Activity of Isavuconazole and Comparator Agents Against Pediatric Isolates From Invasive Fungal Infections

Abstract

Background

Invasive fungal infections (IFIs) are a rising problem among pediatric patients. Antifungal agents have data supporting treatment in pediatric patients of varying ages and infection types but in vitro susceptibility data for large collections of pediatric isolates are lacking, which limits empiric treatment. We describe in vitro antifungal susceptibility testing of 900 contemporary yeast and mold isolates collected from IFIs in patients ≤ 17 years of age as a part of the SENTRY Antifungal Surveillance Program.

Methods

A total of 900 isolates (743 yeasts and 157 molds) were collected from 28 countries between 2017 and 2024. Antifungal susceptibility testing was performed by broth microdilution according to Clinical and Laboratory and Standards Institute (CLSI) methodology.

Results

A total of 33.4% of isolates were from patients ≤ 1 year old, 22.1% from patients aged 2-5 years old, 24.4% from patients aged 6-12 years old, and 20.0% from patients aged 13-17 years old. In total, 63.0% of yeasts were recovered from patients ≤ 5 years old; 79.6% of molds were from patients ≥ 6 years old. Most yeast isolates (72.5%) were from bloodstream infections; most mold isolates (61.8%) were from the respiratory tract. By CLSI breakpoints, 88.5% and 87.5% of Aspergillus fumigatus isolates were susceptible to voriconazole and isavuconazole, respectively. Against Mucorales, the isavuconazole minimum inhibitory concentration (MIC)₅₀ value was 4 mg/L, whereas the MIC₅₀ values for voriconazole and posaconazole were > 8 mg/L. All agents had low MICs against most yeasts collected and there were few resistant or non-wildtype yeasts.

Conclusions

Broth microdilution among a diverse collection of isolates obtained from IFIs among pediatric patients showed low rates of resistance to echinocandins or azoles. Across species, the lowest MICs were observed for voriconazole and isavuconazole in yeasts. All triazoles were highly active against Aspergillus spp. Against non-Aspergillus molds, isavuconazole, posaconazole, and amphotericin B were the most active agents. These data can guide the antifungal treatment of IFIs in pediatric patients when susceptibility testing is unavailable.

INTRODUCTION

With current medical advances, there has been an increase in the prevalence of immunocompromised pediatric patients.¹ The risk of infection rises with increasing immunosuppression, particularly from uncommon pathogens such as fungi. In hospitalized pediatric patients, invasive fungal infections (IFIs) are the second most common attributed cause of sepsis and the most common infectious etiology of death following a diagnosis of cancer or after a transplant procedure.^2–4 The most common etiologies of IFI are Candida spp. and Aspergillus spp., but rates of Mucorales and Fusarium spp. have been rising.^5,6 IFIs drastically increases the risk of in-hospital mortality, with reported mortality rates of 15.8% for pediatric candidemia and 18% for invasive aspergillosis.⁶ Invasive mucormycosis has an even higher mortality rate among the pediatric population at 32%.⁷ There are typically few, if any, pediatric patients in clinical trials and there are no large-scale publications on susceptibility profiles of fungal isolates from pediatric patients. Therefore, the optimal management of pediatric IFI is difficult to determine and often inferred from data on adult patients, which may not be appropriate.

Amphotericin B is the most commonly prescribed agent for the treatment of these infections but comes with significant downsides, including renal toxicity, need for frequent electrolyte monitoring, and intravenous (IV)-only formulation. Posaconazole is approved for prophylaxis of Aspergillus and Candida infections or treatment of oropharyngeal candidiasis in patients 2 years of age or older if the delayed-release oral suspension is used, delayed-release tablets for 13 years of age or older, or 18 years of age or older for injection. Isavuconazole is now the only azole agent with FDA labeling for the treatment of IFIs in pediatric patients as young as 1 year old if the IV formulation is used and for 6 years of age or older for capsules.⁸ Isavuconazole was also approved by the European Medicines Agency (EMA) in 2024 for the treatment of pediatric IFI.⁹ The isavuconazole approvals were based on a phase 2 open-label, non-comparative study, which studied the use of isavuconazole for the treatment of IFI in patients from 1 to less than 18 years old and found that isavuconazole was well tolerated in children, with a successful treatment response of 54.8% for 31 patients.¹⁰ This approval provides an oral option for the treatment of invasive fungal infections in children, which is highly desirable to facilitate outpatient care.

Here, we looked at the in vitro activity of isavuconazole and comparator agents against isolates collected from IFI in pediatric patients (≤17 years old) worldwide from 2017 to 2024. Aspergillus spp. isolates with elevated minimum inhibitory concentrations (MICs) to 1 or more azoles were examined for alterations in CYP51 proteins, Candida spp. isolates with elevated MICs to 1 or more echinocandins were examined for alterations in FKS proteins.

METHODS

Organism Collection and Identification

A total of 900 non-duplicative isolates from IFI in patients ≤ 17 years old were collected globally from 2017 to 2024 as a part of the SENTRY Antimicrobial Surveillance Program (Table S1; Figure S1). Each site submitted a maximum of 40 isolates per year; 92.54% of the sites were academic medical centers or associated with a research institution. Sites were instructed to critically analyze fungal isolates to ensure they represented pathogenic organisms requiring antifungal therapy. A total of 72 medical centers provided isolates, including 25 in North America, 27 in Europe, 13 in the Asia-Pacific region, and 7 in Latin America. There were 28 different countries represented. Dimorphic fungi (Blastomyces spp., Coccidioides spp., Histoplasma spp., Paracoccidioides spp., Talaromyces spp., and Sporothrix spp.) and Cladophialophora bantiana were excluded from collection and testing due to biosafety concerns. Confirmatory identification of isolates was provided by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (Bruker, Billerica, MA, USA). If an isolate did not score ≥ 1.8 (yeast) or ≥ 2.0 (mold) by spectrometry, identification was performed using sequencing of 28S ribosomal subunit, β-tubulin (Aspergillus spp. only), and/or internal spacer regions (ITS, all organisms) as previously described.¹¹ All organisms are identified by their historical nomenclature for simplicity rather than updated identifications (ie, Candida glabrata rather than Nakaseomyces glabratus and Candida auris rather than Candidozyma auris).

