Periorbital necrobiotic xanthogranuloma and stage I multiple myeloma Ultrastructure and response to pulsed dexamethasone documented by magnetic resonance imaging Harold Plotnick, MD,a Yoshiki Taniguchi, MD,a Ken Hashimoto, MD,a William Negendank, MD,b and Liborio Tranchida, MDc Detroit, Michigan We observed a 40-year-old woman with necrobiotic xanthogranuloma from the inception of indurated eyelid and periorbital infiltrates and concurrent stage I multiple myeloma to resolution of infiltrates in skin and bone marrow after pulsed high-dose oral dexamethasone therapy. Ultrastructural studies revealed lipid vacuoles in epidermal keratinocytes, in dermal histiocytic macrophages, and in vascular and lymphatic endothelial cells. The presence oflipid vacuoles in epidermal keratinocytes has not been reported previously in xanthogranuloma. (J AM ACAD DERMATOL 1991;25:373-7.) .
Necrobiotic xanthogranuloma (NXG) is a rare histiocytic disease that was described by Kossard and Winkelmann l in 1980. Clinically, the eruption is characterized by indurated, nontender, dermal, and subcutaneous yellow nodules and plaques that primarily infiltrate the eyelids and periorbital structures, and, to a lesser extent, the flexural extremities and trunk,2 Histologically, NXG forms individual nodules that consist of a syncytium of histiocytic cells and atypical foreign-body cells and Touton giant cells that surround areas of necrobiosis. 3 Other features include anemia, leukopenia, an elevated erythrocyte sedimentation rate, increased serum immunoglobulin, and paraproteinemia. 4 An associated malignant monoclonal gammopathy (multiple myeloma) occurs in approximately 10% of recorded cases. s We report a case of a woman with clinical, histologic, and bone marrow findings consistent with the diagnosis of NXG and concurrent stage I multiple myeloma.
From the Departments of Dermatology· and Medicinec and the Division of Hematology-Oncology,b Wayne State University School of Medicine. Presented in part at the American Academy of Dermatology Summer Session, June 16, 1988, New York City. Reprint requests: Harold Plotnick, MD, Department of Dermatology, Wayne State University School of Medicine, 540 E. Canfield, Detroit, MI 48201. 16/4/23690
CASE REPORT
A 40-year-old white woman who was receiving intermittent treatment for a chronic and recurring adult-type cystic acne with intralesional corticosteroid injections had eyelid and periorbital tissue changes during a 3-month period. The lesions were characterized by nontender, firm, bandlike infiltrates in both upper eyelids; a subcutaneous hard nodule (0.6 em) in each eyebrow arch; an infiltrated mass (2.5 X 1.8 em) in the right paranasal fold; and several yellow indurated inilltrates iri the medial half of the left lower eyelid. There was no bone pain, lymphadenopathy, or hepatosplenomegaly. Histopathology and electron microscopy (EM) Punch biopsy specimens were taken from a xathomatous area of the left lower eyelid and from the nodules in the right and left eyebrow arch. Each specimen was divided into three parts. The first part was processed for hematoxylin and eosin staining; the second was frozen in OCT compound (Miles Laboratories, Elkhart, Ind.) and used for immunohistochemical staining; the third was processed for electron microscopy as previously described. 6 Lipid staining, 7 periodic acid-Schiff staining (PAS), and iron staining were done in the usual manner. Monoclonal and polyclonal antibody staining was done according to the avidin-biotin immunoperoxidase method (Vecstatin ABC Kit, Vector Laboratory, Burlingame, Calif.). The left eyebrow nodule showed large masses of granulomatous infiltrate with a dense proliferation of histiocytes, small lobules of epithelioid cells, and mariy multinucleated epithelial giant cells with a few scattered lymphocytes and occasional eosinophils within a matrix
373
Journal of the American Academy of Dermatology
374 Plotnick et al.
PAS and iron stains were negative in the cytoplasm ofinfiltrates. Electron microscopy showed that the cytoplasm ofkeratinocytes in the epidermis and hair follicles was swollen and contained some vacuoles surrounded by an electrondense amorphous material; vacuoles were also seen in intercellular spaces (Fig. 2, A). These changes were extensive in the basal cell layer (Fig. 2, B). Almost all histiocytes in the dermis contained small vacuoles in their cytoplasm. Vacuoles were observed in the cytoplasm of vascular and lymphatic endothelial cells (Fig. 3). No plasma cells were seen.
Clinical laboratory findings
Fig. 1. A, Low magnification view shows clustered giant cells in granulomatous foci (g). Hematoxylin-eosin stain X2S.) B. Touton-type giant cells are seen just outside necrobiotic foci (n). (Xl00). C, Necrobiotic foci (n) surrounded by palisading granulomatous cell infiltration. (X1SO.)
The results of the fasting blood sugar test, the fourstage glucose tolerance test, the serum lipid profile, the serumcalcium assay, and the creatinine assay were within normal limits. The patient's hemoglobin level was 11.4 gm/dl (normalll.S to 14.8 gm/dl); RBe count, 3.39 X 106/mm3 (normal 3.9 to 5.1S X 106 /mm3); WBC; 3.0 X 103/mm3 (normaI3.S to 9.8 X 103/mm3); granulocytes, 1.3 X 103 /mm3 (normal 1.9 to 7.1 X 103 /mm3); erythrocyte sedimentation rate, 80 rom/hr. A bone marrow biopsy specimen revealed 6.2% plasmacytes (normal < 2%); the plasma cells showed a mature chromatincontent and no nucleoli. The bone marrow hematopoietic cell differential was normal. The antinuclear antibody (HEp2 cells) was nonreactive. Quantitativeserumimmunoglobulin analysis showed the following values: IgG, 4000 mg/dl (normal 1350 mg/dl); IgM, 150mg/dl (normal S5 to 350 mg/dl); IgA, 370 mg/dl (normal 70 to 310 mg/dl). Serum immunoelectrophoresis revealed IgG-K light-chain monoclonal gammopathy. The urine electrophoresis showed a trace of free K light chains. X-ray film revealed a generalized osteopenia. There was no evidence of osteolysis.
Magnetic resonance imaging (MRI) of collagen tissue that surrounded areas of necrobiosis (Fig. 1). The two other specimens had similar cellular changes but showed no evidence of necrobiosis. Cholesterol clefts were not found in sections taken from the three specimens. Immunohistochemical profiles of dermal infiltrates were as follows: Leu-2a (CD8) (+), Leu-3a (CD4) (++), Leu-4 (CD3) (++), OKMI (++), OKT6 (-), lysozyme (-), al-anti-chymotrypsin (-), vimentin (+), and S-100 protein (- ). Results offat staining with Sudan III, Sudan IV, Sudan Black B, Oil Red 0, and Nile blue stains were positive in dermal histiocytic cells and epidermis. The Sudan Black B stain identified aggregates of small lipid droplets along the basem~nt membrane area and in the intercellular and intracellular spaces of the epidermis.
The MRI scans showed diffuse and focal decreases of signal intensity in the thoracic and lumbar vertebrae, the sacral-pelvic region, and the proximal femurs (Fig. 4, A).
Treatment Although the bone marrow did not reveal signs of multiple myeloma, other test results supported a diagnosis of stage I multiple myeloma according to the Durie-Salmon classification8: free light chains in the urine, typical abnormalities revealed by MRI, diffuse osteopenia, leukopenia, and a high level of monoclonal IgG. The inherent association of NXG with neutropenia necessitated moderation in the choice of chemotherapy.3,4 Therefore, therapy was started with mel-
Volume 25 Number 2, Part 2 August 1991
Necrobiotic xanthogranuloma 375
Fig. 2. A, Prickle cell layer: some vacuoles surrounded by electron-dense amorphous material are seen in the intercellular space (arrowhead); some are found in the keratinocyte (*). Nu, Nucleus. (X4500.) B, Basal cell layer reveals chains ofsmall vacuoles in the intercellular space (arrowheads); some vacuoles are seen in the cytoplasm of the basal cell (*). (X4000.)
Fig. 3. A, Dermal infiltrate: almost all histiocytes have small vacuoles in the cytoplasm. Nu, Nucleus. (X3000.) B, Small vacuoles are seen in endothelial cells of dermal lymphatic vessel. Lu, Lumen; Nu, nucleus. (X3000.)
phalan (Alkeran), 6 mg a day for 4 days. Marked improvement in the periorbital infiltrates was noted within 1 week. However, there was gradual enlargement ofthe periorbital lesions during the fourth and fifth weeks after therapy. A second course of melphalan was given. The NXG infiltrates improved, but leukopenia (WBC 2.6 X 103 /rom3) and granulocytopenia (570/mm3) per-
sisted for 7 weeks, and the bone marrow lesions did not improve. When the periocular inilltrates returned, a regimen of high-dose, pulsed glucocorticoids was initiated.9 The protocol consisted of oral dexamethasone administered in courses of 32 mg a day for 3 days, 16 mg the fourth day, 8 mg the fifth, and then 4 mg on the sixth and seventh days. The first course of the drug resulted in
Journal of the American Academy of Dermatology
376 Plotnick et al.
Fig.' 4. Tt-weighted MRI images ofthe sacrum and posterior iliac wings. A, At time of diagnosis. Diffuse and focal regions of low signal intensity signify replacement of normal, partially fatty marroW with cellular material (arrowhead) B, After four courses of high-dose dexamethasone. Replacementof the large lesion in left sacral wing with normal fatty marrow that is homogeneously bright on the T1-weighted image (arrOws).
qomplete regression ofthe periorbital infiltrates. Subsequentcourses were given whe.11 the periorbital infiltrates recurred, usWdlYjit4-~o.11-week intervals. After four c()l,lrsesofpuls¢(j.~erap¥,serum IgG was reducedto 2600 mg/ell, WBCs~nc~eas¢dto4.1 X 103/ mm,3 and granu19CYies increase<1to2;7~.X 103/mm3. Healing of the sacral-~lvic my~loma Ie$ionswas documented by MRI (Fig. 4, jJ). DISCUSSION
The histologic pattern of T-cell granuloma, the characteristic findings on electron microscopy, the negative S-100 protein for Langerhans cells, the high concentration of monocyte/macrophage phenotype (OKMI), and monoclonal antibody studies confirm that NXG is a non-X histiocytosis. 4, 10 Epidermal and endothelial lipid deposits that we have noted in our EM studies have never been observed in necrobiosis lipoidica diabeticorum, a condition that may be histologically confused with NXG.2 Lipid materials seem to derive from the dermis through the intercellular spaces of the epideqnal keratinocytes and are phagocytosed by them. Signals from the bone in MRI are not from cortexor trabeculae, but· from the marrow. ll The intensity of the signals is determined by the proportions of fat and hematopoietic tissue. Decreased
signal intensity signifies replacement of normal fatcontaining marrow by a cellular process. On T 1weighted images, fat in the marrow gives a relatively bright signal, like that of the subcutaneous tissue, whereas hematopoietic tissue appears less brightan important point in the detection of marrow disease, including multiple myeloma. 12 The apparent clinical discrepancy between a bone marrow plasmacyte level of 6.2% and the decreased signal intensity in MRI that represents myelomatous cellular replacement offat in multiple myeloma seems to suggest that the bone marrow is not uniformly involved by the cellular infiltrate; Because the bone marrow biopsy specimen represents a sample of only a small area, the initial findings may not always reveal variability in bone marrow involvement. Our experience in treating NXG with intermittent pulsed high doses of dexamethasone held the NXG inflammatory response in check without further depressing the patient's leukopenia. Moreover, partial remission of the patient's myeloma was also produced.
REFERENCES 1. Kossard S, Winkelmann RK. Necrobiotic xanthogranuloma with paraproteinemia. J AM ACAD DERMATOL
1980;3:257-70.
Volume 25 Number 2, Part 2 August 1991
2. Finan MC, Winkelmann RK. Histopathology of necrobiotic xanthogranuloma with paraproteinemia. J Cutan Pathol 1987;14:92-9. 3. Robertson DM, Winkelmann RK. Opthalmic features of necrobiotic xanthogranuloma with paraproteinemia. Am J OphthalmoI1984;97:173-83. 4. Finan MC, Winkelmann RK. Necrobiotic xanthogranuloma with paraproteinemia. Medicine 1986;65:376-88. 5. Kyle RA, Greiopp PRo Smoldering multiple myeloma. N Engl J Med 1980;302:1347-9. 6. Hashimoto K, Burk JD, Bale GF, et al. Transient bullous dermolysis of the newborn: two additional cases. 1 AM ACAD DERMATOL 1989;21:708-13. 7. Clark G. Staining procedures. 4th ed. Baltimore: Williams & Wilkins 1981:189-95.
Necrobiotic xanthogranuloma 8. Durie BGM, Salmon SE. A clinical staging system for multiple myeloma. Cancer 1975;36:842-54. 9. Alexanian R, Berlogie B, Dixon D. High-dose glucocorticoid treatment of resistant myeloma. Ann Intern Moo 1983;105:8-11. 10. Venencie PY, Puissant A, Verola 0, et al. Necrobiotic xanthogranuloma with myeloma: a case report. Cancer 1987;59:588-92. 11. Bernstein MJ. Magnetic resonance imaging. Consensus Conference. lAMA 1988;259:2132-8. 12. Daffner RH, Lupetin AR, Dash N, et al. MRI in the detection of malignant infiltration of bone marrow. Am J RoentgenoI1986;146:353-8.
Proteus syndrome Ultrastructural study of linear verrucous and depigmented nevi Vincenzo Nazzaro, MD,t Stefano Cambiaghi, MD,a Alessandro Montagnani, MD,b Alberto Brusasco, MD,a Amilcare Cerri, MD,a and Ruggero Caputo, MDa Milan, Italy Proteus syndrome is a rare hamartomatous disorder charact~erized by multifocal overgrowths that can involve any structure of the body. Clinical manifestations include macrodactyly, hemihypertrophy, subcutaneous masses, exostosis, cerebroid thickening of palms and soles, and linear skin lesions. About 50 cases have been described, but the ultrastructural features of the linear skin lesions have not been characterized. We describe the clinical, histologic, and ultrastructural findings for a 30-year-old patient who had a mild form of Proteus syndrome with linear lesions characterized by a mixed pattern of hyperkeratosis and depigmentation. Light microscopy of the linear nevus showed acanthosis and hyperorthokeratosis. Electron microscopy revealed extensive vacuolation at the interface between melanocytes and keratinocytes, with large aggregations of densely packed granules in the intercellular space. Melanocytes showed only slight degenerative changes. An immunohistochemical study of the expression of epidermal growth factor receptors revealed no significant abnormalities. (J AM ACAD DERMATOL 1991;25:377-83.) Proteus syndrome is a hamartomatous disorder characterized by multiple focal overgrowths that can involve any structure of the body. The name of the syndrome was suggested by Wiedemann et a1. 1 in 1983 from the name of the Greek god "Proteus," the polymorph, to emphasize the wide range of clio-
From the Centro Malattie Cutanee Ereditarie, Dermatology Clinie I, University of Milan,' and the Centro Auxologico Italiano.b Reprint requests: Vincenzo Nazzaro, MD, Centro Malattie Cutanee Ereditarie, Clinica Dermatologica Prima, via Pace 9, 20122 Milan, Italy. tDeceased.
16/4/23681
ical manifestations. About 50 cases have been reported; this number includes several cases described before 1983 and subsequently discovered to be Proteus syndrome. 2-8 Typical clinical features include progressive and asymmetric megalodactyly, hemihypertrophy, subcutaneous masses, localized cerebroid thickening of palms and sales, and linear skin lesions. 9- 11 It has been suggested that growth-factor abnormalities are responsible for the tissue overgrowths, but no growth-factor studies have been reported to date. Apart from four cases in adults 9, 10 aU reported cases have been in children. The long-term prognosis is not well known. Linear skin lesions, which have been observed in
377