The performance of a pan-fungal PCR-based technique was evaluated to assess the aetiology of invasive fungal diseases (IFDs). A total of 89 formalin-fixed paraffin-embedded biopsy samples from 84 patients with proven IFD were studied. Culture of tissue was performed in 68 (81%) patients. The sensitivities of the PCR-based technique and microbiological culture of tissues were 89% and 56%, respectively (p <0.01). According to PCR results, Aspergillus species accounted for 67%, Candida species for 13%, zygomycetes for 11%, and rare and emerging fungi for 9%. Aspergillus species were significantly associated with lung samples (79.6%, p <0.01), Mucorales were associated with skin/subcutaneous samples, and Candida species were associated with gastrointestinal samples. Regarding biopsy samples with Aspergillus species, Aspergillus fumigatus DNA was detected in 43 of 50 (86%), and Aspergillus flavus in six of 50 (12%). PCR was positive in 24 of 30 (80%) cases with negative culture. In nine of the 84 patients, the PCR technique failed to amplify the DNA. Six also had negative cultures, and in the remaining three cases culture was positive (Rhizopus microsporus, Rhizopus arrhizus, and Sakseneae vasiformis), suggesting that the PCR technique was not as effective in amplifying the DNA of some species of Zygomycetes. In five cases, there was no correlation between culture results and those obtained with DNA amplification, indicating the possibility of a mixed infection or the presence of colonizer/contaminant microorganisms. In summary, PCR-based techniques for DNA amplification should be implemented in histopathology and microbiology departments, as they appear to be complementary to conventional methods for IFD detection.

Summary Autopsy studies show that IA is among the most commonly missed diagnoses in critically ill patients. And, because of lack of unequivocal diagnostic criteria, a timely diagnosis remains challenging. We investigate the epidemiology of and the clinical risk factors for IA in critically ill patients. We conducted a retrospective, observational study of all consecutive ICU patients with evidence of IA in the postmortem examination. During the period of the study (25 years), 893 postmortem examinations were performed in the ICU. Twenty‐five patients (2.8%) were diagnosed with IA in autopsy. Only ten (40%) were classified as IA ante‐mortem, based on the initiation of antifungal treatment. The most common comorbid conditions were corticosteroid treatment (n = 14, 56%), chronic obstructive pulmonary disease (COPD) (n = 11, 44%), immunosuppression (n = 6, 24%) and haematological malignancy (n = 5, 20%). Twenty‐three patients (92%) had three or more risk factors for IA. Critically ill patients with pulmonary infiltrates, treated with high doses intravenous corticosteroids (even for a short period of time), particularly COPD patients who developed worsening respiratory insufficiency despite appropriate treatment were at the highest risk of IA.

No abstract

No abstract