Whilst the Aspergillus Website & the Fungal Research Trust have taken every precaution in compiling this site, neither it nor any contributors or charity Trustees can be held responsible for using the information held herein. Medical information changes constantly and the information presented here should not be considered complete or exhaustive and should not be relied upon as the sole or primary source on which to base diagnosis or treatment for any individual.

This site is not designed to and does not provide medical advice, professional diagnosis, opinion, treatment or services to you or to any other individual. Through this site and links to other sites, the Aspergillus Website provides general information for educational purposes only. The information provided in this site, or through links to other sites, is not a substitute for medical or professional care, and you should not use the information in place of a visit, call, consultation or the advice of your physician or other healthcare provider.

Neither The Aspergillus Website nor the FRT is liable or responsible for any advice, course of treatment, diagnosis or any other information, services or product you obtain through this site.

I have read this, proceed to the article

Table of Contents

• Authors
• Overview
• Predisposition of the neonate
  • Immunological
  • Mechanical
• Clinical Diseases
  • Primary cutaneous aspergillosis
  • Invasive pulmonary aspergillosis
  • Aspergillosis of the central nervous system
  • Aspergillosis of the gastrointestinal tract
• Miscellaneous single-site aspergillosis
  • Disseminated invasive aspergillosis
• Recognition and Diagnosis
• Approaches to treatment
• Prevention
• References

Overview

As of July of 2000, 53 cases of proven (cultural and histological evidence) or probable (histological or cultural evidence) invasive aspergillosis have been reported in neonates and infants ≤3 months of age [Groll et al. 1998; Meessen et al. 1998; Singer et al. 1998; Reddy et al. 1999; Amod et al. 2000; van Landeghem et al. 2000]. Most patients had primary cutaneous aspergillosis (n=18), invasive pulmonary aspergillosis (n=10), or disseminated aspergillosis (n=14). Prematurity, chronic granulomatous disease, and a complex of diarrhea, dehydration, malnutrition and invasive bacterial infections accounted for the majority of underlying conditions. Routes of infection, clinical symptoms, fungal isolates and patterns of disease were similar to those observed pediatric [Walmsley et al. 93] and adult [Bodey & Vartivarian 1989; Denning 1996] populations. Only one of the reported infants had been neutropenic prior to diagnosis, but at least half had been given corticosteroids in pharmacological doses prior to diagnosis. Almost all cases of primary cutaneous aspergillosis occurred in extremely premature neonates; pulmonary and disseminated infection mainly occurred in term neonates. The disease was uniformly fatal in untreated patients. In contrast, outcome was relatively favorable in patients receiving appropriate systemic antifungal and/or surgical therapy [Groll et al. 1998; Singer et al. 1998; Reddy et al. 1999].

Predisposition of the neonate

Immunological Prolonged and profound neutropenia and administration of high doses of corticosteroids are the most important clinical risk factors for development of invasive aspergillosis [Bodey & Vartivarian 1989; Denning 1996]. While neutropenia was reported in only one single patient, functional impairment of phagocytosis through either prematurity, a diagnosis of chronic granulomatous disease, or therapy with corticosteroids was present in 80% of the reported cases [Groll et al. 1998; Meessen et al. 1998; Singer et al. 1998].

Mono- and polymorphonuclear phagocytes of term- and in particular preterm neonates have diminished chemotactic, phagocytic and microbicidal activity, especially under conditions of severe illness and stress [Lewis & Wilson 1995]. In comparison to this increased general susceptibility to invasive bacterial and also, fungal infections, chronic granulomatous disease (CGD) of childhood is a congenital disorder that is characterized by the inability of phagocytic cells to provide antimicrobial oxidants and intracellular killing [Quie 1993]. CGD is associated with a cumulative lifetime incidence of invasive aspergillosis of 16 to 40% [Cohen et al. 1981; Mouy et al. 1989], and in the 1990s, invasive aspergillosis has emerged as leading cause of mortality in patients with this disorder [Hasui 1999; Winkelstein et al. 2000].  That corticosteroids are an important risk factor for invasive fungal infections in the neonatal setting has been demonstrated for preterm infants with invasive forms of candidiasis [Botas et al. 1995; Rowen et al. 1995]. Indeed, neonatal mononuclear phagocytes may have an increased sensitivity to the effects of corticosteroids Lortie et al. 1990; Kavelaars et al. 1995]. In addition, pharmacokinetic aspects such as a comparatively lower systemic clearance of dexamethasone [Jusko & Ludwig 1992; Charles et al. 1993], low serum albumin and liver function abnormalities, both commonly encountered in neonatal intensive care, can all lead to an increased systemic exposure to the drug [Jusko & Ludwig 1992].
Back to top

