic trichrome features appeared, possibly because of slow progression of the disease. Melanocyte density and skin thickness could also be factors contributing to the development of trichrome features. For example, the trunk shows the lowest melanocyte density of the skin with about 1100 to 1250/mm2 and is the thickest portion of the body.6 Therefore it is feasible to hypothesize that such factors eventually result in slow progression of the disease process. However, truncal lesions of patients with vitiligo are not always trichrome, and because trichrome vitiligo is known to occur on other parts of the body, factors other
For abbreviations see legend to Table IV. *Number of Langerhans cells per 6 high-power fields (×400). †LBS and PLNS compared with VS and NS: P < .05.
than sunlight, melanocyte density, and skin thickness seem to be associated with its pathogenesis. Blacks have a relatively higher frequency of trichrome vitiligo compared with whites and similarly most of the patients in our study had skin type IV or darker. Therefore dark skin also seems to be a contributing factor to the pathogenesis of trichrome vitiligo. The histologic findings of trichrome vitiligo showed the most dense distribution of melanin granules in the perilesional normal skin, followed by normal skin, light brown skin, and vitiliginous skin, in descending order. As such, the characteristic trichrome color may be an expression of changes in melanin granules rather than melanocyte numbers. Hyperpigmentation seen around the periphery of white patches is typically found in vitiligo. This was also observed in the trichrome lesions in which perilesional normal skin showed a slightly darker color compared with normal skin and histologically a higher density of melanin granules. Other histologic findings such as vacuolar degeneration of the basal cell layer, mononuclear cell infiltration of the epidermis and dermis, and melanophage deposition in the dermis were more prominent in light brown skin and perilesional normal skin than in vitiliginous and normal skin. Among these changes, vacuolar degeneration of the basal cell layer and inflammatory cell infiltration were especially accentuated around the melanocytes in the basal cell layer. However, overall destruction of keratinocytes coexisted, and we presume that the target of destruction is not limited to melanocytes but also involves keratinocytes as well. Perilesional normal skin shows more severe vacuolar
594 Hann et al
J AM ACAD DERMATOL APRIL 2000
A
B
C
D Fig 3. S-100 immunohistochemical staining of trichrome vitiligo. Number of melanocytes is decreased in light brown skin (B) compared with perilesional normal skin (C) and normal skin (D) and in vitiliginous skin (A) compared with light brown skin (B). A few melanocytes are observed even in vitiliginous skin of trichrome vitiligo (A). (Original magnification ×200.)
change of the basal layer and inflammatory cell infiltration than light brown skin and histologically is already in the process of vitiligo evolution, despite its normal-looking appearance. Therefore, without appropriate treatment, a change clinically into light brown skin can be predicted. Light brown skin clinically and histologically shows features of active vitiligo, although the degree of histologic change is subtle compared with perilesional normal skin. The histologic features of light brown skin and perilesional normal skin are congruent with findings of active vitiligo lesions of more than 1 year’s duration showing vacuolar degeneration of basal cell layer, mononuclear cell infiltration of the epidermis and dermis, and melanophage deposition.7,8 In view of such similarities, trichrome vitiligo could be a variant form of active vitiligo. The number of melanocytes in trichrome vitiligo was greatest in perilesional normal skin followed by light brown skin and vitiliginous skin, with vitiliginous skin showing at least a few melanocytes, albeit
less than that of normal skin. Our results are contradictory with other studies of vitiligo,7,9,10 which report absence of melanocytes in the depigmented patches confirmed by immunohistochemical staining or electron microscopy. The clinical and histologic findings of trichrome vitiligo suggest a slower progression of lesions than typical vitiligo, and this could be why melanocytes remain in the white patches. However, studies of the functional capacity of the remaining melanocytes should be conducted to elucidate this matter. Langerhans cells may play a major immunologic role in vitiligo. Interaction between keratinocytes, melanocytes, and Langerhans cells is thought to initiate depigmentation, but the exact mechanism is unknown. In patients with nonsegmental type vitiligo, a marked depletion of Langerhans cells was noted in active lesions and a repopulation of Langerhans cells was noted in stable lesions.11 In inflammatory vitiligo, an increase in Langerhans cells was observed in adjacent normal skin compared
Hann et al 595
J AM ACAD DERMATOL VOLUME 42, NUMBER 4
A
B
C
D Fig 4. CD1a immunohistochemical staining of trichrome vitiligo. Number of Langerhans cells is increased in light brown skin (B) and perilesional normal skin (C) compared with vitiliginous skin (A) and normal skin (D). (Original magnification ×100.)
