Fungal Infections in Preterm Infants 

Updated: Jun 25, 2014
  • Author: David A Kaufman, MD; Chief Editor: Ted Rosenkrantz, MD  more...
  • Print

Introduction and Pathogenesis


The risk for invasive fungal infections is high in very low birth weight (VLBW) infants (< 1500 g) and highest for infants born at the youngest gestational ages who survive past the immediate postnatal period. [1, 2] These immunocompromised infants usually require invasive therapies, such as central vascular catheters and endotracheal tubes, and are exposed to broad-spectrum antibiotics and parenteral nutrition. In addition, they occasionally receive postnatal steroids and gastric acid inhibitors. All of these factors place them at high risk for fungal infection. The incidence had been increasing in infants less than 1000 g with the resuscitation and survival of more and more infants prior to the studies of and broad institution of antifungal prophylaxis in high-risk preterm infants. [3, 4]

Most fungal infections in preterm neonates are due to Candida species; a much smaller number of infections may be attributed Malassezia, Zygomycetes, or Aspergillus pathogens. Candida species are commensal organisms that colonize the skin and mucosal surfaces and adhere to catheter surfaces. Candidaalbicans and parapsilosis account for 80-90% of infections. Candida can invade the bloodstream and disseminate in these infants because of their immature immune systems, complicated by the inevitable need to compromise their developing skin and mucosal membrane barrier defenses. For these reasons, fungal infections are often difficult to eradicate in the preterm infant.

Although these immunocompromised infants are at increased risk during most of their hospital stay, they are at the highest risk of acquiring invasive fungal infections during the first weeks of life, when the most invasive therapies are performed and remain in place. Although an index of suspicion must always remain high, infection control, prophylaxis, and aggressive treatment (antifungal therapy and central catheter removal) during this period have the greatest potential to improve the outcome of this population.


The pathogenesis of fungal infections in preterm infants involves adherence, colonization, and dissemination (as is shown in the image below).

Pathogenesis and invasive fungal infections in ver Pathogenesis and invasive fungal infections in very low birth weight infants. From Kaufman and Fairchild 2004, with permission.

Adherence and the slow-growing nature of Candida facilitate its ability to colonize and disseminate into the bloodstream and body tissues before clinical signs and symptoms of infection become apparent. Surface glycoproteins play a role in fungal adherence. One such surface adherence glycoprotein is INT1p, which binds to beta-integrins present on the endothelium and WBCs. The absence of a functional INT1 gene diminishes adherence in yeast cells but not filamentous forms.

The preterm infant is immunocompromised and frequently exposed to broad-spectrum antibacterial medications. Investigators have studied the effect of steroids and antibiotics in mice orally inoculated with Candida albicans to mimic conditions in the preterm infant. [5] Antibiotic treatment alone led to increased Candida colonization but did not affect dissemination. When dexamethasone was added to the antibiotic regimen (presumably amplifying the inherent immunoincompetence), both colonization and dissemination increased in these animal models. Dexamethasone plus antibiotics led to an increase in the percentage of filamentous forms in the GI tract compared with antibiotics alone. In addition, introduction of C albicans strains with 2 functional copies of the INT1 gene increased the number of fungi colonizing the cecum and disseminating to extraintestinal sites.

C albicans is dimorphic, having both yeast and filamentous forms (eg, hyphae, pseudohyphae, germ tubes), and is assumed to have increased virulence in immunocompromised patients because of the filamentous forms. Filamentous forms may contribute to colonization and infection, although species that do not form filaments, such as Candida glabrata, colonize and cause invasive disease in VLBW infants.

To further examine the role of yeast and filamentous forms, researchers intravenously or orally infected antibiotic-treated and dexamethasone-treated mice using 3 strains of C albicans: (1) a wild-type strain that had both yeast cell and filamentous forms, (2) a strain with only yeast cells, and (3) a strain that was constitutively filamentous. [6] The mortality rate was significantly greater in both the wild-type (92%) and yeast-cell (56%) strains compared with the filamentous strain alone (0%). The filamentous strain had no dissemination, and cecal colonization was significantly less than that of the other 2 strains. The wild-type strain had diffuse hyphal invasion with increased tissue necrosis compared with the yeast-cell strain. The researchers speculated that the yeast forms are critically important for adherence and tissue dissemination and that hyphal formation in the tissues contributes to parenchymal destruction.

In preterm infants, vertical and horizontal transmission leads to colonization of the skin, mucosal membranes (GI and respiratory tracts), and central vascular catheters (as is shown in the image below).

Percutaneous intravenous central catheter and fung Percutaneous intravenous central catheter and fungal biofilm formation.

After exposure, patient factors, such as degree of prematurity, skin condition, endotracheal intubation, central vascular access, diseases (eg, necrotizing enterocolitis [NEC], focal bowel perforation [FBP]), and abdominal surgery, can contribute to fungal infection. Fungal factors that contribute to infection include the size of the inoculum and factors that favor colonization and proliferation (eg, use of broad-spectrum antibiotics, postnatal steroids, histamine type-2 [H2] antagonists, parenteral nutrition, or lipid emulsions [Malassezia species]).

Invasive infection of the blood, urine, cerebrospinal fluid (CSF), or peritoneal fluid can lead to disseminated infection, which most commonly involves the heart, kidneys, CNS, eyes, and/or liver.


Risk Factors

Invasive fungal infection risk factors are shown in the image below.

Risk factors for candidemia. PICC = peripherally i Risk factors for candidemia. PICC = peripherally inserted central catheter.

In the very low birth weight (VLBW) infant, colonization of the skin, mucosal membranes, and/or vascular catheters commonly precedes infection. Biofilm formation on catheters inhibits the host's defense mechanisms and the penetration of antifungal agents. Infusates may also become contaminated and directly seed the bloodstream.

