Neonatal Non-Invasive Respiratory Support: Overview and Challenges Jeffrey S. Gerdes, MD, MBA Associate Chair, Department of Pediatrics Children’s Hospital of Philadelphia Associate Professor, Perelman School of Medicine Emidio M. Sivieri, MSE Research Bioengineer, CHOP Newborn Care at Pennsylvania Hospital

Why Non-Invasive Support? Airway distending pressure to maintain FRC is the cornerstone upon which all other support modalities are built Less Chronic Lung Disease Fewer airway complications Fewer Infections Reduced inflammation Less stress for babies, families, and staff Lower cost Arguably, CPAP for RDS (Gregory) was the single most important support strategy for RDS. Perhaps after surfactant, which is a drug not a device Beyond CPAP, even fancier modes of ventilation such as HFOV vs. CMV have not shown differences among them, and all have MAP is cornerstone

All Patients during NICU Stay All Patients during NICU Stay Increasing HFNC Use All Patients during NICU Stay All Patients during NICU Stay Courtesy of Dr. Reese Clark, Pediatrix

HFNC NCPAP NIV using RAM Cannula NIPPV HFONC SiPap NAVA HFONC Mechanisms of Action Physiologic Rationale Challenges for Clinical Research Question of the day ??????????????????

Nasal CPAP: Physiologic Rationale Distending pressure recruits lung volume Increases and stabilizes FRC Splints upper airways and compliant chest wall Improves lung Compliance Reduces airway Resistance Decreases Work of Breathing Improves gas exchange Reduces apnea Choosing a CPAP device should address all of these factors, and results may vary depending on lung pathology vs. apnea as major clinical problem Reduces apnea as combo of splinting airway and stimulating receptors (Head’s paradoxical reflex)

NCPAP: Mechanisms of Delivery Constant flow: flow opposition (conventional vent) Constant flow: liquid seal (bubble) Variable flow: “fluidic flip” (Infant Flow) Patient expiratory flow opposes system flow and an expiratory resistor valve Expiratory limb submerged in liquid, with oscillatory nature Inspiratory and expiratory flow compensation via fluidics engineering

All NCPAP devices are not created equal Examples: Variable flow reduces WOB compared to conventional vent CPAP (Lipsten et al, J Perinatol 2005) Variable flow may compensate better for mouth or seal leakage Bubble oscillations may improve lung recruitment (Pillow et al, Am J Crit Care Med 2007) Bubble CPAP pressure delivery is flow dependent and not reliably delivered relative to liquid depth (Kahn et al, Pediatr Res 2007) Bubble CPAP pressure delivery varies more than variable flow CPAP (Kahn et al, Pediatr 2008) Nasal interfaces vary in resistance to flow (DePaoli et al, Arch Dis Child Fetal Neonatal Ed 2002) Variable safety regarding alarms and expiratory limb obstruction risk

Pressure-drop across prongs - by manufacturer at 6 L/min RAM (micro-preemie ) RAM (preemie) RAM (Newborn) (All at 6 L/min Flow) 2 4 6 8 21 Pressure drop (cmH2O) Modified from: De Paoli, Morley, Davis et al. Arch Dis Child Neonatal Ed 2002

Effect of mouth leak Pharyngeal pressure Set CPAP (cm H2O) 8 7 6 5 4 3 1 Pharyngeal pressure (cm H2O) 3 4 5 6 7 8 Set CPAP (cm H2O) De Paoli, et al. Arch Dis Child Neonatal Ed 2005

Bubble CPAP vs. CPAP with Mech. Ventilator Bubble CPAP pressure is flow dependent Bubble CPAP Ventilator 12 10 8 8 Mean Pressure (±SD) (cm H2O) Set CPAP 6 4 4 2 No Leak 4 6 8 10 12 Bubble Bias Flow (L/min) Kahn, Habib, Courtney, Pediatrics 2008

Clinical Correlations of different CPAP Drivers and Interfaces Clinical studies often do not offer clear actionable results Do mechanistic differences matter, or have they not been discoverable with current studies? Babies have widely variable pathophysiology and severity of illness Do the large pragmatic trials of NIRS strategies deliver specific answers for relative efficacy when the “gold standard” of CPAP is delivered with widely divergent modalities?

NIPPV or SNIPPV: Physiologic Rationale Benefits of NCPAP + additional tidal ventilation and/or lung recruitment from higher MAP May reduce WOB relative to CPAP (Aghai et al, Pediatr Pulmonol 2006; Chang et al, Ped Res 2011) Airway pressures actually delivered vary considerably from settings (Owen et al, Arch Dis Child Fetal Neonat Ed. 2010 ) Is variable pressure delivery favorable to lung function? Is apnea Improved through cyclic receptor stimulation? Synchronization may improve efficacy Unclear if NIPPV or bi-level CPAP is more efficacious (Roberts et al, Pediatrics 2013) Are these modalities “Super CPAP”?

