The Identification of Volatile Organic Compunds Profiles Predictive of Invasive Pulmonary Aspergillosis in Breath Samples

BACKGROUND The diagnosis of invasive pulmonary aspergillosis (IPA) bears grave implications for the prognosis and treatment plan of the immunosuppressed patient. The omnipresence of Aspergillus spp in the environment, and by extension - the ubiquity of such isolates in the respiratory system of healthy subjects, poses a great diagnostic challenge. Namely, the differentiation of IPA from non-invasive colonization of the respiratory tract, may be challenging. The vast increase in prevalence of patients who are immunosuppressed, including malignancies and oncological treatments, hematopoietic stem cells and solid organ transplantations, and immunomodulatory medications, accentuates this dilemma.(1) Particularly, patients with acute myelogenous leukemia (AML) are at high risk of developing IPA, with 5-15% of such patients developing IPA during the first 3 months after AML diagnosis.(1-3) As up to 5% of patients with AML may have IPA at presentation, patients admitted to Rambam Medical Center (RMC) with a new diagnosis of AML routinely undergo chest computerized tomography (CT), which serves as screening for the presence of IPA at diagnosis and a baseline for further reference should IPA be suspected later on.(3) Diagnosis of probable or proven IPA in an AML patient requires suggestive imaging findings along with microbiologic evidence. While in some cases the diagnosis can be made using biomarkers (mainly the d-galactomannan antigen and or Aspergillus spp polymerase chain reaction) in peripheral blood, the sensitivity of these assays is low and often this minimally invasive approach does not yield sufficient diagnostic certainty. In such cases, bronchoscopy with bronchoalveolar lavage is usually performed. This invasive procedure is particularly hazardous in patients with suspected IPA, the vast majority of whom are very frail due to their underlying illness.(4) These considerations have led the quest for additional, non-invasive modalities in the diagnosis of IPA. (5) One such approach has been volatile organic compounds (VOC) analysis in exhaled air samples. Characterized by low vapor pressure at room temperature, VOC profile analysis of exhaled air is an emerging field of diagnostics, with the potential of non-invasive screening and diagnosis of various diseases, including neoplasms,(6) infections(7) and inflammatory processes.(8) Previous studies have identified potential VOC markers of metabolically active Aspergillus spp,(9) with one small prospective study in humans(10) showing encouraging diagnostic performance for VOC based IPA identification. However, these attempts have been hampered by several pitfalls. Firstly, VOC profiles of in vitro Aspergillus spp seem to have little relation with VOCs found in breath analysis of colonized patients.(5) Secondly, exhaled VOCs tend to differ greatly between individuals, making the absence of temporal controls (i.e., VOC profile changes in the same individual before and after IPA development) a crucial caveat.(10) Thirdly, to our knowledge no longitudinal studies have assessed the response of VOC profiles to IPA treatment.

In this prospective longitudinal study we aim to overcome these hurdles taking advantage of two key characteristics of hematologic patients treated in RMC. First, patients with newly diagnosed AML have a relatively high prevalence of IPA. Second, a chest CT is performed at the diagnosis of AML in all patients. Thus, collecting breath samples from patients with normal baseline CT will serve as negative controls, allowing for the comparison of VOC profiles before and after the development of IPA in the same patient. Further sampling after the patient has recovered from IPA will provide additional longitudinal control.

GOALS The identification of VOC profiles predictive of IPA in breath samples

METHODS

Study design: