Introduction

Invasive mold infections (IMIs) are a major cause of mortality among immunosuppressed hosts [1,2,3,4,5]. Despite progress in diagnostics, treatment, and prophylactic strategies, incidence has been increasing in the last decades, mainly because of the emergence of breakthrough IMI (bIMI) [6]. Although IMI definitions have been clearly established, accurately defining bIMI remains challenging [7,8,9]. It has been proposed that bIMI should be considered when an IMI is diagnosed after at least 7 days of administration of a mold-active agent, considering that most antifungal drugs will have reached a steady state in that time [8]. Although logical and practical, this definition fails to include the specific characteristics of different antifungal agents, which could have a significant impact on the development of bIMI. Cornely et al. have proposed that bIMI should be defined based on the exposure to a certain antifungal drug, considering the individual pharmacokinetic and other properties of different agents or classes [9]. For instance, time to steady state heavily differs between echinocandins and azoles and also between different azoles, depending on the administration or not of a loading dose, not always administered by clinicians in the setting of antifungal prophylaxis. Considering the frequency of posaconazole as primary or secondary prophylaxis in high-risk hematological patients, as evidenced by existing guidelines and recent surveys, we will predominately discuss posaconazole-associated bIMI in this article, which we believe represent the vast majority of bIMI [10, 11].

Epidemiology

The incidence of bIMI has been steadily increasing, with recent data suggesting that bIMI represent the majority of IMI amongst allogeneic hematopoeitic cell transplant (HCT) recipients [3, 12]. This can, in part, be attributed to the increasing pool of high-risk hematology patients in the last decades and the routine administration of posaconazole as universal antifungal prophylaxis in patients with acute myelogenous leukemia and allogeneic HCT recipients with acute graft-versus-host disease (GvHD) [10, 13, 14]. Similar to what was observed after the introduction of fluconazole as primary prophylaxis in high-risk patients, with a significant drop in the incidence of invasive candidiasis but the emergence of breakthrough infections due to non-albicans Candida spp. and Aspergillus spp., the routine use of posaconazole antifungal prophylaxis has led to higher numbers of bIMI. The epidemiology of bIMI is different from primary IMI, often due to pathogens with variable susceptibility profiles and for which treatment options are limited and which were rarely seen in the recent past. For instance, increasing reports have shown the large variety of pathogens associated with bIMI, including non-fumigatus Aspergillus spp., such as A. ustus/callidoustus, A. nidulans, or A. udagawe, Mucorales (Rhizomucor pucillus, Rhizopus, Absidia, Cunninghannella), and other non-Mucorales rare molds: Fusarium spp., Scedosporium spp., Alternaria spp., Schizophyllum commune, Scopulariopsis spp., Hormographiella aspergillata [3, 12]. Recent reports suggest that bIMIs may present as either affecting one system, such as sinusitis or lung disease, or as disseminated infection, affecting more than one sites, such as in cases of infections due to the so called “cryptic”-species of Aspergillus and other non-Aspergillus pathogens [15,16,17]. Furthermore, bIMIs have been associated with higher mortality rates, when compared to primary IMIs in hematologic patients [12, 17].

Pathophysiology

Mechanisms leading to bIMI may be related to pharmacokinetic drug characteristics, pathogen, and/or host factors [8, 9].

