Current evidence of antifungal prophylaxis and therapy in pediatric patients

Invasive fungal infections (IFI) are an important complication in pediatric haematological and oncological patients who undergo intensive chemotherapy for leukemia, solid tumour at advanced stage or relapsed, and hematopoietic stem cell transplantation. The incidence of IFI is lower than bacterial infection but mortality rate remains high. This review is designed to help paediatric oncologists in choosing the appropriate anti-fungal strategy and agents for prophylaxis, empirical, pre-emptive and specific therapy on the basis of published evidence.


Introduction
Despite the advent of advanced anti-fungal agents, invasive fungal infections (IFI) remain a major cause of morbidity and mortality in neutropenic cancer patients. Risk classes are well stratified for adult patients but not for children. 1 In a recent prospective survey of Italian Association of Pediatric Hematology Oncology (AIEOP) showed that IFI were more frequent in pediatric patients with acute myeloid leukaemia (AML) during induction therapy, acute lymphoblastic leukaemia (ALL), relapsing non-Hodgkin Lymphoma (NHL) and advanced stage or relapsing solid tumours. 2 Prophylaxis, empirical and pre-emptive therapy Few data are available in paediatric patients. In epidemiological reports on children the incidence of IFI ranged from 5-13% with a peak of up to 24% in AML patients. Yeast infections have represented for several decades the prevalent etiology of proven IFI but in recent years invasive aspergillosis (IA) is continuously increasing. 2,3 Mortality rates remains high, with IA-related mortality ranging from 33-55.6%. 3,4 Given this mortality rate, behavioural measures strategies such as cooking food, washing hands, wearing face masks and using HEPA filters are needed to prevent the risk of developing IFI. Recommendations for adult patients have been recently issued by leading scientific societies without covering the pediatric aerea. [5][6][7] In adult patients, meta-analyses indicated that anti-fungal prophylaxis significantly reduced the incidence of IFI but not the overall mortality. A randomised controlled study reported that prophylaxis with posaconazole significantly reduced the incidence of proven and probable IFI in patients with AML. 8 Regarding empirical therapy a recent metaanalysis indicated that this strategy reduced significantly the incidence of proven and probable IFI in neutropenic patients but it was not able to reduce overall mortality. 9 The comparison of different anti-fungal agent used empirically, particularly azoles vs liposomal Amphotericin B, did not show any difference in efficacy or mortality.
To minimise prescription of unnecessary anti-fungal agents every diagnostic efforts should tried before. In the neutropenic adult patients, Maertens et al., 10 observed that a preemptive strategy, based on daily serum galactomannan antigenemia testing, computed tomography (CT) scan, and bronchoalveolar lavage reduced anti-fungal treatment by 78% compared with empirical therapy without affecting negatively survival. This approach is feasible for IA but not studied adequately so far for zygomycetes.
Only one study 11 compared pre-emptive and empirical strategies. The pre-emptive approach was associated with a significantly greater incidence of IFI, particularly in patients with acute leukemia in induction, but it had no impact upon overall mortality. The study by Girmenia et al., 12 which was based on an intensive diagnostic work-up that included a daily galactomannan test for 3 consecutive days and a chest CT scan, reported the preemptive approach was associated with a 43% reduction in anti-fungal agents compared with the empirical. There was no undue delay in starting therapy and a 63% survival rate was achieved in patients with IFI.
In children the level of evidence for the use of antifungal drugs is inferior than the adults because the published data are limited and based mainly on retrospective studies or on studies of pharmacokinetic; moreover, the antifungal drugs marketed recently have no paediatric indication (posaconazole, anidulafungin), or have an indication narrower than the in vitro activity of the molecule (micafun-gin). Consequently, differently from adults, no specific guidelines are available for IFI in children. We performed a literature search with the aim assess the level of evidence for the use of antifungal drugs in paediatric patients affected by IFI. The key words used for the selection of the studies were antifungal prophylaxis, antifungal empiric therapy, antifungal pre-emptive therapy, antifungal therapy, fungal infections, and pediatric malignancy. The literature search was limited to English language papers and the period analyzed was from 1 January 1990 to 31 July 2010. Reference list of papers selected by literature search was also used as source, if necessary. The