Pharmacokinetics of l-Citrulline in Neonates at Risk of Developing Bronchopulmonary Dysplasia-Associated Pulmonary Hypertension

Abstract

Background

Options to treat pulmonary hypertension (PH) in neonates with bronchopulmonary dysplasia (BPD) are few and largely ineffective. Improving the bioavailability of nitric oxide (NO) might be an efficacious treatment for BPD-PH. When administered orally, the NO-l-arginine precursor, l-citrulline, increases NO production in children and adults, however, pharmacokinetic (PK) studies of oral l-citrulline have not been performed in infants and children.

Objectives

This study characterized the PK of enterally administered l-citrulline in neonates at risk of developing BPD-PH to devise a model-informed dosing strategy.

Methods and results

Ten premature neonates (≤ 28 weeks gestation) were administered a single dose of 150 mg/kg (powder form solubilized in sterile water) oral l-citrulline at 32 ± 1 weeks postmenstrual age. Due to the need to limit blood draws, time windows were used to maximize the sampling over the dosing interval by assigning neonates to one of two groups (ii) samples collected pre-dose and at 1- and 2.5-h post-dose, and (ii) pre-dose and 0.25- and 3-h post-dose. The l-arginine concentrations (μmol/L) and the l-citrulline (μmol/L) plasma concentration-time data were evaluated using non-compartmental analysis (Phoenix WinNonlin version 8.1). Optimal dosage strategies were derived using a simulation-based methodology. Simulated doses of 51.5 mg or 37.5 mg/kg given four times a day produced steady-state concentrations close to a target of 50 μmol/L. The volume of distribution (V/F) and clearance (CL/F) were 302.89 ml and 774.96 ml/h, respectively, with the drug exhibitinga half-life of 16 minutes. The AUC from the time of dosing to the time of last concentration was 1473.3 h*μmol/L, with C_max and T_max of 799 μmol/L and 1.55 h, respectively.

Conclusion

This is the first PK study in neonates presenting data that can be used to inform dosing strategies in future randomized controlled trials evaluating enteral l-citrulline as a potential treatment to reduce PH associated with BPD in premature neonates.

Registration

Clinical trials.gov Identifier: NCT03542812.

1. Introduction

Bronchopulmonary dysplasia (BPD) is a chronic lung disease that affects up to 35% of very low birth weight neonates (<1500 g) [1, 2]. It is estimated that 8–42% of neonates with BPD will develop pulmonary hypertension (PH) [3–7]. Echocardiographic evidence of PH in infants with BPD is associated with up to 40% mortality [5–9]. Moreover, these neonates require lengthy initial hospitalizations, prolonged respiratory support before and after initial discharge, and frequent readmissions. Thus, the individual, familial and societal costs are alarmingly high for neonates who develop BPD-PH [10, 11].

The current primary therapeutic strategy for neonates with BPD-PH is to resolve the underlying respiratory disorder and provide optimal nutrition for lung growth and development. The other long-standing therapy has been to use oxygen as a pulmonary vasodilator [12, 13]. However, clinical evidence supports that vasodilators other than oxygen may be beneficial for treating BPD-PH. In particular, inhaled nitric oxide (iNO) has been shown to function effectively as an acute pulmonary vasodilator in neonates with BPD-PH [14, 15]. Unfortunately, only limited data show clinical improvement in neonates with BPD-PH when iNO is used long-term, for 8–90 days [16]. However, as yet, no randomized clinical trials (RCTs) have been performed to evaluate the efficacy of long-term use of iNO as a treatment to inhibit or reverse the development of BPD-PH in neonates.