Risk factors for Candida colonization and sepsis are similar. [7, 8] Central vascular catheters, vaginal delivery, use of third-generation cephalosporins, and high acuity are risk factors for C albicans infection. H2 antagonists, third-generation cephalosporins, central vascular catheters, parental nutrition and lipid emulsions, and high acuity are risk factors for Candida parapsilosis infection. GI disease (eg, necrotizing enterocolitis [NEC], focal bowel perforation [FBP]), exposure to fluconazole or antibiotics, prolonged hospitalization, and infection with other fungi increase the risk of sepsis due to C glabrata. GI mucosal injury, antibiotic suppression of bacterial flora, neutropenia, and parenteral nutrition increase risk of sepsis due to Candida tropicalis.

Patient risk factors and odds ratios (ORs) summarized from 2 multicenter studies are as follows: [9, 7]

  • Gestational age: For patients born at less than 25 weeks' gestation, the OR was 4.2. For patients born at less than 25-27 weeks' gestation, the OR was 2. For patients born at less than 32 weeks' gestation, the OR was 4.
  • Antibiotics: For patients who received third-generation cephalosporin or carbapenem treatment within 7 days prior to infection, the OR was 1.8. For patients who received 2 or more antibiotics prior to infection, the OR was 3.8.
  • Invasive therapies: For patients who received mechanical ventilation therapy, the OR was 10.7. For patients with a central venous catheter, the OR was 3.9.
  • Intravenous nutrition: For patients who received parenteral nutrition for longer than 5 days, the OR was 2.9. For patients who received lipid emulsion longer than 7 days, the OR was 2.9.
  • Medications: For patients using H2 antagonists, the OR was 2.4.

Diseases that increase risk for fungal sepsis are as follows:

  • Prior bloodstream infection: Patients with prior bloodstream infection may be more susceptible to infections and/or the effect of antibiotics on skin and mucosal microflora (OR, 8.02). [10]
  • NEC: Studies have found that as many as 16.5% of VLBW infants with NEC developed candidemia at presentation or during treatment for NEC. [11, 12]
  • FBP: If the GI tract is colonized with Candida species, Candida peritonitis and sepsis can complicate bowel perforation in affected infants. [13]
  • GI disease: Complicated GI disease in which infants receive nothing by mouth (not enterally feed) and/or antibiotics for longer than 7 days increases the risk for fungal sepsis. Examples include gastroschisis, omphalocele, intestinal atresias, tracheoesophageal fistula, and Hirschsprung disease. Complicated GI disease increases risk in both preterm and term infants (OR, 4.57). [10]
  • Candida dermatitis: As discussed above, skin inflammation caused by Candida species may precede invasion into the bloodstream in preterm infants because of the immature host defenses. Treatment with systemic antifungal therapy in preterm infants who weigh less than 1000 g has been suggested. [14, 15]

Candidal Infections

Candida species

Any Candida species may cause disease in neonates.

C albicans remains the most frequently isolated yeast species in infected neonates, followed by C parapsilosis infections in centers not using antifungal prophylaxis. [16, 17] Because fluconazole prophylaxis is highly effective in preventing C albicans colonization and infection, those NICUs have few infections and they are often non-ablicans species. Less common Candida species include C glabrata and C tropicalis and a small percentage of infections are due to C lusitaniae, C guilliermondii, or C dubliniensis.

Candidal infections

See the list below:

  • Congenital cutaneous candidiasis:
    • This may manifest with pustules, vesicles, skin abscesses, and/or an erythematous maculopapular rash when the patient is born (or within a few days after birth). The skin involvement covers 1 or more of the following areas: face/scalp, chest, abdomen, perineal area, 1 or more extremity, and/or back. These lesions occasionally lead to desquamation. [18, 14]
    • Very low birth weight (VLBW) infants with congenital Candida infection are more likely to present with severe infection, such as pneumonia and widespread dermatitis with focal areas of superficial erosion and desquamation. [19, 20, 15]
    • This infection is invasive, with dissemination in preterm infants who weigh less than 2500 g and should be treated with systemic antifungals for a minimum of 14 days. Dissemination can also occur in term infants (~10%); treatment should be for 7-14 days. For affected term infants, discussing the need for evaluation of the infant's immune system for a primary immunodeficiency may be indicated.
  • Congenital cutaneous candidiasis
    • An example of congenital cutaneous candidiasis is shown in the image below.
      Congenital cutaneous candidiasis in a 26-week-old Congenital cutaneous candidiasis in a 26-week-old infant.
    • This usually manifests as an erythematous papulopustular rash. A diffuse, burnlike, erythematous, macular dermatitis with skin exfoliation is more likely to manifest with blood, urine, or cerebrospinal fluid (CSF) involvement than with a papulopustular rash in patients of any gestational age. [14]
    • Extremely low birth weight (ELBW) infants (< 1000 g) with congenital cutaneous candidiasis are at greater risk of developing invasive fungal infection (66%) than LBW (33% in those 1000-2500 g) or term infants (11%). For this reason, all should be treated with systemic antifungal therapy.
  • Mucocutaneous candidiasis: This manifests postnatally with an erythematous papulopustular rash similar to congenital cutaneous candidiasis. As many as 70% of ELBW infants with candidal dermatitis develop bloodstream dissemination. [20, 21, 22] ELBW infants with mucocutaneous candidiasis in the first weeks of life should be treated with systemic antifungal therapy for a minimum of 14 days.
  • Bloodstream infection: Bloodstream infection with fungal species demonstrates clinical signs and symptoms similar to bacterial sepsis. The incidence of candidemia is reported as 2-6.8% among VLBW infants. [23, 24, 25, 16, 26, 27, 28, 15, 7] The incidence is higher in ELBW infants, ranging from 4-16%. [8, 16, 29, 30, 31] The incidence increases in an inverse linear pattern, from around 3% at 28 weeks' gestation to 24% at 23 weeks' gestation. [30, 31] Most importantly, candidemia often represents disseminated disease. Evaluation of cardiac, renal, ophthalmologic, and central nervous systems is warranted.
  • Urinary tract infection (UTI): This infection is extremely common. Evaluation of late-onset sepsis should include a urine culture obtained via sterile catheterization (or suprapubic bladder aspiration). If the urine culture is positive for fungus, renal ultrasonography should be performed to detect fungus in the collecting system. [32] Candiduria develops in approximately 2.4% of VLBW infants and up to 6% of ELBW infants. [1] Studies have demonstrated similar mortality in infants with Candida UTI alone (26%) compared to Candida BSIs (28%) in ELBWs. [33, 34] These findings emphasize the need for prompt treatment for a duration of a minimum length of 14 days in preterm infants.
  • Meningitis: The reported frequency of fungal meningitis among VLBW infants is 1.6%. [35, 31] The true incidence is likely higher because lumbar punctures are not obtained in many VLBW infants at the onset of sepsis. Cell counts in preterm infants may not always be helpful because the results may not be abnormal in the presence of meningitis.

