The influence of norepinephrine and phenylephrine on cerebral perfusion and oxygenation during propofol–remifentanil and propofol–remifentanil–dexmedetomidine anaesthesia in piglets

Background Vasopressors are frequently used to increase blood pressure in order to ensure sufficient cerebral perfusion and oxygenation (CPO) during hypotensive periods in anaesthetized patients. Efficacy depends both on the vasopressor and anaesthetic protocol used. Propofol–remifentanil total intravenous anaesthesia (TIVA) is common in human anaesthesia, and dexmedetomidine is increasingly used as adjuvant to facilitate better haemodynamic stability and analgesia. Little is known of its interaction with vasopressors and subsequent effects on CPO. This study investigates the CPO response to infusions of norepinephrine and phenylephrine in piglets during propofol–remifentanil and propofol–remifentanil–dexmedetomidine anaesthesia. Sixteen healthy female piglets (25–34 kg) were randomly allocated into a two-arm parallel group design with either normal blood pressure (NBP) or induced low blood pressure (LBP). Anaesthesia was induced with propofol without premedication and maintained with propofol–remifentanil TIVA, and finally supplemented with continuous infusion of dexmedetomidine. Norepinephrine and phenylephrine were infused in consecutive intervention periods before and after addition of dexmedetomidine. Cerebral perfusion measured by laser speckle contrast imaging was related to cerebral oxygenation as measured by an intracerebral Licox probe (partial pressure of oxygen) and transcranial near infrared spectroscopy technology (NIRS) (cerebral oxygen saturation). Results During propofol–remifentanil anaesthesia, increases in blood pressure by norepinephrine and phenylephrine did not change cerebral perfusion significantly, but cerebral partial pressure of oxygen (Licox) increased following vasopressors in both groups and increases following norepinephrine were significant (NBP: P = 0.04, LBP: P = 0.02). In contrast, cerebral oxygen saturation (NIRS) fell significantly in NBP following phenylephrine (P = 0.003), and following both norepinephrine (P = 0.02) and phenylephrine (P = 0.002) in LBP. Blood pressure increase by both norepinephrine and phenylephrine during propofol–remifentanil–dexmedetomidine anaesthesia was not followed by significant changes in cerebral perfusion. Licox measures increased significantly following both vasopressors in both groups, whereas the decreases in NIRS measures were only significant in the NBP group. Conclusions Cerebral partial pressure of oxygen measured by Licox increased significantly in concert with the vasopressor induced increases in blood pressure in healthy piglets with both normal and low blood pressure. Cerebral oxygenation assessed by intracerebral Licox and transcranial NIRS showed opposing results to vasopressor infusions. Electronic supplementary material The online version of this article (10.1186/s13028-018-0362-z) contains supplementary material, which is available to authorized users.


