Antifungal pre-emptive strategy for high-risk neutropenic patients: why the story is still ongoing

Transcription

1 REVIEW / Antifungal pre-emptive strategy for high-risk neutropenic patients: why the story is still ongoing C. Cordonnier 1, C. Robin 1, A. Alanio 2,3 and S. Bretagne 2,3,4 1) Haematology Department, Henri Mondor Hospital, Assistance Publique-H^opitaux de Paris (AP-HP) and University Paris-Est-Creteil, Creteil, 2) Parasitology Mycology Department, Saint-Louis Hospital, AP-HP and University Paris Diderot, 3) Institut Pasteur, Molecular Mycology Unit, CNRS URA 3012 and 4) Institut Pasteur, Centre National de Reference des Mycoses Invasives et Antifongiques, Paris, France Abstract Neutropenic patients with haematological malignancies are at high risk of invasive fungal disease (IFD). Due to limitations in specific procedures to establish an early diagnosis of IFD, two historical unpowered studies suggested, three decades ago, that giving an empirical antifungal treatment to patients with persistent or recurrent fever under broad-spectrum antibacterials, could reduce the risk of IFD. For cost and toxicity reasons, this strategy became debated when modern imaging and indirect biological markers became available. Different pre-emptive strategies, either based on lung imaging, galactomannan antigenaemia, fungal PCR, or a combination of several parameters, were designed with the goal of restricting the administration of antifungals to the more at-risk patients with early signs of IFD. Almost all pre-emptive studies showed or suggested a reduction of administration and cost of antifungals during neutropenic phases. However, the clinical pertinence and safety of the strategy, and mainly its optimal design, are still pending. This paper reviews the evolution of these strategies and how they may be implemented in the haematology ward. Keywords: Empirical antifungal therapy, fungal PCR, galactomannan, invasive fungal disease, invasive fungal infection, neutropenia, pre-emptive antifungal therapy Article published online: 27 November 2013 Clin Microbiol Infect 2014; 20 (Suppl. 6): Corresponding author: C. Cordonnier, Haematology Department, Henri Mondor Hospital, Creteil, France s: carlcord@club-internet.fr; catherine.cordonnier@hmn.aphp.fr Introduction Prolonged neutropenia is a main risk factor for invasive fungal disease (IFD). Patients with haematological malignancies, and especially those with acute myeloid leukaemia (AML), or those who undergo allogeneic, myeloablative, haematopoietic stem cell transplantation (HSCT) have chemotherapy-induced neutropenic phases of long duration classically more than 10 days which make them at higher risk of IFD [1]. Although new antifungal drugs are available, the 12-week mortality rate of IFD remains between 29% [2] and 47.8% [3] in invasive aspergillosis (IA), and between 29.6% [4] and 40% [5] in invasive candidiasis. Additionally, in patients with acute leukaemia, even though the patient initially survives the infection, IFD has a negative impact on the overall outcome, by modifying the choice and timing of chemotherapy or of the conditioning regimen of stem cell transplantation [6]. To avoid the onset of IFD in neutropenic patients, three approaches have been developed, which are often combined: antifungal prophylaxis, empirical antifungal therapy and pre-emptive antifungal approaches. Before efficient antifungal prophylaxis was available, and before indirect biological markers and effective imaging were assessed, the only acceptable approach was empirical antifungal therapy in patients with persistent or recurrent unexplained fever refractory to broad-spectrum antibiotics [7,8]. The present manuscript will review the development of pre-emptive antifungal approaches in high-risk neutropenic patients. Clinical Microbiology and Infection ª2013 European Society of Clinical Microbiology and Infectious Diseases

2 28 Clinical Microbiology and Infection, Volume 20 Supplement 6, June 2014 CMI Empirical Antifungal Therapy in High-risk Neutropenic Patients: an Old, and Debated Story Identified a long time ago from autopsy series, the risk of fungal death is a major concern in haematology [3,9 12]. Following the success of the antibacterial empirical approach in febrile neutropenic patients [13], it was hypothesized that an early empirical antifungal treatment could reduce the risk of IFD in this setting, at a time when there were no sensitive imaging or biological procedures for IFD, no routine use of bronchoalveolar lavage, and almost no other efficient anti-mould therapy than deoxycholate amphotericin-b. This approach was explored in two historical, open, studies whose conclusions are criticized 30 years later for methodological reasons [14,15]. In the first study, 50 febrile neutropenic patients on antibacterial agents for 7 days were randomized to continuing with the same antibacterials only, stopping the antibacterials, or adding deoxycholate amphotericin-b to the antibacterials [8]. The administration of amphotericin-b was associated with a decreased number of IFD and of fungal deaths (1/18 vs 4/16 and 3/16 in the other groups, respectively). This was the starting point for a larger study including 132 neutropenic patients who were randomized from day 4 of fever to additionally receive, or not receive, deoxycholate amphotericin-b [7]. Although there was a lower number of IFD in the deoxycholate amphotericin-b group (1/68 vs 4/64), the difference was not significant but there were fewer fungal deaths (0/68 vs 4/64, p 0.05). Although short on facts, the strategy was widely adopted by the haematology community and was later reinforced by papers showing that any delay in the treatment of IFD is deleterious [16,17]. However, these studies have many limitations: unblinded design, few patients, no difference in the IFD rate [15]. Almost all new antifungal