with vitiliginous skin and normal skin.12 We observed changes in Langerhans cells in 9 patients with trichrome vitiligo and found that light brown skin and perilesional normal skin exhibit an increase in Langerhans cell number compared with vitiliginous skin and normal skin. From our findings, an increased number of Langerhans cells may be involved in actively spreading vitiligo. The pathogenesis of vitiligo is difficult to explain with a single theory; among the various theories set forth, autoimmunity is currently the most supported. However, recently not only humoral immunity but also cell-mediated immunity have been recognized as playing an important role and abnormalities of T lymphocytes in vitiligo have been studied.13,14 Our results, such as inflammation of the epidermis and dermis in light brown skin and perilesional normal skin as well as changes in melanocytes and Langerhans cells, all suggest that cell-mediated immunity may be involved in the development of trichrome vitiligo. Vitiligo lesions of the trunk are known to respond favorably to systemic PUVA therapy in comparison to
systemic steroid therapy, and the existence of inactive melanocytes in the epidermis or follicles is a decisive factor influencing treatment results.15,16 Trichrome vitiligo responded especially well to systemic PUVA treatment and we think this is because a few melanocytes were still remaining in the white lesions, thereby contributing to the repigmentation process. Therefore early systemic PUVA therapy should be considered in patients with trichrome features to shorten treatment duration and achieve satisfactory end results. REFERENCES 1. Fitzpatrick TB. Hypomelanosis. South Med J 1964;57:995-1005. 2. Fargnoli MC, Bolognia JL. Pentachrome vitiligo. J Am Acad Dermatol 1995;33:853-6. 3. Mosher DB, Fitzpatrick TB, Ortonne JP, Hori Y. Hypomelanoses and hypermelanoses. In: Freedberg IM, Eisen AZ,Wolff K, Austen KF, Goldsmith LA, Katz SI, et al, editors. Dermatology in general medicine. 5th ed. New York: McGraw-Hill; 1999. p. 945-1017. 4. Ortonne JP, Mosher DB, Fitzpatrick TB, editors.Vitiligo and other hypomelanoses of hair and skin. New York: Plenum Publishing; 1983. p. 147-8. 5. Pincus H. Vitiligo: what is it? J Invest Dermatol 1959;32:281-4.
596 Hann et al
6. Rosdahl I, Rorsman H. An estimate of the melanocyte mass in human. J Invest Dermatol 1983;81:278-81. 7. Hann SK, Park YK, Lee KG, Choi EH, Im S. Epidermal changes in active vitiligo. J Dermatol 1992;9:217-22. 8. Gokhale BB, Mehta LN. Histopathology of vitiliginous skin. Int J Dermatol 1983;22:477-80. 9. Le Poole IC, Das PK, van den Wijngaard RMJGJ, Bose JD, Westerhof W. Review of the etiopathomechanism of vitiligo: a convergence theory. Exp Dermatol 1993;2:145-53. 10. Le Poole IC, van den Wijngaard RMJGF, Westerhof W, Dutrieux RP, Das PK. Presence or absence of melanocytes in vitiligo lesions: an immunohistochemical investigation. J Invest Dermatol 1993;100:816-22. 11. Kao CH, Yu HS. Depletion and repopulation of Langerhans cells in nonsegmental type vitiligo. J Dermatol 1990;17:280-96.
J AM ACAD DERMATOL APRIL 2000
12. Le Poole IC, van den Wijngaard RMJGJ, Westerhof W, Das PK. Presence of T cells and macrophages in inflammatory vitiligo skin parallels melanocyte disappearance. Am J Pathol 1996; 148:1219-28. 13. Grimes PE, Ghoneum M, Stockton T, Payne C, Kelly AP, Alfred L. T cell profiles in vitiligo. J Am Acad Dermatol 1986;14:196-201. 14. Hann SK, Park YK, Chung KY, Kim HI, Im S, Won JH. Peripheral blood lymphocyte imbalance in Koreans with active vitiligo. Int J Dermatol 1993;32:286-9. 15. Ortonne JP, Schmitt D, Thivolet J. PUVA-induced repigmentation of vitiligo: scanning electron microscopy of hair follicles. J Invest Dermatol 1980;74:40-2. 16. Cui J, Shen L,Wang G. Role of hair follicles in the repigmentation of vitiligo. J Invest Dermatol 1991;97:410-6.
T
he term trichrome vitiligo was first suggested in 1964 by Fitzpatrick.1 The lesions have an intermediate zone of hypochromia located between the achromic center and the peripheral unaffected skin. This results in 3 shades of color—brown, tan, and white—in the same patient1,2 (Fig 1). The trichrome lesion naturally evolves to a typical vitiligo macule.3 The significance of trichrome is unknown, but it is clearly a metastable or transitional pigmentary state, though it may persist for months to years with little change.4 Fitzpatrick1 and Pincus5 interpreted From the Department of Dermatology, Yonsei University College of Medicine,a and the Pochon CHA Medical College, Pundang CHA Hospital.b Accepted for publication Dec 9, 1999. Reprint requests: Seung-Kyung Hann, MD, Department of Dermatology, Yonsei University College of Medicine, CPO Box 8044, Seoul, Korea. E-mail: [email protected]. Copyright © 2000 by the American Academy of Dermatology, Inc. 0190-9622/2000/$12.00 + 0 16/1/104896 doi:10.1067/mjd.2000.104896
Fig 1. This patient shows the characteristic light brown skin between normal and vitiliginous skin.
trichrome as suggestive of a gradual centrifugal spread of hypomelanosis or a stepwise depigmentation. However, other reports pointed out that the sharp demarcation between the 3 areas in their cases, 589
590 Hann et al
J AM ACAD DERMATOL APRIL 2000
Table I. Clinical characteristics of trichrome vitiligo Patient No.
Sex
Age (y)
Duration (y)
Type
Lesion site
Activity
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
M F F F M F M F M F M M M M M F F M M F F
60 18 11 32 45 51 32 30 19 62 34 29 40 33 26 35 20 25 34 56 9
4 4 3 14 7 15 1 13 2 3 8 10 25 0.6 14 20 3 10 5 3 1
Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Vulgaris Focal
Back Back Abdomen Arm Buttock Back Buttock Back Abdomen Abdomen Back Back Back Abdomen Back Back Back Back Chest Back Leg
Stable Spreading Spreading Spreading Spreading Spreading Spreading Spreading Spreading Spreading Spreading Stable Spreading Spreading Stable Spreading Spreading Spreading Spreading Spreading Spreading
MATERIAL AND METHODS
Table II. Location of trichrome vitiligo Location
No. of patients (%)
Back Abdomen Buttock Chest Arm Leg Total
12 (57) 4 (19) 2 (9.5) 1 (4.8) 1 (4.8) 1 (4.8) 21 (100)
as well as the lack of gradual changes of color and the stability of the lesion, is inconsistent with the interpretation of trichrome vitiligo as an active centrifugal spreading lesion.4 Therefore, whether trichrome vitiligo is a temporary phenomenon of active spreading vitiligo or a hypomelanosis showing an unusual progression pattern, remains to be defined. The present study was designed to observe detailed changes of the epidermis and dermis in vitiliginous skin, light brown skin, perilesional normal skin, and normal skin as far as 5 cm from the nearest vitiligo spot in trichrome vitiligo. The sections were stained with hematoxylin-eosin, and immunohistochemical staining was performed in selected cases. The clinical and histopathologic features of trichrome vitiligo were studied to clarify its characteristics and pathogenesis.
