Effectiveness and safety of isavuconazole prophylaxis for invasive fungal infections in the haematologic setting

ABSTRACT
 Objectives Patients with haematologic malignancies are at high risk of developing invasive fungal infections (IFIs). Current guidelines recommend the use of azoles for IFI prophylaxis; however, in many clinical situations, antifungal prophylaxis is used off-label. We conducted a systematic literature review to provide haematologists with the available evidence on the effectiveness and safety of isavuconazole in IFI prophylaxis in interventional and real-world, observational studies. Methods Embase, MEDLINE and Cochrane Library databases, and relevant conference proceedings and clinical trial registries, were searched for studies on the effectiveness and safety of isavuconazole prophylaxis in adults at high risk of IFIs. Studies were assessed for inclusion and risk of bias. Results Nine studies were eligible for inclusion in the review, eight of which were in haematologic populations (patients undergoing haematopoietic stem cell transplantation or with acute myeloid leukaemia or myelodysplastic syndromes; n = 5) or included haematologic populations (n = 3). Evidence from these studies suggests isavuconazole is effective for IFI prophylaxis in the haematologic setting. However, the studies frequently lacked safety data, most were based on small patient populations from single centres and risk of bias could not be assessed for five studies. Discussion These findings provide evidence for isavuconazole as an alternative azole for prophylaxis in high-risk populations. Limitations include lack of applicability of risk of bias assessment tools, level of filtering applied in the search strategy and focus on English-language publications. Conclusion Isavuconazole may be an effective azole for IFI prophylaxis in high-risk haematologic populations, although further studies are needed.

Despite guideline recommendations and available trial data, posaconazole and voriconazole are the only azoles approved for prophylaxis in the EU and the US (posaconazole) [9,10] or the EU only (voriconazole) [11], with approvals limited to specific patient populations. Posaconazole prophylaxis is indicated for patients at high risk of IFIs, namely HSCT recipients undergoing high-dose immunosuppressive therapy for GvHD or patients receiving remissioninduction chemotherapy for AML or MDS expected to result in prolonged neutropenia; voriconazole is indicated for IFI prophylaxis in high-risk allogeneic HSCT recipients [10,11]. In the US, posaconazole prophylaxis is indicated for invasive Aspergillus and Candida infections in patients at high risk of these infections due to being severely immunocompromised, such as HSCT recipients with GvHD or those with haematologic malignancies with prolonged neutropenia from chemotherapy [9]. Prophylactic treatments are not approved in other haematologic indications associated with a high risk of IFIs [3][4][5]12]; therefore, antifungal prophylaxis is frequently administered off-label. Furthermore, in situations where the approved azoles cause adverse drugdrug interactions, alternative therapies are lacking. Posaconazole poses a risk of overexposure of midostaurin, which is prescribed in patients with advanced systemic mastocytosis or FMS-like tyrosine kinase-3mutated AML. Isavuconazole may be an alternative option but is not formally recommended [13]. Since isavuconazole is not associated with QTc prolongation, it may provide value in patients using QT-prolonging medications [14]. Given these challenges, collecting data is essential to inform the appropriate use of azoles.
Isavuconazole is an intravenous or oral azole antifungal that was approved in 2015 to treat invasive aspergillosis or mucormycosis in the US and the EU [14,15]. Isavuconazole had not been approved when the current antifungal prophylaxis guidelines were developed, and there was limited clinical evidence for its effectiveness and safety in IFI prophylaxis [3][4][5][6][7].
European Medicines Agency guidance [21] states that 'studies that assess the use of an antifungal agent for prophylaxis of invasive fungal diseases would be conducted only after an agent has demonstrated satisfactory clinical efficacy in the treatment of several types of invasive fungal diseases.' Isavuconazole has demonstrated non-inferiority to voriconazole for treating aspergillosis [19] and to amphotericin B for treating mucormycosis [22].
While isavuconazole is not currently approved for prophylaxis, interventional [23][24][25] and real-world observational studies [20,[26][27][28][29][30][31] support its effectiveness and safety for this purpose. In a large retrospective study of isavuconazole or voriconazole prophylaxis following lung transplantation [28], isavuconazole was comparable with voriconazole in preventing breakthrough (b)IFIs and mortality at 1-year post-lung transplantation. Furthermore, discontinuation rates due to adverse events (AEs), including hepatic toxicity, were significantly lower with isavuconazole [28]. To understand how these findings translate to haematologic settings, we conducted a systematic literature review (SLR) to identify all currently available evidence from interventional and real-world studies on the effectiveness and safety of isavuconazole for IFI prophylaxis.

