Designing a Quantitative Videofluoroscopic Analysis Approach in Infants with Spinal Muscular Atrophy Type 1

Abstract

Background: Spinal Muscular Atrophy Type I (SMA 1) is a progressive neuromuscular disorder that causes rapid feeding deterioration in infants. Recent FDA approval of disease-modifying therapies have led to improvements in survival and motor function; however, their effects on bulbar physiology remains unclear. In this study, we aimed to establish a videofluoroscopic swallow study (VFSS) analysis method that optimizes analysis time yet is sensitive to detecting small changes in swallowing physiology to elucidate these deficits. We explored this by comparing kinematic and timing measures obtained using three VFSS analysis approaches: analyzing every swallow observed, selecting a sample of swallows, and manually selecting the worst integrity swallow for physiologic measures. Methods: 30 VFSS from infants with SMA 1 were quantitatively analyzed by a trained rater. To determine the optimal analysis method, the rater measured biomechanical measures on every swallow including oropharyngeal transit time (OPT), hypopharyngeal transit time (HPT), total pharyngeal transit time (TPT), duration of pharyngoesophageal segment opening (PESdur), and pharyngeal constriction ratio (PCR) to represent the ‘every swallow’ analysis approach. A sample of swallows, consisting of 7 swallows occurring in the middle of the VFSS exam, was generated to represent a ‘mid-feed sample’ analysis approach. A manually selected sample of swallows representing the swallow that exhibited the worst physiology for each biomechanical measure was selected to represent the ‘manually selected worst swallow’. Descriptive statistics were used to compare the difference in the worst score an infant received on each physiologic parameter between analysis approaches. Results: The ‘every swallow’ analysis approach revealed global impairments in swallowing physiology with greatest impairment in OPT (0.62 ± 0.51sec), HPT (0.79 ±1.03 sec), TPT (1.41 ±1.48 sec), and PESdur (0.21 ±0.04 sec). Comparison of results between approaches showed similar outcomes in average biomechanical performance captured by the ‘every swallow’ and ‘mid-feed’ sample analysis approaches (p > 0.05). However, the ‘mid-feed’ approach was not able to detect the full extent of deficits across all biomechanical outcomes (p < 0.001). Furthermore, the ‘every swallow’ analysis approach demonstrated significant differences in outcomes compared to the ‘manually selected worst swallow’ approach for both average score and worst score, under identifying deficits in all outcomes (p < 0.05) with the exception of average PESdur (p = 0.966). Despite the attempt to identify the worst integrity using the ‘manually selected worst swallow approach’, it under-identified the magnitude of the worst impairment in all biomechanical measures (p <0.001). Conclusions: Failure to analyze every swallow in an exam of a patient with SMA 1 may confound research results as a result of missed data. Analysis of every swallow visualized during VFSS offers critical information regarding SMA 1 swallowing physiology that is otherwise missed using alternative approaches. Differences in outcomes based on analysis methods may provide research significance but may not provide clinical significance.