SARCOIDOSIS IN TAIWAN: CLINICAL CHARACTERISTICS

Transcription

1 Sarcoidosis and Atypical Mycobacteria in Taiwan SARCOIDOSIS IN TAIWAN: CLINICAL CHARACTERISTICS AND ATYPICAL MYCOBACTERIA Julia Yu-Yun Lee, Sheau-Chiou Chao, Mei-Hui Yang, and Jin-Jou Yan 1 Background and Purpose: Sarcoidosis is rare in Taiwan, although increasingly more cases are being diagnosed. The purpose of this study was to further characterize the clinical features of all biopsy-proven cases of sarcoidosis diagnosed between 1988 and 2002 in a national teaching hospital in southern Taiwan, and to detect mycobacterial DNA in sarcoid lesions. Materials and Methods: A total of 28 biopsy-proven cases of sarcoidosis were reviewed. Formalin-fixed, paraffin-embedded tissues were used as template for nested polymerase chain reaction (PCR) detection of the mycobacterial hsp65 gene. This was followed by direct DNA sequencing to identify the species in positive cases. Results: Eleven males and 17 females, aged 19 to 79 years (mean, 50 yr) were enrolled. Bilateral hilar lymphadenopathy, pulmonary lesions or both, were detected in 14 patients (50%). Of the 22 patients (78%) with skin involvement, nine manifested angiolupoid lesions on the face, a rare form of cutaneous sarcoidosis. Eight patients had papular lesions and five had plaque lesions, usually widespread on the extremities and trunk. No patients presented with erythema nodosum. Fifteen of the 22 patients with skin involvement had had the disease for more than 2 years. The hsp65 gene was detected in seven cases, four of which showed greater than 90% homology to four different species of atypical mycobacteria. Conclusion: Our study showed that sarcoidosis tended to affect women older than 40 years, often manifested angiolupoid lesions on the face, and ran a subacute to chronic course. Atypical mycobacteria might play a role in the etiology in a minority of our patients, but more cases and other potential etiologic agents need to be studied. (J Formos Med Assoc 2002;101:749 55) Key words: sarcoidosis angiolupoid mycobacterium nested PCR Sarcoidosis is a multisystem granulomatous disorder that commonly affects young adults and frequently presents with bilateral hilar lymphadenopathy (BHL), pulmonary infiltration, and skin and ocular lesions. The diagnosis is based on suggestive clinical and/or radiologic features and histologic evidence of non-caseating epithelioid cell granulomas, after exclusion of granulomas of known causes [1]. Sarcoidosis manifesting erythema nodosum (EN) or asymptomatic BHL usually follows a self-limiting course, while insidious onset of multiple extrapulmonary lesions is often associated with progressive fibrosis of the lungs and other organs [2]. Sarcoidosis is considered rare in Taiwan, but has been reported more frequently in recent decades [3 5], partly due to heightened interest and diagnostic efforts among dermatologists and pulmonary specialists. Despite extensive research, the etiology of sarcoidosis remains unknown. Infectious agents, mycobacteria in particular, have long been suspected as the main etiology. Recent etiologic studies using bacterial culture or polymerase chain reaction (PCR) reported divergent results [6 13]. We found no evidence of Mycobacterium tuberculosis complex DNA in sarcoid skin lesions from 12 patients [5]. Since that report, we have seen more cases. The purpose of this study was to analyze both the clinical spectrum of all biopsy-proven cases diagnosed in our hospital, and the results of PCR study to detect mycobacteria in lesions. This work was presented at the 29th World Congress of Dermatology in Paris, July 1 5, Departments of Dermatology and 1 Pathology, National Cheng-Kung University, Tainan. Received: 29 July Revised: 21 August Accepted: 1 October Reprint requests and correspondence to: Dr. J. Yu-Yun Lee, Department of Dermatology, National Cheng Kung University Hospital, 138 Sheng-Li Road, Tainan, Taiwan. 749