Materials and methods
To identify evidence on the prophylactic use of isavuconazole and other antifungals for IFIs in high-risk populations, Embase, MEDLINE and Cochrane Library databases (Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews) were searched on 23 September 2020 (Table S1-S3) and relevant conference proceedings (2018-2020), clinical trial registries and bibliographic reference lists of included studies and relevant SLRs/network metaanalyses were also hand-searched.
Eligibility criteria, including patient populations, are displayed in Table 1. Studies were initially identified by including five interventions, then narrowed down to only studies including isavuconazole. Patient populations were determined in each study.
Results from database searches were downloaded in Covidence® (Melbourne, VIC, Australia) and duplicates were removed. Titles and abstracts were screened for relevance by a single reviewer and confirmed by a second reviewer if required. Full papers were evaluated by a single reviewer to confirm relevance and verified by a second reviewer; disputes regarding eligibility were referred to a third reviewer. Data from relevant publications were extracted into standardized dataextraction tables and validated through an independent internal data check. For bIFIs, proven or probable IFIs were extracted.
Risk of bias was assessed using the seven-criteria checklist provided in the National Institute for Health and Care Excellence (NICE) single technology appraisal user guide [32] and the Newcastle-Ottawa Scale [33] for RCTs and observational studies, respectively.
No ethical approval was required as this review article has no original research data.

Results
The initial search identified 2792 records, of which 152 studies were eligible for inclusion ( Figure 1). This was narrowed down to nine studies that included isavuconazole (Table 2). Efficacy results are summarized in Table 3. Key findings by patient population, including safety results, are presented in the following sections.

HSCT
Two studies investigated isavuconazole in patients undergoing HSCT [25,26]. Cupri et al [26] conducted a study, published as a conference abstract, comparing isavuconazole prophylaxis with historical fluconazole prophylaxis. Isavuconazole and fluconazole groups had similar bIFI rates. Mucositis (grade 3-4) was reported in 62% and 61% of patients in the isavuconazole and fluconazole groups, respectively. No other safety outcomes were reported. Stern et al [25] reported a single-arm study in which micafungin was administered followed by isavuconazole. Three of 95 patients (3.2%) developed bIFIs and six patients (6.3%) died during the study, including two deaths due to Candida infection. Seven patients (7.4%) discontinued isavuconazole due to toxicity, most commonly liver function abnormalities (n = 5). No other safety outcomes were reported.

AML/MDS
Three studies in patients with AML/MDS included isavuconazole [20,23,27]. Bose et al [23] reported an open-label, Phase 2 study of isavuconazole prophylaxis in patients with AML/MDS [23]. Four of 65 patients (6.2%) had bIFIs: one proven and three probable. The survival rate was 92%. Three AEs were reported, possibly attributable to isavuconazole: grade 1 transaminitis (n = 2) and elevated bilirubin (n = 1). Two patients discontinued due to deaths (leukaemia progression, n = 1; cardiac arrest, n = 1) and three due to mild-to-moderate elevations of aminotransferases or total bilirubin. No other safety outcomes were reported.
DiPippo et al [20] reported a study of isavuconazole in patients with leukaemia with prior toxicity to posaconazole. Only a minority of patients received isavuconazole prophylaxis, and data were not reported separately for this group.

Mixed study populations
Three studies in mixed adult populations included isavuconazole [29][30][31]. Vu et al [31] reported a study at a single centre of hospitalized patients with autologous and allogeneic HSCT, AML, myeloma, B-cell acute lymphoblastic leukaemia, acute promyelocytic leukaemia, natural killer/T-cell lymphoma with MDS, solid organ transplants, chronic granulomatous disease, systemic lupus erythematous, and no underlying immunocompromised condition. Among patients who received English-language publications (English-language abstracts of foreignlanguage publications considered) Non-English-language publications without English abstract AE, adverse event; ALL, acute lymphocytic leukaemia; AML, acute myeloid leukaemia; CLL, chronic lymphocytic leukaemia; HSCT, haematopoietic stem cell transplantation; IFI, invasive fungal infection; NA, not applicable; PICOS, population, intervention, comparator(s), outcomes and study design.
primary or secondary isavuconazole prophylaxis, there were two bIFIs (7.7%) during primary prophylaxis: one probable bIFI in a patient with AML and one proven bIFI in a patient who had undergone allogeneic HSCT. Both patients died. Mortality rates and safety outcomes were not reported. Bowen et al [29] studied hospitalized patients at a single centre who had received isavuconazole prophylaxis. Underlying conditions included AML, GvHD, MDS, other types of leukaemia and aplastic anaemia. There were six bIFIs (6.1%; three proven and three probable). While discontinuation rates due to AEs were not reported, five patients discontinued due to a suspected drug-induced liver injury and one due to a suspected drug rash. Mortality rates and other safety outcomes were not reported.
Fontana et al [30] reported a study at a single centre of patients with a haematologic malignancy or undergoing HSCT, who received primary isavuconazole prophylaxis. In the main analysis, the rate of bIFIs of invasive pulmonary aspergillosis exceeded expectations based on experience: seven patients (4.8%) died within a median of 12 days from bIFI onset. Subsequently, bIFIs were retrospectively analyzed among patients who received isavuconazole, posaconazole or voriconazole prophylaxis during chemotherapy for AML. In patients with de novo AML, bIFIs occurred during courses of isavuconazole (8%), posaconazole (3%; p = 0.6 vs isavuconazole) and voriconazole (0%; p = 0.04 vs isavuconazole). In patients with relapsed/ refractory AML, bIFIs occurred at similar rates with no significant differences between isavuconazole, posaconazole and voriconazole. Mortality rates and safety outcomes were not reported.

