High-Risk Neutropenic Fever and Invasive Fungal Diseases in Patients with Hematological Malignancies

Invasive fungal diseases (IFDs) still represent a relevant cause of mortality in patients affected by hematological malignancies, especially acute myeloid leukaemia (AML) and myelodysplastic syndrome (MDS) undergoing remission induction chemotherapy, and in allogeneic hematopoietic stem cell transplantation (allo-HSCT) recipients. Mold-active antifungal prophylaxis (MAP) has been established as a standard of care. However, breakthrough IFDs (b-IFDs) have emerged as a significant issue, particularly invasive aspergillosis and non-Aspergillus invasive mold diseases. Here, we perform a narrative review, discussing the major advances of the last decade on prophylaxis, the diagnosis of and the treatment of IFDs in patients with high-risk neutropenic fever undergoing remission induction chemotherapy for AML/MDS and allo-HSCT. Then, we present our single-center retrospective experience on b-IFDs in 184 AML/MDS patients undergoing high-dose chemotherapy while receiving posaconazole (n = 153 induction treatments, n = 126 consolidation treatments, n = 60 salvage treatments). Six cases of probable/proven b-IFDs were recorded in six patients, with an overall incidence rate of 1.7% (6/339), which is in line with the literature focused on MAP with azoles. The incidence rates (IRs) of b-IFDs (95% confidence interval (95% CI), per 100 person years follow-up (PYFU)) were 5.04 (0.47, 14.45) in induction (n = 2), 3.25 (0.0013, 12.76) in consolidation (n = 1) and 18.38 (3.46, 45.06) in salvage chemotherapy (n = 3). Finally, we highlight the current challenges in the field of b-IFDs; these include the improvement of diagnoses, the expanding treatment landscape of AML with molecular targeted drugs (and related drug–drug interactions with azoles), evolving transplantation techniques (and their related impacts on IFDs’ risk stratification), and new antifungals and their features (rezafungin and olorofim).


Introduction
Major improvements have been achieved in the prophylaxis, treatment and outcome of invasive fungal diseases (IFDs); however, they still represent an important cause of mortality in patients affected by hematological malignancies (HM) who are undergoing remission induction chemotherapy for acute leukaemia and allogeneic hematopoietic stem cell transplantation (allo-HSCT).Patients with long-lasting neutropenia, such as patients with acute myeloid leukaemia (AML) or myelodysplastic syndrome (MDS), continue to represent the population at highest risk of developing neutropenic fever (NF) and IFDs [1].The most frequently identified fungal pathogens are Aspergillus spp.and Candida spp., which significantly contribute to mortality in these patients [2][3][4][5][6]; therefore, mold-active antifungal prophylaxis (MAP) has been established as a standard of care [7][8][9][10][11][12].However, breakthrough IFDs (b-IFDs) in patients receiving antifungal prophylaxis (AFP), including MAP, is an issue for clinicians, alongside with the rising resistance to antifungals and the continuous change in IFD epidemiology, due to both the increased use of mold-active drugs and the improved survival of HM patients.

Search Strategy
In this narrative review, we discuss the literature of the last decade that concerns IFDs in HM patients with NF who are undergoing induction chemotherapy for acute leukaemia and allo-HSCT.
A structured literature search in PubMed/MEDLINE database was performed in October 2023 (update freezing date: 31 October) of all English-written full articles with the following search terms: ("Hematopoietic Stem Cell Transplantation" OR "Acute Leukaemia") AND ("Invasive Fungal Infections" OR "Breakthrough Invasive Fungal Infections" OR "Antifungal Prophylaxis").Studies published prior to 2010 (except randomized controlled trials; RCTs), including pre-prints, abstracts, letters, reviews, commentary articles, opinion articles and animal studies were excluded.This timeframe was selected to include more recent studies that highlight an updated epidemiology and outcome of IFDs.Fifty studies were included, all assessed to be of moderate to high quality and with moderate risk of bias.
Then, we report the single-center experience of San Raffaele Scientific Institute on b-IFDs in AML/MDS patients undergoing induction, consolidation and salvage chemotherapy while receiving posaconazole.

Diagnosis and Definitions of Invasive Fungal Diseases
The diagnosis of IFDs, especially the mycological one, is a crucial issue for the management of patients with high-risk NF.A recent trial highlights the pursuit of diagnosis, demonstrating the safety of a pre-emptive strategy compared to empiric antifungal therapy (AFT), in 549 AML/MDS patients undergoing induction chemotherapy or allo-HSCT with fluconazole prophylaxis, who are randomly assigned to receive caspofungin empirically or pre-emptively.The overall survival (OS) at day 42 was non-inferior (pre-emptive arm 96.7%, empiric arm 93.1%) and IFD rates at day 84 were superimposable (pre-emptive arm 7.7%, empiric arm 6.6%), while the exposure to caspofungin was significantly lower in the pre-emptive arm (27% vs. 63%), without excess mortality [13].
In 2020, the European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education (EORTC/MSG) consensus group updated the IFD definitions published in 2008 [14], and we highlight the main revisions [15].

Proven Invasive Fungal Diseases
Proven IFDs apply to patients regardless of their immune status.The culture-based and histopathological detection of fungi, in a specimen from a normally sterile site, remains the gold standard for diagnosis.Wherever feasible, when fungi are seen in formalin-fixed paraffin-embedded tissue, the amplification of fungal DNA by a polymerase chain reaction (PCR) combined with DNA sequencing has become a criterion for proven diagnosis.

Probable Invasive Fungal Diseases
Probable IFDs require a host risk factor, alongside clinical, radiological and mycological findings.
The main revisions of their radiological features are as follows: • Among patients with pulmonary invasive aspergillosis (IA), nodules with halo sign remain useful among neutropenic patients, but they are nonspecific for IA in other groups.Conversely, in invasive mucormycosis (IM), consolidation is the most frequent presentation, followed by mass lesions and nodules.Multiple nodules (>10) and pleural effusions appear more frequently in IM than in IA and reverse halo sign is more specific for IM.• Among non-neutropenic patients, multiple nodules and various nonspecific findings are present (bronchopneumonia, consolidation, cavitation, pleural effusions, ground glass or tree-in-bud opacities and atelectasis).
Below are the main revisions of the host factors, including new additions: • Hematologic malignancy (acute leukaemia, lymphoma, multiple myeloma) intended as active cancer receiving treatment or remission in the recent past.