Background
Vasopressors are frequently used to increase blood pressure in order to ensure sufficient cerebral perfusion and oxygenation (CPO) during hypotensive periods induced by e.g. anaesthesia or septic shock [1][2][3][4]. Two of the most commonly used vasopressors are norepinephrine and phenylephrine [5]. The quality of the response in cerebral blood flow (CBF) to vasopressor treatment has been shown to rely on the type of vasopressor used [3, 6, 7] and on the concurrent anaesthetic protocol [3].
Vasopressors have a well-documented systemic cardiovascular effect [8]. The primary benefit of vasopressors on CPO is believed to be due to the concomitant elevation of cerebral perfusion pressure following elevation of the systemic blood pressure [9]. However, if systemic pressure is kept within the limits of cerebral autoregulation, vasopressor treatment should have little effect on cerebral perfusion, despite an increase in blood pressure [9, 10]. Porcine cerebral arteries and veins have been reported to have dense sympathetic innervation and to be susceptible to vasopressor induced vasoconstriction in vitro [11]. Vasopressor-induced vasoconstriction has also been reported in vivo in healthy humans [12]. The sympathomimetic effect of norepinephrine is mediated by binding to both α-(α 1 ) and β-(β 1 and β 2 ) adrenergicreceptors, whereas phenylephrine is a highly selective α 1 -agonist. The cerebral veins are more sensitive to sympathetic activation than cerebral arteries, and vasoconstriction are more specifically mediated by α 2 rather than α 1 adrenoceptors [11]. These differences may have contributed to the varying vasopressor effect on cerebral arteries versus veins observed in humans [12]. Despite the presence of α-adrenoceptors in the cerebral arteries, the vasoconstrictive effect of vasopressors on the cerebral arteries has been reported as clinically insignificant, since maximal stimulation has been shown to only reduce CBF by 5-10%, in healthy humans [10,13].
General anaesthesia with propofol in combination with a potent opioid, such as remifentanil, has become increasingly popular [14] and the preferred total intravenous anaesthesia (TIVA) protocol in neuroanaesthesia and paediatric intensive care units [4,15]. Propofol alone or in combination with remifentanil has been shown to preserve cerebral autoregulation in anaesthetic doses in both human [16][17][18] and animal studies [19,20], and may thus be favourable in experimental studies requiring intact cerebral autoregulation. The α 2 -agonist, dexmedetomidine, has been receiving increasing attention as anaesthetic adjuvant in human anaesthesia and intensive care because of its abilities to preserve cerebral autoregulation and because of its near-ideal sedation. It has furthermore been recognized for facilitating better haemodynamic stability, analgesia and neuroprotection [15,[21][22][23][24]. In veterinary anaesthesia, α 2 -agonists have been widely used for many years as premedication, sedation and analgesia [25][26][27]. The hemodynamic effect of dexmedetomidine has shown to be both dose-and species dependent [28]. Dexmedetomidine has been shown to decrease systemic blood pressure in both humans and animals and to cause a generalized decrease in CBF [29][30][31][32][33][34]. This has not consistently been associated with decreased cerebral metabolic rate of oxygen [30,31] and concern has been raised that dexmedetomidine may have the potential to cause cerebral vasoconstriction [23,35].
The potential interactive vasoconstrictive effect of vasopressors and anaesthetics on the cerebral vasculature may be of concern in regard to CBF in hemodynamically compromised or neurocritical patients [36,37]. The effect of vasopressors on CBF may be influenced by anaesthesia if cerebral autoregulation is preserved (intravenous anaesthesia) or impaired (inhalation anaesthesia) [3, [16][17][18], the latter making CBF more blood pressure-dependent with high gas concentrations [38], or when used with anaesthetics that might precondition cerebral vasoconstriction (α 2 -agonists) [39]. The systemic adrenoceptor-mediated properties of norepinephrine and phenylephrine produce different circulatory effects [40,41]. Norepinephrine has been shown to improve myocardial function whereas phenylephrine decreases ventricular performance. In addition, norepinephrine appears to decrease microcirculatory blood flow to the abdominal organs, whereas phenylephrine does not [5]. Therefore, the choice of anaesthetic protocol in experimental animal studies should include consideration of such interactions to avoid adverse effects on CPO.
The objective of this study was to investigate the CPO response to vasopressor infusion with norepinephrine and phenylephrine during propofol-remifentanil and propofol-remifentanil-dexmedetomidine TIVA, in healthy piglets with normal and lowered blood pressure.

Study design and animals
Full study details and data regarding the entire study, and results regarding the effect of dexmedetomidine on CPO have been reported elsewhere [42] and are presented in Additional file 1. The same animals were used for the study of vasopressor effect on CPO in this study.
In summary: This study was designed as a non-blinded, randomized, two-arm parallel group, experimental animal trial (Fig. 1). Sixteen female slaughter piglets (Danish Landrace/Yorkshire/Duroc) with a body weight of 25-34 kg were used. In one group, low blood pressure (LBP) was induced using caval block [43,44], whereas in the other group normal blood pressure (NBP) was maintained. All animals were subjected to the same anaesthetic protocol and did not receive premedication prior to induction on the day of the experiment.

