Voriconazole for prophylaxis of invasive fungal infections after allogeneic hematopoietic stem cell transplantation

ABSTRACT

Introduction: Invasive fungal infections (IFIs) following allogeneic hematopoietic stem cell transplantation (alloHSCT) are associated with a high mortality, and accordingly most alloHSCT recipients receive prophylaxis with antifungal agents. Despite some improvement in outcomes of IFIs over time, they continue to represent substantial clinical risk, mortality, and financial burden.

Areas covered: We review the main pathogens responsible for IFIs in recipients of alloHSCT, current treatment recommendations, and discuss clinical and economic considerations associated with voriconazole prophylaxis of IFIs in these patients.

Expert commentary: The clinical efficacy of voriconazole appears to be at least equivalent to other antifungal treatments, and generally well tolerated. Overall, benefit-risk balance is favorable, and findings from cost-effectiveness analyses support the use of voriconazole prophylaxis of IFIs in recipients of alloHSCT.

KEYWORDS:

• Allogeneic hematopoietic stem cell transplantation
• invasive aspergillosis
• invasive candidiasis
• prophylaxis
• triazoles
• voriconazole

1. Introduction

Invasive fungal infections (IFIs) following allogeneic hematopoietic stem cell transplantation (alloHSCT) remain a leading cause of morbidity and mortality in alloHSCT recipients [1,2]. Based on the findings from several studies, the estimated incidence of proven/probable (according to revised European Organization for Research and Treatment of Cancer [EORTC] and the Mycoses Study Group [MSG] definitions) [3] IFIs following alloHSCT ranges from 8% to 12% in the first year post-alloHSCT [4,5].

Pathogens causing IFIs can be broadly categorized into three groups: Candida species; Aspergillus species; and other molds (including zygomycetes, Fusarium species, and Scedosporium species). Over time, Aspergillus species have replaced Candida species as the most common fungal pathogen in recipients of alloHSCT [6]. The estimated incidence of EORTC/MSG proven/probable invasive aspergillosis (IA) is approximately 6–7% in the year following alloHSCT, while invasive candidiasis (IC) accounts for approximately 1–5% of cases [2,4,5]. IA accounts for approximately 60% of IFIs in patients who have received alloHSCT, and IC for approximately 25% [1]. More than 80% of IFIs in recipients of alloHSCT occur within 100 days of transplantation, with approximately 57% of observed cases occurring in the first 40 days (early phase) and 24% occurring between 40 and 100 days (late phase) after alloHSCT [4]; it must, however, be considered that data reporting late-phase infections are inherently less reliable, as patients have typically left the transplant clinic and are no longer subject to the same monitoring and observation.

The emergence of reduced-intensity (non-myeloablative) conditioning (RIC) regimens as an alternative to conventional (myeloablative) regimens prior to alloHSCT has improved clinical outcomes in patients who may be precluded from conventional myeloablative approaches due to tolerability issues. Unlike conventional conditioning which uses high-intensity chemotherapy and radiotherapy regimens to minimize disease and eradicate the bone marrow (thereby suppressing immune response to the transplant), RIC applies a lower dose of chemotherapy and/or radiotherapy and consequently is associated with less toxicity [7], and relies on the donor immune response to eliminate residual disease [8]. A pan-European survey suggested that RIC use has increased dramatically from 2000 to 2011, and the percentage of all alloHSCT recipients who had received RIC exceeded 50% in several countries, including France and the UK [9]. Comparison between patients receiving conventional conditioning and RIC is difficult as patients receiving RIC alloHSCT tend to be older and have more comorbidities, but evidence suggests that transplant-related mortality in these patients is improved with RIC [10–13]. Despite the favorable neutropenia and toxicity outcomes observed with RIC compared with conventional conditioning, the post-alloHSCT incidence of IFIs has not been demonstrated to be reduced using RIC [14,15]. It appears that RIC may be most suitable for use in patients with a pretransplant history of IFI for whom additional toxicity and neutropenia is a high-risk factor [16].

Multiple factors have been observed to increase the risk of IFI during neutropenia in alloHSCT recipients. Individual risk factors include: older age; use of broad-spectrum antibiotics (and the effect on gut microbiota); iron overload; use of central venous catheter; diabetes; prior IFI (or species colonization); and lower respiratory tract infection [17,18]. Additionally, patients receiving transplantation from a haploidentical donor have been shown to be at an elevated risk of IFI compared with HLA-matched siblings [19]. Furthermore, the condition of the patient at the time of infection (as measured by Acute Physiology and Chronic Health Evaluation II scores) [20] or at the time of transplant (as measured by the hematopoietic cell transplantation-specific comorbidity index) [21] are considered to be important predictors of patient outcome.