2 J.Y.Y. Lee, S.C. Chao, M.H. Yang, et al 750 Materials and Methods A total of 28 cases diagnosed in our hospital from 1988 to 2002 were retrieved from the medical records. Diagnosis of sarcoidosis was based on characteristic clinical, radiologic and pathologic features after the exclusion of other granulomatous disease. Clinicopathologic features and PCR identification of M. tuberculosis-complex IS6110 in the first 12 cases have been described [5]. Formalin-fixed, paraffin-embedded tissues were used as template for detection of mycobacterial hsp65 DNA using nested PCR. Amplicons were directly sequenced using an ABI 377 automatic sequencer (Advanced Biotechnologies, Columbia, MD, USA). Ten to 20 thin (5 µm) sections were obtained from each paraffin block and placed into 1.5-mL Eppendorf tubes. To avoid cross-contamination of the specimens, the microtome blade was changed between blocks. To increase the yield, DNA was extracted according to the method described by Li et al [12]. Specifically, tissue was digested with proteinase K 200 µg/ ml at 37 C for 7 days, during which time 20 µl of proteinase K was added daily. After digestion, the sample was heated to 95 C for 10 minutes, then subjected to 10 cycles of heat/cold shock at 95 C and -70 C to destroy any residual mycobacterial cell wall. The primers for the first round of PCR amplification were Tb-11b (5'-CCCCACGATCACCAACGAT-3') and Tb-12a (5'-CTTGTCGAACCGCATACCCT-3'). For the nested amplification, primers Tb-11a (5'- ACCAACGATGGTGTGTCCAT-3 ) and Tb-12a were used. The first PCR amplification was carried out in a 50-µL reaction mixture containing 10 µl of DNA and 1 µl of dimethylsulfoxide (DMSO). Amplification was performed with an initial 5 minute denaturation step at 94 C, followed by 40 cycles of 1 minute denaturing at 94 C, 1 minute annealing at 60 C, and 1 minute extension at 72 C; a final 10-minute extension at 72 C was added at the end. For nested amplification, 10 µl of first-round PCR product was added to 40 µl of second-round PCR mixture containing the internal primers. β-actin was amplified in each specimen as an internal control for adequacy and integrity of template DNA. Positive controls included cultures of M. tuberculosis ATCC25177 (H37Rv) (American Type Culture Collection, Rockville, MD, USA) and skin lesions of tuberculosis verrucosa cutis and Mycobacterium haemophilum cellulitis. Negative tissue controls (n = 16) included deep mycosis, foreign body granuloma, granuloma annulare, eczema and psoriasis. Results Eleven males and 17 females, aged 19 to 79 years (mean, 50 yr), were enrolled in the study. Five patients were younger than 40 years. The clinicopathologic findings and PCR results are summarized in the Table. Fourteen patients (50%) had chest lesions, 11 had BHL, and four had pulmonary infiltration or fibrosis. Patient 11 had both BHL and lung lesions. Twenty-two patients had cutaneous sarcoidosis, eight of whom also manifested chest lesions. Other organ involvement included the eye (1 patient), peripheral lymphadenopathy (2), salivary and lacrimal glands (2), heart and polyneuropathy (1), and diplopia (1). Biopsy specimens from skin lesions (22 cases), lymph nodes (7), and liver (1) showed non-caseating epithelioid or sarcoidal granulomatous infiltrate (Fig. 1) without foreign bodies of microorganisms identifiable using special stains. Areas of tuberculoid infiltrate were noted in nine cases (Patients 1, 4, 6, 7, 9, 14 17). A tuberculin test or intradermal