drugs, especially liposomal amphotericin-b [18], voriconazole [19] and caspofungin [20], were assessed in the empirical indication. The investigators initially used deoxycholate amphotericin-b as the standard [18], then liposomal amphotericin-b [19,20], which was as effective and less toxic than amphotericin-b and was associated with less breakthrough IFD [18]. Caspofungin was as effective as liposomal amphotericin-b but generally better tolerated [20]. A large study failed to show that voriconazole was non-inferior to liposomal amphotericin-b [19]. Other antifungals were also evaluated [15]. All these trials raised methodological concerns, using sophisticated, composite endpoints, mainly based on drop in fever, or focusing on toxicity as a primary endpoint [21 23]. Years later, the empirical strategy is widely used, easy to apply, does not need costly equipment, and is still endorsed by international guidelines [15,24]. With such a strategy, 30 70% of febrile neutropenic patients receive antifungals, usually from day 4 5 of persistent fever to the end of neutropenia. Overall, one may consider that the empirical approach is more to be debated than the choice of the optimal antifungal drug in this indication, but we now have indirect markers or imaging procedures to develop a more targeted strategy. The Concept of a Pre-emptive, Diagnosticdriven Strategy There is no consensual definition of a pre-emptive antifungal strategy, but one common goal is to restrict the administration of antifungals to patients at the very early phase of IFD, and so decrease toxicity, costs and risk of emerging resistance [25,26]. Some authors prefer alternative wordings like diagnosis-driven or targeted, but this is a sterile debate because the optimal strategy is not defined. Most of these approaches have mainly focused on the risk of filamentous infections, mainly IA, using galactomannan (GM) screening and computed tomography (CT) scans. The wording pre-emptive has been adopted by the haematology community on the experience of pre-emptive therapy for cytomegalovirus infection after HSCT, a situation where infection and disease are clearly distinguished, and where giving an antiviral drug on the basis of a biomarker without symptoms prevents the onset of disease [27 29]. The fungal issue is not the same: Many studies have explored the benefit of GM or PCR screening in neutropenic patients. However, no study compared a strategy where such a marker triggered antifungal administration with a strategy of wait-and-see for IFD. The only comparative trials were done with an empirical, fever-based strategy, as the control arm [30 34]. Some authors found that GM does not precede major signs of IA on CT scan [35]. Similarly, there is no proof that serum or blood fungal PCR, performed according to the more recent guidelines [36], either precedes or clearly diagnoses IA. Indeed, the fungal risk going from no IFD to proven/probable/ possible IFD during neutropenia has to be thought of as a daily increasing process because time (duration of neutropenia) is essential in the fungal risk [37 39]. This tiny margin is well illustrated by this definition of pre-emptive therapy [40]: Between over- and undertreatment of invasive fungal disease. In fact, in most pre-emptive studies, the investigators combined biological, clinical and imaging triggers to administer

3 CMI Cordonnier et al. Pre-emptive antifungal treatment during neutropenia 29 antifungals, in order to be in the best position to catch IFD at a very early stage. mortality rates (22.5% vs 7.1%, and 15.9% vs 6.3%, respectively, p 0.002) were significantly higher in the pre-emptive group. Feasibility, and Observational or Retrospective Studies on Pre-emptive Antifungal Strategies One of the main concerns in assessing pre-emptive strategies is the optimal methodology to use. Blinded or placebocontrolled studies have not been performed and would be practically difficult. Several studies were described as observational but in fact were retrospective (Table 1). The feasibility of a pre-emptive approach in high-risk neutropenic patients was first shown by Maertens et al. [41]. The authors assessed 136 high-risk neutropenic episodes in 88 haematology patients. Most patients had AML, all received fluconazole prophylaxis, and were managed in HEPA-filtered rooms. Patients were assessed daily by a blood GM test. Chest CT scan, and eventually a bronchoalveolar lavage were performed in the case of two consecutive GM tests (index >0.5), persistent fever for >5 days, or recurrent fever, new pulmonary infiltrate on chest X-ray, or mould colonization. Patients received liposomal amphotericin-b only where there was IFD. The strategy reduced the use of antifungals from 35% (the rate expected in an empirical approach) to 7.7%. Additionally, the GM test was positive in 10 afebrile patients. One case of zygomycosis and two cases of invasive candidiasis were not detected by the strategy. Finally, the incidence of proven/probable IFD was 15% per episode, and 24% per patient. From this preliminary experience, the haematologists explored alternative strategies to empirical therapy. Several studies were reported, based on imaging only [42], combinations of GM and CT scan [43], purely clinical criteria [44], or on an intensive diagnostic work-up triggered by persistent fever or clinical suspicion of IFD [45] (summarized in Table 1). However, many methodological issues preclude clear conclusions, the main caveat being the lack of comparison to alternative approaches and especially to a purely empirical, or prophylactic approach. Most authors agree on the hypothesized reduction (>50%) of antifungal consumption by the pre-emptive approach [41,45]. However, three issues deserve comment: (i) the IFD incidence rates in these studies were often in the higher range for such populations in the pre-emptive studies [1,41,45]; (ii) survival (when given) was never compared with that of other strategies; (iii) the larger observational study with 397 patients [43] managed either with an empirical or a pre-emptive approach, showed that the rate of IFD (23.7% vs 7.4%, p <0.001) and the attributed and overall Comparative Trials Few randomized studies, comparing a pre-emptive with an empirical strategy, have been published (Table 2). Hebart et al. assessed the use of a panfungal PCR during the first 3 months following allogeneic HSCT [32]. They compared a fever-driven antifungal strategy with a hybrid both fever-- driven and PCR-driven strategy in 403 transplant recipients. Although survival was better at day 30 in the PCR group, this benefit was no longer found at day 100. Moreover, more patients received antifungals in the PCR-based group when compared with the empirical group. This can be explained in part by the addition of an empirical approach after 5 days of persistent febrile neutropenia in the PCR-driven group, which led to the provision of antifungals to PCR-negative patients, a strategy that disagrees with the negative predictive value of the PCR shown by other authors [46]. Additionally, the high rate of PCR positivity (28.3% of the patients) might suggest false positives with the endpoint format of the PCR used at that time [47,48]. Another study based on a weekly whole-blood real-time PCR targeting Candida and Aspergillus spp. evaluated 99 patients who received a reduced-intensity conditioning allogeneic HSCT [30]. In the case of a positive PCR, the patients were planned to be randomized for liposomal amphotericin-b treatment. Patients with fever of unknown origin for >3 days despite administration of antibacterial agents were also permitted to receive antifungals, regardless of the PCR results. Only 21 of the 41 PCR-positive patients were randomized. The rate of proven or probable IFD was not different in the patients who received liposomal amphotericin-b on the basis of PCR (0/8) when compared with the patients who did not (1/13). Finally, the small number of patients, the lack of sensitivity of the pan-fungal PCR, and the once-a-week screening do not permit conclusions to be drawn on the relevance of the strategy. In France, we ran a randomized trial comparing a purely empirical versus a pre-emptive approach in 293 high-risk neutropenic patients [31,39]. With 67% of AML patients, the median duration of neutropenia <500/mm 3 was 17.5 days. Patients undergoing allogeneic HSCT were excluded. In the empirical group, the administration of conventional or liposomal amphotericin-b was exclusively fever-driven. In the pre-emptive group, patients had a twice-weekly GM test from the start of chemotherapy, and received deoxycholate amphotericin-b or liposomal amphotericin-b only in the case of one GM index being 1.5, or in the case of predefined clinical

4 30 Clinical Microbiology and Infection, Volume 20 Supplement 6, June 2014 CMI TABLE 1. Summary of the main observational studies on antifungal pre-emptive approaches in high-risk neutropenic patients Reference Design of the study No. of patients or episodes Inclusion of allogeneic HSCT recipients (%) Biomarker (s) used and frequency of screening Indication of antifungal (s) in the pre-emptive group Antifungal used Incidence (%) of IFD (Proven + probable) Main limitations Maertens et al. [41] Prospective, open, monocentric Daily screening with serum GM, eventually thoracic ( sinus) CT scan and BAL Dignan et al. [42] Retrospective, consecutive, monocentric CT scan in case of persistent fever (>72 h) under broad-spectrum antibacterials. CT scan repeated days later Aguilar-Guisado et al. [44] Girmenia et al. [45] Prospective, open, monocentric Prospective, open, monocentric Pagano et al. [43] Registry, multicentric study, retrospective comparison of patients managed with an empirical or a pre-emptive approach, choice left to the initiative of the investigator 136 episodes/117 febrile neutropenic episodes/88 patients 99 patients All (myeloablative: 36; reduced-intensity: 63) 66 patients with persistent febrile neutropenia 220 episodes, 146 patients 32/88 (23.5%) patients Daily GM Two consecutive positive (index 0.5) GM or CT suggestive of IFD, with positive culture or microscopy for moulds None Caspofungin started if CT scan and/or at the discretion of the physician 9 (14%) patients None Clinically driven (severe sepsis, septic shock, focused infection), or individualized clinical decision None GM for 3 consecutive days in case of persistent fever after 4 days of antibacterials, or recurrent fever, or any clinical finding evocative of IFD 397 patients None GM: 1 or 2/week and CT scan Proven, probable, or possible IFD Left to the discretion of the investigator in the pre-emptive group Persistent fever with an unknown source in the empirical group L-ampho-B 15%/episode 24%/patient Caspofungin (eight patients received an alternative 1st line treatment) L-ampho-B or D-ampho-B or voriconazole Left to the discretion of the investigator Left to the discretion of the investigator Daily performance of GM difficult to implement in many centres Re-inclusions High rate of IFD 4% No GM performed Pneumonia investigation by BAL at the discretion of the physician 15 patients with a previous IFD 4.5% No GM performed Overall mortality at day 30: 15.1% 14.5% per episode Re-inclusions Different antifungals High rate of IFD Overall mortality at 3 months: 24.6% 7.4% in the empirical group vs 23.7% in the pre-emptive group (p <0.001) Strategy and choice of antifungal left to the local investigator No protocol fixed Different screening time for GM from centre to centre BAL, bronchoalveolar lavage; CT, computed tomography; D-ampho-B, deoxycholate amphotericin-b; GM, galactomannan; HSCT, haematopoietic stem cell transplant; IFD, infectious fungal disease; L-ampho-B, liposomal amphotericin-b.