Antifungal Susceptibility Testing

Antifungal susceptibility testing was performed using frozen-form broth microdilution panels as recommended in Clinical and Laboratory and Standards Institute (CLSI) M27 and M27M44S (yeast) and CLSI M38 and M38M51S (mold).^12–15 The interpretive criteria used were published by CLSI.^14–17 Quality control was performed with appropriate organisms, as indicated in CLSI M27 and M38, and readings were within range.^12,13

Whole Genome Sequencing and Gene Analysis

Mutations in the sterol 14α-demethylase-encoding gene (cyp51A and cyp51B) were analyzed for 14 Aspergillus spp. isolates that were non-wildtype (NWT) to one or more azoles as previously described.¹⁸ Sequences were compared to the GenBank sequences available under the accession numbers AAK73659.1 for cyp51A and AAK73660.1 for cyp51B. Some of these mutations have been previously published and analyzed.¹⁹

All 5 Candida spp. isolates that were NWT to 1 or more echinocandins were selected for sequencing and analysis of FKS alterations as previously described.^20,21

RESULTS

Organism and Patient Characteristics

Among the 900 tested fungal isolates, 721 (80.1%) were Candida spp., 22 (2.4%) were non-Candida yeasts, and 157 (13.9%) were molds, including 125 (125/157, 79.6%) Aspergillus spp. and 32 (32/157, 20.4%) other molds (Table S1). The most common yeast organism identified in the pediatric population was Candida albicans (32.7%), followed by C. parapsilosis (17.6%), C. tropicalis (9.1%), C. glabrata (5.8%), C. lusitaniae (5.2%), C krusei (1.8%), C. dubliniensis (1.7%), C. guilliermondii (1.2%), and C. orthopsilosis (1.2%; Table 1). There were also 5 Cryptococcus spp., 7 Trichosporon asahii, 3 Rhodotorula spp., 2 Saccharomyces cerevisiae, and 4 other yeasts represented (Table 1). The Aspergillus spp. included 96 A. fumigatus (10.7% of organisms), 12 Aspergillus section Flavi (1.4%), 7 Aspergillus section Nigri (0.8%), 6 Aspergillus section Terrei (0.7%), and 4 other Aspergillus spp. There were 32 other molds tested, including 9 Scedosporium spp. (1.0%), 8 Fusarium spp. (0.9%), 7 Mucorales (0.8%), and 8 other organisms (Table 1).

Table 1.

Infection Source of All Isolates Collected from Pediatric Patients in Surveillance Program

	Infection type
Organism/organism group	BSI	RTI	SSSI	IAI	UTI	Other	Total
Yeasts	539	37	27	6	25	109	743
Candida spp.	531	35	26	6	25	98	721
Candida albicans	193	21	12	1	15	52	294
Candida parapsilosis	135	4	3	2	1	13	158
Candida tropicalis	62	4	3	1	1	11	82
Candida glabrata	42	0	1	1	4	4	52
Candida lusitaniae	37	2	1	0	0	7	47
Candida krusei	13	0	1	0	0	2	16
Candida dubliniensis	10	1	1	0	0	3	15
Candida guilliermondii	8	1	1	0	0	1	11
Candida orthopsilosis	8	1	0	0	0	2	11
Other Candida spp.1	23	1	3	1	4	3	35
other non-Candida yeast2	8	2	1	0	0	11	22
Molds	2	97	26	0	0	32	157
Aspergillus spp.	0	86	16	0	0	23	125
Aspergillus fumigatus	0	71	12	0	0	13	96
Aspergillus section Flavi	0	7	1	0	0	4	12
Aspergillus section Nigri	0	2	2	0	0	3	7
Aspergillus section Terrei	0	2	1	0	3	3	9
Other Aspergillus spp.3	0	4	0	0	0	0	4
Fusarium spp4	1	2	3	0	0	2	8
Mucorales5	0	0	4	0	0	3	7
Scedosporium/Lomentospora spp.6	0	7	2	0	0	0	9
Other	1	2	1	0	0	4	8
Total	541	134	53	6	25	141	900

Abbreviations: BSI, bloodstream infection; IAI, intra-abdominal infection; RTI, respiratory tract infection; SSSI, skin and skin structure infection; UTI, urinary tract infection

For molds, isolates were from respiratory tract infections (61.8%), skin and skin structure infections (16.6%), bloodstream infections (1.3%), and other infection sites (20.4%; Table 1). For yeasts, isolates were from bloodstream infections (72.5%), respiratory tract infections (5.0%), skin and skin structure infections (3.6%), urinary tract infections (3.4%), intra-abdominal infections (0.8%), and other infection sites (14.7%; Table 1).

The largest proportion (33.4%) of isolates were from patients ≤ 1 year of age, while 22.1% of isolates were from patients 2-5 years old, 24.4% were from patients 6-12 years of age, and 20% of isolates were from patients 13-17 years old (Table 1). Most yeasts (63.0%) were from patients ≤ 5 years old. In contrast, 79.6% of molds were from patients ≥ 6 years old. Female patients provided 44.2% of isolates and male patients provided 55.2% of isolates; 5 isolates did not have patient gender indicated. A total of 34.8% isolates were categorized as community-acquired (defined as admitted ≤ 2 days prior to culture collection) and 50.9% were nosocomial (Table S2). In total, 57.2% of yeast isolates were nosocomial (67.2% for the subset of Candida spp.). For molds, 77.9% were community-acquired. There were 132 isolates that did not have admission dates defined and were not included in either the nosocomial or community-acquired groups. Isolates were from diverse service lines (Table S3).

Antimicrobial Susceptibility Testing of Aspergillus spp.

