Lung Ultrasound for Infants' Swallowing Disorders

The dynamic use of LUS for monitoring feeding related aspiration can be effective in detecting meal-related pulmonary abnormalities in neurologically impaired infants and that LUS-management will improve medical/neurological/behavioural status, and reduce chronic pain/discomfort and stress for caregivers. LUS effectiveness is expected to reduce the need for x-Ray, invasive and more expensive techniques (VFSS and FEES). Results may be translated into clinical practice in the management of swallowing disorders in a very vulnerable population, by reducing invasiveness, possible x-Ray early exposition risks, costs, medical complications, and by assessing risk categories based on neuroimaging MRI markers, to rationalize prevention and intervention.

Preliminary data. LUS examinations have been performed on 6 infants with neurological impairment (mean age 15 ± 8 months; Gross Motor Function Classification System (GMFCS) range 2-5, median=4) and in 3 healthy infants (mean age 25 ± 3 months) before and right after a meal. A LUS score ranging from 0 to 18 was established, according to previous literature (0= normal pattern, 1= separated B-lines, 2= confluent B-lines or B-lines with pleural alterations, 3= consolidations; the score 0-3 was then calculated in all 6 thoracic scanning areas and summed up). In the 6 infants with neurological impairment, at paired sample t test, LUS score changed before and after meal (pre-meal 2.50±2.95 vs post-meal 5.17±3.25, p=0.01), with very large effect size (Cohen d=1.6). No change was detected in the control group. Changes in LUS score pre and post-meal correlated with GMFCS (p=0.03, r=0.64) and with adapted Eating and Drinking Ability Classification System (EDACS) scores (p=0.04, r=0.59), as more severe clinical pictures resulted in worse LUS scores. The LUS score pre- meal was worse in infants with neurological impairment compared to controls (2.50±2.95 vs 0.67±0.58, p=0.14). Results make LUS extremely promising as a tool to monitor pulmonary aeration changes in infants with neurological impairment. Feasibility of dorsal and ventral oromotor tract reconstruction has been tested in an infant with neurological impairment.

Specific Aim 1:

• To establish the feasibility and effectiveness of LUS-guided management in the detection of aspiration related to swallowing disorders in children with cerebral palsy and other developmental disabilities, determining advantages of LUS-guided management over standard care on medical (respiratory illness and growth rate) LUS findings themselves and behavioural/neurological outcomes
• To compare diagnostic accuracy of LUS versus X-ray VFSS and invasive FEES

Specific Aim 2:

• To determine specific parameters to estimate aspiration entity by LUS (eg cut-off values for the diagnosis of clinically significant silent or overt aspiration, responsiveness of LUS monitoring to medical, postural or food consistencies adaptation) and their relationship with standardised clinical feeding evaluation and medical/behavioural/neurological measures in order to define risk groups for aspiration

Specific Aim 3:

• To evaluate other applications/potentialities of LUS in children with cerebral palsy and other developmental disabilities.
• To uncover the impact of brain abnormalities detected by brain MRI on LUS findings, for establishing risk categories for unsafe swallowing disorders on early imaging markers, and potential clinical recommendations for early management.

Experimental Design Aim 1:

Double-blinded randomized parallel-designed controlled trial (Consort checklist), with block randomization (blocks of size 4), in one of 2 groups: 1) LUS-monitored management (LUS-m); 2) Standard care management (SC-m). Both groups will undergo an experimental 6-months follow-up. In the first 3 months, participants will be evaluated a minimum of 1 time per month, in-hospital, for a total of 3 evaluations (T1, T2 and T3), plus baseline (T0). At any time point, they will undergo at least one LUS-monitored (before and after) feeding trial (different consistencies might be tested in separate repeated trials according to clinical evaluation). A further LUS evaluation will be performed at a distance of 3 hours, before the next meal to check for resolution of after-meal abnormalities. Both the clinicians who will perform and score LUS, and the patients' family will be blinded to randomization. The child neurologist and the speech and language pathologist (SLP) will be informed of the result of LUS only in the LUS-m group and will include that result to impact on feeding care (postural, thickening fluids or with drugs available for GERD). LUS results in the SC-m group will be available only at the time of data analyses. All meals will be monitored by pulse-oximetry and video recording. A further 6-months LUS-monitored and clinical assessment (T4) will be delivered for collecting outcome measures in both treatment groups. At T0, T3 and T4 time points, infants will perform standardized medical/behavioural/neurological, chronic pain/discomfort assessments. At T0 and T4, a blood sample will be collected, to evaluate general and nutritional status, and a quantitative ultrasound (QUS) will be performed to evaluate bone density. A parent stress questionnaire (Parenting Stress Index (PSI)) will be delivered (T0, T3 and T4). Between T0 and T4, parents will be asked to fill a respiratory infection-diary for respiratory illness rate calculation. VFSS and FEES data will be collected, compared in frequency of execution between groups, and their results will be compared to LUS findings. Primary and secondary outcome measures will be collected at both T3 (short-term) and T4 (long-term).

Experimental Design Aim 2:

Aim 2 is based on the same design as Aim 1. We will explore the frequency, severity and distribution of basal and feeding related LUS parameters and their relationship with clinical (feeding assessment/medical/behavioural/neurological/chronic pain) measures, interventions (postural, thickening fluids or with drugs available for GERD), inter- and intra- group at any time point.

Other possible applications of LUS in the enrolled population will be collected (e.g. pneumonia detection and monitoring).

Experimental Design Aim 3:

Brain MRI is a clinical procedure in the diagnostic approach to children with neurological impairment. Brain abnormalities and corticobulbar tract integrity for lips, tongue, larynx motor and sensory control will be checked to uncover the neural circuits that may underlay swallowing problems leading to silent or overt aspiration and pulmonary risk estimated by LUS. Brain lesions topography and severity will be assessed by using a standardized MRI acquisition protocol according to the diagnostic procedure in use at Stella Maris Scientific Institute MRI Lab. Diffusion-weighted Imaging (DWI) will be used with the aim of revealing structural integrity and connectivity along white matter tracts involved in swallowing.

Expected outcomes:

the LUS-m group having better outcomes, at short- and long-term, including: having lower rate of pulmonary illness, better growth curve, reduction of execution rate of VFSS/FEES. Also, better results at blood sample and bone metabolism, lower pain indices, better scores at neurological/behavioural clinical measures, and less stressful interaction with caregivers.