British Journal of Dermatology 2002; 146: 228237.

Dermatopathology
Melasma: histopathological characteristics in 56 Korean patients
W.H.KANG, K.H.YOON, E-S.LEE, J.KIM, K.B.LEE,* H.YIM,* S.SOHN AND S.IM
Departments of Dermatology and *Pathology, and Laboratory of Cell Biology, Institute for Medical Sciences, Ajou University School
of Medicine, 5 Wonchon-dong, Paldal-ku, Suwon 442-721, Korea

Accepted for publication 30 August 2001

Summary Background Melasma is a common acquired symmetrical hypermelanosis characterized by

irregular light to dark brown macules and patches on sun-exposed areas of the skin. Its
histopathological characteristics are not fully understood.
Objectives To characterize the histopathological features of facial melasma skin in comparison
with adjacent normal skin.
Methods Biopsies were taken from both melasma lesional skin and adjacent perilesional normal
skin in 56 Korean women with melasma. The sections were stained using haematoxylin and eosin,
FontanaMasson, diastase-resistant periodic acid-Schiff, Masson trichrome and Verhoeffvan Gie-
son stains, and immunostaining for melanocytes. Data on the changes in number of melanocytes
and melanin contents of the epidermis were analysed by a computer-assisted image analysis pro-
gram. The ultrastructure of the skin was also examined.
Results The amount of melanin was signicantly increased in all epidermal layers in melasma
skin. The staining intensity and number of epidermal melanocytes increased in melasma lesions.
Lesional skin showed more prominent solar elastosis compared with normal skin. Melanosomes
increased in number and were more widely dispersed in the keratinocytes of the lesional skin.
Lesional melanocytes had many more mitochondria, Golgi apparatus, rough endoplasmic reticulum
and ribosomes in their cytoplasm. A dihydroxyphenylalanine reaction was apparent in the cis-
ternae and vesicles of the trans-Golgi network in melanocytes from lesional skin.
Conclusions Melasma is characterized by epidermal hyperpigmentation, possibly caused both by an
increased number of melanocytes and by an increased activity of melanogenic enzymes overlying
dermal changes caused by solar radiation.
Key words: adjacent control skin, histopathological differences, melasma

Melasma is a common acquired symmetrical hypermel-
anosis characterized by irregular light to dark brown
macules and patches on sun-exposed areas of the skin.
Although no sex, race or age is exempt from developing
melasma, it is more common in Oriental or Hispanic
women living in sunny locations.1 Its aetiological
factors are genetic background, exposure to ultravio-
let (UV) radiation, pregnancy, hormonal therapies,
cosmetics, phototoxic drugs and antiseizure medica-
tions.25 Its pathogenesis is not yet clearly dened.
A knowledge of the morphological and physiological
changes in melasma skin is essential to understanding
Correspondence: Sungbin Im.
E-mail: sungbio@yahoo.com
the pathogenesis of melasma. However, only a few
reports are available on histopathological changes in
melasma.2 To characterize these, we investigated the
histopathological differences between melasma skin
and adjacent control skin.
Patients and methods
Patients
We examined 56 Korean women with melasma who
attended the Department of Dermatology, Ajou
University Hospital (Suwon, Korea), between October
1998 and September 1999. Their mean age was
37 years and the mean duration of melasma was

