Abstract 908P
Background
Invasive Aspergillosis (IA) is a major cause of early mortality in acute leukemia and stem cell transplants. Establishing a proven histo pathological diagnosis is often difficult in these patients. European organization for research and treatment of cancer (EORTC) classifies diagnosis of IA into proven, probable and possible.
Methods
Episodes of febrile neutropenia (FN) among inpatients aged ≥18 years with acute leukemia or those undergoing Hematopoietic stem cell transplantation (HSCT) were prospectively evaluated over 2 years (N=200). Serum galactomannan was analyzed on day 0, 4, 7 and then weekly till fever persisted. High Resolution Computed Tomography (HRCT) of chest was done on day 4 of fever. Optical density Index (ODI) > 0.5 for ≥ 2 samples was defined as positive. IA was diagnosed as per EORTC 2008 guidelines.
Results
Among 200 enrolled FN episodes, IA was diagnosed in 61 episodes (30.5%): Possible 27 (13.5%), Probable IA 34 (17%), Proven IA 0 (0%). 55.7% (27/61) of possible cases were reclassified as probable based on positive Galactomannan (GM) assay. False positivity rate was 49% but dropped t0 26% when positivity cut off of GM was increased to > 1 ODI. GM assay was able to detect 73% of probable IA cases earlier than CT scans. Use of antifungal was similar in GM positive and negative cases. However, episodes with positive GM had significantly higher use of > 1 antifungals. Overall mortality in all the enrolled episodes was 66 (33%). The mortality in those tested positive for GM was 33 (47.8%) and was 33 (50.7%) in those who tested negative. It was not statistically significant.
Conclusions
In a high risk population of acute leukemia and HSCT, GM assay could identify probable IA earlier than HRCT Chest. Although high number of false positives is a concern; it can be overcome by increasing cut off to 1, especially with known confounders.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.