The responsiveness to iNO suggests that other agents that improve nitric oxide (NO) bioavailability might prove to be efficacious therapies for BPD-PH. For example, an alternate source of NO is the NO-l-arginine precursor, l-citrulline [17]. The l-citrulline can help to boost l-arginine, and subsequently, this enhances NO production. Compared to the challenges of long-term administration of inhaled therapies such as iNO, l-citrulline can be administered orally or enterally. Indeed, oral administration of l-citrulline has been shown to increase metrics of NO production in children [18] and adults [19] as well as in a newborn animal model of chronic hypoxia-induced PH [20]. In addition, pharmacokinetic (PK) studies are needed to guide the doses and treatment intervals that could be used in a RCT to evaluate oral/enteral l-citrulline as an efficacious treatment for neonates with BPD-PH. Although PK studies of oral l-citrulline have been performed in adults [21, 22], to date, no studies of oral l-citrulline have been performed in human infants and children. Therefore, this study aimed to evaluate the PK of an easy-to-administer oral formulation of l-citrulline to attain a target blood concentration of 50–80 μmol/L in neonates at risk of developing BPD-PH with optimal sampling developed through a model-informed dosing strategy.

2. Methods

2.1. Patient Population

Each hospital’s institutional review board human subjects committee approved the study protocol. Participants were from two neonatal intensive care units (NICU), the University of Utah Hospital NICU and the Intermountain Medical Center NICU, both in Salt Lake City, UT. Eligibility criteria included infants born prematurely at ≤ 28 weeks gestation and requiring invasive, mechanical ventilation or non-invasive positive pressure support (nasal continuous positive airway pressure or high flow nasal cannula ≥ 1 l/min) at 32 ± 1 weeks postmenstrual age. Additional inclusion criteria were that on the day of sample collection, the infant had to be tolerating at least 60 mL/kg/day (one-half of full volume) oral/gavage tube feedings (no patients had gastrostomy tubes), have a continuous need for some form of respiratory support for the prior 14 days, and have hemoglobin ≥ 10 g/dL.

Neonates were excluded if they had any known significant fetal anomalies, chromosomal aneuploidy, or clinical evidence of congenital heart disease, with the exception of patent ductus arteriosus (PDA), atrial septal defect, or ventricular septal defect. Neonates were excluded if they had urine output < 1 mL/kg/h or if they had a history of or were known to have renal dysfunction or liver failure, necrotizing enterocolitis, significant feeding intolerance beyond the first week of life, or the presence of any acute illness as defined by a fever >100.4, vomiting, or diarrhea. Multiple births and futile NICU cases (i.e., neonates with anticipated death prior to hospital discharge) were also excluded. It is generally known that sepsis [23, 24] and renal dysfunction [25] are factors that can influence the plasma concentrations and PK of l-citrulline. Infants with any evidence of an acute infection or who were being treated or evaluated for sepsis were excluded from this study. At the time of enrollment in the study, no infants were receiving nutrition in forms that would increase l-citrulline, including total parenteral nutrition (TPN), or from breastfeeding (human milk) as l-citrulline is negligible in the normal adult diet, or infant formulas which do not contain l-citrulline. Four patients diagnosed by transthoracic echocardiography were identified with a PDA during the screening. Although two patients with PDA received pharmacological treatment with ibuprofen, no patient was receiving ibuprofen while receiving l-citrulline. No patients received indomethacin or aspirin while being administered l-citrulline.

2.2. Study Design

These analyses are for the single-dose PK arm of NCT03542812 (PK of l-citrulline in infants at substantial risk of developing PH associated with BPD). The multiple-dose steady-state arm will be published separately.

Each patient received a single dose of 150 mg/kg oral l-citrulline. The l-citrulline, donated by Asklepion Pharmaceuticals (Baltimore, MD), was provided in powder form, and was solubilized in sterile water to achieve a 50 mg/mL concentration. Each participant received the solubilized l-citrulline of 3 mL/kg (total dose range 150 mg to 285 mg for a weight range of 0.96–1.9 kg). This gave a concentration and volume per day of 3 mL/kg, which was used for the single-dose PK study design. The dose of l-citrulline was delivered through a nasogastric tube 30 minutes prior to feeding.