Anaesthesia
General anaesthesia was induced while the animals were still in their pen to minimise stress. A dosage of 4-8 mg/ kg propofol was given through a catheter placed in an auricular vein the day before the experiment. After endotracheal intubation, the animals were connected to a mechanical ventilator. End tidal CO 2 (EtCO 2 ) was maintained between 35 and 50 mmHg. Fraction of inspired oxygen (FiO 2 ) was maintained around 0.8, and was centrally supplied by an in-house generator.
General anaesthesia was maintained using a TIVA protocol with separate syringe pumps (Terumo, Terufusion Syringe Pump TE-331, Belgium) for propofol (12-20 mg/ kg/h) and remifentanil (20-45 µg/kg/h) TIVA. Doses were individually regulated to accomplish unresponsiveness to noxious stimuli (dewclaw pinching), with propofol doses adjusted to control anaesthetic depth (assessed as lack of movement) and remifentanil doses adjusted to eliminate responses to noxious stimuli. Once animal preparation was completed, anaesthetic doses were kept unchanged. Dexmedetomidine was supplemented after the first half of the experiment (Fig. 1) with a bolus of 1 µg/kg given over 10 min followed by a fixed dose of 0.7 µg/kg/h iv.

Surgical preparation and instrumentation
All invasive procedures were conducted after infiltration with a mixture of lidocaine and bupivacaine. After surgical cut down, the femoral artery was cannulated for invasive blood pressure monitoring and intermittent blood collection for blood gas analysis. An eight French Fig. 1 Experimental design. The experimental design and flow for the two groups (NBP normal blood pressure group, LBP low blood pressure group). Key time points are marked with arrows and vertical bars. Solid vertical bars show pre-or post-intervention baselines, where the red arrows indicate supplemental blood gas readings. Open vertical bars show vasopressor challenges. Yellow horizontal bar indicates period with propofolremifentanil TIVA, green horizontal bar indicates period with propofol-remifentanil-dexmedetomidine, and blue horizontal bar indicates period with induced hypotension. PCB pre-caval block, PR-1 baseline during propofol-remifentanil, NE-1 norepinephrine during propofol-remifentanil, PR-intvas after norepinephrine and wash-out period during propofol-remifentanil, PE-1 phenylephrine during propofol-remifentanil, PR-2 after phenylephrine wash-out period/pre-dexmedetomidine during propofol-remifentanil, PRD propofol-remifentanil-dexmedetomidine, NE-2 norepinephrine during propofol-remifentanil, PRD-intvas after norepinephrine and wash-out period during propofol-remifentanil-dexmedetomidine, PE-2 phenylephrine during propofol-remifentanil-dexmedetomidine, PRD-end after phenylephrine and wash-out period during propofol-remifentanil-dexmedetomidine (end of experiment) balloon-tipped catheter was placed in the femoral vein in all animals, with the balloon positioned in the caudal vena cava just below the heart. A urinary catheter with a closed collecting bag was placed to prevent bladder distension. Isotonic glucose solution was administered throughout the experiment at 2.5 mL/kg/h. A multiparametric bedside monitor recorded haemodynamic and respiratory variables every 30 s, and data were transferred to a personal computer using Datex-Ohmeda S/5 ™ Collect software (GE Healthcare, Helsinki, Finland). The collected variables were pulse rate, and mean arterial blood pressure (MAP), body temperature (oesophageal probe), fractionated inspired oxygen (FiO 2 ), and EtCO 2 (Additional file 1). Electrocardiogram and peripheral oxygen saturation by pulse oximetry (S p O 2 ) measured on the tail or the lower lip were monitored for continuous assessment.

Cerebral perfusion and oxygenation (CPO) measures
A circular craniotomy (20-30 mm) was performed over the right parietal lobe with a 5 mm craniotome, and dura was removed. A laser speckle contrast imaging (LSCI) camera (MoorFLPI-2, Moor Instruments, Devon, UK) was used to measure cerebral perfusion semi quantitatively in laser speckle perfusion units (LSPU). The position of the head of the animal remained unchanged throughout the experiment and the focus distance was 25 cm. Cerebral partial pressure of oxygen (P br O 2 ) was measured by an intracerebral Clark-type probe (Licox, Integra LifeSciences, New Jersey, USA) which was placed 25 mm subdurally into the white matter and secured to the craniotomy edge with bone wax. Non-invasive measurement of cerebral oxygen saturation (S br O 2 ) was obtained by near infrared spectroscopy (NIRS) (Invos 5100, Covidien/Medtronic, Minneapolis, USA). A sensor was attached to the skin of the forehead on the left side of the animal, on the contralateral side to the Licox probe and the LSCI camera, and isolated from external light.

