- Email: [email protected]
Benign lymphoplasia of the earlobes induced by gold earrings: Immunohistologic study on the cellular infiltrates
>Benign lymphoplasia of the earlobes induced by gold earrings: Immunohistologic study on the cellular infiltrates Keiji Iwatsuki, M.D., Mizuho Yamada, M.D., Masahiro Takigawa, M.D., Kunio Inoue, M.D., and Kichiro Matsumoto', M.D. Hamamatsu, Japan We report three patients with peculiar nodules of the earlobes that developed a few months after the wearing of pierced-type gold earrings. Biopsy specimens showed dense infiltration of lymphoid cells in the dermis and subcutaneous tissue associated with the formation of lymphoid follicles. The clinical and histologic pictures were similar to those of lymphocytoma cutis. Immunohistologically, the nodular lesions were characterized by the proliferation of T cells, mainly suppressor/cytotoxic T cells, and the presence of histiocytic cells displaying Leu-3a antigens on the surface. Patch tests of aqueous gold compound induced a strong skin reaction. A histologic study of the patch test reaction sites revealed a picture of ordinary allergic contact dermatitis showing that T cells, including both helper and suppressor T cells, invaded the spongiotic epidermis and the papillary dermis. Eosinophils and a few Langerhans cells were also present in the dermal infiltrates. These findings suggest that pierced-type gold earrings induced a long-term dermal response that resulted in producing benign 1ymphoplasia in the sensitized individuals. (J AM ACAD DERMATOL 1987;16:83-8.)
The sensitization potency of gold chloride has been demonstrated under conditions of "maximization" test. 1 In fact, several reports have clinically described the occurrence of contact sensitivity to gold ornaments. 2 •3 Previous investigators and we have reported peculiar nodular lesions of the earlobes that developed after the wearing of pierced-type gold earrings. 4-7 The histologic observation of the nodules demonstrated dense infiltration of lymphoid cells in the dermis and subcutaneous tissue without notable change in the epidermis_ Favorable sections showed lymphadenoid From the Department of Dermatology, Hamamatsu University School of Medicine. Accepted for pUblication July 8, 1986. Reprint requests to: Dr. K. Iwatsuki, Department of Dennatology, Hamamatsu University School of Medicine, 3600 Handa-cho, Hamamatsu, 431-31, Japan.
structure as observed in lymphocytoma cutiS. 7 It is surprising that most patients studied showed an extremely strong delayed-type hypersensitivity reaction to gold compounds. These findings suggest that a long-lasting allergic reaction to gold seems to provoke lymphoplasia in the sensitized individuals. Moreover, these cases may provide the pathomechanism underlying the development of lymphocytoma cutis. In the present study, we report three patients with nodules of the earlobes induced by piercedtype gold earrings and attempt to clarify the cel~ lular constituents of the lesions as compared with those of the positive patch test sites by light and immunoelectron microscopy. CASE REPORTS All three patients studied showed similar nodular lesions of the earlobes and had a history of wearing
83
84 Iwatsuki et al
Journal of the American Academy of Dermatology
Fig. 1. a, Case 3. A reddish nodule of the earlobe resulting from the wearing of an 18carat gold earring. b, Case 1. A strong patch test reaction to 1% gold sodium thiomalate
at 48 hours. C, Case 3. Dense infiltration of lymphoid cells in the earlobe lesions. Note the presence of grenz zone beneath the atrophic epidermis. (Hematoxylin-eosin stain; X 100.) d, Case 3. The patch test reaction at 1 week shows mononuclear cell infiltration in the dermis and spongiotic epidennis. (Hematoxylin-eosin stain; X 100.) pierced-type I8-carat gold earrings (Fig. 1, a). Clinical details of the patients are summarized in Table I. Although two of them noticed nonspecific, acute inflammation of the earlobes soon after piercing, the reaction faded away spontaneously within several days. One to 5 months after the wearing of gold earrings, subcutaneous nodules began to develop in the earlobes. The nodules became larger in size and had been present for about 2 years without improvement. The patch test results on the three patients revealed
an extremely strong reaction to 1% gold sodium thiomalate (Fig. 1, b). This compound did not induce any abnormal skin reaction in five nonnal volunteers. The patch tests with other metallic solutions, including nickel, chromium, and copper, were all negative. In Patient 1, the positive patch test reaction persisted for a few months and faded slowly. In the remaining patients, the patch test sites were surgically removed a week later. Biopsy specimens taken from nodular lesions of the
Volume 16 Number 1, Part 1
January 1987
Benign lymphoplasia induced by gold
85
Fig. 2. a, Case 2. Many Leu-2a-positive lymphocytes infiltrating in the nodular lesions of earlobes. (Semithin section stained with methylene blue; x 600.) b, Case 2. Immu~ nolabeling with Leu-3a. The number of Leu-3a-positive cells is smaller than that of Leu2a~positive ones. (Semithin section stained with methylene blue; X 600.) Note the presence of large histiocytic cells displaying Leu-3a antigens on the surface (arrowheads). c and d, Case 2. Immunoelectron microscopic findings of a Leu-3a-positive histiocytic celL Reaction products are seen on the surface. There is no Birbeck granule in the cytoplasm. Arrowheads indicate Leu-3a-pbsitive lymphocytes. (Uranyl and lead stain; C, X 4,800; d, X 11,600.)