Mechanical. Preceding local trauma appears to be a prerequisite for primary cutaneous aspergillosis [Walsh 1998]. Immature barrier functions of the skin, reduced thickness of the epidermis and maceration by tape, armboards,  cutaneous sensors, and other devices may be responsible for most cases of primary cutaneous aspergillosis in very immature neonates of ≤ 3 weeks of postgestational age [Rowen et al. 1995; Meesen et al. 1998; Singer et al. 1998; Amod et al. 2000].
Back to top

Clinical Diseases

Primary cutaneous aspergillosis  [Granstein et al. 1980; Roth et al. 1991; Rowen et al. 1992; Lackner et al. 1992; Perzigian et al. 1993; van den Anker et al. 1993; Walsmley et al. 1993; Rowen et al. 1995; Papouli et al. 1995; Gupta et al. 1996; Meessen et al. 1998; Singer et al. 1998; Amod et al. 2000]. Primary cutaneous aspergillosis is a disease of very premature infants and tends to occur early during the postpartum period: Sixteen of the 18 reported patients were premature neonates with a mean birth weight of 766g (range, 440-1500g), and the age at the time of diagnosis ranged from 3 days to 30 days (mean, 11 days). Prior exposure to corticosteroids was documented in 8 cases. None of the patients was neutropenic (absolute neutrophil count ≤500/µL) or had a documented immunodeficiency disorder [Groll et al. 1998; Meessen et al. 1998; Singer et al. 1998; Amod et al. 2000].

In all but 2 cases, PCA was diagnosed during lifetime. The most frequently encountered species was A.fumigatus (n=10), followed by A.flavus (n=5), A.niger (n=2) and Aspergillus spp. (n=1 case). The clinical presentations included single or multiple plaques, papules, pustular lesions, small abscesses, focal and spreading skin necroses, and crusting ulcerations. In 4 cases, skin lesions were located on the infant's back, and thus probably related to prolonged positioning as part of the minimal handling approach to the care of these highly vulnerable infants [Groll et al. 1998]. In a series of 4 extremely immature male neonates, skin lesions started from the penis and the perigenital area, and were ultimately due to the utilization of contaminated fingerstalls for collection of urine specimens [Singer et al. 1998]. In two further cases, skin lesions originated from facial and nuchal macerations caused by tape serving to secure the infants endotracheal tubes [Amod et al. 2000]. Other documented sites of entry included extracranial skin defects from adhesive tape (n=2), an oxymeter sensor, or a contaminated tongue depressor used for fixation of intravenous catheters [Groll et al. 1998; Meessen et al. 1998; Singer et al. 1998].

In two cases, no information was available regarding treatment and ultimate outcome. One patient underwent complete excision of the lesion and died later from unrelated complications with no evidence for aspergillosis at autopsy. Nine patients received systemic antifungal treatment with intravenous amphotericin B  ± 5-flucytosine, and 1 patient was treated with systemic antifungals and surgical incision of the lesion. Seven of these patients were cured and eventually survived. Among the remaining three patients, one had Aspergillus-endophthalmitis diagnosed 3 weeks after apparently successful completion of a 4-weeks-course of amphotericin B (AmB), one died with PCA few days after initiation of antifungal therapy from refractory pneumonitis of unknown etiology [Amod et al. 2000], and one died with PCA from unrelated causes without evidence for dissemination at postmortem. One patient, who was treated with 5-flucytosine alone, died early after initiation of antifungal therapy from multiorgan failure and had disseminated aspergillosis at autopsy. In two cases, the diagnosis was made antemortem, but therapy was withheld due to the overall extremely poor prognosis of these infants. In two further patients, the diagnosis of PCA was established only at autopsy, and both had disseminated disease at postmortem examination [Groll et al. 1998; Meessen et al. 1998; Singer et al. 1998; Amod et al. 2000].
Back to top