5 CMI Cordonnier et al. Pre-emptive antifungal treatment during neutropenia 31 TABLE 2. Randomized studies comparing antifungal pre-emptive (PE) approaches versus empirical (E) approaches Reference Design of the study Control arm Primary endpoint No. of patients or episodes Percentage of allogeneic HSCT recipients Biomarker(s) used and frequency of screening Indication of antifungal (s) in the pre-emptive group Antifungal used Incidence (%) of IFD (Proven + probable) Conclusion on the primary objective Main limitations Hebart et al. [32] Cordonnier et al. [31] Blennow et al. [30] Morissey et al. [33] Randomized unblinded Randomized unblinded Randomized unblinded Randomized unblinded Empirical (5 days of fever and/or pneumonia) Empirical (4 days of persistent fever or recurrent fever) Whole blood PCR once a week. If +, randomization to receive, or not to receive L-ampho-B Standard diagnostic approach: persistent fever (3 5 days), blood, urine and stool cultures, and high-resolution chest CT scan Incidence of IFD till day 100 after HSCT 403 patients 100% Panfungal PCR 92/w 5 days of fever or pneumonia or one PCR + Survival at day patients 0% GM 92/week GM 1.5 or septic shock, pneumonia, sinusitis Incidence of proven and probable IFD till day 100 after HSCT Proportion of patients receiving at least one course of empirical antifungals during the 26 weeks after enrolment 99 patients, all receiving reduced-intensity conditionings 100% Panfungal PCR 91/w 240 patients 80% GM and nested PCR 92/week in inpatients, and 91/week in outpatients Lung CT-scan if persistent fever (even if negative biological markers) PCR +=> L-ampho-B at random or 3 days of fever of unknown origin refractory to antibacterials Criteria of probable or possible IFD L-ampho-B 8.1% (PE) vs 8.2%) (E) Using PCR does not provide any significant difference on final IFD incidence D- or L-ampho-B according to renal function 9.1% (PE) vs 2.7% (E) Survival not inferior, but IFD more frequent in the PE group L-ampho-B Proven + probable: 12% at 1 year after HSCT No difference in the PCR + group between patients who received (0/8) or who did not receive (1/13) L-ampho-B PCR is a poor indicator of IFD early after RIC HSCT Not fixed 24.5% (PE) vs 4.1% (E) Patients significantly received less empirical antifungals in the PE group (15%) than in the classical group (32%) Missing PCR samples during the first 5 weeks Addition of empirical approach and PCR use in the PE group No allogeneic HSCT patient Prophylaxis in only 17% of the cohort Some low-risk patients (15% autologous HSCT). Could not rule out the superiority of the E approach on survival in the more at-risk population (AML induction) Low number of PCR+ randomized patients PCR once a week Modified definitions of IFD, different in both groups Several antifungals used Heterogeneous prophylaxis AML, acute myeloid leukaemia; D-ampho-B, deoxycholate amphotericin-b; GM, galactomannan; HSCT, haematopoietic stem cell transplant; IFD, invasive fungal disease; L-ampho-B, liposomal amphotericin-b; RIC, reduced-intensity conditioning.

6 32 Clinical Microbiology and Infection, Volume 20 Supplement 6, June 2014 CMI symptoms or septic shock. Overall survival at day 42 after inclusion (primary endpoint) was not statistically different in both groups (97% in the empirical group vs 95% in the pre-emptive group). As expected, amphotericin-b was given to fewer patients and for shorter durations in the pre-emptive group. However, two main issues pushed our group to go further: (i) the incidence of proven/probable IFD was significantly higher in the pre-emptive group both in the whole population (9.1% vs 2.7%; p <0.02) and in the higher risk subgroup (AML inductions) (16.4% vs 3.8%; p <0.01), and (ii) in this subgroup, we could not rule out the inferiority of the pre-emptive approach on overall survival. Knowing the high risk of IFD in this subgroup of patients, any doubts concerning a loss in survival or gain of IFD risk makes it difficult to implement the pre-emptive strategy without further studies. Second, the difference in IFD incidence cannot be considered as incidental because this may impact on outcome [6]. A similar trial was run by Tan et al. [34] but was underpowered to draw any conclusion. Recently, Morrissey et al. compared a hybrid empirical approach (fever and CT scan driven), with a hybrid pre-emptive approach using twice-weekly blood testing with GM and two-nested PCR qualitative assays targeting Aspergillus spp. A CT scan was performed in the case of positive biomarker(s) or of persistent fever [33]. Among the 240 included patients, most were allogeneic HSCT recipients. The choice of antifungals was left to the initiative of the investigator. The use of empirical antifungal drugs at week 26 (the primary endpoint) was significantly lower in the pre-emptive compared with the empirical group (15% vs 32%; p 0.002). Overall survival was not different between groups but the trial was not powered to assess survival. IFD, and especially probable IA, were significantly more frequent in the pre-emptive group than in the standard group (24.5% vs 4.1%; p <0.0001) but different definitions of IFD were used in the two groups. Additionally, 11 of the 16 probable IA were documented through biological markers only (GM and PCR: three; PCR alone: eight) without any site of infection. Because the formats of the PCR assays were nested and because the primers were designed to also detect environmental Aspergillus species, the risk of false PCR results from amplicon contamination and from environmental mould DNA may have been high. Therefore, the results of this trial are difficult to interpret. Is There Any Optimal Biomarker for Neutropenic Patients? Due to the predominance of IA over candidiasis in haematology patients, GM has been logically the most explored method, both as an indirect marker of IA [49 51] and as a guide for pre-emptive antifungal therapy in neutropenic patients [31,33,41,43,45]. Due to the diversity of the PCR assays [52], the role of PCR, both in the diagnosis of IA and in guiding a pre-emptive approach, remains investigative. Despite pitfalls, and especially false-positive results, the advantage of GM over PCR here is the result of the standardization of the GM test and its strong validation by autopsies [53]. Although PCR has been also explored for the early diagnosis of IA, the main concerns about laboratory procedures [36,52] clearly moderate the conclusions of the meta-analysis [54] performed on studies published before the European consensus on PCR [36]. Consequently, we consider its use as still exploratory, pending large studies fitting with the European Aspergillus PCR Initiative guidelines [55]. b-d-glucan, which is a cell wall component of most fungi including Aspergillus spp. and Candida spp. has been the purpose of two meta-analyses in haematology patients [56,57]. The sensitivity of the test appears to be lower in the haematology ward than in the intensive care unit. b-d-glucan has so far not been used for guiding a pre-emptive strategy in neutropenic patients. One important issue to consider in assessing the diagnostic yield of a fungal biomarker is the time-variation of the fungal risk. What we expect from indirect biomarkers in this setting is to anticipate overt disease. This is why we chose to perform a time-dependent analysis in the translational research of the French study [31]. Serum samples were frozen twice-weekly for b-d-glucan and quantitative PCR, but the results were unknown to the clinicians. Later, we assessed their performance for the early detection of IA through a multi-state model in 185 high-risk patients with AML, taking into account the daily evolution of the risk over time during neutropenia [39]. With this approach, quantitative PCR did accelerate the early detection of IA, while b-d-glucan did not. In the setting of high-risk neutropenia, and especially in patients receiving yeast prophylaxis, the main concern remains IA, and the b-d-glucan test probably does not add a lot to GM, whereas PCR could be useful in combination with GM [39]. This hypothesis, however, requires confirmation. How to Go Further? The reasons why several trials did not solve the question of whether to give up the empirical approach are multiple. The design of prospective pre-emptive antifungal studies is a recent issue and not yet fixed. Many methodological concerns emerge.

7 CMI Cordonnier et al. Pre-emptive antifungal treatment during neutropenia 33 First, assessing the predictive value of biomarkers on the diagnosis of IFD, and assessing a new clinical strategy based on the same biomarkers and its impact on outcome (survival, IFD incidence, cost) are different goals. Second, because of the evolution of the practices and availability of new procedures, the control, so-called empirical, group is often no longer the same as in the historical empirical approach, i.e. purely fever-based [7,8]. In most centres, the empirical approach is now combined with CT scans and/or GM use, whereas in pre-emptive approaches, fever is sometimes added to the list of the criteria for when to give antifungals [30,33]. This creates some overlap between the two strategies, and some confusion to implement conclusions of prospective trials in routine practice. Additionally, the use of an approach initially designed for neutropenic patients [8] in patients who may not be neutropenic is confusing. This is the case for allogeneic HSCT recipients, when this strategy is applied until 3 months or more after transplant [32,33,42]. Administration of antifungals in immunocompromised but non-neutropenic patients may often be delayed by 1 3 days to allow direct diagnostic procedures, especially bronchoalveolar lavage in the case of pneumonia, to cover several infectious causes, and the strategy should not be the same as in neutropenic patients. The optimal endpoints of such studies are debated. From a haematological point of view, survival, and preferably fungal-free survival, of deep neutropenia is the first requisite to benefit from complete remission or engraftment. IFD is a tricky endpoint because the use of biomarkers and early CT scans in pre-emptive groups may favour the identification of IFD, especially of probable IFD, because the second version of the European Organization for Research and Treatment of Cancer-Mycoses Study Group (EORTC-MSG) definitions [58] where one positive GM test and a lung lesion are enough for a diagnosis of probable IA. As for antifungal use and cost, there is no doubt that any pre-emptive approach decreases antifungal use during neutropenia [31,33,41]. However, the administration of antifungals to treat IFD after neutrophil recovery and the costs of diagnostic procedures should also be considered. Whether prophylactic measures impact on the strategy efficacy is unclear. Marr et al. have suggested that mould-active antifungals could decrease the sensitivity of GM [59]. In our experience, laminar-air-flow rooms were protective against IA [39]. The choice of prophylactic measures should logically impact on the choice of the strategy during neutropenia, but few data are available. Similarly, the use of scores to assess the fungal risk should help [60]. In our experience [31], by combining GM and clinical criteria, the pre-emptive administration of antifungals was more often triggered by pulmonary symptoms than by positive GM tests. On the other hand, 40% of our patients with persistent or recurrent fever in the empirical group had some concomitant clinical symptom, in addition to fever, when antifungals were started. So, daily clinical assessment remains essential, even when using the most sophisticated diagnostic procedures. What Next? Using the experience of the French trial and its pitfalls [31], a large EORTC prospective trial comparing a purely empirical approach to a GM and CT scan based pre-emptive approach has been designed (NCT ) with some differences to improve the power of the study: All the patients receive fluconazole prophylaxis. Only high-risk patients, namely patients with AML at induction and allogeneic myeloablative transplant recipients at the neutropenic phase, are included. Caspofungin is used in both arms instead of amphotericin-b. However, caspofungin and liposomal amphotericin-b are both licensed for the empirical indication. We hope that this study will clarify the conflicting results of the previous studies and provide the clinician with a useful message for daily practice. Conclusion Many studies assessing pre-emptive antifungal strategies have been published or are running. This occurs in parallel with considerable progress in antifungal prophylaxis of neutropenic patients. Two large studies showed a significantly higher incidence of IFD with a pre-emptive strategy when compared with an empirical strategy [31,43] and no large study has so far shown that the overall survival is not impaired by a pre-emptive approach in the more at-risk patients. Pending the results of additional trials, the empirical approach remains easy, reproducible and safe when neutropenia lasts more than days. It is a costly strategy in terms of antifungals, but cheap in terms of diagnostic procedures. For neutropenia of shorter duration (<10 days), the benefit of both strategies is similar, and both are even debatable considering the low risk of IFD in that setting [31]. In the case of efficient anti-mould prophylaxis by drugs or environmental procedures, the low risk of IFD

8 34 Clinical Microbiology and Infection, Volume 20 Supplement 6, June 2014 CMI should favour a pre-emptive strategy but a low performance of the biomarkers should be expected. Pre-emptive strategies in other settings should be explored prospectively and compared, and eventually combined, with the more efficient prophylaxis. Transparency Declaration C. Cordonnier has received research grants from Pfizer and Gilead Sciences. She has been an advisor or speaker for Astellas, BioMerieux, Gilead Sciences, MSD, Pfizer, Schering-Plough, and Zeneus; C. Robin has nothing to declare; A. Alanio has received travel grants from MSD and Gilead Sciences and speaking honoraria from Gilead Sciences; S. Bretagne is a consultant for Gilead Sciences, and has received speaking honoraria from Pfizer and Gilead Sciences and travel grants from Astellas, Pfizer and Schering-Plough. References 1. Pagano L, Akova M, Dimopoulos G, Herbrecht R, Drgona L, Blijlevens N. Risk assessment and prognostic factors for mould-related diseases in immunocompromised patients. J Antimicrob Chemother 2011; 66 (Suppl. 1): i5 i Herbrecht R, Denning DW, Patterson TF et al. Voriconazole versus amphotericin B for primary therapy of invasive aspergillosis. N Engl J Med 2002; 347: Nivoix Y, Velten M, Letscher-Bru V et al. Factors associated with overall and attributable mortality in invasive aspergillosis. Clin Infect Dis 2008; 47: Pappas PG, Rotstein CM, Betts RF et al. Micafungin versus caspofungin for treatment of candidemia and other forms of invasive candidiasis. Clin Infect Dis 2007; 45: Kuse ER, Chetchotisakd P, da Cunha CA et al. Micafungin versus liposomal amphotericin B for candidaemia and invasive candidosis: a phase III randomised double-blind trial. Lancet 2007; 369: Even C, Bastuji-Garin S, Hicheri Y et al. Impact of invasive fungal infection on the chemotherapy schedule and event-free survival in acute leukemia patients who survive to invasive fungal infection: a case control study. Haematologica 2011; 96: EORTC International Antimicrobial Therapy Cooperative Group. Empiric antifungal