earlobes demonstrated a dense infiltration of lymphoid cells in the dermis and subcutaneous tissue (Fig. 1, c). A small number of eosinophils and plasma cells were also recognized in the lesions. A grenz zone was observed beneath the atrophic epidermis. In the favorable sections, the formation oflymphoid follicles with germinal centerlike structure was seen. In contrast to true lymphoid follicles in lymph nodes, the germinal centerlike· structure observed in the present patients was composed mainly of esterase-positive histiocytic cells but not of immunoblasts. It was interesting that both
the clinical and the histologic pictures were essentially similar to those of lymphocytoma cutis. 8 The histologic examination of the positive skin reaction sites to the gold solution at 48 hours showed spongiosis and exocytosis of mononuclear cells in the epidermis. The patch test reaction 1 week later still demonstrated spongiotic epidermis with moderate acanthosis and infiltration of mononuclear cells in the papillary dermis (Fig. 1, d). In addition, many eosinophils infiltrated in the dermis. These reaction patterns were of an ordinary allergic contact dermatitis and distinct
Journal of the American Academy of
86 Iwatsuki et al
Dermatology
Fig. 3. a, Case 3. OKT6-positive dendritic cells in the epidermis of the patch test site. A few OKT6-positive cells are found in the dermal infiltrates. (x 200.) b, Case 3. Immunolabeling with Leu,l shows dense infiltration of T lymphocytes in the patch test site. Both Leu-2a-and Leu-3a-positive cells are infiltrating in the lesions. (X 200.)
Table I. Details of clinkal aspects of our patients Time to onset of acute dermatitis
Time to onset of nodules (rna)
Period before visiting our office (yr)
Results of patch tests
Case
Age/sex
1*
211F
2-3 days
1
2
Au, + +; Ni, Cr3' -; Cu,
2 3
20tF
Unknown 2-3 days
1 5
2
Au, ++ Au, + +; Ni, , Cr3' -; Cr6' -; Co, -; Cu, -
32/F
1.5
Therapy
Surgical removal Topical· injection of sterpid Surgical removal Surgical removal
Au: 1% gold sodium thiomalate; Cr3: 2% chromium sulfate; Cu: 1% cupric sulfate; Ni: 5% nickel sulfate; Cr.: 0.5% potassium dic!ll:omate; Co: 2% cobalt chloride. *This case was described previously in Reference 7.
from the nodular lesions of earlobes in which lymphopI asia was produced in the dermis and subcutaneous tissue in the absence of notable change in the epidermis.
