To the Editor.
We are very interested to see this review published recently,1 and would endorse the general message regarding clinical presentation and investigation. In our series of 7 children all under 1000 g,2 we felt that early recognition and aggressive therapy might reduce the mortality from this condition.
Our practice is now to take early specimens from potential cases, with a particular focus on those children below 26 weeks and below 750 g. With any suspicion of early Candida infection, we commence systemic fluconazole, 8–12 mg/kg 72 hourly. Should the fluconazole not be successful, or if cultures prove positive, our next line of therapy is systemic liposomal amphotericin, since this preparation has an improved therapeutic index compared with standard amphotericin B. In particular, potassium loss is much less of a problem. With this approach, we have had no further severe cases of this condition and no further deaths among affected patients.
It is likely that a high index of suspicion and vigorous early treatment can improve the prognosis for this vulnerable group.
- Copyright © 2001 American Academy of Pediatrics
In Reply.
We thank Drs Melville and Kempley for their interest in our article1 and for sharing their perspective on the management of very low birth weight (VLBW) neonates with congenital cutaneous candidiasis and suspected invasive infection.
As discussed in our article, we concur that early recognition and aggressive therapy are paramount in managing neonates with suspected invasive candidiasis in association with congenital cutaneous candidiasis. Drs Melville and Kempley advocate initial empiric therapy in suspected cases with fluconazole, and a small but growing body of evidence supports its use for treatment of systemic candidiasis. Efficacy and safety for treatment of disseminated neonatalCandida infection was equivalent to amphotericin B in a small randomized prospective trial,2 and favorable efficacy and safety data has been reported in additional patients.3–5 Fluconazole is synergistic with flucytosine, and may be particularly advantageous for treatment of cases complicated by thrombus formation or central nervous system involvement. Little data is yet available, however, on use of fluconazole for first-line treatment of invasive candidal infections in neonates, particularly low birth weight infants, and most experts still advocate amphotericin B as the treatment of choice.6–9In recognition of this, Drs Melville and Kempley suggest that when invasive Candida infection is documented, therapy should be switched from fluconazole to amphotericin. When used, fluconazole serum levels should be monitored,10,,11 and it should not be used concurrently with amphotericin B due to possible antagonism.12,,13 Candidal species also vary in their susceptibility to fluconazole, with resistance particularly common among isolates of Candida kruzei.14,,15
Drs Melville and Kempley advocate use of liposomal amphotericin in preference to amphotericin B for first-line treatment of documented invasive infection. Although infusion-associated and renal toxicities are reduced with lipid amphotericin preparations, dosages of these products for use in preterm infants have not been established. Use of liposomal amphotericin preparations may be warranted, however, for those in whom toxicity has become limiting, disease has progressed despite optimal first-line therapy, or the risk of renal impairment is considered unacceptable.16,,17 Further randomized clinical studies are needed before liposomal amphotericin products are considered first-line agents for documented systemic fungal infections in neonates.9 Thus, we advocate continued use of amphotericin B as the principal agent for treatment of VLBW neonates with disseminated candidiasis.
REFERENCES
- Copyright © 2001 American Academy of Pediatrics
In this issue