•
Receipt of a solid organ transplant (SOT).• Treatment with B-cell immune-suppressants, such as Bruton's tyrosine kinase inhibitors.

•
Inherited severe immunodeficiency, not only chronic granulomatous disease, but also STAT-3 or CARD9 deficiency, STAT-1 gain of function or severe combined immunodeficiency.
The main revisions of the mycological criteria, according to the new available literature: • Aspergillus galactomannan (GM) detection using serum, bronco-alveolar lavage (BAL) and cerebrospinal fluid (CSF) remains essential.The thresholds of the Platelia Aspergillus assay have been standardized to support the diagnosis of IA as follows: single serum ≥ 1, BAL ≥ 1, CSF ≥ 1 or single serum ≥ 0.7 plus BAL ≥ 0.8.The performance of GM remains clearly lower in non-neutropenic patients and those receiving MAP.

•
Aspergillus PCR assays obtain sufficiently robust data to be performed on plasma, serum, whole blood and BAL from adults with HM or from allo-HSCT recipients for screening and diagnosis confirmation.Any of the following is a criterion for probable IA: ≥2 consecutive positive tests on plasma, serum or whole blood; ≥2 positive tests on BAL or ≥1 positive test on plasma, serum or whole blood plus one positive test on BAL.There are few commercial tests (mostly in-house assays); despite technologic variability, PCR performance is comparable with that for detecting GM.The Fungal PCR Initiative developed recommended criteria for Aspergillus PCR test rather than a standardized method per se • Beta-D-glucan (BDG) detection is suitable for diagnosing IFDs in specific settings (HM with and without neutropenia, pre-engraftment phase after allo-HSCT, intensive care unit patients at high risk for invasive candidiasis (IC)).The recommended assay is Fungitell, with a single threshold (>80 pg/mL).Confidence for true positive results increases with repeated positive tests (≥2 consecutive positive serum samples constitute a criterion for probable IC) and for values that greatly exceed the threshold.BDG displays an extremely high negative predictive value (NPV) for candidiasis, aspergillosis, pneumocystosis or fusariosis • T2Candida, performed on blood specimens, has a very high NPV and a variable positive predictive value (PPV) (62% in patients with sepsis, shock or an intensive care stay of >3-7 days; 92% in neutropenic leukemic patients or post allo-HSCT while not receiving AFP).It represents a criterion for diagnosis of probable IC.

