Limits...
Systematic review and mixed treatment comparison meta-analysis of randomized clinical trials of primary oral antifungal prophylaxis in allogeneic hematopoietic cell transplant recipients.

Bow EJ, Vanness DJ, Slavin M, Cordonnier C, Cornely OA, Marks DI, Pagliuca A, Solano C, Cragin L, Shaul AJ, Sorensen S, Chambers R, Kantecki M, Weinstein D, Schlamm H - BMC Infect. Dis. (2015)

Bottom Line: Relative to fluconazole, prophylaxis with itraconazole (odds ratio [OR]: 0.52; interquartile range [IQR]: 0.35-0.76), posaconazole (OR: 0.56; IQR: 0.32-0.99), and voriconazole (OR: 0.46; IQR: 0.28-0.73) reduced incidence of overall proven/probable IFI.All-cause mortality was similar across all mould-active agents.The paucity of comparative efficacy data suggests that other factors such as long-term tolerability, availability of intravenous formulations, local IFI epidemiology, and drug costs may need to form the basis for selection among the mould-active azoles.

View Article: PubMed Central - PubMed

Affiliation: CancerCare Manitoba, 675 McDermot Ave, Winnipeg, MB, Canada. EJBow@cancercare.mb.ca.

ABSTRACT

Background: Antifungal prophylaxis is a promising strategy for reducing invasive fungal infections (IFIs) in allogeneic hematopoietic cell transplant (alloHCT) recipients, but the optimum prophylactic agent is unknown. We used mixed treatment comparison (MTC) meta-analysis to compare clinical trials examining the use of oral antifungals for prophylaxis in alloHCT recipients, with the goal of informing medical decision-making.

Methods: Randomized controlled trials (RCTs) of fluconazole, itraconazole, posaconazole, and voriconazole for primary antifungal prophylaxis were identified through a systematic literature review. Outcomes of interest (incidence of IFI/invasive aspergillosis/invasive candidiasis, all-cause mortality, and use of other antifungals) were extracted from eligible RCTs and incorporated into a Bayesian hierarchical random-effects MTC.

Results: Five eligible RCTs, randomizing 2147 patients in total, were included. Relative to fluconazole, prophylaxis with itraconazole (odds ratio [OR]: 0.52; interquartile range [IQR]: 0.35-0.76), posaconazole (OR: 0.56; IQR: 0.32-0.99), and voriconazole (OR: 0.46; IQR: 0.28-0.73) reduced incidence of overall proven/probable IFI. Posaconazole (OR: 0.31; IQR: 0.17-0.58) and voriconazole (OR: 0.33; IQR: 0.17-0.58) prophylaxis reduced proven/probable invasive aspergillosis more than itraconazole (OR: 0.68; IQR: 0.42-1.12). All-cause mortality was similar across all mould-active agents.

Conclusion: As expected, mould-active azoles prevented IFIs, particularly invasive aspergillosis, more effectively than fluconazole in alloHCT recipients. The paucity of comparative efficacy data suggests that other factors such as long-term tolerability, availability of intravenous formulations, local IFI epidemiology, and drug costs may need to form the basis for selection among the mould-active azoles.

Show MeSH

Related in: MedlinePlus

Treatment effect of antifungal agents on A) proven/probable invasive fungal infection (IFI) and B) proven/probable invasive aspergillosis (IA) at 180 days, compared between the base-case mixed treatment comparison (MTC) and the sensitivity analysis MTC using an empirical prior, expressed in log odds. Estimates less than zero indicate a reduced probability of IFI relative to fluconazole. The vertical bar of the box plot represents the posterior median value (probability <50%); the outer limits of the box plot represent the posterior interquartile range (probability 25%–75%); whiskers represent the most extreme Markov Chain Monte Carlo values of the posterior no more than 1.5 times the width of the interquartile range above or below the upper or lower bounds of the interquartile range.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4374298&req=5

Fig3: Treatment effect of antifungal agents on A) proven/probable invasive fungal infection (IFI) and B) proven/probable invasive aspergillosis (IA) at 180 days, compared between the base-case mixed treatment comparison (MTC) and the sensitivity analysis MTC using an empirical prior, expressed in log odds. Estimates less than zero indicate a reduced probability of IFI relative to fluconazole. The vertical bar of the box plot represents the posterior median value (probability <50%); the outer limits of the box plot represent the posterior interquartile range (probability 25%–75%); whiskers represent the most extreme Markov Chain Monte Carlo values of the posterior no more than 1.5 times the width of the interquartile range above or below the upper or lower bounds of the interquartile range.

Mentions: The post-hoc sensitivity analysis using an empirical Bayesian prior for the heterogeneity parameter (see Methods section for details), yielded estimated treatment effects comparable to the base case (Table 3). However, credible intervals in this additional analysis were considerably less wide than in the base case; to illustrate this point, estimated posterior credible intervals (on the log-odds scale) for IFI overall and IA from both analyses are depicted in Figure 3A and B, respectively. Some of the notable differences in posterior probabilities observed in the base-case analysis became even more pronounced in the sensitivity analysis: for instance, the probability of itraconazole and voriconazole being better than fluconazole for prevention of IFI and the probability of posaconazole and voriconazole being better than fluconazole for prevention of IA were now found to be ≥95% and thus reached traditional thresholds of statistical significance. Results for all outcomes were very similar in a sensitivity analysis that excluded the single posaconazole RCT (see online Additional file 1), and was therefore limited to fluconazole, itraconazole, and voriconazole.Table 3


Systematic review and mixed treatment comparison meta-analysis of randomized clinical trials of primary oral antifungal prophylaxis in allogeneic hematopoietic cell transplant recipients.

Bow EJ, Vanness DJ, Slavin M, Cordonnier C, Cornely OA, Marks DI, Pagliuca A, Solano C, Cragin L, Shaul AJ, Sorensen S, Chambers R, Kantecki M, Weinstein D, Schlamm H - BMC Infect. Dis. (2015)

Treatment effect of antifungal agents on A) proven/probable invasive fungal infection (IFI) and B) proven/probable invasive aspergillosis (IA) at 180 days, compared between the base-case mixed treatment comparison (MTC) and the sensitivity analysis MTC using an empirical prior, expressed in log odds. Estimates less than zero indicate a reduced probability of IFI relative to fluconazole. The vertical bar of the box plot represents the posterior median value (probability <50%); the outer limits of the box plot represent the posterior interquartile range (probability 25%–75%); whiskers represent the most extreme Markov Chain Monte Carlo values of the posterior no more than 1.5 times the width of the interquartile range above or below the upper or lower bounds of the interquartile range.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4374298&req=5

Fig3: Treatment effect of antifungal agents on A) proven/probable invasive fungal infection (IFI) and B) proven/probable invasive aspergillosis (IA) at 180 days, compared between the base-case mixed treatment comparison (MTC) and the sensitivity analysis MTC using an empirical prior, expressed in log odds. Estimates less than zero indicate a reduced probability of IFI relative to fluconazole. The vertical bar of the box plot represents the posterior median value (probability <50%); the outer limits of the box plot represent the posterior interquartile range (probability 25%–75%); whiskers represent the most extreme Markov Chain Monte Carlo values of the posterior no more than 1.5 times the width of the interquartile range above or below the upper or lower bounds of the interquartile range.
Mentions: The post-hoc sensitivity analysis using an empirical Bayesian prior for the heterogeneity parameter (see Methods section for details), yielded estimated treatment effects comparable to the base case (Table 3). However, credible intervals in this additional analysis were considerably less wide than in the base case; to illustrate this point, estimated posterior credible intervals (on the log-odds scale) for IFI overall and IA from both analyses are depicted in Figure 3A and B, respectively. Some of the notable differences in posterior probabilities observed in the base-case analysis became even more pronounced in the sensitivity analysis: for instance, the probability of itraconazole and voriconazole being better than fluconazole for prevention of IFI and the probability of posaconazole and voriconazole being better than fluconazole for prevention of IA were now found to be ≥95% and thus reached traditional thresholds of statistical significance. Results for all outcomes were very similar in a sensitivity analysis that excluded the single posaconazole RCT (see online Additional file 1), and was therefore limited to fluconazole, itraconazole, and voriconazole.Table 3

Bottom Line: Relative to fluconazole, prophylaxis with itraconazole (odds ratio [OR]: 0.52; interquartile range [IQR]: 0.35-0.76), posaconazole (OR: 0.56; IQR: 0.32-0.99), and voriconazole (OR: 0.46; IQR: 0.28-0.73) reduced incidence of overall proven/probable IFI.All-cause mortality was similar across all mould-active agents.The paucity of comparative efficacy data suggests that other factors such as long-term tolerability, availability of intravenous formulations, local IFI epidemiology, and drug costs may need to form the basis for selection among the mould-active azoles.

View Article: PubMed Central - PubMed

Affiliation: CancerCare Manitoba, 675 McDermot Ave, Winnipeg, MB, Canada. EJBow@cancercare.mb.ca.

ABSTRACT

Background: Antifungal prophylaxis is a promising strategy for reducing invasive fungal infections (IFIs) in allogeneic hematopoietic cell transplant (alloHCT) recipients, but the optimum prophylactic agent is unknown. We used mixed treatment comparison (MTC) meta-analysis to compare clinical trials examining the use of oral antifungals for prophylaxis in alloHCT recipients, with the goal of informing medical decision-making.

Methods: Randomized controlled trials (RCTs) of fluconazole, itraconazole, posaconazole, and voriconazole for primary antifungal prophylaxis were identified through a systematic literature review. Outcomes of interest (incidence of IFI/invasive aspergillosis/invasive candidiasis, all-cause mortality, and use of other antifungals) were extracted from eligible RCTs and incorporated into a Bayesian hierarchical random-effects MTC.

Results: Five eligible RCTs, randomizing 2147 patients in total, were included. Relative to fluconazole, prophylaxis with itraconazole (odds ratio [OR]: 0.52; interquartile range [IQR]: 0.35-0.76), posaconazole (OR: 0.56; IQR: 0.32-0.99), and voriconazole (OR: 0.46; IQR: 0.28-0.73) reduced incidence of overall proven/probable IFI. Posaconazole (OR: 0.31; IQR: 0.17-0.58) and voriconazole (OR: 0.33; IQR: 0.17-0.58) prophylaxis reduced proven/probable invasive aspergillosis more than itraconazole (OR: 0.68; IQR: 0.42-1.12). All-cause mortality was similar across all mould-active agents.

Conclusion: As expected, mould-active azoles prevented IFIs, particularly invasive aspergillosis, more effectively than fluconazole in alloHCT recipients. The paucity of comparative efficacy data suggests that other factors such as long-term tolerability, availability of intravenous formulations, local IFI epidemiology, and drug costs may need to form the basis for selection among the mould-active azoles.

Show MeSH
Related in: MedlinePlus