Identifying Gaps in the International Consensus Case Definitions for Invasive Aspergillosis: A Review of Clinical Cases Not Meeting These Definitions.

Background:International consensus definitions for invasive aspergillosis (IA) in research are rigorous, yet clinically significant cases are often excluded from clinical studies for not meeting proven/probable IA case definitions. To better understand reasons for the failure to meet criteria for proven/probable infection, we herein review 47 such cases for their clinical and microbiological characteristics and outcomes. Methods:Data on 47 cases that did not meet consensus IA definitions but were deemed significant were derived from a retrospective, observational, multicenter survey of 382 presumed IA cases across Australasia, of which findings of 221 proven/probable infections were recently published. The clinical, microbiological, and radiologic characteristics of these cases were analyzed. Mortality outcomes were compared with those of 221 proven/probable cases. Results:Of 47 cases studied, 15 lacked classical host factors; 22 exhibited only a single positive Aspergillus polymerase chain reaction result; 7 lacked typical IA radiologic findings on chest computed tomography; and 3 had borderline galactomannan optical density indices (<1.0 but ≥0.5) in bronchoalveolar lavage fluid. The median age of patients was 61 years (IQR, 52-68); 34 were male (72%). Seven patients (15%) required intensive care admission. All patients had lung as the primary site of infection. Antifungal treatment was initiated in 42 patients (89%). All-cause 90-day mortality was 33%, similar to the 30% mortality in the comparative cohort (n = 221). Conclusions:Our findings highlight the limitations of current consensus definitions for IA. Notably, the mortality of patients not meeting these definitions was similar to that of patients with proven/probable IA. Further studies, especially of patients with a single positive Aspergillus polymerase chain reaction result and those without host factors, are needed to determine if future consensus definitions may benefit from modifications.