Advances in Microbiological Diagnostic Tools and Radiological Techniques
After the update to the EORTC/MSG criteria [15], further studies have been published on new diagnostic tools, to expand our ability to diagnose IFDs in high-risk NF patients, especially mycological tests (PCR-based molecular methods, plasma microbial cell-free DNA sequencing (mcfDNA-Seq)) and radiological ones ([¹ 8 F]FDG-PET-CT scan).
Starting from molecular tests, the AsperGenius PCR assay, which detects A. fumigatus, A. terreus, other Aspergillus species and mutations in the A. fumigatus cyp51A gene (TR34/L98H, TR46/T289A/Y121F), was evaluated in a prospective study on 323 HM patients with NF and pulmonary infiltrates, who were undergoing BAL within 48 h, of which 36% were diagnosed as having probable IA.The assay turned out positive in 40% of BAL, more frequently in samples with increasing GM positive values, and A. fumigatus DNA was detected in 30%.PCR, used to detect mutations associated with azole-resistant A. fumigatus, was conclusive in 58/89 samples and positive in 8/58 (1/8 treatment failure).Interestingly, while GM positivity predicted mortality, an isolated positive Aspergillus PCR did not [16].Molecular tools seem promising, especially in the improvement of our ability to diagnose IM, whose diagnosis is based on direct examination using low sensitivity special stains and culture [17].The MODIMUCOR study evaluated a quantitative PCR (qPCR) assay, targeting Rhizomucor, Lichtheimia and Mucor/Rhizopus, performed in serum twice weekly in 232 and 13 patients, respectively, with the suspicion of invasive mold disease (IMD) (cohort 1) and proven/probable IM (cohort 2).Overall, 40 cases of proven/probable IM were diagnosed (cohort 1 n = 27, cohort 2 n = 13; overall, 67% HM and 52% grade IV neutropenia).In cohort 1, 23/27 patients with IM had a positive qPCR (serum sampling was suboptimal for four patients with negative qPCR); furthermore, an additional 18 patients had a positive qPCR (PCR-only group).Overall, the sensitivity was 85% and the specificity was about 90%, with a NPV of almost 98%.Among proven/probable IM cases, 25/40 patients had Mucorales-positive cultures and more fungal identifications were achieved directly from tissues by molecular tests; 36/40 had a positive serum qPCR that was 100% in accordance with the species identified from tissue or BAL.The first positive qPCR was observed at a median of 4 days before the positive mycological or histological specimen.Survival at 30 days and 6 months was significantly higher among patients whose qPCR results became negative within 7 days after amphotericin-B initiation than among patients for whom qPCR remained positive, both in the proven/probable IM and the Mucorales PCR-only groups [18].Currently, there are three commercial PCR assays for Mucorales: MucorGenius (PathoNostics), MycoGenie Aspergillus-Mucorales (Ademtech) and Fungiplex (Bruker) [17].The MucorGenius, which amplifies 28S ribosomal RNA of Rhizopus, Mucor, Rhizomucor, Lichtheimia and Cunninghamella, is the one with more evaluable data and good sensitivity, but it may be less efficient at detecting Lichtheimia corymbifera [19,20].Its clinical relevance, in comparison to in-house assays and conventional mycology, was evaluated in 319 immunocompromised patients (42% HM, 11% allo-HSCT, 19% SOT) with proven/probable IM (n = 6), proven/probable IA (n = 63), Aspergillus-Mucorales co-infections (n = 4), possible IMD (n = 152) and no IMD (n = 94).In proven/probable IM, the in-house and MucorGenius assays showed a sensitivity of 100% and 90% and a specificity of 95.7% and 97.9%, respectively.Moreover, MucorGenius allowed the detection of Mucorales DNA in samples from 10 possible IMDs, all missed by culture [21].The plasma mcfDNA-Seq also displays a high specificity, NPV and PPV for IMD, with a moderate sensitivity.It was evaluated in HSCT recipients with pneumonia (n = 51 proven/probable IA, n = 24 proven/probable non-Aspergillus IMD, n = 20 possible IMD, n = 19 non-IMD controls); among proven/probable IMD, it showed a sensitivity of 51%, which increased to 61% in samples obtained within <3 days of diagnosis and 79% in non-Aspergillus IMD, and was negative in non-IMD controls [22].Another study on HSCT recipients with pulmonary IMDs (16 IA, 23 non-Aspergillus IMDs) showed that mcfDNA-Seq detected ≥1 pathogenic molds in 38%, 26%, 11% and 0% of plasma samples collected during the first, second, third, and fourth week before clinical diagnosis, respectively.Additionally, in non-Aspergillus IMDs, plasma mcfDNA concentrations may correlate with extra-pulmonary spread and mortality at the 6 week mark [23].
Moving to imaging techniques, a RCT highlighted [¹ 8 F]FDG-PET computed tomography (CT) as a valid tool to support decision-making regarding antibiotics cessation or de-escalation in leukemic patients with persistent or recurrent NF, after induction chemotherapy or allo-HSCT, almost all of whom (90%) were receiving MAP.Patients were randomly assigned to [¹ 8 F]FDG-PET-CT (n = 65) or conventional CT (n = 69); the primary endpoint was a composite of starting, stopping or changing the spectrum of antimicro-bials within 96 h.Antimicrobial rationalization occurred in 82% of patients in the [¹ 8 F] FDG-PET-CT group and 65% in the CT group (p = 0.03), mainly with the de-escalation of carbapenems and glycopeptides.There were six and four cases of proven/probable IFDs in the [¹ 8 F]FDG-PET-CT and CT groups, respectively; while zero and five cases of possible IFDs occurred in the [¹ 8 F]FDG-PET-CT and CT groups, respectively.Due to its metabolic component, [¹ 8 F]FDG-PET-CT seemed more accurate in the identification of true infection than conventional CT, which identified many pneumonias-pneumonitis in which no causative pathogen was identified, and which might never have represented infection [24].
Finally, it is of paramount importance to correctly define b-IFDs.Thus, consensus definitions of persistent, refractory, relapsed, and b-IFDs were developed, to facilitate epidemiologic analysis and comparisons between clinical studies [25]: ➢ Breakthrough IFD: an IFD occurs during exposure to an antifungal, irrespective of the treatment intention (prophylactic, empiric, pre-emptive, targeted), including fungi outside the drug's spectrum of activity.The time-point of b-IFD is the first attributable clinical sign/symptom, mycological findings or radiological feature.As a pre-existing IFD is not a b-IFD, b-IFDs can be diagnosed only if the time-point of b-IFD occurs after a minimum antifungal exposure, according to drugs' pharmacokinetic and pharmaco-dynamic properties (time to steady state: Anidulafungin, 1 day; Caspofungin, 4-7 days; Micafungin, 4-5 days; Fluconazole, 5-10 days without a loading dose; Voriconazole, 1 day with a loading dose, 5 days without a loading dose; Posaconazole, 3-7 days; Isavuconazole, 4-7 days with a loading dose, 10-14 days without a loading dose; liposomal Amphotericin-B, 4-7 days; Olorofim, 1-2 days).➢ Relapsed IFD: the IFD was caused by the same pathogen at the same site, occurring after AFT discontinuation.For probable IFD, the same clinical picture (imaging results, an increase in non-culture-based fungal biomarkers like GM in case of IA) is sufficient.
A differentiation between relapse and flare during immune reconstitution is essential.➢ Refractory IFD: an IFD with worsening or new attributable clinical signs, or radiological findings attributable to IFD while on treatment.Immune reconstitution can complicate assessment, as it may also lead to radiological and clinical progression that coincides with immune system recovery, so it needs to be ruled out.➢ Persistent IFD: IFD unchanged from baseline, may precede treatment success.Persistent IFDs may vary by patient group (e.g., persistent disease may represent a therapeutic response in persistently neutropenic hosts, while it could represent a lack of response in patients who have become more immune-competent during the course of IFD).

Invasive Fungal Diseases in Patients Undergoing Remission Induction Chemotherapy for Acute Leukaemia
For AML patients receiving induction chemotherapy, the concomitant presence of multiple predisposing factors usually concurs to IFD development [26][27][28][29][30][31].These risk factors can be divided into four main categories: leukemia-related (an advanced stage of the disease, unfavorable genetic pattern or a failure to obtain complete remission (CR)), host-related (performance status, comorbidities, older age or organ dysfunction), treatmentrelated (deep and prolonged neutropenia or mucositis) and fungal exposure-related factors (rooms without HEPA filters or previous IFDs) [1,32].It is challenging to provide precise estimates of the incidence and etiology of IFDs, as they change over time and differ greatly among countries.This emphasizes the need for local epidemiological data and differences in at-risk patient populations, AFP practices and environmental exposure [4,26,31,33].
In 2007, a RCT with posaconazole in patients with AML undergoing induction chemotherapy resulted in 2% proven/probable IFDs, compared to 8% with fluconazole or itraconazole, translating to an OS benefit and leading to FDA approval for IFD prevention [34].This and other subsequent studies [35][36][37][38][39] placed a large emphasis on the role of MAP in this setting and established posaconazole as the preferred agent for primary AFP undern international guidelines [7][8][9][10]40].Yeasts, and especially Candida spp., have historically been the most frequent cause of IFDs [41].However, recent epidemiological studies clearly demonstrated that IC rarely occurs in HM patients, probably due to effective AFP [42].In contrast, molds are the leading cause of IFDs nowadays, with a major role of Aspergillus spp.; Zygomycetes, Fusarium and other molds are reported, but less frequently involved [26,31,33].In addition, there has been a steady increase in IFDs caused by nonalbicans Candida species and non-Aspergillus molds, which could be directly related to the extensive use of AFP with azoles [43].These epidemiological shifts are reported in Table 1, which summarizes the recent real-life cohorts describing b-IFD's incidence in patients undergoing induction, consolidation and/or salvage chemotherapy.
Finally, the treatment landscape for AML has recently shifted to molecular targeted therapies, such as BLC2-inhibitors (venetoclax), FLT3 inhibitors (midostaurin, gilteritinib, quizartinib and sorafenib), CD33-directed antibodies (gemtuzumab ozogamicin) and IDH inhibitors (ivosidenib and enasidenib), rather than or in addition to cytotoxic chemotherapeutics or hypomethylating agents (HMAs).These targeted therapies led to an improvement in OS, with a higher achievement of CR in subgroups of patients with historically poor outcomes [66][67][68][69][70][71][72][73][74].The incidence of IFDs with these new therapies is largely variable [55,75,76], and few data are available on the need for and efficacy of AFP.In addition, DDIs between azoles and these new molecules, mainly driven by the strong inhibition of cytochrome P450 3A4 [77], make the choice of AFP even more difficult and heterogeneous between centers.In the case of co-administration of venetoclax and triazoles, it is recommended to reduce the venetoclax dose by 75% and 50% when co-administered with posaconazole and isavuconazole, respectively; conversely, HMAs (azacitidine or decitabine) and midostaurin require no dose alterations while monitoring for eventual toxicities [77,78].Table 1 includes studies on patients receiving both midostaurin [55] or HMAs; in the latter setting, the incidence of IFDs ranged between 3% and 12.6%, with a significant variability of AFP and a high percentage of patients receiving echinocandins or fluconazole [46,57,75,79].