Of 323 patients with vitiligo seen at the vitiligo clinic in Severance Hospital of Yonsei Medical Center from January through October 1996, evaluation of 21 patients with trichrome vitiligo was performed. Responses to questionnaires regarding type, activity, sex, age at onset, age at initial visit, involved sites, family history, and associated disease were recorded. Four-millimeter punch biopsy specimens of the skin were obtained from 21 patients with trichrome vitiligo. Each patient underwent 4 biopsies: vitiliginous skin, light brown skin, perilesional normal skin, and normal skin as far as 5 cm from the nearest vitiligo spot. The sections were stained with hematoxylin-eosin to study the inflammatory cell infiltration in the epidermis and dermis, vacuolar degeneration of basal cell layer, and number of melanophages. The number of melanophages was counted in two separate sections of each 4-mm punch biopsy specimen, and the mean value was obtained. Immunohistochemical staining with S-100 protein and CD1a was performed in 10 patients showing inflammatory cell infiltration in hematoxylin-eosin staining. Positive cells with S-100 and CD1a were evaluated on 2 sections for each antibody with 3 fields per each section (×400). To count the accurate number of melanocytes in the basal cell layer, we stained the paired slide with CD1a and compared
Hann et al 591
J AM ACAD DERMATOL VOLUME 42, NUMBER 4
Table III. Histologic findings of trichrome vitiligo Inflammatory cell infiltration
Vitiliginous skin Light brown skin Perilesional normal skin Normal skin
Epidermis
Dermis
Vacuolar degeneration of basal cells
1 (4.8)* 13 (61.9) 16 (76.2)
6 (28.6) 12 (57.1) 16 (76.2)
5 (23.8) 13 (61.9) 19 (90.5)
2 (9.5)
3 (14.3)
7 (33.3)
*Number (percentage in parentheses) of positive cases.
the two paired slides for exclusion of CD1a+ Langerhans cells of the basal cell layer. The statistical significance of the results was calculated by the paired t test. Patients with trichrome vitiligo were treated with systemic PUVA, topical PUVA, systemic steroid therapy, or topical steroid therapy for at least 2 months depending on each patient’s situation.
RESULTS Clinical characteristics of trichrome vitiligo Table I summarizes the clinical characteristics of trichrome vitiligo. Of the 21 patients with trichrome vitiligo, 11 were male and 10 were female; their ages ranged from 9 to 62 years. The mean age at disease onset was 33.4 ± 14.8 years, and the mean duration was 7.9 ± 6.8 years. There were 20 cases of generalized type and 1 localized type disease. All the 20 patients with generalized type disease showed features of trichrome vitiligo restricted to a localized area, whereas the remaining areas displayed white spots with well-demarcated borders of typical vitiligo. At the time of the visit, 18 cases were active vitiligo (85.7%) and 3 cases were stable vitiligo (14.3%). The most commonly involved site was the back (57%) followed by abdomen (19%), buttock (9.5%), chest (4.8%), arm (4.8%), and leg (4.8%) (Table II). Histopathologic findings Light microscopic findings of the 4 discrete sites of biopsies revealed inflammatory cell infiltration in the epidermis occurring in 4.8% of vitiliginous skin, 61.9% of light brown skin, 76.2% of perilesional normal skin, and 9.5% of normal skin. Inflammatory cell infiltration in the dermis was 28.6% in vitiliginous skin, 57.1% in light brown skin, 76.2% in perilesional normal skin, and 14.3% in normal. Inflammatory cell infiltration was more prominent in light brown skin and perilesional normal skin than in vitiliginous skin and normal skin (Fig 2). Vacuolar degeneration of the basal cell layer was more prominent in light
Table IV. Number of melanophages in skin from patients with trichrome vitiligo Patient No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Mean ± SD†
VS
LBS
PLNS
NS