Of the 125 Aspergillus spp., the MIC distributions and MIC_50/90 values are in Table 3. Using CLSI breakpoints,²² 88.5% and 87.5% of the 96 A. fumigatus isolates were susceptible to voriconazole and isavuconazole, respectively; 6.2% and 7.3% were intermediate to voriconazole and isavuconazole, respectively; 5.2% (5 isolates) were resistant to both agents. By epidemiological cutoff value (ECV), 94.8% of A. fumigatus isolates were wildtype to voriconazole, 87.5% were wildtype to isavuconazole, and 100.0% were wildtype to amphotericin B. All 12 Aspergillus section Flavi isolates (100.0%) were wildtype by ECV for voriconazole, isavuconazole, posaconazole, and amphotericin B. The 7 Aspergillus section Nigri were wildtype by ECV for all antifungal agents.

Table 3.

Antimicrobial Activity of Antifungal Agents in mg/L Tested Against Molds Using the CLSI Method^a

	Anidulafungin		Micafungin		Voriconazole		Isavuconazole		Posaconazole		Itraconazole		Amphotericin B
Organism (n)	MEC50/90	Range (%S, %WT)	MEC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)
Molds (157)	0.015/>4	≤0.002 to >4	≤0.008/>4	≤0.002 to >4	0.5/4	0.12 to >8	1/8	0.12 to >8	0.25/2	0.03 to >8	1/>8	0.12 to >8	1/2	≤0.03 to >4
Aspergillus spp. (125)	0.015/0.03	≤0.002 to 0.12	≤0.008/0.015	≤0.002 to 0.06	0.5/1	0.12 to 4	0.5/2	0.12 to 8	0.25/0.5	0.06 to 1	1/2	0.25 to 4	1/2	0.25 to >4
Aspergillus fumigatus (96)	0.015/0.03	≤0.008 to 0.06 (NA, NA)	≤0.008/0.015	≤0.008 to 0.03 (NA, NA)	0.5/1	0.12 to 4 (88.5, 94.8)	0.5/2	0.25 to 8 (87.5, 87.5)	0.25/0.5	0.06 to 1 (NA, NA)	1/2	0.25 to 4 (NA, 87.5)	1/2	0.25 to 2 (NA, 100.0)
Aspergillus section Flavi (12)	0.004/0.015	0.004 to 0.03 (NA, NA)	0.004/0.015	≤0.002 to 0.03 (NA, NA)	0.5/1	0.25 to 1 (NA, 100.0)	0.5/1	0.25 to 1 (NA, 100.0)	0.25/0.5	0.12 to 0.5 (NA, 100.0)	0.5/0.5	0.25 to 1 (NA, 100.0)	2/2	1 to 2 (NA, 100.0)
Aspergillus section Nigri (7)	0.004/NA	≤0.002 to 0.008 (NA, NA)	0.004/NA	≤0.002 to 0.015 (NA, NA)	0.5/NA	0.5/2 (NA, 100.0)	2/NA	0.5 to 4 (NA, 100.0)	0.5/NA	0.25 to 0.5 (NA, 100.0)	1/NA	0.5 to 4 (NA, 100.0)	0.25/NA	0.25 to 1 (NA, 100.0)
Other Aspergillus spp. (10)b	0.015/0.05	≤0.002 to 0.12	0.004/0.015	≤0.002 to 0.06	0.25/0.5	0.25 to 2	0.25/0.5	0.12 to 4	0.25/0.5	0.12 to 0.5	0.5/1	0.25 to 2	2/4	1 to >4
Fusarium spp. (8)c	>4/NA	>4	>4/NA	>4	4/NA	2 to 8	8/NA	4 to >8	2/NA	1 to >8	>8/NA	>8	2/NA	0.5 to >4
Mucorales (7)d	>4/NA	>4	>4/NA	>4	>8/NA	4 to >8	4/NA	1 to >8	>8/NA	0.25 to >8	8/NA	0.5 to >8	0.25/NA	≤0.03 to 0.5
Scedosporium/Lomentospora spp. (9)e	2/NA	1 to >4	0.25/NA	0.06 to >4	1/NA	0.5 to >8	8/NA	2 to >8	2/NA	1 to >8	>8/NA	4 to >8	>4/NA	1 to >4
Other molds (8)f	0.06/NA	≤0.002 to 0.25	0.03/NA	≤0.002 to 2	0.5/NA	0.12 to >8	0.5/NA	0.25 to >8	0.12/NA	0.03 to 2	0.25/NA	0.12 to >8	0.5/NA	0.12 to >4

Abbreviations: CLSI, Clinical and Laboratory and Standards Institute; ECV, epidemiological cutoff value; MIC, minimum inhibitory concentration; WT, wildtype.

^aNA indicates either there are no breakpoints or ECVs available or, for organisms with less than 10 isolates, the MIC₉₀ value cannot be calculated.

^bIncludes Aspergillus alabamensis (1), A. caelatus (1), A. lentulus (1), A. ochraceus species complex (1), A. terreus (4), A. terreus species complex (1), and A. unguis (1).

^cIncludes Fusarium oxysporum species complex (4), F. solani species complex (2), and Gibberella fujikuroi species complex (2).

^dIncludes Lichtheimia corymbifera (1), Mucor circinelloides (1), Rhizopus microsporus (2), Mucor spp. (2), and Syncephalastrum spp. (1).

^eIncludes Lomentospora prolificans (2), Scedosporium apiospermum (1), S. apiospermum/Scedosporium boydii (5), and Shigella boydii (1).

^fIncludes Alternaria alternata (1), Exophiala dermatitidis (1), Penicillium citrinum (1), Purpureocillium lilacinum (1), Rasamsonia argillacea (1), and Curvularia spp. (3).

Antimicrobial Susceptibility Testing of Other Molds

For the non-Aspergillus molds, the isavuconazole MIC₉₀ was > 8 mg/L with a range from 0.25 to > 8 mg/L (Table 3). The MIC₉₀ was the same for all agents (above the highest dilution tested) although the highest dilution tested for amphotericin B was 4 mg/L and for the other agents it was 8 mg/L. Of the 7 Mucorales (Lichtheimia corymbifera, Mucor circinelloides, Rhizopus microsporus, Mucor spp., and Syncephalastrum spp.), the isavuconazole MIC range was from 1 to > 8 mg/L; for amphotericin B, it was ≤ 0.03-0.5 mg/L. Five of the Mucorales isolates had MICs > 8 mg/L for voriconazole.

