ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 23, No. 5 Copyright © 1993, Institute for Clinical Science, Inc.

Malassezia furfur Folliculitis in Cancer Patients The Need for Interaction of Microbiologist, Surgical Pathologist, and Clinician in Facilitating Identification by the Clinical Microbiology Laboratory*

RAMON L. SANDIN, M.D., M.S.,f TZANN-TARN FANG, M.D.,* JOHN W. HIEMENZ, M.D.,t JOHN N. GREENE, M.D.,§ LINA CARD, M.T. (ASCP),t ALEXANDRA KALIK, M.D.,t and JENO E. SZAKACS, M.D.t Departments of Pathology, f Bone Marrow Transplant Service,t and Infectious Diseases,§ H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, Florida 33612-9497

ABSTRACT Malassezia furfur (MF) is a lipophilic yeast which can be found as a member of the indigenous microbiota of human skin. In immunocompro­ mised transplant patients, MF can cause a distinctive folliculitis which is a clinical look-alike to Candida folliculitis, the latter of more potentially devastating significance. Recovery of MF in culture is dependent upon the addition to culture media of an exogenous source of fatty acids, such as olive oil. The addition of an extra Sabourauds plate with an olive oil overlay to the routine set of media used to inoculate all skin biopsy specimens in order to detect MF is labor-intensive and not cost-effective. Thus, MF may not be isolated in cases of MF folliculitis unless the clinical microbiology laboratory is put on alert by the clinical suspicions of the attending physi­ cian, or by histopathologic findings suggestive of folliculitis revealed by review of surgical pathology slides. The clinical, pathological, and micro­ biological findings of two cases of MF folliculitis are presented where an interactive approach featuring communication between the microbiologist, the surgical pathologist, and the clinician guided the microbiology labora­ tory to the isolation and identification of isolates of MF that were clinically- relevant. These cases underscore how a combined approach which features communication between the laboratory and the clinical services always provides superior guidance in the diagnosis and therapy of infec­ tious diseases.

* Send reprint requests to: Ramon L. Sandin, M.D., Room 2071 Pathology, H. Lee Moffitt Cancer Center, 12902 Magnolia Drive, Tampa, FL 33612-9497. 377 0091-7370/93/0900-0377 $01.20 © Institute for Clinical Science, Inc. 3 7 8 SANDIN, FANG, HIEMENZ, GREENE, CARD, KALIK, AND SZAKACS

Introduction from these other conditions on purely clinical grounds is not always feasible, Malassezia furfur (MF) is a lipophilic one alternate approach to diagnosis yeast, which is a member of the indige­ would be to add an extra Sabourauds nous microbiota of the skin of many plate with an olive oil overlay to the rou­ warm-blooded hosts.5 Growth of MF on tine set of media inoculated with speci­ culture media is dependent upon the mens from the skin. addition of an exogenous source of This global approach, however, would medium- to long-chain fatty acids, such be labor-intensive and neither wise nor as olive oil. Malassezia furfur can cause cost-effective, since the mere recovery of an erythematous papulopustular folliculi­ MF from skin biopsies may be hard to tis that may be more common than is rec- interpret in the absence of clinical suspi­ ognized in immunosuppressed cion, patient history or of histopathologi- patients4,5,7 and which must be differen­ cal findings suggestive of a frank follicu­ tiated from the macronodular lesions of litis. Subsequently, an interactive disseminated candidiasis, as well as approach which features communication other conditions.1,5,6, Since distinction between the microbiologist, the surgical

