Factors to Consider to Study Preductal Oxygen Saturation Targets in Neonatal Pulmonary Hypertension

There are potential benefits and risks to the infant with higher and lower oxygen saturation (SpO2) targets, and the ideal range for infants with pulmonary hypertension (PH) remains unknown. Targeting high SpO2 can promote pulmonary vasodilation but cause oxygen toxicity. Targeting lower SpO2 may increase pulmonary vascular resistance, especially in the presence of acidosis and hypothermia. We will conduct a randomized pilot trial to compare two ranges of target preductal SpO2 in late-preterm and term infants with hypoxic respiratory failure (HRF) and acute pulmonary hypertension (aPH) of the newborn. We will assess the reliability of a newly created HRF/PH score that could be used in larger trials. We will assess trial feasibility and obtain preliminary estimates of outcomes. Our primary hypothesis is that in neonates with PH and HRF, targeting preductal SpO2 of 95–99% (intervention) will result in lower pulmonary vascular resistance and pulmonary arterial pressures, and lower the need for pulmonary vasodilators (inhaled nitric oxide—iNO, milrinone and sildenafil) compared to targeting SpO2 at 91–95% (standard). We also speculate that a higher SpO2 target can potentially induce oxidative stress and decrease response to iNO (oxygenation and pulmonary vasodilation) for those patients that still require iNO in this range. We present considerations in planning this trial as well as some of the details of the protocol design (Clinicaltrials.gov (NCT04938167)).


Introduction
Successful transition at birth is dependent on establishment of lungs as the organ of gas exchange [1]. Breathing at birth and an increase in alveolar oxygen tension (PAO 2 ) leads to an 8-10 fold increase in pulmonary blood flow with a marked reduction in pulmonary vascular resistance (PVR) [2]. Failure to decrease PVR at birth results in hypoxemic respiratory failure (HRF) and acute pulmonary hypertension (aPH) of the newborn [3]. Hypoxic pulmonary vasoconstriction exacerbates PH by increasing PVR [4]. However, administration of excess oxygen can lead to free radical formation, increase pulmonary vascular reactivity and reduce response to pulmonary vasodilators such as inhaled nitric oxide (iNO) [5][6][7]. There is potential for poor outcomes at both higher and lower oxygen saturations (SpO 2 ) measured by pulse-oximetry in aPH, with the ideal target SpO 2 range being unknown. Guidelines from the American Heart Association (AHA) and American Thoracic Society (ATS) suggest that extreme hyperoxia (fraction of inspired oxygen (F i O 2 ) > 0.6) may aggravate lung injury, and be ineffective owing to extrapulmonary shunting [8]. The European Pediatric Pulmonary Vascular Disease Network (EPPVDN) recommends maintaining preductal SpO 2 between 91 and 95% when persistent PH of the newborn (PPHN) is