Experimental protocol
After instrumentation the Licox probe was equilibrated for a period of 2 h or until P br O 2 > 25 mmHg, and followed by baseline data collection (PR-1-NBP and PCB-LBP) for all animals (Fig. 1). The blood pressure was lowered by caval block in the LBP group (by inflation of the balloon catheter in the vena cava) until a stable MAP of 50-60 mmHg was achieved, and an additional baseline was recorded in this group (PR-1-LBP). The caval block was maintained throughout the experiment for the LBP group, and was not subjected to further adjustments.
Vasopressor intervention followed the same sequence (norepinephrine followed by phenylephrine) in both groups and was repeated after initiation of dexmedetomidine infusion. Baseline recordings were obtained before and after each intervention or washout period (Fig. 1). The standard 30 min washout periods were conservatively chosen to allow the blood pressure and the CPO measures to return to baseline values between vasopressor interventions, and were chosen based on the longest reported clinical effect time of 15 min for phenylephrine [45]. Norepinephrine (1 mg/ mL Noradrenalin "SAD", Amgros I/S, Copenhagen, Denmark) was administered by bolus (100 µg) and followed by infusion of (0.6-2.0 µg/kg/min) to a target effect of either MAP 130-140 mmHg or 100% increase in MAP (primarily for the LBP group) from the baseline. Similarly, phenylephrine (1 mg/mL Metaoxedrin "SAD", Amgros I/S, Copenhagen, Denmark) was administered by bolus (200 µg) and followed by infusion of 5.0-13.5 µg/kg/min to a target effect of either MAP 130-140 mmHg or 100% increase (primarily for the LBP group) in MAP from the baseline. During the experiment, arterial blood samples were collected at PCB, PR-1, PR-2, PRD and PRD-end for blood gas and acid-base evaluation ( Fig. 1 and Additional file 1). All animals were euthanized with pentobarbital IV at the end of the experiment.

Statistics
Statistical analysis was performed using SPSS 24.0 software (IBM ® SPSS ® Statistics for Mac, IBM Corp. ©, Armonk, NY, USA), and Microsoft ® Excel ® for Mac 2011 version 14.3.9 (2010 Microsoft Corporation). Nonparametric statistical tests were used, since normal distribution of data could not be assumed due to the small sample size. Data were reported as medians and range (min-max), and differences between groups were analysed by the independent-samples median test. Median changes for all outcome variables were reported with 95% confidence intervals (95% CI) using Hodges-Lehmann estimates where appropriate. Primary (Licox, NIRS and LSCI) and secondary (MAP, pulse rate and EtCO 2 ) outcome variables at time points PR-1, NE-1, PR-intvas, PE-1, PRD, NE-2, PRD-intvas and PE-2 were compared using Friedman's ANOVA with post hoc pairwise comparisons. Significance levels for the four comparisons of interest (before vasopressor and during vasopressor administration) were controlled using Holm-Bonferroni's correction before reporting. Significance was set at the 5% level.
A sample size of 16 animals, divided into two groups of 8, was calculated using conservative estimates based on earlier studies [46] with expected power of 80% in detecting a minimum of 30% difference in MAP with a two-tailed significance level of 5% after supplementation of dexmedetomidine.