F ig u r es 1A a n d IB. The clinical mani­ festations of Malassezia furfur (M F ) folliculitis in the two described patients were similar: an erythematous, pruritic, papulopustular rash on the back and shoulders, which can be appreciated in low and high magnifi­ cation photographs. Some lesions became unroofed and were replaced by scabs. The raised (papu­ lar) character of these lesions can be appreci­ ated in the high-power photograph. MALASSEZIA FURFUR FOLLICULITIS 3 7 9 pathologist and the clinician would ture were inoculated onto Sabourauds appear to be the most reasonable way in dextrose agar (SAB), SAB with chloram­ which to guide the laboratory staff phenicol, and brain heart infusion agar. towards isolation and identification of All were incubated at 30°C. For isolation clinically-relevant isolates of this organ­ of MF, a few drops of sterile olive oil ism, while decreasing labor and cost. were added to the SAB plate. Tentative Case reports of two patients with fol­ identification of MF was carried out by licular rashes will be presented which the observation of growth in the pres­ illustrate the convenience of such an ence, but not in the absence, of olive oil. interactive approach, without which iso­ Definitive identification relied on colo­ lation and identification of MF as the nial morphology suggestive of a yeast, etiological agent of folliculitis in these direct smears of colonial growth with the two patients would have been missed. expected microscopic morphology, lack of reactivity in the Vitek YBC yeast iden­ Materials and Methods tification card* and positive results fol­

Tissues from skin biopsy specimens submitted to microbiology for fungal cul­ * Vitek Systems, Hazelwood, MO.

