Abnormal Oscillometry in a Multicenter Study of Subjects Following Hematopoietic Stem Cell Transplantation (HSCT) Identified as 'Normal' at Enrollment

Abstract

Rationale: Oscillometry is a tidal breathing test which may be useful to monitor lung function and screen for early manifestations of pulmonary abnormalities after HSCT. We previously reported that oscillometry is abnormal in patients preparing to undergo HSCT. However, little is known about the sensitivity of oscillometry to identify sub-clinical abnormal pulmonary function in this population. Methods: The TRANSPIRE study (NCT04098445) is an ongoing multi-center prospective observational cohort study of lung injury after HSCT in children. We performed oscillometry (Tremoflo © C100, Thorasys, Montreal, CA) at baseline in a cohort of 148 subjects at 4 centers. Children between the ages of 4 and 21 years were eligible for study, of whom 69 had oscillometry pre HSCT. Subjects were identified as either screen normal (sN) or screen abnormal (sAbn) at time of enrollment by the site PI. sAbn subjects had abnormalities in one or more of the following categories: clinical pulmonary function testing (spirometry and DLCO), radiology, clinical history, and bronchoalveolar lavage (BAL). Following ERS 2020 Guidelines for quality control, oscillometry was measured over recording periods of 30 seconds during tidal breathing using commercial equipment (Tremoflo, Thorasys Ltd, 5-37Hz) in at least three trials targeting a coefficient of variation of <15% in respiratory system resistance at5 Hz(R5) for the lowest frequency reported. Variables reported were X5, X11, LogAX, Fres, LogR5-R19. Linear regression analysis was used to compare Transpire subjects to a cohort of 80 normal control patients studied at Children's Hospital of Philadelphia. Differences in oscillometry outcomes between groups were expressed as a height adjusted coefficient which quantitated the mean difference, or 'offset' across the height range measured. Results: TRANSPIRE subjects (sN + sAbn) differed significantly from controls (C) in all measured parameters. Similarly, sN vs controls and sAbn vs controls differed significantly on all parameters. However, there were no statistically significant differences in X5, X11, Log Ax, Fres, or Log R5-R19 between sN and sAbn subjects. Conclusions: In a multicenter study of oscillometry in a cohort of children undergoing HSCT, we found that subjects identified as normal by screening at enrollment in fact had abnormal oscillometry compared to controls. Furthermore, there was not good concordance between our screening tool and oscillometry results, with abnormal oscillometry seen in both screen- normal and screenabnormal subjects. These findings suggest that patients undergoing HSCT have dysfunction of the respiratory system which may not be recognized when evaluating by spirometry, imaging, BAL, and clinical history.