therapy in febrile granulocytopenic patients. Am J Med 1989; 86: Pizzo P, Robichaud KJ, Gill FA, Witebsky FG. Empirical antibiotic and antifungal therapy for cancer patients with prolonged fever and granulocytopenia. Am J Med 1982; 72: Chamilos G, Luna M, Lewis R et al. Invasive fungal infections in patients with hematologic malignancies in a tertiary care cancer center: an autopsy study over a 15-year period ( ). Haematologica 2006; 91: Lewis R, Cahyame-Zuniga L, Leventakos K et al. Epidemiology and sites of involvement of invasive fungal infections in patients with haematological malignancies: a 20-year autopsy study. Mycoses 2013; 56: Marr KA, Carter RA, Crippa F, Wald A, Corey L. Epidemiology and outcome of mould infections in hematopoietic stem cell transplant recipients. Clin Infect Dis 2002; 34: Pagano L, Caira M, Candoni A et al. Invasive aspergillosis in patients with acute myeloid leukemia: SEIFEM-2008 registry study. Haematologica 2010; 22: Bodey GP, Buckley M, Sathe YS, Freireich EJ. Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intern Med 1966; 64: Klastersky J. Empirical antifungal therapy. Int J Antimicrob Agents 2004; 23: Marchetti O, Cordonnier C, Calandra T. Empirical antifungal therapy in neutropaenic cancer patients with persistent fever. Eur J Cancer 2007; S5: Chamilos G, Lewis R, Kontoyiannis D. Delaying amphotericin-b based frontline therapy significantly increases mortality among patients with hematologic malignancy who have zygomycosis. Clin Infect Dis 2008; 47: von Eiff EM, Roos N, Schulten R et al. Pulmonary aspergillosis: early diagnosis improves survival. Respiration 1995; 62: Walsh TJ, Finberg RW, Arndt C et al. Liposomal amphotericin B for empirical therapy in patients with persistent fever and neutropenia. N Engl J Med 1999; 340: Walsh TJ, Pappas P, Winston DJ et al. Voriconazole compared with liposomal amphotericin B for empirical antifungal therapy in patients with neutropenia and persistent fever. N Engl J Med 2002; 346: Walsh TJ, Teppler H, Donowitz GR et al. Caspofungin versus liposomal amphotericin B for empirical antifungal therapy in patients with persistent fever and neutropenia. N Engl J Med 2004; 351: Prentice H, Hann I, Herbrecht R et al. A randomized comparison of liposomal versus conventional amphotericin B for the treatment of pyrexia of unknown origin in neutropenic patients. Br J Haematol 1997; 98: White M, Bowden R, Sandler E et al. Randomized, double-blind clinical trial of amphotericin B colloidal dispersion vs. amphotericin B in the empirical treatment of fever and neutropenia. Clin Infect Dis 1998; 27: Wingard JR, White MH, Anaissie E, Raffali J, Goodman J, Arrieta A. A randomized, double-blind comparative trial evaluating the safety of liposomal amphotericin B versus amphotericin B lipid complex in the empirical treatment of febrile neutropenia. Clin Infect Dis 2000; 31: Freifeld AG, Bow EJ, Sepkowitz KA et al. Clinical practice guidelines for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the infectious diseases society of America. Clin Infect Dis 2011; 52: e56 e Alanio A, Cabaret O, Sitterle E et al. Azole preexposure affects the Aspergillus fumigatus population in patients. Antimicrob Agents Chemother 2012; 56: Alanio A, Cordonnier C, Bretagne S. Azole resistance in Aspergillus fumigatus current epidemiology and future perspectives. Current Fungal Infect Rep 2011; 5: Boeckh M, Gooley T, Myerson D, Cunningham T, Schoch G, Bowden R. Cytomegalovirus pp65 antigenemia-guided early treatment with ganciclovir versus ganciclovir at engraftment after allogeneic marrow transplantation: a randomized double-blind study. Blood 1996; 88: Ljungman P, De La Camara R, Cordonnier C et al. Management of CMV, HHV6, HHV7 and Kaposi-sarcoma herpesvirus infections in patients with hematological malignancies including stem cell transplantation. Bone Marrow Transplant 2008; 42: Ljungman P, Griffiths P, Paya C. Definitions of cytomegalovirus infection and disease in transplant recipients. Clin Infect Dis 2002; 34: Blenow O, Remberger M, Klingspor L et al. Randomized PCR-based therapy and risk factors for invasive fungal infection following

9 CMI Cordonnier et al. Pre-emptive antifungal treatment during neutropenia 35 reduced-intensity conditioning and hematopoietic SCT. Bone Marrow Transplant 2010; 45: Cordonnier C, Pautas C, Maury S et al. Empirical versus pre-emptive antifungal strategy in high-risk febrile neutropenic patients: a prospective randomized study. Clin Infect Dis 2009; 48: Hebart H, Klingspor L, Klingebiel T et al. A prospective randomized controlled trial comparing PCR-based and empirical treatment with liposomal amphotericin B in patients after Allo-SCT. Bone Marrow Transplant 2009; 43: Morrissey O, Chen S, Sorrell T et al. Galactomannan and PCR versus culture and histology for directing use of antifungal treatment for invasive aspergillosis in high-risk hematology patients: a randomised controlled trial. Lancet Infect Dis 2013; 13: Tan B, Low J, Chlebicka N et al. Galactomannan-guided preemptive vs. empirical antifungals in the persistently febrile neutropenic patient: a prospective randomized study. Int J Infect Dis 2011; 15: Weisser M, Rausch C, Simcock