Whole Genome Sequencing of NWT Aspergillus fumigatus Isolates

There were 14 A fumigatus isolates that were NWT to one or more azoles. WGS was performed to identify alterations in the cyp51 genes (Table 5). Five (5/14, 35.7%) did not have identified amino acid changes in CYP51A or CYP51B. Of the isolates with identified CYP51 changes, 3 had only CYP51B Q42L and were all from New Jersey, USA. There were 4 isolates with CYP51A TR₃₄/L98H; these isolates were from Belgium (1 isolate) and Italy (3 isolates). There was 1 isolate from Indiana, USA with CYP51A I242V. There was 1 isolate from Vermont, USA with CYP51A F46Y, M172V, N248T, D255E, and E427K.

Table 5.

Summary of CYP Alterations Detected Among Azole-NWT Aspergillus spp. Isolates and FKS Alterations Among Echinocandin-NWT Candida spp.

State and/or country	Organism	MEC/MIC according to CLSI method (ug/mL)								CYP51 or FKS alterations a
		Anidulafungin	Caspofungin	Micafungin	Isavuconazole	Voriconazole	Itraconazole	Posaconazole		CYP51A		CYP51B
Italy	Aspergillus fumigatus	0.015	0.015	≤0.008	2	0.5	1	0.5		Wild type		Wild type
Italy	A fumigatus	0.015	0.06	≤0.008	2	0.5	0.5	0.25		Wild type		Wild type
Italy	A fumigatus	0.03	0.015	≤0.008	8	4	4	1		Wild type		Wild type
IN, USA	A fumigatus	0.008	0.03	0.015	1	1	2	1		I242V		Wild type
Belgium	A fumigatus	0.008	0.03	0.015	4	2	4	1		L98H,TR34		Wild type
Italy	A fumigatus	≤0.002	0.004	≤0.002	4	2	4	1		L98H,TR34		Wild type
Italy	A fumigatus	0.008	0.008	0.008	4	1	4	0.5		L98H,TR34		Wild type
Italy	A fumigatus	0.008	0.008	0.004	4	2	4	1		L98H,TR34		Wild type
NJ, USA	A fumigatus	0.004	0.008	0.004	2	1	2	0.5		Wild type		Q42L
NJ, USA	A fumigatus	0.015	0.008	0.008	2	1	2	0.5		Wild type		Q42L
VT, USA	A fumigatus	0.03	0.015	0.015	1	0.5	2	0.5		F46Y, M172V, N248T, D255E, E427K		Wild type
NJ, USA	A fumigatus	0.03	0.03	0.008	2	2	4	0.5		Wild type		Q42L
NJ, USA	A fumigatus	0.03	0.03	0.008	2	1	2	0.5		Wild type		Wild type
VA, USA	A fumigatus	0.03	0.06	0.015	1	0.5	2	0.5		Wild type		Wild type
		Anidulafungin	Caspofungin	Micafungin	Isavuconazole	Voriconazole	Fluconazole	Posaconazole		fks1 HS1	fks1 HS2	fks2 HS1	fks2 HS2
Hungary	Candida albicans	0.015	0.03	0.06	≤0.008	≤0.008	≤0.12	0.03		R647G	Wild type	NT	NT
Hungary	Candida glabrata	1	1	0.5	0.5	0.25	16	0.5		Wild type	Wild type	F659_del	Wild type
Taiwan	C albicans	0.25	0.25	0.25	0.004	0.004	0.12	0.03		Wild type	R1361G	NT	NT
VA, USA	C albicans	0.5	2	2	≤0.002	0.004	0.25	0.03		S645P	Wild type	NT	NT
FL, USA	C albicans	0.12	0.12	0.06	≤0.002	≤0.002	0.06	0.008		Wild type	Wild type	NT	NT

^aNT, not tested.

Gray indicates an isolate that was NWT by ECV.

Antimicrobial Susceptibility Testing of Yeast

For the 743 yeasts that were tested, MIC_50/90 values for all tested agents are shown in Table 4.

Table 4.

Antimicrobial Activity of Antifungal Agents in mg/L Tested Against Yeasts Using the CLSI Method^a