Disseminated infection

Patients with disseminated infection may present with several entities. [11, 36, 37] These infections require longer duration of treatment of 4-6 weeks or longer until resolution. Consultation with pediatric infectious disease specialists should be considered to help guide treatment decisions.

  • Endocarditis: Endocarditis has been reported in 5-15% of candidemia cases.
  • Renal abscess: Renal abscess is detected in 5% of patients with candidemia. It may occur in as many as 36.6% of VLBW infants with fungal urinary tract infections.
  • CNS abscess/ventriculitis: CNS abscess occurs in 4% of patients with candidemia. [36] This may be a complication in as many as one third of infants with fungal meningitis. [38]
  • Endophthalmitis: This occurs in 3-6% of patients with candidemia. Endophthalmitis occurs as multiple or single, yellow-white, raised lesions with indistinct (fluffy) or circular edges located in the posterior fundus or vitreous. It may affect one or both eyes. [39] Possibly due to earlier treatment, appropriate dosing, and prophylaxis, the incidence has been lower in recent years.
  • Liver abscess
    • Liver abscess occurs in 3% of patients with candidemia.
    • Liver ultrasonography is recommended when candidemia first manifests but is particularly indicated if hepatomegaly or significant change in liver enzymes results or in patients with persistent candidemia (>5 d). [11, 12]
    • Hepatic candidiasis is shown in the image below.
      Hepatic candidiasis. Note the white pustules of Ca Hepatic candidiasis. Note the white pustules of Candida albicans on the surface of the liver.
  • Splenic abscess: Splenic ultrasonography is recommended in patients with candidemia and should be performed if splenomegaly occurs or candidemia persists longer than 5 days.
  • Cutaneous abscess: Skin abscesses should be cultured and drained, if indicated.
  • Osteomyelitis: Evaluation should be considered in infants with infection who are not moving, have limited range of motion, or have swelling of an extremity.
  • Septic arthritis: Septic arthritis manifests with joint swelling. It may also occur several months after antifungal treatment. [40]
  • Peritonitis: Peritonitis may occur with any bowel perforation, focal bowel perforation (FBP), and necrotizing enterocolitis (NEC). It can be a complication of any abdominal surgery.

Clinical Presentation and Differential Diagnosis

Clinical presentation

Although, the very low birth weight (VLBW) infant with candidiasis can present with many of the nonspecific signs and symptoms associated with invasive bacterial infection, symptoms are often more subtle and indolent. Cultures should be obtained whenever sepsis is suspected. Cultures should be repeated after the initial evaluation if the infant does not clinically improve within 48 hours or if the infant's condition worsens. New-onset thrombocytopenia (< 100 X 109/L [< 100 X 103/µL, or < 100,000/µL]) is present in most cases of fungal sepsis and decrease an additional 50% to a mean platelet count of less than 50 X 109/L (< 100 X 103/µL, or < 100,000/µL). [41] Persistent thrombocytopenia may indicate therapeutic failure.

Signs and symptoms in VLBW infants with candidemia are summarized according to incidence, as follows: [42]

  • Thrombocytopenia with count of less than 100 X 10 9/L (< 100 X 10 3/µL, or < 100,000/µL) - 84%
  • Immature-to-total neutrophil ratio of 0.2 or higher - 77%
  • Increase in apnea and/or bradycardia - 63%
  • Increase in oxygen requirement - 56%
  • Increase in assisted ventilation - 52%
  • Lethargy and/or hypotonia - 39%
  • GI symptoms (eg, gastric aspirates, distention, bloody stools) - 30%
  • Hypotension - 15%
  • Glucose concentration of more than 140 mg/dL - 13%
  • WBC count of more than 20 X 10 9/L (>20 X 10 3/µL, or >20,000/µL) - 12%
  • Absolute neutrophil count less than 1.5 X 10 9/L (< 1500/µL) - 3%

Differential diagnosis

See the list below:

  • Gram-positive or gram-negative sepsis
  • Necrotizing enterocolitis (NEC)
  • Focal bowel perforation (FBP)
  • Abscess
  • Intracranial hemorrhage
  • Central vascular catheter thrombosis

Morbidity, Mortality, and Follow-up


Neurodevelopmental impairment (NDI) is more common in infants with fungal sepsis who weigh less than 1000 g than in extremely low birth weight (ELBW) infants without infection. [43, 44, 38] NDI does not appear to be more common in infants who weigh 1000-1500 g, but a more detailed study of this subgroup is needed.

In a study of NDI associated with these infections in ELBW infants, Stoll et al examined mental and psychomotor developmental indexes, cerebral palsy (CP), and hearing or visual impairment. [44] Forty-one percent of infected infants (any clinical sepsis, bloodstream infection, or meningitis) and 57% of infants with fungal sepsis had at least one adverse neurodevelopmental outcome. The prevalence of adverse neurodevelopmental outcomes in infants with fungal sepsis were as follows:

  • Mental developmental index of less than 70 - 34%
  • Psychomotor developmental index of less than 70 - 24%
  • CP - 18%
  • Visual impairment - 14%
  • Hearing impairment - 5%

The rate of NDI in infants with fungal sepsis did not differ significantly from those with bloodstream infection with other microorganisms (coagulase-negative staphylococcus [CONS], non-CONS, and gram-positive and gram-negative organisms).

