Maturation of Intestinal Oxygenation: A Review of Mechanisms and Clinical Implications for Preterm Neonates

Nutrient requirements of preterm neonates may be substantial, to support growth and maturation processes in the presence of challenging post-natal circumstances. This may be accompanied by substantial intestinal oxygen requirements. Preterm neonates may not be able to meet these oxygen requirements, due to a developmental delay in intestinal oxygenation regulation mechanisms. This review summarizes the available literature on post-natal maturation of intestinal oxygenation mechanisms and translates these changes into clinical observations and potential implications for preterm neonates. The different mechanisms that may be involved in regulation of intestinal oxygenation, regardless of post-natal age, are first discussed. The contribution of these mechanisms to intestinal oxygenation regulation is then evaluated in newborn and mature intestine. Finally, the course of clinical observations is used to translate these findings to potential implications for preterm neonates.


INTRODUCTION
The major functions of the intestine are immunological and digestion-absorption (1,2). The immunological function includes a wide variety of cells and strategies, beyond the scope of this review, and among other things prevents bacterial translocation across the epithelium in the presence of microbial colonization of the gut (1,3,4). The digestive-absorptive function includes production of digestive enzymes and absorption of carbohydrates, lipids, proteins, and vitamins (2). Ultimately, the intestine provides required nutrients to support body growth and function. The nutrient requirements of neonates may be substantial, consequent to maturation processes and tissue growth (5,6). In preterm neonates, these nutritional requirements may be even greater, due to challenging post-natal circumstances, e.g., infection and respiratory distress (7). To meet ongoing nutritional demands, sufficient intestinal oxygenation is essential for intestinal function (8).
After birth, the intestine transitions from a relatively dormant organ to the sole site for nutrient absorption, thus requiring a concomitant increase in oxygen supply (9). This transition is accompanied by rapid tissue growth relative to the whole body, as the intestine increases its weight by 40-70% within 24 h and 4-fold within 10 days (10). Additionally, the change from continuously swallowing amniotic fluid to tolerating intermittent enteral feeds may alter intestinal physiology (11). Therefore, the intestine and its associated circulation are subject to a considerable fetal-to-neonatal adaptation which continues to mature during early life (9,11). Although little is known about post-natal changes in the intestinal circulation of preterm neonates, it can be speculated that these occur with increasing gestational age and post-natal age, as was recently described for the cerebral circulation (12). Among other factors, this may predispose preterm neonates to gastrointestinal complications, such as feeding issues, necrotizing enterocolitis (NEC), and poor growth (13)(14)(15).
Monitoring of intestinal oxygenation may facilitate early identification of gastrointestinal complications. Intestinal blood flow velocity can be measured using Doppler, but this provides only momentary information on oxygen delivery and not oxygen consumption (16). Near-infrared spectroscopy (NIRS) is a non-invasive, bedside technique used to continuously monitor regional oxygen saturation (rSO 2 ) (17). Previous studies have reported promising results for splanchnic rSO 2 (r s SO 2 )monitoring for detecting hemodynamic changes that accompany physiologic and pathophysiologic conditions (18,19). Currently, the use of splanchnic NIRS remains mostly limited to research settings. A better understanding of the different mechanisms regulating intestinal oxygenation may facilitate interpretation of r s SO 2 and advance clinical applications of r s SO 2 -monitoring.
In this review, we will first discuss the presumed mechanisms regulating intestinal oxygenation regardless of post-natal age, derived primarily from animal models. Next, we will discuss maturation of these mechanisms and elaborate on potential clinical implications for preterm neonates.