	Anidulafungin		Micafungin		Fluconazole		Voriconazole		Isavuconazole		Posaconazole		Itraconazole		Amphotericin B
Organism (n)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)	MIC50/90	Range (%S, %WT)
Yeast (743)	0.06/2	≤0.008 to >4	0.03/1	≤0.002 to >4	0.25/4	≤0.12 to >128	≤0.008/0.06	≤0.008 to 2	≤0.008/0.06	≤0.008 to 2	0.06/0.25	0.004 to 1	0.12/0.5	≤0.06 to 4	0.5/1	0.12 to 2
Candida spp. (721)	0.06/2	≤0.008 to >4	0.03/1	≤0.002 to >4	0.25/4	≤0.12 to >128	≤0.008/0.06	≤0.008 to 8	≤0.008/0.06	≤0.008 to 2	0.06/0.25	0.004 to 1	0.12/0.5	≤0.06 to 4	0.5/1	0.12 to 2
Candida albicans (294)	0.015/0.06	≤0.008 to 0.5 (99.7, 99.3)	0.015/0.03	≤0.008 to 2 (99.7, 98.6)	≤0.12/0.25	≤0.12 to 2 (100.0, 98.0)	≤0.008/0.06	≤0.008 to 0.25 (99.7, 98.6)	≤0.008/≤0.008	≤0.008 to 0.03 (NA, NA)	0.03/0.06	0.004 to 0.12 (NA/98.0)	≤0.06/0.12	≤0.06 to 0.25 (NA/NA)	0.5/1	0.25 to 1 (NA, 100.0)
Candida parapsilosis (158)	2/2	0.015 to 4 (90.5, 100.0)	1/2	0.015 to 2 (100.0, 100.0)	0.5/1	≤0.12 to 32 (98.1, 98.1)	≤0.008/0.015	≤0.008 to 0.25 (98.7, 96.8)	≤0.008/0.015	≤0.008 to 0.12 (NA, NA)	0.06/0.12	0.015/0.25 (NA, 100.0)	0.12/0.25	0.015 to 0.5 (NA, 100.0)	0.5/1	0.25 to 1 (NA, 100.0)
Candida tropicalis (82)	0.03/0.06	0.004 to 0.12 (100.0, 100.0)	0.03/0.06	0.008 to 0.12 (100.0, 98.8)	0.25/0.5	≤0.12 to 64 (98.8, 97.6)	0.03/0.03	≤0.008 to 1 (98.8, 98.8)	0.015/0.03	≤0.008 to 0.25 (NA, NA)	0.06/0.12	0.015 to 0.5 (NA, 97.6)	0.12/0.25	≤0.06 to 1 (NA, 98.8)	1/1	0.25 to 1 (NA, 100.0)
Candida glabrata (52)	0.06/0.12	0.008 to 1 (98.1, 98.1)	0.015/0.03	≤0.008 to 0.5 (98.1, 98.1)	4/4	0.25 to 32 (100.0, 92.3)	0.06/0.12	≤0.008 to 1 (NA, 94.2)	0.06/0.12	≤0.008 to 0.5 (NA, NA)	0.5/0.5	0.06 to 1 (NA, 100.0)	0.5/1	≤0.06 to 1 (NA, 100.0)	1/1	0.25 to 1 (NA, 100.0)
Candida lusitaniae (47)	0.25/0.5	0.06 to 0.5 (NA, 100.0)	0.12/0.25	0.03 to 0.25 (NA, 100.0)	0.25/1	≤0.12 to 64 (NA, 93.6)	≤0.008/0.015	≤0.008 to 0.5 (NA, NA)	≤0.008/0.03	≤0.008 to 0.5 (NA, NA)	0.06/0.12	0.008 to 0.5 (NA, 83.0)	0.12/0.25	0.06 to 1 (NA, 100.0)	0.25/0.5	0.12 to 1 (NA, 100.0)
Candida krusei (16)	0.03/0.06	0.008 to 0.12 (100.0, 100.0)	0.06/0.12	0.06 to 0.25 (100.0, 100.0)	32/64	8 to 64 (NA, NA)	0.25/0.5	0.06 to 0.5 (100.0, 100.0)	0.25/0.5	0.03 to 0.5 (NA, NA)	0.25/0.5	0.03 to 0.5 (NA, 100.0)	0.25/0.5	0.06 to 1 (NA, 100.0)	1/2	1 to 2 (NA, 100.0)
Candida dubliniensis (15)	0.06/0.12	0.015 to 0.12 (NA, 100.0)	0.03/0.06	≤0.002 to 0.12 (NA, 100.0)	≤0.12/0.25	≤0.12 to 2 (NA, 93.3)	≤0.008/0.015	≤0.008 to 0.015 (NA, NA)	≤0.008/≤0.008	≤0.008 to 0.06 (NA, NA)	0.03/0.12	0.015 to 0.12 (NA, 100.0)	0.06/0.25	0.015 to 0.25 (NA, 100.0)	0.25/0.5	0.12 to 0.5 (NA, 100.0)
Candida guilliermondii (11)	2/4	1 to 4 (81.8, 100.0)	0.5/1	0.5 to 1 (100.0, 100.0)	4/128	1 to >128 (NA, 81.8)	0.06/2	0.03 to 8 (NA, NA)	0.12/2	0.06 to 2 (NA, NA)	0.25/1	0.12 to 1 (NA, 72.7)	0.5/4	0.25 to 4 (NA, 81.8)	0.5/0.5	0.5 to 1 (NA, 100.0)
Candida orthopsilosis (11)	0.5/1	0.12 to 2 (NA, 100.0)	0.5/1	0.12 to 1 (NA, 100.0)	0.5/1	0.25 to >128 (NA, 90.9)	0.015/0.03	0.008 to 2 (NA, 90.9)	0.015/0.06	0.008 to 1 (NA, NA)	0.06/0.12	0.03 to 0.5 (NA, 90.9)	0.12/0.25	0.06 to 2 (NA, 90.9)	0.5/1	0.25 to 2 (NA, 100.0)
Other yeast (22)	>4/>4	0.03 to >4	>4/>4	0.03 to >4	4/128	1 to >128	0.12/1	0.015 to 2	0.12/1	0.015 to 2	0.25/1	0.03 to 1	0.5/1	0.03 to 1	1/1	0.25 to 2

Abbreviations: CLSI, Clinical and Laboratory and Standards Institute; ECV, epidemiological cutoff value; MIC, minimum inhibitory concentration; WT, wildtype.

^aNA, indicates no breakpoint or ECV available.

For C. albicans, the highest observed MIC for isavuconazole was 0.03 mg/L, which is lower than for any comparator agents (Table 4). For the comparator agents, 99.7% were susceptible to anidulafungin and micafungin, 100% were susceptible to fluconazole, and 99.7% were susceptible to voriconazole. There are no breakpoints for posaconazole or amphotericin B, but 98.0% and 100.0% were wildtype to each agent, respectively. There are no breakpoints or ECVs for isavuconazole for any yeast species due to lack of regulatory approval for treatment.

In the 158 C. parapsilosis isolates, a similar pattern was observed (Table 4). A total of 100.0% of isolates were susceptible to micafungin, 90.5% to anidulafungin, 98.1% to fluconazole, and 98.7% to voriconazole; whereas 100.0% of isolates were wildtype for posaconazole and amphotericin B. Again, there is no breakpoint or ECV for isavuconazole against C. parapsilosis due to lack of regulatory approval for treatment.

For the 82 C. tropicalis isolates (Table 4), 100.0% of isolates were susceptible to anidulafungin and micafungin and 98.8% were susceptible to fluconazole and voriconazole. By ECV, 98.8% were wild-type for micafungin and voriconazole, 97.6% were wild-type for fluconazole and posaconazole, and 100.0% were wild-type for anidulafungin and amphotericin B. Again, no interpretative criteria were available for isavuconazole, but MICs were similar to or lower than comparators.