Assessment of bias
Assessment of bias was possible for four studies following PRISMA guidelines [20,27,28,30]; there was a low risk of bias from their inclusion (Table S4). As all studies were observational, the NICE checklist was not used. The assessment tools were not applicable to the other studies, as they were non-comparative or did not provide enough published data.

Discussion
This SLR identified evidence on the effectiveness and safety of isavuconazole for IFI prophylaxis in haematologic populations, including real-world evidence.

HSCT
One isavuconazole prophylaxis study reported bIFI rates; 1-year survival rates and AE rates were similar to those with fluconazole prophylaxis [26]. Furthermore, the single-arm study by Stern et al [25] concluded that the data supported isavuconazole for antifungal prophylaxis after HSCT [25]. A follow-up to this study was published after the SLR was conducted, in which the efficacy and safety of isavuconazole versus voriconazole as the primary antifungal prophylaxis were evaluated in a frequency-matched cohort analysis in allogeneic HSCT recipients (voriconazole, n = 210; isavuconazole, n = 95) [34]. Patients discontinued voriconazole significantly earlier and more frequently than isavuconazole, most commonly because of biochemical hepatotoxicity (voriconazole, n = 48/210 [22.9%] and isavuconazole, n = 5/95 [5.3%]; p = 0.0002). None developed bIFIs in the voriconazole group, while three patients (3.2%) Table 3. bIFI and mortality rates in isavuconazole studies. developed candidaemia in the isavuconazole group at a median of 48 days post-HSCT. There were no significant differences between the voriconazole and isavuconazole groups in overall survival or the number of probable or proven IFIs at 180 days post-HSCT. Primary antifungal prophylaxis with isavuconazole showed a more favourable safety profile and comparable efficacy versus voriconazole in HSCT recipients; however, this was an unblinded, non-randomized, observational study. A 2021 single-centre case report, published after the SLR was conducted, reported on five patients who received isavuconazole prophylaxis following allogenic HSCT [12]. The five patients had malignancies at various remission stages: ALK-positive anaplastic lymphoma, AML, Hodgkin lymphoma, diffuse large B-cell lymphoma and acute lymphoblastic leukaemia [12]. One patient on voriconazole prophylaxis changed to isavuconazole due to hepatic toxicity [12]. Following the change to isavuconazole, toxicities resolved [12]. Another patient remained on isavuconazole prophylaxis at the time the paper was written (December 2020), and no IFIs were reported in any of the patients [12].

AML/MDS
An open-label, Phase 2 study concluded that isavuconazole is effective for antifungal prophylaxis in patients with newly diagnosed AML/MDS undergoing remission-induction chemotherapy [23]. These findings are supported by a retrospective study that explored risk factors for IFIs in patients with AML treated with venetoclax with hypomethylating agents [27]. In this study, the type of antifungal prophylaxis did not influence the development of IFIs; however, group sample sizes were small and safety outcomes were not reported [27]. In a retrospective study that treated patients with prior posaconazole toxicity prophylactically with isavuconazole [20], data were also not provided separately for these patients, limiting interpretation [20].
Finally, a Phase 2 study was not included in the SLR as it was not identified by the search strategy but rather during a hand-search of the reference lists of citations identified by the SLR; this study explored the safety of low-and high-dose isavuconazole prophylaxis in a small group of patients with AML, and concluded that isavuconazole was well tolerated at both doses [24].