1.5* 4 1 0 0.5 0 1 0 2 0 0 2 0 0.5 0 3 7 1 0 0 3
5 16.5 14 3.5 6 18 10 25 23 9 7 6.5 0 7 8 4 13 5.5 15 2 17
3 13 13 8.5 8 17 16 25 5 19 4 3.5 0 8 24 12 3 0.5 5 5 15
0 0.5 0.5 0 0.5 0 0 8 3 2 2 0 0 0 3.5 0 10 2 7 1.5 0
9.9 ± 8.0
2.8 ± 4.0
1.3 ± 1.8 10.2 ± 6.8
LBS, Light brown skin; NS, normal skin; PLNS, perilesional normal skin; VS, vitiliginous skin. *Number of melanophages in two sections of 4-mm punch biopsy specimen. †VS and LBS: P < .05; PLNS and NS: P < .05.
brown skin (61.9%) and perilesional normal skin (90.5%) than in vitiliginous skin (23.8%) and normal skin (33.3%) (Table III). The mean number of melanophages in the 4-mm punch biopsy specimens was 1.3 ± 1.8 in vitiliginous skin, 10.2 ± 6.8 in light brown skin, 9.9 ± 8.0 in perilesional normal skin, and 2.8 ± 4.0 in normal skin. There was a significant increase in melanophages in light brown skin and perilesional normal skin compared with vitiliginous and normal skin (P < .05) (Table IV). Immunohistochemical findings The number of S-100+ cells in the basal cell layer of the epidermis was counted per 6 high-power fields (×400) and is summarized in Table V. The mean number of melanocytes was 4.8 ± 2.2 in vitiliginous skin, 11.9 ± 3.3 in light brown skin, 16.8 ± 3.0 in perilesional normal skin, and 16.3 ± 3.5 in normal skin. There was no statistically significant difference between perilesional normal skin and normal skin (P > .05). The number of melanocytes was decreased in light brown skin compared with perilesional nor-
592 Hann et al
J AM ACAD DERMATOL APRIL 2000
A
B
C
D Fig 2. Hematoxylin-eosin staining of trichrome vitiligo. Vacuolar degeneration of the basal cell layer and mild inflammatory cell infiltration in epidermis and dermis is more prominent in light brown skin (B) and perilesional normal skin (C) than vitiliginous skin (A) and normal skin (D) of trichrome vitiligo. (Original magnification ×200.)
mal skin (P < .05) and in vitiliginous skin compared with light brown skin (P < .05). A few melanocytes were observed even in vitiliginous skin of trichrome vitiligo (Fig 3). The number of CD1a+ Langerhans cells in the epidermis was counted per 6 high-power fields (×400); these data are summarized in Table VI and representative findings are illustrated in Fig 4. The mean number of CD1a+ Langerhans cells was 29.5 ± 9.8 in vitiliginous skin, 55.7 ± 11.1 in light brown skin, 54.3 ± 10.4 in perilesional normal skin, and 36.7 ± 8.6 in normal skin. There was no statistically significant difference between light brown skin and perilesional normal skin (P > .05). The number of Langerhans cells was increased in light brown skin and perilesional normal skin compared with vitiliginous and normal skin (P < .05). Treatment and response In trichrome vitiligo, treatment results were excellent with systemic PUVA therapy. The 21 patients
enrolled in this study received systemic PUVA, topical PUVA, systemic steroid therapy, or topical steroid therapy for at least 2 months based on their clinical manifestations, biopsy findings, and patient compliance. Among the 21 patients, 6 patients dropped out of the study and the remaining 15 patients were followed up for periodic observation of treatment results. Three patients (20.0%) experienced complete resolution of the lesions or marked improvement (excellent), 2 patients (13.3%) showed improvement of a moderate degree (good), 2 patients (13.3%) had some improvement (fair), 2 patients (13.3%) had slight or no improvement (poor), and 6 patients (40.0%) showed aggravation of the lesions (very poor). All 3 patients who showed excellent treatment results had been treated with systemic PUVA therapy. Eight patients were treated with systemic PUVA therapy and 3 (37.5%) of them showed an excellent response. Among the 5 patients treated with systemic steroid therapy, 4 patients (80%) responded very poorly.