Heated and Humidified High Flow Nasal Cannula

HFNC: Mechanisms of Action Gas Conditioning Effects Flow Effects Pressure Effects

HFNC: Mechanisms of Action Gas Conditioning Effects Reduces patient’s metabolic workload needed for Heating and Humidifying (Waugh et al, Granger, RespCare 2004) Improves mechanics - H&H Increases Compliance and Conductance (Greenspan et al, J Peds 1991)

HFNC: Mechanisms of Action Flow Effects Improves lung mechanics Reduced inspiratory resistance => Reduced resistive Work of Breathing (Saslow et al, J Perinatol 2006) Expiration: possible “Coanda” effect (Dysart et al, Resp Med 2009) Washout of Nasopharyngeal dead space => Improved gas exchange (Frizzola et al, Pediatr Pulm 2011) As flow increases, CO2 elimination > increase in MAP Analogous to Trans Tracheal Catheter TGI

Washout of anatomical dead space Continuous flow washes out the upper airways and leads to improved oxygenation Reservoir of fresh gas in upper airway Avoids rebreathing of high- CO2 gas in dead space Proposed flushing of dead space at higher flows Schibler et al, Int Care Med 2011 Washout of nasopharyngeal cavity Washout flow exiting the mouth Nasal cannula Courtesy of Walsh et al, Resp Care 2009

HFNC: Mechanisms of Action Pressure Effects HFNC has been shown to provide positive distending pressure Highly dependent on prong-to-nares diameter ratio and degree of mouth closure (Locke et al, Pediatrics 1993; Sivieri et al, Pediatr Pulm 2013)

HFNC Pressure Delivery: Effect of flow rate, % nares occlusion, and mouth leak Mouth Fully Closed Mouth Fully Open Intra-cannula pressure (74% occl.) Percent Occlusion Pressure Limiting Safety Valve opens Percent Occlusion 100% 86% 74% 74% 68% 68% 55% 55% 44% 44% 36% 36% Sivieri, Gerdes, Abbasi. Pediatr. Pulm. 2013

HFNC delivered pressure Pharyngeal Pressure (cmH2O) Flow 6 Lpm Flow 4 Lpm Flow 2 Lpm 1 minute Wilkinson et al, J Perinatol 2008

Causes of variability of delivery of HFNC % nasal occlusion by catheter Characteristics of catheter design Characteristics of gas conditioning with heat and humidity Mouth leak Variability in disease state of the patient

RAM CannulaTM (Neotech Products) Soft nasal prong interface, relatively large ID FDA approved as a Class I medical device, as a nasal cannula for delivering oxygen: CPAP interfaces are Class 2 Medical Devices 3 sizes: micro-preemie (ID 2.0 ,OD 2.7), preemie (ID 2.0, OD 2.9), newborn (ID 2.5, OD 3.1 ) “Expiratory limb” much smaller diameter than CPAP interfaces Recommended % nasal occlusion: 60-80% Connector attaches to standard ventilator tubing Has been used to deliver CPAP, NIMV, HFONV

RAM CannulaTM (Neotech Products)

RAM CannulaTM (Neotech Products) Soft nasal prong interface, relatively large ID FDA approved as a Class I medical device, as a nasal cannula for delivering oxygen: CPAP interfaces are Class 2 Medical Devices 3 sizes: micro-preemie (ID 2.0 ,OD 2.7), preemie (ID 2.0, OD 2.9), newborn (ID 2.5, OD 3.1 ) “Expiratory limb” much smaller diameter than CPAP interfaces Recommended % nasal occlusion: 60-80% Connector attaches to standard ventilator tubing Has been used to deliver CPAP, NIMV, HFONV

RAM Cannula Pressure Delivery: Bench Test ` Mouth Fully Closed Percent Occlusion Mouth Fully Open Dräger BabyFlow M prongs 4.5mm O.D. RAM Preemie Cannula 3.1 mmO.D. prongs Nares I.D. 100% 100% 78% 60% 48% 38% Gerdes, Sivieri, Abbasi, EAPS Congress 2014

RAM Cannula used for “NCPAP” In bench tests, RAM cannula interface delivers 60% less than the MAP set on the ventilator, even with mouth closed With mouth open, delivered MAP is further reduced By design, RAM cannula is neither NCPAP nor HFNC Using 60-80% nares occlusion does not provide sufficient seal to deliver CPAP. Due caution should be applied if using RAM cannula interface to provide CPAP, pending further studies and classification as a Class II Medical Device

Nasal High Frequency Ventilation Infant Star, Single Nasopharyngeal tube Frequency 10Hz Amplitude 50 torr n=14 Infants GA 27 (25-30) wks BW 1.6 (1.1-2.9) kg 75 70 65 60 55 50 45 40 35 30 pCO2 (mm Hg) p = 0.011 2 hours NHFV Initial pCO2 Final pCO2 Colaizy et al, Acta Paediatr. 200

Does it matter which NIRS strategy is used in a given patient or NICU? HFNC NCPAP NIV using RAM Cannula NIPPV SiPap NAVA HFONC Mechanisms of Action Physiologic Rationale Challenges for Clinical Research Question of the day Does it matter which NIRS strategy is used in a given patient or NICU?

Pulmonary Factors to Consider When Evaluating NIRS Systems Physiologic mechanisms imply that: - NCPAP/NIPPV might be better for lung recruitment - HFNC might be better for CO2 retention - NIPPV might be better for apnea BUT Are we reasonably able to study these differences, And are these differences clinically important?

Extra-pulmonary Factors to Consider When Evaluating NIRS Systems Minimizing nasal septal injury Ease of nursing and respiratory therapy care Comfort for baby and family Noise pollution Promotion of infant development Cost –interface and disposables, amortized cost of the driver, RRT and RN time, and possible impacts on length of stay or severity of illness

NIRS: Summary and Challenges for this conference NCPAP, NIPPV, and HFNC all have physiologic rationales and feasible mechanisms of action All provide varying degrees of respiratory support which improve physiologic parameters and gas exchange in neonates Little is known about which devices may be better for different patho-physiologies or different severities of illness Do the clinical trials have sufficient numbers or stratification of diagnoses and severity of illness to differentiate the utility of these modalities? OR,

NIRS: Summary and Challenges for this conference Are we left with the current state in which clinicians apply well-studied, safe NIRS techniques according to unit or provider preference, matching the device to the baby’s needs, and response to therapy? Are “more randomized trials” the answer, or should we steer towards other research methodologies?