Figure IB. Continued. 380 SANDIN, FANG, HIEMENZ, GREENE, CARD, KALIK, AND SZAKACS lowing the rapid urea test. Skin biopsy enteritis and mucositis. Owing to persistent fevers, specimens sent to surgical pathology empiric antifungal coverage with amphotericin B was begun five days post transplant. By this time, were formalin-fixed, paraffin-embedded, the skin rash had begun to resolve. Skin biopsy sug­ and stained with hematoxylin and eosin gested fungus; however, the rash had significantly improved before cultures of the skin biopsy grew (H & E), gomori-methenamine silver Malassezia furfur. By day 10 post transplant, WBC stain (GMS), and periodic acid-Schiff was 1,200 and the ANC was >500 per mm3. The stain (PAS). patient defervesced and was subsequently dis­ charged from hospital 19 days post-transplant with her rash resolved. Case Reports Case #2 Case #1 A 22-year-old white female presented with a his­ A 34-year-old white female with breast cancer tory of Burkitt’s lymphoma of the distal ileum diag­ metastatic to bone was first diagnosed in January nosed at laparotomy. After resection of the tumor, 1991. The patient had initial response to cyclophos­ the patient was treated with five cycles of cyclo­ phamide, adriamycin, and 5-fluorouracil systemic phosphamide, adriamycin, vincristine, and pred­ chemotherapy but showed progressive disease by nisone along with prophylactic intrathecal metho­ October 1991, as evidenced by increased bone pain, trexate. The patient did well off therapy from Jan­ worsening metastatic disease on bone scan, and ris­ uary to March 1992 when she was noted to have a ing tumor markers (CEA and CA 15-3). The patient mandibular and intra-abdominal relapse with retro­ was referred to the H. Lee Moffitt Cancer Center at peritoneal and mesenteric adenopathy. Reinduc­ the University of South Florida and relapse therapy tion was accomplished with high dose cytosine ara- was begun with standard dose ifosfamide, carbo- binoside and cisplatinum in April 1992. The patient platinum, and etoposide. After documenting subsequently was referred to the H. Lee Moffitt response the patient underwent peripheral stem Cancer Center and received two cycles of standard cell harvest after intravenous cyclophosphamide. dose chemotherapy with ifosfamide, carbo­ The patient was admitted to the bone marrow platinum, and etoposide achieving a com­ transplant unit on April 1, 1992 for high dose che­ plete remission. motherapy and peripheral blood stem cell rescue. In order to minimize the risk or relapse, the On admission, the patient was found to be well patient was admitted to the bone marrow transplant developed and in no acute distress. She was afe­ unit in July 1992 for high dose chemotherapy and brile with stable vital signs. No skin rash or lesions autologous bone marrow transplantation. On admis­ were seen. There was no adenopathy. Admission sion, examination revealed the patient to be a well laboratory exam revealed a white blood cell (WBC) developed, well nourished young female in no dis­ count of 3,100 and an actual neutrophil count (ANC) tress. There were no skin rash or lesions. Head and of 2,100. Renal and liver functions were normal. neck exam showed resolution of the mandibular Preadmission scans revealed boney metastatic dis­ tumor and no other abnormalities. Admission labo­ ease and were otherwise unremarkable. ratory evaluation revealed WBC 7,200 and ANC Therapy was begun with high dose ifosfamide, 6,300. Chemistry panel was unremarkable with nor­ carboplatinum, and etoposide daily for six days. mal renal and liver functions. Preadmission com­ Treatment was delivered in a single patient hospital puterized tomography showed resolution of previ­ room equipped with laminar air flow and HEPA ous adenopathy. filters. Nystatin was used for oropharyngeal candi­ Upon admission, the patient underwent harvest diasis prophylaxis, and aerosolized pentamidine for of 922 ml of bone marrow from the posterior iliac Pneumocystis carinii prophylaxis. No systemic anti­ crests. The following day she began high dose che­ bacterial or antifungal prophylaxis was utilized. motherapy with ifosfamide, carboplatinum, and After completion of chemotherapy, two days were etoposide given daily for six days. After two days of allowed for drug elimination, and cryopre- rest for drug elimination, cryopreserved bone mar­ served stem cells were then thawed and rein­ row was thawed and reinfused intravenously. All fused intravenously. therapy was delivered in a single bed hospital room The patient tolerated chemotherapy relatively equipped with laminar air flow and HEPA filters. well except for nausea controlled with antiemetics Nystatin was used for local oropharyngeal Candida and dexamethasone. Peripheral blood counts fell prophylaxis, and aerosolized pentamidine