Hann et al 593
J AM ACAD DERMATOL VOLUME 42, NUMBER 4
Table V. Numbers of S-100+ melanocytes in patients with trichrome vitiligo
Table VI. Number of CD1a+ Langerhans cells in patients with trichrome vitiligo
Patient No.
VS
LBS
PLNS
NS
Patient No.
1 2 3 4 5 6 7 8 9 10
— 5 5 2 4 5 8 8 4 2
— 14 15 14 8 15 13 8 13 7
23* 19 18 13 14 16 15 17 16 11
23 — 20 15 16 17 16 12 16 14
1 2 3 4 5 6 7 8 9
VS
LBS
PLNS
NS
— — 39 36 21 33 34 — 14
62* 52 67 28 50 65 — 52 34
73 58 54 44 50 47 60 44 36
36 47 37 — 45 24 31 33 28
Mean ± SD† 29.5 ± 9.8 55.7 ± 11.1 54.3 ± 10.4 36.7 ± 8.6
Mean ± SD† 4.8 ± 2.2 11.9 ± 3.3 16.8 ± 3.0 16.3 ± 3.5 For abbreviations see legend to Table IV. *Number of melanocytes per 6 high-power fields (×400). †PLNS and LBS: P < .05; LBS and VS: P < .05.
DISCUSSION In this study, we investigated whether the tan lesions of trichrome vitiligo and its surrounding perilesional normal skin show features of active vitiligo, both clinically and histologically. Among the 21 vitiligo patients showing trichrome lesions, 95.2% were classified as having vitiligo vulgaris and 85.7% had spreading lesions clinically. Histopathologic findings also showed the characteristics of active spreading vitiligo. Therefore trichrome vitiligo was regarded as a phenomenon occurring in restricted areas of active vitiligo. Of the trichrome lesions, 85.5% were localized to the trunk region, including the abdomen, back, and buttock, leading to the assumption that trichrome vitiligo predominates in unexposed skin. According to previous reports, sun-exposed areas are the predilection sites of vitiligo and lesions in these areas commonly show rapid progression. In contrast, the fact that trichrome vitiligo lesions of our patients predominated in unexposed skin could be one of the reasons the characteristic trichrome features appeared, possibly because of slow progression of the disease. Melanocyte density and skin thickness could also be factors contributing to the development of trichrome features. For example, the trunk shows the lowest melanocyte density of the skin with about 1100 to 1250/mm2 and is the thickest portion of the body.6 Therefore it is feasible to hypothesize that such factors eventually result in slow progression of the disease process. However, truncal lesions of patients with vitiligo are not always trichrome, and because trichrome vitiligo is known to occur on other parts of the body, factors other
For abbreviations see legend to Table IV. *Number of Langerhans cells per 6 high-power fields (×400). †LBS and PLNS compared with VS and NS: P < .05.