Rationale for Higher (95-99%) SpO2 in PH
a. Acidosis: Infants with aPH often have associated respiratory and metabolic acidosis that can exacerbate hypoxic pulmonary vasoconstriction [4]. Maintaining a higher SpO2 target (95-99%) may limit hypoxic pulmonary vasoconstriction when pH is <7.2. b. Optimal FiO2: We have previously shown in the lamb model of asphyxia, MAS and PH, that targeting 95-99% SpO2 (achieving a median of 97%) resulted in lower PVR compared to the 90-94% target (achieving a median of 92%- Figure 1). This reduction in PVR was not associated with a statistically significant difference in preductal PaO2 (56 ± 11 mmHg with 90-94% target and 58 ± 19 mmHg with 95-99% target SpO2- Figure 1). However, the mean FiO2 to achieve 95-99% SpO2 was significantly higher than the 90-94% group (0.5 ± 0.21 vs. 0.29 ± 0.17) [10]. Given the importance of alveolar PAO2 in mitigating hypoxic pulmonary vasoconstriction [17], we speculate that FiO2 (in addition to SpO2 or PaO2) plays an important role in pulmonary vasodilation in PH. c. Skin pigmentation, race and SpO2: Skin pigmentation can underestimate hypoxemia by pulse oximetry [18,19]. In neonates the discrepancy between arterial saturation (SaO2) and pulse oximetry (SpO2) is higher in Black infants compared to white infants, especially when SpO2 is below 95%. The incidence of occult hypoxemia (defined as SaO2 < 85% when SpO2 is ≥90%) is more common in Black infants (9.2% of samples) compared to white infants (7.7%) [19]. Targeting SpO2 in the low 90s might increase the risk of occult hypoxemia in infants with darker skin. Interestingly, conditions exacerbated by hypoxemia (such as necrotizing enterocolitis-NEC) and aPH are more common in Black infants, and conditions exacerbated by hyperoxia (such as ROP) are more common among white infants ( Figure 2) [20][21][22][23]. b. Optimal F i O 2 : We have previously shown in the lamb model of asphyxia, MAS and PH, that targeting 95-99% SpO 2 (achieving a median of 97%) resulted in lower PVR compared to the 90-94% target (achieving a median of 92%- Figure 1). This reduction in PVR was not associated with a statistically significant difference in preductal PaO 2 (56 ± 11 mmHg with 90-94% target and 58 ± 19 mmHg with 95-99% target SpO 2 - Figure 1). However, the mean F i O 2 to achieve 95-99% SpO 2 was significantly higher than the 90-94% group (0.5 ± 0.21 vs. 0.29 ± 0.17) [10]. Given the importance of alveolar PAO 2 in mitigating hypoxic pulmonary vasoconstriction [17], we speculate that F i O 2 (in addition to SpO 2 or PaO 2 ) plays an important role in pulmonary vasodilation in PH. c.
Skin pigmentation, race and SpO 2 : Skin pigmentation can underestimate hypoxemia by pulse oximetry [18,19]. In neonates the discrepancy between arterial saturation (SaO 2 ) and pulse oximetry (SpO 2 ) is higher in Black infants compared to white infants, especially when SpO 2 is below 95%. The incidence of occult hypoxemia (defined as SaO 2 < 85% when SpO 2 is ≥90%) is more common in Black infants (9.2% of samples) compared to white infants (7.7%) [19]. Targeting SpO 2 in the low 90s might increase the risk of occult hypoxemia in infants with darker skin. Interestingly, conditions exacerbated by hypoxemia (such as necrotizing enterocolitis-NEC) and aPH are more common in Black infants, and conditions exacerbated by hyperoxia (such as ROP) are more common among white infants ( Figure 2) [20][21][22][23]. d. Therapeutic hypothermia: Infants with moderate to severe hypoxic ischemic encephalopathy (HIE) undergoing therapeutic hypothermia exhibit features of aPH (~25% of neonates cooled down to a core temperature of 33.5 • C and~34% of neonates cooled to 32 • C received iNO) [24]. Konduri  be impaired in HIE and hypothermia. HIE is often associated with left ventricular dysfunction, which can lead to pulmonary vascular congestion from increased left atrial and pulmonary venous pressure resulting in a poor response to iNO [26][27][28].
In addition, during whole body hypothermia, the hemoglobin-oxygen dissociation curve shifts to the left (Figure 3), resulting in higher SpO 2 values for the same PaO 2 range. Targeting 91-95% preductal SpO 2 might lead to hypoxemia (lower PaO 2 ), increased PVR and increased need for ECLS. Finally, the use of IV vasodilators such as milrinone may be associated with severe systemic hypotension during hypothermia, and may contribute to an increased need for ECLS [29]. The optimal target SpO 2 , physiologic basis of hemodynamic and oxygenation response to iNO, sildenafil and milrinone, during whole-body hypothermia are not known.  [26][27][28].
In addition, during whole body hypothermia, the hemoglobin-oxygen dissociation curve shifts to the left (Figure 3), resulting in higher SpO2 values for the same PaO2 range. Targeting 91-95% preductal SpO2 might lead to hypoxemia (lower PaO2), increased PVR and increased need for ECLS. Finally, the use of IV vasodilators such as milrinone may be associated with severe systemic hypotension during hypothermia, and may contribute to an increased need for ECLS [29]. The optimal target SpO2, physiologic basis of hemodynamic and oxygenation response to iNO, sildenafil and milrinone, during whole-body hypothermia are not known.   Response to pulmonary vasodilators and target SpO 2 : The target SpO 2 that results in optimal vasodilation in response to iNO, sildenafil or milrinone is not known. Nitric oxide reacts with superoxide anions to form toxic peroxynitrite ( Figure 4). The bioavailability of iNO is determined by local concentration of superoxide anions [30]. Ventilation with 100% oxygen increases superoxide anion production in pulmonary arterial smooth muscle cells and impairs response to iNO in lambs [5,31]. Gitto et al. ventilated 60 term neonates with aPH with an initial F i O 2 of 0.45 or 0.8. All infants received iNO. Serum IL-6, IL-8 and TNF-α levels were measured over 72 h. Infants in the 0.45 FiO 2 group showed progressive decrease in these inflammatory markers. Infants in the 0.8 F i O 2 group saw an increase in serum IL-6, IL-8 and TNF-α levels [32]. The reduction in oxygenation index (mean airway pressure × F i O 2 × 100 ÷ PaO 2 ) was similar between both groups. These findings suggest a combination of high F i O 2 and iNO can trigger inflammatory cytokines but does not alter improvement in oxygenation [32].  PaO2 = partial arterial oxygen tension, SpO2 = oxygen saturation. To achieve the clinically accepted range of PaO2 50-80 mmHg, SpO2 target is in the low-to-mid 90s during normothermia but in the high 90s during whole body hypothermia. Copyright Satyan Lakshminrusimha.