Anidulafungin and micafungin were active against 98.1% of the 52 C. glabrata isolates using the CLSI breakpoints (Table 4), caspofungin was active against 96.1%, and 100.0% of isolates were susceptible-dose-dependent to fluconazole. By ECV, 100.0% of isolates were wildtype for posaconazole and amphotericin B, 98.1% were wildtype for anidulafungin and micafungin, 94.2% were wildtype for voriconazole, and 92.3% were wildtype for fluconazole.

Among 47 C. lusitaniae, no breakpoints are available for this organism. By ECV, 100.0% were wildtype to anidulafungin, micafungin, and amphotericin B, 93.6% were wildtype to fluconazole, and 83.0% were wildtype to posaconazole.

For Candida spp. with smaller numbers represented (Table 1, C. krusei, C. dubliniensis, C. guilliermondii, and C. orthopsilosis) isavuconazole MICs were also similar to or lower than azole comparators but a higher percentage of azole NWT were observed among C. guilliermondii and C. orthopsilosis. There were no C. auris isolates collected from pediatric patients in the surveillance program.

In the 22 non-Candida yeasts, isavuconazole and voriconazole maintained low MICs (Table 4). There were 8 Trichosporon spp. tested (7 Trichosporon asahii and 1 Trichosporon ovoides, data not shown). For these isolates, isavuconazole MIC₅₀ was 0.12 mg/L; this was lower than fluconazole, echinocandins, and amphotericin B. There were 5 Cryptococcus spp. (1 C. laurentii, 3 C. neoformans var. grubii, and 1 C. neoformans var. neoformans, data not shown) with an isavuconazole MIC₅₀ of 0.03 mg/L, lower than all comparator agents. All Cryptococcus spp. isolates had MICs > 4 mg/L for echinocandins and were wildtype for fluconazole by ECV. Minimum inhibitory concentrations for isavuconazole were the lowest for all agents against 3 Rhodotorula spp. (2 R. mucilaginosa and 1 Rhodotorula spp.); all micafungin MICs were > 4 mg/L and fluconazole MICs were ≥ 32 mg/L (data not shown).

Whole Genome Sequencing and FKS Analysis of Candida spp.

Five Candida spp. were NWT for 1 or more echinocandin and had WGS performed (Table 5). These 5 isolates included 4 C. albicans and 1 C. glabrata. One of the 4 C. albicans isolates did not have changes detected in FKS1. Of the C. albicans isolates with alterations detected, 1 was from Hungary with R647G in FKS1 HS1, 1 was from Taiwan with R1361G in FKS1 HS2, and 1 was from Virginia, USA with S645P in FKS1 HS1. The C. glabrata isolate had a F659 deletion in FKS2 HS1. The isavuconazole MICs among these isolates ranged from ≤ 0.002 to 0.5 mg/L, which was comparable to voriconazole and lower than fluconazole and posaconazole.

DISCUSSION

We presented contemporary surveillance data on IFI, including both yeasts and molds, from all infection sources in patients under 18 years of age. There are limited prior data on large collections of fungal isolates from pediatric patients. Older publications include 203 bloodstream infections (BSIs) from Spain from 2009 to 2010 with only yeast identified²³ and approximately 400 isolates from 1997 to 2000 again with only yeast from BSI included.²⁴ These data provide essential information on in vitro susceptibility patterns of yeasts and molds, including multiple antifungal classes encompassing those with indications for pediatric infection treatment.

Organism Prevalence and Characteristics

In our 8-year surveillance of fungal isolates from IFI in pediatric patients, we found that yeasts were more common than molds (Table 1). Among yeasts, C. albicans was the most common organism, followed by C. parapsilosis and C. tropicalis. This result is notably different from the epidemiology in adults, where C. glabrata is the second most common etiology of IFI.²⁵ However, multiple studies have found that C. parapsilosis is the second most common yeast in IFI among pediatric patients and the highest risk of C. parapsilosis infection development is in patients ≤ 2 years old.^24,26–28 This outcome has been postulated to be the result of higher colonization rates of C. parapsilosis among neonates, but its exact etiology remains unknown. Most Candida spp. infections were from BSIs (73.6%), which is consistent with their main route of infection, intravascular devices. In contrast, most mold infections were from respiratory tract infections (61.8%), which is consistent with the inhalational route of infection.

The organism split by patient age was consistent with prior studies. In patients ≤ 1 years old, 96.3% of IFIs were from yeasts; in patients 2-5 years old, 89.4%; in patients 6-12 years old, 70.9%; and, in patients 13-17 years old, 66.1% (Table 2). This outcome likely reflects the different exposure risks among the different age groups. Intravascular devices among premature or low-birth weight infants represent the main risk factors for IFI among pediatric patients ≤ 2 years old. Risk factors for IFI among older pediatric patients are related to immunocompromising conditions for which fluconazole prophylaxis is prescribed, leading to infections from fluconazole-resistant environmental agents such as molds.

Table 2.

Age Group and Organism Breakdown of Pediatric Isolates Collected from Invasive Fungal Infections

Organism/organism group	Patients ≤ 1 years old	Patients 2-5 years old	Patients 6-12 years old	Patients 13-17 years old	Total
All organisms	301	199	220	180	900
Yeasts	290	178	156	119	743
Candida spp.	285	174	152	110	721
Candida albicans	123	67	64	40	294
Candida parapsilosis	68	43	28	19	158
Candida tropicalis	20	23	17	22	82
Candida glabrata	20	10	15	7	52
Candida lusitaniae	27	11	5	4	47
Candida krusei	2	5	6	3	16
Candida dubliniensis	2	3	6	4	15
Candida guilliermondii	5	3	1	2	11
Candida orthopsilosis	4	4	0	3	11
Other Candida spp.a	14	5	10	6	35
other non-Candida yeastb	5	4	4	9	22
Molds	11	21	64	61	157
Aspergillus spp.	9	13	53	50	125
Aspergillus fumigatus	6	11	51	38	96
Aspergillus section Flavi	0	1	7	4	12
Aspergillus section Nigri	3	0	1	3	7
Aspergillus section Terrei	0	0	3	3	6
Other Aspergillus spp.c	0	1	1	2	4
Fusarium spp.d	0	3	4	1	8
Mucoralese	2	2	1	2	7
Scedosporium/Lomentospora spp.f	0	1	4	4	9
Other molds (8)g	0	2	1	5	8
Total	301	199	220	180	900

^aIncludes Candida fabianii (3), C. fermentati (3), C. haemulonii (1), C. inconspicua (1), C. intermedia (1), C. kefyr (6), C. lipolytica (2), C. metapsilosis (5), C. nivariensis (2), C. pararugosa (3), C. pelliculosa (1), C. pseudohaemulonii (1), C. rugosa (3), C. spencermartinsiae (1), C. sphaerica (1), and C. utilis (1)

^bIncludes Blastobotrys adeninivorans (1), Cryptococcus laurentii (1), C neoformans var. grubii (3), C neoformans var. neoformans (1), Cyberlindnera rhodanensis (1), Geotrichum clavatum (1), Magnusiomyces capitatus (1), Rhodotorula mucilaginosa (2), Saccharomyces cerevisiae (2), Trichosporon asahii (7), T. ovoides (1), and Rhodotorula spp. (1)

^cIncludes Aspergillus caelatus (1), A. lentulus (1), A. ochraceus species complex (1), and A. unguis (1).

^dIncludes Fusarium oxysporum species complex (4), F. solani species complex (2), and Gibberella fujikuroi species complex (2).

^eIncludes Lichtheimia corymbifera (1), Mucor circinelloides (1), Rhizopus microsporus (2), Mucor spp. (2), and Syncephalastrum spp. (1).

^fIncludes Lomentospora prolificans (2), Scedosporium apiospermum (1), S. apiospermum/S. boydii (5), and S. boydii (1).

^gIncludes Alternaria alternata (1), Exophiala dermatitidis (1), Penicillium citrinum (1), Purpureocillium lilacinum (1), Rasamsonia argillacea (1), and Curvularia spp. (3).

Yeasts were more commonly found in nosocomial infections; molds in community-acquired infections. This result is also a reflection on the risk factors associated with each infection. Invasive fungal infections due to yeast are associated with disruption of normal barriers by intravascular devices, urinary catheters, or mechanical ventilation.²⁶ In contrast, mold IFI are due to an underlying immunocompromising condition, but not typically to specific hospital-associated interventions.

In contrast to prior reports,¹ we did not see an increase in mold isolates collected over the years of this study (Figure S2). There was year-to-year variability in the number of Aspergillus spp. and other molds, but no trend of increase over time. We did note higher numbers of Candida spp. collected in 2017, 2018, and 2019, at more than 100 isolates annually in those years, compared to 60-80 isolates annually in the other years. There has been a decrease in dedicated children’s hospitals participating in the SENTRY Surveillance Program since 2019, which may explain this reduction in pediatric isolate number as the overall number of fungal isolates has remained unchanged.

Antimicrobial Susceptibility Testing and Whole Genome Sequencing of Molds

Often it is difficult to obtain antimicrobial susceptibility testing of molds due to their slower growth rate and the limited number of reference labs with specialized expertise in mold broth microdilution testing. Additionally, guidelines recommend antifungal susceptibility testing for Aspergillus isolates only for patients suspected to have an azole-resistant isolate, who are unresponsive to antifungal agents, or for epidemiological purposes.²⁹ Unlike bacteria, there are no automated test systems, approved gradient diffusion strips, or approved Kirby–Bauer disks for mold susceptibility testing. Therefore, large surveillance studies such as this study are useful in providing data about empiric antifungal choices for the treatment of IFI. Without this type of data, amphotericin B is often the empiric treatment of choice. Amphotericin B has significant downsides in terms of its toxicity and its lack of an oral route of administration. Additionally, given the high mortality rate of IFI, which increases with a delay in appropriate therapy, it is imperative to choose an active agent for treatment.

From our susceptibility testing of 157 mold isolates collected from IFI among pediatric patients, the lowest MICs were observed for amphotericin B and posaconazole when all isolates were examined together (Table 3), similar to prior publications that did not separate isolates by age of patient.³⁰ This result is also consistent with EUCAST guidance, in which the posaconazole breakpoints are several dilutions below the isavuconazole and voriconazole breakpoints against same mold species.³¹ Since the posaconazole breakpoints are lower, it is important to compare the isolates by the percentage susceptible or wildtype rather than the MIC values. However, there are no breakpoints or ECVs for posaconazole against A. fumigatus due to interlaboratory variability of antimicrobial susceptibility testing and concern about the inability to separate wildtype and non-wildtype isolates.^32,33

Our data suggest the following recommendations for pediatric patients based only on MICs, ECVs, and breakpoints. Isavuconazole, voriconazole, or amphotericin B would be the preferred agent of treatment for Aspergillus spp. infections. Isavuconazole or amphotericin B would be the preferred treatment for Mucorales. Posaconazole or amphotericin B would be the preferred treatment for Fusarium spp. Voriconazole or posaconazole would be the preferred treatment for Scedosporium/Lomentospora spp. For the rare molds, dual empiric therapy with any azole and amphotericin B while always performing susceptibility testing is likely preferred given the range of MICs observed across all of the antifungal agents.

Additional factors contribute to antifungal choice, including dosing route, side effect profile, FDA indication, and ease of dosing or need for therapeutic monitoring. Posaconazole is approved at ≥ 2 years of age but only for prophylaxis of Aspergillus and Candida infections in contrast to isavuconazole at ≥ 1 year of age for treatment of Aspergillus and Mucorales infections. Posaconazole is also difficult to dose and requires therapeutic drug monitoring unlike isavuconazole, which has more predictable pharmacokinetic and pharmacodynamic parameters in both adult and pediatric patients.^10,34,35 Posaconazole pharmacokinetics have been explored in pediatric patients and it has been found to be safe with better exposure for intravenous formulation compared to an oral suspension.³⁶ Therefore, although posaconazole MICs trend lower, the lack of breakpoints, lack of approval for IFI treatment, and dosing difficulty make isavuconazole the preferred azole therapeutic agent. Despite its wide-spectrum of activity, amphotericin B including lipid formulations is also less preferred due to its IV-only formulation and toxicity profile.

Antimicrobial Susceptibility Testing and Whole Genome Sequencing of Yeasts

Despite having one of the lowest MIC_50/90 values against the yeast isolates from pediatric IFI (Table 4), isavuconazole is not included in Infectious Disease Society of America (IDSA) treatment guidelines for candidemia or the US Department of Health and Human Services guidelines on Candida in children living with HIV and there is no approved indication for the treatment of candidemia or invasive candidiasis with isavuconazole.^37,38 This is partially based on a prior phase 3 randomized double-blind clinical trial on the treatment of patients with candidemia or invasive candidiasis with either isavuconazole or caspofungin (ACTIVE trial) that did not show non-inferiority of isavuconazole for the treatment of these infections.³⁹ Therefore, the IDSA treatment guidelines recommend only fluconazole, echinocandins, amphotericin B, or voriconazole as options for the management of candidiasis. In contrast, isavuconazole is marginally recommended after echinocandins, fluconazole, voriconazole, and liposomal amphotericin B in the global guidelines for the management of candidiasis from the European Confederation of Medical Mycology, International Society for Human & Animal Mycology, and the American Society for Microbiology.⁴⁰ The marginal recommendation for isavuconazole is based on case reports and case series showing the successful cure of a variety of candidal infections with isavuconazole treatment. Both itraconazole and posaconazole are discouraged by these guidelines if alternatives are available. Isavuconazole and posaconazole are included as alternatives to fluconazole for refractory esophageal candidiasis in the US Department of Health and Human Services guidelines on Candida in children living with HIV.³⁸

Minimum inhibitory concentrations for voriconazole, isavuconazole, and posaconazole were notable for the non-Candida yeasts where both echinocandin and fluconazole MICs were off-scale high for many organisms. Isavuconazole may be a desirable agent for treatment of Cryptococcus spp. given its favorable drug–drug interaction profile compared to the other azoles. A prior phase 3 clinical trial also showed successful treatment of Cryptococcus infections with isavuconazole.⁴¹

This study had several advantages: isolates were collected over a multiyear period, including 2024; isolates were from a wide geographic spread; and isolates were relatively equally distributed across the pediatric age ranges. However, we are only able to provide in vitro susceptibility results and there is no data on what treatment was provided to the patients or their outcomes, which limits our conclusions. Additionally, there are a small number of non-Aspergillus molds and non-Candida yeasts included in this study, making inferences about susceptibility patterns among these genera difficult. Future work should focus on these less common organisms to create a more robust database of susceptibility patterns.

Conclusions

Reference broth microdilution susceptibility testing was performed on 900 fungal isolates from IFI in patients ≤ 17 years old from 2017 to 2024. Yeast IFI were more common than mold IFI. Among molds, Aspergillus spp. were the most common organisms and triazoles were highly active. Few isolates among the pediatric patients were resistant or NWT to azoles (Aspergillus spp.) or echinocandins (Candida spp.). The lowest MICs were observed for voriconazole and isavuconazole against Candida spp. and non-Candida yeasts. These data will be useful for empiric treatment choices among pediatric patients with IFI in situations where amphotericin B or echinocandins cannot be used.

Author contributions

M.L.W. was involved in conceptualization, data curation, formal analysis, methodology, and writing. P.R. was involved in methodology, project administration, data curation, and writing. K.F. was involved in data curation and methodology. A.K. was involved in methodology and data curation. S.E. was involved in methodology and data curation. M.C. was involved in conceptualization, data curation, formal analysis, methodology, and writing.

Funding

This study was performed by Element Iowa City (JMI Laboratories) and sponsored by Pfizer. Medical writing support for the development of this manuscript, under the guidance of the authors, was provided by Element Iowa City (JMI Laboratories) and funded by Pfizer. Element Iowa City was contracted to perform services in 2024 for AbbVie, Inc., Adamed Pharma S.A., AimMax Therapeutics, Allecra Therapeutics, Amicrobe, Inc., AN2 Therapeutics, Inc., Apnimed, Astellas Pharma, Inc., Basilea Pharmaceutica AG, Baxis Pharmaceuticals, Inc., Beckman Coulter, Inc., bioMérieux, Biosergen AB, Blacksmith Medicines, Bugworks, Carnegie Mellon University, Center for Discovery and Innovation, Cerba Research NV, Cidara Therapeutics, Cipla USA Inc., ContraFect Corporation, CorMedix Inc., Crestone, Inc., Discoveric Bio Beta Ltd., Dr Falk Pharma Gmbh, Entasis Therapeutics, Evopoint Biosciences, Fedora Pharmaceuticals, Forge Therapeutics, GARDP Foundation, Genentech, Gilead Sciences, Inc., GlaxoSmithKline plc, Harvard University, Institute for Clinical Pharmacodynamics, Iterum Therapeutics plc, Janssen Biopharma, Johnson & Johnson, Kbio, Inc., Lakewood-Amedex Inc., Locus Biosciences, Inc., McGill University, Medpace, Inc., Meiji Seika Pharma, Melinta Therapeutics, Menarini Group, Merck & Co., MetCura Pharmaceuticals, Inc., MicuRx Pharmaceutical Inc., Mundipharma International Ltd., Nabriva Therapeutics, National Cancer Institute, National Institutes of Health, NovoBiotic Pharmaceuticals, LLC, Ohio State University, Omnix Medical Ltd., Paratek Pharmaceuticals, Pfizer, Inc., PPD Global Central Labs, LLC, Pulmocide Ltd., Qpex Biopharma, Inc., Revagenix, Inc., Roche Holding AG, Scynexis, Inc., Seed Health, SeLux Diagnostics, Shionogi & Co., Ltd., Spero Therapeutics, Sumitovant Biopharma, Inc., Swedish National Reference Laboratory, TenNor Therapeutics, Thermo Fisher Scientific, U.S. Food and Drug Administration, University of Wisconsin, VenatoRx Pharmaceuticals, Wockhardt, and Zoetis, Inc.

Conflicts of interest

At the time of writing all authors were employees of Element Iowa City (JMI Laboratories) which was a paid contractor to Pfizer in connection with the development of this manuscript and study design/management.