than sunlight, melanocyte density, and skin thickness seem to be associated with its pathogenesis. Blacks have a relatively higher frequency of trichrome vitiligo compared with whites and similarly most of the patients in our study had skin type IV or darker. Therefore dark skin also seems to be a contributing factor to the pathogenesis of trichrome vitiligo. The histologic findings of trichrome vitiligo showed the most dense distribution of melanin granules in the perilesional normal skin, followed by normal skin, light brown skin, and vitiliginous skin, in descending order. As such, the characteristic trichrome color may be an expression of changes in melanin granules rather than melanocyte numbers. Hyperpigmentation seen around the periphery of white patches is typically found in vitiligo. This was also observed in the trichrome lesions in which perilesional normal skin showed a slightly darker color compared with normal skin and histologically a higher density of melanin granules. Other histologic findings such as vacuolar degeneration of the basal cell layer, mononuclear cell infiltration of the epidermis and dermis, and melanophage deposition in the dermis were more prominent in light brown skin and perilesional normal skin than in vitiliginous and normal skin. Among these changes, vacuolar degeneration of the basal cell layer and inflammatory cell infiltration were especially accentuated around the melanocytes in the basal cell layer. However, overall destruction of keratinocytes coexisted, and we presume that the target of destruction is not limited to melanocytes but also involves keratinocytes as well. Perilesional normal skin shows more severe vacuolar
594 Hann et al
J AM ACAD DERMATOL APRIL 2000
A
B
C
D Fig 3. S-100 immunohistochemical staining of trichrome vitiligo. Number of melanocytes is decreased in light brown skin (B) compared with perilesional normal skin (C) and normal skin (D) and in vitiliginous skin (A) compared with light brown skin (B). A few melanocytes are observed even in vitiliginous skin of trichrome vitiligo (A). (Original magnification ×200.)
change of the basal layer and inflammatory cell infiltration than light brown skin and histologically is already in the process of vitiligo evolution, despite its normal-looking appearance. Therefore, without appropriate treatment, a change clinically into light brown skin can be predicted. Light brown skin clinically and histologically shows features of active vitiligo, although the degree of histologic change is subtle compared with perilesional normal skin. The histologic features of light brown skin and perilesional normal skin are congruent with findings of active vitiligo lesions of more than 1 year’s duration showing vacuolar degeneration of basal cell layer, mononuclear cell infiltration of the epidermis and dermis, and melanophage deposition.7,8 In view of such similarities, trichrome vitiligo could be a variant form of active vitiligo. The number of melanocytes in trichrome vitiligo was greatest in perilesional normal skin followed by light brown skin and vitiliginous skin, with vitiliginous skin showing at least a few melanocytes, albeit
less than that of normal skin. Our results are contradictory with other studies of vitiligo,7,9,10 which report absence of melanocytes in the depigmented patches confirmed by immunohistochemical staining or electron microscopy. The clinical and histologic findings of trichrome vitiligo suggest a slower progression of lesions than typical vitiligo, and this could be why melanocytes remain in the white patches. However, studies of the functional capacity of the remaining melanocytes should be conducted to elucidate this matter. Langerhans cells may play a major immunologic role in vitiligo. Interaction between keratinocytes, melanocytes, and Langerhans cells is thought to initiate depigmentation, but the exact mechanism is unknown. In patients with nonsegmental type vitiligo, a marked depletion of Langerhans cells was noted in active lesions and a repopulation of Langerhans cells was noted in stable lesions.11 In inflammatory vitiligo, an increase in Langerhans cells was observed in adjacent normal skin compared
Hann et al 595
J AM ACAD DERMATOL VOLUME 42, NUMBER 4
A
B
C
D Fig 4. CD1a immunohistochemical staining of trichrome vitiligo. Number of Langerhans cells is increased in light brown skin (B) and perilesional normal skin (C) compared with vitiliginous skin (A) and normal skin (D). (Original magnification ×100.)