IMMUNOmSTOLOGIC FINDINGS
Biopsy specimens taken from the nodules of the earlobes and the positive patch test sites were fixed in periodate-lysine-parafonnaldehyde and processed for immunostaining. The technic for monoclonal antibody staining follows. Cryostat sections, 6 /-1m, were placed on
egg albumin-coated glass slides and allowed to air dry. The sections were incubated with 1: 10 diluted OKT6 (OrthoDiagnostic Systems Inc., Raritan, NJ), Leu-I, Leu-2a, Leu-3a, and HLADR (Becton Dickinson Immunocytometry Systems, Mountain View, CA) monoclonal antibodies for 60 minutes at 37° C. They were then reacted with biotinylated goat antimouse IgG (Tago, Inc., Burlingame, CA; working dilution, 1 :40), followed by incubation of the mixture of peroxidase-labeled streptavidin (Bethesda Researc!l Lab-
Volume 16 Number 1, Part 1 January 1987
oratories, Inc., Gaithersburg, MD; working dilution, 1: 50) and mouse peroxidase-antiperoxidase complex (Medical & Biological Laboratories Co., Ltd., Nagoya, Japan; working dilution, 1:50). Binding sites of the peroxidase conjugates were visualized by incubation of the sections in 0.05% diaminobenzidine for immunoelectron microscopy or in 0.04% aminoethyl carbazole for light microscopy in the presence of 0.03% hydrogen peroxide. The sections for immunoelectron· microscopy were fixed with 2.5% glutaraldehyde, postfixed with 1% osmium tetroxide, and embedded in Epon. Ultrathin sections were stained with uranyl acetate and lead citrate and observed in a Hitachi H-300 electron microscope. Semithin sections were also made out of the Epon-embedded materials and stained with methylene blue for observation with a light microscope. These sections enabled us to study detailed morphology of the infiltrating cells and to count the accurate number of immunoreactive cells among the infiltrates. In the nodular lesions of earlobes studied (Cases 2 and 3), dense infiltration of Leu-I-positive cells (T cells) was the major finding. Among these infiltrating T cells, the number of Leu-2a-positive cells (suppressor/cytotoxic T cells) was greater than that of Leu-3a-positive cells (helper/inducer T cells) (Fig. 2, a and b). The ratio of Leu-3apositive cells to Leu-2a-positive cells was about D.22" in Patient 2 and 0.72 in Patient 3. Another characteristic finding was the presence of large histiocytic cells expressing Leu-3a-related antigens on the surface (Fig. 2, b). Electron microscopic observation showed that these histiocytic ~ells had an invaginated nucleus, many lysosomes, all(~ well-developed cytoplasmic organelles (Fig. 2,9 and d). They did not contain Birbeck granules in the cytoplasm. Although the distribution of OKT6-positive dendritic cells in the epidermis was normal, there was a small number of OKT6-positive round cells, suggestive of Langerhans cells, in the dermal infiltrates, some of which possessed Birbeck granules in the cytoplasm by immunoelectron microscopy. Biopsy specimens taken from the positive patch test sites at 1 week demonstrated infiltration of T cells, including both Leu-2a-positive and Leu-3apositive cells. The latter cell type was slightly
Benign lymphoplasia induced by gold 87
greater in number among the dermal infiltrates. Many infiltrating eosinophils were seen in the dermis as well. A few dermal Langerhans cells, which reacted with OKT6, were present in the lesions (Fig. 3, a and b). There were no histiocytic cells expressing Leu-3a antigens in the patch test sites. DISCUSSION
The present study clearly shows the fact that the wearing of gold earrings through the pierced earlobes has a risk of formation of lymphocytoma cutis-like nodules. Because all patients demonstrated a strong delayed-type hypersensitivity to aqueous gold but not to other metal solutions, the n,odular lesions are considered to be formed on the basis of an allergic reaction to gold. In the present cases the gold earrings were worn immediately after piercing of the earlobes, which permitted direct contact between the earrings and the dermis, with continuous solubilization of a small amount of gold into the tissue fluid. Moreover, nonspecific, acute inflammation noticed soon after piercing might initiate the establishment of hypersensitivity to gold and contribute to the late formation of a lymphadenoid structure. A previous report, however, described the occurrence of nodular lesions of the earlobe~, resulting from hypersensitivity to nickeJ.9 Therefore, we cannot deny the possibility that metals other than gold induce the similar reaction under different conditions. Microscopic