was and the patient had her first fever to 38°C the day given for Pneumocystis carinii prophylaxis. No sys­ before transplantation. Cultures were obtained and temic antibacterial or antifungal prophylaxis broad spectrum antibiotic coverage with ceftazi­ was utilized. dime and vancomycin was begun. The day follow­ By completion of chemotherapy, two days before ing transplantation a follicular skin rash over the transplantation, WBC had fallen to 2,000 and a fol­ shoulders, neck and back appeared (figure 1A & licular rash was noted on the shoulders. The first IB). Fevers persisted and metronidazole was added fever occurred on the day of transplantation and for anaerobic coverage with the development of empiric broad spectrum coverage with ceftazidime MALASSEZIA FURFUR FOLLICULITIS 381 and vancomycin was begun after cultures were were set up and were completely unreac­ obtained. The follicular rash had spread to involve the upper chest but was otherwise asymptomatic. tive. A rapid urea test was set up and the Daily fevers from 38°C to 38.5°C persisted, and met­ medium turned positive (red-pink) ronidazole was added for anaerobic coverage in the within a few hours. Besides MF, several setting of worsening mucositis. Skin biopsy was obtained and clotrimazole cream was begun. Owing species within the genera Cryptococcus, to persistent fevers in the setting of prolonged neu­ Rhodotorula, and Trichosporon are also tropenia, empiric antifungal coverage was initiated urease positive. The latter three genera, 10 days after transplant with fluconazole. By this time, the follicular rash had already begun to fade however, can be ruled out on the basis of: before systemic antifungal therapy had been (1) growth in the absence of added olive started. Over the next week the rash continued to oil, (2) their own particular colonial and resolve and bone marrow recovered with an ANC of microscopic morphologic characteristics, >500 per mm3 by day 18 after transplantation. All antibiotics were discontinued and the patient was and (3) specific patterns of reactivity in discharged for outpatient followup. The rash had the VITEK YBC card. begun to fade once culture results identified Malas- sezia furfur. Specimens from the skin biopsy on case #2 were inoculated onto fungal Microbiologic and Pathologic Findings media with olive oil from the start, given the clinical suspicion of the astute clini­ Tissue sent to microbiology from the cian who alerted the microbiologist prior skin biopsy on patient # 1 failed to show to the biopsy procedure. Within three growth after two days of incubation in the days, yeast colonies were identified with absence of olive oil (figure 2A). At that characteristics identical to those of time, review of the H & E slides by the patient #1. The surgical pathology microbiologist, upon requests by the sur­ specimen from patient # 2 , however, gical pathology resident, revealed acute failed to reveal the presence of fun­ fungal folliculitis, with fibrinoid necrosis gal microorganisms. of one of the walls of the hair follicles, but without inflammatory cells. Within the Discussion infundibulum of the hair follicle in the necrotic areas, there were multiple round Malassezia furfur is a common yeasts which were highlighted by the lipophilic fungal saprophyte of human PAS (figure 3) and GMS stains. The skin that is known to cause the superficial yeasts were small with unipolar and dermatosis tinea (or pityriasis) versicolor, broad-based budding, and a small cir­ as well as a distinctive folliculitis.5 On cumferential thickening at the bud occasion, a deep-seated infection may attachment. Hyphae and pseudohyphae occur particularly in debilitated hosts were absent. with central catheters who are receiving Given such histopathologic findings intravenous lipid therapy.5 Several stud­ suggestive of MF, the microbiologist ies have established the high incidence requested that the medical technologist of MF skin carriage on normal-appearing add a few drops of sterile olive oil to the skin of the scalp, shoulders and chests plates and spread them over the inocu­ of adults.5 lum. Within two days following olive oil The entity of MF folliculitis is proba­ addition, the previously negative fungal bly a more common clinical problem than plates were overrun with small, white, is currently appreciated and is of par­ creamy colonies of variable sizes (figure ticular importance to clinicians because 2B). A wet mount prepared from the of its potential for confusion with life- growth on the plate showed small yeasts threatening fungal infections.1’5,6,7 It with unipolar and broad-based budding, occurs predominately in post-adolescents and collarettes. The VITEK YBC cards as opposed to acne vulgaris which it 3 8 2 SANDIN, FANG, HIEMENZ, GREENE, CARD, KALIK, AND SZAKACS