Posaconazole Pharmacokinetic Considerations

Suboptimal concentrations of antifungal drugs may play a role in selecting for bIMI. This may be related to suboptimal dosing, suboptimal absorption in the setting of mucositis or gastro-intestinal GvHD, enhanced metabolism due to drug-drug interactions, or still non-adherence to treatment. Therapeutic drug monitoring (TDM) is recommended for patients treated with voriconazole and posaconazole with target trough blood concentrations well defined for both agents [18, 19]. For instance, target trough concentrations for posaconazole have been proposed at > 0.7 and > 1 mg/L for prophylaxis and treatment, respectively [19, 20]. Although those proposed cut-offs are widely accepted and part of consensus guidelines, they are predominately based on post hoc analyses on an exposure–response study on the relationship between posaconazole levels and effective prophylaxis [20]. However, recent data appear to challenge the association between “appropriate” blood trough posaconazole concentrations and effective prophylaxis against IMI. In a single-center study from the US including 343 high-risk patients (adult patients with hematological malignancy receiving posaconazole as primary IMI prophylaxis), eight bIMI were observed [21]. Posaconazole through concentrations were measured in six of them and they were all within the desired target. In another recent multi-center cohort study from the Swiss Transplant Cohort Study (SCTS) including 288 allogeneic HCT recipients who received posaconazole as prophylaxis or treatment, 1944 posaconazole trough levels were measured, 1317 of which in patients who received posaconazole as prophylaxis [22]. Nine (3.1%) breakthrough fungal infections were reported: three with invasive candidiasis and six bIMI. The median time to diagnosis of breakthrough fungal infections was 14 days post posaconazole-initiation. Notably, there was no difference in posaconazole levels between patients with and without breakthrough fungal infections, and in 8/9 patients posaconazole levels were between 0.97 and 1.56 mg/L.

In addition, measuring drug levels in the blood may not necessarily reflect what is actually happening in the target tissue, such as the sinuses or the lungs. First, as posaconazole is a high-protein binding antifungal agent (98–99%), the free plasma posaconazole concentration (or active fraction of the drug) may be subtherapeutic [23]. Second, lung tissue penetration appears to be quite high for posaconazole, reported to concentrate to high levels within cells (up to 40- to 50-fold) and attain therapeutic concentrations in pulmonary epithelial lining fluid and alveolar cells with more than five times the concentration in plasma [24, 25]. Furthermore, posaconazole levels have been found to be high in human peripheral blood mononuclear cells (PBMCs) and polymorphonuclear neutrophils (PMNs) [26]. The above point out the complexity of assessing the efficacy of posaconazole prophylaxis simply depending on blood trough concentrations, an easy and fairly accessible diagnostic tool, which does not necessarily provide clinicians with all pertinent information associated with what actually happens in the tissue on a cellular level and may lead to simplified and -at times-false perceptions.

Fungal Pathogen Considerations

Breakthrough IMI may be the result of a combination of fungal-associated variables. For instance, the exposure to a high fungal inoculum, which could potentially overcome the protection provided by an effective antifungal prophylaxis and which is not feasible to measure or quantify. In addition, broad-spectrum antifungal prophylaxis may lead to selection for drug-resistant pathogens, such as azole-resistant Aspergillus spp. (e.g. A. ustus/callidoustus) or other non-Aspergillus, non-Mucorales molds (e.g. Scedosporium spp., Lomentospora spp., Hormographiella aspergillata, etc.) [3, 12]. Last, the emergence of azole-resistant A. fumigatus reported initially in the Netherlands and Belgium, but currently worldwide, may also lead to even higher rates of bIMI in the future [27].

Host Factor Considerations

Host factors play an essential role in the pathophysiology of bIMI. As a result of the progress achieved in the field of Hematology and Transplantation, there is a constantly increasing highly-immunocompromised patient population, receiving prolonged courses of immunosuppression, leading to a growing “at risk” population, eventually developing a bIMI. Furthermore, those patients are back in the community, not in isolated positive-pressure hospital rooms anymore, regaining their everyday life activities, and eventually exposed to all types of pathogens, including fungal spore inhalation. In that sense, new forms of treatments for patients with acute myelogenous leukemia, including venetoclax and hypomethylating agents, have as a result large numbers of patients treated as outpatients, with variable neutrophil counts, often neutropenic for prolonged periods of time, without protection against environmental pathogens, such as molds, and without antifungal prophylaxis strategies well defined currently.