examination of the biopsy specimens taken from the nodules of earlobes disclosed a picture of benign lymphoplasia. Immunohistologically, cellular constituents of the nodules were characterized by the proliferation of suppressor/ cytotoxic T cells and the presence of histiocytic cells expressing Leu-3a antigens on the surface, although the exact nature of the latter cell type remains obscure. The ultrastructural findings of those histiocytic cells, along with the fact that Leu3a (T-4) antigens are expressed by Langerhans cells in normal and diseased conditions, 10 may provide the evidence for Langerhans cell origin. However, we cannot exclude the possibility that these histiocytic cells are Leu-3a-positive, activated T cells that morphologically mimic histiocytes as observed in the lesions of lymphomatoid papulosis .11 Other cell types, such as helper/inducer T cells,
Journal of tbe American Academy of Dermatology
88 Iwatsuki et al
plasma cells, and a small number of eosinophils, were also found in the lesions in various pro~ portions. When the golq compound penetrated through the epidermis, however, epidermal changes, such as spongiosis'and exocytosis ofmononuc1ear cells, were induced. The cellular infiltrates of the patch test reaction were composed of T cells, induding both helper and suppressor phenotypes and a few dermal Langerhans cells. Eosinophils appeared to infiltrate at the later stage. Among the cellular infiltrates of positive patch test sites, one can occasiomilly observe the aggregates of B cells and dendritic reticulum cells similar to lymphoid follicles. 12 In the cases reported here, however, we could not :find such follicle formation in the positive patch test sites even a week later, while in the earlobe lesions we observed IYIIlphocytes surrounding aggregates of esterase-positive histiocytes, resulting in structures that mimic lymphoid follicles in lymph nodes. 7 .Since lymphoplasia has been reported to occur under peculiar conditions such as vaccinations, 13 repeated hyposensitization injections,14 and tattoos, IS it is possible that immunologic reaction in the dennis with little contribution by epidermal Langerhans cells has the capability of formation of benign lymphoplasia. The similar pathomechanism seems to be responsible for the development of lymphoplasia in the cases reported here. In addition, gold compound itself has the potency of stimulating or inhibiting immunologic responses. 16 Therefore, such immunoreactive properties of gold may prolong or modify the inflammatory reaction to induce benign lymphoplasia.
2.
3. 4. 5. 6. 7. 8. 9. 10. 11.
12. 13.
14. 15.
16.
for screening and rating contact sensitizers. J Invest Dermatol 47:393-409, 1966. Comaish S: A case of contact hypersensitivity to metallic gold. Arch Dermatol99:720-723, 1969. Shelly WB, Epstein E: Contact-sensitivity to gold as a chronic papular eruption. Arch Dermatol 87:388-391, 1963. Petros H, Macmillan AL: Allergic contact sensitivity to gold with unusual features. Br J Dermato188:505-508, 1973. Young E: Contact hypersensitivity.to metallic gold. Dermatologica 149:294-298, 1974. Fisher AA: Metallic gold: The cause of persistent allergic "dermal" contact dermatitis. Cutis 14:177-180, 1974. Iwatsuki K, Tagami H, Moriguchi T, Yamada M: Lymphadenoid structure induced by gold hypersensitivity. Arch Dennatol118:608-611, 1982. Lever WF, SChaumburg-Lever G: Histopathology of the skin, ed 6. Philadelphia, 1983, J. B. Lippincott Co., pp. 753-756. Gaul LE: Development of allergic nickel dermatitis from earrings. JAMA 200:176-178, 1967. Groh V, Tani M, Harrer A, et al: Leu-3/T4 expression on epidermal Langerhans' cells in normal and diseased skin. J Invest Dermato1 86:115-120, 1986. Kaudewitz P, Stein H, Burg G, et al: Atypical cells in lymphomatoid papulosis express the Hodgkin cell-associated antigen Ki-I. J Invest Dermatol 86:350-354, 1986. . Palfkier E, Wantzin GL: In situ immunological characterization of the infiltrating cells in positive patch tests. Br J Dermatol 111:13-22, 1984. Martsock R1: Postvaccinial lymphadenitis hyperplasia of lymphoid tissue that stimulates malignant lymphoma. Cancer 21:632-649, 1968. Bernstein H, ShupackJ, Ackerman AB: Cutaneous pseudolymphoma resulting from antigen injections. Arch DermatolllO:756-757, 1974. Blumental 0, Okun MR, Ponitch JA: Pseudolymphomatous reaction to tattoos. JAM ACAD DERMATOL 6:485 7 488, 1982. Walz DT, Griswold DE: Immunopharmacology of gold sodium thiomalate and auranofin (SK & FD-39162): Effects on cell-mediated immunity. Inflammation 3:117128, 1978.