In order todetermine optimal sampling times, PK data were obtained from an existing publication where data were from young, healthy adult males and scaled to provide initial estimates for a model in neonates [21]. Subsequently, the PK parameter estimates from the literature were used to generate the individual and mean PK profiles from the simulated PK parameters [i.e., apparent clearance (CL/F), volume (V/F)] for premature neonates (< 28 weeks gestation), which included testing covariates including mean weights (1.5 kg and 3.5 kg) and gestational ages (26 and 28 weeks). Therefore, the mean PK profile from the adjusted simulation (represented by the solid black line in Fig. 1) was undertaken for an optimal sampling analysis, and then a limited sampling strategy was tested using WinPOPT [26] and PFIM [27]. Based on the limited sampling results, sampling windows were determined to predict the optimal time to obtain the PK samples. Roberts et al [28] describe these methods in more detail. This study collected PK samples at times based on the optimal sampling windows determined above. On an alternating basis, neonates were assigned to one of two groups to allow different time windows across neonates, as only sparse sampling was possible. For the neonates in group 1, a pre-dose blood sample (to account for endogenous l-citrulline concentrations) was collected 24–48 h prior to administration of the oral l-citrulline dose, and then two post-dose samples were collected at 1 h ± 10 min and 2.5 h ± 10 min post-dose. For the neonates in group 2, a pre-dose blood sample was collected 24–48 h prior to administering the dose of l-citrulline, and two post-dose samples were collected at 15 ± 10 min and 3 h ± 10 min.

Fig. 1.

l-Citrulline concentrations (μmol/L) versus time profiles. Black solid line: simulated PK profile using CL/F and V/F obtained from pooled NCA analysis. Black solid line: simulated PK profile using CL/F and V/F obtained from pooled NCA analysis. Black dashed line: simulated PK profile using CL/F and V/F obtained from the literature. Grey lines with grey circles: Individual profiles of 10 subjects. CL/F apparent clearance, NCA non-compartmental analysis, PK pharmacokinetic, V/F apparent volume of distribution

All blood samples were collected via heel stick into EDTA microtainers. Within 1 hour of collection, blood samples were centrifuged at 1500 g for 5 min at 4 °C. The plasma was placed into cryovials within 30 min, and then the cryovials were placed into a − 80 °C freezer for storage. Samples were analyzed for l-citrulline (μmol/L) and l-arginine concentrations (μmol/L) by ARUP Laboratories (Salt Lake City, UT) via Liquid Chromatograph-Tandem Mass Spectrometry (LC-MS/MS) using the Sciex aTRAQ^™ labeling method for amino acid analysis and an LC-MS/MS system comprising a Sciex API 4000 triple quadrupole mass spectrometer and a Shimadzu High Performance Liquid Chromatography System. Amino acids were separated using a C18 chromatographic column and identified using their characteristic retention times and mass transitions. A calibration curve for each reported amino acid was included in each analytical run to allow accurate quantification, and internal standards were used to correct for ionization differences and ion suppression.

2.3. Pharmacokinetic Analysis and Simulation

l-citrulline plasma concentration-time data were analyzed using non-compartmental analysis (NCA) conducted using Phoenix WinNonlin 8.1 (Certara USA, Inc. Princeton, USA) software. For all data analyses, the molecular weight of 175.2 g/mol was used for l-citrulline. For NCA, the plasma concentration of l-citrulline at the time of dosing (time = 0 h) was assumed to be the same as that of the basal pre-dose concentration. The primary parameters, apparent volume of distribution (V/F) and apparent clearance, were obtained with a pooled NCA analysis. The area under the concentration-time curve (AUC) from the time of dosing to the time of last concentration measurement (AUC _last) was obtained using the linear trapezoidal method for ascending concentrations and the log trapezoidal method for descending concentrations. The AUC _last, maximum observed concentration (C_max), and time to maximum observed concentration (T_max) were calculated by the software. In addition, two sets of simulated datasets corresponding to a single dose administration of l-citrulline were obtained from R (version 4.0.0) via package “linpk” using: (i) PK parameters from the literature [29] (ii) PK parameters from the pooled NCA from this study. The simulated datasets were overlaid with the individual data for visual inspection of simulated versus observed data and prediction of optimal dosing regimens (Fig. 1).