with vitiliginous skin and normal skin.12 We observed changes in Langerhans cells in 9 patients with trichrome vitiligo and found that light brown skin and perilesional normal skin exhibit an increase in Langerhans cell number compared with vitiliginous skin and normal skin. From our findings, an increased number of Langerhans cells may be involved in actively spreading vitiligo. The pathogenesis of vitiligo is difficult to explain with a single theory; among the various theories set forth, autoimmunity is currently the most supported. However, recently not only humoral immunity but also cell-mediated immunity have been recognized as playing an important role and abnormalities of T lymphocytes in vitiligo have been studied.13,14 Our results, such as inflammation of the epidermis and dermis in light brown skin and perilesional normal skin as well as changes in melanocytes and Langerhans cells, all suggest that cell-mediated immunity may be involved in the development of trichrome vitiligo. Vitiligo lesions of the trunk are known to respond favorably to systemic PUVA therapy in comparison to
systemic steroid therapy, and the existence of inactive melanocytes in the epidermis or follicles is a decisive factor influencing treatment results.15,16 Trichrome vitiligo responded especially well to systemic PUVA treatment and we think this is because a few melanocytes were still remaining in the white lesions, thereby contributing to the repigmentation process. Therefore early systemic PUVA therapy should be considered in patients with trichrome features to shorten treatment duration and achieve satisfactory end results. REFERENCES 1. Fitzpatrick TB. Hypomelanosis. South Med J 1964;57:995-1005. 2. Fargnoli MC, Bolognia JL. Pentachrome vitiligo. J Am Acad Dermatol 1995;33:853-6. 3. Mosher DB, Fitzpatrick TB, Ortonne JP, Hori Y. Hypomelanoses and hypermelanoses. In: Freedberg IM, Eisen AZ,Wolff K, Austen KF, Goldsmith LA, Katz SI, et al, editors. Dermatology in general medicine. 5th ed. New York: McGraw-Hill; 1999. p. 945-1017. 4. Ortonne JP, Mosher DB, Fitzpatrick TB, editors.Vitiligo and other hypomelanoses of hair and skin. New York: Plenum Publishing; 1983. p. 147-8. 5. Pincus H. Vitiligo: what is it? J Invest Dermatol 1959;32:281-4.
596 Hann et al
6. Rosdahl I, Rorsman H. An estimate of the melanocyte mass in human. J Invest Dermatol 1983;81:278-81. 7. Hann SK, Park YK, Lee KG, Choi EH, Im S. Epidermal changes in active vitiligo. J Dermatol 1992;9:217-22. 8. Gokhale BB, Mehta LN. Histopathology of vitiliginous skin. Int J Dermatol 1983;22:477-80. 9. Le Poole IC, Das PK, van den Wijngaard RMJGJ, Bose JD, Westerhof W. Review of the etiopathomechanism of vitiligo: a convergence theory. Exp Dermatol 1993;2:145-53. 10. Le Poole IC, van den Wijngaard RMJGF, Westerhof W, Dutrieux RP, Das PK. Presence or absence of melanocytes in vitiligo lesions: an immunohistochemical investigation. J Invest Dermatol 1993;100:816-22. 11. Kao CH, Yu HS. Depletion and repopulation of Langerhans cells in nonsegmental type vitiligo. J Dermatol 1990;17:280-96.
J AM ACAD DERMATOL APRIL 2000
12. Le Poole IC, van den Wijngaard RMJGJ, Westerhof W, Das PK. Presence of T cells and macrophages in inflammatory vitiligo skin parallels melanocyte disappearance. Am J Pathol 1996; 148:1219-28. 13. Grimes PE, Ghoneum M, Stockton T, Payne C, Kelly AP, Alfred L. T cell profiles in vitiligo. J Am Acad Dermatol 1986;14:196-201. 14. Hann SK, Park YK, Chung KY, Kim HI, Im S, Won JH. Peripheral blood lymphocyte imbalance in Koreans with active vitiligo. Int J Dermatol 1993;32:286-9. 15. Ortonne JP, Schmitt D, Thivolet J. PUVA-induced repigmentation of vitiligo: scanning electron microscopy of hair follicles. J Invest Dermatol 1980;74:40-2. 16. Cui J, Shen L,Wang G. Role of hair follicles in the repigmentation of vitiligo. J Invest Dermatol 1991;97:410-6.