228 2002 British Association of Dermatologists

 HISTOPATHOLOGY OF MELASMA IN KOREAN PATIENTS
7 years. In each patient, the diagnosis of melasma
was determined by physical examination. Two-milli-
metre punch biopsies from lesional and adjacent
(usually within 1 cm) non-lesional facial skin were
obtained from each patient under local anaesthesia.
Acquired bilateral naevus of Ota-like macules
(ABNOM) characterized by dermal melanocytosis were
excluded in this study based on positive dermal
NKI-beteb immunostaining. Tissues were prepared for
light microscopic study by 10% formalin xation. For
ultrastructural evaluation, a further 10 pairs of
lesional and normal skin specimens from Korean
women aged 2852 years were used. The specimens
were xed in a Karnovsky's cacodylate-buffered glu-
taraldehyde-paraformaldehyde mixture.6
Stains
Parafn-embedded tissue sections of 3 lm thickness
were processed for light microscopic examination. A
haematoxylin and eosin (H&E) stain was used for
studying the general histopathological changes in the
melasma skin. Melanin pigment was visualized with
the FontanaMasson stain, performed by the usual
methods without an eosin background stain for image
analysis. The sections were also stained using diastase-
resistant periodic acid-Schiff (D-PAS) for basement
membranes, Masson trichrome for collagen and Ver-
hoeffvan Gieson for elastic bres.
Analysis of staining results
Sections were evaluated independently by four different
observers and scored as: 0, none (normal); 1, minimal
(equivocal); 2, mild; 3, moderate; and 4, severe. Mean
scores were analysed with the Statistical Package for
Social Science Version 80 (SPSS, Chicago, IL, U.S.A.).
The data of individual groups were evaluated with the
paired t-test; P < 005 indicated a statistically signi-
cant difference.
Table 1. Antibodies used, and their work-
3 Antibody
ing dilutions
NKI-beteb
(melanocytes)
Mel-5
(melanocytes)
CD1a (Langerhans
cells)
S100
(melanocytes)
RT, room temperature.
2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237
229
Immunohistochemistry
Immunohistochemistry was performed as described
elsewhere,7 and is briey summarized below. Four-
micrometre parafn-embedded sections of both lesional
and control skin were mounted on a Polysine micro-
1 scope slide (Menzel-Glaser, Germany) coated with 01%
poly D-lysine. Tissues were deparafnized and rehy-
drated by sequential immersion in xylene, graded
concentrations of ethanol, and distilled water. They
were incubated for 30 min at room temperature in a
solution of 05% hydrogen peroxidase in methanol to
quench endogenous peroxidase activity, followed by
washing three times in Tris-buffered saline (TBS,
01 mol L)1, pH 74, Dako, Carpinteria, CA, U.S.A.).
They were subsequently incubated in 005% pepsin in
TBS for 20 min at 37 C. After washing three times in
TBS, they were ooded with a protein-blocking agent
(PBA; Immunon, Pittsburgh, PA, U.S.A.) for 10 min at
room temperature. Excess PBA was drained and the
primary antibodies NKI-beteb, Mel-5, CD1a and S100
were applied to tissue sections (Table 1). The slides
were then incubated for 1 h at room temperature and
for 30 min at 37 C in a humid chamber. Following
three washes in TBS, sections were incubated for
30 min at room temperature while being ooded with
a biotinylated universal secondary antibody reagent
(Immunon). The slides were then washed in TBS,
followed by incubation in streptavidin alkaline phos-
phatase reagent (Immunon) for 30 min. After washes
in TBS, sections were incubated in fast red chromogen
(Immunon) for 10 min. The sections were counter-
stained with haematoxylin modied solution (Merck,
Darmstadt, Germany) and mounted in an aqueous
mounting medium (Biomeda, Foster City, CA, U.S.A.).
Image analysis
A CCD camera (CCD-IRIS, Sony, Tokyo, Japan)
mounted on a microscope (Olympus BX50F, Olympus
Dilution Incubation Temperature Manufacturer
1 : 20 2h RT Monosan
1 : 40 45 min RT Signet
Prediluted 2h RT Immunotech
1 : 20 60 min RT Castra
230 W . H . K A N G et al.
Optical Co., Tokyo, Japan) was connected to an IBM
personal computer (PC). The image signals taken by
the PC were evaluated using Image Pro Plus Version
30 (Media Cybertics Co., Silver Spring, MD, U.S.A.).
The image analysis was performed on a representative
area of each specimen. In FontanaMasson staining,
the amount of melanin pigment was measured under
