Recalcitrant trichophytic granuloma associated with NK-cell deficiency in a SLE patient treated with corticosteroid

ARTICLE

Case report

A 28-year-old man who had shown butterfly erythema, pancytopenia, and high titers of anti nuclear antibody and anti DNA antibody, filling the American College of Rheumatology (ACR) criteria for SLE twelve years before, was treated with a maintenance dose of prednisolone (22.5 mg/day). He had also received pulse therapy of 1,000 mg of methylprednisolone several times during the two years after the onset because of thrombocytopenia. With the exception of thrombocytopenia, cutaneous and systemic manifestations suggestive of SLE were controlled successfully. He had no episodes of atopic dermatitis, asthma, or susceptibility to viral and bacterial infections. There was no familial history of immunological deficiencies. When he was 20 years old he experienced athletes foot and the lesion gradually expanded despite therapy with topical antifungal cream. Pyodermic nodules developed on his left leg and thigh, and increased in number and size during the last 5 years.

He was referred to our clinic because of numerous purulent nodules and ulcers on his left lower extremities. On examination, blackish necrotic crusts adhered to the lesions (Fig. 1A), and yellowish pus drained from the fistulae. Scaly erythematous plaques covered approximately 70% of his body surface (Fig. 1B). Hyperkeratosis was prominent on both of his soles. No cutaneous lesions suggestive of SLE were observed.

Laboratory tests results are shown in Table I. Total protein, albumin, total bilirubin, glutamic oxaloacetic transaminase (GOT), glutamic pyruvic transaminase (GPT), serum urea nitrogen, creatinine, urinalysis, antinuclear antibody, anti HTLV-1 antibody, and anti HIV antibody were negative or within the normal range.

Hematoxylin and eosin (HE)-stained sections of biopsy specimens from the edge of the ulcer showed a granulomatous reaction consisting of plasma cells, neutrophils, lymphocytes, and giant cells (Fig. 2A, B). Small tiered and branched hyphae were observed in Grocott-stained sections (Fig. 2C). Fungal elements in the tissue were negative for anti Candida albicans and Fusarium oxysporum antibodies. Purulent exudates from the cutaneous lesions contained hyphae with smooth and thin walls (Fig. 3A). The exudates were cultured in Sabouraud's dextrose at 27° C. Two different kinds of colonies grew. One was white to cream and had a soft texture (Fig. 3B), characteristic of Candida albicans and the other was white and had a downy or cottony surface. The under side of the colony produced a deep red pigment (Fig. 3C) characteristic of T. rubrum, and small tear-shaped microconidia arranged along the sides of the hyphae were observed by microscopy.

The patient was diagnosed as having Majocchi's granuloma due to T. rubrum and Candida albicans coinfection. He was given 400 mg/day of miconazole intravenously for six months and topical therapy with 1% of clotrimazole cream. Candida albicans disappeared from the lesions soon after the treatment, whereas the purulent discharge continued because of persistent trichophytic granuloma. Oral itraconazole (200 mg/day) was then added to the initial treatments four months later. The minimum inhibitory concentration (MIC) of itraconazole for T. rubrum isolated from the patient measured by using flat agar plates containing various concentrations of itraconazole was 60 ng/ml. The concentrations of itraconazole in the patient's serum and pus were 198 ng/ml (non hydrated) in serum, 430 ng/ml (hydrated) in serum, and 210 ng/ml (non hydrated) in pus. Both flucytosine (4,000 mg/day) and griseofulvin (500 mg/day) were administered for two months to prevent coinfection of Candida albicans and the dissemination of Candida albicans to the internal organs in a 26-year-old. This treatment, however, also failed to resolve the lesions. Since the ulcers and nodules were connected to form fistulae, the fistulae were washed with miconazole, amphotericin B, and saline solution. Amphotericin B was discontinued because of a severe decrease in the patient's platelet count ten days later.