2.4. Pharmacokinetically-Guided Dose Optimization

Optimal dosage strategies were derived using a simulation-based methodology via the Phoenix Non-Linear Mixed Effects tool in Phoenix WinNonlin. Based on the available data, a one-compartment model with extravascular administration was employed using 1000 data points. Optimal dosing strategies were generated to determine the doses and intervals of treatment needed to achieve plasma l-citrulline steady-state concentrations of 50–80 μmol/L in neonates administered l-citrulline via the enteral route of administration for 72 h. The targeted steady-state of plasma l-citrulline concentrations of 50–80 μmol/L reflect at least a 50–100% increase from median basal plasma l-citrulline concentrations in the study participants. Dosing strategies were generated using the literature values for volume of distribution (V/F) of 0.9 L and clearance (CL/F) of 0.623 L/h [29]. The total daily dose was capped to a maximum value of 150 mg/kg/day (average weight of 1.374 kg), which reflects the desire to limit the total daily dosing volume to 4–5 mL/day. The total dose was based on the solution concentration of 50 mg/mL, which was a minor increase in daily fluid intake. The absorption constant (K_a) of 0.95 for the simulations was calculated using the formula T_max = ln(k_a/k_e)/(k_a − k_e). The elimination rate constant (k_e) value of 0.693 (based on a half-life of 60 min) was also obtained from the literature [29]. Time to maximum observed concentration (T_max) for a single dose was obtained from the ten subjects in this study.

3. Results

Maternal characteristics are presented in Table 1. All the mothers received prenatal care; the majority (60 %) received an entire course of antenatal steroids, while 30% did not receive the entire course. The majority of mothers were multiparous. In addition, several mothers were diagnosed with chorioamnionitis (20%) or pregnancy-induced hypertension (30%); none had gestational diabetes.

Table 1.

Characteristics associated with the mothers of the neonates

Characteristics	Median (min, max) or count (%)
Age (years)	29(17, 40)
Gravity (G)
G1	1 (10%)
G2	5 (50%)
G3	0 (0%)
G4	3 (30%)
G5	0 (0%)
G6	0 (0%)
G7	1 (10%)
Marital status
Married	7 (70 %)
Single	3 (30%)
Prenatal care	10 (100%)
Gestational diabetes	0 (0%)
Chorioamnionitis	2 (20%)
Delivery mode
C-section	6 (60%)
Vaginal	4 (40%)
Antenatal steroids
Full course	6 (60%)
Incomplete course	3 (30%)
None	1 (10%)
Pregnancy-induced hypertension	3 (30%)

Demographic characteristics of the neonates providing the l-citrulline plasma concentrations are presented in Table 2. Most of the neonates were Caucasian, and there was an equal gender distribution in the patient population. The median basal plasma concentration for the ten participants (before the study agent was administered) was 31 μmol/L, a range of 9–36 μmol/L.

Table 2.

Patient characteristics and baseline plasma l-citrulline concentration

Characteristic	Median (min, max) or count (%)
Postmenstrual age at birth (weeks)	27 0/7 (23 6/7, 28 5/7)
Postmenstrual age at study date (weeks)	31 3/7 (31 1/7, 32 3/7)
Birth weight (kg)	0.85 (0.61, 1.36)
Birth weight (Z-scores)	0.185 (−1.29, 1.44)
Weight at study date (kg)	1.37 (0.96, 1.92)
Sex
Female	5 (50%)
Male	5 (50%)
Race
Not Hispanic or Latino	8 (80%)
Other	1 (10%)
Asian	0
Black	0
Unknown	1 (10 %)
Ethnicity
Caucasian	8 (80%)
Hispanic or Latino	2 (20%)
Respiratory support at study date
IMV	2 (20%)
CPAP	2 (20%)
High-flow nasal cannula	6 (60%)
FiO2 (%)	26.5 (21, 35)
Volume of feedings (mL/kg/day)	150 (130, 160)
Site
UUMC NICU	8 (80%)
IMC NICU	2 (20%)
Baseline plasma l-citrulline concentration μmol/L	31 (9, 36)