constant magnication ( 200). The pigmented area
per millimetre epidermal length (PA 1E), the pigmented
area per millimetre rete ridge length (PA 1R) and the
ratio of pigmented area to measured epidermal area
(PA MA) were measured in lesional and control skin.
In the immunohistochemical stain for melanocytes
(NKI-beteb), three serial sections were made in each
specimen, in order to count basal keratinocytes and
melanocytes. The number of melanocytes was
estimated using three methods: the number of mela-
nocytes per millimetre epidermal length (MC 1E), the
number of melanocytes per millimetre rete ridge length
(MC 1R) and the ratio of melanocytes to basal kera-
tinocytes (MC KC). Each measurement was evaluated
under constant magnication. Careful examination
was performed and each melanocyte was counted as
one cell when its nucleus was conrmed. All morpho-
metric procedures were performed by manually tracing
the borders of the epidermis and rete ridges, and the
epidermal area that contained hair follicles was
excluded from this tracing. For each frame, the tracing
was repeated three times and the mean was used for
evaluation; all morphometric measurements were per-
formed by the same person.
Electron microscopic evaluation
For electron microscopic evaluation, 10 pairs of
lesional and normal skin specimens were used. In
seven cases, the tissues were washed with a cacodylate
buffer and were incubated in a 01% dihydroxyphenyl-
alanine (DOPA) solution for 4 h at 37 C (in the other
three cases DOPA staining was omitted), postxed in
1% osmium tetroxide and then dehydrated in an
ethanol gradient and propylene oxide. Tissues were
embedded in Epon 812, and 1-lm semithin sections
were cut and stained with toluidine blue to select
appropriate areas for observation. Serial ultrathin 50-
nm sections were made using a Reichert-Ultracut S
ultramicrotome and placed on 100 mesh grids. The
sections, counterstained with 4% uranyl acetate and
05% lead citrate, were observed and photographed in a
Zeiss EM 902A transmission electron microscope at an
accelerated voltage of 50 kV.
2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237
Results
Distribution
Two patterns of melasma were recognized on the basis
of clinical examination: centrofacial and malar (Fig. 1).
The centrofacial pattern was observed in 52% of cases:
the melasma involved the cheeks, forehead, upper lip
and chin. The malar pattern refers to those patients in
whom the melasma was localized to the malar region,
and was seen in 48% of cases.
General histopathological features of melasma
The general histopathological features of melasma
were compared in H&E-stained sections with those of
perilesional normal facial skin. All specimens of melas-
ma and normal skin showed a mild to moderate degree
of rete ridge attening and epidermal thinning. Solar
elastosis was signicantly increased in melasma skin.
In 52 of 56 cases (93%), melasma skin showed
moderate to severe solar elastosis, while in 39 of 56
cases (70%), the normal skin showed mild to moderate
solar elastosis (Table 2, Fig. 2A,B).
In FontanaMasson-stained sections, the amount of
melanin was signicantly increased in all epidermal
layers in melasma skin. Melasma skin had more free
melanin and melanophages in the dermis. However,
three of 56 (5%) cases had dermal melanophages
conned only to melasma skin. Twenty of 56 cases
(36%) had dermal melanin in both melasma and
normal skin and only four of these 20 cases (7%)
showed more dermal melanophages in melasma skin
than in normal skin. The remaining 33 of 56 cases
(59%) showed no dermal melanin either in melasma or
in normal skin. Overall, only seven cases (12%) showed
increased dermal melanin and melanophages in mel-
asma skin (Table 3, Fig. 2C,D).
CD1a immunostaining showed no changes in the
number of epidermal Langerhans cells in either lesional
or normal skin. No destruction of basement membranes
was noted in D-PAS staining. Masson trichrome
staining for collagen and S100 immunostaining for
dermal nerves did not show any signicant changes
(not shown). In contrast, Verhoeffvan Gieson staining
showed thick, highly curled, and more fragmented
elastic bres in melasma skin than in normal skin
taken from the perilesional, retroauricular or breast
area (Fig. 3).
NKI-beteb detection of epidermal melanocytes
showed the intensity of staining in melasma lesions
 HISTOPATHOLOGY OF MELASMA IN KOREAN PATIENTS 231

