Fatal Rhinoorbital Mucormycosis Caused by Saksenaea Vasiformis in an Immunocompromised Child

Mucormycosis is a term used for infections caused by saprophytic fungi of the class Zygomycetes, order Mucorales. The most common clinical forms are rhinocerebral, pulmonary, cutaneous and disseminated infections. Risk factors include immunosuppression, diabetes mellitus, therapy with deferoxamine, renal failure, burns, trauma and malnutrition.1-3 Rhinocerebral infection is the most distinctive form of mucormycosis. It begins in the nasal sinuses and extends to the paranasal sinuses, before spreading to involve the face, orbit and brain.1, 4 Rhinoorbital and paranasal mucormycosis appear to represent early stages of rhinocerebral disease.1, 4Rhizopus arrhizus is responsible for 90% of cases of rhinocerebral infection.5 We report here the first case in a child of rhinoorbital (and probable rhinocerebral) infection caused by Saksenaea vasiformis. Case report. A 5 1/2-year-old girl was admitted with periorbital cellulitis. She had a past history of acute lymphoblastic leukemia, which had been diagnosed 14 months previously. Having initially achieved remission she relapsed with bone marrow disease 2 weeks before this presentation and induction chemotherapy was recommenced, which led to profound neutropenia. Her initial presentation was to her local doctor with a 24-h history of worsening headache, fever and a painful, itchy right eye. The eye then became swollen and increasingly painful. She was given intravenous ceftriaxone and was transferred to the Royal Children's Hospital. On admission she was pale and unwell, with a temperature of 38.9°C. There was marked edema and erythema of the right eyelid and infraorbital tissue, extending onto the right side of the nose. The eye was almost closed with swelling and there was chemosis of the bulbar conjunctiva laterally. The rest of the examination was normal. The white blood cell count was 0.1 × 109/1 (100% neutrophils); hemoglobin was 75 g/l; platelet count was 107 × 109/l; and the erythrocyte sedimentation rate was 80 mm in the first hour. Blood culture yielded no growth. Computerized tomography of the orbit and sinuses showed right upper cheek and preseptal lower lid cellulitis, associated with extensive right anterior ethmoidal and bilateral maxillary sinus opacification. Intravenous flucloxacillin was added to the ceftriaxone, and she was transfused with 1 unit of irradiated packed red blood cells. During the next 48 h she remained febrile and unwell; the area of swelling and erythema extended to the left periorbital area. The skin of the inner canthus of the right eye had become necrotic, and a black, necrotic lesion was noted on the hard palate. At this stage her treatment was changed to imipenem and amphotericin B. Examination of her eyes under anesthetic revealed bilateral extensive chemosis. Her maxillary and ethmoid sinuses were drained, and biopsies were taken of necrotic periorbital skin and palate as well as from the ethmoid sinus and the middle nasal meatus. Pathologic examination of these samples showed a fibrinous exudate with scattered polymorphonuclear cells and histiocytes and multiple fragments of nonseptate fungal hyphae with irregular diameter and branching, intermixed with necrotic cells. There was evidence of angioinvasion by fungal hyphae in sections of palatal tissue. The morphology of the hyphae was suggestive of mucormycosis. A repeat computerized tomography scan showed extension of the disease process but did not include views of the brain. Because of the site and extensive nature of the disease, it was felt that curative surgery would be technically difficult, if not impossible. This was discussed with the parents, who requested that she be given supportive treatment only. Treatment with amphotericin B and imipenem was maintained, but the disease progressed rapidly. Her fever continued and her conscious state declined. She died 7 days after her initial presentation. An autopsy was not performed. Mycology. Palate, facial skin, maxilla, eye and ethmoid sinus tissue were submitted for microbiologic investigation. Tissues were mounted in a KOH-Quink® ink preparation (70% 1.78 M KOH, 20% glycerol, 10% Quink® ink; Parker Pen Co., Newhaven, UK). Fungal elements were observed in direct preparations of all tissues except the maxilla. Large ribbon-like nonseptate hyphae up to 12.5 μm in diameter were seen. Whole tissue was cultured on bacteriologic media and Sabouraud's dextrose agar, incubated at 37°C. Within 24 h a white mycelial growth had appeared on all the plates from all of the specimens. Because the fungus failed to sporulate on Sabouraud's dextrose agar it was referred to the Microbiological Diagnostic Unit, University of Melbourne, for identification. Sporulation is essential for correct identification and S. vasiformis usually requires special conditions for sporulation to occur.6 Sporulation was eventually induced by floating blocks of Sabouraud's dextrose agar containing mycelial growth in sterile distilled water enriched with 0.2 ml of 10% filter-sterilized yeast extract and incubating at 37°C in the dark.6 The finding of flask-shaped sporangia with globular columellae arising from dichotomously branched darkly pigmented rhizoids allowed the fungus to be identified as S. vasiformis(Fig. 1).Fig. 1: Characteristic flask-shaped sporangium of S. vasiformis. ×400.Discussion.S. vasiformis is an infrequent cause of mucormycosis. It usually causes cutaneous lesions after traumatic implantation. Thus infection generally occurs in previously healthy people who have sustained traumatic injury, such as head7 or scalp trauma,8 burns,9 insertion of an arterial catheter10 or a tattoo.11 All such patients have recovered after surgical debridement and treatment with amphotericin B or removal of the arterial catheter. One case of localized cutaneous mucormycosis in a 3-month-old healthy girl was cured by surgical debridement alone.12 The intact immune response in these patients was probably a major factor in determining a successful outcome. Few fatal infections with S. vasiformis have been reported. One such case was an immunocompromised patient with preleukemic syndrome13; another was a previously healthy child.14 Both patients presented with necrotic skin lesions, although it is not clear whether these were primary lesions or a manifestation of disseminated infection. Fatal disseminated infection with this fungus has also occurred after a cutaneous mucormycotic lesion at an intramuscular injection site. The lesions in this immunocompetent woman progressed despite surgical and medical treatment.15 Disseminated mucormycosis, regardless of the etiologic agent, is a rapidly fatal disease with a mortality rate of 98%.16 Rarely has rhinocerebral or rhinoorbital mucormycosis been attributed to a fungus other than R. arrhizus.5Apophysomyces elegans was responsible for rhinocerebral mucormycosis in a previously healthy man.17Cunninghamella bertholletiae caused a fatal rhinocerebral infection in a diabetic with sideroblastic anemia and hemachromatosis, who was receiving deferoxamine therapy.18 Only one previous case of rhinocerebral infection with S. vasiformis has been described. This was in an immunocompetent man who died despite surgical debridement and therapy with amphotericin B.19 To our knowledge this case represents the first of confirmed rhinoorbital (and probable rhinocerebral) mucormycosis caused by S. vasiformis in a child. Although we did not conclusively establish that this patient had cerebral involvement, this seems highly likely given the rapid progression and fatal outcome. Rhinocerebral mucormycosis has been described previously in patients with leukemia, most of whom had coexisting diabetes or were receiving treatment with cytotoxic drugs.1 Our patient was neutropenic as a result of disease relapse and chemotherapy. Her persistent neutropenia was probably the main factor contributing to the rapid progression of her infection, despite early diagnosis and treatment with amphotericin B. In this patient, as in others, failure to reverse the underlying condition was probably the key factor leading to a fatal outcome. Acknowledgments. We gratefully acknowledge the support of Alan Woodger from the Microbiological Diagnostic Unit, for identifying the fungus and providing slide cultures of it, and of Professor Roy Robins-Browne for reviewing the manuscript. Gena Gonis, B.Sc. Hons. Michael Starr, M.B., B.S.