CPAP continuous positive airway pressure, FIO₂ fractional inspired oxygen content, IMC NICU Intermountain Medical Center Newborn Intensive Care Unit, IMV intermittent mandatory ventilation, UUMC NICU University of Utah Medical Center Newborn Intensive Care Unit

l-Citrulline plasma concentration versus time plots for each of the ten patients in the study are shown in Fig. 2. The V/F and CL/F were 302.89 mL mL and 774.96 mL/h, respectively (Table 3), with the drug exhibiting a half-life of 16 min. The CL/F is when the ratio of CL to bioavailability is assumed. The AUC from the time of dosing to the time of last concentration was 1473.3 h*μmol/L, with C_max and T_max of 799 μmol/L and 1.55 h, respectively (Table 3).

Fig. 2.

l-Citrulline concentrations (μmol/L) of individual subjects collected at different time points. Solid black lines withblack circles represent concentrations collected at respective time points. Lines indicate concentrations from an individualsubject

Table 3.

Data from the pooled non-compartmental analysis

Variable	Units	Value
AUClast	h*μmol/L	1473.3
C max	μmol/L	799
T max	h	1.55
V/F	mL	302.89
CL/F	mL/h	774.96

AUC _last area under the concentration-time curve from the time of dosing to the time of the last concentration measurement, CL/F apparent clearance, C_max maximum observed concentration, T_max time to maximum observed concentration, V/F apparent volume of distribution

From visual inspection, it appears that the concentrations simulated with the PK parameters obtained from the Barr et al [29] study in infants with intensive sampling (5 concentrations per dosing interval) fit the study data better than those estimated from the sparse (2 concentrations post-dose per dosing interval) dataset in ten individuals in the current study (Fig. 1). Dosing scenarios (including the dosing regimen and quantity of the agent to be administered) with a maximum total daily dose of 150 mg/kg/day, reflecting a mean total daily dosing volume of 3.0 mL/kg/day, are presented in Table 4. Additionally, the output from the modeling was not normalized to body weight. Simulations for predicted concentrations following a dose of 206.1 mg/day (150 mg/kg/day) divided into varying dosing frequencies that include once a day to eight times a day dosing is shown in Fig. 3 in order to achieve the desired steady-state concentrations target of 8.76–14 μg/mL (or 50–80 μmol/L). l-arginine plasma concentrations versus time plots for each of the ten patients in the study are shown in Fig. 4.

Table 4.

Dosing scenarios with various dosing regimens and target concentrations

Dosing regimen	Quantity to be administered (mg/kg)
QD	206.1 mga (150 mg/kg)
BID	103.1 mga (75 mg/kg)
TID	68.7 mga (50 mg/kg)
4 times per day	51.5 mga (37.5 mg/kg)
5 times per day	41.2 mga (30 mg/kg)
6 times per day	34.4 mga (25 mg/kg)
8 times per day	25.8 mga (18.75 mg/kg)

BID twice per day, QD once per day, TID three times per day

^amg doses are scaled by the mean weight of 1.374 kg for infants in this study

Fig. 3.