REFERENCES L Kligman AM: The identification of contact allergens by human assay. Ill. The maximization test: A procedure
CORRECTION
In the article entitled "Cyclosporin A" by EH Page and associates, which appeared on pages 785-791·of the May 1986 issue of the JOURNAL, the tenth line in the second paragraph should read " . . . including 'V-interferon and interleukin 2." On page 788 the second line in the first paragraph should read " . . . suppression of 'V-interferon release . . ."
The sensitization potency of gold chloride has been demonstrated under conditions of "maximization" test. 1 In fact, several reports have clinically described the occurrence of contact sensitivity to gold ornaments. 2 •3 Previous investigators and we have reported peculiar nodular lesions of the earlobes that developed after the wearing of pierced-type gold earrings. 4-7 The histologic observation of the nodules demonstrated dense infiltration of lymphoid cells in the dermis and subcutaneous tissue without notable change in the epidermis_ Favorable sections showed lymphadenoid From the Department of Dermatology, Hamamatsu University School of Medicine. Accepted for pUblication July 8, 1986. Reprint requests to: Dr. K. Iwatsuki, Department of Dennatology, Hamamatsu University School of Medicine, 3600 Handa-cho, Hamamatsu, 431-31, Japan.
structure as observed in lymphocytoma cutiS. 7 It is surprising that most patients studied showed an extremely strong delayed-type hypersensitivity reaction to gold compounds. These findings suggest that a long-lasting allergic reaction to gold seems to provoke lymphoplasia in the sensitized individuals. Moreover, these cases may provide the pathomechanism underlying the development of lymphocytoma cutis. In the present study, we report three patients with nodules of the earlobes induced by piercedtype gold earrings and attempt to clarify the cel~ lular constituents of the lesions as compared with those of the positive patch test sites by light and immunoelectron microscopy. CASE REPORTS All three patients studied showed similar nodular lesions of the earlobes and had a history of wearing
83
84 Iwatsuki et al
Journal of the American Academy of Dermatology
Fig. 1. a, Case 3. A reddish nodule of the earlobe resulting from the wearing of an 18carat gold earring. b, Case 1. A strong patch test reaction to 1% gold sodium thiomalate
at 48 hours. C, Case 3. Dense infiltration of lymphoid cells in the earlobe lesions. Note the presence of grenz zone beneath the atrophic epidermis. (Hematoxylin-eosin stain; X 100.) d, Case 3. The patch test reaction at 1 week shows mononuclear cell infiltration in the dermis and spongiotic epidennis. (Hematoxylin-eosin stain; X 100.) pierced-type I8-carat gold earrings (Fig. 1, a). Clinical details of the patients are summarized in Table I. Although two of them noticed nonspecific, acute inflammation of the earlobes soon after piercing, the reaction faded away spontaneously within several days. One to 5 months after the wearing of gold earrings, subcutaneous nodules began to develop in the earlobes. The nodules became larger in size and had been present for about 2 years without improvement. The patch test results on the three patients revealed
an extremely strong reaction to 1% gold sodium thiomalate (Fig. 1, b). This compound did not induce any abnormal skin reaction in five nonnal volunteers. The patch tests with other metallic solutions, including nickel, chromium, and copper, were all negative. In Patient 1, the positive patch test reaction persisted for a few months and faded slowly. In the remaining patients, the patch test sites were surgically removed a week later. Biopsy specimens taken from nodular lesions of the
Volume 16 Number 1, Part 1
January 1987
Benign lymphoplasia induced by gold
85
Fig. 2. a, Case 2. Many Leu-2a-positive lymphocytes infiltrating in the nodular lesions of earlobes. (Semithin section stained with methylene blue; x 600.) b, Case 2. Immu~ nolabeling with Leu-3a. The number of Leu-3a-positive cells is smaller than that of Leu2a~positive ones. (Semithin section stained with methylene blue; X 600.) Note the presence of large histiocytic cells displaying Leu-3a antigens on the surface (arrowheads). c and d, Case 2. Immunoelectron microscopic findings of a Leu-3a-positive histiocytic celL Reaction products are seen on the surface. There is no Birbeck granule in the cytoplasm. Arrowheads indicate Leu-3a-pbsitive lymphocytes. (Uranyl and lead stain; C, X 4,800; d, X 11,600.)