Results
All 16 animals completed the experimental protocol. Data from three (NBP group n = 2, LBP group n = 1) piglets were excluded from analysis. One piglet developed signs of brain oedema with a severe reduction in CPO following craniotomy (LBP group). In the NBP group, one piglet had persistently and unexplainably high pulse rate, EtCO 2 , P a CO 2 and a low pH, which were expected to produce an atypical CPO response. The other piglet was excluded due to technical difficulties.
The remaining 13 piglets reached the target MAP with vasopressor administration of > 130 mmHg or a 100% increase over pre-treatment MAP. The animals of the two groups revealed no significant demographic differences, nor were any significant differences revealed in anaesthesia time, preparation time, anaesthetic doses, baseline P br O 2 or LSPU measurements after equilibration [42]. Additionally, there was found no significant difference between the two groups in S br O 2 measured by NIRS (P = 0.59), which were 65% (range 59-72) in the LBP group and 62% (range 51-70) in the NBP group at PCB and PR-1 respectively. Both groups reached normal P br O 2 and S br O 2 levels [47,48] after equilibration.
Pulse rate did not alter significantly in either group with either vasopressor in the NBP group, but fell significantly following phenylephrine in the LBP group [P = 0.05, median decrease − 25 (95% CI − 66; 34)] (Fig. 2b).
The LBP group continued to exhibit increased variability in EtCO 2 readings compared to the NBP group, but no significant changes were observed due to vasopressor administration (Fig. 2c).

Discussion
Cerebral partial pressure of oxygen (P br O 2 ) was found to increase during vasopressor challenges when assessed by Licox, while cerebral oxygen saturation (S br O 2 ) decreased when assessed by NIRS. Cerebral perfusion (LSPU) was not found to change significantly in concert with the vasopressor induced increases in MAP. This pattern of findings was similar in both groups and during both TIVA protocols (Fig. 3c), suggesting that the cerebral autoregulation did remain intact throughout the experiment. This is consistent with the findings of Bruins et al.
The increase in P br O 2 during norepinephrine or phenylephrine has been previously reported in pigs with both uninjured brains [49] and traumatic brain injury [7]. In the present study, the significant increase in P br O 2 during vasopressor infusion was not accompanied by a concurrent increase in perfusion, suggesting that changes in P br O 2 do not simply reflect changes in CBF [50]. Decreasing S br O 2 with increasing P br O 2 was observed during vasopressor infusions, in both the propofol-remifentanil and propofol-remifentanil-dexmedetomidine TIVA groups (Fig. 3a, b). These results are consistent with findings in human NIRS studies, where a decrease in S br O 2 in response to blood pressure elevation by norepinephrine [51] or phenylephrine [6, 12, 52, 53] has been reported. Human studies of NIRS have also reported increase in S br O 2 after nitroprusside induced blood pressure decrease and decrease in S br O 2 in response to vasopressor induced increase in blood pressure [54,55]. Some authors speculated that this response in cerebral oxygen saturation was part of a normal cerebral autoregulatory response [54], while others have questioned the validity of NIRS technology and suggested that it primarily reflects skin perfusion rather than cerebral oxygenation [12,37,[56][57][58][59]. Cerebral oxygen saturation values (S br O 2 ) and cerebral partial pressure of oxygen values (P br O 2 ) are not directly comparable in absolute values, since NIRS reflects levels of oxygen-saturated haemoglobin in the venous, capillary and arterial blood [60] and Licox has been described as a measure of "the pool of oxygen" that accumulates in the brain tissue and thus reflects the balance between oxygen delivery, diffusion and consumption [50,61,62]. The distribution ratio of arteries versus veins in the cerebral cortex is approximately 30:70, and NIRS therefore predominantly reflects the cerebral venous oxygen reserve [60,63]. Transcranial assessment of the cerebral oxygen saturation in piglets will furthermore depend on factors like skull thickness and pneumatisation of the frontal sinus, since NIRS has limited penetration. Due to the size and age of the animals used in this study, the pneumatisation of the large frontal sinus, which starts at approximately 3-4 months of age in domestic swine [64], was not expected to affect the results. NIRS purportedly reflects S br O 2 in the grey matter of the cerebral cortex, whereas Licox measures P br O 2 in the less metabolically active white matter of the CNS, areas with different metabolic activity and blood flow [9, 65,66].
The differences in cerebral oxygenation assessment by Licox and NIRS could indicate that vasopressor treatment affects the cortex and the white matter differently. Whether this is a normal response attributable to a preserved cerebral autoregulation [54] or related to limitations of the methods used remains unanswered. Since cutaneous vessels predominantly have α-adrenergic innervation, treatment with vasopressors having high affinity for α-adrenergic receptors (such as norepinephrine and phenylephrine), results in vasoconstriction [67] and decreases in skin blood flow. Thus, decreases in cerebral oxygen saturation values during vasopressor infusion could therefore be a reflection of extra-cranial rather than cerebral oxygen saturation.
In summary, the CPO response to vasopressor challenge in piglets was found to be qualitatively similar for the two TIVA protocols used, and the concern regarding the potential additive vasoconstrictive effect of dexmedetomidine during vasopressor infusions could not be confirmed in this study.