OVERVIEW OF MECHANISMS REGULATING INTESTINAL OXYGENATION
Intestinal oxygenation represents the balance between oxygen supply and oxygen demand. Intestinal oxygen supply can be divided into convective delivery, from mesenteric arteries to intestinal capillaries, and diffusive delivery, from intestinal capillaries to parenchymal cells (20). Convective delivery depends on intestinal blood flow and arterial oxygen content and is modulated by resistance vessels, i.e., terminal mesenteric and submucosal arterioles (20,21). In contrast, diffusive delivery depends on functional capillary density and capillary-to-cell local PO 2 gradients, and is modulated by precapillary sphincters (20,21).
Intestinal oxygenation is regulated at both local and systemic levels (21,22). Local mechanisms provide real-time modulation of intestinal oxygenation and reflect an intestinal intrinsic capacity (21,23). Systemic, or extrinsic, mechanisms integrate intestinal circulation into the systemic circulation (22). In the next sections, we will discuss these mechanisms in more detail. These mechanisms are summarized in Figure 1.

Intrinsic Regulation of Intestinal Oxygenation
Intrinsic mechanisms for regulating intestinal oxygenation are divided into myogenic factors, metabolic factors, and endothelial vasoactive compounds (24). Both myogenic and metabolic factors participate in pressure-flow autoregulation, i.e., the ability to maintain vascular flow during changes in perfusion pressure. However, the contribution of metabolic factors to this phenomenon seem to exceed those of myogenic factors (25). Additionally, overall vascular responses may be modified by endothelial vasoactive compounds (26).
The myogenic mechanism aims to sustain capillary pressure and transcapillary fluid exchange during changes in transmural pressure (23,27). Increases in intravascular pressure lead to vasoconstriction in resistance vessels and closure of precapillary sphincters (28). Myogenic vasoconstriction in response to circumferential stretch of vascular smooth muscle is mediated by Ca 2+ influx, Ca 2+ release from the sarcoplasmic reticulum, and increased Ca 2+ -sensitivity of contractile myofilaments (29). Intracellular signaling pathways leading to these events may involve protein kinase C (30).
The metabolic mechanism aims to sustain blood flow and oxygen delivery during changes in tissue metabolism (23). Increases in tissue metabolism leads to vasodilation of resistance vessels and relaxation of precapillary sphincters, by reduction of tissue PO 2 and interstitial accumulation of vasoactive metabolites, such as H + , K + , and adenosine (23). Moderate increases in tissue metabolism seem to be associated with augmented diffusive oxygen delivery, whereas greater increases in tissue metabolism seem to be supported by augmentation of convective oxygen delivery (31).
Endothelial vasoactive compounds modulate vascular resistance during changes in shear stress generated by blood flow against the static endothelium (26). The principal relaxing factor is nitric oxide (NO) (32)(33)(34). The principal constricting factor is endothelin-1 (ET-1) (35). Although activation of both ET A -and ET B -receptors on smooth muscle cells leads to vasoconstriction, activation of ET B -receptors on endothelial cells leads to NO-mediated vasodilation (35).
It has been suggested that the enteric nervous system participates in regulation of intestinal oxygenation, however, as this seems to be mediated via endothelial release of nitric oxide (NO), this will not be discussed separately (36). A direct effect of gastrointestinal hormones and peptides in the regulation of intestinal oxygenation has not been clearly established and will therefore not be further addressed in this review (23).

