Outbreak of Invasive Mycoses Caused by Paecilomyces lilacinus from a Contaminated Skin Lotion

  1. Bernd Orth, MD;
  2. Reno Frei, MD;
  3. Peter H. Itin, MD;
  4. Michael G. Rinaldi, PhD;
  5. Bruno Speck, MD;
  6. Alois Gratwohl, MD; and
  7. Andreas F. Widmer, MD, MS
  1. From University Hospital, Basel, Switzerland; and the University of Texas Health Science Center, San Antonio, Texas. Acknowledgments: The authors thank the following persons from the University Hospital, Basel, Switzerland: J. Wuhrmann, H. Leisinger, C. Grasslin, K. Haupt, and the staff of the hematology-oncology isolation and bone marrow transplantation unit; A. Tichelli, M. Bargetzi, J. Passweg, and the staff of the Division of Hematology; W. Stauffacher and M. Uhr of the Department of Internal Medicine; C. Haberthur and I. Fey of the Division of Clinical Epidemiology; T. Renz, G. Dietsche, and the staff of the Bacteriology Laboratory; W. Zimmerli of the Division of Infectious Diseases; and H.L. Kain, J. Messerli, and the staff of the Clinics of Ophthalmology. The authors also thank G. Schar of the Institute of Medical Microbiology, University of Zurich, Zurich, Switzerland, and J. Pettypool for secretarial support. Requests for Reprints: Andreas F. Widmer, MD, MS, Division of Clinical Epidemiology, University Hospital, Petersgraben 4, CH-4031 Basel, Switzerland. Current Author Addresses: Dr. Orth: Division of Hematology, Klinik Bavaria, An der Wolfsschlucht 1-2, Kreischa bei Dresden 01731, Germany.
     
    Next Section

    Abstract

    Background: Invasive mycoses are an important cause of illness and death in immunocompromised patients. Infections with molds other than aspergilli have been increasingly seen in patients with hematologic cancers, but epidemics of these infections have not yet been reported.

    Objective: To describe an outbreak of invasive mycoses with Paecilomyces lilacinus in severely neutropenic patients.

    Design: An outbreak investigation.

    Setting: The hematology-oncology isolation and bone marrow transplantation unit of the University Hospital, Basel, Switzerland.

    Patients: 25 consecutive patients admitted between 17 August 1993 (the date of the first manifestation of P. lilacinus infection) and 31 October 1993 (when the unit was closed).

    Measurements: Clinical and microbiological data, including histologic findings; cultures from several patient sites; and environmental examinations of potential airborne, parenteral, enteric, and horizontal routes of transmission. Infections were defined by the isolation of P. lilacinus from clinically evident skin eruptions.

    Results: 12 of the 25 patients (48%) were infected or colonized. Nine patients (36%), including all bone marrow transplant recipients, had documented invasive P. lilacinus infections. All 9 infected patients had papular, pustular, or necrotic skin eruptions. Two patients with severe graft-versus-host disease died with refractory fungal disease; 1 also had microbiologically documented endophthalmitis and kidney infiltrates. Seven affected patients no longer had P. lilacinus after recovery of bone marrow function. The organism was resistant in vitro to amphotericin B, itraconazole, and fluconazole. Patients did not respond clinically to these agents. The outbreak was ultimately traced to a contaminated, commercially available, pharmaceutically prepared skin lotion. The outbreak ended after the skin lotion was recalled and has not recurred after a follow-up period of 2 years.

    Conclusion: Contaminated skin lotion is a potential cause of opportunistic fungal infections in immunocompromised hosts. Paecilomyces lilacinus is a common saprophytic mold that can cause, by direct cutaneous inoculation, invasive infections associated with illness and death.

    Invasive mycoses present a major challenge in patients with prolonged neutropenia [1]. Impaired cellular immunity and suppressed polymorphonuclear leukocyte function are the main risk factors for fungal disease. Protracted neutropenia, graft-versus-host disease, therapy for graft-versus-host disease, severe mucositis, skin toxicity after chemoradiotherapy, broad-spectrum antimicrobial prophylaxis, parenteral nutrition, and central venous access devices all predispose patients to locally invasive infection and fungemia. Candida species are the most frequent cause of fungal infections, followed by aspergilli and Fusarium species, in patients after bone marrow transplantation and in patients with leukemia [2]. Outbreaks caused by aspergilli are frequently associated with ongoing hospital construction work [3]. Infection with molds other than aspergilli, such as Paecilomyces species, are rare. These organisms are found in soil, in decaying vegetation [4], and in animal reservoirs worldwide [5]. They are also frequently encountered as contaminants in laboratories [6-8] and in food processing and are resistant to many disinfectants [9] and standard sterilizing techniques. Occasionally, Paecilomyces species have been reported as infective agents in immunosuppressed patients [10]. We report an epidemic of Paecilomyces lilacinus infection, which was traced to the use of contaminated skin lotion, at the hematology-oncology isolation and bone marrow transplantation unit of the University Hospital, Basel, Switzerland.

    Previous SectionNext Section

    Methods

    Hospital and Unit Setting

    The University Hospital is a 1400-bed tertiary care center. The hematology-oncology isolation and bone marrow transplantation unit consists of a reverse isolation ward and a laminar airflow isolation ward for bone marrow transplant recipients. The unit is separated from other parts of the hospital by locked doors. It was restructured and renovated in 1988 [11]. The air in the reverse isolation ward is filtered (pore size of the filters, 1 to 10 µm). Air with fewer than 10 particles/m3 is provided for patients receiving a bone marrow transplant. Three hundred sixty-seven bone marrow transplantations were done in Basel between 1 May 1973 and 31 December 1993; no epidemic caused by horizontal transmission was detected during that time. Routine surveillance of patients includes weekly cultures of skin, tissue from the throat and nose, urine, stool, and blood. The water supply is checked for bacteria three times weekly, and the air is sampled monthly.

    Patients

    All 25 consecutive patients who were admitted to the unit between 17 August and 31 October 1993 were included in this analysis. Five patients were admitted to the laminar airflow ward for bone marrow transplantation, and 20 were admitted to the reverse isolation ward. Bone marrow transplant recipients were placed in laminar airflow units; patients who did not receive bone marrow transplants were placed in single rooms with reverse isolation. Antimicrobial therapy and precautions against infection in neutropenic patients were applied according to published guidelines [1]. All patients began receiving oral nystatin (2.4 × 106 units, 3 times daily) on admission. Central venous catheters were inserted and parenteral nutrition was given to patients receiving myeloablative therapy.

    Bone marrow transplant recipients were cared for under sterile conditions as described elsewhere [12]. Low-dose amphotericin B (10 mg/d) was given as prophylaxis beginning 8 days before transplantation [13]. Standardized conditioning therapy contained etoposide, cyclophosphamide, and total body irradiation [14]. Prophylaxis for graft-versus-host disease consisted of cyclosporin with or without methotrexate [12]. All patients were given a commercially available moisturizing skin lotion to prevent and treat xeroderma and skin toxicity secondary to chemotherapy. Bone marrow transplant recipients continued to use this skin lotion after discharge.

    Epidemiology

    On 17 August 1993, a filamentous fungus was isolated from a skin lesion of a patient in the laminar airflow ward. The identification of P. lilacinus in two other patients on the laminar airflow ward by 13 September suggested the occurrence of an outbreak. The division of clinical epidemiology was notified, and a working group was convoked. Weekly sessions with a written protocol were initiated and held throughout the epidemic. As a first intervention, isolation precautions and surveillance cultures were intensified. Clinically, the skin lesions suggested bloodborne spread of fungus; epidemiologically, however, most fungal outbreaks are airborne. Therefore, bloodborne and airborne routes of transmission were systematically investigated. Daily samples of all infusates and transfusion sets, all replaced intravascular catheters, and air samples were analyzed in the microbiology laboratory. This analysis failed to show the source of the outbreak. The investigation was therefore extended to the horizontal and enteral modes of transmission. Medications, skin preparations (including two samples that turned out to be the source of the outbreak), and nursing aids were collected and cultured. Since 1992, samples of all food items from the hospital kitchen had been stored in the freezer for 7 days after meals had been distributed. Patients' menu plans were checked, and all items ingested at this unit were selectively cultured for P. lilacinus. Hand cultures from health care workers in the unit and from patients (taken at admission and weekly thereafter) were screened for P. lilacinus.

    Despite major efforts, the outbreak continued. By 31 October 1993, six patients had become infected with P. lilacinus but no epidemiologic or microbiological evidence of the origin or source had been found. Consequently, new patients were not admitted for 3 months, and many of the 60 health care workers from the unit were reassigned to other units in the hospital. A temporary epidemic unit was set up in another wing of the hospital for patients who were infected or colonized. On 4 January 1994, the source of infection was identified. The unit was then reopened after high-level disinfection, change of air conditioner filters, and implementation of a vigorous surveillance system for early detection of P. lilacinus.

    Microbiology

    Skin biopsy specimens were stained with hematoxylin-eosin and periodic acid-Schiff stains. Needle aspirations and cutaneous lavage were done as described elsewhere [15]. All aspirations were done by the same investigator. Clinical specimens were processed according to standard methods [16]. For fungal cultures, Sabouraud dextrose agar plates with and without chloramphenicol and gentamicin were inoculated and incubated at 28 °C for at least 7 days. Growth was subcultured onto potato dextrose agar. To detect P. lilacinus in clinical and environmental specimens contaminated with other fungi, a selective Sabouraud dextrose agar containing 4 mg of amphotericin B per liter was developed and used in parallel with standard Sabourad agar that did not contain amphotericin B. Hand cultures were done with the broth-bag technique using tryptone soya broth supplemented with 3% polysorbate 80, 3% saponin, 0.1% histidine, and 0.1% cysteine [17]. The organism was identified by its morphologic characteristics, including 1) floccose white colonies that gradually turned lilac on the front and purple to brown on the reverse side and 2) a microscopic form that resembled Penicillium organisms (except for flask-shaped phialides with elongated tapered necks bearing divergent chains of elliptic conidia). Confirmation of the identification of the isolates and the susceptibility test results were done at the Fungus Testing Laboratory (University of Texas Health Science Center, San Antonio, Texas), according to modifications of the proposed National Committee for Clinical Laboratory standards [18].

    Patients were defined as having infection if P. lilacinus was cultured from clinically evident skin eruptions. Colonization was defined as evidence of P. lilacinus in any surveillance culture in the absence of lesions. Outcome was either death related to infection with P. lilacinus or elimination of fungal disease documented by at least one negative culture.

    Statistical Analysis

    We collected data on 42 variables from 25 patients; the variables included age, sex, underlying disease and treatment, day of admission, start of aplasia, duration of hospitalization, and antibiotic and antimycotic therapy. Univariate analysis of the case–control study was done using the Student t-test or the chi-square statistic. If fewer than five observations were available, the Fisher exact test was used. A two-tailed P value less than 0.05 was considered significant. All data were analyzed by using the computer software StatCalc (Epi-Info, Atlanta, Georgia), Systat 5 (SYSTAT, Inc., Evanston, Illinois), or EGRET (Statistics and Epidemiology Research Corp., Seattle, Washington).

    Previous SectionNext Section

    Results

    Twelve of the 25 patients (48%) admitted between 17 August and 31 October 1993 were infected or colonized with P. lilacinus. All 12 had protracted neutropenia (absolute neutrophil count <0.5 × 109/L). Patient characteristics are shown in Table 1. Ten other patients never had neutropenia and were not infected. None of the 9 patients who had severe aplastic anemia, the myelodysplastic syndrome, or solid tumor were involved in the outbreak.

    View this table:
    Table 1. Underlying Disease and Clinical Signs of Patients Infected with Paecilomyces lilacinus

    Invasive infections developed in all 5 (100%) recipients of allogeneic bone marrow grafts and in 4 of 11 (36%) patients treated with induction chemotherapy for leukemia and lymphoma. Three additional patients who received chemotherapy (12%) were colonized only on their hands. The median time between the onset of neutropenia and clinical evidence of infection was 49 days (range, 13 to 107 days) in bone marrow transplant recipients and 14 days (range, 6 to 27 days) in patients after induction chemotherapy. Skin eruptions appeared in 6 of 9 patients during neutropenia. In 3 bone marrow transplant recipients, lesions appeared after recovery from aplasia (13, 63, and 90 days after marrow engraftment). These patients continued to use the skin lotion until its contamination was detected. In 2 bone marrow transplant recipients, the first cutaneous manifestations were detected 2 and 5 days before engraftment and 13 and 22 days after onset of neutropenia. The case–control study did not show clues to the source of the outbreak. In the univariate analysis, bone marrow transplantation, diagnosis of acute leukemia, treatment with ablative chemotherapy, and presence of neutropenia were associated with P. lilacinus infection (Table 2).

    View this table:
    Table 2. Main Characteristics of Patients Exposed to Paecilomyces lilacinus

    Clinical Signs and Symptoms

    Clinical presentation and outcome of infections and colonizations are listed in Table 1. Clinical performance status was not impaired, and only one patient had fever at the initial manifestation of skin eruptions. All nine infected patients developed solitary or disseminated skin eruptions (Figure 1 and Figure 2). These lesions were the only microbiologically documented manifestations of P. lilacinus infection in eight of the nine infected patients. All lesions appeared asymmetrically on the extremities and on the trunk and face (in two patients). The appearance of the lesions varied, consisting of erythematous papules and nodules measuring 3 to 15 mm with necrotic centers. Lesions were not tender or pruritic.

    Figure 1.
    View larger version:
      Figure 1. Discrete erythematous papulovesicular lesions in patient 5.
      Figure 2.
      View larger version:
        Figure 2. Necrotizing papulovesicular lesions in patient 7.

        Eight of the nine infected patients showed transient interstitial pulmonary infiltrates before skin lesions appeared. Paecilomyces lilacinus could not be identified in bronchoalveolar lavage fluid from these patients. All patients were initially treated with high-dose amphotericin B (1 mg/kg of body weight) and other antifungal agents as clinical failure became obvious and susceptibility testing became available. One patient had a protracted course with hematogenous dissemination of P. lilacinus: Several skin eruptions first appeared during aplasia, 13 days after bone marrow transplantation and under empirical treatment with amphotericin B (>0.5 mg/kg). After initial improvement, endophthalmitis was diagnosed and was associated with several chorioretinitis lesions (Figure 3) that required vitrectomy. Paecilomyces lilacinus grew in cultures from needle aspirates of skin lesions and from vitreal granulomas. The patient died despite receiving aggressive treatment with amphotericin B (1.0 mg/kg·d−1), administration of several imidazoles according to the results of susceptibility testing, and administration of granulocyte- and granulocyte-macrophage-stimulating factors to improve severely impaired bactericidal activity [19, 20]. At autopsy, blood culture results were positive, and the kidneys were infiltrated by P. lilacinus. A second patient died with disseminated necrotizing skin eruptions and acute graft-versus-host disease (grade III to IV) 57 days after bone marrow transplantation and 32 days after first clinical evidence of P. lilacinus infection. The request for autopsy was declined.

        Figure 3.
        View larger version:
          Figure 3. Endophthalmitis and appearance of several chorioretinitic lesions in patient 3.

          Microbiology

          Paecilomyces lilacinus was isolated from nine infected patients in 68 clinical specimens taken from skin (56 samples from nine patients), hair (3 samples from one patient), blood (at autopsy of one patient), throat (one patient), nose (one patient), vitreal chamber (3 samples from one patient), central venous catheters (2 samples from one patient), and kidney (at autopsy of one patient). Repeated cultures with positive results ranged from 2 to 21 per patient (median, 7 per patient). Two hand cultures from the three colonized patients were positive for P. lilacinus. Further isolates were detected in 7 biopsy specimens (from four infected patients), 20 needle aspirations (from seven infected patients), and 13 swabs (from six infected patients). In seven of the nine infected and in the three colonized patients, P. lilacinus was identified in 23 cultures of hands. Specimens from the 25 patients (n = 625) were negative, including all blood, stool, urine, vaginal, penile, and sputum cultures.

          Susceptibility testing of P. lilacinus indicated presumptive resistance to amphotericin B (minimum inhibitory concentration [MIC], 16 mg/L), fluconazole (MIC, 64 to >64 mg/L), and itraconazole (MIC, 4.0 mg/L). Possible borderline susceptibility to ketoconazole (MIC, 0.5 mg/L) and miconazole (MIC, 0.5 to 1.0 mg/L) was found. None of these compounds had fungicidal activity in vitro (minimum lethal concentration, ≥ 8 mg/L). Infected patients did not respond clinically to any of these agents.

          Histology

          Seven skin biopsies were done in four patients with cutaneous manifestations. In two patients with hemorrhagic and necrotic lesions suggestive of septic ecthyma, histologic examination showed hyphae in periodic acid-Schiff stained sections (Figure 4). Hyphae were found in the stratum corneum and in the perivascular area within a dense mixed infiltrate. The histologic findings were otherwise nonspecific, as is the case with ecthyma. In one additional specimen from a subcutaneous lesion without clinical involvement of the epidermis, a granulomatous infiltrate with numerous hyphae was found but no hyphae were present in the stratum corneum. This patient showed well-documented hematogenous spread of P. lilacinus. In three patients, mycotic invasion of the skin could not be found by dermatohistologic examination. In these patients, the skin lesions were more widespread and erythematous with fewer hemorrhagic-necrotic components. The lymphocytic infiltrate was less pronounced in these patients than in patients in whom hyphae were documented by periodic acid-Schiff stain.

          Figure 4. Several inflammatory cells can be seen. (Periodic acid-Schiff stain; original magnification, × 32.).
          View larger version:
            Figure 4. Several inflammatory cells can be seen. (Periodic acid-Schiff stain; original magnification, × 32.). Hyphae in the stratum corneum epidermidis and the deep dermis in patient 3.

            Epidemiologic Investigations

            All potential routes of transmission were systematically analyzed after the case–control study did not identify a clue for a targeted investigation. More than 1000 samples of all parenteral prescriptions, infusate bottles, and transfusion sets were cultured to rule out a bloodborne source of infection. Paecilomyces lilacinus did not grow in cultures of 130 food samples that represented all food items eaten by affected patients during 1 week of the epidemic. All topical agents administered to patients were investigated. Twice monthly hand cultures of health care workers during the epidemic were negative for P. lilacinus. Cultures of hair from health care workers were done because P. lilacinus was identified from the scalp of one patient; the results of these cultures were negative. Two of 252 sedimentation plates placed at the borders of the unit and one environmental culture from a patient room were contaminated by P. lilacinus. Air samples and in-depth examination of the air conditioning and central ventilation system were negative. No clue to the source of the outbreak was found after 3500 cultures had been examined.

            On 21 December 1993, colonization with P. lilacinus was detected in an immunocompetent patient on an open general medicine ward. The patient had a malleolar ulcer from which P. lilacinus grew. This ulcer had been treated with the same moisturizing skin lotion in use on the isolation ward. Simultaneously, surveillance cultures of a nurse of the hematology-oncology unit who was temporarily working in the surgery department became positive for P. lilacinus. The nurse had applied the same skin lotion to the surgical patients with her bare hands. The lotion in use was found to be contaminated with P. lilacinus.

            With this strong epidemiologic evidence, a second, extended screening of unopened bottles of the skin lotion was initiated. The bacteriology laboratory isolated P. lilacinus in 12 of 16 sealed bottles. The results of the screening of the first bottle in August 1993 were negative, even after rechecking with selective culture media in December 1993. Contamination of the samples varied between 6 and 12 500 colony-forming units per milliliter of lotion. These findings were confirmed at the Institute of Medical Microbiology of the University of Zurich, Zurich, Switzerland. The pharmaceutical components of the skin lotion include 36% lipids, 40 mg of urea per milliliter, and triclosan (0.3%) and chlorhexidine dihydrochloride (0.34%) as preservatives. Investigations done by the manufacturer of the skin lotion showed that the ingredients of the lotion were not contaminated. Paecilomyces lilacinus was finally detected in empty containers that were awaiting bottling.

            The unit was reopened on 4 January 1994 after all lots of skin lotion had been recalled. Continued, unchanged, microbiological surveillance of patients, health care workers, and the environment for P. lilacinus has shown no further cases of colonization and infection for 2 years. The intensified microbiological surveillance used during the outbreak was discontinued 3 months after no further evidence of P. lilacinus infection was found.

            Previous SectionNext Section

            Discussion

            We describe an epidemic with P. lilacinus that was traced to a rare mode of transmission—direct inoculation of the microorganism onto the skin—with secondary hematogenous spread in one case. The setting (a seriously ill patient population), the microorganism (a rare mold), and an unknown mode of transmission severely challenged the abilities of the hospital epidemiology staff. The case–control study provided no clues. Therefore, systematic microbiological investigation of the potential transmission routes—airborne, parenteral, enteral, and horizontal—was warranted. A total of 3500 cultures taken from parenteral, enteral, and topical items (such as nursing aids) did not show the source of infection. However, needle aspirations combined with cutaneous lavages from skin eruptions recovered P. lilacinus from all infected patients.

            Eventually, a positive hand culture from a nurse who was temporarily working in a surgery department and a wound colonization with P. lilacinus in a patient from a general medicine ward suggested a link to a skin lotion that was widely used in the hospital. Use of the skin lotion was a basic standard of care and was therefore not documented on the patients' charts. Only certain bottles of the skin lotion were contaminated, and microbial density among contaminated bottles varied greatly. Results of testing of the two containers examined at the beginning of the outbreak were still negative after retesting was done with selective culture media. Moreover, the amount of skin lotion that was applied depended on the damage done to the skin by induction or conditioning therapy and differed greatly among patients. Bone marrow transplant recipients, especially those affected by graft-versus-host disease, continued to use the moisturizing lotion extensively, even after discharge. These factors may explain the fluctuating course of the epidemic and the late onset of infection in two of the five bone marrow transplant recipients.

            Seven of nine P. lilacinus infections manifested as cutaneous lesions without evidence of dissemination. Hematogenous spread was documented in one patient and clinically suspected in another. Skin lesions erupted insidiously without fever and varied greatly. They appeared as discrete erythematous papules and molluscum contagiosum-like lesions similar to those seen in Penicillium infections [21]. The lesions were also similar in appearance to the hemorrhagic necrotizing papulovesicular and ecthyma gangrenosum-like eruptions seen in Fusarium infections [22] and in disseminated aspergillosis. Lesions were located on all sites of the skin but were found mainly on the lower extremities and, in two cases, were disseminated on the trunk and face. This distribution pattern can be explained by application of the skin lotion to the extremities by the patients themselves. Biopsies of the skin lesions showed locally invasive disease with hyphae in the stratum corneum and in the dermis, especially in the perivascular regions. These findings indicate that infection was caused primarily by cutaneous inoculation but in at least one case, the cutaneous lesions appeared to have been caused by hematogenous spread.

            Reports have been published of sporadic Paecilomyces infections in immunosuppressed patients after solid organ transplantation [7, 23, 24], after cytotoxic chemotherapy [25], during chronic granulomatous disease [26], or as the result of a specific immunologic deficiency [27]. Isolation of the organism from toenails and fingernails in immunocompetent patients shows the keratinous properties of P. lilacinus[28]. To our knowledge, articles on infections with P. lilacinus after bone marrow transplantation have not been published.

            Patients need moisturizing lotion to prevent and treat xeroderma and skin toxicity secondary to chemotherapy. Health care workers need to use skin lotion regularly to prevent dry, irritated skin and dermatitis after frequent handwashing. Unfortunately, as our experience shows, even such a basic treatment may put patients at risk for nosocomial infection. This epidemic emphasizes the importance of appropriate quality standards for skin lotion used for severely immunocompromised patients.

            Paecilomyces lilacinus was apparently resistant in vitro to amphotericin B, itraconazole, and fluconazole and did not respond clinically to these agents or to miconazole. Recovery from myelosuppression was crucial for clinical resolution. Skin lesions resolved rapidly after the recovery of neutrophil counts in all patients except those with graft-versus-host disease.

            These findings are consistent with recent observations of emerging, drug-resistant fungal pathogens, which have previously been considered only as colonizing organisms or laboratory contaminants [29]. Drug-resistant fungal pathogens have been described with intensified antifungal prophylaxis [30]. Hyalohyphomycosis has no standard treatment [29]. Fusarium, Penicillium, and Paecilomyces species differ in their sensitivity to antifungal drugs. Resistance of P. lilacinus in vitro to amphotericin B and flucytosine has been documented [7, 31], although the response of P. lilacinus infections in vivo to amphotericin B alone or combined with flucytosine [32] has been reported in single cases.

            This outbreak highlights the importance of combined epidemiologic and microbiological strategies to control an epidemic. Both initially failed: Epidemiologic tools could not identify an association with a factor present only in affected patients. The microbiological strategy could not find the source of the epidemic despite initial culture of two samples of the skin lotion. An observational link, namely the analysis of data from a nonimmunocompromised colonized patient and a health care worker outside the affected unit, led to extensive microbiological investigations of unopened, sealed containers of skin lotion and finally to the discovery of P. lilacinus in 75% of examined containers.

            The information gleaned from this epidemic supports previous reports that contaminated topical agents are a potential nosocomial hazard [33]. Skin lotions are frequently used in hospitals, and a high degree of suspicion is needed to recognize contaminated products. Isolation of P. lilacinus from clinical specimens does not always indicate laboratory contamination but should direct attention to early detection of invasive and disseminated disease. Paecilomyces lilacinus is a microorganism that is resistant to many drugs and can cause severe and fatal opportunistic infections. The combined efforts of clinical epidemiology and microbiology can identify the source of a rare disease and an uncommon mode of transmission, but these efforts need to include sensitive and sophisticated observation as well as more traditional laboratory measures.

            Dr. Frei: Department of Laboratory Medicine, Bacteriology Laboratory, University Hospital, Petersgraben 4, CH-4031 Basel, Switzerland.

            Dr. Itin: Department of Dermatology, University Hospital, Petersgraben 4, CH-4031 Basel, Switzerland.

            Dr. Rinaldi: University of Texas Health Science Center, San Antonio, TX 78284-7756.

            Drs. Speck and Gratwohl: Division of Hematology, University Hospital, Petersgraben 4, CH-4031 Basel, Switzerland.

            Dr. Widmer: Division of Clinical Epidemiology, University Hospital, Petersgraben 4, CH-4031 Basel, Switzerland.

            Previous Section
             

            References

            1. 1.↵
              Pizzo PA. Management of fever in patients with cancer and treatment-induced neutropenia. N Engl J Med. 1993; 328:1323-32.
            2. 2.↵
              Analssie E. Opportunistic mycoses in the immunocompromised host: experience at a cancer center and review. Clin Infect Dis. 1992; 14(Suppl 1):S43-53.
            3. 3.↵
              Opal SM, Asp AA, Cannady PB Jr, Morse PL, Burton LJ, Hammer PG 2d. Efficacy of infection control measures during a nosocomial outbreak of disseminated aspergillosis associated with hospital construction. J Infect Dis. 1986; 153:634-7.
            4. 4.↵
              Pandey A, Agrawal GP, Singh SM. Pathogenic fungi in soils of Jabalpur, India. Mycoses. 1990; 33:116-25.
            5. 5.↵
              Ali Shtayeh MS, Arda HM, Hassouna M, Shaheen SF. Keratinophilic fungi on sheep hairs from the West Bank of Jordan. Mycopathologia. 1996; 106:95-101.
            6. 6.↵
              Raper KB, Thom C. Manual of the Penicillia. Baltimore: Williams & Wilkins; 1949:284-8.
            7. 7.↵
              Castro LG, Salebian A, Sotto MN. Hyalohyphomycosis by Paecilomyces lilacinus in a renal transplant patient and a review of human Paecilomyces species infections. J Med Vet Mycol. 1990; 28:15-26.
            8. 8.↵
              Pettit TH, Olson RJ, Foos RY, Martin WJ. Fungal endophthalmitis following intraocular lens implantation. A surgical epidemic. Arch Ophthalmol. 1980; 98:1025-39.
            9. 9.↵
              Volna F, Maderova E. [Paecilomyces lilacinus—sensitivity to disinfectants.] Cesk Epidemiol Mikrobiol Imunol. 1990; 39:315-7.
            10. 10.↵
              Tan TQ, Ogden AK, Tillman J, Demmler GJ, Rinaldi MG. Paecilomyces lilacinus catheter-related fungemia in an immunocompromised pediatric patient. J Clin Microbiol. 1992; 30:2479-83.
            11. 11.↵
              Passweg J, Gratwohl A, Stauffer G, Wuhrmann J, Buser U, Tichelli A, et al. [Isolation ward: initial experiences after 4 years.] Schweiz Med Wochenschr. 1994; 124:62-8.
            12. 12.↵
              Speck B, Gratwohl A, Tichelli A, Nissen C. Cyclosporine in bone marrow transplantation. Transplant Proc. 1992; 24(4 Suppl 2):88-90.
            13. 13.↵
              Riley DK, Pavia AT, Beatty PG, Petersen FB, Spruance JL, Stokes R, et al. The prophylactic use of low-dose amphotericin B in bone marrow transplant patients. Am J Med. 1994; 97:509-14.
            14. 14.↵
              Petersen FB, Buckner CD, Appelbaum FR, Sanders JE, Bensinger WI, Storb R, et al. Etoposide, cyclophosphamide and fractionated total body irradiation as a preparative regimen for marrow transplantation in patients with advanced hematological malignancies: a phase I study. Bone Marrow Transplant. 1992; 10:83-8.
            15. 15.↵
              Sachs MK. The optimum use of needle aspiration in the bacteriologic diagnosis of cellulitis in adults. Arch Intern Med. 1990; 150:1907-12.
            16. 16.↵
              Merz WG, Roberts GD. Detection and recovery of fungi from clinical specimes. In: Murray PR, ed. Manual of Clinical Microbiology. 6th ed. Washington, DC: Amer Soc Microbiol Pr; 1995:709-22.
            17. 17.↵
              Doebbeling BN, Pfaller MA, Houston AK, Wenzel RP. Removal of nosocomial pathogens from the contaminated glove. Implications for glove reuse and handwashing. Ann Intern Med. 1988; 109:394-8.
            18. 18.↵
              Espinel-Ingroff A, Dawson K, Pfaller M, Anaissie E, Breslin B, Dixon D, et al. Comparative and collaborative evaluation of standardization of antifungal susceptibility testing for filamentous fungi. Antimicrob Agents Chemother. 1996; 39:314-9.
            19. 19.↵
              Zlmmerli W, Zarth A, Gratwohl A, Speck B. Neutrophil function and pyogenic infections in bone marrow transplant recipients. Blood. 1991; 77:393-9.
            20. 20.↵
              Bodey GP, Analssie E, Gutterman J, Vadhan-Raj S. Role of granulocyte-macrophage colony-stimulating factor as adjuvant therapy for fungal infection in patients with cancer. Clin Infect Dis. 1993; 17:705-7.
            21. 21.↵
              Supparatpinyo K, Chiewchanvit S, Hirunsri P, Uthammachai C, Nelson KE, Sirisanthana T. Penicillium marneffei infection in patients infected with human immunodeficiency virus. Clin Infect Dis. 1992; 14:871-4.
            22. 22.↵
              Nelson PE, Dignani MC, Analssie EJ. Taxonomy, biology, and clinical aspects of Fusarium species. Clin Microbiol Rev. 1994; 7:479-504.
            23. 23.↵
              Benedict LM, Kusne S, Torre-Cisneros J, Hunt SJ. Primary cutaneous fungal infection after solid-organ transplantation: report of five cases and review. Clin Infect Dis. 1992; 15:17-21.
            24. 24.↵
              Paya CV. Fungal infections in solid-organ transplantation. Clin Infect Dis. 1993; 16:677-88.
            25. 25.↵
              Bernacer M, Gadea I, Esteban J, Gegundez MI, Kamal K, Soriano F. Catheter-related fungemia due to Paecilomyces lilacinus in a leukemic child. Medical Microbiology Letters. 1992; 1:207-12.
            26. 26.↵
              Williamson PR, Kwon-Chung KJ, Gallin JI. Successful treatment of Paecilomyces varioti infection in a patient with chronic granulomatous disease and a review of Paecilomyces species infections. Clin Infect Dis. 1992; 14:1023-6.
            27. 27.↵
              Takayasu S, Akagi M, Shimizu Y. Cutaneous mycosis caused by Paecilomyces lilacinus. Arch Dermatol. 1977; 113:1687-90.
            28. 28.↵
              Abdel-Hafez AI, el-Sharouny HM. Keratinophilic and saprophytic fungi isolated from students' nails in Egypt. J Basic Microbiol. 1990; 30:3-11.
            29. 29.↵
              Vartivarian SE, Anaissie EJ, Bodey GP. Emerging fungal pathogens in immunocompromised patients: classification, diagnosis, and management. Clin Infect Dis. 1993; 17(Suppl 2):S487-91.
            30. 30.↵
              Wingard JR, Merz WG, Rinaldi MG, Johnson TR, Karp JE, Saral R. Increase in Candida krusei infection among patients with bone marrow transplantation and neutropenia treated prophylactically with fluconazole. N Engl J Med. 1991; 325:1274-7.
            31. 31.↵
              Gordon MA. Paecilomyces lilacinus (Thom) Samson, from systemic infection in an armadillo (Dasypus novemcinctus). J Med Vet Mycol. 1984; 22:109-16.
            32. 32.↵
              Jade KB, Lyons MF, Gnann JW Jr. Paecilomyces lilacinus cellulitis in an immunocompromised patient. Arch Dermatol. 1986; 122:1169-70.
            33. 33.↵
              Morse LJ, Williams HL, Grenn FP Jr, Eldridge EE, Rotta JR. Septicemia due to Klebsiella pneumoniae originating from a hand-cream dispenser. N Engl J Med. 1967; 277:472-3.
            « Previous | Next Article »Table of Contents

            This Article

            1. Ann Intern Med November 15, 1996 vol. 125 no. 10 799-806
            1. AbstractFREE
            2. Figures
            3. » Full Text
            4. Full Text (PDF)

            Rapid Responses

            1. Submit a response
            2. No responses published

            Navigate This Article

            Current Issue

            1. November 17, 2009, 151 (10)
            1. Alert me to current issues