Immunological examinations

Several additional laboratory tests were done to evaluate more precisely the patient's immunological state. The results are described in Table II. Surface markers on his peripheral lymphocytes were evaluated with flow cytometry. To evaluate neutrophil functions, fluorescent monodisperse carboxylated microspheres phagocytized by neutrophils were measured by flow cytometry, and oxidative product formation by measuring oxidized dichlorofluorescein. NK cell function was determined by a ⁵¹Cr-release assay using the K-562 cell line. The activation of the complement cascade in the patient's and normal control sera was evaluated as described previously [4]. Sera chelated with either ethylene glycol-bis (beta-amino-ethyl ether) N, N'-tetraacetic acid (EGTA) or ethylenediamine tetraacetic acid (EDTA), were incubated at 37° C for 1 hr with extracts from the isolated T. rubrum strain or with 1% zymosan (Sigma, Missouri, USA). The supernatant obtained after centrifugation was analyzed by immunoelectrophoresis against anti-human C3c (DAKO, Denmark). The results demonstrated the conversion of C3, from beta1a to beta1c, in both sera. This conversion was absent in the EDTA-chelated serum, and present in the EGTA-chelated serum, indicating that complement was activated by the alternative pathway. Lymphocytes (1 x 10⁶ cells/ml) were stimulated with 0.125% phytohaemagglutinin (PHA), 5 mug/ml concanavalin A (Con A), or 50 mug/ml trichophytin (Kaken Pharmaceutical, Tokyo in Japan), and ³H thymidine uptake was measured 96 hrs later. The results are expressed as the stimulation index (SI) (stimulated lymphocyte (c.p.m.)/unstimulated lymphocyte (c.p.m.)). A delayed-type cutaneous response to PPD (0.05 mug) or trichophytin (50 mug) was not clearly demonstrated because of wide spreading of the tinea lesions, a tendency to bleed at the injection site and the administration of 12.5 mg/day of prednisolone. We examined the patterns of cytokine production by peripheral blood mononuclear cells (PBMCs) stimulated with trichophytin using a reverse transcriptase-polymerase chain reaction (RT-PCR) method. PBMCs were obtained from the patient and from a volunteer who had an episode of tinea pedis and a positive skin test reaction to trichophytin as a control. Messenger RNA was extracted from the mononuclear cells after stimulation with trichophytin (65 mug/ml) for 24 hrs, then RT-PCR was performed. The following primer sets were used: IL-2; upstream, 5'- ATGTACAGGATGCAACTCCTGTCTT -3'; downstream, 5'-GTCAGTGTTGAGATGATGCTTTGAC -3' (Ratagene, La Jolla, CA, USA); IL-5; upstream, 5'- ATGAGGATGCTTCTGCATTTG -3'; downstream, 5'- TCAACTTTCTATTATCCACTC -3'[5]; IL-10; upstream, 5'- ATGCCCCA AGCTGAGAACCAAGACCCA -3'; downstream, 5'- TCTCAAGGGGCTGG GTCAGCTATCCCA -3'[6], IFN-gamma; upstream, 5'- ATGAAATATACAAGTTA TATCTTGGCTTT -3'; downstream, 5'- GATGCTCTTCGACCTCGAAACA GCAT-3' (continental laboratory products incorporation, Burlington, MA, USA) and beta-actin; upstream, 5'- TGACGGGGTCACCCACACTGTGCCCA TCTA -3'; downstream, 5'- CTAGAAGCATT TGCGGTGGACGATGGAGGG -3' (Ratagene, La Jolla, CA, USA). The results of two independent experiments demonstrated expression of IL-2 and IFN-gamma mRNAs by the patient's lymphocytes stimulated with trichophytin. Expression of IL-5 and IL-10 mRNAs was not observed.

A remission has not been achieved yet, although systemic antifungal treatments with 200 mg/day of oral itraconazole have continued for two years, the dose of prednisolone has tapered every 6 months (12.5 mg/day), and the lymphopenia has improved except for the NK-cell lymphopenia (data not shown) and dysfunction as in Table III.

Discussion

Host defense against fungi depends mainly on the innate and adaptive immune systems [7]. Adaptive immunity has been focused on by many investigators and is thought to be the most important immune defensive mechanism against fungal infection. Many interesting discoveries in innate immunity, however, are reported recently. Innate immune recognition regulates adaptive responses by up-regulating the expression of costimulatory molecules on antigen-presenting cells through toll-like receptors [8]. Lemaitre et al. described how high susceptibility to fungal infection was induced in Drosophila with a loss-of-function mutation in the toll gene [9]. It is likely that trichophytic granuloma occurred in our case from immunodeficiency caused by SLE and the long term prednisolone treatment. Various approaches to treating the fungal infection were unsuccessful.

To find out the reasons for the incurability, we first investigated the sensitivity of the causative T. rubrum to itraconazole. We confirmed that our strain was sensitive to itraconazole and that the concentration of itraconazole in the lesion was higher than the MIC. Smith et al. [1] described how organisms in the dermis can change their morphology and adapt to the environment. Therefore, the T. rubrum in the lesion may become more resistant to itraconazole than the in vitro strain. Next, we evaluated T cell-mediated immunity and innate immunity mediated by neutrophils, NK cells, and complements against the fungus infection. Lymphopenia (655 lymphocytes/mul) has improved twelve years after (1,974 lymphocytes/mul) although the lesions remain with several antifungal treatments. This implies that lymphopenia was a cause of the onset of fungal infection but other anti-fungal mechanisms might be still unable to improve the lesions. Delayed-type hypersensitivity (DTH) to dermatophyte antigens is mediated mainly by Th1 type cells secreting IL-2 and IFN-gamma, and is correlated with the clinical course [10]. On the other hand, immediate skin test responses are associated with chronic fungal infections [11]. The patient's PBMCs were activated and secreted Th1 type cytokines upon stimulation with trichophytin. In addition, immediate skin responses were not observed and anti-trichophyton IgE was slightly increased (0.56 UA/ml). IgG increased without M-peak. These results indicate that T cell-mediated immunity to T. rubrum is present in our patient.