earlobes demonstrated a dense infiltration of lymphoid cells in the dermis and subcutaneous tissue (Fig. 1, c). A small number of eosinophils and plasma cells were also recognized in the lesions. A grenz zone was observed beneath the atrophic epidermis. In the favorable sections, the formation oflymphoid follicles with germinal centerlike structure was seen. In contrast to true lymphoid follicles in lymph nodes, the germinal centerlike· structure observed in the present patients was composed mainly of esterase-positive histiocytic cells but not of immunoblasts. It was interesting that both
the clinical and the histologic pictures were essentially similar to those of lymphocytoma cutis. 8 The histologic examination of the positive skin reaction sites to the gold solution at 48 hours showed spongiosis and exocytosis of mononuclear cells in the epidermis. The patch test reaction 1 week later still demonstrated spongiotic epidermis with moderate acanthosis and infiltration of mononuclear cells in the papillary dermis (Fig. 1, d). In addition, many eosinophils infiltrated in the dermis. These reaction patterns were of an ordinary allergic contact dermatitis and distinct
Journal of the American Academy of
86 Iwatsuki et al
Dermatology
Fig. 3. a, Case 3. OKT6-positive dendritic cells in the epidermis of the patch test site. A few OKT6-positive cells are found in the dermal infiltrates. (x 200.) b, Case 3. Immunolabeling with Leu,l shows dense infiltration of T lymphocytes in the patch test site. Both Leu-2a-and Leu-3a-positive cells are infiltrating in the lesions. (X 200.)
Table I. Details of clinkal aspects of our patients Time to onset of acute dermatitis
Time to onset of nodules (rna)
Period before visiting our office (yr)
Results of patch tests
Case
Age/sex
1*
211F
2-3 days
1
2
Au, + +; Ni, Cr3' -; Cu,
2 3
20tF
Unknown 2-3 days
1 5
2
Au, ++ Au, + +; Ni, , Cr3' -; Cr6' -; Co, -; Cu, -
32/F
1.5
Therapy
Surgical removal Topical· injection of sterpid Surgical removal Surgical removal
Au: 1% gold sodium thiomalate; Cr3: 2% chromium sulfate; Cu: 1% cupric sulfate; Ni: 5% nickel sulfate; Cr.: 0.5% potassium dic!ll:omate; Co: 2% cobalt chloride. *This case was described previously in Reference 7.
from the nodular lesions of earlobes in which lymphopI asia was produced in the dermis and subcutaneous tissue in the absence of notable change in the epidermis.
IMMUNOmSTOLOGIC FINDINGS
Biopsy specimens taken from the nodules of the earlobes and the positive patch test sites were fixed in periodate-lysine-parafonnaldehyde and processed for immunostaining. The technic for monoclonal antibody staining follows. Cryostat sections, 6 /-1m, were placed on
egg albumin-coated glass slides and allowed to air dry. The sections were incubated with 1: 10 diluted OKT6 (OrthoDiagnostic Systems Inc., Raritan, NJ), Leu-I, Leu-2a, Leu-3a, and HLADR (Becton Dickinson Immunocytometry Systems, Mountain View, CA) monoclonal antibodies for 60 minutes at 37° C. They were then reacted with biotinylated goat antimouse IgG (Tago, Inc., Burlingame, CA; working dilution, 1 :40), followed by incubation of the mixture of peroxidase-labeled streptavidin (Bethesda Researc!l Lab-
Volume 16 Number 1, Part 1 January 1987
oratories, Inc., Gaithersburg, MD; working dilution, 1: 50) and mouse peroxidase-antiperoxidase complex (Medical & Biological Laboratories Co., Ltd., Nagoya, Japan; working dilution, 1:50). Binding sites of the peroxidase conjugates were visualized by incubation of the sections in 0.05% diaminobenzidine for immunoelectron microscopy or in 0.04% aminoethyl carbazole for light microscopy in the presence of 0.03% hydrogen peroxide. The sections for immunoelectron· microscopy were fixed with 2.5% glutaraldehyde, postfixed with 1% osmium tetroxide, and embedded in Epon. Ultrathin sections were stained with uranyl acetate and lead citrate and observed in a Hitachi H-300 electron microscope. Semithin sections were also made out of the Epon-embedded materials and stained with methylene blue for observation with a light microscope. These sections enabled us to study detailed morphology of the infiltrating cells and to count the accurate number of immunoreactive cells among the infiltrates. In the nodular lesions of earlobes studied (Cases 2 and 3), dense infiltration of Leu-I-positive cells (T cells) was the major finding. Among these infiltrating T cells, the number of Leu-2a-positive cells (suppressor/cytotoxic T cells) was greater than that of Leu-3a-positive cells (helper/inducer T cells) (Fig. 2, a and b). The ratio of Leu-3apositive cells to Leu-2a-positive cells was about D.22" in Patient 2 and 0.72 in Patient 3. Another characteristic finding was the presence of large histiocytic cells expressing Leu-3a-related antigens on the surface (Fig. 2, b). Electron microscopic observation showed that these histiocytic ~ells had an invaginated nucleus, many lysosomes, all(~ well-developed cytoplasmic organelles (Fig. 2,9 and d). They did not contain Birbeck granules in the cytoplasm. Although the distribution of OKT6-positive dendritic cells in the epidermis was normal, there was a small number of OKT6-positive round cells, suggestive of Langerhans cells, in the dermal infiltrates, some of which possessed Birbeck granules in the cytoplasm by immunoelectron microscopy. Biopsy specimens taken from the positive patch test sites at 1 week demonstrated infiltration of T cells, including both Leu-2a-positive and Leu-3apositive cells. The latter cell type was slightly
Benign lymphoplasia induced by gold 87
greater in number among the dermal infiltrates. Many infiltrating eosinophils were seen in the dermis as well. A few dermal Langerhans cells, which reacted with OKT6, were present in the lesions (Fig. 3, a and b). There were no histiocytic cells expressing Leu-3a antigens in the patch test sites. DISCUSSION
The present study clearly shows the fact that the wearing of gold earrings through the pierced earlobes has a risk of formation of lymphocytoma cutis-like nodules. Because all patients demonstrated a strong delayed-type hypersensitivity to aqueous gold but not to other metal solutions, the n,odular lesions are considered to be formed on the basis of an allergic reaction to gold. In the present cases the gold earrings were worn immediately after piercing of the earlobes, which permitted direct contact between the earrings and the dermis, with continuous solubilization of a small amount of gold into the tissue fluid. Moreover, nonspecific, acute inflammation noticed soon after piercing might initiate the establishment of hypersensitivity to gold and contribute to the late formation of a lymphadenoid structure. A previous report, however, described the occurrence of nodular lesions of the earlobe~, resulting from hypersensitivity to nickeJ.9 Therefore, we cannot deny the possibility that metals other than gold induce the similar reaction under different conditions. Microscopic examination of the biopsy specimens taken from the nodules of earlobes disclosed a picture of benign lymphoplasia. Immunohistologically, cellular constituents of the nodules were characterized by the proliferation of suppressor/ cytotoxic T cells and the presence of histiocytic cells expressing Leu-3a antigens on the surface, although the exact nature of the latter cell type remains obscure. The ultrastructural findings of those histiocytic cells, along with the fact that Leu3a (T-4) antigens are expressed by Langerhans cells in normal and diseased conditions, 10 may provide the evidence for Langerhans cell origin. However, we cannot exclude the possibility that these histiocytic cells are Leu-3a-positive, activated T cells that morphologically mimic histiocytes as observed in the lesions of lymphomatoid papulosis .11 Other cell types, such as helper/inducer T cells,
Journal of tbe American Academy of Dermatology
88 Iwatsuki et al
plasma cells, and a small number of eosinophils, were also found in the lesions in various pro~ portions. When the golq compound penetrated through the epidermis, however, epidermal changes, such as spongiosis'and exocytosis ofmononuc1ear cells, were induced. The cellular infiltrates of the patch test reaction were composed of T cells, induding both helper and suppressor phenotypes and a few dermal Langerhans cells. Eosinophils appeared to infiltrate at the later stage. Among the cellular infiltrates of positive patch test sites, one can occasiomilly observe the aggregates of B cells and dendritic reticulum cells similar to lymphoid follicles. 12 In the cases reported here, however, we could not :find such follicle formation in the positive patch test sites even a week later, while in the earlobe lesions we observed IYIIlphocytes surrounding aggregates of esterase-positive histiocytes, resulting in structures that mimic lymphoid follicles in lymph nodes. 7 .Since lymphoplasia has been reported to occur under peculiar conditions such as vaccinations, 13 repeated hyposensitization injections,14 and tattoos, IS it is possible that immunologic reaction in the dennis with little contribution by epidermal Langerhans cells has the capability of formation of benign lymphoplasia. The similar pathomechanism seems to be responsible for the development of lymphoplasia in the cases reported here. In addition, gold compound itself has the potency of stimulating or inhibiting immunologic responses. 16 Therefore, such immunoreactive properties of gold may prolong or modify the inflammatory reaction to induce benign lymphoplasia.
2.
3. 4. 5. 6. 7. 8. 9. 10. 11.
12. 13.
14. 15.
16.
for screening and rating contact sensitizers. J Invest Dermatol 47:393-409, 1966. Comaish S: A case of contact hypersensitivity to metallic gold. Arch Dermatol99:720-723, 1969. Shelly WB, Epstein E: Contact-sensitivity to gold as a chronic papular eruption. Arch Dermatol 87:388-391, 1963. Petros H, Macmillan AL: Allergic contact sensitivity to gold with unusual features. Br J Dermato188:505-508, 1973. Young E: Contact hypersensitivity.to metallic gold. Dermatologica 149:294-298, 1974. Fisher AA: Metallic gold: The cause of persistent allergic "dermal" contact dermatitis. Cutis 14:177-180, 1974. Iwatsuki K, Tagami H, Moriguchi T, Yamada M: Lymphadenoid structure induced by gold hypersensitivity. Arch Dennatol118:608-611, 1982. Lever WF, SChaumburg-Lever G: Histopathology of the skin, ed 6. Philadelphia, 1983, J. B. Lippincott Co., pp. 753-756. Gaul LE: Development of allergic nickel dermatitis from earrings. JAMA 200:176-178, 1967. Groh V, Tani M, Harrer A, et al: Leu-3/T4 expression on epidermal Langerhans' cells in normal and diseased skin. J Invest Dermato1 86:115-120, 1986. Kaudewitz P, Stein H, Burg G, et al: Atypical cells in lymphomatoid papulosis express the Hodgkin cell-associated antigen Ki-I. J Invest Dermatol 86:350-354, 1986. . Palfkier E, Wantzin GL: In situ immunological characterization of the infiltrating cells in positive patch tests. Br J Dermatol 111:13-22, 1984. Martsock R1: Postvaccinial lymphadenitis hyperplasia of lymphoid tissue that stimulates malignant lymphoma. Cancer 21:632-649, 1968. Bernstein H, ShupackJ, Ackerman AB: Cutaneous pseudolymphoma resulting from antigen injections. Arch DermatolllO:756-757, 1974. Blumental 0, Okun MR, Ponitch JA: Pseudolymphomatous reaction to tattoos. JAM ACAD DERMATOL 6:485 7 488, 1982. Walz DT, Griswold DE: Immunopharmacology of gold sodium thiomalate and auranofin (SK & FD-39162): Effects on cell-mediated immunity. Inflammation 3:117128, 1978.
REFERENCES L Kligman AM: The identification of contact allergens by human assay. Ill. The maximization test: A procedure
CORRECTION
In the article entitled "Cyclosporin A" by EH Page and associates, which appeared on pages 785-791·of the May 1986 issue of the JOURNAL, the tenth line in the second paragraph should read " . . . including 'V-interferon and interleukin 2." On page 788 the second line in the first paragraph should read " . . . suppression of 'V-interferon release . . ."