Rationale for Standard (91-95%) SpO2 Range in PH
a. Response to pulmonary vasodilators and target SpO2: The target SpO2 that results in optimal vasodilation in response to iNO, sildenafil or milrinone is not known. Nitric oxide reacts with superoxide anions to form toxic peroxynitrite ( Figure 4). The bioavailability of iNO is determined by local concentration of superoxide anions [30]. Ventilation with 100% oxygen increases superoxide anion production in pulmonary arterial smooth muscle cells and impairs response to iNO in lambs [5,31]. Gitto et al. ventilated 60 term neonates with aPH with an initial FiO2 of 0.45 or 0.8. All infants received iNO. Serum IL-6, IL-8 and TNF-α levels were measured over 72 h. Infants in the 0.45 FiO2 group showed progressive decrease in these inflammatory markers. Infants in the 0.8 FiO2 group saw an increase in serum IL-6, IL-8 and TNF-α levels [32]. The reduction in oxygenation index (mean airway pressure x FiO2 x 100 ÷ PaO2) was similar between both groups. These findings suggest a combination of high FiO2 and iNO can trigger inflammatory cytokines but does not alter improvement in oxygenation [32]. b. Oxygen toxicity: Use of 100% inspired oxygen is associated with free radical formation in animal models and is associated with reduced response to pulmonary vasodilators such as iNO. The combination of iNO with high FiO2 is proinflammatory due to oxidant injury. However, iNO can exhibit anti-inflammatory and antioxidant effects when FiO2 is low. There are no studies evaluating different target SpO2 ranges with response to iNO.
There is no evidence-based optimal SpO2 target in aPH. While 100% inspired oxygen and SpO2 of 100% with PaO2 values > 150 mmHg have been associated with oxidative stress and impaired response to iNO in animal studies, there are no studies showing different oxidative stress levels or any other negative effects with targeting SpO2 in the 91-99% range. To our knowledge, this will be the first clinical trial evaluating two SpO2 target ranges (91-95% and 95-99%) in late-preterm and term infants with PH.  However, inhaled nitric oxide (iNO) may be more effective due to less inactivation by superoxide anions (O2 .− ) and there may be less oxidative stress and improved cerebral blood flow. Higher oxygen target may be associated with need for higher FiO2, higher PAO2 and enhanced pulmonary vasodilation decreasing RV afterload and right-to-left (R → L) shunts. However, higher inspired oxygen can lead to increased superoxide anion formation that may interact with nitric oxide to form peroxynitrite (OONO .− ), a toxic substance that can inactivate surfactant, cause inflammation and pulmonary vasoconstriction. Copyright Satyan Lakshminrusimha.