The effect of invasive fungal infection on other morbidities is still being studied. In ELBW infants, several studies have described an association with retinopathy of prematurity. One study demonstrated an increased incidence of bronchopulmonary dysplasia. [38]

Because of the infants' maturing immune systems, outcomes may better correlate with corrected gestational age at the time of infection. Infants who develop infection later in their hospital stay (ie, after 6 wk) may have better outcomes, but this requires further study.


In very low birth weight infants (VLBW) infants, candidemia is associated with a mortality rate of 21-32% in multicenter studies. [45, 46, 47, 48] Less than 26 weeks' gestation and a birth weight of less than 1000 g correlate with increased mortality rates (40-50%). The mortality rate is significantly higher when sepsis is due to C albicans (nearly 44%) than when it is due to C parapsilosis (15%) or other Candida species. [46] Results vary from center to center, and several single-center studies have reported no mortality; thus, intensive care management of the septic infant and other factors may play an important role in survival. [27, 1]


All preterm infants with infection should receive neurodevelopmental follow-up in the first few years of life and early intervention services, if needed.


Workup And Evaluation and Future Diagnostics

Workup and evaluation

Workup and evaluation for fungal infections in preterm infants includes the following tests: blood, urine, and cerebrospinal fluid (CSF) cultures. Clearance of blood stream infection should be documented with 3 or more negative blood culture results. Each negative culture result should be obtained at least 24 hours apart.

  • Laboratory studies at presentation
  • Laboratory studies during the infection
    • Thrombocytopenia is extremely common and can persist until clearance of Candida infection and should be closely monitored during treatment until resolution.
    • Liver and renal function should be evaluated at the time of diagnosis (or if candidemia is persistent) because they may suggest liver or renal dissemination and the need for ultrasonography.
    • Antifungal treatment can affect serum electrolytes and the hematologic, hepatic, and renal systems and should be closely monitored during treatment.
  • Screening tests for dissemination
    • Screening for end-organ dissemination should be performed at the time of diagnosis in all sepsis cases and repeated if fungemia persists for longer than 5 days. Dissemination affects length of treatment.
    • The optimal timing of surveillance is not well defined, but persistent fungemia (>5 d) is associated with increased dissemination. [12, 11] Consequently, appropriate cultures and surveillance for end-organ dissemination should occur at onset and after 5-7 days if persistent fungemia is present.
    • The screening for dissemination includes the following:
      • Echocardiography
      • Renal ultrasonography
      • Head ultrasonography
      • Indirect ophthalmoscopy
      • Peritoneal cultures if laparotomy is performed to manage necrotizing enterocolitis (NEC) or focal bowel perforation (FBP)
  • Laboratory testing in patients with persistent candidemia
    • In patients with persistent candidemia that lasts more than 2 days, central catheters should be removed if they still remain (central catheters should be removed when bloodstream infection is diagnosed).
    • In patients with persistent candidemia of longer than 5 days, repeat screening tests for vegetation or abscess, including the following:
      • Echocardiography and renal and head ultrasonography
      • Liver and spleen ultrasonography
      • Lumbar puncture
      • Bone scan or joint aspiration (if clinical symptoms warrant)
      • Ultrasonography or laparotomy of the abdomen (in patients with a history of abdominal surgery, NEC, or FBP, to evaluate for abscesses)
      • Ultrasonography, venography, or magnetic resonance venography (MRV) of the previous location of the catheter tip (if the patient had any vascular catheters prior to or at the time of diagnosis, to evaluate for a thrombus)

Other diagnostic tests

Currently, polymerase chain reaction (PCR) and fungal markers such as beta-glucan can be extremely helpful for diagnosis and for following the response to treatment of fungal infections. PCR is useful in detecting fungal infections at other sites (urine, peritoneum, abscesses complicating NEC, or focal bowel perforation), when fungal infection suspicion is high despite negative cultures. Beta-glucan can be helpful when additional information is desired to trend treatment response. Beta-glucan levels decrease over time with antifungal therapy.

Investigators are studying molecular techniques to identify fungi (and other microorganisms) and fungal susceptibility with higher sensitivity and more rapidly than with blood cultures. Examples include PCR and DNA microarray technology. The hope is that these techniques will allow for the rapid detection of small numbers of organisms in minute volumes of blood, even after antimicrobial treatment is started.

Fungal PCR to detect the gene for 18S ribosomal RNA (rRNA) in very low birth weight (VLBW) infants has yielded promising results but requires additional study to be used with every infection evaluation. [49] PCR results detect a broader number of infections as they not only detect patients with candidemia but are also positive in those with Candida peritonitis and those with candiduria. [50]

In addition, investigators are examining the role of monitoring markers of fungal disease to diagnose and evaluate responses to antifungal therapy. These markers include beta-glucan of the cell wall, anti-Candida antibodies, D-arabinitol (candidal metabolite), and fungal chitin synthase (assessed with PCR).

Microarray technology and gene chips are being studied to rapidly determine susceptibility and resistance patterns at the time of diagnosis. These will facilitate the initiation of therapy with an appropriate antifungal agent when resistance occurs and, hopefully, improve outcomes.


Treatment, Empiric Treatment, and Prophylaxis

Prompt initiation of systemic antifungal therapy and central vascular catheter removal (if bloodstream infection) at the time of diagnosis are needed to optimize successful eradication, prevent dissemination, and improve outcomes. The image below shows antifungal mechanisms.

Antifungal mechanisms. Yeast cell and targets of a Antifungal mechanisms. Yeast cell and targets of antifungal therapy. From Kaufman 2004, with permission.