Extrinsic Regulation of Intestinal Oxygenation
Extrinsic mechanisms include neural factors and circulating vasoactive compounds (22). The physiological role of these compounds, including norepinephrine, angiotensin II, vasopressin, histamine, and bradykinin, is uncertain, as these were mostly studied using exogenous administration (23).
Splanchnic nerve stimulation produces a pattern of changes in the intestinal vasculature that is characterized by three phases: a constrictor phase, an escape phase, and a hyperemic phase (23). First, nerve stimulation leads to constriction of resistance vessels and closure of precapillary sphincters. However, during continued sympathetic stimulation, blood flow partially recovers. It is suggested that accumulation of local metabolites and/or release of vasodilators from sensory nerves causes this autoregulatory escape by relaxation of previously constricted resistance vessels (37). Cessation of sympathetic stimulation is followed by a hyperemic phase before blood flow gradually returns to baseline (38). This post-stimulatory hyperemia may be explained by vasodilator metabolite release during the escape phase (39).
The intestine is extensively innervated by parasympathetic fibers originating from the vagus nerve. Although these fibers may have an indirect effect on intestinal oxygenation via changes in intestinal motility and secretion, there does not appear to be a direct vasoactive effect (23).
Several reviews on post-natal maturation of these intrinsic and extrinsic mechanisms describe that some may not be functionally mature at birth, whereas others may functionally decline in the post-natal period (9,(40)(41)(42)(43)(44). However, these reviews do not provide guidance for clinical practice, specifically with regard to preterm neonates. Therefore, our aim was to review the literature on maturation of intestinal oxygenation mechanisms and translate these changes into clinical observations with potential implications for preterm neonates.

METHODS
A literature search was conducted to evaluate reports on the post-natal maturation of intestinal oxygenation mechanisms in neonates. The search strategy is presented in Table 1. English-language articles were selected only if they included a comparison between newborn and mature intestinal oxygenation. Furthermore, as we were mostly interested in baseline intestinal hemodynamics, we excluded articles in which external influences or interventions were investigated. In addition to the database search, we screened the reference lists of all relevant articles for additional publications. To identify clinical studies in preterm neonates, the search strategy was repeated, following the same stepwise procedure, with several additional search terms, as presented in Table 1. We excluded articles that included only sick infants or studied the effect of external influences. In case a control group was included, we included the article, but only present results for the control group.

RESULTS
Our initial search resulted in 9,837 articles. We assessed titles and abstracts of all articles, of which 114 appeared relevant. One additional publication was ascertained using the reference lists within these articles. After reading the full texts, 27 articles were included in this review (Figure 2A). The main findings are presented in Table 2.

Postnatal Maturation of Mechanisms Regulating Intestinal Oxygenation
In the neonatal intestine, basal vascular resistance seems to be determined by myogenic factors, endothelial vasoactive compounds, and neural factors, based on observations in newborn swine (45,51,55,58,61,66,67). Using a swine model, Nowicki et al. (46) demonstrated in several studies that pressure-flow autoregulation is absent in neonatal intestine (54). Therefore, decreases in arterial pressure result in decreased intestinal blood flow. Moreover, these studies indicate that subsequent increases in oxygen extraction are insufficient to meet intestinal oxygen demand (48,50,53). The absence of pressure-flow autoregulation may indicate immaturity of the myogenic and/or metabolic mechanisms described above. However, myogenic vasoconstriction has been observed in newborn swine in response to increases in venous pressure (45,47,54,58,62). Therefore, it seems more likely that the absence of pressure-flow autoregulation reflects immaturity of the metabolic mechanism. Neonatal intestine may thus lack an important vasodilator mechanism. Moreover, there may be a greater amount of constricting factors, as demonstrated by a greater quantity of ET A -and ET Breceptors in newborn swine compared to mature swine (59,63). In the presence of these vasoconstrictor influences, i.e., the myogenic mechanism and ET-1, there seems to be an important role for NO as a vasodilator to maintain intestinal oxygenation in neonates. This assumption is supported by findings of Reber et al. (60,61) who demonstrated that NO production is considerably greater in newborn swine compared with mature swine under both basal and stimulated conditions.
In mature intestine, basal vascular resistance seems to be determined by passive-elastic characteristics of the vasculature rather than active constrictor or dilator tone, based on observations in mature swine (42, 46,53). Pressure-flow autoregulation is present in mature swine and seems to be associated with venous PO 2 , consistent with the metabolic mechanism described above (46, 48-50, 53, 54). Therefore, during decreases in arterial pressure, blood flow is still maintained in mature swine compared to newborn swine and concomitant increases in oxygen extraction may enable mature intestine to more consistently meet oxygen demand (46, 48-50, 53, 54). In contrast to newborn swine, myogenic vasoconstriction in response to venous pressure elevation seems to be absent in mature swine (47,54,58,62). As the influence of vasoconstrictor forces, i.e., the myogenic mechanism and ET-1, decreases with advancing post-natal age, it follows that mature intestine may not require the same vasodilator forces as neonatal intestine. Indeed, several studies showed that NO production and the degree of flow-induced vasodilation are considerably smaller in mature swine than in newborn swine (51,54,60,61).
In conclusion, during post-natal maturation, the contribution of the metabolic mechanism in regulation of intestinal oxygenation increases, whereas the influences of the myogenic mechanism and endothelial vasoactive compounds decrease. In neonatal intestine, pressure-flow autoregulation is not yet functional, possibly due to immaturity of the metabolic mechanism.

