Invasive fungal infections in immunocompromised hosts : epidemiology and diagnostics

Abstract: This doctoral project aimed to attain a more in-depth knowledge of the epidemiology and diagnosis of invasive fungal infections. In Study I, we evaluated the diagnostic performance of PCR/ESI-MS. This novel broad- spectrum molecular test combines multiplex PCR with electrospray ionisation for diagnosing mould lung infections in haematological patients. We analysed bronchoalveolar lavage samples from 27 patients, and PCR/ESI-MS was able to correctly identify all cases of proven and probable invasive fungal infection. It also detected additional cases missed by the standard diagnostic tests, including four cases of Mucorales. PCR/ESI-MS had a clinical impact on the antifungal therapy in 44% of the cases. In Study II, we prospectively investigated the incidence of influenza-associated pulmonary aspergillosis (IAPA) in 12 Swedish intensive care units. Patients were screened for IAPA at inclusion using serum Galactomannan, β-d-Glucan, and a respiratory sample for fungal culture. Bronchoalveolar lavage was performed when feasible. The study included 55 patients, of which five (9%) were classified as probable IAPA according to a recently published case definition. An important finding was that four of these patients lacked traditional risk factors. Patients diagnosed with IAPA had a statistically higher ICU mortality, 60%, compared to 8% in patients without IAPA. Our study is the first prospective study to confirm the association between influenza and IAPA and the first to report data from the Nordic region. Based on our findings, we recommend implementing screening for IAPA for all influenza patients receiving mechanical ventilation in Sweden. In Study III, we present the development of a novel porcine model of invasive candidiasis using monitoring and interventions relevant to intensive care. The final model utilised corticosteroids, repeated Candida inoculation, and continuous endotoxin to simulate the clinical situation where invasive candidiasis occurs. The model consistently demonstrated quantifiable growth of Candida in blood and organs. The proposed model could be a valuable tool for studying pathophysiology and testing new therapies for invasive candidiasis. In Study IV, we prospectively evaluated the efficacy of the commercial T2Candida test for diagnosing intraabdominal candidiasis in surgical patients admitted to intensive care. T2Candida is a new direct-on-blood test that combines PCR with magnetic resonance targeting five clinically relevant Candida species and delivers a result within a few hours. In total, 134 patients were tested with T2Candida and 1,3-β-D-glucan parallel to blood cultures. T2Candida had lower sensitivity for non-candidemia intraabdominal candidiasis compared to 1,3-β-D-glucan, but comparable performance for intraabdominal candidiasis with concurrent candidemia. Still, T2Candida could serve as a valuable complement to 1,3-β-D-glucan in situations where a rapid result can help guide treatment or surgical intervention. In Study V, we retrospectively evaluated the efficacy of reduced-dose trimethoprim- sulfamethoxazole (7.5-15 mg trimethoprim /kg/day) compared to the standard dose (>15-20 mg trimethoprim/kg/day) for treating Pneumocystis jirovecii pneumonia in haematological patients. The study included 113 patients from six Swedish University hospitals, with 80 patients receiving reduced and 33 receiving the standard dose. The groups had no clinically relevant or statistically significant difference in respiratory function or mortality. In patients with mild to moderate pneumonia, defined as PaO2/FiO2 >200 mmHg, all 44 patients receiving the reduced dose were alive at 30 days. The results indicate that the reduced dose is effective and safe for treating mild to moderate Pneumocystis jirovecii pneumonia.