Invasive Fungal Diseases in Patients Undergoing Allogeneic Hematopoietic Stem Cell Transpantation
In 2007, a RCT of allo-HSCT recipients with a-GVHD proved the superiority of posaconazole compared to fluconazole, in preventing proven/probable IA (2.3% vs. 7.0%) and b-IAs (1.0% vs. 5.9%) and reducing IFD-related mortality (1% vs. 4%) [80].While there is international consensus on the indication for posaconazole AFP in the a-GVHD phase, in the pre-engraftment phase there is more heterogeneity on the use of AFP.In 2015, Bow and colleagues performed a mixed treatment comparison meta-analysis, to compare RCTs examining the use of oral drugs for AFP in allo-HSCT recipients.Five RCTs were included with 2147 patients, and it was found that, compared to fluconazole, itraconazole (OR 0.52 (0.35-0.76)), posaconazole (OR 0.56 (0.32-0.99)), and voriconazole (OR 0.46 (0.28-0.73)) reduced the incidence of proven/probable IFDs.Moreover, posaconazole (OR 0.31 (0.17-0.58)) and voriconazole (OR 0.33 (0.17-0.58)) reduced proven/probable IA compared to itraconazole (OR 0.68 (0.42-1.12)), while all-cause mortality was similar across all mold-active agents [81].However, the included studies were heterogeneous in terms of study design, patient population and risk of IFDs, thus limiting the interpretations of the treatment effects, due to those biases.Winston and colleagues observed the highest IFD rate, but a greater proportion of fluconazole compared to itraconazole recipients received an unrelated donor transplant and experienced acute and chronic GVHD [82].The percentage of patients receiving myeloablative conditioning (MAC) was different across the studies (100% [83,84], 78% [82] and 58% [85]), thus influencing the pre-engraftment myelosuppression and cytotoxic therapy-induced intestinal damage.Other significant variances among the studies were the AFP initiation (beginning of conditioning [83], the day of transplant [84,85], the day after transplant [82], the day of GVHD onset [80]) and the incidence of an a-GVHD grade ≥ two (100% [80], 64% [83], 46% [85], 41% [84] and 37% [82]).Finally, there were also significant variations in toxicity or intolerance-driven drug withdrawal rates, and the percentage of transplant from mismatched/unrelated donors was limited, not exceeding 41-48% [83][84][85].Nowadays, the main criteria that guide clinicians to set up MAP in the pre-engraftment and early post-engraftment phases are the following: active disease at transplant, cord blood transplant, previous allo-HSCT, an a-GVHD grade ≥ 2, a mismatched related donor (MMRD), a matched or mismatched unrelated donor (MUD or MMUD), grade IV neutropenia before allo-HSCT or Cytomegalovirus reactivation within 100 days [12].
An Austrian study on recipients at low-moderate risk of IFDs (59% of whom had reduced intensity conditioning, 80% of whom had a matched donor), with only 43% receiving AFP during the pre-engraftment phase, reported that 11.6% of patients developed IFDs displaying a significantly higher mortality (48% vs. 25%) [86].The Swiss cohort (5% of whom had had prior transplant, 23% of whom had a mismatched donor) reported a cumulative incidence of proven/probable IMD after 1 year of 7%, of which 69% were b-IMDs; overall, 29% of IMDs occurred in the pre-engraftment phase, with most patients receiving fluconazole, and the all-cause 1-year mortality rate was 48% and 31% in patients with and without IMD [87].Moving to studies with a universal use of MAP with azoles, the cumulative incidence of IFDs did not exceed 3.2% [88,89].Two papers described a cohort of 95 recipients receiving AFP with isavuconazole, reporting a favorable safety profile (premature discontinuation: 15% isavuconazole vs. 44% voriconazole) [88], but a trend towards a higher incidence of b-IFDs (mostly candidemia) [88,90].Two studies evaluated a tailored AFP strategy according to IFD risk: high-risk patients received posaconazole in the study by Malagola and micafungin in the study by Busca, while fluconazole was administered to low-risk patients.The cumulative incidence of proven/probable IFDs ranged from 9.6% to 18% [91,92].
Among studies not entirely focused on allo-HSCT recipients, two analyzed the effectiveness of posaconazole (oral suspension) and isavuconazole AFP.In the first study, allo-HSCT recipients with a-GVHD represented 35% of the population: 3.2% of patients (9/270) developed IFDs (1/9 allo-HSCT recipients), a sub-optimal exposure to posaconazole was frequent and more than one third of deaths were caused by IFDs [94].Considering the study on isavuconazole, allo-HSCT recipients were 43% of the sample; overall, 7.6% of patients (11/145) developed proven/probable b-IFDs (1/11 allo-HSCT recipients) with more than half dead at 6 weeks [52].Another study focused on candidemia in HM patients: 133 episodes were analyzed, 42% of which were breakthrough infections while on AFP (more frequently C. krusei or fluconazole-resistant isolates), and the 30-day mortality was considerable (36%, 25/48 deaths were classified as Candida-related) [2].A Spanish study enrolled HM patients diagnosed with b-IFDs, reporting 121 cases (42% of which were allo-HSCT recipients); among microbiologically documented b-IFDs, any fungus susceptible to the prior antifungal administered at good therapeutic levels was isolated (except some catheter-related fungemia).Again, the mortality after 100 days was high (47%, with 61% of those deaths caused by b-IFD), especially in mucormycosis and other rare yeasts.Interestingly, the molecular testing of biopsy samples identified a high number of non-Aspergillus molds [95].Lastly, two studies analyzed b-IFDs in immunocompromised patients receiving MAP with azoles, and the incidence of proven/probable IFDs was 4.9% [96] and 6.3% [97].
Another important issue is the change in transplant platforms, including the advent of post-transplant cyclophosphamide (PT-Cy) to allow transplantation from alternative donors, especially haploidentical donors, while modulating the consistent risk of GHVD.The rate and characteristics of infectious complications with PT-Cy have been reported in different studies, but few were dedicated specifically to IFDs [98].A retrospective study (2011-2021) analyzed 205 haploidentical HSCT recipients receiving PT-Cy, mainly after reduced intensity conditioning (75%) and without the routine administration of AFP.Overall, 29 patients (14%) developed proven/probable IFDs (14/29 within 30 days post-HSCT (13/14 invasive yeast infections), 15/29 after 30 days post-HSCT (7/15 IMDs, of which 6/7 were IA, 6/15 were invasive yeast infections, 1 was pneumocystosis and 1 was coccidioidomycosis)).The 30-day cumulative incidence of proven/probable invasive yeast infection was 6.3%, with significantly increased non-relapse mortality occurring in these patients (54% vs. 11%) [99].Another retrospective study (2013-2015) investigated 381 haploidentical HSCT recipients receiving PT-CY (38% of whom had had a prior transplant): IFDs were reported in 78 patients (mostly caused by Aspergillus (43%) and Candida (33%)), the rate of IFDs was 6% and IFDs occurred at a median of 20 days post-HSCT.Notably, MAP with posaconazole was provided only during GVHD, which might have contributed to cases of early IA; finally, among 37 patients who died of infection-related mortality, only 4/37 deaths were caused by IFDs [100].In a Spanish multicenter study (2013-2018) on 236 PT-CY haploidentical HSCT recipients, the incidence of IFDs within 30 days post-HSCT was 4%, reaching 11% within 3 years; the most frequent IFD was proven/probable IA (n = 16), followed by candidiasis (n = 16) and pneumocystosis (n = 2).In this cohort, AFP regimens differed between centers, with some using fluconazole and others MAP [101].Finally, the CAESAR study, which developed a clinical risk score for predicting IFDs within 6 months from allo-HSCT, reported haploidentical HSCT as a risk factor [102].In conclusion, haploidentical HSCT with PT-Cy seems to be associated with a high risk of IFDs, particularly early IFDs, and dedicated prophylactic strategies might be warranted.

Single-Center Experience with Posaconazole Prophylaxis in Patients with Acute Myeloid Leukaemia
We retrospectively reviewed the medical records of 662 adults diagnosed with AML/MDS at San Raffaele Hospital between 2010 and June 2023, undergoing induction, consolidation and/or salvage chemotherapy.We excluded patients who did not receive intention-to-cure treatment, acute promyelocytic leukemia, subjects receiving secondary AFP for pre-existing IFD and those who received AFP other than posaconazole.Thus, we analyzed the data of 184 patients (totaling 339 chemotherapy treatments).