F ig u r e s 2A and 2B. Tissue from the skin biopsy on patient # 1 was cultured on Sabourauds dextrose agar (SAB) medium but failed to show growth after two days of incubation at 30°C in the absence of olive oil (2A). Two days following the addition of a few drops of sterile olive oil to the same plate, it became overrun with small, white creamy colonies of vari­ able size (2B). MALASSEZIA FURFUR FOLLICULITIS 3 8 3

F i g u r e 3 . Surgical pathology tissue slides from patient # 1 showed fibrinoid necrosis of the hair follicle walls and the presence within the infundibulum of multiple round yeasts with unipo­ lar broad-based budding. This was highlighted by the periodic acid-Schiff (PAS) and gomori-methe- namine silver (GMS) stains. (PAS stain, lOOOx magnification).

resembles.5 Many patients with MF fol­ Within the hair follicle, numerous bud­ liculitis have underlying debilitative dis­ ding yeasts are usually present without eases or conditions, such as diabetes mel- hyphal or pseudohyphal forms, as litus, cancer, bone marrow opposed to C andida which features both transplantation, and steroid or broad- filamentous and yeast forms in tissue in spectrum antibiotic administration.1,5 cases of deep or disseminated infection As represented by both cases, lesions (figure 3). An intensely pruritic eosino­ of MF folliculitis are usually multiple philic pustular folliculitis associated with and are distributed over the back, shoul­ abundant MF yeasts inside hair follicles ders and/or upper chest.1,5 Biopsy of has also been associated with AIDS affected hair follicles shows a spectrum patients.2 Studies utilizing electron of pathological findings, from plugging of microscopy have suggested that follicular the hair follicle to frank destruction of the occlusion may be a primary event in follicular wall with fibrinoid necrosis, as the development of MF folliculitis, seen in patient #1. An infundibular infil­ with yeast overgrowth as a second­ trate consisting of polymorphonuclear ary phenomenon.3 and mononuclear leukocytes may or The lesions of MF folliculitis are to be may not be present and was absent in contrasted with those of disseminated patient #1. candidiasis, which are clinical look- 384 SANDIN, FANG, HIEMENZ, GREENE, CARD, KALIK, AND SZAKACS alikes but are of a much graver signifi­ This approach can be adapted by some cance.1,5,6 The macronodular lesions of institutions, and is perhaps already being disseminated candidiasis may be located practiced in many others. It underscores anywhere on the body but have predilec­ to laboratorians that a combined tion for the extremities, whereas candidal approach to patient care, where continual papulopustulosis is most commonly communication exists between pathology found in heroin addicts and most fre­ and the clinical services, provides supe­ quently in the bearded area, shoulders, rior guidance in the diagnosis and and chest.5 Candida also favors intertrigi- therapy of infectious and neoplas­ nous moist areas such as the groin, axilla tic diseases. and skin folds, particularly in diabetics and obese patients. Satellite lesions are common as well. Both types of candidal References lesions yield the responsible fungus if 1. B u f i l l , J. A., L u m , L. G., C a y a , J. G., C h i - biopsy material is placed on standard TAMBAR, C . R ., RlTCH, P. S., ANDERSON, T ., fungal culture media, whereas tissue and ASH, R. C.: Pityrosporum folliculitis after bone marrow transplantation. Clinical observa­ from MF folliculitis requires the lipid tions in five patients. Ann. Int. Med. 108:560— supplementation already alluded to. Sur­ 563, 1988. gical pathological examinations in MF 2. F e r r a n d iz , C ., R ib e r a , M., Ba r r a n c o , J. C., C l o t e t , B ., and L o r e n z o , J. C .: Eosinophilic folliculitis may disclose the presence of pustular folliculitis in patients with acquired yeasts within the follicular infundibulum immunodeficiency syndrome. Intern. J. Der­ and perhaps within the surrounding epi­ matol. 31:193-195, 1992. dermis, but without the deep dermal 3. H il l , M. K., G o o d f ie l d , M. J. D., R o d g e r s , F. G ., C r o w l e y , J. L., and Sa ih a n , E. M.: invasion which may be observed fre­ Skin surface electron microscopy in Pityrospo­ quently with infections by C andida.6 rum folliculitis. The role of follicular occlusion D estruction of the hair follicle is in disease and the response to oral ketocona- zole. Arch. Dermatol. 226:1071-1074, 1990. also possible. 4. J a c in t o -Ja m o r a , S., T a m e s is , J., and Ka t ig - The approach followed in the clinical b a k , M. L.: Pityrosporum folliculitis in the microbiology laboratory at our institution Philippines: Diagnosis, prevalence and man­ agement. J. Am. Acad. Dermatol. 24:693-696, is a conservative one. It relies on feed­ 1991. back from the surgical pathology resident 5. Kl o t z , S. A.: Systemic fungal infections: Diag­ or attending, and/or clinical attending or nosis and treatment. II. Malassezia furfur. fellow, in order to alert the laboratory Infect. Dis. Clin. North Amer. 3:53—64, 1989. that a particular specimen from skin is to 6 . Kl o t z , S. A., D r u t z , D . J., H u p p e r t , M., and JOHNSON, J. E.: Pityrosporum folliculitis. Its be tested for the presence of this organ­ potential for confusion with skin lesions of sys­ ism. This is superior to any approach temic candidiasis. Arch. Intern. Med. 142: requiring that an additional fungal plate 2126-2129, 1982. 7. MARCON, M. J. and POWELL, D. A.: H u m a n with an olive oil overlay be part of the infections due to Malassezia spp. Clin. Micro­ work-up of all skin biopsy specimens. biol. Revs. 5:101-119, 1992.