Materials and Methods
We will conduct a randomized nonblinded pilot trial to compare two ranges of target preductal SpO2 in late-preterm and term infants with HRF and PH. We will assess the reliability of an HRF/PH score ( Figure 5) that could be used in larger clinical trials. We will also assess trial feasibility and obtain preliminary estimates of outcomes. Our primary hypothesis is that in neonates with PH and HRF, targeting preductal SpO2 of 95-99% (intervention) will result in lower PVR and lower need for non-oxygen-based pulmonary vasodilators (iNO, milrinone and sildenafil) compared to a target of 91-95% (standard). We also hypothesize that although the need for iNO and other vasodilators will be lower in the 95-99% SpO2 target range, when these vasodilators are utilized, the pulmonary vasodilator and oxygenation response will be lower in the intervention group (95-99% target) compared to standard group (91-95% target). b.
Oxygen toxicity: Use of 100% inspired oxygen is associated with free radical formation in animal models and is associated with reduced response to pulmonary vasodilators such as iNO. The combination of iNO with high F i O 2 is proinflammatory due to oxidant injury. However, iNO can exhibit anti-inflammatory and antioxidant effects when F i O 2 is low. There are no studies evaluating different target SpO 2 ranges with response to iNO.
There is no evidence-based optimal SpO 2 target in aPH. While 100% inspired oxygen and SpO 2 of 100% with PaO 2 values > 150 mmHg have been associated with oxidative stress and impaired response to iNO in animal studies, there are no studies showing different oxidative stress levels or any other negative effects with targeting SpO 2 in the 91-99% range. To our knowledge, this will be the first clinical trial evaluating two SpO 2 target ranges (91-95% and 95-99%) in late-preterm and term infants with PH.

Materials and Methods
We will conduct a randomized nonblinded pilot trial to compare two ranges of target preductal SpO 2 in late-preterm and term infants with HRF and PH. We will assess the reliability of an HRF/PH score ( Figure 5) that could be used in larger clinical trials. We will also assess trial feasibility and obtain preliminary estimates of outcomes. Our primary hypothesis is that in neonates with PH and HRF, targeting preductal SpO 2 of 95-99% (intervention) will result in lower PVR and lower need for non-oxygen-based pulmonary vasodilators (iNO, milrinone and sildenafil) compared to a target of 91-95% (standard). We also hypothesize that although the need for iNO and other vasodilators will be lower in the 95-99% SpO 2 target range, when these vasodilators are utilized, the pulmonary vasodilator and oxygenation response will be lower in the intervention group (95-99% target) compared to standard group (91-95% target).

Objectives/Specific Aims
Aim 1: Understand the variation of the HRF/PH score among term and late-preterm neonates and inter-rater reliability of the score for future trial planning. We hypothesize that the HRF/PH score will have a substantial between-infant variance component relative to the within-infant and between-rater variance components, supporting its inter-rater reliability and informing the design of future trials.
Aim 2: Estimate the relationship of the HRF/PH score to outcomes such as duration of mechanical ventilation, pulmonary vasodilator use, incidence of ECLS, and survival.
Aim 3: Assess the feasibility of future trials in regard to patient enrollment, withdrawal or crossover between standard and intervention arms.

Objectives/Specific Aims
Aim 1: Understand the variation of the HRF/PH score among term and late-preterm neonates and inter-rater reliability of the score for future trial planning. We hypothesize that the HRF/PH score will have a substantial between-infant variance component relative to the within-infant and between-rater variance components, supporting its inter-rater reliability and informing the design of future trials.
Aim 2: Estimate the relationship of the HRF/PH score to outcomes such as duration of mechanical ventilation, pulmonary vasodilator use, incidence of ECLS, and survival.
Aim 3: Assess the feasibility of future trials in regard to patient enrollment, withdrawal or crossover between standard and intervention arms.
Aim 4: Obtain preliminary estimates of the intervention's impact on the HRF/PH score, length of stay, duration of mechanical ventilation and need for ECLS. To assist with future trial design, we will estimate the impact of the intervention on outcomes. Even if the HRF/PH score is determined to be invalid or have poor reliability, we will be able to measure the impact of SpO 2 target on PH, and above-mentioned outcomes to assist with future larger, definitive multicenter trial design.

Study Design
This is a randomized single-center pilot trial. Neonates with evidence of HRF and PH will be randomized to one of two SpO 2 target ranges, 95-99% (intervention) vs. 91-95% (standard). This pilot trial will permit development and validation of a novel HRF/PH score and provide preliminary outcome estimates for larger trial planning. Figures 6 and 7 display overall study design and process. measure the impact of SpO2 target on PH, and above-mentioned outcomes to assist with future larger, definitive multicenter trial design.

Study Design
This is a randomized single-center pilot trial. Neonates with evidence of HRF and PH will be randomized to one of two SpO2 target ranges, 95-99% (intervention) vs. 91-95% (standard). This pilot trial will permit development and validation of a novel HRF/PH score and provide preliminary outcome estimates for larger trial planning. Figures 6 and  7 display overall study design and process.

Screening and Eligibility
Neonates in the Neonatal Intensive Care Unit (NICU) will be screened for eligibility. All the following inclusion criteria should be met: corrected gestational age (postmenstrual age) > 34 6/7 weeks; postnatal age < 28 days; on respiratory support with invasive mechanical ventilation, non-invasive mechanical ventilation (including continuous positive airway pressure (CPAP), non-invasive positive pressure ventilation (NIPPV/BIPAP), or high flow nasal cannula (flow rates > 2 L per minute) with F i O 2 ≥ 0.3); and echocardiography findings suggestive of PH (or score > 0 for PH- Figure 5b). If an echocardiogram has not been completed but other inclusion criteria are met without exclusion criteria, then following consent, a research-dedicated echocardiogram can be obtained to determine eligibility.

Screening and Eligibility
Neonates in the Neonatal Intensive Care Unit (NICU) will be screened for eligibility. All the following inclusion criteria should be met: corrected gestational age (postmenstrual age) > 34 6/7 weeks; postnatal age < 28 days; on respiratory support with invasive mechanical ventilation, non-invasive mechanical ventilation (including continuous positive airway pressure (CPAP), non-invasive positive pressure ventilation (NIPPV/BIPAP), or high flow nasal cannula (flow rates > 2 L per minute) with FiO2 > 0.3); and echocardiography findings suggestive of PH (or score > 0 for PH- Figure 5b). If an echocardiogram has not been completed but other inclusion criteria are met without exclusion criteria, then following consent, a research-dedicated echocardiogram can be obtained to determine eligibility.
The following are exclusion criteria: <32 weeks at birth; weight < 2000 g at time of enrollment; severe HRF with oxygenation index (OI) > 35 or SpO2 < 75% on FiO2 = 1.0 on mechanical ventilation for >60 min in spite of correction of reversible factors such as pneumothorax; a condition or congenital anomaly known to be lethal or associated with high likelihood of death during infancy (e.g., chromosomal anomalies such as Trisomy 18 or 13); congenital heart disease other than atrial septal defect (ASD), patent foramen ovale (PFO), patent ductus arteriosus (PDA) or ventricular septal defect (VSD-single or multiple, <2 mm). Patients of varying race and ethnicity will be included. This may result in patients with darker skin pigmentation randomized to the lower SpO2 target. Given prior data regarding possible underestimation of SpO2 by pulse oximetry in darker skin patients, we will also measure melanin-index, a non-invasive measurement of skin pigmentation, in each patient to take this into consideration if differences in arterial blood saturation and pulse oximetry are noted.
Infants with congenital diaphragmatic hernia (CDH), Trisomy 21, or HIE on therapeutic hypothermia may be included.

Criteria for Exit from the Trial Intervention
Exit from the active intervention phase will occur if the infant meets any of the following criteria (whichever is earlier). Once exit criteria are met, the infant will follow unit or discharge protocols for SpO2 targets if needed (such as going home on oxygen, etc.).
1. Weaned to nasal cannula oxygen with FiO2 = 0.21 and flow < 2 LPM; 2. Significant deterioration with OI > 35 or preparation for ECLS cannulation; The following are exclusion criteria: <32 weeks at birth; weight < 2000 g at time of enrollment; severe HRF with oxygenation index (OI) > 35 or SpO 2 < 75% on F i O 2 = 1.0 on mechanical ventilation for >60 min in spite of correction of reversible factors such as pneumothorax; a condition or congenital anomaly known to be lethal or associated with high likelihood of death during infancy (e.g., chromosomal anomalies such as Trisomy 18 or 13); congenital heart disease other than atrial septal defect (ASD), patent foramen ovale (PFO), patent ductus arteriosus (PDA) or ventricular septal defect (VSD-single or multiple, <2 mm). Patients of varying race and ethnicity will be included. This may result in patients with darker skin pigmentation randomized to the lower SpO 2 target. Given prior data regarding possible underestimation of SpO 2 by pulse oximetry in darker skin patients, we will also measure melanin-index, a non-invasive measurement of skin pigmentation, in each patient to take this into consideration if differences in arterial blood saturation and pulse oximetry are noted.
Infants with congenital diaphragmatic hernia (CDH), Trisomy 21, or HIE on therapeutic hypothermia may be included.