The Italian Group for Bone Marrow Transplantation (GITMO) recommendations stratify patients by the level of post-transplant risk (high, intermediate, low) and the post-transplant phase the patient is in (early: 0–40 days; late: 41–100 days; very late: >100 days) [22]. Risk factors considered to indicate that a patient is at high risk of IFI during the early and late phases include: grade III–IV acute graft-versus-host disease (GVHD); transplant from unrelated/mismatched donor in combination with grade II acute GVHD; >1 week of steroid use; cytomegalovirus disease/recurrent infection; prolonged neutropenia; iron overload (engraftment phase only); or steroid-refractory/-dependent acute GVHD. Additional risk factors during the early phase include active acute leukemia at the time of transplantation or cord blood transplantation. In both the early and late phases, all remaining patients not fulfilling the criteria outlined for high risk must be considered to be at least at intermediate risk of IFI (no patient is considered at low risk of IFI during the first 100 days after transplant). Risk factors during the very late phase after transplantation include persistent or late-onset grade III acute GVHD; steroid-refractory/-dependent acute GVHD; or grade II acute GVHD after transplant from a mismatched/unrelated donor. Extensive chronic GVHD, when preceded by acute GVHD, is also considered a high-risk indicator. Patients in the very late phase who experience de novo chronic GVHD without steroid-based immunosuppression are considered at intermediate risk; absence of GVHD and steroid therapy suggests that the patient is at low risk, and antifungal prophylaxis may not be required.

The broad-spectrum polyene amphotericin B deoxycholate (AmB) was used in the 1990s as prophylaxis against IFIs [23], but effectiveness was limited and infusion-related effects and nephrotoxicity were common [24]. AmB was soon superseded by oral triazoles such as fluconazole, which was shown to have appropriate bioavailability [25] and resulted in significantly improved protection (against candidiasis) and survival compared with placebo when given to patients following alloHSCT [26]. Fluconazole, however, does not exhibit efficacy against opportunistic mold infections, necessitating development of broader spectrum alternatives. Itraconazole was subsequently shown to provide improved protection against aspergillosis [27] and IFIs [28] over 180 days after alloHSCT compared with fluconazole; however, tolerability issues are present, and absorption and bioavailability are unpredictable [27,28]. Newer-generation triazoles provide improved prophylactic options for IFIs in immunocompromised patients [18], with voriconazole demonstrating similar efficacy to fluconazole [29] and itraconazole [30] with improved tolerability compared with the latter; and with posaconazole therapy, which was non-inferior to fluconazole for prophylaxis against IFIs and IA events in patients with GVHD [31]. Use of prophylactic agents has resulted in improved patient outcomes over time [32,33]. Furthermore, early use of prophylactic triazoles has been shown to reduce IFI-related morbidity and mortality in recipients of alloHSCT [27–31]. Despite improvements in prophylaxis and treatment, IFI-related morbidity and mortality continues to present a substantial clinical and economic burden [34,35]. Of note, the recent development of isavuconazole, a new second-generation triazole with a broad spectrum of activity, offers protection against Aspergillus species and has been reported to be non-inferior to voriconazole for the treatment of adults with IA [36]; however, isavuconazole has not yet demonstrated prophylactic efficacy for the prevention of IFIs.

The objective of this review is to provide an overview of recommended prophylactic strategies for IFIs in recipients of alloHSCT, and to assess the clinical and economic aspects of voriconazole as prophylaxis against IFIs in this setting.

2. Recommendations for antifungal prophylaxis of IFIs in recipients of alloHSCT

Several major international bodies have published recommendations for prophylaxis of IFIs in recipients of alloHSCT; a summary of selected guidelines and recommendations are presented in Table 1 [18,22,37–44]. In general, fluconazole, itraconazole, and voriconazole are recommended for prophylaxis of IFI in recipients of alloHSCT who have no additional risk factors, with posaconazole recommended in patients with GVHD (although the level of supporting evidence varies between agents). In patients at high risk of IFI (please see previous description of IFI risk factors), primary prophylaxis with mold-active triazoles such as itraconazole, posaconazole, or voriconazole is typically recommended [18,22,37,38,45]. However, the local epidemiology of yeast/mold infections should also be taken into account when selecting appropriate prophylaxis [37]. Antifungal prophylaxis should be continued for at least 75 days post-alloHSCT [18,22] (at least 3 months in Chinese guidelines [38]). If gastrointestinal issues are a concern, then voriconazole may be preferred to posaconazole [22]. For patients experiencing GVHD requiring corticosteroid treatment, posaconazole is the recommended prophylactic agent [37,44]. Secondary prophylaxis is recommended to prevent recurrence or onset of an IFI during a prolonged neutropenic period in patients with a prior history of IFI [40].