Predicted l-Citrulline concentrations (μg/mL) profiles following varying dosing regimens. Black solid and dashed lines represent the 8.76 μg/ml (50 μmol/L) and 14 μg/ml (80 μmol/L) concentrations, respectively. DV Dependent variable [l-Citrulline concentrations (μg/mL)]

Fig. 4.

l-arginine concentrations (μmol/L) of individual subjects collected at different time points. Solid black lines with black circles represent concentrations collected at respective time points. Lines indicate concentrations from an individual subject

4. Discussion

This study is the first to provide essential information to design and perform an RCT evaluating l-citrulline as a PH-targeted medication in premature neonates at risk of developing BPD-PH. Specifically, the data characterize the PK profile of enterally administered l-citrulline in newborn neonates at 32 ± 1-week postmenstrual age born prematurely at birth ≤ 28 weeks gestation. As the study was undertaken in preterm neonates, a limited sampling strategy was used. The initial simulations used were from published PK parameters [29] to determine the PK starting dose for this study. The subsequent PK analysis reconciled the differences between the published PK parameters, which used intravenous (IV) dosing, and this study, which used enteral dosing, and found that they matched remarkably well, supporting the initial dosing determined from the simulations.

The need to identify efficacious treatments for neonates who develop BPD-PH is well acknowledged [6, 30]. However, few PK studies or clinical trials have been performed in neonates with BPD-PH [30–32]. One reason for the paucity of studies in this fragile group of neonates is the reluctance of parents to grant consent for studies that necessitate taking blood or performing experimental therapies. In addition, the number of patients at any one center with BPD-PH are so low that truly informative clinical trials need to be multicenter trials. Due to these and other challenges, the current use of PH-targeted medications in neonates with BPD-PH is almost entirely off-label. It is based on PK studies and clinical trials performed in adult patients [30–32]. However, striking differences exist in the pathophysiology and responsiveness to therapies between adults and neonates with PH, leading to an appreciation that PK studies must be performed in infants with BPD-PH to provide the underpinnings for well-designed RCTs in this patient population [31, 32].

Evidence that impaired NO signaling plays a pivotal role in the pathobiology of PH in humans [33, 34] and underlies, at least in part, the reason for evaluating the use of medications that target the NO pathway as a therapy for neonates with BPD-PH. Data from previous non-randomized clinical studies provide evidence that iNO might be an efficacious therapy for neonates with BPD-PH [14–16]. As expected, there is an increase in plasma l-arginine concentrations that lags behind the l-citrulline concentrations, which could be explained by the fact that there is a sizable portion of circulating l-citrulline converted in the kidney to l-arginine (Fig. 4). Compared to iNO, the l-arginine-NO precursor, l-citrulline, can be administered orally/enterally and therefore has the potential to provide a therapy that is less challenging to administer in the long term than inhaled therapies like iNO. However, dosing regimens requiring administration multiple times a day, i.e., exceeding four times a day, are challenging and can lead to compliance issues. Studies in a newborn piglet model showing that treatment with oral l-citrulline increases pulmonary vascular NO production and inhibits both the onset and progressive development of PH [20, 35], provides proof of concept to evaluate enterally administered l-citrulline to enhance endogenous NO production in human infants. Additional rationale to evaluate l-citrulline as a PH-targeted medication for neonates with BPD comes from a recent study showing that plasma l-citrulline concentrations were lower in human neonates with BPD-PH than in neonates with BPD and no PH [36]. Furthermore, oral l-citrulline given to children with cardiovascular disease undergoing cardiopulmonary bypass to prevent post-operative PH [37] demonstrated that patients with naturally high baseline concentrations of l-citrulline or achieved concentrations exceeding 37 μmol/L did not develop post-operative PH [37]. Furthermore, their findings suggest that achieving l-citrulline plasma concentrations greater than 37 μmol/L might be needed to prevent post-operative PH and perhaps other forms of PH in children.