Genetics

Finally, genetic factors may play a role in immune responses to fungal infections. Single nucleotide polymprphisms (SNPs) in toll-like receptors (TLR) 2 or 4 and dectin-1 have been identified as important predisposing factors for infections due to Aspergillus spp. in allogeneic HCT recipients [28, 29]. However, the prevalence of specific dectin-1 and TLR2 SNPs in high-risk hematology patients was reported between 3 and 10% [30]. Similarly, the frequency of specific TLR4 haplotypes in donors of allogeneic HCT has been reported in the range of 6% [28]. Initial studies have shown strong associations between some of those polymorphisms and invasive aspergillosis [25, 26, 28, 30]. However, the relatively low prevalence of most of those polymorphisms in the general population may not necessarily allow for meaningful observations and conclusions [29]. In contrast, SNPs in certain pattern recognition receptors (PRR), particularly the pentraxin 3 (PTX3) have been found to have a much higher prevalence in the general population and associated with higher rates of invasive aspergillosis in different patient populations, including allogeneic HCT and solid organ transplant recipients and patients with acute myeloid leukemia [31,32,33,34]. Targeting patients harboring those polymorphisms in key-genes, such as that of PTX3, may allow us to stratify antifungal prophylactic strategies based on underlying genetic risk factors and hence decrease the exposure of large patient populations to high-risk antifungal prophylactic agent administration and selection for bIMI. In an ongoing prospective, randomized, stratified, double-blind clinical trial patients with acute myelogenous leukemia receiving intensive chemotherapy are stratified based on the underlying PTX3 SNPs, as high- and low-risk for IMI, and subsequently randomized to either posaconazole or fluconazole (1:1 and 1:3 for the high- and low-risk patients, respectively) [35]. This study enrolls patients in 9 different sites in three European countries and is expected to complete enrollment in 2025.

Diagnostic and Treatment Considerations

Management of bIMI is challenging, mainly because of the lack of robust evidence to guide clinical decisions and because of the complexity and variety of real-life scenarios. Diagnostics should be performed as routinely recommended for IMI with early imaging (chest computed tomography, CT), serum galactomanan enzyme immunoassay (GM EIA), and a diagnostic procedure, such as bronchoscopy, when feasible. There are some important details in the diagnosis of bIMI, which merit special mention. First, a positive GM EIA may not be adequate enough to guide therapeutic decisions, considering the increasing prevalence of azole-resistant A. fumigatus and non-fumigatus Aspergillus spp., with variable susceptibility profiles. Second, a negative GM EIA is not adequate enough to rule out a bIMI, considering the high numbers of non-Aspergillus bIMI, due to Mucorales and other non-GM producing molds. Third, a bronchoalveolar lavage (or other specimens) on which a fungal culture can be performed is crucial and represents the backbone of the diagnosis, because it may allow for the identification of a fungal pathogen and subsequent susceptibility testing. Fourth, molecular testing using A. fumigatus, Mucorales multiplex, or panfungal polymerace chain reaction (PCR) assays on blood and BAL samples may increase the diagnostic yield and allow for the identification of an organism [36, 37]. Finally, antifungal TDM at the time of bIMI diagnosis may be informative and hence still recommended to complete the relevant diagnostic work-up of patients with bIMI.

Antifungal treatment of posaconazole-bIMI includes initiation of empirical treatment with an amphotericin-B formulation, with close clinical, biological, and radiological monitoring of the patient. This empiric approach might allow for treatment of either broad-spectrum resistant Aspergillus spp. or Mucorales. In case a pathogen is identified and a susceptibility profile becomes available, antifungal treatment should be tailored accordingly. If a pathogen is identified by PCR and no culture or antibiogram are available, treatment should be adjusted based on susceptibility data from the existing body of literature. Finally, in case that no pathogen is identified, transition from amphotericin-B to another broad-spectrum azole (e.g. isavuconazole) could be considered, after several weeks of empirical treatment and considering the clinical and radiological evolution of the patient. Immunosuppression treatment should be tapered, if and when feasible.

Conclusion

The incidence of bIMI is increasing in the last years with more rare and challenging fungi identified as significant pathogens, leading to high mortality rates, despite progress in the diagnostic and therapeutic domains. Establishing a microbiological diagnosis remains problematic, with most infections remaining possible or probable, based on PCR rather than culture-based assays. As a result, treatment is based on local epidemiology and expertise, further underscoring the complexity of the management of bIMI. Factors leading to bIMI are variable, predominantly associated with the administration of routine universal broad-spectrum antifungal prophylaxis in high-risk patient populations, changing epidemiology worldwide, and the underlying patient immune status. Recent research efforts, focusing on fully understanding individual susceptibility to IMI and tailoring prophylactic strategies based on underlying genetic risk factors may shed some more light and contribure further to the ongoing battle between host, pathogen, and environment.