• 
• 

Three Part Question

In [preterm very low birth weight infants] is [intravenous fluconazole prophylaxis effective and safe] in [reducing invasive fungal infection as well as mortality]

Clinical Scenario

A 26 week gestation premature neonate is born with a birthweight of 650g. He is intubated and ventilated from birth. Both arterial and venous umbillical catheters are inserted. He is at risk of sepsis from maternal chorioamnionitis and prolonged rupture of membranes and is therefore commenced on broad spectrum antibiotics. Invasive candidiasis is an increasingly recognised problem in such infants. Would intravenous antifungals be effective and safe at preventing this?

Search Strategy

EMBASE online database. 1974-date
OVID online database. No date restricition
PUBMED medline database. No date restriction
{Fluconazole OR Antifungal} AND {Prophylaxis OR Prevention} AND {Fungal infection OR Candida} AND {Preterm infants OR Very Low Birthweight Infants}

Search Outcome

132 articles found. Duplicates, irrelevant articles and studies of insufficient quality were removed. Six studies remain.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Kaufman et al 2001 United States	50 extremely low birth weight infants (<1000g) compared to 50 placebo controls.	Prospective randomised double-blind placebo controlled clinical trial. 6 weeks treatment intravenous fluconazole 3mg/kg every third day for 2 weeks then 48hrly for 2 weeks then daily in weeks 5 and 6. Surveillance cultures taken weekly. Drug or placebo stopped when intravenous access removed.	Fungal colonisation	Reduced in treatment group (22%) compared to placebo group (60%) p=0.002	Diagnosis of systemic fungal infection based upon positive blood, CSF, urine cultures. How much effect on culture sensitivity does fluconazole have?
			Fungal infection	Infection developed in 10 of 50 in the placebo group. None in treatment group. Difference in risk 0.20. (95% CI 0.04 to 0.36; p=0.008)
			Mortality	4/50 in fluconazole group. 10/50 in placebo group. Not significant p=0.22
Kicklighter et al 2001 United States	53 very low birth weight infants (<1500g) less than 72hrs old compared with 50 placebo controls.	Prospective randomised double blinded placebo controlled trial. Intention to treat analysis. Power calculation included. Intravenous fluconazole 6mg/kg given every 72hrs for first 7 days then 24hrly until day 28. Changed to oral once tolerating feeds. Surveillance rectal swabs on day of enrollment then DOL 7,14,28 (and DOL 35,49,56 in babies <1250g).	Candidal colonisation	8 of 53 in study group (15.1%) 23 of 56 in control group (46%). p=0.0005	Primary outcome only rectal colonisation rates. Not powered or designed to look at systemic fungal infection rates.
			Invasive candidiasis	No difference between groups
McGuire et al 2004 North America	Very low birth weight infants (<1500g) with a subset of extremely low birth weight infants (<1000g). Metanalysis of 3 randomised controlled trials. 214 infants in total.	Cochrane Systematic Review. Meta-analysis of Kaufman et al and Kicklighter et al trials as above and a smaller trial by Cabrera (11 patients - abstract only). Heterogeneity analysis completed. Prespecified outcomes: death before discharge; longterm neurodevelopment; incidence of invasive fungal infection; emergence of antifungal resistance; adverse drug reactions.	Death	When Kaufman and Kicklighter trials combined - reduced risk of death in treatment groups. Relative risk 0.44 (95%CI 0.21 to 0.91). Typical risk reduction 0.11 (95% CI 0.02 to 0.21). NNT 9.	Doesn't just look at IV fluconazole, some studies use oral once enteral feeds tolerated. Subsequent to this meta analysis further large studies have been published. Significant statistical heterogeneity in risk difference calculations for invasive candidiasis.
			Invasive fungal infection	Reduced risk in treatment group. Typical relative risk 0.2 (0.07 to 0.64). Typical risk reduction 0.12 (0.05 to 0.2). NNT 8.
Healy et al 2005 United States	446 extremely low birthweight infants (<1000g) over 4 year time period.	Cohort analysis of ELBW infants born in 2 yr period before fluconazole prophylaxis introduced and in 2 year period after introduction of a protocol for its administration (82% of ELBW received fluconazole). Invasive candidiasis (defined by positive blood or CSF cultures) compared between two time frames. Dosage as per Kaufman study. 6 weeks treatment, stopped early if intravascular access no longer required,or developed invasive candidiasis. (54% stopped early).	Invasive candidiasis	Reduced in the fluconazole group (5 of 240 (2%)) compared with pre-fluconazole group (15 of 206 (7%)). p=0.01	Not a randomised controlled trial. Authors noted a trend for IC to develop in larger infants later in their neonatal course. Such infants are more likely to have a good outcome however.
			Invasive candidiasis related deaths	Fell in the FP period from 2% to 0% (p=0.04)
			All cause mortality	No significant difference
Bertini et al 2005 Italy	Very low birth weight preterm infants (<1500g) with central vascular access over a six year period. N=155	Cohort study comparing three year baseline period with subsequent three year period where all infants <1500g with central vascular access were eligible for fluconazole prophylaxis. Dosage as Kicklighter to day 28. Invasive fungal infection defined by positive blood, CSF or urine cultures.	Invasive fungal infection	7.6% in baseline group. 0% in FP group. Risk reduction 0.076 (0.03 to 0.122 p=0.003)	Not a randomised controlled trial.
			Mortality	Trend for decreasing mortality. 12.6% vs 8.1% (p=0.32)
Manzoni et al Jan 2006 Italy	465 very low birth weight infants (<1500g). Over 6 year period.	Retrospective non randomised interventioanl study. 240 VLBW infants admitted between 1998 and 2000 did not receive fluconazole. 225 born between 2001 and 2003 did. Dosage as per Kicklighter study until day 30 (1000-1500g) or day 45 (<1000g). Oral route used if tolerating feeds and no intravenous line. Categorises infection into proven (positive cultures) and presumed (clinical picture of fungal sepsis in infants heavily colonised).	Fungal colonisation	24% in prophylaxis group compared to 43.8% in the baseline group. RR 0.406 (95% CI 0.273 to 0.605)	Although this is a well designed study with large patient numbers it is not a randomised controlled trial. There may be other confounding factors such as increased awareness of infection control measures over the time of the study.
			Systemic fungal infection	4.4% in prophylaxis group compared to 16.7% in the baseline group. RR .233 (0.113 to 0.447)
			Rate of progression from colonisation to systemic infection	Reduced from 0.38 to 0.17. Relative risk 0.369 (0.159 to 0.815)
			Overal Mortality	Similar in both groups
			Mortality in colonised infants	3.7% in prophylaxis group compared to 18.1% in baseline group. Relative risk 0.20 (0.039 to 0.778)

Comment(s)

The above papers suggest that intravenous fluconazole is effective in reducing colonisation, systemic infection and mortality from candidal species. Potential side effects of fluconazole are skin eruptions, hepatotoxicity and thrombocytopaenia. Kicklighter et al found slight elevation of serum ALT levels at day 14 but no difference at day 28. The other studies did not report any adverse effects or appreciable effect on LFTs. Certain authors express concerns that widespread use of fluconazole could induce resistance to this antifungal. The above studies monitored MICs of all candidal isolates and found no evidence of emergent resistance. Whilst the roles of oral nystatin and miconazole prophylaxis have been looked at previously, the use of oral fluconazole prophylaxis has not systematically evaluated. All the above trials look at intravenous administration but in some cases a switch to the oral preparation was made mid-treatment.

Clinical Bottom Line

Fluconazole appears to be effective and safe in preventing candidal colonisation of immature neonates and the development of invasive fungal infection. A multi-centred, well designed prospective clinical trial to confirm safety,efficacy and reduction in mortality, of both intravenous and oral fluconazole prophylaxis would be beneficial. We should continue to monitor for signs of resistance.

References

1. Kaufman D, Boyle R, Hazen KC, Patrie JT, Robinson M, Donowitz LG Fluconazole prophylaxis against fungal colonization and infection in preterm infants. N Engl J Med 2001 Dec 6;345(23):1660-6.
2. Kicklighter SD, Springer SC, Cox T, Hulsey TC, Turner RB Fluconazole for prophylaxis against candidal rectal colonization in the very low birth weight infant. Pediatrics 2001 Feb;107(2):404-5
3. McGuire W, Clerihew L, Austin N Prophylactic intravenous antifungal agents to prevent mortality and morbidity in very low birth weight infants. (Review) Cochrane Database of Systematic Reviews 2004, Issue 1. Art No: CD003850.pub2
4. Healy CM, Baker CJ, Zaccaria E, Campbell JR Impact of fluconazole prophylaxis on incidence and outcome of invasive candidiasis in a neonatal intensive care unit. J Pediatr 2005 Aug;147(2):166-71
5. Bertini G, Perugi S, Dani C, Filippi L, Pratesi A, Rubaltelli FF Fluconazole prophylaxis prevents invasive fungal infection in high risk, very low birth weight infants. J Pediatr 2005 Aug;147(2):162-5
6. Manzoni P, Arisio R, Mostert M, Leonessa M, Farina D, Latino MA, Gomirato G Prophylactic Fluconazole is Effective in Preventing Fungal Colonisation and Fungal Systemic Infections in Preterm Neonates: A single center, 6-year, retrospective cohort study. Pediatrics 2006;117;22-32