Figure 1. Patients with melasma. (A) centrofacial pattern: melasma involving the cheeks, forehead, upper lip and chin; (B) malar pattern:
melasma localized to the malar region.

to be much stronger than in control skin. The melasma
melanocytes exhibited more dendrites (Fig. 4A,B). Mel-
5 immunostaining showed a similar staining pattern
and number of epidermal melanocytes as did NKI-beteb
immunostaining (not shown).
Quantitative analysis of melanin content and number
of melanocytes in melasma
Image analysis showed increases in PA 1E and PA 1R
of 73% and 39%, respectively, in melasma lesional skin
as compared with normal skin. A signicant increase
(83%) in PA MA was observed in melasma skin. In
these preliminary observations, three serial sections
from the same specimen gave a good correlation with
image analysis of the number of melanocytes. Signif-
icant increases were observed in MC 1E (24%) and
MC 1R (27%) in melasma skin as compared with
normal skin. Although four of 56 cases (7%) showed
no increase in melanocytes, and ve cases (9%) showed
a slightly decreased number of melanocytes in melasma
skin, 47 cases (84%) showed an increased number
of melanocytes in melasma skin. Melasma skin
also showed a signicant increase (33%) in MC KC
(Table 4).
Ultrastructural changes of melasma
In the stratum corneum, melanosomes were more
abundant in lesional skin than in control skin. Mela-
nosomes transferred into keratinocytes were distributed
mostly in a single array, and occasionally membrane-
bound complexes were observed in both lesional and
perilesional normal skin. Lesional melanocytes were
lled with more mitochondria, Golgi apparatus,
rough endoplasmic reticulum and ribosomes in the

2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237

232 W . H . K A N G et al.

Table 2. Comparative histological features

Perilesional Melasma
(mean SD) of biopsy specimens from
normal skin skin
Korean patients with melasma
Epidermis (haematoxylin and eosin; n 56)
Hyperkeratosis 0 0
Parakeratosis 0 0
Hypergranulosis 0 0
Acanthosis 0 0
Spongiosis 0 0
Exocytosis 0 0
Basal cell damage 0 0
Rete ridge attening 239 011 229 107
Basal 123 047* 241 065*
hyperpigmentation
Dermis
Supercial vessels 0 0
Telangiectasia 0 0
Inammatory cell 152 079 157 085
inltration
Solar elastosis 273 094* 346 074*
Collagen degeneration 205 120 223 128
Dermal melanophages 049 039 063 126
Dermal melanocytes 0 0

0, none (normal); 1, minimal (equivocal); 2, mild; 3, moderate; 4, severe. *P < 005.

Figure 2. Photomicrographs of biopsy specimens from a patient with melasma. (A,C) Perilesional skin; (B,D) melasma skin. Note that the melanin
pigment (C,D) is located in a cap overlying the keratinocyte nuclei (A,B, haematoxylin and eosin; C,D, FontanaMasson; original magnication
200).

2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237

 HISTOPATHOLOGY OF MELASMA IN KOREAN PATIENTS 233

Table 3. Comparative histological features

Perilesional Melasma skin
(mean SD) of biopsy specimens from
normal skin (% increase)
Korean patients with melasma (Fontana
Masson stain; n 56) Melanin amount
Stratum corneum 196 074* 284 068* (45%)
Stratum granulosum 138 049* 218 058* (58%)
Stratum spinosum 145 057* 252 071* (74%)
Stratum basalis 241 056* 359 053* (49%)
Dermal melanin
Free melanin 064 062* 082 103*
Melanophages 048 054* 113 145*
Dermal melanocytes 0 0

0, none (normal); 1, minimal (equivocal); 2, mild; 3, moderate; 4, severe. *P < 005.