Mixed study populations
All three observational studies in mixed populations were retrospective and conducted in a single centre. Vu et al [31] and Bowen et al [29] provided limited dosing information and, in all studies, follow-up duration was unclear. Two of the studies were supportive of isavuconazole prophylaxis [29,31], but one noted a possible increased incidence of bIFIs, particularly invasive pulmonary aspergillosis [30]. All three studies noted that further data are required to assess isavuconazole for antifungal prophylaxis in these populations.
Findings from this SLR in mixed study populations have provided evidence regarding isavuconazole prophylaxis in indications, namely autologous HSCT [31] and haematologic malignancies such as other leukaemias [29]: acute promyelocytic leukaemia, natural killer/T-cell lymphoma and myeloma [31], that currently have no approved prophylactic azole. This is noteworthy as isavuconazole has fewer drug-drug interactions, a weaker cytochrome P450 3A4 inhibitory effect and potentially lower risk of hepatic and renal toxicity than posaconazole and voriconazole [9, 10, 13, 15, 17-20, 35, 36]. A risk-benefit analysis is recommended where intravenous posaconazole or voriconazole are indicated in patients with moderate-to-severe renal impairment; however, no dose adjustment of intravenous isavuconazole is required, as <1% is renally excreted [36].

Strengths and limitations
Our SLR followed a robust design. Multiple databases were searched, and bias was assessed using checklists specific to RCTs and observational studies. Nonetheless, the methodology was subject to limitations. The risk of bias assessment tools were unsuitable for noncomparative studies. Additionally, Cupri et al [26], published as a conference abstract, was not assessed for bias due to a lack of information reported. The search strategy included validated filters for publication type, which meant one potentially relevant study was not identified by the database searches [24].
The focus on English-language publications may have excluded relevant non-English records. Conference proceedings were hand-searched for the period 2018-2020 on the basis that older records would have been published as manuscripts by the time the SLR was conducted, which may have excluded studies. However, no date restriction was applied to conference abstracts indexed in Embase, reducing the likelihood of excluding older records.
Results were not directly compared across studies due to variations in study designs, patient populations, additional interventions, prophylaxis and follow-up durations, and outcomes. Within-study comparisons were possible in some cases, but many studies did not include a comparator arm [26][27][28]30].

Conclusions
Several interventional and real-world observational studies have looked at isavuconazole prophylaxis for IFIs. These include investigations in haematologic settings, where isavuconazole prophylaxis was generally effective, but safety data were often lacking. Although the limited evidence available supports the use of isavuconazole as a prophylactic treatment, further studies are needed, particularly to provide insight into its potential safety value. This review may assist treatment decisions for patients at high risk of IFIs with no approved prophylactic azole or those requiring an alternative option.

Authorship contributions
CB, JAA and MI conceptualized the study design; NW and TM acquired the data; and all authors analyzed the data; substantially revised and critically reviewed the article; reviewed and agreed on all versions of the article before submission, including the final version accepted for publication and any significant changes introduced at the proofing stage; agreed on the journal to which to submit to; agreed to take responsibility and be accountable for the contents of the article and to share responsibility to resolve any questions raised about the accuracy or integrity of the published work.

Availability of data and materials
Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Disclosure statement
YP and JH report no conflicts of interest. CB is an employee of Pfizer Pharma GmbH. MI and JAA are employees of Pfizer Inc. TM and NW are employees of Source Health Economics. OP has previously received honoraria from Pfizer Inc; however, OP did not receive an honorarium for his work on the current manuscript. OP has received honoraria or travel support from Astellas, Gilead, Jazz, MSD, Neovii Biotech, Novartis and Therakos. He has received research support from Gilead, Incyte, Jazz, Neovii Biotech and Takeda. He is a member of advisory boards for Gilead, Jazz, MSD, Omeros, Priothera, Shionogi and SOBI.

Funding
This systematic literature review was funded by Pfizer Inc, New York, NY, USA.

Notes on contributors
Yang Ping is Deputy Chief Physician and Doctor of Medicine in the Hematology Department at Peking University Third Hospital, Beijing, China. She has been engaged in the clinical diagnosis and treatment of haematological diseases for a long time and has experience in the diagnosis and treatment of various benign and malignant diseases of the blood system, particularly haematological malignancies such as lymphoma and multiple myeloma. At the same time, she has rich clinical experience in immunotherapy and haematopoietic stem cell transplantation. She has participated in several national and international clinical studies on lymphoma. Her academic appointments are as follows: Chapter. Mónica's research interests are related to the application of health economics, namely health technology assessment and outcomes research studies.
Tom Macmillan joined Source Health Economics in 2020 having previously worked as an Information Specialist and Health Technology Assessor at King's College London, London, UK. He has worked in the health economics and outcomes research field since 2013 and has experience delivering a wide variety of health technology assessment (HTA) projects, including several years working on the NICE medical technologies evaluation programme. He has previously worked on systematic reviews for the Royal College of Surgeons at the National Surgical Commissioning Centre and current awareness bulletins for surgical specialities. His disease experience comprises vascular access, diabetes, oncology, neurodegenerative disease, wound healing and surgery and he has worked extensively on HTAs for medical devices. He completed an MSc in Information Management in 2014 and holds a BA in English Literature and Linguistics.