skin test for cell-mediated immunity (CMI) with seven recall antigens including old tuberculin was performed in 15 of the 28 patients; 13 showed reduced or negative reaction. Of the 22 patients (78%) with skin sarcoidosis, lesions were limited to the face in 12 patients (55%). The other 10 patients (45%) had multiple or disseminated lesions outside the face, mostly on the extremities and trunk, and a few on the scalp. Skin lesions were classified as angiolupoid in nine patients, papular in eight, and plaque in five. Annular lesions were noted in three patients. Angiolupoid lesions, a rare form of cutaneous sarcoidosis, manifested violaceous infiltrating papules or plaques involving the side of the nasal bridge, the medial canthus, and adjacent areas (Fig. 2). Patient 9 (Fig. 2A) had a few infiltrating papules and plaques on the forehead and cheek in addition to the lesion on the nasal bridge. She also had BHL and a negative CMI test. Asthma and glaucoma had been diagnosed before this visit to the hospital for the angiolupoid lesions on her face. All the manifestations resolved after treatment with hydroxychloroquine and low-dose prednisolone. Patient 3 (Fig. 2B) had depressed CMI tests. Eight of the nine patients with angiolupoid lesions were women older than 40 years. Appearance of chest lesions was least common in patients with papular cutaneous sarcoidosis. There were no cases of EN or lupus pernio. The disease in most of our patients ran a subacute to chronic course. Twentytwo patients were followed for up to 9 years (mean, 4 yr); 15 had the disease for more than 2 years. PCR-competent DNA was obtained in 21 cases. Of the remaining seven cases, paraffin blocks were not

3 Sarcoidosis and Atypical Mycobacteria in Taiwan Table. Clinicopathologic findings and polymerase chain reaction (PCR) study of 28 patients with sarcoidosis Patient no./ Skin lesion BHL/lung Other lesions Biopsy site PCR results DNA sequencing Treatment/response Remarks age (yr)/sex type hsp65/is6110 1/42/M Plaques & Fibrotic change, Liver, LN Skin, LN, +/ Rhodococcus equi (89%) PRD, MTX/I CMI all (-), tumors R upper lung liver Lymphoma 2/29/M Papules Normal Diplopia R Skin +/ MAC (99%) PRD & KA local inj/i CMI, incl TC (w+) 3/58/F Angiolupoid Normal None Skin / Topical steroid/r CMI incl TC (w+) 4/32/M Papules Normal None Skin +/ Mycobacterium gordonae (84%), Topical steroid/i CMI (-), TC (+) MAC (83%) 5/59/F Angiolupoid Normal None Skin +/ Poor data KA local inj, PRD/R CMI (w+), TC (+) 6/54/F Angiolupoid Normal Skin IS/ None/R 7/79/M Papules & Reticular nodular None Skin / Predn/I CMI (-) plaques pattern 8/64/M Angiolupoid BHL Left calf Skin +/ M. gordonae (90%) KA local inj/i CMI (w+), TC (+) 9/52/F Angiolupoid BHL Lacrimal glands Skin ND/ND PRD & Plaquenil/I CMI all (-) 10/53/F Papules Normal None Skin +/ Poor data Topical steroid/lost FU 11/64/M Plaques & Diffuse interstitial BHL, salivary Skin IS/ PRD & Plaquenil/R CMI (-), TC (+) nodules fibrosis & lacrimal glands 12/54/M Papules Normal None Skin IS/ PRD/R CMI (-), TC (+) 13/28/F Papules Normal None Skin ND/ND PRD & Plaquenil/loss to FU PPD (-) 14/53/F Angiolupoid Fibrotic change, None Skin / KA local inj/i PPD (+) bil upper lungs 15/55/M Plaques Normal None Skin +/ M. scrofulaceum (94%) PRD & topical steroid/i 16/19/M Papules Skin /ND No treatment/stationary 17/51/F Angiolupoid Normal None Skin +/ND M. gordonae (85%), MAC (84%) KA local inj/stationary CMI all (-) 18/70/F Papules Normal Skin IS/ND PRD, Plaquenil/I 19/64/F Angiolupoid Skin +/ND Nocardia farcinica (83%), KA local inj & topical PPD (-) M. simiae strain (83%) steroid/r 20/47/F Plaques Normal Skin / Topical steroid & Minocycline/not I 21/58/F Angiolupoid BHL Complete