Antifungal therapy


Amphotericin B deoxycholate (Fungizone) remains the primary antimicrobial medication for invasive fungal infection. This drug binds to the sterol component (ergosterol) of the cell membrane, creating a pore that leads to cell death. Although test doses have preceded administration in the past based on pediatric and adult responses to the drug, this is not necessary and delays appropriate treatment. Enough safety data now support initiating administration with a starting dose of 1 mg/kg/d without need for lower test doses. [51, 52]

If the initial treatment is ineffective, studies have demonstrated safety with dosing of amphotericin B deoxycholate increased to 1.5 mg/kg per day. [53] Poor outcomes when treating patients may be related to the delay in reaching appropriate antifungal dosing. It should be intravenously administered once daily over 2-6 hours. In infants receiving parenteral nutrition, a second intravenous line is often not needed. Parenteral nutrition can be paused, the line flushed with D5W or D10W and amphotericin B deoxycholate can be mixed in D10W and given in a similar volume per hour to maintain euglycemia.

Several studies have examined lipid formulations of amphotericin B. [54, 52] Lipid formulations distribute to the mononuclear phagocytic system, and doses of 5 mg/kg are required for efficacy similar to that of amphotericin B deoxycholate. One study examined doses of 5-7 mg/kg of lipid amphotericin B formulations in 36 very low birth weight (VLBW) infants and reported no adverse effects. [54] Lipid formulations include liposomal amphotericin, amphotericin B colloidal dispersion (ABCD), and amphotericin B lipid complex (ABLC).

One special circumstance is worth discussing. In patients with urinary tract infections or renal abscesses, amphotericin B deoxycholate has higher renal penetration compared with the lipid preparations and may be more effective. This is based on rodent studies that demonstrated significantly higher renal concentrations of deoxycholate, which has led most experts to recommend preference over lipid formulations for UTIs and renal candidiasis. None of the studies has compared optimal dosing of 1-1.5 mg/kg of amphotericin deoxycholate with 5-7 mg/kg of lipid preparations of amphotericin B.

A rodent study examining 1 mg/kg of both amphotericin B deoxycholate and liposomal amphotericin B found kidney tissue concentrations of 735 ng/g compared with 298 ng/g, respectively. With dosing having a linear concentration for liposomal amphotericin B, a fivefold increase would be expected with dosing of 5 mg/kg, likely resulting in similar kidney concentrations. Clinical studies are desired examining the efficacy of amphotericin products for fungal UTIs and for those with renal involvement.

Amphotericin resistance is extremely rare. Most C lusitaniae strains are susceptible, and infections due to these organisms clear with amphotericin. However, C lusitaniae resistance has been reported. Susceptibility testing can help guide therapeutic choices.


Fluconazole, an azole that inhibits the enzyme C-14 lanosterol demethylase in the formation of ergosterol, has demonstrated similar efficacy to amphotericin B deoxycholate. Fluconazole has excellent tissue penetration. [55] New data recommend a dose of 12 mg/kg/day. [56, 57] Some experts would use every-48-hour dosing in infants younger than 29 weeks in the first 7 days of life and then change to every-day dosing. Studies are also examining the need for a loading dose. Although these recommendations are based on pharmacokinetic data, safety of the higher dosing needs further study and close monitoring. Fluconazole is available as a parenteral for intravenous infusion or as a powder for oral suspension. The oral products are 100% bioavailable; therefore, the same dose may be used for oral or intravenous administration.

The frequency of dosing varies with the patient's gestational and postnatal age. Resistance can occur, and susceptibility testing should be performed if resistance is a concern. Most C krusei isolates are intrinsically resistant to fluconazole. As with all azoles, the drug has many potential interactions and should be closely monitored if administered with cisapride, cotrimoxazole, cyclosporine, phenytoin, rifampin, or macrolides.

Voriconazole is an azole derived from fluconazole with a broader spectrum of antifungal activity. To date, it has not been studied in neonates. Unlike fluconazole, voriconazole is 58% protein bound and contains a cyclodextrin carrier that is cleared by the kidney and can accumulate in infants with renal insufficiency. A rare complication is torsades de pointes, and 13% of pediatric patients have reported visual disturbances (ie, photophobia, blurred vision, color changes). Until further study is completed, administration should be considered only in patients with aspergillosis. Similar to all azoles, voriconazole should be closely monitored if administered with cisapride or macrolides.

Drug levels should be monitored because pharmacokinetic studies in this population are currently lacking if there is renal insufficiency or decreased urine output. Exact dosing may vary by gestational and postconceptional age and birth weight (recommend 6 mg/kg q12h for first 2 doses, followed by 4mg/kg/dose q12h). [58, 59]


A new class of antifungals is the echinocandins, which inhibit 1,3-beta-glucan synthesis of the cell wall. Caspofungin acetate (Cancidas) and micafungin sodium (Mycamine) are now approved in the United States. Caspofungin is approved to treat aspergillosis and infection with Candida species. Micafungin is approved for prophylaxis of Candida infections in patients undergoing hematopoietic stem cell transplantation and for treatment of esophageal candidiasis. Another drug in this class undergoing study is anidulafungin (Eraxis). Because these agents inhibit an enzyme, resistance and safety need to be studied along with efficacy.

Studies are underway to determine the effectiveness of echinocandins in pediatric and neonatal patients. For the echinocandins, caspofungin dosing recommendation is around 2.5 mg/kg and micafungin dosing around 10 mg/kg. [60, 61] A micafungin RCT is currently in progress ( Identifier: NCT00815516). Dosing and central line removal are key in analyzing studies. This is evident in in 2 small casposfungin studies in which the drug has shown promise and some efficacy, but its optimal dosage and safety needs further study.

Caspofungin therapy was studied in 10 neonates with candidemia that persisted 13-49 days despite treatment with amphotericin. [62] Nine infants survived, including one who had a relapse after 15 days of treatment that cleared after caspofungin was administered for another 15 days. Central venous catheters were removed as soon as blood-culture results were known. The dosage was 1 mg/kg/d for 2 days then 2 mg/kg/d. The limitations of the study included small size, lack of pharmacokinetic data, and lack of attempted combination therapy. Another study of 13 patients had a much lower success rate, with 1 mg/kg/d of caspofungin combined with other antifungals. [63] This study was complicated by delayed catheter removal.

