698 - Feasibility of breath collection in neonates for volatile organic compound detection.

Publication Number: 698.4654

Sanika Kamble, Childrens Hospital of Philadelphia, Philadelphia, PA, United States; Amalia Z. Berna, Childrens Hospital of Philadelphia, Philadelphia, PA, United States; Sagori Mukhopadhyay, Childrens Hospital of Philadelphia, Philadelphia, PA, United States; Yang Liu, Childrens Hospital of Philadelphia, Philadelphia, PA, United States; Kathryn Hafertepe, Childrens Hospital of Philadelphia, New York, NY, United States; Andrew Kau, Washington University, St. Louis, MO, United States; Audrey R. Odom John, Childrens Hospital of Philadelphia, Philadelphia, PA, United States

Background: Breath metabolite profiling is a rapidly emerging discipline that can provide a real-time, noninvasive window into health. One potential application is to use breath volatile organic compounds (VOCs) to evaluate gut microbiome changes. In neonates, gut microbiome is dynamic and is associated with multiple health outcomes. Breath VOCs in neonates may therefore provide an innovative way to study gut microbiome and health. However, human breath collection methods typically require subjects to purposely exhale into sample bags, limiting neonatal studies. We developed a unique system to collect and transport neonatal breath samples. To our knowledge this standardized procedure has not been previously described.

Objective: To demonstrate feasibility of breath collection procedure in neonates.

Design/Methods: This is a prospective cohort study of healthy infants born ≥37 weeks gestational age. Breath collection procedure: We employed neonatal facemasks (2 sizes) with a dual one-way valve system to direct 1L of exhaled breath into a 3L SamplePro Flexfilm sample collection bag (Figure 1). As a control sample, ambient air was collected using a battery-operated pump run for 6 minutes. Both samples were then discharged into concentrating (sorbent) traps and analyzed by thermal desorption Gas Chromatography Mass Spectrometry (GCxGC-MS). Breath sample was deemed adequate in quality if the level of isoprene (a canonical mammalian breath metabolite), had a >5 fold change compared to levels in air sample.

Results: We enrolled 38 neonates with mean gestational age of 39 weeks (SD 1.3). Over half (58%) were male. Samples were collected day 1-3 after birth. Challenges with breath collection included the need to change face mask size in three infants due to inadequate seal and need for repeat collection in one infant due to inadequate volume. While crying during sample collection occurred in 74% neonates, none required aborting the procedure and no serious respiratory events occurred. Two breath samples and one ambient air sample were discarded due to transport errors. To date, we have analyzed 17 breath samples and 17 concomitantly collected air samples. As expected, we found significantly higher abundance of isoprene in breath compared to air samples (Figure 2), confirming successful breath collection.

Conclusion(s): Breath collection in newborns is procedurally feasible and VOCs can be determined from the collected samples. The data presented is part of a larger study aimed at testing the hypothesis that gut bacterial metabolism produces VOCs that are absorbed systemically, distributed, and partitioned into exhaled breath.

Figure1: Demonstration of breath collection.
breath collection demonstration (1).jpegPhotograph of breath collection methodology on infant model.

Figure 2: Abundance of isoprene is increased in infant breath compared to ambient air.
VOCs levels in breath..jpegBox and whisker plot showing distribution of Isoprene level in 15 breath and concomitantly collected 15 air samples. The median isoprene levels in the breath samples (green) were 0.408, compared to the median level in air samples (blue) 0.048, P < 0.001(***).

Figure1: Demonstration of breath collection.
breath collection demonstration (1).jpegPhotograph of breath collection methodology on infant model.

Figure 2: Abundance of isoprene is increased in infant breath compared to ambient air.
VOCs levels in breath..jpegBox and whisker plot showing distribution of Isoprene level in 15 breath and concomitantly collected 15 air samples. The median isoprene levels in the breath samples (green) were 0.408, compared to the median level in air samples (blue) 0.048, P < 0.001(***).