AV block, Skin +/ M. austroafricanum (93%), PRD/I Neuropathy M. vaccae (94%) 22/43/F Plaque BHL None Skin /ND PRD/I PPD (+) 23/45/F None BHL None LN /ND PRD/R CMI (w+), TC (-) 24/49/F None BHL None LN /ND None/I 25/49/F None BHL None LN /ND None/stationary 26/47/M None BHL LN /ND PRD/I 27/45/M None BHL Subretinal hemorrhage, LN /ND Anti-TB therapy: no effect, Inf vitreous snowball changed to PRD/I 28/37/F None BHL Lacrimal glands, neck LN LN ND/ND Baktar/lost to FU BHL = bilateral hilar lymphadenopathy; R = right; LN = lymph node; PRD = prednisolone; MTX = methotrexate; I = improved; CMI = cell mediated immunity skin test, containing 7 recall antigens including tuberculin (old); MAC = Mycobacterium avium complex; KA = Kenacort (triamcinolone acetonide); TC = tuberculin; R = resolved; w+ = weak positive; IS = insufficient sample; ND = not done; FU = follow-up; bil = bilateral; PPD = tuberculin test; AV = atrioventricular; inf = inferior; TB = tuberculosis; Baktar = sulfamethoxazole/trimethoprim. 751

4 J.Y.Y. Lee, S.C. Chao, M.H. Yang, et al A B Fig. 1. A) Biopsy of a plaque lesion (Patient 1) revealing extensive non-caseating epithelioid granulomas in the dermis and subcutaneous tissue. B) A lymph node (Patient 25) revealing diffuse infiltrate of noncaseating epithelioid granulomas (Hematoxylin & eosin, A x 40; B x 100). available in three, and the DNA extracted was inadequate in quantity or quality in four. The 439-bp amplicon of the mycobacterial hsp65 gene was detected in 10 cases (Fig. 3). Seven of these showed DNA sequences related to atypical mycobacteria or Mycobacterium other than tuberculosis (MOTT), but only four A showed homology greater then 90%: Mycobacterium avium complex (MAC; 99% homology), Mycobacterium scrofulaceum (94% homology), Mycobacterium vaccae (94% homology), and Mycobacterium gordonae (90% homology) (Table). These were found in four patients with cutaneous sarcoidosis: two had angiolupoid B Fig. 2. A) Angiolupoid cutaneous sarcoidosis in a 52-year-old woman (Patient 9) presenting as a violaceous infiltrating plaque across the nasal bridge and extending to the right medial canthus. B) A 58-year-old woman (Patient 3) presenting with a violaceous plaque on the left side of the nasal bridge with prominent telangiectasia on the surface. 752

5 Sarcoidosis and Atypical Mycobacteria in Taiwan Patient no. N bp P M 439 bp Fig. 3. Polymerase chain reaction amplification of hsp65. Lane 1: 100-bp size ladder; lane 2: positive control showing the 439-bp hsp65 band; lane 3: negative reagent control; lanes 4 12 positive for hsp65 in Patients 1, 2, 4, 5, 7, 9, 15, 17 and 19. lesions (Patients 8 and 21), one (Patient 2) had papules, and one (Patient 15) had plaques. Of the remaining three cases, amplicons from two specimens from Patient 1 showed 89% homology to Rhodococcus equi. This patient had a history of frequent cleaning of lesion sites with a harsh brush prior to onset. The specificity of the amplicons from the other two patients could not be further identified. All controls gave appropriate results. Discussion Sarcoidosis is more prevalent (> 10/100,000 population), in general, in developed countries [2, 14]. It is rare in Asia, except in Japan, where the annual incidence is about 1.4/100,000 [15]. Recent reports suggest that sarcoidosis is increasing in Asia, probably due to increased interest and diagnostic efforts from doctors [15]. In Korea, national surveys revealed a gradual increase in incidence from 14 cases in 1993 (0.027/100,000) to 59 cases in 1998 (0.125/100,000) [15]. The first two Taiwanese cases of sarcoidosis were reported in the 1960s [16, 17]. In 1992, we reported the first case of cutaneous sarcoidosis [3]. Since then, only two series and a few sporadic cases have been reported from Taiwan [4, 5, 18 20]. In the present study, we reviewed all 28 cases that were diagnosed in our 800-bed teaching hospital over 14 years, two per year on average in the Tainan area (regional population, about 800,000). In the report by Perng et al from northern Taiwan, 38 patients were admitted to a large medical center over 35 years, about one per year [4]. Since there has been no island-wide survey, the true annual incidence of sarcoidosis in Taiwan remains unknown. Our study shows that sarcoidosis more commonly affected females aged more than 40 years. Similar findings were observed in the other 41 Taiwanese cases [4, 18 20]. Three different forms of sarcoid skin lesions were found in our patients, namely papular, plaque, and angiolupoid. The first two are common forms of cutaneous sarcoidosis, usually associated with acute to subacute and chronic disease, respectively [21]. Angiolupoid lesions are a rare chronic form affecting women predominantly [21]. They typically manifest as violaceous papules or plaques involving the side of the nasal bridge towards the inner canthus, or the adjacent area of the cheek. Angiolupoid lesions have a more livid hue than other forms due to the marked telangiectatic component. It is worth noting that angiolupoid lesions were noted in about one-third of our patients with cutaneous sarcoidosis, a unique aspect of our series. Clinically, angiolupoid sarcoidosis lesions may be confused with other facial infiltrating erythematous lesions, such as basal cell carcinoma, lupus vulgaris, and lupus erythematosus. These diseases can be readily differentiated pathologically. Another unique aspect of our series is the total absence of EN, a common nonspecific lesion of sarcoidosis typically associated with BHL [21]. This type of sarcoidosis usually occurs in younger patients (mean age, 30 years) and resolves completely within 2 years, in contrast to infiltrative sarcoid lesions, which tend to affect old people (mean age, 48 yr). A worldwide survey revealed that EN was the presenting sign in 17% of patients with sarcoidosis [22], varying from 53% in Scandinavia, through 31% in England, to 2 to 4% in Japan [2]. It appears unlikely that we have missed cases of EN associated with BHL. EN is not uncommon in our practice. We routinely perform chest roentgenography in such cases and have not found any cases with BHL. Thus, genetic and geographic factors might contribute to the rarity of EN in our patient population. In the series reported by Perng et al, there was a higher frequency of patients with chest lesions (97%), followed by skin involvement (24%) [4]. In our series, 78% had sarcoid skin lesions, followed by chest lesions (50%). The higher frequency of skin lesions probably reflected easier accessibility of skin lesions for biopsy and greater diagnostic efforts from dermatologists in our hospital. It is also possible that some cases with BHL and/or pulmonary infiltration were not diagnosed, in particular when extrathoracic lesions were not obvious or missed. It is not clear whether the pattern of skin lesions observed in our series is representative for Taiwanese patients. Perng et al did not specify the types of skin lesions in their report [4]. In the other three female patients [18 20], one had BHL and angiolupoid lesions, and the other two had BHL and subcutaneous 753

6 J.Y.Y. Lee, S.C. Chao, M.H. Yang, et al 754 sarcoidosis on the extremities. One of the authors (JYYL) has also seen several cases of angiolupoid lesions at consultations. Thus, we suspect that the angiolupoid form is probably a common manifestation in Taiwanese female patients. The etiology of sarcoidosis is still unknown despite extensive research. Many infectious agents, including Proprionibacterium acnes and particularly mycobacteria, have been proposed as the cause, but without firm proof [14, 21, 23]. Cell wall-defective acid-fast bacteria have been isolated from skin lesions of sarcoidosis and from blood [6, 7]. Mycobacterial DNA, mostly atypical mycobacteria or MOTT, has been detected in 20 to 70% of various types of specimens, including bronchial washings, lung, lymph node, and skin [8 10]. Popper et al found mycobacterial hsp65 DNA in 11 of 35 lung sarcoid specimens, four of which were M. avium and none was M. tuberculosis complex [11]. Li et al used nested PCR to detect mycobacterial hsp65 in a total of 20 cases and identified M. tuberculosis complex in two cases, M. avium-intracellulare in four, Mycobacterium kansasii in two, and other MOTT in eight cases [12]. Ghossein et al, on the other hand, did not find mycobacterial chaperonin in 10 cases of sarcoidosis [13]. P. acnes is indigenous in the skin and intestines of healthy humans. It has been isolated in culture from 21% of 180 non-sarcoid samples. P. acnes is a strong adjuvant and can cause granulomas when injected into sensitized rats and rabbits. Recently, P. acnes or Proprionibacterium granulosum DNA was detected in 106 of 108 lymph node sarcoidosis specimens and in 0 to 60% of tuberculosis and control samples using realtime PCR [24]. In the same study, M. tuberculosis was found in 0 to 9% of sarcoid samples. The authors suggest that propionibacteria might be more likely to be the etiologic agent in sarcoidosis than mycobacteria in both Japan and Europe. MOTT DNA was detected in only a few of our cases. There was no apparent correlation between the clinical and pathologic features and the presence of MOTT DNA. This suggests that mycobacteria probably did not play a major role in our cases. However, there was a small number of cases in our study and more cases and other potential etiologies need to be studied to draw a definite conclusion. Despite accumulating laboratory data suggesting an infectious cause in sarcoidosis, sarcoidosis responds to corticosteroid and other immunosuppressive therapy without activating an occult infection [21]. Recently, Bachelez et al have demonstrated good therapeutic responses of sarcoidosis to minocycline 200 mg/day for a median of 12 months [25]. Although the positive response supported an etiologic role for atypical mycobacteria or P. acnes, it should be noted that tetracyclines also possess anti-inflammatory and immunosuppressive properties independent of their antimicrobial effects [26]. The pathogenesis of sarcoidosis involves accumulation of CD4+ T-lymphocytes resulting in a Th-1 type immune response [23, 27]. Positive associations with HLA types are found in different populations [28], indicating genetic predisposition. Because most microbes or their DNA detected in sarcoid lesions are of low pathogenic potential, the microbes might be behaving in a non-infectious fashion to trigger a sarcoidosis granulomatous response in genetically predisposed individuals. Moreover, these microbes seem to be in a biologic state where they cannot be activated even when patients are under immunosuppressive therapy. Thus, in a practical sense, sarcoidosis may be regarded as a disorder of the hypersensitivity reaction to various microbes of low virulence rather than a true infectious disorder. In summary, our study showed that sarcoidosis tended to affect females older than 40 years, often manifested as angiolupoid lesions on the face, and ran a subacute to chronic course. Atypical mycobacteria might be involved in a minority of our patients, but further study involving more cases and other potential etiologic agents is needed. References 1. Yamamoto M, Sharma OP, Hosoda Y: The 1991 descriptive definition of sarcoidosis. Sarcoidosis 1992;9(Suppl 1): Hosoda Y, Yamaguchi M, Hiraga Y: Global epidemiology of sarcoidosis. What story do prevalence and incidence tell us? Clin Chest Med 1997;18: Lee JYY, Mak CP, Kao HF: Extrathoracic sarcoidosis in a Chinese man presenting with multiple, large plaques and tumors. J Formos Med Assoc 1992;91: Perng RP, Chen JH, Tsai TT, et al: Sarcoidosis among Chinese in Taiwan. J Formos Med Assoc 1997;96: Chao SC, Yan JJ, Lee JYY: Cutaneous sarcoidosis among Taiwanese. J Formos Med Assoc 2000;99: Kon OM, du Bois RM: Mycobacteria and sarcoidosis. Thorax 1997;52(Suppl 3):S Graham DY, Markesich DC, Kalter DC, et al: Isolation of cell wall-defective acid-fast bacteria from skin lesions of patients with sarcoidosis. In: Grassi C, Rizzato G, Pozzi E, eds. Sarcoidosis and Other Granulomatous Disorders: Proceedings of the XI World Congress on Sarcoidosis and Other Granulomatous Disorders, Milan, 6-11 September Amsterdam: Elsevier, 1988: Saboor SA, Johnson NM, McFadden J: Detection of mycobacterial DNA in sarcoidosis and tuberculosis with polymerase chain reaction. Lancet 1992;339:

7 Sarcoidosis and Atypical Mycobacteria in Taiwan 9. Mangiapagan G, Hance AJ: Mycobacterium and sarcoidosis: an overview and summary of recent molecular biological data. Sarcoidosis 1995;12: El-Zaatari FAK, Naser SA, Markesich DC, et al: Identification of Mycobacterium avium complex in sarcoidosis. J Clin Microbiol 1996;34: Popper HH, Klemen H, Hodfler G, et al: Presence of mycobacterial DNA in sarcoidosis. Hum Pathol 1997;28: Li N, Bajoghli A, Kubba A, et al: Identification of mycobacterial DNA in cutaneous lesions of sarcoidosis. J Cutan Pathol 1999;26: Ghossein R, Ross DG, Salomon RN, et al: A search for mycobacterial DNA in sarcoidosis using the polymerase chain reaction. Am J Clin Pathol 1994;101: English JC III, Patel PJ, Greer KE: Sarcoidosis. J Am Acad Dermatol 2001;44: Kim DS: Sarcoidosis in Korea: report of the second nationwide survey. Sarcoidosis Vasc Diffuse Lung Dis 2001; 18: Hsing CT, Han FC, Liu HC: Sarcoidosis among Chinese. Am Rev Respir Dis 1964;89: Wu MC: Sarcoidosis: one case report with a general discussion: J Formos Med Assoc 1960:59: Lin TH, Chuan MT, Hsu CY, et al: Sarcoidosis a case report. Dermatol Sinica 1992;10: Lee HG, Cheng ST, Yu HS: Subcutaneous sarcoidosis. Dermatol Sinica 1999;16: Chiu PC, Chu CY, Chiu HC: Subcutaneous sarcoidosis. In: Proceedings, The 26th Annual Meeting of the Chinese Dermatological Society, Taipei, 2000:79. [Abstract] 21. Savin JA: Sarcoidosis. In: Champion RH, Burton JL, Ebling FJG, eds. Textbook of Dermatology. 5th ed. Oxford: Blackwell Scientific Publications, 1992: James DG, Neville E, Siltzbach TE: A worldwide review of sarcoidosis. Ann NY Acad Sci 1976;278: Thomas PD, Hunninghake GW: Current concepts of the pathogenesis of sarcoidosis. Am Rev Respir Dis 1987;135: Eishi Y, Suga M, Ishige I, et al: Quantitative analysis of mycobacterial and proprionibacterial DNA in lymph nodes of Japanese and European patients with sarcoidosis. J Clin Microbiol 2002;40: Bachelez H, Senet P, Cadranel J, et al: The use of tetracyclines for the treatment of sarcoidosis. Arch Dermatol 2001;137: Humber P, Treffel P, Chapuis JF, et al: The tetracyclines in dermatology. J Am Acad Dermatol 1991;25: McGrath DS, Goh N, Foley PJ, et al: Sarcoidosis: genes and microbes soil or seed? Sarcoidosis Vasc Diffuse Lung Dis 2001;18: Martinetti M, Tinelli C, Kolek V, et al: The sarcoidosis map: a joint survey of clinical and immunogenetic findings in two European countries. Am J Respir Crit Care Med 1995;152: