Portal de Periódicos da CAPES

PRIMARY CUTANEOUS ASPERGILLOSIS IN VENTILATED NEONATES

2000; Lippincott Williams & Wilkins; Volume: 19; Issue: 5 Linguagem: Inglês

10.1097/00006454-200005000-00022

1532-0987

Farida Amod, Yacoob Coovadia, Thillagavathie Pillay, Graham Ducasse,

Fungal Infections and Studies

Primary cutaneous aspergillosis (PCA) is being reported with increased frequency and, in the overwhelming majority, in premature neonates. Of the 16 described cases 15 have been reported in the last decade, and all but 1 were premature babies.1-6 Association with the use of adhesive tape has been reported in 2 cases: one in relation to a chest tube; and the other in an area macerated by adhesive tape.5,66 We report the first 2 cases of primary cutaneous aspergillosis occurring on the face after the use of adhesive strapping to secure endotracheal tubes. Case reports.Case 1. A 900-g male, preterm second twin was admitted to the neonatal intensive care unit (NICU) in respiratory distress requiring ventilation. Chest roentgenogram showed bilateral pulmonary infiltrates. The differential diagnoses of hyaline membrane disease and acute bacterial sepsis were considered. Therapy was commenced empirically with penicillin and gentamicin. The hemoglobin was 15.7g/dl, white blood cells 15.5/nl and platelets 182/nl. Blood and cerebrospinal fluid cultures were sterile. He remained ill and pyrexial and developed sclerema and a progressive pneumonia. The antibiotics were changed to ceftazidime and subsequently to imipenem because nosocomial sepsis was suspected. On Day 5 of imipenem treatment (Day 17 in the nursery), the infant's overall condition was unchanged despite the use of broad spectrum antibiotics. It was also noted that he had developed ulcerated lesions with black eschars on the left cheek, side of nose and ear, beneath the adhesive tape used to secure the endotracheal tube. The diagnosis of disseminated candidiasis with embolic skin lesions was considered, and fluconazole (5 mg orally) was added. A swab of the lesion was taken and Aspergillus niger was isolated in pure culture. Therapy was amended to amphotericin B deoxycholate in an initial dose of 0.25 mg/kg/day and increased at daily increments of 0.125 mg with the aim of reaching 1 mg/kg/day. Further investigations to exclude disseminated disease (endotracheal aspirates and blood cultures, processed via a lysis-centrifugation method) were negative. The patient remained ventilator-dependent and scleremic with a nonresolving pneumonia. He died 6 days after commencement of amphotericin B. The parents refused an autopsy. Case 2. A term, female newborn with Apgar scores of 3, 4 and 5 at 1, 5 and 10 min, respectively, required ventilatory and inotrope support soon after birth. Hypoxic ischemic encephalopathy and neonatal sepsis were suspected. Therapy was empirically commenced with penicillin and gentamicin. Specimens of gastric aspirate, blood culture, cerebrospinal fluid and ear swab sent for bacterial culture were negative. By Day 3 the infant had developed pancytopenia (hemoglobin 5.6 g/dl, platelets 10/nl, leukocyte count 2.4/nl with 56% neutrophils) with an abnormal coagulation profile (INR 8.7, partial thromboplastin time >180 s, fibrin degradation products >2000 ng/ml) consistent with worsening sepsis. Therapy was changed to cefotaxime. On Day 10 in the NICU the neonate developed three distinct lesions on the left neck, right cheek and lip beneath the adhesive tape securing the endotracheal tube. These began as hemorrhagic bullae that rapidly evolved into ulcers with indurated edges and central necrotic bases. A clinical diagnosis of ecthyma gangrenosum was made. Swab of the cheek lesion yielded a pure culture of A. flavus. Histology of the subsequent biopsy revealed ulceration with necrotizing acute inflammation. No specific pathogens were demonstrated and the periodic acid-Schiff stain for fungal elements was negative. Culture of the biopsy, however, grew A. flavus. The chest roentgenogram showed a progressive bilateral bronchopneumonia, the possible etiology of which included bacteria, viruses or systemic dissemination of PCA. Further investigations (endotracheal tube aspirates, blood) of cultures to exclude these were negative. Fontanel ultrasonography revealed a frontal fossa hemorrhage presumed to be related to the bleeding diathesis and/or severe birth asphyxia. Owing to the overall poor prognosis antifungal therapy was not commenced, and antibiotics were discontinued. She died 4 days after the onset of PCA. The infant's parents did not consent to an autopsy. In both patients the HIV antibody tests were negative. Glucocorticoid therapy had not been administered antenatally or during the course of their illness. Both infants appeared to have inflammatory conditions, possibly caused by infection, that progressively worsened from the time of birth. PCA in this context most likely represents an epiphenomenon. The cases were temporally unrelated and occurred in NICUs in two hospitals. Culture of samples of the tapes used in the NICUs at the respective times did not yield Aspergillus, and the environmental sources were not established. Discussion. Cutaneous aspergillosis can be a site-specific entity or the initial focus for fatal disseminated disease.2,68 In our cases we failed to retrieve the organism from extracutaneous sites (lung, blood). The yield of blood cultures in disseminated disease, however, is low,1 even with blind terminal subculture and lysis-centrifugation. Thus invasive biopsies are ultimately required to establish the diagnosis. In the absence of these and autopsy findings we were unable to reliably exclude the possibility of disseminated invasive disease. Preterm infants have an almost unique predisposition to neonatal PCA. This may be ascribed to the increased vulnerability of their skin to minor trauma associated with intensive care as well as immaturity of host defenses. Of the 16 reported cases 15 have been described in premature neonates. The first patient in this report was preterm but the second was a term baby. To our knowledge this is the second reported case of PCA occurring in a term baby. Other recognized risk factors for primary cutaneous aspergillosis include neutropenia, glucocorticoid administration and prior use of antibiotics.4,67 Although neutropenia appears to be an important risk factor in certain immunocompromised groups, including bone marrow transplant patients and those with hematologic malignancies, our patients were not neutropenic. This observation is in keeping with that of Groll et al.1 who noted neutropenia to be an infrequent occurrence in neonatal aspergillosis. Instead a qualitative defect in neutrophil chemotaxis and phagocytosis is thought to predispose this group of patients to aspergillosis.8 Glucocorticoids, on the other hand, have been shown to impair macrophage and neutrophil killing of Aspergillus spores and hyphae, respectively.8 Neither of our patients had received glucocorticoids during the course of their illness. The role of antibiotics in the pathogenesis of PCA is unclear. Of the 16 cases of neonatal PCA reported thus far, 13 had previously received antibiotics.1,62 Likewise our patients received prolonged broad spectrum antibiotics (>10 days). It may be that systemic antibiotics contribute to the disturbance in the ecology of the skin flora.4,67 Mechanical disruption of the barrier function of the skin is involved in all cases of PCA.8 In one-half of all reported cases of neonatal PCA, the source was not determined. In the remaining eight cases, these included a contaminated hand splint (one),3 macerated skin associated with an oxymeter sensor (one),9 contaminated latex finger stalls (four)2 and adhesive tape (two).5,66 The skin of newborns is exceedingly fragile,8 especially on the face, and minor friction can thereby predispose to a breach in the epithelium allowing Aspergillus conidia to enter. Fragility of skin in the newborn might explain the lack of reports of PCA localized to the face in high risk adult populations requiring prolonged intubation. We, however, cannot explain why cases similar to ours have not been reported previously, given the large number of premature neonates who have endotracheal tubes secured by adhesive tape. It is likely that more such cases will be described in the future. The lesions of PCA may be nonspecific, ranging from induration to pustules, hemorrhagic bullae, ulcers, small abscesses and eschars.1,67 Thus any new skin lesion in a neonate at risk should raise the suspicion of PCA. Confirmation requires histology and culture of a biopsy taken from the center of the lesion.7 Because neonates tend not to tolerate skin surgery well, treatment entails antifungal therapy often without surgical excision.8 Amphotericin B is the recommended choice with the adjunctive use of flucytosine being controversial but widely practiced. Limited reports exist on the use of itraconazole in neonates, but it may have a place in localized PCA, not related to catheter exit sites.7 In contrast to disseminated invasive aspergillosis, the prognosis for PCA is favorable probably owing to easier and earlier diagnosis and treatment.1 These are the first two reported cases of PCA occurring on the face after the use of adhesive tape to secure endotracheal tubes. Although neonatal PCA is relatively uncommon, early recognition and treatment require a high index of suspicion in patients at risk. Farida C. Amod, M.B., Ch.B. Yacoob M. Coovadia, M.B., Ch.B., F.F.Path. Thillagavathie Pillay, M.B., Ch.B., F.C.P.(Paed.) Graham Ducasse, M.B., Ch.B., F.C.P.(Paed.) Departments of Microbiology (FCAm YMC) and Paediatrics (TP, GD); University of Natal; Medical School; Congella, South Africa