Optimising the management of invasive fungal infections in adult immunocompromised patients

The incidence of opportunistic invasive fungal infections (IFIs) is on the rise, largely driven by the growing number of immunocompromised patients who are at high risk of contracting fungal infections. Patients with haematological malignancies and lung transplantation are particularly vulnerable to IFIs caused by moulds, with Aspergillus species being the predominant fungal pathogen. At the same time, the less well-known moulds such as members of the Scedosporium genera are an emerging cause of IFIs. To date, the management of IFIs in immunocompromised patients remains to be optimised from the aspects of diagnostics, therapeutic antifungal drug monitoring and pharmacoeconomics of antifungal therapy. This thesis aimed to derive critical data to improve the clinical management of IFIs in immunocompromised patients. Specifically, the work in this thesis addresses: the utility of newer diagnostic platforms i.e. Aspergillus galactomannan (GM) and polymerase chain reaction (PCR) assays in haematology patients, voriconazole concentrations in lung transplant recipients, and the pharmacoeconomic of antifungal prophylaxis and invasive scedosporiosis in the haematology population. Non-culture-based diagnostic tests may assist the diagnosis of invasive aspergillosis and guide directed antifungal therapy. However, the clinical utility of Aspergillus GM antigen and Aspergillus genome when tested on bronchoalveolar lavage (BAL) fluid of patients with haematological malignancies remains uncertain. A systematic review and meta-analysis evaluating the diagnostic performance of BAL GM assay in haematology patients was performed. Results from the meta-analysis confirmed the usefulness of GM quantification in BAL fluid for the diagnosis of invasive aspergillosis, yielding a sensitivity, specificity, positive likelihood ratio and negative likelihood ratio of 92%, 98%, 53.7 and 0.08, respectively, at the index cutoff value of 1.5. Combining BAL GM with serum GM or PCR led to a marginal increase in overall sensitivity. Test specificity was significantly decreased by administration of a β-lactam antibiotic at the time of BAL, but sensitivity was not significantly reduced by receipt of mould-active antifungal agents. Many of the previous studies (included in the meta-analysis aforementioned) evaluating the diagnostic accuracy of BAL GM or PCR have restricted study population or bias, which could affect the test performance estimated. As such, to assess the utility of BAL GM and PCR for the diagnosis of invasive pulmonary aspergillosis in routine clinical practice, a multicentre retrospective study involving high-risk haematology patients undergoing BAL was conducted. Our data showed that, in this unselected group of haematology patients in receipt of mould-active antifungal agents and β-lactam antibiotic, BAL GM had lower sensitivity than PCR (61% versus 78%) at an index cutoff of 0.8, but the specificity was higher (93% versus 81%). The optimal OD index cutoff for BAL GM determined in this observational study was lower than that in the aforementioned meta-analysis, probably related to the case-mix of haematology patients, use of therapeutic agents, and BAL protocol. Similar to the findings in our meta-analysis, the sensitivity of both tests was not significantly decreased by administration of mould-active antifungal agent, and the detection rate of invasive pulmonary aspergillosis was not improved by a combination of BAL GM and PCR diagnostic modalities. The thesis also investigated the relationship between trough plasma and pulmonary epithelial lining fluid (ELF) concentrations of voriconazole in lung transplant recipients, with the intention of exploring if plasma voriconazole concentration can be used as a surrogate for the corresponding concentration in ELF. A high-performance liquid chromatographic fluorescence detection method was developed and validated to facilitate measurement of voriconazole concentration in human BAL fluid; a modification of the method was used to quantify voriconazole in plasma. Simultaneous trough concentrations of voriconazole in plasma and ELF (the latter determined from BAL concentration) of 12 lung transplant recipients were determined in a prospective pilot study. Voriconazole achieved consistently higher concentration in the ELF than plasma (mean ± SD ELF:plasma ratio = 12.5 ± 6.3). A strong positive linear relationship between trough plasma and ELF voriconazole concentrations was observed (r2=0.868), which can be described by the equation [VRC]ELF=15.4*[VRC]plasma–1.16, where [VRC]ELF is the voriconazole concentration in ELF and [VRC]plasma is the plasma voriconazole concentration. Findings from this preliminary study suggested the potential for using trough plasma voriconazole concentration as a surrogate for the corresponding concentration in ELF. Another major work of this thesis focuses on the pharmacoeconomics of managing IFIs in the haematology population. Antifungal therapy has a large impact on health resource utilisation and there is currently a paucity of data to guide the selection of antifungal agent for prophylactic use in patients undergoing consolidation chemotherapy for acute myeloid leukaemia. Hence, a pharmacoeconomic evaluation comparing fluconazole, posaconazole and voriconazole prophylaxis was conducted, using retrospective data collected from medical records and cost inputs obtained from Australian databases. The decision-analytical modelling was performed from the perspective of Australian public hospitals. Fluconazole prophylaxis was the most cost-effective approach, with a mean per-patient cost saving of AU$8430 (95% CI AU$5803–AU$11 054) versus posaconazole, and AU$3681 (95% CI AU$990–AU$6319) versus voriconazole. The robustness of the model was confirmed in one-way sensitivity analyses. Little is known about the costs of treating scedosporiosis despite its increasing importance. A multicentre retrospective case-control study was undertaken. Case patients with scedosporiosis were matched (1:2) to controls. Median regression modelling was used to adjust for variables that were not accounted for in the matched-pairs analysis. Invasive scedosporiosis was shown to impose substantial impact on the hospital resources at an adjusted median excess cost of AU$23611, increased median duration of hospitalisation by 13 days and result in a 38-fold elevation in the mortality rate in haematology patients. This is the first report of the attributable economic and clinical outcomes of invasive scedosporiosis in patients with haematological malignancies. In conclusion, this thesis has explored several critical aspects related to the management of IFIs in immunocompromised patients. Importantly, the data generated will assist healthcare providers such as clinicians and pharmacists in moving towards enhanced utilisation of newer fungal diagnostics and antifungal therapy to manage IFIs.