Translation to Clinical Observations
To the best of our knowledge, there are no studies available that investigate post-natal maturation of intestinal oxygenation mechanisms in preterm intestine, nor studies that compared maturation of intestinal oxygenation mechanisms between preterm and term intestine.
Therefore, we used the clinical studies that were identified by the additional search strategy to translate the results into clinical observations for preterm neonates. Our additional search resulted in 145 articles. We assessed titles and abstracts of all articles, of which 29 appeared relevant. Three additional publication were ascertained using the reference lists within these articles. After reading the full texts, 18 articles were included in this review ( Figure 2B). The main findings are presented in Table 3.

INTERPRETATION AND DISCUSSION
Our review of the literature shows that pressure-flow autoregulation is only present in mature intestine, as demonstrated in animal studies. In contrast, neonatal intestine relies on increases in oxygen extraction to meet oxygen demand during decreases in arterial pressure. Clinical observations demonstrate lower baseline hemodynamic and oxygenation characteristics in preterm compared to term neonates. These results suggest a developmental delay of vasodilator forces and a smaller reserve to increase intestinal oxygen extraction in preterm neonates that may endanger intestinal oxygenation during decreases in arterial pressure.
Oxygen requirements of neonatal intestine, and specifically those of preterm neonatal intestine, may be substantial, due to a high nutritional demand to support growth and maturation processes (7). Our review of the literature shows that preterm neonates may not be able to meet these requirements, as pressureflow autoregulation is still absent and increases in oxygen extraction are insufficient to meet tissue oxygen demand (46, 48-50, 53, 54). Clinical studies showed that r s SO 2 initially decreases after birth, possibly indicating that oxygen extraction is already maximized under basal conditions (83,84,86,87). Although the course of intestinal blood flow and r s SO 2 in early postnatal life has not been studied with simultaneous Doppler and NIRS measurements, the initial decrease in r s SO 2 suggests that the increase in oxygen extraction is greater than the increase in blood flow during the first days of life. We hypothesize that this could be due to patency of the ductus arteriosus. This hypothesis is supported by findings of Ledo et al. (87) who investigated the effect of ductal patency on the course of r s SO 2 in preterm neonates and found that the increases in r s SO 2 from day 3 of life are paralleled by ductal closure. The initial decrease in r s SO 2 may thus be explained by ductal steal, resulting in reduced diastolic flow in the descending aorta, resulting in a decreased intestinal perfusion pressure, and in the absence of intestinal pressure-flow regulation, preterm intestine relies on increases in oxygen extraction (73,74,80,82,87). This may enable preterm neonates to meet intestinal oxygen requirements during baseline conditions, yet creates unfavorable conditions during periods of additional stress. A recent review by Chaaban et al. (44) showed that the predominant response of neonatal intestine to decreases in oxygen delivery or increases in oxygen demand is increased oxygen extraction. In preterm neonates, however, it may not be possible to further increase oxygen extraction. Therefore, during periods of additional oxygen requirements, preterm neonates may fail to meet intestinal oxygen requirements, leading to disruption of the intestinal barrier and reduced nutrient absorption.
There may be an important role for NO to facilitate oxygen delivery in neonatal intestine during baseline conditions. Our review of the literature demonstrates that NO counteracts active vasoconstrictor tone, induced by myogenic factors and ET-1. Clinical studies showed that intestinal blood flow increases with advancing post-natal age, possibly indicating maturation of vasodilator forces. Although no causal relation between intestinal blood flow and NO production has been demonstrated, Reber et al. (60,61) found that the post-natal increase in intestinal blood flow is paralleled by increases in NO production in neonatal swine. Reber et al. (9) hypothesized that loss of NO production may compromise intestinal oxygen delivery and thus contribute to intestinal injury in neonates. This hypothesis is supported by findings of Nowicki et al. (89) who showed that NO-mediated vasodilation was disrupted in human intestine resected for NEC. The pathophysiology of NEC is complex Blood flow decreased in response to reductions in perfusion pressure.
Oxygen uptake increased in response to reductions in perfusion pressure. Pressure-flow autoregulation was present.
Pressure-flow autoregulation was only present in older animals. Blood flow decreased and oxygen uptake increased to a similar extent in both age groups. As a consequence, oxygen uptake was more effectively maintained in older animals.

1990
Nowicki et al. Oxygen uptake was maintained during arterial pressure reduction. Pressure-flow autoregulation was present and associated with venous PO 2 , but not with blood flow.
Pressure-flow autoregulation was only present in older animals. Oxygen uptake was only maintained in older animals.

1992
Nowicki et al. Myogenic vasoconstriction in response to increased intravascular pressure was absent. Flow-mediated dilation in response to increased flow was present. A modest degree of pressure-flow autoregulation was observed.
Myogenic vasoconstriction was only observed in newborn animals. A greater degree of flow-induced dilation was observed in newborn animals. Vasodilation in response to combined increases in pressure and flow was only observed in newborn animals. Pressure-flow autoregulation was only present in older animals.

1999
Nowicki (55) Swine 50 3, 35 d Vasoconstriction was observed in response to reduction of flow rate. Inhibited NO production increased vascular resistance. Low flow conditions caused increased vasoconstriction in response to ET-1 and these effect were even greater during inhibited NO production.
Vasoconstriction was observed in response to reduction of flow rate. Inhibited NO production increased vascular resistance. Low flow conditions did not alter the response to ET-1.
In newborn animals, low flow conditions caused greater vasoconstriction and a greater vasoconstrictor response to ET-1. Inhibited NO production increased vascular resistance to a greater extent in newborn animals.

(Continued)
Frontiers in Pediatrics | www.frontiersin.org Myogenic vasoconstriction in response to increased arterial pressure was absent, instead vasodilation was observed. Blockade of ET A -receptors, ET B -receptors and NO production produced no effect on vessel diameter.
Myogenic vasoconstriction in response to increased arterial pressure was only observed in newborn animals. In newborn animals, ET-1 participates in setting basal vascular tone, independent of the myogenic mechanism, that is offset by activation of ET B -receptors. Expression of eNOS protein was present. Blockade of NO production increased vascular resistance at day 10, but not at day 30.
Expression of eNOS protein increased until day 10, but then decreased until day 30, whereas eNOS mRNA remained stable. Compared to 1-day-old animals, vascular resistance was higher in 30-day-old animals. Oxygen uptake increased until day 3, but then decreased until day 30.

(Continued)
Frontiers in Pediatrics | www.frontiersin.org An increase in vascular resistance was observed in response to inhibition of the baroreceptor reflex, achieved by occlusion of the carotid arteries. The circulation is under neural vasoconstrictor tone, as evidenced by decreased vascular resistance in response to section of the splanchnic nerve. SpNS produced vasoconstriction.
An increase in vascular resistance was observed in response to inhibition of the baroreceptor reflex, achieved by occlusion of the carotid arteries. The circulation is under neural vasoconstrictor tone, as evidenced by decreased vascular resistance in response to section of the splanchnic nerve. Increased vascular resistance was observed in response to SpNS and MNS.
The mesenteric circulation participated in the baroreceptor reflex in all age groups. Neural factors participate in setting basal vascular tone from birth onwards, but the decrease in vascular resistance observed in response to splanchnic nerve section was greater in older animals. Vascular resistance increased to a greater extent in older animals in response to SpNS.

1987
Buckley et al. Inhibition of the baroreceptor reflex increased vascular resistance. Severing the major components of the innervation increased flow. Vasoconstriction was observed in response to mesenteric nerve stimulation. Autoregulatory escape in response to sustained MNS was not observed.
Inhibition of the baroreceptor reflex increased vascular resistance. Severing the major components of the innervation increased flow. Vasoconstriction was observed in response to mesenteric nerve stimulation. Autoregulatory escape in response to sustained MNS was observed.
From birth, the mesenteric circulation is under neural vasoconstrictor tone and participates in the baroreceptor reflex. The increase in vascular resistance was greater and the latencies for the onset of vasoconstriction in response to MNS were smaller in older animals. From the age of 2 weeks, autoregulatory escape during sustained MNS is demonstrable and it is well-established by the end of the first month.

(Continued)
Frontiers in Pediatrics | www.frontiersin.org     and has not been fully elucidated, but may include impaired intestinal microcirculation (14). In preterm neonates, loss of NO production may thus predispose the intestine to hypoxic tissue injury, possibly contributing to the development of NEC. We acknowledge some limitations. First, our review describes the mechanistic strategies available to neonatal and mature intestine to maintain adequate oxygenation, but does not include the contribution of these strategies during external influences that may alter intestinal oxygen supply or demand, e.g., anemia and enteral feeding. Nevertheless, our speculations on implications of these external influences on intestinal oxygenation are supported by recent reviews in both animal models and preterm neonates (18,19,44). Second, we purposely did not take into account factors that may influence the maturation processes described. These factors may include structural maturation and growth of intestinal tissue and vascularization networks, microbial colonization, and increasing volumes of enteral feeding (6,(90)(91)(92). Third, inherent limitations in Doppler and NIRS techniques complicate the translation to clinical implications. Doppler requires trained personnel, is prone to operator-dependent bias and provides only momentary blood flow velocity measurements of large vessels, whereas NIRS is challenged by intraindividual variability and interference of other tissues, intestinal contents and bowel movements with splanchnic oxygen saturation measurements (18). Finally, translation of our findings, derived from animal studies, to implications for preterm neonates is complicated by interspecies differences. By using NIRS to more regularly monitor r s SO 2 in preterm neonates, we may be able to learn more about the maturational phenomena described previously. For instance, it may provide insight in the influence of GA and post-natal age on basal intestinal oxygenation and it may be used to study the response to external influences that alter oxygen supply or oxygen demand. At the same time, a better understanding of post-natal maturation of intestinal oxygenation mechanisms may facilitate interpretation of r s SO 2 and advance the use of NIRS in a clinical setting. Ultimately, bedside r s SO 2 -monitoring may lead to more streamlined and personalized care practices for at-risk neonates.
In conclusion, preterm intestine may have a smaller reserve for perturbations in intestinal oxygen delivery and oxygen demand, as oxygen extraction may be already be maximized under baseline circumstances. Developing additional understanding on this delicate balance of oxygen supply and demand may help in guiding clinical management to prevent intestinal tissue hypoxia.

AUTHOR CONTRIBUTIONS
BD and EK conceptualized and designed the study. BD screened databases for eligible studies, drafted the initial manuscript, and revised the manuscript after feedback from coauthors. JMi, JMo, JH, AB, and EK critically reviewed the article. All authors contributed to the article and approved the submitted version.