results of the search were first discussed and scored by the authors and then presented for the final discussion and approval to an expert meeting held in Florence on 12 and 13 October 2010 where the representatives of centers belonging to Italian Association of Paediatric Hematology Oncology (AIEOP) attended. The scoring system proposed by Infectious Diseases Society of America was adopted to define the quality of evidence and strength of recommendations. Table 1 summarizes the published studies on antifungal prophylaxis. Most studies were retrospective or phase II with a limited sample size. Moreover, the only randomised study 13 with a large cohort of patients, dating back to 1994, did not satisfy current EORTC criteria for fungi classification. Overall, the studies on prophylaxis were scored as C with an evidence level of between II and III. The need of prophylaxis is still debated and depends on the type of patients managed (high-risk vs. low-risk), local incidence of IFI, type of fungi isolated, attitude of the centre to intensive diagnostic work-up in case of IFI is suspected, and finan- cial resources. We concluded that no strong recommendation is possible for antifungal prophylaxis in pediatric patients. However, It should be considered for all high risk patients such as AML during induction, acute leukaemia and NHL in relapse, advanced stage or relapsing solid tumours when local epidemiology data show an incidence of IFI higher than usually reported in other centres or higher than in a recent past. The antifungal drugs most commonly used and resulted safe and active, in analyzed studies, were fluconazole and itraconazole, liposomal amphotericin-B, and recently micafungin. The determination of blood level is recommended when itraconazole or other azoles are used. 5 Differently from prophylaxis, the empiric use of antifungal drugs has been investigated in prospective controlled randomized studies. In a multi-centre, randomised study recruiting adults and children Prentice 14 reported that liposomal amphotericin B was more efficacious than standard amphotericin B in patients with fever of unknown origin (FUO). In a double-blind multi-center study Maertens et al. 15 showed that caspofungin was not inferior to liposomal amphotericin B in children. A recent prospective, randomised, controlled study by Italian Association of Pediatric Hematology Oncology (AIEOP) that evaluated the empiric antifungal therapy in patients at low and high-risk of IFI reported that no significant differences in the tolerability and efficacy of liposomal amphotericin B and caspofungin (Caselli D., personal communication). Moreover, in the low-risk arm the patients were randomised to three different treatment i.e. caspofungin vs. liposomal amphotericin B vs. no antifungal therapy. Success rate and 30day survival resulted similar in all three arms, suggesting that the use of empiric antifungal treatment is no more fundamental in this category of patients.
We concluded that empiric antifungal therapy has a level of evidence of BII in high-risk patients. Liposomal amphotericin B and caspofungin are the recommended drugs. In the lowrisk patients, this strategy is not of proven efficacy and should be discouraged, while a careful diagnostic work-up is recommended Pre-emptive treatment is the new frontier of early antifungal therapy because has the potential for reducing unnecessary antifungal therapy given empirically on the basis of a persisting neutropenic fever refractory to antibiotic treatment. The adoption of a pre-emptive strategy needs a prompt diagnostic evaluation with CT scan, galactomannan antigen, and 1-3 b-D-glucan antigen, as well as the performing of brochoalveolar lavage (BAL). This diagnostic panel may be difficult to perform in children for several reasons. CT may need sedation in younger children, may increase the exposure to radiation in case of repeated exams, or fre-  quently may results in a specific sign. Burgos et al. reported that nodules and a specific infiltrates were the most common findings in children and, differently from adults, halo sign was detected in only 6.5% of children with IA. 4 Serum fungal antigen tests may be useful for the diagnosis of IA in pediatric cancer patients. The performance of serum galactomannan antigen has been show to be good in children as in adults with regard to specificity and NPV (negative predictive value). The less sentivity of the test than adult patients could be attributed to a high incidence of false positives due to the type of diet, the cross-reactivity with b-lactamic antibiotics commonly used in children (penicillin, acid clavulanic/amoxicillin), and cross-reactions with Bifidobacteria lipoteic acid in intestinal micro-flora. 16,17 Moreover, there is less experience of testing galactomannan antigen in children who performed a BAL. Scarce information is available on the use of 1-3 b-D-glucan in children.