Criteria for Exit from the Trial Intervention
Exit from the active intervention phase will occur if the infant meets any of the following criteria (whichever is earlier). Once exit criteria are met, the infant will follow unit or discharge protocols for SpO 2 targets if needed (such as going home on oxygen, etc.).
Parental decision to withdraw from the study; 5.
Provider (attending physician) decision to withdraw from the study; 6. Death.

Randomization
Patients will be randomized in a 1:1 ratio to the higher 95-99% range (intervention) or the lower 91-95% range (standard). The random allocation was generated in blocks (number of treatments per block will be kept confidential to avoid prediction of future patients' allocation). Randomization was not stratified. To conceal allocation, the randomization list was generated by a statistician and uploaded into the REDCap randomization module, which then reveals assignment as each patient is enrolled.

Intervention
Following enrollment and consent, patients will be randomized to one of two SpO 2 targets. The preductal (right upper extremity) SpO 2 will be targeted at either 95-99% (intervention study arm) with alarm limits at 93% and 100% or target 91-95% (comparison standard arm) with alarm limits at 89% and 96% (Figure 8). Since a patient's SpO 2 monitor often leads to titration of F i O 2 by nurses and respiratory therapists, the assigned SpO 2 target range will be posted on the monitor and the patient's respiratory support device (i.e., ventilator) to remind staff of assigned goal. Respiratory support management such as non-invasive or mechanical ventilation titration will be at the discretion of the treating team.
Patients will be randomized in a 1:1 ratio to the higher 95-99% range (intervention) or the lower 91-95% range (standard). The random allocation was generated in blocks (number of treatments per block will be kept confidential to avoid prediction of future patients' allocation). Randomization was not stratified. To conceal allocation, the randomization list was generated by a statistician and uploaded into the REDCap randomization module, which then reveals assignment as each patient is enrolled.

Intervention
Following enrollment and consent, patients will be randomized to one of two SpO2 targets. The preductal (right upper extremity) SpO2 will be targeted at either 95-99% (intervention study arm) with alarm limits at 93% and 100% or target 91-95% (comparison standard arm) with alarm limits at 89% and 96% (Figure 8). Since a patient's SpO2 monitor often leads to titration of FiO2 by nurses and respiratory therapists, the assigned SpO2 target range will be posted on the monitor and the patient's respiratory support device (i.e., ventilator) to remind staff of assigned goal. Respiratory support management such as noninvasive or mechanical ventilation titration will be at the discretion of the treating team.

Blinding
Since the intervention will be administered to critically ill neonates and requires staff to adjust therapies (e.g., FiO2) to achieve the assigned SpO2 range, the treating medical team, investigators, or participants (guardians) will not be blinded. Given prior problems with the use of pulse oximeters with altered algorithms [33], we decided to use standard clinical pulse oximeters without masking for the current pilot trial.

Outcomes
Primary outcome: Change in HRF/PH score ( Figure 3) from baseline (on day 1/time of recruitment) to day 7. Echocardiograms and respiratory details will be collected daily between days 1-7, and the twice weekly until meets exit criteria to assign these scores. If the patient is having an echocardiogram for their routine care, that echocardiogram will be used and overread by members of the research team. Otherwise, a limited researchspecific echocardiogram will be obtained. The echocardiograms will be reviewed/scored by two research personnel independently using the PH portion of the score. The HRF

Blinding
Since the intervention will be administered to critically ill neonates and requires staff to adjust therapies (e.g., F i O 2 ) to achieve the assigned SpO 2 range, the treating medical team, investigators, or participants (guardians) will not be blinded. Given prior problems with the use of pulse oximeters with altered algorithms [33], we decided to use standard clinical pulse oximeters without masking for the current pilot trial.