Methods, Objectives and Statistical Analysis
Patients gave written informed consent for chemotherapy and the use of their medical records for research and were treated according to the institutional standard of care.Baseline patients' characteristics, hematological disease features, details about chemotherapy cycles and hematological response assessments were analyzed.Data were anonymized and managed according to the Good Clinical Practice guidelines published by the World Medical Association's Declaration of Helsinki.The analyses were performed considering the hematological treatments' episodes: remission induction chemotherapy, consolidation chemotherapy and salvage chemotherapy.Hematological treatment platforms were uniform throughout the study period; conversely, anti-leukemia targeted therapies were progressively introduced over time.
The primary objective was to estimate the incidence rates of proven/probable b-IFDs; the secondary objectives included the description of characteristics and outcomes of patients who experienced proven/probable b-IFDs.
Baseline (day 0) was defined as the date of chemotherapy cycle initiation; the followup of each hematological treatment episode was censored at patient's death, subsequent eventual chemotherapy cycle or HSCT, or on the follow-up visit on day 180.Neutropenia was defined as ANC < 500 cells/mm 3 [11,40].The hematological response was assessed for induction chemotherapy cycles at day 14 to evaluate the eventual indication to early re-induction; otherwise, it was assessed at day 30.High-risk neutropenia before chemotherapy was defined as ANC < 500 cells/mm 3 at baseline, for at least 7 days.Engraftment was defined as the first of three consecutive days with ANC > 500 cells/mm 3 [11,40].Proven/probable b-IFDs were defined according to EORTC/MSG definitions, and b-IFD diagnosis was dated at the time of symptoms' onset, radiological signs or mycological evidence of IFDs; whichever occurred first [25].Posaconazole was started at baseline.PPCs were assessed weekly and they were considered inadequate if <0.7 mg/L, adequate for prophylaxis if they measured 0.7-1 mg/L and adequate for both prophylaxis and treatment if >1 mg/L [7,8].The clinical outcome of patients with b-IFD was assessed at 6 and 12 weeks after diagnosis.
Incidence rates (IRs) of proven/probable b-IFDs in each chemotherapy category were calculated with univariable Poisson regressions; rates per 100 person years follow-up (PYFU) were reported; 95% confidence intervals (95% CI) were estimated with a normal approximation to binomial.Two-sided p-values < 0.05 were considered significant.Incidence rate ratios (IRRs) were calculated to compare IRs.Statistical analyses were performed using R (R Foundation for Statistical Computing, Vienna, Austria, version 4.2.3).

Results
Overall, 339 chemotherapy treatments were analyzed from 184 patients (a patient may have received more than one chemotherapy treatment): 150 patients received 153 remission induction chemotherapy treatments, 90 patients underwent 126 consolidation chemotherapy treatments and 54 patients received 60 salvage chemotherapies.The total number of treatment cycles analyzed was 1052 (1 to 12 maximum cycles per patient).The characteristics of these chemotherapy treatments are reported in Table 3.
Three probable IA occurred in patients with adequate PPCs at the time of diagnosis, while two patients were under-exposed to posaconazole; most patients (5/6) received voriconazole as AFT.At 12 weeks post-b-IFD diagnosis, treatment success was achieved in all cases.Moreover, no deaths were attributable to b-IFDs and all patients regularly proceeded in their course of leukemia treatments.The neutrophils engraftment was achieved in 86.3% of treatments, although with different percentages among groups (87.3% in induction, 94.4% in consolidation and 66.7% in salvage chemotherapy; p < 0.001).Similarly, a complete response to treatment occurred in 78.7% of cases (248/339), with significantly different proportions between treatments groups (78.7% in induction, 89.1% in consolidation and 56.4% in salvage chemotherapy; p < 0.001).Overall, 19% of patients (35/184) died during follow-up, more frequently in the salvage chemotherapy group (29.6%) compared to the induction (7.3%) and consolidation chemotherapy (8.9%) groups.The most common cause of death in the induction and salvage chemotherapy groups was relapsed refractory disease, while in the consolidation chemotherapy group half of the deaths were due to bacterial infections.
Overall, PPCs were available for 95.1% of patients (175/184) and 95.3% of chemotherapy treatments (323/339), with a total of 936 measurements (0 to 19 maximum measurements per patient).Table 3 reports details about PPC measurements and levels, according to different chemotherapy treatment groups; interestingly, at least two consecutive inadequate PPCs were recorded in only 6.8% of chemotherapy treatments (22/323).
In our cohort, the proportion of patients who developed b-IFDs during induction chemotherapy was 1.3% (2/150), one of the lowest among similar studies, even considering those that tested isavuconazole [50][51][52], thus, supporting the efficacy of posaconazole AFP in this context.Moving to consolidation chemotherapy, the proportion of our patients experiencing b-IFDs was 1.1% (1/90), lower than in studies conducted before extensive AFP use [26,[60][61][62][63] and lower than in the recent study by Del Principe [64], where AFP was not provided for 44% of patients and the b-IFD incidence among those receiving MAP was 1.5%.Notably, two out of six probable IA in our cohort occurred in patients treated with VEN-HMAs, with 10.5% of patients receiving VEN-HMAs and experiencing b-IFD (2/19), which is similar to the proportion reported in one study [57] but higher than in other cohorts [46,75,76,79].In this setting, some studies [46,57] suggested AFP during initial cycles of induction and in relapsed/refractory AML, while other studies [75,76,79] stated there was no need for AFP.The addition of VEN to HMAs improves OS and rates of CR, compared to HMAs alone, but is associated with prolonged neutropenia [66].The role of AFP in patients receiving VEN-HMAs deserves careful consideration, at least until further studies are available.Analyzing the outcomes of our patients with b-IFDs, at week 12 all subjects achieved treatment success and continued their hematological treatments.These findings are appreciable, thus suggesting that a broad-spectrum AFP may reduce the burden of IFDs, resulting in less severe forms.Previous studies reported a similar positive effect of AFP on the effectiveness of AFT [64,103].Even if our first-line AFT did not comply with most guidelines, by favoring an antifungal of a different class [104][105][106][107][108], the treatment success was favorable, underlying how the optimal management of b-IFDs is largely unknown and the clinical studies investigating different treatment modalities are scarce [106,[109][110][111].To the best of our knowledge, our study is the one with the largest PPC measurements.Overall, in 20.1% of chemotherapy treatments, the PPCs totaled <0.7 mg/L at one time-point (only 6.5% at two consecutive time-points), and this percentage is lower than that reported by several studies [44,49,112], possibly due to a more extensive use of tablet formulations, but still indicates considerable interpatient variability.Although there are concerns about the suboptimal efficacy of azoles in the context of neutropenia [113], the importance of achieving PPCs > 0.7 mg/L for prophylaxis and >1 mg/L for treatment [7,8,114], and the absence of a definite correlation between PPCs and toxicity, are well known [114,115].A recent meta-analysis indicated a C min of 0.5 mg/L as the cut-off that separates successful from failed prophylaxis [116].These data confirm the importance of performing TDM to optimize MAP and treatment.
Our retrospective study has several limitations.Firstly, the low number of events did not allow us to investigate predictive factors for b-IFDs, or to draw conclusions regarding b-IFD management and clinical outcomes, especially IFD-attributable mortality, which was remarkably low in our cohort.Since it was not designed as a case-control study, we did not compare different AFP strategies.We also did not collect medical conditions potentially affecting posaconazole pharmacokinetics.Our cohort included a low percentage of patients treated with HMAs (8.3%), VEN-HMAs (5.6%) and target therapies (8.5%), so it is unfeasible to draw meaningful conclusions about the incidence of b-IFDs in these novel settings of AML treatment, considering that larger studies have been published (Table 1).
Finally, an effective containment of IFD occurrence in patients undergoing induction chemotherapy, as emerged in our cohort, is fundamental, because several patients will be candidates for allo-HSCT, and a history of previous IFDs is a risk factor for relapsing IFDs after transplantation.

Final Remarks and Future Perspectives on Novel Antifungal Drugs
IFDs represent one of the causes of persistent NF, alongside multi-drug resistant bacterial infections, viral infections and other non-infectious reasons.At least 15% of patients with high-risk neutropenia develop recurrent NF, mainly caused by microbiologically documented infections (31%) and clinically documented infections (30%), with IFDs representing almost half of microbiologically documented infections; notably, the risk of death is higher among patients with recurrent NF (5.4% vs. 1.4%) [117].

•
Early b-IFDs on MAP with azoles occur in AML patients during induction chemotherapy due to a high fungal inoculum exposure, previous colonization or sub-optimal pharmacokinetics of triazoles; late b-IFDs, caused by Aspergillus and non-Aspergillus IMDs, are typical of relapsed refractory AML or prolonged corticosteroid treatment for GVHD.In the absence of RCTs to guide therapeutic decisions in the different scenarios of b-IMDs, a rational AFT is liposomal amphotericin-B eventually associated with isavuconazole or posaconazole, while trying to obtain a microbiological diagnosis [107]; • The incidence of IC in HM patients has decreased with AFP, leading to an increased proportion of non-albicans Candida with decreased azole susceptibility (C.glabrata, C. krusei); among patients receiving AFP with azoles, the prevalence of candidemia and hepato-splenic candidiasis is estimated around 20% and 3%, respectively, and the emergence of echinocandin resistance (C.glabrata) and novel pathogenic species (C.auris) is an issue, although the prevalence is low [2,118]; • The epidemiology of b-IMDs in patients receiving AFP with azoles shows a shift towards non-fumigatus Aspergillus (i.e., A. ustus complex, usually associated with extrapulmonary infection and high mortality >50%) [119], non-Aspergillus molds (Fusarium, Mucorales, Scedosporium apiospermum) and rare resistant molds [120]; It is of paramount importance to pursue a mycological diagnosis of b-IFDs in highrisk NF patients; Aspergillus real-time PCR assays on blood and BAL have been included in EORTC/MSG 2020 criteria [15] and novel diagnostic tools are promising (i.e., Septifast [121], plasma mcfDNA sequencing [22,23] and real-time PCR-based methods for Mucorales [18][19][20] or IA sustained by mixtures of triazole-resistant and susceptible strains of A. fumigatus [16,122]); • Exposure to mold-active agents could select already resistant Aspergillus strains (i.e., cryptic species) and non-Aspergillus molds, rather than promote the induction of resistance "in vivo"; not only the molecular identification of non-Aspergillus molds, but also that of cryptic Aspergillus species, whose frequency stands at 11-15%, could be clinically relevant, since antifungal drug resistance is common (40%) [123,124]; • Antifungal susceptibility testing is recommended for Candida spp.(azoles, echinocandins) and A. fumigatus (azoles) because the mechanisms of resistance are well-described (acquired mutations: for azoles/Candida mutations in the ERG11 target gene, the overexpression of target genes and/or drug transporters; for echinocandins/Candida FKS hot-spot mutations; for azoles/A.fumigatus, cyp51A gene mutation) and the performance of clinical break-points is acceptable for their identification with a high probability of treatment failure (clinical break-points to predict clinical success (CLSI for Candida, EUCAST for Aspergillus): echinocandins/C.glabrata MIC ≤0.25-0.5 for caspofungin, ≤0.12 for anidulafungin; azoles/A.fumigatus MIC ≤2 mg/L for voriconazole, ≤0.25 mg/L for posaconazole, ≤2 mg/L for isavuconazole); conversely, for non-Aspergillus molds, antifungal susceptibility testing should be interpreted with caution, because MIC are unreliable predictors of outcome [125]; Guidance on imaging for pulmonary IA and IM exists [15,126]; high resolution lung CT is fundamental in the diagnostic work-up of IFDs; although not pathognomonic, in patients with HM and neutropenia, the halo sign is present at the symptoms' onset in >70% of IA, but its prevalence decreases rapidly over time (<40% after 7 days, <20% after 14 days).The air-crescent sign is not specific and occurs in later stages of IFD; the reversed halo sign and hypodense sign are typical of pulmonary IM but occur less frequently.

Future Considerations
• The optimal management of AFP, including the necessity, type and dosage of antifun- gals, in the novel treatment landscape of AML based on molecular targeted therapies; • The development and clinical application of optimal risk stratification strategies, to prevent IFDs in allo-HSCT recipients, through an individualized choice of AFP (MAP vs. non MAP) according to patients' features (their hematopoietic comorbidity index, disease status at HSCT, whether they have undergone previous allo-HSCT, whether they have had previous IFDs), the type of transplant (conditioning, donor (haploidentical, cord blood), GVHD prophylaxis (PT-Cy)), and an individualized AFP length (engraftment, poor graft function or a-GVHD).Early diagnosis strategies based on local prevalence, risk factors and phase after HSCT, are warranted;

•
Prospective studies should include careful diagnosis of proven/probable IFDs and b-IFDs and correct detailing of the outcomes of IFDs, to discern the exact burden of IFDs, b-IFDs and the attributable mortality from the role of other factors (i.e., non-fungal infections, neutrophils recovery, underlying disease response, GVHD and persistent immune-suppression).
Finally, because of some limitations of current antifungals (e.g., route of administration, toxicity or DDIs) and the emergence of resistant fungi, there is still an urgent need to expand our armamentarium and several compounds are reaching the stage of phase II or III clinical assessment.These include new drugs within the existing antifungal classes or displaying a similar mechanism of activity with improved pharmacologic properties (rezafungin and ibrexafungerp) or first-in-class drugs (olorofim and fosmanogepix).Rezafungin, a long acting intravenous echinocandin with a weekly schedule of administration, has recently been FDA-approved for the treatment of IC, after the results of the STRIVE and ReSTORE trials (with the comparator caspofungin) [127,128].Its spectrum includes Candida spp.(including azole-resistant strains), Aspergillus spp.(including azole-resistant isolates) and P. jirovecii, but it does not cover non-Aspergillus molds.Pooling the data from the two trials (139 patients for rezafungin, 155 patients for caspofungin), the 30-day all-cause mortality rates were comparable between groups (19% rezafungin, 19% caspofungin), and mycological eradication occurred by day 5 in 73% and 65% of patients treated with rezafungin and caspofungin, respectively [129].Rezafungin is currently also being studied for prophylaxis; the phase III ReSPECT trial is ongoing to compare rezafungin vs. fluconazole as AFP in allo-HSCT recipients (drug exposure: from transplant until week 14), while considering some conditions at high risk of IFDs opt be exclusion criteria (active disease at transplant, cord blood transplant or previous allo-HSCT).Olorofim is the first compound from the novel orotomide class, which selectively inhibits fungal dihydroorotate dehydrogenase, a key enzyme in pyrimidine biosynthesis that results in the loss of essential substrates for cell wall integrity and DNA replication.It can be administered both intravenously and orally, the therapeutic concentrations are achieved in all targeted organs and tissues (including lung, brain, liver and kidney) and it is metabolized by several cytochrome P450 isoenzymes, but DDIs may be less clinically relevant than those observed with mold-active triazoles.Its spectrum includes wild-type and azole-resistant Aspergillus spp.and some uncommon fungi (Scedosporium spp., L. prolificans, Fusarium spp.), while it is inactive against Candida and Mucorales [130].Both the phase II open label single-arm trial (NCT03583164) evaluating olorofim for IMDs with limited therapeutic options, and the phase III OASIS trial (NCT05101187) comparing olorofim vs. liposomal amphotericin B for proven/probable IA refractory to azoles, or breakthrough on MAP with azoles, or in the case of DDIs that make azoles unsuitable, are recruiting.Fosmanogepix is a first-in-class antifungal agent inhibiting Gwt1, an enzyme in the glycosylphosphatidylinositol biosynthesis pathway.It can be administered both intravenously and by the oral route, and it distributes in many difficult-to-treat body sites (including the brain, eyes and abdominal abscesses).It is highly active against Candida spp.(including C. auris, but not C. krusei and C. kefyr) and the majority of echinocandin and azole-resistant Candida.Regarding molds, fosmanogepix displays potent activity against Aspergillus, including azole-resistant A. fumigatus and cryptic species, and it is also active against most Fusarium, in particular F. solani, and most Scedosporium and L. prolificans isolates.In contrast, in vitro antifungal activity against most Mucorales is limited [130].
In conclusion, these novel drugs are promising for the treatment of emerging resistant Candida and Aspergillus spp., as well as some notoriously refractory IMDs (disseminated fusariosis, scedosporiosis/lomentosporiosis); however, they do not seem to be suitable for the treatment of IM, which remains a major issue.

Not available;
Not available; Incidence rates of probable/proven IFDs did not differ significantly between patients with newly diagnosed AML and those with r/r AML, nor between AFP and no-AFP groups.
Table 2. Studies reporting the incidence of invasive fungal diseases and/or the antifungal prophylaxis practice in allogeneic hematopoietic stem cell transplant recipients (eight studies entirely on transplant recipients, six studies on immune-compromised hosts including allogeneic hematopoietic stem cell transplant recipients).IFD was associated with a higher mortality rate compared to patients without IFD (48% vs. 25%, p = 0.02), and the higher mortality rate was even more pronounced for IA (62%, p = 0.003); Patients with IFD were admitted to the ICU more often (18/28, 64%) than patients without IFD (26/214, 12%); Intensified GVHD therapy (≥1 mg/kg corticosteroids and basiliximab or etanercept) and TAM were associated with an increased risk for IFD and IA, while AFP was associated with a decreased risk;

Table 1 .
Studies reporting the incidence of invasive fungal diseases in patients with AML treated with PSZ-AFP undergoing induction, consolidation or salvage chemotherapy *.

Design Study Period Study Population Antifungal Prophylaxis Incidence of Invasive Fungal Diseases Therapeutic Drug Monitoring Type of Breakthrough Invasive Fungal Diseases Clinical Outcome Risk Factors for Invasive Fungal Diseases Other Analysis/Considerations
Proven IFDs: five IC (one C. albicans, two C. glabrata, one C. krusei, one Candida spp.), three IA (one A. fumigatus, two Aspergillus spp.);Probable IFDs: two IC (two C. albicans), three P. jirovecii, three IA (one A. fumigatus, one A. niger, one Aspergillus spp.), twelve IA with positive galactomannan antigen.

Table 2 .
Cont.Three proven IFD, five probable IFD (type of b-IFDs not precisely reported (in the ISA group there was one Zygomycetes and one Aspergillus spp., in the PSZ-VCZ group there was one Candida spp.and three Aspergillus spp.)).Patients in the PSZ-VCZ group experienced more hepatotoxicity, compared to patients receiving ISA (16.7% vs. 4.8%).

Table 3 .
Characteristics of chemotherapy treatments, posaconazole plasma concentration data, follow-up time and causes of death.