Invasive pulmonary aspergillosis [Brass 1975; Burger et al. 1978; Krasinski et al. 1985; Mouy et al. 1985; Schoumacher et al. 1987; Gonzalez et al. 1994; Mouy et al. 1995; Groll et al. 1998]. In contrast to PCA, eight of the 10 reported neonates and infants with invasive pulmonary aspergillosis (IPA) were delivered at term. Five had documented, and one probable CGD, and 2 patients had diarrhea, dehydration and invasive bacterial infections as their underlying condition. The mean age at diagnosis was 31 days (range, 01-57 days). At least half of the patients had community-acquired disease. The mean duration of prior hospitalization in the cases considered as hospital acquired was 19 days (range, 01-45 days). Prior exposure to corticosteroids was reported in 3 cases. None of the patients had been neutropenic; instead, 4 presented with marked leukocytosis [Groll et al. 1998].

Similar to other settings, A.fumigatus accounted for the majority of cultured isolates. The clinical presentation of IPA consisted of lower respiratory tract symptoms and/or new pulmonary infiltrates on chest x-ray. Of note, fever was an inconsistent feature.  At least 1 patient had cardiovascular instability, and 1 patient presented with massive diffuse pulmonary hemorrhage. In retrospect, the presumable portal of entry was the respiratory tract in all cases. Construction work at or nearby the hospital was reported in 2 circumstances [Groll et al. 1998].

In 4 of the reported cases, IPA was diagnosed only at autopsy, and no antifungal treatment was given. In 3 of those cases, IPA was an incidental finding and stood in no relationship to the cause of death. The fourth patient died from suffocation by pulmonary mass bleeding in the presence of overwhelming necrotizing Aspergillus infection of the lung. Six of the 10 patients received medical treatment alone; one of these patients died after eight days of AmB therapy with bilateral Aspergillus pneumonia and cerebral venous thrombosis without evidence for fungal involvement. The remaining 5 patients, representing the cases with proven CGD, responded to prolonged medical treatment, and at least 4 of them were cured and survived [Groll et al. 1998].
Back to top

 Aspergillosis of the central nervous system  [Wittig et al. 1973; Rhine et al. 1986; Green et al. 1991; Rowen et al. 1992; van Landeghem et al. 2000]. Five cases of isolated Aspergillus infection of the central nervous system (CNS) in the first three months of life are reported. The lung was the most likely portal of entry in all but one case with aplasia cutis of the scalp, in whom the organism may have entered through the integumental defect, the respiratory tract or during open heart surgery [van Landeghem et al. 2000]. Underlying conditions were prematurity, immunosuppression after liver transplantation, staphylococcal pneumonia and septicemia, and heart malformations with complex cardiac surgery and infectious complications (2 cases). The infection was hospital acquired in 4 cases, and the mean age at diagnosis was 44 days (range, 17 to 84 days). While none of the infants had been neutropenic, four had received corticosteroids with additional cyclosporin A and anti-OKT3 antibodies in one case [Groll et al. 1998; van Landeghem et al. 2000].

Instability, fever, seizures, culture-negative but unspecifically abnormal cerebral spinal fluid (CSF) findings, ring enhancing lesions and hydrocephalus internus were the presenting symptoms. In 3 patients, diagnosis was established by needle aspiration; A.fumigatus was cultured in two and A.flavus in one case. With prolonged systemic, intralesional and/or intraventricular antifungal treatment and surgical drainage or excision, all 3 patients were ultimately cured and survived, although with considerable neurologic sequelae. One patient with aplasia cutis of the scalp and status post open heart surgery for hypoplastic left heart syndrome [van Landeghem et al. 2000] developed hydrocephalus internus, meningitis and multiple brain abcesses. Diagnosis was established by CSF culture that grew A.fumigatus. Despite a partial response to treatment with liposomal amphotericin B (4mg/kg) plus flucytosine and amphotericin B instillations after shunt placement, the patient died 52 days post diagnosis from autopsy proven cerebral aspergillosis. In the remaining patient [Wittig et al. 1973], who had widespread invasive staphylococcal infection and possibly CGD as underlying disorder, diffuse Aspergillus- meningoencephalitis was an incidental autopsy finding [Groll et al. 1998].
Back to top

Aspergillosis of the gastrointestinal tract  [Steiner et al. 1997; Bruyere et al. 1983; Rowen et al. 1992]. Two of the 3 reported cases of gastrointestinal aspergillosis occurred in preterm neonates. In the third case, although not formally diagnosed, a congenital T-cellular immunodeficiency disorder appears as the most likely retrospective diagnosis.