perikaryon than were melanocytes in normal skin,
suggesting increased cell activity. In lesional melano-
cytes, stage III or IV melanosomes were dispersed singly
throughout the cytoplasm, and the number of mela-
nosomes was increased in comparison with normal
melanocytes (Fig. 5A,B). Vacuolar degeneration was
observed both in melasma and normal melanocytes. A
portion of the dendritic process of the melanocyte in
melasma lesions contained more numerous stage IV
melanosomes. More melanosomes were found in mem-
brane-bound clusters and were more densely packed in
the basal and suprabasal keratinocytes in melasma skin
than in normal facial skin (Fig. 5C,D). There were no
differences in the size of melanosomes between lesional

Figure 3. Changes in elastic bres in melasma skin. (A) Perilesional normal skin; (B) pigmented lesional skin; (C) retroauricular normal skin;
(D) breast control skin. Black, elastic bres; red, collagen. Melasma skin (B) showed thick, highly curled, and more fragmented elastic bres than in
normal skin taken from the perilesional, retroauricular or breast area (Verhoeffvan Gieson; original magnication 200).

2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237

234 W . H . K A N G et al.
Figure 4. Immunostaining for melanocytes in melasma skin. (A)
Perilesional skin; (B) melasma skin. Epidermal melanocytes detected
by NKI-beteb showed a stronger intensity of staining in melasma
lesions than in control skin (NKI-beteb; original magnication 200).
and normal skin. The dermoepidermal junction of
lesional skin was covered by an intact, continuous
sheet of basal lamina. Dermal melanophages were
variably distributed in both lesional and normal skin.
After DOPA histochemistry of melasma melanocytes,
an extensive amount of reaction product was apparent
in the cisternae and vesicles of the trans-Golgi network
(TGN), in contrast to control melanocytes. There were
more numerous DOPA-positive trafcking vesicles,
approximately 50 nm in diameter, scattered in areas
near the TGN in melasma melanocytes (not shown).
Discussion
Melasma is dened as a light to dark brown, irregular
hypermelanosis of the face. It develops slowly and is
usually symmetrical.1 Although it may occur in men,8
it is much more common in women, particularly
Asians and those of Hispanic origin. There have been
2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237
Table 4. Quantitative analysis (mean SD) of melanin content and
number of melanocytes in melasma (n 56)
Perilesional Melasma skin
normal skin (% increase)
Melanin content
PA 1Ea 830 549* 1437 603* (73%)
PA 1Rb 953 144* 1328 516* (39%)
PA MAc (%) 1221 740* 2239 947* (83%)
Number of melanocytes
MC 1Ed 1164 390* 1438 394* (24%)
MC 1Re 1117 388* 1424 394* (27%)
MC KCf 012 004* 016 005* (33%)
a
PA 1E, (pigmented area mm1 epidermal length) 1000; bPA 1R,
(pigmented area mm1 rete ridge length) 1000; cPA MA, (ratio of
pigmented area to measured epidermal area) 100; dMC 1E, num-
ber of melanocytes mm1 epidermal length; eMC 1R, number of
melanocytes mm1 rete ridge length; fMC KC, ratio of melanocytes to
basal keratinocytes. *P < 005.
few studies of the histopathological characteristics of
melasma, and none comparing histopathological nd-
ings between melasma skin and adjacent normal facial
control skin. The present investigation of the histopa-
thology of melasma, in comparison with normal skin,
is a prerequisite for the understanding of its patho-
genesis.
In our study, a centrofacial pattern was observed in
52% of cases, and a malar pattern in 48% of cases. The
observed predilection sites of melasma in our patients
were areas receiving greatest UV exposure, e.g. fore-
head, upper lip, infra- and supraorbital margins, and
zygomatic process. We did not nd any histopatholog-
ical differences between the two clinical types.
It is well known that sun-exposed skin of the face
contains signicantly more epidermal melanocytes
than the rest of the body.9 For this reason, skin from
the trunk or the extremities is not an adequate control.
Melasma skin is characterized by a different colour
from the adjacent facial skin under similar environ-
mental stresses, such as UV exposure. To understand
the mechanism of colour changes in melasma lesional
skin, we took control biopsies from normal facial skin
just adjacent to the lesion.
The epidermis of the facial skin was attened in both
lesional and perilesional areas. This attened epidermis
suggested changes of chronic solar damage.10,11 These
ndings suggest that keratinocyte proliferation is not
the major event in the pathogenesis of melasma, and
therefore we suspect the causative factor to be some
stimulus causing melanocyte hyperfunction without
inducing keratinocyte proliferation. The increased
solar elastosis in melasma skin may suggest that
 HISTOPATHOLOGY OF MELASMA IN KOREAN PATIENTS 235