Table 1. Summary of clinical guidelines and recommendations for prophylactic treatment of IFIs in alloHSCT recipients.

European Conference on Infections in Leukaemia (ECIL-5) [37] I: evidence from ≥1 properly randomized, controlled trial; II: evidence from ≥1 well-designed clinical trial, without randomization, from cohort or case-controlled analytic studies (preferably from >1 center), from multiple time-series studies, or from dramatic results from uncontrolled experiments; III: evidence from opinions of respected authorities, based on clinical experience, descriptive studies, or reports of expert committees. A: good evidence to support a recommendation for or against use; B: moderate evidence to support a recommendation for or against use; C: poor evidence to support a recommendation.
Pre-engraftment, low risk for molds
Fluconazole	A-I
Itraconazole; micafungin; voriconazole		B-I
Posaconazole		B-II
Aerosolized AmB; liposomal AmB			C-II
Pre-engraftment, high risk for molds
Fluconazole	A-III against
Itraconazole; voriconazole		B-I
Aerosolized AmB; posaconazole		B-II
Micafungin			C-I
Liposomal AmB			C-II
Post-engraftment, GVHD
Fluconazole	A-III against
Itraconazole; voriconazole		B-I
Posaconazole	A-I
Liposomal AmB; micafungin			C-II
Royal Australian College of Physicians (RACP) [18] A: evidence can be trusted to guide practice; B: evidence can be trusted to guide practice in most situations; C: evidence provides some support for recommendations but care should be taken in its application; D: evidence is weak and recommendation must be applied with caution (note: none of the listed specific antifungal prophylaxis options were given a D-grade recommendation)
High risk of IFI
Posaconazole	A
Itraconazole; micafungin; voriconazole		B
Caspofungin; liposomal AmB			C
Low risk of IFI
Fluconazole		B
Echinocandins; itraconazole			C
No prophylaxis is recommended for patients at very low risk of IFI
Chinese invasive fungal infection workgroup [38]
Caspofungin; fluconazole; itraconazole; micafungin; posaconazole	Recommended
Choice of secondary prophylaxis should be based on response to primary prophylaxis	Recommended
Italian Group for Bone Marrow Transplantation (GITMO) [22] Strength of recommendation – A: strongly supports; B: moderately supports; C: marginally supports Quality of evidence – I: evidence from ≥1 properly designed randomized controlled trial; II: evidence from ≥1 well-designed nonrandomized trial, cohort/case-controlled study (from ≥1 center), from multiple time series, or from dramatic results of uncontrolled experiments; III: evidence from opinions of respected authorities, based on clinical experience, descriptive case studies, or reports of expert committees
High risk of IFI
Posaconazole (in patients with GVHD)	A-I
Voriconazole		B-I
Aerosolized AmB plus fluconazole; caspofungin; liposomal AmB; micafungin			C-III
Standard risk of IFI
Fluconazole	A-I
Itraconazole; micafungin; voriconazole		B-I
No prophylaxis is recommended for patients at low risk of IFI
Infectious Diseases Society of America (IDSA) [39,40] Strength of recommendation – weak or strong Quality of evidence – very low, low, moderate, or high
High-risk patients in intensive care units with a high rate (>5%) of invasive candidiasis
Fluconazole	Weak recommendation; moderate-quality evidence
Anidulafungin; caspofungin; micafungin	Weak recommendation; low-quality evidence
Patients at high risk of invasive aspergillosis
Posaconazole	Strong recommendation; high-quality evidence
Voriconazole	Strong recommendation; moderate-quality evidence
Micafungin	Weak recommendation; low-quality evidence
Caspofungin	Weak recommendation; low-quality evidence
Itraconazole – limited by tolerability and absorption issues	Strong recommendation; moderate-quality evidence
Patients with GVHD at high risk of invasive aspergillosis
Posaconazole	Strong recommendation; high-quality evidence
Voriconazole has not improved survival in clinical trials	Strong recommendation; moderate-quality evidence
Itraconazole – limited by tolerability and absorption issues	Strong recommendation; high-quality evidence
European Society of Clinical Microbiology and Infectious Diseases (ESCMID) [41] Strength of recommendation – A: strongly supports; B: moderately supports; C: marginally supports; D: opposes (note: none of the listed specific antifungal prophylaxis options were given a D-grade recommendation) Quality of evidence – I: evidence from ≥1 properly designed randomized controlled trial; II: evidence from ≥1 well-designed non-randomized trial, cohort/case-controlled study (from ≥1 center), from multiple time series, or from dramatic results of uncontrolled experiments; III: evidence from opinions of respected authorities, based on clinical experience, descriptive case studies, or reports of expert committees
Prophylaxis against Candida in the neutropenic phase after alloHSCT with the intention of improving survival
Fluconazole	A-I
Posaconazole		B-II
Itraconazole; micafungin; voriconazole			C-I
Caspofungin; liposomal AmB			C-III
As prophylaxis against candidemia in the first 100 days after alloHSCT without GVHD with the intention of improving survival
Fluconazole	A-I
Itraconazole; voriconazole			C-I
Caspofungin			C-II
Liposomal AmB; micafungin; posaconazole			C-III
As prophylaxis against candidemia during GVHD with the intention of improving survival
Posaconazole		B-I
Fluconazole; itraconazole; voriconazole			C-I
Infectious Diseases Working Party of the German Society for Haematology and Oncology [42,44] Strength of recommendation – A: strongly supports; B: moderately supports; C: marginally supports; D: opposes Quality of evidence – I: evidence from ≥1 properly designed randomized controlled trial; II: evidence from ≥1 well-designed non-randomized trial, cohort/case-controlled study (from ≥1 center), from multiple time series, or from dramatic results of uncontrolled experiments; III: evidence from opinions of respected authorities, based on clinical experience, descriptive case studies, or reports of expert committees
Prophylaxis against mold infections in patients without GVHD in first 100 days after alloHSCT
Posaconazole		B-II
Itraconazole; micafungin; voriconazole			C-I
Prophylaxis against candidemia in patients without GVHD in first 100 days after alloHSCT
Fluconazole	A-I
Micafungin; posaconazole; voriconazole		B-II
Itraconazole			C-I
Prophylaxis against aspergillosis during GVHD
Posaconazole	A-I
Secondary prophylaxis against IFIs
Voriconazole		B-II
Caspofungin; posaconazole		B-III
AmB is not recommended for prophylaxis against IFIs due to its unacceptable toxicity profile (recommendation grade: DII)
Korean National Evidence-based Healthcare Collaborating Agency (NECA) [43] Strength of recommendation – A: good evidence; B: moderate evidence; C: poor evidence Quality of evidence – I: evidence from ≥1 properly designed randomized controlled trial; II: evidence from ≥1 well-designed nonrandomized trial, cohort/case-controlled study (from ≥1 center), from multiple time series, or from dramatic results of uncontrolled experiments; III: evidence from opinions of respected authorities, based on clinical experience, descriptive case studies, or reports of expert committees
Prophylaxis in alloHSCT recipients
Posaconazole; fluconazole	A-I
Itraconazole; micafungin	B-I