As a general rule, neutrophils are the most effective killers of fungi through phagocytosis and the release of oxidative products and lysosomal enzymes. To show full-cytotoxic abilities, firstly, neutrophils must be attracted to the infected lesions by chemotactic factors and signal to kill the fungus through CD18 on neutrophils [7]. The phagocytic function of the patient's neutrophils was within the normal range with a slight decrease in the generation of oxidative products. Several fungi, including T. rubrum, directly activate complement via the alternative pathway causing the release of complement-derived chemotactic factors which results in the accumulation of neutrophils in the lesions. The fungi are opsonized by complement, and phagocytized by the neutrophils [12-14]. Each complement, C3 and C4 was in a normal range in the patient. In our experiment, the patient's serum C3 was activated by zymosan and the fungal extract to a similar extent as that from a normal control. We estimated that both the complement activation pathway and the neutrophil functions were normal in this patient since he had no episodes of susceptibility to infections, although the defects of neutrophil chemotaxis, the expression of CD18 molecule, and C5 have not yet evaluated.

NK cells with CD16 and CD56 surface antigens are aggressive against fungi. NK cells binding to targets produce cytotoxic molecules and lymphokines which stimulate other effector cells [7]. The number and the function of NK cells were very low although lymphopenia has improved twelve years later. In the assay of NK cell function, however, there is a possibility that this assay might not show clinically relevant results because a cytotoxicity to K-562 cell line, not to fungus, is evaluated. We need to compare our case with other identical cases to confirm that NK cell deficiency is a reason for incurability of T. rubrum infection whereas we have not yet seen other cases like this nor found any previous report describing the implications of severe NK cell deficiency in deep mycosis. To analyse the function of NK cells directly in fungal infection NK cell knockout mice may be a valuable tool. Our immunological evaluations imply that the deficiency of innate immunity mediated by NK cells might be responsible for a part of the recalcitrant trichophytic granuloma in our case.

CONCLUSION

Acknowledgements

We would like to thank Dr. M. Ito (Shinshu University School of Medicine) for performing the immunohistological staining.

REFERENCES

1. Smith KJ, Neafie RC, Skelton HD, Barrett TL, Graham JH, Lupton GP. Majocchi's granuloma. J Cutan Pathol 1991; 18: 28-35.

2. Majocchi D. Sopra una nuova tricofizia (granuloma tricofitico), studi clinici microgici. Bull R Acd Med Roma 1883; 9: 22-223.

3. Morikawa T. Granuloma trichophyticum Majocchi, hervorgerufen von Sabouraudites uber (Castellani), Trichophyton purpureum (Bang). Arch Dermatol Syph 1937; 176: 265-81.

4. Tagami H, Natsume N, Aoshima T, Inoue F, Suehisa S, Yamada M. Analysis of transepidermal leukocyte chemotaxis in experimental dermatophytosis in guinea pigs. Arch Dermatol Res 1982; 273: 205-17.

5. Vowels BR, Lessin SR, Cassin M, Jaworsky C, Benoit B, Wolfe JT, et al. Th2 cytokine mRNA expression in skin in cutaneous T-cell lymphoma. J Invest Dermatol 1994; 103: 669-73.

6. Frankenberger M, Sternsdorf T, Pechumer H, Pforte A, Ziegler-Heitbrock HW. Differential cytokine expression in human blood monocyte subpopulations: a polymerase chain reaction analysis. Blood 1996; 87: 373-7.

7. Murphy JW. Mechanisms of natural resistance to human pathogenic fungi. Ann Rev Microbiol 1991; 45: 509-38.

8. Medzhitov R, Janeway C Jr. Innate immunity. N Engl J Med 2000; 343: 338-44.

9. Lemaitre B, Nicolas E, Michaut L, Reichhart JM, Hoffmann JA. The dorsoventral regulatory gene cassette spatzle/toll/cactus controls the potent antifungal response in Drosophila adults. Cell 1996; 86: 973-83.

10. Wagner DK, Sohnle PG. Cutaneous defenses against dermatophytes and yeasts. Clin Microbiol Rev 1995; 8: 317-35.

11. Jones HE, Reinhardt JH, Rinaldi MG. A clinical, mycological, and immunological survey for dermatophytosis. Arch Dermatol 1973; 108: 61-5.

12. Swan JW, Dahl MV, Coppo PA, Hammerschmidt DE. Complement activation by trichophyton rubrum. J Invest Dermatol 1983; 80: 156-8.

13. Tagami H, Kudoh K, Takematsu H. Inflammation and immunity in dermatophytosis. Dermatologica 1989; 1: 1-8.

14. Kozel TR. Activation of the complement system by pathogenic fungi. Clin Microbiol Rev 1996; 9: 34-46.