Other antifungals

In the future, agents such as nikkomycins, which inhibit chitin synthase of the cell wall, may be added to the antifungal armamentarium.

Central catheter removal with candidal bloodstream infections

Central catheter removal is critical in the treatment of neonatal candidemia. The catheter should be removed upon the first positive blood culture result. Prompt removal, within 24 hours of documented positive blood culture results, is associated with lowered mortality rates, reduced end-organ dissemination, improved neurodevelopmental outcomes, and increased scores on the Bayley scale. [12, 11, 43] The most recent study demonstrated decreased mortality with prompt catheter removal and candidemia (21% vs 37%, P = 0.024) and a trend toward decreased neurodevelopmental impairment (NDI) alone (45% vs 63%, P = 0.08). [43]

While catheter removal at the time of diagnosis and replacement if needed after bloodstream clearance is documented if recommended, cases in which central access is required (eg, on vasopressors or peripheral access unobtainable), removing the line and replacing at a different site is acceptable.

Combination therapy

Amphotericin B with the addition of flucytosine (Ancobon) has been used to treat meningitis in infants who can tolerate the oral formulation of flucytosine. However, efficacy of this regimen has not been shown to be superior to that of amphotericin B alone. Flucytosine is a fluorine analog of cytosine that is converted to 5-fluorouracil, leading to inhibition of thymidylate synthetase and disruption of DNA synthesis. Flucytosine monotherapy rapidly leads to resistance, so flucytosine cannot be used alone. For meningitis or in patients with CNS abscess, the addition of fluconazole (because of its excellent CSF penetration) is a better therapeutic option.

One study examined the use of a second antifungal agent (fluconazole) in combination with amphotericin B in patients with fungal sepsis. The second agent was administered immediately upon discovery of an abscess or a positive urine culture result and also administered in patients with a persistent culture-positive infection for longer than 10 days. [52] Infants received 1 mg/kg of amphotericin B deoxycholate (n=34) if their creatinine level was less than 1.2. If the creatinine level was more than 1.2, they received 5 mg/kg of liposomal amphotericin B (n=6) or ABCD (n=14). Patients were treated for 14 days after negative culture result or until radiographic resolution of abscess. Sterilization occurred in 36 patients (67%) with monotherapy and increased to 52 patients (96%) with polytherapy.

Another issue is the treatment of presumed invasive fungal infection in the absence of positive fungus culture results. Although postmortem diagnosis of invasive candidiasis was common in the past, 2 recent studies demonstrated that only 2.7% of cases were diagnosed at autopsy. [11, 47]

Empiric treatment

In the VLBW infant, an evaluation for signs and symptoms of late-onset sepsis is typically accompanied by antibacterial treatment for at least 48 hours. Some studies have reported on the use of empiric antifungals pending culture results. Some authors propose that starting empiric antifungal therapy while culture results are pending may decrease the high mortality rate associated with candidemia in VLBW infants, especially those born at less than 28 weeks' gestation. [27, 9] In other studies, empiric therapy has improved outcomes in VLBW infants. [27, 64] Neither approach has been treated in a randomized controlled trial.

In a prospective study using empiric antifungal therapy (amphotericin B deoxycholate) in infants less than 1500 g or "very ill” with clinical signs of infection plus vancomycin and/or a third-generation cephalosporin for 7 days and one of the following: parenteral nutrition, mechanical ventilation, postnatal steroids, H2 blocker, or candidal rash or trush.

A scoring system has been proposed that includes thrombocytopenia, a gestational age of less than 28 weeks, and broad-spectrum antibiotic treatment; however, this system has not been prospectively studied for safety or efficacy. [9] Infants with necrotizing enterocolitis (NEC) or focal bowel perforation (FBP) are also at increased risk. Further study is needed to investigate the efficacy and safety of empiric antifungal therapy.

Most fungi are isolated from cultures within 48 hours. [64] Therefore, some experts do not recommend empiric antifungal therapy. They recommend prompt initiation of antifungal treatment and removal of any central venous catheters upon positive culture results. In a study by Noyola et al, the start of antifungal therapy and the removal of central vascular catheters within 2 days after blood cultures were obtained was not associated with increased morbidity or mortality in episodes of fungal sepsis. [11]

In certain circumstances, empiric antifungal therapy for 48-72 hours may be warranted in infants with negative initial culture results who still have signs and symptoms of sepsis after 48 hours of antibacterial treatment and who are recultured. In addition, the infants must have one of the following criteria:

  • Thrombocytopenia (< 100 X 10 9/L [< 100 X 10 3/µL, or < 100,000/µL])
  • NEC or FBP
  • Weight of less than 750 g or a gestational age of less than 26 weeks

Preemptive treatment

Preemptive treatment has been used in a few neonatal studies following early detection of endotracheal or respiratory Candida colonization by culture or mannan antigen in infants 28 weeks’ gestation or younger. Respiratory colonization is a high-risk site in preterm infants. [65, 66] With the approach of early identification of Candida species from tracheal aspirates followed by 14 days of antifungal treatment, invasive Candida infections were reduced from 75% (12 of 16 with positive growth of Candida in culture) to 0% (0 of 16 with Candida mannan detected). [67] This approach was associated with no Candida -associated mortality. Patients with NEC in whom Candida is isolated in their gastrointestinal tract may be another group that would benefit from preemptive treatment.


Because of the high mortality rate and NDI associated with fungal sepsis in VLBW infants, prevention with nystatin, miconazole, and fluconazole has been studied in the highest-risk patients.

Fluconazole studies

Fluconazole is an excellent drug for prophylaxis because of its long half-life, high tissue concentration, low lipophilicity, and low protein binding. One concern with fluconazole prophylaxis though the years has been the emergence of resistance over time, but with dosing of 3 mg/kg twice while high-risk patients have central or peripheral access, resistance has not been seen after several years of prophylaxis.