Article
Altogether, we concluded that pre-emptive therapy strategy, though a valid option, has been not investigated adequately in the pediatric patients so it is not possible to formulate any recommendation on it. Therefore, prospective study are needed.

Therapy for probable/proven mycosis
Although prospective phase III studies are almost lacking in children, several retrospective, case-control, or prospective phase II studies have been published on proven/probable IFI in children. Table 2

Polyenes
Liposomal amphotericin B has a broad-spectrum activity against the most common yeast and mould species, while Aspergillus terreus, Aspergillus versicolor, Aspergillus flavus (some), Scedosporium and Candida lusitaniae are reported to be resistant.
Liposomal amphotericin B was scored as BII for the first and second line treatment of IA, BI for invasive candidiasis, and BIII for nonaspergillus mould infections. [18][19][20][21] Azoles Triazoles are a class of drugs with a narrower spectrum of action and less toxicity than amphotericin B. Voriconazole is licensed for children over 2 years old. It was scored BII both for IA, candidiasis, and invasive infections by Scedosporium and Fusarium. The use of voriconazole as targeted therapy in children was investigated by Walsh,22 in 58 patients who reported 42% complete and partial responses and 7% disease stabilization. Cesaro 23 and Yilmaz 24 achieved similar results in a total of 34 patients with refractory IA. Recently, the therapeutic drug monitoring has been suggested to optimize the efficacy and prevent toxicity. In fact, voriconazole has many interactions with other drugs that may influence its escretion and require the modulation of the daily dose. Posaconazole (PSZ), a second generation azole, has a spectrum of activity even broader than voriconazole, being active also against zygomycetes, and a better safety profile. The published data on pediatric use are limited to 3 studies [25][26][27] one assessed the pharmacokinetic in a group of patients of 8-12 yearold, one described the clinical response in 12 patients treated as salvage therapy and the last one evaluated the clinical response on 15 pediatric patients with proven/probable infections. This last study reported an efficacy rate of 57% against zygomycetes, 75% against moulds, 50% against Aspergillus and 50% against candida strains. 26 Considering that posaconazole is currently not indicated for the pediatric age and the limited data available, we scored it as CIII, both for IA, candidiasis, and other mould infections.

Echinocandins
This new class of drugs inhibits membrane beta 1,3 D-glucan and in fact blocks membrane formation and stabilisation. Efficacy and toxicity profiles are excellent although some strains of candida (Candida guilliermondii and parapsilosis) and rare fungi (Zygomycetes, Cryptococcus, Trichosporon, Fusarium, Scedosporium) intrinsincally less sensitive or even resistant. Caspofungin, the first of this new class of drugs, is associated with rare drug-drug interactions. One multi-centre study reported an 81% success rate of caspofungin as first-line therapy against documented candida infections and a 50% success rate as rescue therapy for IA. 28 We scored caspofungin as BII for the therapy of Candida and CII for the therapy of IA Micafungin, which was approved as first-line therapy against invasive and oropharyngeal candidiasis in the newborn and older children, was recently introduced into clinical onco-haematology practice. Micafungin resulted non inferior to liposomal amphotericin B in a recent prospective randomized study on invasive candidiasis. 29 Moreover, Denning et al. found a response rate of 50% to micafungin used alone or in combination with other antifungal drugs for IA. 30 We scored micafungin as BII for invasive candidiasis and CII for IA.

Combination therapy
Combination therapy (particularly polyenes and azoles or polyenes plus echinocandins) is used in several paediatric centres. Drug combinations are hypothesized to exert a synergic action because different mechanisms of action on the fungal cell are exploited, the spectrum of action is widened and complementary pharmacokinetic and dynamic properties come into play. Most of reports on combination therapy are retrospective [2][3][4]31,32 and showed that this strategy is safe and active both for IA, candidiasis, and other mould infection but evidence of superior efficacy than monotherapy has been found so far. Considering the higher costs, combination therapy is optional and a matter of further investigation, we scored its use as CII as second line treatment for IA and CIII for all other infections.
In attempting to orientate paediatric oncologists in appropriate choice of anti-fungal agents for their patients, the evidence provided in this review leads to the conclusion that the risks, benefits and costs of modern antifungal therapy in children need to be carefully weighed up, given the state of the art in adults and the few reports in paediatric patients.