alloHSCT: allogeneic hematopoietic stem cell transplantation; AmB: amphotericin B; GVHD: graft-versus-host disease; IFI: invasive fungal infection.

3. Clinical impact of voriconazole prophylaxis

Table 2 presents a summary of key randomized trials comparing triazoles for prophylaxis of IFIs in patients receiving alloHSCT. A systematic review and meta-analysis of antifungal prophylaxis against IFI in recipients of alloHSCT reported that systemic antifungal treatments improved mortality compared with placebo, no treatment, or nonsystemic antifungals (relative risk [95% confidence interval (CI)]: 0.62 [0.45–0.85]) [46]. However, at the time of this analysis, data on the effectiveness of newer-generation triazoles such as posaconazole and voriconazole were unavailable. More recent clinical trials have assessed prophylaxis with these newer generation agents [29–31]. A double-blind, randomized trial [29] compared prophylactic voriconazole (n = 305) with fluconazole (n = 295) given for 100 days post-alloHSCT, and found no significant difference in the primary endpoint of fungal-free survival at 180 days (hazard ratio [95% CI]: 1.07 [0.82–1.40]); however, in a subset of patients receiving alloHSCT for therapy of acute myeloid leukemia, fewer IFIs were observed with voriconazole prophylaxis than with fluconazole (hazard ratio [95% CI]: 2.03 [1.29–3.19]). An open-label, randomized trial compared voriconazole (n = 234) and itraconazole (n = 255) in 489 recipients of alloHSCT following treatment for 100–180 days [30]. The primary objective was a composite endpoint incorporating survival without proven/probable IFI at 180 days post-alloHSCT transplant and ability to tolerate therapy for at least 100 days with fewer than 14 days’ interruption. The voriconazole arm met the criteria for superiority in the primary endpoint when compared with the itraconazole arm (48.7% vs. 33.2%, respectively; p < 0.01). In addition, the median duration of prophylaxis was longer with voriconazole (96 days) than with itraconazole (68 days); there were no differences between groups in terms of survival or incidence of IFI. In both of these trials, the study populations represented patients with standard risk and not high risk, based on the number of IFIs detected; in the Marks trial, patients were excluded if they had prior IFIs within 6 months, and in the Wingard trial patients were excluded if they had prior mold infections within 4 months of transplant. Additionally, approximately 90% of the population in the latter study were deemed ‘standard risk.’ These trials may therefore underestimate the effect of prophylaxis in a higher-risk population.

Table 2. Key comparative trials of primary triazole prophylaxis of IFIs in patients undergoing alloHSCT.

Survival	IFIs	IA	Other end point of note	Treatment-related AEs	Duration of prophylaxis
Marks 2011 (alloHSCT recipients) [30] Voriconazole (n = 224) vs. itraconazole (n = 241)
100 d: 91.9% vs. 92.3% 180 d: 81.9% vs. 80.9% 1 y: 73.5% vs. 67.0%	180 d: 1.3% vs. 2.1%	180 d: 0.4% vs. 2.1% (NSD)	Percentage of patients who tolerated agent for ≥100 d and had IFI-free survival 100 d: 54.0% vs. 39.8% (p < 0.001) 180 d: 48.7% vs. 33.2% (p < 0.01)	180 d: All serious AEs – 47.8% vs. 37.3% (p < 0.05) All (adjusted for treatment duration) – 1.7 per 30 d vs. 2.0 per 30 d (NSD) Gastrointestinal – 14.7% vs. 42.7% (p < 0.01) Hepatotoxicities – 12.9% vs. 5.0% (p < 0.01) Headache – 4.5% vs. 5.0% (NSD) Visual impairment – 5.4% vs. 0% (p < 0.01)	Median: 96 d vs. 68 d (proportion of patients completing >100 d of prophylaxis: 53.6% vs. 39.0%; p < 0.01)
Marr 2004 (alloHSCT recipients) [27] Itraconazole (n = 151) vs. fluconazole (n = 148)
250 d: 61% vs. 69% (NSD)	180 d: 17.9% vs. 18.9% (p < 0.05)	180 d: 5.4% vs. 11.3%	Discontinuation: 36% vs. 16% (p < 0.01)	180 d: Renal toxicities – 7% vs. 3% (NSD) Tripling of baseline total bilirubin – 95% vs. 86% (NSD)	Median: 89 d vs. 120 d
Ullman 2007 (patients with GVHD) [31] Posaconazole (n = 301) vs. fluconazole (n = 299)
112 d: 74.8% vs. 71.9% (NSD) 168 d: 80.7% vs. 80.3% (NSD)	112 d: 5.3% vs. 9.0% (p = 0.07) Exposure period (duration of prophylaxis plus 7 d): 2.4% vs. 7.6% (p < 0.01)	112 d: 2.3% vs. 7.0% (p < 0.01) Exposure period (duration of prophylaxis plus 7 d): 1.0% vs. 5.9% (p < 0.01)	Discontinuation: 34.2% vs. 38.1% (NSD)	168 d: All – 35.5% vs. 38.4% (NSD)	Mean: 80 d vs. 77 d Median: 111 d vs. 108 d
Wingard 2010 (alloHSCT recipients) [29] Voriconazole (n = 305) vs. fluconazole (n = 295)
180 d: 81.2% vs. 80.0% (NSD)	180 d: 7.3% vs. 11.2% (NSD) 365 d: 12.7% vs. 13.7% (NSD)	180 d: 3.0% vs. 5.8% (NSD) 365 d: 5.6% vs. 7.1% (NSD)	Fungal-free survival at 180 d: 78.2% vs. 74.9% (NSD) 365 d: 64% vs. 65% (NSD)	NSD in any CTCAE v3 grade 3–5 toxicity	100 d per protocol; 40.7% vs. 44.4% discontinued prematurely (median time to discontinuation was 29 d in both groups)
Winston 2003 (alloHSCT recipients) [28] Itraconazole (n = 71) vs. fluconazole (n = 67)
180 d: 54.9% vs. 58.2% (NSD)	180 d: 8.5% vs. 25.4% (p < 0.05)	180 d: 4.2% vs. 11.9% (p = 0.12)	IFI-related mortality at 180 d: 8.5% vs. 17.9% (NSD)	180 d: All – 46.5% vs. 20.9% (p < 0.001) Gastrointestinal – 23.9% vs. 9.0% (p < 0.05) Hepatotoxicities – 15.5% vs. 10.4% (NSD)	Mean: 78 d vs. 85 d

AE: adverse event; alloHSCT: allogeneic hematopoietic stem cell transplantation; CTCAE: Common Terminology Criteria for Adverse Events; GVHD: graft-versus-host disease; IA: invasive aspergillosis; IFI: invasive fungal infection; NSD: no significant difference.

Owing to a lack of randomized controlled trials directly comparing more than two oral triazoles, a mixed-treatment comparison was conducted to indirectly compare voriconazole, posaconazole, itraconazole, and fluconazole as prophylaxis for IFIs in recipients of alloHSCT [47]. Mixed-treatment comparisons allow for comparison of interventions that are not directly assessed in the same trial, by assessing them relative to comparator interventions that are common between trials. In this case, the studies compared posaconazole with fluconazole [31], fluconazole with voriconazole [29], itraconazole with fluconazole [27,28], and voriconazole with itraconazole [30]. The authors concluded that voriconazole was the agent most likely to reduce the incidence of IFI within 180 days post-alloHSCT, with a treatment-effect odds ratio (95% CI) of 0.46 (0.28–0.73) relative to the base case of fluconazole (itraconazole: 0.52 [0.35–0.76]; posaconazole: 0.56 [0.28–0.73]).

In neutropenic patients receiving alloHSCT the relapse rate of previous IFIs is high; therefore, the use of secondary prophylaxis is recommended [18,22,37–39,41,48]. A single-arm, multicenter trial in a cohort of 45 recipients of alloHSCT with a previous history of IFI (of whom 69% had prior probable/proven IA) evaluated secondary prophylaxis with voriconazole [49] and found that only 6.7% (3/45) of patients developed an IFI during the first year after transplant. Additionally, a prospective multicenter trial assessed efficacy and safety of secondary antifungal prophylaxis with voriconazole, itraconazole, caspofungin, or liposomal AmB in patients with a history of IA in alloHSCT, and found that the 1-year cumulative incidence of IFI was 8.8% (12/136), with no discontinuation due to adverse events (AEs) [50]. No differences in the incidence of breakthrough IFIs were observed between these agents or between patients with active or stable IA at the time of transplant.

3.1. Adverse events

No significant difference in the incidence of AEs was observed with voriconazole or fluconazole prophylaxis in a randomized, double-blind clinical trial; tolerability was similar between the two arms (no Common Terminology Criteria for Adverse Events [CTCAE] Grade 3–5 AEs varied by more than 5% between arms) [29]. In another trial, treatment-related gastrointestinal AEs such as vomiting (3.6% vs. 16.6%; p < 0.01), nausea (7.1% vs. 15.8%; p < 0.01), and diarrhea (4.0% vs. 10.4%; p < 0.01) were observed to be significantly lower with voriconazole than with itraconazole, respectively. Conversely, treatment-related AEs such as hepatotoxicity/liver function (12.9% vs. 5.0%; p < 0.01) and visual impairment (5.4% vs. 0.0%; p < 0.01), and serious all-causality AEs (47.8% vs. 37.3%; p = 0.02), were higher with voriconazole than with itraconazole [30]. Hepatotoxicity-related discontinuation was higher with voriconazole (17.0%) than with itraconazole (4.1%), although no formal statistical analysis was conducted [30]. However, overall treatment discontinuation was significantly lower with voriconazole than itraconazole (46.4% vs. 61.0%, respectively; p < 0.01). When the AE incidence was corrected for duration of observation, the incidence per 30 days of treatment was similar (1.7 vs. 2.0 AEs per 30 days; p = 0.53) [30]. A retrospective analysis of this trial assessing drug tolerability and the impact on resource utilization concluded that the longer duration of tolerated prophylaxis observed with voriconazole was associated with fewer days spent in hospital and with decreased use of other antifungal therapies, with obvious economic benefits [51].

As voriconazole is inactive against mucormycetes [52], it has been suggested that the incidence of mucormycosis may be increased with voriconazole prophylaxis compared with other agents. However, this has not been the case. No mucormycosis infections were reported among 224 patients receiving voriconazole within 180 days after alloHSCT [30]; however, interpretation is limited by the relatively small sample size and short follow-up used in this study. Additionally, no differences in the rate of mucormycosis infections were observed between patients receiving fluconazole (4 patients) or voriconazole (5 patients) in the 365 days after alloHSCT [29]. In a recent retrospective review of risks for mucormycosis, type of prophylaxis was not an independent risk [53].

Voriconazole use has been associated with skin photosensitivity, which manifests as prominent sunburn-like erythema on sun-exposed surfaces, particularly the head, neck, hands, and forearms [54]. Most reports have been in children, with incidence reported to be as high as 20% [55], and 10.5% in patients aged over 12 years [56]. However, neither study was specific to patients who had received alloHSCT. Voriconazole-induced skin photosensitivity has also been associated with development of subsequent squamous cell carcinoma (SCC) [57] in adults and in pediatric patients; although most evidence for this has come from adult patients receiving solid organ transplants [58,59], it seems prudent to monitor the potential association between voriconazole prophylaxis in alloHSCT recipients and SCC.

Owing to the nonlinear pharmacokinetics of voriconazole, dosage adjustment may be desirable, particularly in children where some studies have suggested that oral bioavailability may be lower and plasma concentrations less predictable [60,61]. Lack of control of plasma voriconazole concentrations has been shown to be associated with negative outcomes; in a retrospective assessment of plasma trough voriconazole concentrations, low plasma voriconazole was associated with negative clinical outcomes [62]. Efficacy has been observed to be near optimal at plasma trough voriconazole concentrations 1–2 mg/L, and a higher incidence of AEs is associated with trough concentrations >5 mg/L [62–66]. Accordingly, a clinical trial reported that patients receiving therapeutic drug monitoring (TDM) with the aim of regulating plasma concentrations at 1.0–5.5 mg/L) had better clinical outcomes and less discontinuation due to AEs than patients who did not receive TDM [63]. In light of these findings, guidelines suggest that TDM with the aim of regulating trough voriconazole concentrations in a therapeutic range of 1–6 mg/L (with an upper limit of 4 mg/L in Japanese patients owing to observations of increased hepatotoxicity above this threshold [67]) should be performed within 5 days of initiating therapy and regularly thereafter in the majority of patients receiving voriconazole [68,69]. Although some authors have previously suggested that sufficient evidence to warrant routine TDM is lacking [70,71], TDM appears to be a prudent approach when using voriconazole as prophylaxis against IFIs, particularly in pediatric patients.

Voriconazole is primarily metabolized by the hepatic cytochrome P450 enzyme system, which is subject to wide intra- and interindividual variability [72], and genetic polymorphisms in the CYP2C19 gene account for up to 50% of the variability seen [73]. In individuals with genotypically identified polymorphisms predictive of poor metabolism of voriconazole, the area under the concentration–time curve (AUC) is nearly five-times higher than in extensive metabolizers [74]. Voriconazole also expresses inhibitory function on the CYP2C19, CYP2C9, and CYP3A4 enzyme pathways, although polymorphisms in CYP2C9 are not thought to contribute greatly to dysfunction in voriconazole metabolism, as they are responsible for only a relatively small proportion of total voriconazole metabolism [75]. As several drugs share enzymatic pathways with voriconazole, concomitant medications can often be problematic, and concomitant sirolimus [76] or St John’s Wort [77] are not recommended during therapy with voriconazole. Also contraindicated are the antibiotics rifampin and rifabutin, the antiretrovirals efavirenz and ritonavir, the anticonvulsants carbamazepine and phenytoin, and long-acting barbiturates [78].

High plasma voriconazole concentrations have been hypothesized to negatively affect liver function, and have been shown to be associated with discontinuation in a trial assessing voriconazole treatment of confirmed IA [79]. However, a retrospective cohort study evaluating hepatotoxicities in high- versus labeled-dosing strategies did not report any dose-dependent association of liver function markers [80]. When administered intravenously, voriconazole is formulated in solution with the excipient sulfobutyl-ether cyclodextrin [81], which places high demand on the renal clearance system. Accordingly, oral voriconazole is preferred in patients with renal dysfunction (creatinine clearance rates of <50 mL/min) [48,81].

4. Economic impact of voriconazole prophylaxis

Beyond the reported clinical considerations, prevention of IFIs in recipients of alloHSCT has substantial economic implications. A retrospective analysis of prescription records from a single hospital in France reported that the cost of treating 192 patients who had received antifungal treatment over the course of a year was €1,731,300, with the daily cost of treatment for proven/probable cases of IFI estimated at €459 per patient [34]. An analysis of voriconazole and fluconazole prophylaxis in recipients of alloHSCT in the USA demonstrated that voriconazole was more cost-effective in a subset of patients with underlying acute myeloid leukemia (at a cost per life-year gained threshold of $50,000), most likely because of an inherently higher risk of IA in these patients [82]. More recently, a cost-effectiveness analysis comparing fluconazole, itraconazole, posaconazole, and voriconazole in recipients of alloHSCT in the Spanish healthcare system concluded that savings with voriconazole were likely to be in the region of €4707 per patient compared with oral posaconazole, with fewer deaths and IFIs with voriconazole [83]. Although the same study reported that prophylaxis with voriconazole results in higher total costs than with itraconazole, it was also associated with substantially fewer IFIs, resulting in an incremental cost of €212,101 per IFI avoided with voriconazole compared with itraconazole [83]. Both of these economic evaluation analyses reported that costs relating to empirical therapy of breakthrough IFIs were reduced with voriconazole/mold-active prophylaxis. Additionally, a retrospective observational cohort study assessing patients with chemotherapy-induced neutropenia in a US hemato-oncology unit demonstrated that, although prophylaxis was more expensive with voriconazole than AmB, the former was preferable due to an associated reduction in incidence of breakthrough IFIs [84]. It is important to note that differences between countries in terms of clinical practice, costs of treatment, and use of resources mean that conclusions drawn from one country are not readily applicable to others because of differing costs of inpatient stays, stays in intensive therapy units, drug costs, and other associated costs. Thus, the findings from cost-effectiveness analyses should be interpreted with caution.

5. Conclusions

It is widely accepted that recipients of alloHSCT should receive antifungal prophylaxis and that anti-mold activity should be a consideration. However, the optimal agent of choice remains unclear. Voriconazole is a broad-spectrum triazole with proven prophylactic activity against Candida and Aspergillus species in patients at risk of IFI after alloHSCT. The demonstrated clinical efficacy of voriconazole appears to be at least equivalent to other antifungal treatments, and is generally well tolerated. Overall, benefit–risk balance is favorable and findings from cost-effectiveness analyses provide further evidence supporting the use of voriconazole prophylaxis of IFIs in recipients of alloHSCT.

6. Expert commentary

Current evidence-based clinical practice guidelines recommend that antifungal prophylaxis should be used in patients receiving alloHSCT who are at risk of neutropenia and those with either chronic extensive or grade 3/4 acute GVHD; in circumstances where mold infections are a risk, oral triazoles are commonly used. Although transplant practices have changed over time and the risk of neutropenia has been lessened with improvements in pretransplant conditioning such as reduced intensity conditioning, overall risk of IFI has not decreased and remains a substantial clinical and economic risk. In terms of the duration of prophylaxis, there is consensus that duration should cover the neutropenic period before engraftment. However, less is known about the most appropriate duration of prophylaxis in GVHD, and the optimal agents in primary and secondary prophylaxis remain unclear. Mold-active, newer-generation triazoles can be used in patients at high risk such as: those receiving a transplant from an alternative donor; those experiencing GVHD requiring immunosuppressive therapy; those for whom prolonged neutropenia is anticipated; or those who have previously experienced IFIs. As risk factors are increasingly elucidated, prophylaxis can be better personalized to the requirements of patients, and prophylactic agents can be reserved for those at the highest risk of IFI following alloHSCT.

7. Five-year view

The prophylactic efficacy of newer-generation oral triazoles such as voriconazole and posaconazole has been confirmed in several randomized trials in patients receiving alloHSCT. Future research should aim for early diagnosis of IFIs, and assess comparative tolerability and toxicity in triazole agents. Increased understanding of risk factors and the precise nature of the immune response to IFIs and transplant will permit better patient selection and improve patient outcomes. More recent work assessing genetic polymorphisms and their association with risk of IFI may yield further improvements in patient selection. As diagnostic techniques and procedures improve, lower-risk patients can be identified and hopefully treated pre-emptively when possible, reducing reliance on prophylaxis.

Key issues

• Patients receiving alloHSCT may experience prolonged severe neutropenia in the post-transplant phase or require post-engraftment immunosuppression to treat GVHD.

• In alloHSCT, antifungal prophylaxis is commonly used during neutropenia or GVHD, in order to mitigate the clinical and economic burden of IFI.

• Clinical guidelines recommend antifungal prophylaxis with either fluconazole, itraconazole, posaconazole (when GVHD is present), or voriconazole after alloHSCT; the level of supporting evidence varies between agents.

• Voriconazole is a broad-spectrum triazole with proven activity against Candida and Aspergillus species in patients at risk of IFI after alloHSCT, with at least comparable efficacy to other antifungal agents and manageable toxicity.

• After alloHSCT, prophylaxis with mold-active triazoles such as voriconazole and posaconazole is recommended when the risk of IA is high.

• The clinical efficacy of voriconazole has been reported by two large randomized clinical trials to be comparable to fluconazole, and better than itraconazole, with a similar safety profile; of note, the prophylactic efficacy advantage of voriconazole over fluconazole was comparable with that seen in a comparison of posaconazole with fluconazole.

• Economic analyses have reported cost-effectiveness of voriconazole prophylaxis in Spain and USA; voriconazole is particularly likely to provide economic benefit in regions with high risk of IA, or in alloHSCT recipients considered at higher risk for IFI because of prolonged neutropenia, GVHD, unrelated or alternative donor, and relapsed leukemia at the time of transplant.

Declaration of interest

D Marks has received payment for consulting work from Pfizer Inc, MSD and Basilea. M Slavin has received research funding and honoraria from, and has served in a consultant/advisory role for, Gilead, Merck, and Pfizer Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Medical writing support, under the guidance of the authors, was provided by Martin Bell PhD of Complete Medical Communications, funded by Pfizer Inc.

Additional information

Funding

This paper was funded by Pfizer Inc.