The initial randomized controlled trial of intravenous fluconazole prophylaxis was performed in high-risk infants who weighed less than 1000 g and had an endotracheal tube or central vascular catheter and was effective in preventing invasive fungal infection. [29] Investigators studied prophylaxis using 3 mg/kg of intravenous fluconazole every 72 hours on days 1-14, every 48 hours on days 15-28, and then daily administration on days 29-42 for as long as 6 weeks if intravenous access is not required. No adverse effect or fungal resistance was detected during the 30-month study period. The same authors examined dosing with 3 mg/kg twice a week compared with the regimen described above and found similar efficacy. [68]

Manzoni and colleagues in Italy published the results of their multicenter, randomized, placebo-controlled trial investigating 2 different fluconazole doses (3 mg/kg and 6 mg/kg) compared with a placebo group. [69] In 322 infants who weigh less than 1500 g, investigators reported a significant difference between the fluconazole prophylaxis groups compared with the placebo patients. Fluconazole prophylaxis has similar efficacy for both the 3 mg/kg and 6 mg/kg groups. Fungal colonization occurred in 7.7% and 9.8% and invasive fungal infection in 3.8% and 2.7% in the 3 mg/kg and 6 mg/kg groups, respectively. In the placebo group, fungal colonization occurred in 29.2% and invasive infection occurred in 13.2% (P = .005 for 6 mg; P = .02 for 3 mg). The results also showed a significant effect in infants less than 1000 g as well as infants less than 27 weeks.

Two multicenter studies were recently published. A randomized controlled trial in infants less than 750 g demonstrated a statistically significant decrease in invasive candidiasis in the fluconazole-treated patients compared with the placebo group (9% vs 3%). [70] The study was underpowered to examine the primary outcome of infection and all-cause mortality. The composite outcome of invasive infection or death would need approximately 1200 patients to be appropriately powered compared with the 361 infants studied. This study did show a high mortality rate of 41% in this patient population who developed invasive Candida infections, compared with 18% in patients without candidiasis.

The second study was a multicenter case-controlled analysis of efficacy and safety in 95 NICUs that also demonstrated efficacy and safety for fluconazole prophylaxis. [71] Fluconazole prophylaxis was administered to 127 patients [754 ± 163 g birth weight (BW) and 25.4 ± 1.7 weeks gestational age (GA)] and were compared with 399 control patients (756 ± 163 g BW and 25.5 ± 1.8 weeks GA). Invasive Candida infection occurred in 1 (0.8%) of 127 infants who received fluconazole prophylaxis compared with 29 (7.3%) of 399 of matched controls (P = .006). Candida bloodstream infection occurred in 1 (0.8%) of 127 fluconazole prophylaxis infants compared with 22 (5.5%) of 399 of matched controls (P = .02). There were no differences in late-onset sepsis due to gram-positive or gram-negative organisms, focal bowel perforation, necrotizing enterocolitis, cholestasis, or overall mortality.

Several observational studies have been completed to examine fluconazole prophylaxis in both extremely low birth weight (ELBW) and VLBW infants. [72, 73, 66, 74] A meta-analysis of randomized and observational studies with control subjects using Mantel-Haenszel methods demonstrated an 84% reduction in invasive fungal infections among 2111 preterm infants (odds ratio [OR], 0.16; 95% confidence interval [CI], 0.08-0.31; P < .001). [29, 26, 72, 73, 66, 74] For high-risk ELBW infants, the studies demonstrated an 88% reduction in invasive fungal infections (OR, 0.12; 95% CI, 0.05-0.29; P < .001).

Dosing with 3 mg/kg twice weekly is effective and limits exposure, cost, and potential adverse effects. When initiated around birth, prophylaxis should be administered for 6 weeks or less in patients with a birth weight of less than 1000 g or less than 6 weeks if intravenous access is no longer needed. For patients with a birth weight of more than 1000 g, continue prophylaxis until intravenous access is no longer needed and until adequate enteral feedings are achieved.

If antifungal prophylaxis with fluconazole is administered, using a different antifungal agent (eg, amphotericin B) for primary treatment of an invasive fungal infection is important. This ensures treatment with a susceptible antifungal agent and possibly decreases the risk of fungal resistance. Surveillance cultures for fungal resistance are recommended when fluconazole prophylaxis is completed.

Nystatin studies

Nystatin, an enteral or orally administered nonabsorbable antifungal agent, was the first antifungal studied for prophylaxis in preterm infants. Nystatin studies have added to our understanding of antifungal prophylaxis by demonstrating its efficacy and use in NICUs with low rates of infection, demonstrating the greater efficacy of prophylaxis when it is started early (by 72 hours after birth) versus later when colonization is detected, and a decrease in all-cause mortality.

Antifungal prophylaxis started after colonization is detected is not as effective as starting shortly after birth in preventing infections. In a study by Ozturk et al, fungal BSI was decreased by 90% when nystatin prophylaxis was started in the first 72 hours (47 ± 12 hours, mean ± SD) compared with only 62% when started after colonization was detected (12.6 ± 2.4 days). [75]

Two large retrospective studies of nystatin prophylaxis demonstrate the use of antifungal prophylaxis even in NICUs with low Candida bloodstream rates. In the largest multicenter epidemiologic study to date (N = 14,778), NICUs using nystatin prophylaxis had lower rates of BSI or meningitis in infants less than 1500 g (0.54% vs 1.23%, P < .0001) and in those less than 1000 g (1.23% vs 2.67%, P < .0001) compared with NICUs that did not use antifungal prophylaxis. [76]

The second large retrospective study of nystatin (1459 infants) demonstrated similar efficacy of nystatin prophylaxis in reducing ICI in preterm infants younger than 33 weeks. There was a 56% decrease in ICI, from 4.1% to 1.8%, and 54% decrease in those infants less than 1500 grams (5.5% to 2.5%) with nystatin prophylaxis. This study also found that all-cause mortality was lower in the nystatin prophylaxis group (11.8%) compared with patients not receiving prophylaxis (17.8%) (P < .0001).

Who is at high risk and should receive antifungal prophylaxis?

High-risk patients are defined as those patients with significant mortality and neurodevelopmental impairment, as well as those with a combination of risk factors for fungemia. Neonatal Candida studies have consistently demonstrated increased mortality and neurodevelopmental sequelae in preterm infants less than 1000 grams. Several studies of risk factors discussed above can be used to help identify other high-risk patients.

Since prophylaxis works best if initiated in the first days of life, identifying those infants at that time leads to the greatest effect in reducing infection. A recent study of infants less than 1250 g in 95 NICUs examined maternal and neonatal risk factors present at birth. [77] NICUs can examine their invasive Candida infection rates at each gestational age by week to see at what gestational age infections fall to zero. From published data including all invasive Candida infections (BSIs, UTIs, peritonitis, and meningitis), antifungal prophylaxis would be beneficial for infants born at less than 28 weeks' gestation while they have risk factors (see below). [9, 30]

Prophylaxis should be administered while high-risk infants require intravenous access. This targets prevention to the time period infants have risk factors. When intravenous access is no longer required, the risk for invasive fungal infection decreases because the additional risk factors for fungemia (eg, parenteral nutrition, lipid emulsions, broad-spectrum antibiotics, and central venous access) are no longer present. [16, 29] For example a 26-week gestation infant who reaches full enteral feedings at 3 weeks after birth (and no longer needs central or peripheral intravenous access), would receive only 6 doses of fluconazole. Other patients that may benefit and are being studied include preterm and full-term infants with complicated GI disease (e.g. NEC, gastroschisis) who require prolonged periods without enteral feedings.

The AAP has recommendations that are in the Redbook. They advocate for prophylaxis in the highest-risk patients < 1000 grams. [78, 79] The highest-risk patients who may benefit include the following:

  • All infants with a birth weight of less than 1000 g and/or < 28 weeks' gestation
  • Preterm and full-term infants with GI disease, such as NEC, FBP, or complicated cases of gastroschisis, omphalocele, intestinal atresia, and Hirschsprung disease (eg, not receiving or expected to receive enteral feedings for >7 d)

For patients who receive fluconazole prophylaxis, the dose should be 3 mg/kg twice weekly until intravenous access is no longer required. In patients with suspected or documented fungal infection who receive prophylaxis, amphotericin B should be administered as the initial antifungal therapy until Candida species infection and fungal susceptibility is determined.


Malassezia Infections, Aspergillosis, Zygomycosis (Mucormycosis)

Malassezia infections

Presentation of infection with Malassezia organisms is similar to that seen with invasive candidiasis. Infection does not routinely disseminate; therefore, end-organ surveillance is needed only if species are persistently isolated from several cultures.

Malassezia furfur is a lipid-dependent fungus that may colonize central venous catheters when lipid emulsions are infused. It can also colonize the skin and GI tract. Horizontal transmission is common. These fungi readily grow in Sabouraud medium coated with sterile olive oil. Treatment can include any one of the following measures:

  • Stopping lipid infusions for 48-72 hours
  • Stopping lipid infusions for 48-72 hours and administering amphotericin B for 7 days
  • Stopping lipid infusion for 48-72 hours and removing the central venous catheter
  • Removing a central venous catheter

Malassezia pachydermatis is not an obligate lipophilic organism. It has been reported to cause sepsis, urinary tract infection, and meningitis in very low birth weight (VLBW) infants but not in other neonates. Horizontal transmission occurs and can be prevented with handwashing.


Aspergillus infections are rare in neonates but are associated with a high morbidity and mortality rate. Aspergillus species are ubiquitous filamentous fungi (eg, molds) that form spores in the air, soil, decaying vegetation, and dust. For the neonate, transmission usually involves airborne spores. The site of entry may be the respiratory tract, skin, or central vascular catheter. Infection is usually due to exposure to contaminated dust. Invasive aspergillosis in infants can be cutaneous, pulmonary, or systemic infections, with occasional dissemination to the CNS.

Diagnosis is difficult, and a high index of suspicion is needed. Any culture that is positive for Aspergillus must be considered serious in preterm infants. Any skin or oral rashes or lesions should be cultured. Pulmonary presentation should be considered if infection is suspected with negative culture results and persistent signs despite antibacterial treatment.

One presentation involves injured skin areas that rapidly (over 24 h) progress to necrotic eschars. Diagnosis is made by demonstrating septate hyphae with 45° angles characteristic of Aspergillus species. Spores do not readily grow in blood cultures. The organism can be isolated from lung or skin samples when they are cultured on Sabouraud dextrose agar. Bronchoalveolar lavage fluid can also be microscopically examined and cultured.

Treatment has routinely involved amphotericin B, but with little success. Studies in adults have shown that newer antifungals, including voriconazole and echinocandins (eg, caspofungin or micafungin), are more effective than amphotericin B. Because the efficacy, safety, and optimal dosing of these antifungals is currently being determined, consultation with a pediatric infectious disease specialist and pharmacologist is important.

Prevention is crucial and involves filtration of NICU ventilation systems and containment of dust, especially during hospital renovation and construction. High-efficiency particulate air (HEPA) filters are excellent in clearing almost all of these fungi. NICUs should have continuous surveillance programs for mold, especially in and around windows, which can lead to good preventative measures. Ensuring that all ceiling tiles are in proper alignment is critical because all ceilings have some degree of mold.

Any construction in the surrounding area outside of the hospital can also increase the air spore count. During renovation or construction, the air should be tested for Aspergillus with the aid of infection control and microbiology services. HEPA filters should be used to prevent infection if significant levels of Aspergillus are detected.


Zygomycotic infections initially present as a black eschar at a site of local trauma or intravenous catheter insertion or infiltrate and progress to a necrotizing soft tissue infection. [28, 49] Early diagnosis, treatment with amphotericin B, and surgical debridement are needed to prevent ulceration, necrosis, and rapidly fatal dissemination. A high degree of suspicion is needed, and tissue biopsy must be performed to identify the nonseptate hyphae with right-angled branches. The mortality rate associated with these infections is reported to be 61%. [28, 49]