Figure 5. Electron microscopy of perilesional skin and melasma skin. Melanocytes in (A) perilesional skin and (B) melasma skin. Lesional
melanocytes were lled with more mitochondria, Golgi apparatus, rough endoplasmic reticulum and ribosomes in the perikaryon than were
normal skin melanocytes. Basal keratinocytes in (C) perilesional skin and (D) melasma skin. More melanosomes were seen in membrane-bound
clusters and were more densely packed in the basal keratinocytes in melasma skin than in normal facial skin (original magnication: A,B, 7000;
C,D, 4400).

accumulated sun exposure is necessary for the devel- of hyperpigmentation in the overlying epidermis, al-
opment of melasma and that the process of solar though the solar damage may be just a secondary
damage to the dermis may inuence the development epiphenomenon to UV radiation.

2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237

236 W . H . K A N G et al.
FontanaMasson staining gives more denite identi-
cation of melanin.12 Melasma skin had more melanin
in the whole epidermis, including the stratum corne-
um, whereas in normal skin the melanin pigment was
mostly conned to the basal layer. This suggests that
an increased synthesis of melanosomes in the melano-
cytes, and increased transfer and decreased degrada-
tion of melanosomes in the keratinocytes, are essential
for the development of melasma.1315
Melasma was formerly classied histopathologically
as epidermal, dermal and mixed type2,13 by the
locations of pigments. Melanophages were present both
in melasma and in normal facial skin in 36% of our
patients, and melanophages are known to be present in
the normal dermis of Japanese skin.16 Thus, melano-
phages cannot be a hallmark of the dermal type of
melasma. Based on our ndings, we suggest that there
is no true dermal type of melasma, as the previous
histopathological classication came from a study
lacking controls. One other possible explanation for
the presence of dermal melanophages is that ABNOM
have similar clinical characteristics to melasma: nine of
65 cases initially selected under the clinical impression
of melasma in our study were diagnosed as ABNOM on
FontanaMasson and NKI-beteb immunostaining.
Histopathological evaluation of epidermal melano-
cytes on routine H&E staining is difcult and cannot be
made with accuracy. It is still controversial whether
the number of melanocytes is increased in melas-
ma.2,1315 A few reports on the number of melanocytes
using DOPA immunochemistry showed a signicant
increase in the population density of melanocytes in
melasma as compared with normal control skin.
However, in one study most control samples came
from the arm, with only two samples from the face.2
Results of incubation with DOPA are inconsistent, and
the procedure is time consuming. The melanocyte
lineage-specic antigens recognized by the monoclonal
antibody NKI-beteb are among the best diagnostic
markers for human melanocytes available to date. The
antigens recognized by NKI-beteb are glycoproteins of
pmel-17, which are localized at the inner side of
premelanosomal vesicles.1720 Melasma skin showed
an increased staining intensity to NKI-beteb, suggest-
ing that the level of pmel-17 (which is related to
melanogenesis) was higher than in normal skin. These
increases may be related to increased melanogenesis in
melasma skin. Mel-5 specically detects a 75-kDa
glycoprotein, tyrosinase-related protein (TRP)-1.21
Mel-5 also gave increased staining intensity in melasma
skin, suggesting that the enzyme level of TRP-1 was
2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237
higher in melasma melanocytes. Increased expression
of TRP-1 indicates an increased eumelanin synthesis.
Quantication of the skin melanin content would
provide valuable information to assess a clinical grade,
and would be an index for the effects of melasma
treatment.2,5 PA 1E indicates the accumulated mela-
nin colour when the skin is observed from outside: it
increased in melasma skin. PA 1R was evaluated in
terms of whether changes in rete ridge length may
inuence the true melanin colour,22 and this also
increased in melasma skin. The amount of melanin per
unit area, expressed as PA MA,23,24 also increased in
melasma skin. These image analyses correlate well
with visual scoring and give more quantitative infor-
mation. MC 1E and MC 1R also increased in melasma
skin. The increased number of melanocytes in melasma
skin was conrmed by the increase in MC KC. The
increase in the number of melanocytes was due to
an increased density of melanocytes, and not to an
increased length of rete ridge as in seborrhoeic kera-
tosis and lentigines. Our observations clearly demon-
strate that there are more melanocytes and melanin in
the epidermis in melasma skin.
Ultrastructural observation of melasma skin also
revealed increased numbers of melanosomes and mel-
anocytes. There was no defect in the formation of
melanosomes because maturation status, morphology
and distribution of melanosomes were very similar in
melasma and normal skin. The most striking feature of
melasma melanocytes was an active process of protein
synthesis and DOPA-reactive tyrosinase formation.
This suggests that increased synthesis of tyrosinase
with active protein synthesis is responsible for the
formation of the melanosomal matrix and melanin.
Based on our histopathological ndings we suggest
that melasma is histopathologically characterized by
epidermal hyperpigmentation, possibly due to an
increased number of melanocytes and to increased
activity of melanogenic enzymes overlying dermal
changes caused by solar radiation. The melanocytes
are known to be functionally altered and seem to be
type-specic in melasma. What actually triggers the
focal type-specic alterations of melanocytes in
melasma is not known. A possible scenario for the
development of melasma is that certain clones of
melanocytes are activated by a series of interactions
of the skin with UV, with an inuence of female
sex hormones and genetic background.25 In addition,
the melanocytes can be altered by a wide variety
of endogenous factors produced by the melanocyte
itself (autocrine), or a local (paracrine) or systemic
 HISTOPATHOLOGY OF MELASMA IN KOREAN PATIENTS
(endocrine) environment. These factors mediate their
own signals to induce proliferation, death, differentia-
tion, dendricity, morphology, migration and pigmenta-
tion of melanocytes.2628 The possible role of these
factors in the development of melasma could be of great
importance.
Acknowledgments
W.H.K., E-S.L., S.S. and S.I. were supported by a grant
from the Graduate School of Medicine, Ajou University,
1997, and by a grant (HMP-98-M-2-0021) from the
1998 Good Health R&D Project, Ministry of Health and
Welfare, Republic of Korea.
References
1 Newcomer VD, Lindbert MC, Stenbert TH. A melanosis of the face
(chloasma). Arch Dermatol 1961; 83: 28497.
2 Sanchez NP, Pathak MA, Sato S et al. Melasma: a clinical, light
microscopic, ultrastructural, and immunofluorescence study. J Am
Acad Dermatol 1981; 4: 698710.
3 Grimes PE. Melasma etiologic and therapeutic considerations.
Arch Dermatol 1995; 131: 14537.
4 Resnick S. Melasma induced by oral contraceptive drugs. JAMA
1967; 199: 959.
5 Nordlund JJ, Boissy RE, Hearing VJ, King RA, Ortonne JP, eds.
Pigmentary System: Physiology and Pathophysiology. New York:
Oxford University Press, 1997.
2 6 Karnovsky MJ. Formaldehyde-glutaraldehyde fixation of high
osmolarity for use in electron microscopy. J Cell Biol 1965; 27: 137.
7 Adema GJ, de Boer AJ, Vogel AM, Loenen WAM. Molecular
characterization of the melanocyte lineage-specific antigen
gp100. J Biol Chem 1994; 269: 2012633.
8 Vazquez M. Melasma in men: a clinical and histologic study. Int J
Dermatol 1988; 27: 257.
9 Pathak MA, Riley FC, Fitzpatrick TB. Melanogenesis in human
skin following exposure to long-wave ultraviolet and visible light.
J Invest Dermatol 1962; 39: 43543.
10 Wiliam K, Bhawan J. Histopathology of solar lentigines of the
face: a quantitative study. J Am Acad Dermatol 1997; 36: 4447.
11 Toyoda M, Morohashi M. Morphological alterations of epidermal
melanocytes in photoageing: an ultrastructural and cytomor-
phometric study. Br J Dermatol 1998; 139: 44452.
12 Mishima Y. New technique for comprehensive demonstration of
melanin, premelanin, and tyrosinase site. J Invest Dermatol 1959;
33: 35560.
2002 British Association of Dermatologists, British Journal of Dermatology, 146, 228237
237
13 Elder D, Elenitsas R, Jaworsky C, Johnson B Jr, eds. Lever's His-
topathology of the Skin, 8th edn. Philadelphia: Lippincott-Raven,
1997.
14 Moschella SL, Hurley HJ, eds. Dermatology, 3rd edn. Philadelphia:
WB Saunders, 1992.
15 Fitzpatrick TB, Eisen AZ, Wolff K, Freedberg IM, Austen KF, eds.
Dermatology in General Medicine, 4th edn. New York: McGraw-Hill
Inc., 1993.
16 Ohkuma M. Presence of melanophages in the normal Japanese
skin. J Am Acad Dermatol 1991; 13: 327.
17 Adema GJ, de Boer AJ, Hullenaar R et al. Melanocyte lineage-
specific antigens recognized by monoclonal antibodies NKI-beteb,
HMB-50, and HMB-45 are encoded by a single cDNA. Am J Pathol
1993; 143: 157985.
18 Vennegoor C, Hageman P, van Nouhuijs H et al. A monoclonal
antibody specific for cells of the melanocyte lineage. Am J Pathol
1988; 130: 17992.
19 Smit N, Le Poole I, van den Wijngaard R et al. Expression of
different immunological markers by cultured human melano-
cytes. Arch Dermatol Res 1993; 285: 35665.
20 Adema GJ, Bakker ABH, de Boer AJ et al. PMel17 is recog-
nised by monoclonal antibodies NKI-beteb, HMB-45 and
HMB-50 and by anti-melanoma CTL. Br J Cancer 1996; 73:
10448.
21 Bhawan J. Mel-5: a novel antibody for differential diagnosis of
epidermal pigmented lesions of the skin in paraffin-embedded
sections. Melanoma Res 1997; 7: 438.
22 Choi JG, Chang SN, Kim SC et al. Multiple lentigines syndrome: a
comparison of normal skin and lentiginous skin by electron
microscopy and immunohistochemical staining. J Dermatol 1998;
25: 4005.
23 Smolle J, Wellenhof RH. Quantitative assessment of epidermal
Langerhans cells using automated image analysis. Skin Res
Technol 1998; 4: 3740.
24 Ishida O, Jimbow K. Computed image analyzing system for
quantitation of melanocyte morphology in cafe-au-lait macules of
neurofibromatosis. J Invest Dermatol 1987; 88: 28791.
25 Grando SA. Physiology of endocrine skin interrelations. J Am Acad
Dermatol 1993; 28: 98192.
26 Slominski A, Wortsman J, Mazurkiewicz JE et al. Detection of
proopiomelanocortin-derived antigens in normal and pathologic
human skin. J Lab Cutan Med 1993; 122: 65866.
27 Slominski A, Heasley D, Mazurkiewicz JE et al. Expression of
proopiomelanocortin (POMC)-derived melanocyte-stimulating
hormone (MSH) peptides in skin of basal cell carcinoma patients.
Hum Pathol 1996; 30: 20815.
28 Nagahama M, Funasaka Y, Fernandez-Frez ML et al. Immu-
noreactivity of a-melanocyte-stimulating hormone, adrenocorti-
cotrophic hormone and b-endorphin in cutaneous malignant
melanoma and benign melanocytic naevi. Br J Dermatol 1998;
138: 9815.