Outcomes
Primary outcome: Change in HRF/PH score ( Figure 3) from baseline (on day 1/time of recruitment) to day 7. Echocardiograms and respiratory details will be collected daily between days 1-7, and the twice weekly until meets exit criteria to assign these scores. If the patient is having an echocardiogram for their routine care, that echocardiogram will be used and overread by members of the research team. Otherwise, a limited research-specific echocardiogram will be obtained. The echocardiograms will be reviewed/scored by two research personnel independently using the PH portion of the score. The HRF portion of the score will be timed to correlate with the echocardiogram portion given the potential fluctuations in HRF (i.e., FiO 2 administered) throughout the day.
Other outcome data collected will include duration of mechanical ventilation, iNO duration, ECLS use, death before discharge, length of stay, discharge with home O 2 and diagnosis of bronchopulmonary dysplasia (BPD). Follow up at 4, 8 and 12 months will be completed as well.

Sample Size
We set a target sample size of 30 patients (15 patients per arm), which will permit us to address each of our aims, even if 20% of infants are lost to a particular analysis due to missingness. In particular, we would be able to estimate inter-rater reliability of the day-7 HRF/PH score with sufficient precision to provide 90% power to detect a true interrater correlation ≥60% and we would have 84.6% power to detect within-infant, over-time intra-cluster correlation coefficients (ICC) ≥ 30% [34].

Data Monitoring Committee
A data safety monitoring board of neonatologists and a statistician will review deidentified, aggregated data every 6 months or after the first 10 patients complete the study if enrollment is slow. They will review adverse events. Since this is a pilot study, we do not anticipate power to warrant early termination due to undue harm or clear benefit and thus analyses are not planned for that. However, given prior data considering underestimation of SpO 2 by pulse oximetry in darker skin pigmented patients, the data monitoring committee will conservatively evaluate for high risk of infants with darker skin, particularly in the lower SpO 2 target range.

Anticipated Challenges and Limitations
Recruitment of 30 infants with the above specified criteria, particularly the minimum F i O 2 requirement, may be challenging. Due to standard SpO 2 targets, we anticipate the F i O 2 will be weaned as quickly as possible, thus making some infants no longer eligible when we screen. Varying F i O 2 by the treating team prior to enrollment may also limit enrollment. Some potential mitigation measures may include an average F i O 2 over the determined time period before enrollment. Additionally, the achieved SpO 2 may vary compared to the targeted SpO 2 range. We will conduct the analysis based on the intended target SpO 2 group, but we will also extract data for the actual achieved SpO 2 . The OI and OSI ranges in our proposed score may not always result in the same category assignment. Additionally, OSI can be problematic at higher SpO 2 (>97%) as the oxygen-hemoglobin dissociation curve reaches the flat portion. Therefore, if both OI and OSI are available, we will preferentially use OI. We will not measure markers of oxygen toxicity to enhance consenting rates, but we will consider such measurements in the definitive future trial. We will report the PaO 2 data when available from arterial blood gases. For decreased response to iNO in the higher SpO 2 target range, this may be difficult to note considering response to iNO is often classified based on improvement in SpO 2 and the SpO 2 may already be high in this group and thus change very little. Additionally, the relationship between PVR and PaO 2 is not linear. We will attempt to mitigate this issue by applying appropriate transformations (i.e., power/log/inverse) to independent and/or dependent variables.

Conclusions
Current commonly practiced SpO 2 targets for neonates with PH may not be ideal and there are both clinical evidence and animal model data that suggest a higher SpO 2 target may potentially improve HRF/PH outcomes. Higher oxygen saturations can potentially increase oxygen toxicity and limit effectiveness of iNO. There have been no clinical trials assessing SpO 2 targets for term and late-preterm infants with PH to date for a problem that needs answered.

Trial Status
Enrollment started October 2021 and enrollment is expected to take 2 years.

Institutional Review Board Statement:
The study is being conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board at University of California, Davis (protocol code 1766675, date of approval 9 March 2021). The study is registered with ClinicalTrials.gov NCT04938176 and last modified 25 January 2022.