This study aimed to generate PK data for premature neonates at risk of developing BPD-PH. Current definitions of BPD are based on the respiratory support needs of the infant at 36 weeks of postmenstrual age, such that neonates requiring no respiratory support at 36 weeks postmenstrual age are designated as not having BPD [38–40]. Therefore, the eligibility criteria for this study included the need for respiratory support on the day of study, at 32 ± 1 week’s postmenstrual age, and for at least 14 days prior to the day of study. The choice of studying the neonates at 32 ± 1 week postmenstrual age and not waiting until 36 weeks postmenstrual age, i.e., the postmenstrual age at which BPD is diagnosed, is because of the ultimate goal of performing an RCT to evaluate the ability of oral l-citrulline therapy to prevent or reduce the incidence of BPD-PH. The ultimate goal is to determine whether starting oral l-citrulline treatment at or before 32 weeks postmenstrual age will reduce the percentage of neonates with BPD who have evidence of PH when they are ≥ 36 weeks postmenstrual age.

The CL/F estimate obtained from the ten subjects from the current study is similar to that reported by Barr et al [29]; however, there is a marked difference in the estimate of apparent volume of distribution. There are several differences between the current study and the Barr study. The Barr study used IV dosing, and a formulation with a bioavailability of 100% is assumed. In contrast, our study administered an oral dose of l-citrulline, and bioavailability is assumed to be lower. It should also be noted that the study by Barr et al. [29] included intensive sampling after IV administration of l-citrulline, which enabled a better estimate of the volume of distribution. Target citrulline concentrations of 80–100 μmol/L were achieved in the Barr study when infants and children at a median age of six months received l-citrulline as an IV bolus dose followed by a continuous infusion for 48 hours, with the total daily dose on Days 1 and 2 being 330 mg/kg and 216 mg/kg, respectively. Our simulation predicts that target plasma l-citrulline concentrations of 50 μmol/L can be achieved in premature infants at 32±1 week postmenstrual age by enterally administering a total daily dose of 150 mg/kg/day (206.1 mg/day) divided into dosing frequencies ranging from once a day to eight times a day. The study simulated different dosing frequencies because of the awareness that practical issues must be considered when choosing a dosing strategy for fragile infants. For example, the volume of each dose should be small enough to minimize the potential for emesis. In addition, the frequency of dosing should not be overly burdensome for the caregiver since the long-range intent is that the l-citrulline will be given for weeks to months. It is important to note that a 51.5 mg or 37.5 mg/kg dose of l-citrulline given 4 times a day achieves the target steady-state plasma l-citrulline concentrations of 50–80 μmol/L (Fig. 3) and considers the practical issues.

The decision to target l-citrulline plasma concentrations of 50–80 μmol/L is based on the intention to achieve an elevation of at least 50–100% above the median basal plasma l-citrulline concentration of patients in the study, 31 μmol/L (Table 1), a concentration comparable to previously published basal levels in pediatric patients [41]. The dosing and PK parameters for this study were normalized to body weight. In turn, the choice of elevating plasma concentrations by at least 50–100% is based on findings that oral treatment with l-citrulline inhibited PH development in chronically hypoxic piglets who achieved 50–100% increases from their basal plasma l-citrulline concentrations [20, 35]. In addition, Barr et al [29] chose a target plasma concentration of 80–100 μmol/L based on their desire to achieve a value well above a threshold value of approximately 40 μmol/L that they had previously identified as potentially protective against postoperative PH in children undergoing cardiopulmonary bypass [37, 42].

4.1. Limitations

There are limitations associated with this PK study. First, the actual magnitude of increase from baseline l-citrulline concentrations and the steady-state l-citrulline plasma concentrations that will achieve therapeutic efficacy to inhibit BPD-PH are unknown. However, it is also important to realize that our choice of dosing strategies and target plasma l-citrulline concentrations incorporate the need to consider the total volume of any formulation to be administered to the patient each day and the practicality of dosing at home over extended periods of time. The total daily volume of the l-citrulline is a crucial consideration since neonates with BPD are at risk of developing pulmonary edema when administered high daily total fluid volumes, i.e., total daily fluid volumes exceeding 150 mL/kg/d. l-citrulline has limited solubility, so in order to give a higher dose of l-citrulline as a liquid therapy, a larger fluid volume must be given.

Moreover, the total daily fluid volume given to premature infants must consider all volume-related administrations, including any medications and diluents. Therefore, the solubility of agents needs to be considered as low-soluble compounds will require higher volumes, contribute to the overall volumetric load, and limit the amount of nutrition delivered, thereby compromising growth. The goal was to target an increase from basal l-citrulline concentration by 50–100%, and not by a greater percent increase, based on the need to limit the total volume of l-citrulline administered each day to this vulnerable patient group.

The vast majority of circulating l-citrulline is endogenously produced from amino acids, including glutamine and proline, by enterocytes in the proximal intestines [25]. In order to account for endogenous l-citrulline concentrations, pre-dose samples were collected 24–48 h prior to administration of the oral l-citrulline dose. When administering compounds found naturally, such as amino acids, certain factors influencing plasma concentrations and PK should be considered. The normal human adult diet contains almost no l-citrulline [43]. Therefore, the l-citrulline content in human milk is negligible.

Moreover, because it is considered a non-essential amino acid, l-citrulline is not present in either infant formulas or TPN. Hence, dietary sources given in this study should not contain exogenous l-citrulline in amounts sufficient to affect plasma concentrations or the PK of l-citrulline. Although none of the infants received TPN when enrolled in this study, they all received dietary protein via enteral feedings. For that reason, all patients had a baseline measurement pre-dose, and the endogenous concentrations should be accounted for in each individual; therefore, we do not expect an enteral feeding bolus at 30 min to influence the overall contribution of l-citrulline during the study or to affect the study conclusions.

In other words, rather than influencing the PK profile of the dose of exogenously administered l-citrulline, it is possible that differences in total daily amounts of dietary protein might have contributed to the range of basal plasma l-citrulline concentrations found in these infants. However, with enteral administration, binding of the exogenously administered l-citrulline to proteins in the feedings is also possible, which could not be determined in this study. Therefore, future studies investigating BPD should assess several types of feedings and how human milk may interact with these therapies.

It also merits comment that this was a single-dose PK study and did not include any efficacy endpoints. The results of this study were not intended to and cannot be used to provide information about the efficacy of oral l-citrulline as a therapy for infants with BPD-PH. Moreover, the PK profile generated for premature infants at risk of BPD-PH in this study may not apply to patient populations from other age groups or with other disease conditions.

Conclusions

This study provides data that can inform dosing strategies for future RCT that evaluate oral/enteral l-citrulline as a potential treatment to inhibit BPD-PH in premature neonates. The dosing strategies developed from this single-dose PK study were predicted to achieve a steady-state l-citrulline plasma concentration of 50 to 80 μmol/L, an elevation of at least 50–100% above the median basal plasma l-citrulline concentration of 5.4 μg/mL in patients in this study. The l-citrulline concentrations measured from samples at the pre-determined optimal time points in this study showed that sampling timing from the simulation model were reflective of clinically obtained samples, thus supporting the initial dosing strategies derived from a virtual environment. Further multiple doses, steady-state studies are needed to determine if neonates at risk of developing BPD-PH will achieve the target plasma l-citrulline concentrations of 50–80 μmol/L and if these targets achieve efficacy as a novel therapy for BPD-PH.

Key Points.

Current treatments for pulmonary hypertension (PH) in infants with bronchopulmonary dysplasia (BPD) are few, largely ineffective, and almost entirely based on results of pharmacokinetic (PK) studies and clinical trials performed in adult patients.

The urgent need to obtain PK information and perform clinical trials in newborns is driven by the growing awareness that the pathophysiology and responsiveness to therapies differ between adults and neonates with PH.

This study was performed in premature newborns and provides essential PK data needed to design future clinical trials evaluating the effectiveness of l-citrulline as a treatment for PH in infants with BPD.

Availability of Data and Material

All the raw data are available from the first author, who is ready to share it with any researcher.