The exclusive clinical presentation was intestinal perforation, and the presumable portal of entry was the gastrointestinal tract. In one patient, the small bowel was affected, and in another, the stomach. One additional patient [Bruyere et al. 1983] with necrotizing enterocolitis had Aspergillus niger cultured from stool and the peritoneal cavity, but no histologic evidence was provided. All cases were hospital acquired and diagnosed during lifetime at a mean age of 14 days (range, 4-32 days).  One patient had been exposed to corticosteroids.

While the patient with gastric perforation did not survive diagnostic surgery, the second patient was cured and survived after resection of necrotic intestinal tissue and prolonged treatment with AmB. The third patient died from unrelated causes after resection of necrotic bowel and topical antifungal treatment only; autopsy revealed no evidence for invasive aspergillosis [Groll et al. 1998].
Back to top

Miscellaneous single-site aspergillosis.  Sihota et al. (Sihota et al. 1987) reported a case of biopsy proven bilateral Aspergillus - endophthalmitis in an otherwise healthy boy of 30 days. He was reportedly cured and survived after intravenous AmB treatment and enucleation of the left eye. Reddy and colleagues (Reddy et al. 1999) described the case of a 10-days-old, otherwise healty male neonate who presented with ethmoiditis and a left sided orbital abscess. While cultures obtained during  external ethmoidectomy and abscess drainage grew Staphylococcus aureus, microscopical examination of biopsy material revealed dichotomously branching hyphae suggestive of Aspergillus spp. that invaded the bone. The infant was cured without sequelae by treatment with antibacterial agents and a 21-day course of intravenous amphotericin B deoxycholate. Rhaghavan et al. (Rhaghavan et al. 1987), in an autopsy survey, list a case of isolated Aspergillus - myocarditis detected in a child of 42 days of age who had been hospitalized for diarrhea and bacterial bloodstream infection.
Back to top

Disseminated invasive aspergillosis [Zimmermann 1955; Akkoyunlu et al. 1957; Allen & Andersen 1960; Matturi & Fasoli 1962; Luke et al. 1963; Paradis & Roberts 1963; Tan et al. 1966; Raaf et al. 1977 and Case Records of MGH 1976; Mangurten &Fernandez 1979; Gonzalez-Crussi et al. 1979; Schwartz et al. 1988; Pereira et al. 1989; Rowen et al. 1992]. Interestingly, the majority of cases of disseminated invasive aspergillosis (DIA) have been reported before 1980. Similar to IPA, most patients (n=10/14) were born at term. The mean age at diagnosis was 32.5 days (range, 16-73 days) [Groll et al. 1998].

Prematurity was the underlying condition in 4 cases, combined with cytomegaloviral pneumonitis in 1 case. One patient had confirmed CGD and 1 newborn had congenital myeloblastic leukemia treated with antineoplastic chemotherapy. Two patients had acute hepatic failure of unclear etiology. The remainder initially presented with diarrhea and dehydration, failure to thrive, invasive bacterial infections, or a combination thereof. In 3 of these latter patients, the reported findings are in retrospect highly suspicious for CGD as underlying illness. In most patients with sufficient information, the disease was acquired in the hospital; however, there were 3 cases where the infection was clearly community acquired. Exposure to corticosteroids prior to diagnosis was reported in 6 (43%) cases. One patient was deeply neutropenic for prolonged periods of time. The mean length of hospitalization prior to diagnosis was 22.5 days (range, 0-51 days). Construction work was not documented in any of the reports [Groll et al. 1998].

In only 1 patient, the diagnosis of invasive aspergillosis had been entertained during lifetime. The most common clinical symptoms were respiratory and/or central nervous system compromise in 8 patients. Two patients initially presented with small bowel obstruction or perforation, and 2 patients had signs and symptoms of severe hepatitis. Of note, in 2 instances, Aspergillus spp. was isolated from blood cultures (10 days prior to death; during (unsuccessful) cardiopulmonary resuscitation). Cultures from CSF were negative in all 3 cases with CNS-involvement [Groll et al. 1998].

All 14 patients with DIA died, and aspergillosis was related to the cause of death in all instances. None of the patients had received any form of medical or surgical treatment. In all cases, invasive aspergillosis was widely disseminated at autopsy, and in most cases, the respiratory tract was the most likely portal of entry. Pulmonary involvement was reported in 85%, followed by the brain (8/11 patients examined, 73%), the  heart (62%), kidneys and bowels (54% each), liver (46%), thyroid gland (31%), spleen (23%) and various other organs. Infective endocarditis was found in 23%, as was meningitis. Aspergillus endophthalmitis was described in only 1 case (8%). In all patients, the infection was found in at least 3 anatomically distinct sites. In the cases were Aspergillus had been documented microbiologically, A.fumigatus was the predominant organism (n=5), followed by A.sydowi, A.flavus, and A.niger (one case each) [Groll et al. 1998]. 
Back to top

Recognition and Diagnosis

There are no distinct clinical symptoms or specific radiographic signs of neonatal aspergillosis. Similarly, even in widely disseminated disease, blood cultures were only exceptionally positive, and cultures from cerebrospinal fluid were negative in all but one of 8 infants with CNS-involvement [Groll et al. 1998].

Nevertheless, invasive aspergillosis should be considered as differential diagnosis in the presence of all of the following: An apparently infected skin lesions, in particular in a very low birth weight infant; infarct-like lung lesions and pulmonary hemoptysis; intestinal infarction or perforation; and with persistent signs of infection despite antibiotic therapy and negative blood cultures, especially when there is a combination of pulmonary and cerebral findings [Schwartz et al. 1988; Rowen et al. 1992; Papouli et al. 1996]. Isolation of Aspergillus spp in these circumstances should be regarded as proof of infection unless otherwise excluded, but any culture positive for Aspergillus must be considered seriously in the neonate. Ultimately, invasive procedures and bioptic evidence may be required to establish the diagnosis or to avoid unnecessary, potentially toxic treatment. If a diagnosis of invasive aspergillosis has been made, dissemination to the lungs (if not primarily involved), CNS and parenchymatous organs should be investigated and a congenital phagocytic disorder, primarily CGD, should be ruled out.
Back to top

Approaches to treatment

High-dose (1.0-1.5mg/kg/day) amphotericin B deoxycholate (D-AmB; Fungizone™) remains the cornerstone of treatment for both suspected and proven invasive aspergillosis. In neonates, the pharmacokinetics of D-AmB are characterized by an extreme interindividual variability and a comparatively lower clearance rate [Starke et al. 1987; Koren et al. 1988; Baley et al. 1990]. However, no correlations between plasma concentrations and pharmacological effects have ever been established, and there is no evidence that these pharmacokinetic characteristics result in any clinical consequences. Indeed, several case series indicate that D-AmB is usually tolerated without nephrotoxicity at daily dosages of up to 1mg/kg/day, even in very low birth weight infants of ≤ 1500 g [Faix et al. 1984; Johnson et al. 1984; Starke et al. 1987; Baley et al. 1990; Butler et al.  1992; Leibovitz et al.  1992; Glick et al. 1993; Schloesser et al. 1995; Kingo et al.  1997], and daily dosages of up to 1.5mg/kg have been safely administered [Butler & Baker 1988]. In view of the dismal prognosis of invasive Aspergillus infections, we recommend a starting dosage of at least 1 mg/kg and day. Due to its potential to cause cardiac arrhythmia’s, D-AmB should be infused under careful monitoring, in particular in the presence of hyperkalemia and/or renal impairment. [Walsh et al. 1996; Googe & Walterspiel 1988]. Adequate renal blood flow should be maintained and serum levels of co-administered aminoglycosides, vancomycin and flucytosine be monitored [Koren et al. 1988; Goren et al. 1988].

Combination of D-AmB with 5-flucytosin is controversial and has not been systematically investigated. However, the combination has been used in cases with successful outcome and may be appropriate, in particular in cases with CNS involvement [Walsh et al. 1996; Denning & Stevens 1990]. A starting dose of 100mg/kg/day and consecutive adjustment of the dosage to maintain serum levels below 100 µg/mL is recommended [Francis & Walsh 1992; van den Anker et al. 1995].

Liposomal AmB (L-AmB; AmBisome™) has been tolerated at dosages of up to 7mg/kg /d in approximately 100 term and preterm infants [Groll et al. 1998; Scarcella et al. 1998; Weitkamp et al. 1998; Leibovitz et al. 2000]. Published data on other lipid-based formulations of AmB is more limited: Eleven infants between 3 and 13 weeks of age and weighing from 0.8 to 5 kg received amphotericin B lipid complex (ABLC; Abelcet™) at varying dosages without limiting adverse effects and toxicity [Walsh et al. 1999]. Similarly, amphotericin B colloidal dispersion (ABCD; Amphotec™; Amphocil™) was well tolerated by 16 premature infants who received the drug at dosages of 5 mg/kg and day [Linder et al. 2000]. Because of the larger number of patients treated safely with this agent, we consider L-AmB as current salvage agent of choice in neonates refractory to or intolerant of D-AmB. Based on animal data [Francis et al. 1994] and data from randomized studies in adults with D-AmB as comparator [Prentice et al. 1997; Walsh et al. 1999], we recommend a starting dose of 5 mg/kg/day for the treatment of neonatal aspergillosis with dosage adjustment only for limiting toxicity. 

Itraconazole, while not being a choice for initial treatment due to its variable bioavailability, may be indicated in a stable patient with residual lesions who is able to tolerate oral medication. The drug can be absorbed by preterm neonates and be effective in treating fungal infections [Bhandari & Narang 1992; van den Anker 1992; van den Anker et al. 1992]. Published data on the safety and tolerance of the investigational broad-spectrum antifungal triazoles voriconazole, posaconazole, and ravuconazole does not exist, and their advantage over itraconazole for treating invasive aspergillosis remains to be defined. Notably, because of yet unresolved concerns of antagonism [Sugar 1995], at present, antifungal azoles should not be combined with amphotericin B products for treatment of invasive aspergillosis.

Independent of cumulative dosage and agent, effective antifungal treatment must be administered until the complete resolution of all lesions and reversal of the principal underlying deficiencies in host defenses [Denning 1996; Walsh et al. 1996]. Restoration of host defenses is paramount and includes the discontinuation of corticosteroids, if feasible, and treatment with G-CSF or GM-CSF in patients with diminished neutrophil counts to achieve normal levels of circulating granulocytes. Finally, adjunctive surgical treatment is considered essential in Aspergillus endophthalmitis, endocarditis and peritonitis, and should be strongly entertained in lesions of skin and subcutaneous tissues and in Aspergillus infections of the CNS [Rowen et al. 1992; Denning 1996].
Back to top

Prevention

Infections by Aspergillus spp. can occur in neonates and young infants and cause diseases that are life-threatening but potentially amenable to treatment. Thoughtful use of corticosteroids, avoidance of skin trauma, meticulous attention to ventilation systems and appropriate measures during renovation and construction work may prevent most cases of neonatal aspergillosis. A recent study, performed during renovation works in a neonatal intensive care unit in Antwerp, Belgium, investigated the impact of sealing, positive room pressure and air filtration on the spore count in this situation. Renovation work in the unit was associated with a significant increase in the concentration of Aspergillus spores in the air. While the additional use of mobile air filtration devices led to a significant decrease in the spore counts, no relationship was found between air contamination and colonization. Aspergillus spp. was isolated de novo from the nasopharynx in six of a total of 311 neonates admitted during the 11-months observation period, but in all instances, subsequent cultures were negative, and no case of invasive aspergillosis occurred during a total of 6176 days of exposure [Mahieu et al. 2000]. Nevertheless, in our opinion, any culture positive for moulds obtained from a neonate should be considered seriously and prompt empirical treatment should be instituted until infection can be reliably excluded.
Back to top

Andreas H. Groll, M.D.,
Senior Staff Fellow,
Immunocompromised Host Section,
Pediatric Oncology Branch,
National Cancer Institute,
Building 10,
Room 13N240,
10 Center Drive,
Bethesda, MD 20892.
Phone: (301) 435-3355. Fax: (301) 402-0575 E-mail: grolla@mail.nih.gov

August 2000