Strengths and limitations
The current study was strengthened by omission of premedication on the day of experiment, by avoiding a possible vasoactive effect of premedication on CPO, and by using animals of the same sex and with a narrow age span since both age and gender may influences CBF in porcine models [68][69][70]. The results from this animal study best translates into children with neurodevelopmental maturity of approximately 10 months of age [71,72] and translation to other age groups should be made with caution [69,73]. We compared regional with focal CPO measures and at contralateral brain sites in animals undergoing craniotomy, which limits derived conclusions regarding global CPO status and regarding CPO in animals not undergoing craniotomy. Another potential limitation was the possibility that the initial norepinephrine infusion could have affected the physiological and cerebral response to the subsequent phenylephrine infusions. However, the vasopressors were given in the same sequence in all the animals, and any potential preconditioning was therefore expected to be similar in the entire group of animals. The supra clinical target increases in blood pressure and infusion doses of norepinephrine and phenylephrine in this study, were set to ensure an effect on CPO, if present, while retaining the animals within the expected blood pressure range for intact cerebral autoregulation. The doses though, were comparable to doses used in other animal studies [7,74]. This intensive approach to elevate blood pressure and subsequently CPO, when compared to having used more moderate targets and vasopressor doses, may have increased the risk for type I errors. Furthermore, the design of the study was set up to illustrate haemodynamic influences in the normal as well as in hemodynamically compromised patients without initial brain pathologies. Extrapolation to situations with concurrent brain pathology should be made with caution since neurophysiology and cerebral autoregulation may be different. The animals were subjected to prolonged anaesthesia, which for some animals lasted up to 10-11 h, and a mild to moderate hypercapnia was observed in all animals throughout the experiment [42] (Additional file 1). The resulting respiratory acidosis would make the autoregulatory plateau narrower, and thereby the cerebral perfusion more sensitive to changes in blood pressure. This was however not evident from the results of the current study, since no significant variations in cerebral perfusion could be detected. The increased levels of P a CO 2 (represented by EtCO 2 in this study) might however have increased the variability of the perfusion data, which would make the non-significant findings expected. In general, P br O 2 showed greater variability and a tendency to increase in the NBP group than in the LBP group, especially after addition of dexmedetomidine. Finally, the supplemented inspiratory oxygen levels were also relatively high in this study, with a FiO 2 of 0.8 for all animals (Additional file 1). The FiO 2 did not differ throughout the experiment, and is therefore not expected to influence the relative changes in P br O 2 of this study. It cannot be excluded that the CPO response to changes in MAP observed in this study could be different at lower FiO 2 .
Non-significant P values following Holm-Bonferroni correction were noted for several parameters despite their confidence intervals for the change in medians not including zero. The small sample size and large variability for some parameters (P br O 2 , LSPU, pulse rate) increased the risk of type II error. Confidence intervals were therefore reported for both significant and non-significant changes, permitting a more nuanced interpretation.

Conclusions
Cerebral partial pressure of oxygen measured by Licox increased significantly in concert with the vasopressor induced increases in blood pressure in healthy piglets with both normal and low blood pressure. Cerebral oxygenation assessed by intracerebral Licox and transcranial NIRS showed opposing results to vasopressor infusions. The CPO responses, induced by norepinephrine and phenylephrine, were shown to be qualitatively similar during both propofol-remifentanil and propofol-remifentanildexmedetomidine TIVA.