M et al. Galactomannan does not precede major signs on a pulmonary computerized tomographic scan suggestive of invasive aspergillosis in patients with hematological malignancies. Clin Infect Dis 2005; 41: White P, Bretagne S, Klingspor L et al. Aspergillus PCR: one step closer towards standardisation. J Clin Microbiol 2010; 48: Gerson S, Talbot G, Hurwitz S, Strom B, Lusk E, Cassileth PA. Prolonged granulocytopenia: the major risk factor for invasive pulmonary aspergillosis in patients with acute leukemia. Ann Intern Med 1984; 100: Muhlemann K, Wenger C, Zenhausern R, Tauber MG. Risk factors for invasive aspergillosis in neutropenic patients with hematologic malignancies. Leukemia 2005; 19: Schwarzinger M, Sagaon-Teyssier S, Cabaret O, Bretagne S, Cordonnier C. Performance of serum biomarkers for the early detection of invasive aspergillosis in febrile, neutropenic patients: a multi-state model. PlosOne 2013; 8: e De Pauw BE. Between over- and undertreatment of invasive fungal disease. Clin Infect Dis 2005; 41: Maertens J, Theunissen K, Verhoef G et al. Galactomannan and computed tomography-based preemptive antifungal therapy in neutropenic patients at high risk for invasive fungal infection: a prospective feasibility study. Clin Infect Dis 2005; 41: Dignan FL, Evans SO, Ethell ME et al. An early CT-diagnosis based treatment strategy for invasive fungal infection in allogeneic transplant recipients using caspofungin first line: an effective strategy with low mortality. Bone Marrow Transplant 2009; 44: Pagano L, Caira M, Nosari A et al. The use and efficacy of empirical versus pre-emptive therapy in the management of fungal infections: the HEMA e-chart project. Haematologica 2011; 96: Aguilar-Guisado M, Espigado I, Cordero E et al. Empirical antifungal therapy in selected patients with persistent febrile neutropenia. Bone Marrow Transplant 2010; 45: Girmenia C, Micozzi A, Gentile G et al. Clinically driven diagnostic antifungal approach in neutropenic patients: a prospective feasibility study. J Clin Oncol 2010; 28: Barnes RA, White PL, Bygrave C, Evans N, Healy B, Kell J. Clinical impact of enhanced diagnosis of invasive fungal disease in high-risk haematology and stem cell transplant patients. J Clin Pathol 2009; 62: Einsele H, Hebart H, Roller G et al. Early detection and identification of Candida and Aspergillus species in blood using molecular probes. J Clin Microbiol 1997; 35: Loeffler J, Hebart H, Sepe S, Schumacher U, Klingebiel T, Einsele H. Detection of PCR-amplified fungal DNA by using a PCR-ELISA system. Med Mycol 1998; 36: Leeflang M, Debets-Ossenkopp Y, Visser C et al. Galactomannan detection for invasive aspergillosis in immunocompromized patients. Cochrane Database Syst Rev 2008; 8: CD Maertens J, Verhaegen J, Lagrou K, Van Eldere J, Boogaerts M. Screening for circulating galactomannan as a noninvasive diagnostic tool for invasive aspergillosis in prolonged neutropenic patients and stem cell transplantation recipients: a prospective validation. Blood 2001; 97: Pfeiffer C, Fine J, Safdar N. Diagnosis of invasive aspergillosis using a galactomannan assay: a meta-analysis. Clin Infect Dis 2006; 42: White P, Perry M, Loeffler J et al. Critical stages of extracting DNA from Aspergillus fumigatus in whole-blood specimens. J Clin Microbiol 2010; 48: Maertens J, Verhaegen J, Demuynck H et al. Autopsy-controlled prospective evaluation of serial screening for circulating galactomannan by a sandwich enzyme-linked immunosorbent assay for haematological patients at risk for invasive aspergillosis. J Clin Microbiol 1999; 37: Mengoli C, Cruciani M, Barnes R, Loeffler J, Donnelly J. Use of PCR for diagnosis of invasive aspergillosis: systematic review and meta-analysis. Lancet Infect Dis 2009; 9: Marchetti O, Lamoth F, Mikulska M, Viscoli CV, Verweij P, Bretagne S. ECIL recommendations for the use of biological markers for the diagnosis of invasive fungal diseases in leukemic patients and hematopoietic SCT recipients. Bone Marrow Transplant 2012; 47: Karageorgopoulos D, Vouloumanou E, Ntziora F, Micahlopoulos A, Rafailidis P, Falagas M. b-d-glucan assay for the diagnosis of invasive fungal infections: a meta-analysis. Clin Infect Dis 2011; 52: Lamoth F, Cruciani M, Mengoli C et al. b-glucan antigenemia assay for the diagnosis of invasive fungal infections in patients with hematological malignancies: a systematic review and meta-analysis of cohort studies from the third European conference on infections in leukemia (ECIL3). Clin Infect Dis 2012; 54: De Pauw B, Walsh T, Donnelly JP et al. Revised definitions of invasive fungal disease from the European organization for research and treatment of cancer/invasive fungal infections cooperative group and the national institute of allergy and infectious diseases mycoses study group (EORTC/MSG) consensus group. Clin Infect Dis 2008; 46: Marr K, Laverdiere M, Gugel A, Leisenring W. Antifungal therapy decreases sensitivity of the Aspergillus galactomannan enzyme immunoassay. Clin Infect Dis 2005; 40: Portugal R, Garinca M, Nucci M. Index to predict invasive mold infection in high-risk neutropenic patients based on the area over the neutrophilcurve. J Clin Oncol 2009; 27: