THAM reduces CO2-associated increase in pulmonary vascular resistance – an experimental study in lung-injured piglets

Introduction Low tidal volume (VT) ventilation is recommended in patients with acute respiratory distress syndrome (ARDS). This may increase arterial carbon dioxide tension (PaCO2), decrease pH, and augment pulmonary vascular resistance (PVR). We hypothesized that Tris(hydroxymethyl)aminomethane (THAM), a pure proton acceptor, would dampen these effects, preventing the increase in PVR. Methods A one-hit injury ARDS model was established by repeated lung lavages in 18 piglets. After ventilation with VT of 6 ml/kg to maintain normocapnia, VT was reduced to 3 ml/kg to induce hypercapnia. Six animals received THAM for 1 h, six for 3 h, and six serving as controls received no THAM. In all, the experiment continued for 6 h. The THAM dosage was calculated to normalize pH and exhibit a lasting effect. Gas exchange, pulmonary, and systemic hemodynamics were tracked. Inflammatory markers were obtained at the end of the experiment. Results In the controls, the decrease in VT from 6 to 3 ml/kg increased PaCO2 from 6.0±0.5 to 13.8±1.5 kPa and lowered pH from 7.40±0.01 to 7.12±0.06, whereas base excess (BE) remained stable at 2.7±2.3 mEq/L to 3.4±3.2 mEq/L. In the THAM groups, PaCO2 decreased and pH increased above 7.4 during the infusions. After discontinuing the infusions, PaCO2 increased above the corresponding level of the controls (15.2±1.7 kPa and 22.6±3.3 kPa for 1-h and 3-h THAM infusions, respectively). Despite a marked increase in BE (13.8±3.5 and 31.2±2.2 for 1-h and 3-h THAM infusions, respectively), pH became similar to the corresponding levels of the controls. PVR was lower in the THAM groups (at 6 h, 329±77 dyn∙s/m5 and 255±43 dyn∙s/m5 in the 1-h and 3-h groups, respectively, compared with 450±141 dyn∙s/m5 in the controls), as were pulmonary arterial pressures. Conclusions The pH in the THAM groups was similar to pH in the controls at 6 h, despite a marked increase in BE. This was due to an increase in PaCO2 after stopping the THAM infusion, possibly by intracellular release of CO2. Pulmonary arterial pressure and PVR were lower in the THAM-treated animals, indicating that THAM may be an option to reduce PVR in acute hypercapnia.


Introduction
Low tidal volume (V T ) ventilation has been shown to reduce ventilator-induced lung injury (VILI) and to improve survival in patients with acute respiratory distress syndrome (ARDS) [1]. In addition, it seems to reduce the risk of lung complications after surgery in which patients are under general anesthesia [2]. Therefore, low V T ventilation is recommended for ventilating patients with ARDS and has also gained support in anesthesia.
It has been suggested that ventilation with V T even less than 6 ml/kg would further reduce the risk of VILI in ARDS [3]. However, very low V T ventilation may increase arterial carbon dioxide tension (PaCO 2 ) substantially. Although hypercapnic acidosis has been shown to have both negative and positive effects on immune function, it has unequivocal and clinically relevant negative effects on the pulmonary circulation, increasing pulmonary vascular resistance (PVR) [4][5][6][7]. This, in combination with high positive end-expiratory pressure (PEEP) levels, may induce acute right heart failure [8,9]. Different methods to reduce PaCO 2 during low V T ventilation have been suggested, such as reduction of apparatus dead space by a tracheal double-lumen tube, tracheal gas insufflation, expiratory flushing of the dead space by a tracheal catheter, or prolonging the endinspiratory pause [10][11][12][13]. One of the most common methods is to increase minute ventilation by increasing respiratory rate (RR). However, this sometimes results in an unwanted buildup of auto-PEEP. Furthermore, it might be that the increased RR in itself, owing to increased energy transfer to the lungs, induces VILI. Indeed, very high RR combined with very low V T using high-frequency oscillation has not been shown to be beneficial and might even increase mortality in ARDS [14,15].
Other methods are extracorporeal CO 2 removal using an arteriovenous or venovenous approach with a low blood flow through a small membrane oxygenator/CO 2 remover. However, this method has not yet been shown to improve clinical outcome [16]. Thus, in many situations with low V T ventilation, hypercapnic acidosis is unavoidable.
Because the acidosis caused by the increased PaCO 2 may be the main reason for the side effects of "permissive hypercapnia," it has been speculated whether treatment with sodium bicarbonate (NaHCO 3 ) would be useful [1,17,18]. In fact, in order for NaHCO3 to work as a buffer, it has to generate CO 2 , which in turn has to be removed via the lungs increasing ventilatory demands [19]. Because the problem with low V T ventilation is that the CO 2 excretion via the lungs is impaired, NaHCO 3 is theoretically not a good choice for buffering in these conditions. However, Tris(hydroxymethyl)aminomethane (THAM), or trometamol, which is a pure proton acceptor that exerts its buffer effects without generating CO 2 , could be a more logical choice in cases of excessive CO 2 accumulation [20]. In addition, there are experimental data indicating that THAM may attenuate VILI [21]. THAM has already been used with success in cases with severe asthma and as an adjunct in ARDS [5,20,22]. However, in all published studies so far, THAM has been used at a modest dose, for a short period of time, and not with the aim of full pH correction. Using a porcine model, we have previously shown that THAM could be used to correct pH for at least 3 h during total apnea and that PVR was not severely increased during this period [23]. The aims of the present study were to explore how THAM, administered during two limited periods of very low V T ventilation, would affect, first, pH and PVR and, second, lung inflammatory parameters in a porcine lung lavage model. We hypothesized that, during hypercapnia, THAM infusion would keep pH normal by increasing the metabolic base component, would prevent increase of PVR, and would not increase lung injury.

Methods
The study was approved by the Animal Research Ethics Committee at Uppsala University, and the National Institutes of Health guidelines for animal research were followed. The study was performed at the Hedenstierna Laboratory, Uppsala University, Uppsala, Sweden.
During the first hour, 10 ml kg −1 h −1 Ringer's acetate was infused intravenously, after which the infusion rate was lowered to 5 ml kg −1 h −1 intravenously. After open dissection of the neck vessels, an arterial catheter was inserted into the right carotid artery for blood sampling and blood pressure monitoring, and a central venous catheter was inserted via the right external jugular vein. In addition, a pulmonary arterial catheter (Criti Cath No7; Ohmeda, Singapore) for measurement of cardiac output (CO) and pulmonary arterial pressure was introduced via the right external jugular vein, and its correct position was verified by pressure monitoring. CO was obtained as the mean of three values measured by thermodilution after injection of 10 ml of ice-cold saline into the central venous catheter (Siemens SC 9000XL; Dräger, Lübeck, Germany). A bladder catheter was inserted suprapubically to measure hourly urine production. Electrocardiographic monitoring was started, and peripheral capillary oxygen saturation was monitored at the base of the tail by pulse oximetry (Siemens SC 9000XL). Next, a lung recruitment maneuver was performed with I:E 1:1, RR 6 breaths/min, pressure control, inspiratory pressure 40 cmH 2 O, and PEEP 20 cmH 2 O for 60 seconds. If the animal was considered hemodynamically unstable (mean arterial blood pressure [MAP] <50 mmHg) at this point, 50-ml boluses of hydroxyethyl starch (Hesra, Baxter Medical, Kista, Sweden) were given until a MAP of at least 50 mmHg was reached. All recruitment maneuvers were performed in the manner detailed above.

Estimation of required amounts of buffer to equilibrate protons
Using data from a pilot study and previous experiments, we targeted a pH of 7.35 and a PaCO 2 of 15 kPa at the 6-h endpoint. The pH target of 7.35 was chosen because, as seen in the study by Weber et al. [5], already at pH of approximately 7.2 there was a rise in mean pulmonary arterial pressure (MPAP). Inserting these values into the Henderson-Hasselbalch equation and solving for HCO 3 − = 10 (7.2−6.1) •0.23 × 15 = 43 mEq/L, we could find the mEq/L amount of buffer needed as the difference between this value and normal bicarbonate (HCO 3 − ), which was assumed to be 20 mEq/L. The buffering capacity needed for 1 h of infusion was then estimated as weight in kilograms × 23 × 0.4 = 9.2 × weight. The recommended dosing for THAM (Addex-THAM; Fresenius Kabi, Uppsala, Sweden) is normally calculated with a coefficient of 0.3, not 0.4, but in a pilot study a 1h infusion with coefficient 0.3 was found to be an inadequate dose; therefore, the coefficient was raised.

Experimental protocol Preparations
After the instrumentation, a lung recruitment maneuver was performed to homogenize lung volume history, FiO 2 was set to 1.0, and the animal was allowed to stabilize for at least 15 minutes before baseline blood gas values were recorded; that is, arterial and venous blood was sampled for measurement of oxygen tension (PaO 2 ), PaCO 2 , mixed venous oxygen saturation, hemoglobin, lactate, sodium, chloride, potassium, calcium, glucose content, pH, and base excess (BE) (ABL800 FLEX; Radiometer, Copenhagen, Denmark), and oxygen hemoglobin saturation (OSM3; Radiometer). The following baseline hemodynamic parameters were measured: MAP, MPAP, central venous pressure (CVP), pulmonary capillary wedge pressure, and CO. Brody's formula [24] for body surface area of pigs was used for the cardiac index (CI) calculations. PEEP was set to 0 cmH 2 O, and an airway pressure-volume (PV) loop was obtained by slow insufflation to 40 cmH 2 O followed by slow exsufflation via an occluder [25], and functional residual capacity was obtained (FRC) using a sulfur hexafluoride washin/washout method [26,27].
After baseline measurements, lung injury was induced by repeated lung lavages with 30 ml/kg of 37-39°C normal saline (eight in total). Blood gases were sampled again, and MAP, MPAP, and CVP, as well as FRC and compliance of the respiratory system as obtained from the maximum slope of the expiratory PV loop, were recorded [25].
Next, the tracheal tube was replaced with a double lumen endotracheal tube with both distal openings in the trachea [10]. The inspiratory tubing was connected to one of the endotracheal tube lumens and the expiratory tubing was attached to the other lumen, separating inspiration from expiration. Thus, the apparatus dead space was eliminated. A second lung recruitment maneuver was performed, and the ventilator was set to PEEP 10 cmH 2 O, V T 6 ml/kg, and I:E 1:2, and the RR was adjusted to target a pH between 7.38 and 7.42 in two successive blood samples obtained within a 15-minute time window.

Main experiment
After confirming that pH was normal by arterial and venous blood gas sampling, hemodynamic measurements (time point 0) were registered. Hypoventilation was initiated by reducing V T to 3 ml/kg, and THAM infusion was started at the same time through a central venous line, in the two buffering groups. Arterial and mixed venous blood gases were obtained at 5, 10, 15, 30, 60, 90, and 120 minutes and then every hour up to 6 h from time point 0. In addition, the hemodynamic variables were registered at similar time points, except at 5 and 10 minutes.
After 6 h of low V T ventilation, further data were collected by drawing venous samples for cytokine analyses, measuring FRC and compliance and performing a lavage of the basal right lung with 30 ml of normal saline. Thereafter the left lung was removed and dissected to obtain tissue samples from the ventral, medial, and dorsal parts at the hilar level. The tissue samples were frozen in liquid nitrogen and stored at −80°C. The animals were killed with an intravenous dose of potassium chloride under deep anesthesia.

Statistics
A power analysis (analysis of variance [ANOVA]) indicated that six animals in each group would be enough to detect a pH difference of 0.1 with a standard deviation of 0.05 with a p <0.05 and a power of 0.8, assuming a normal distribution.
If there was a clear resemblance with the exponential distribution, a log transformation was performed to improve normality. If the variables did not pass the Shapiro-Wilk normality test, Kruskal-Wallis one-way ANOVA was performed. Tukey's honestly significant difference was assessed post hoc. The data are presented as mean and standard deviation. The statistical calculations were performed using R 3.1.1 [28] and SigmaPlot 11.0 (Systat Software, San Jose, CA, USA).

Results
All values reported in groups of three are in the order controls (NT), 1-h infusion (1T), 3-h infusion (3T) of THAM. p-Values reported are with regard to NT unless otherwise specified.

Blood gas and acid-base status
Blood gas and acid-base data are reported in Table 1 and graphed in Fig. 1.
As evident in Fig. 1, arterial pH was largely different between the groups during THAM infusion. These differences diminished over time, showing no differences between groups at 360 minutes. PaCO 2 rose in all groups. At the end of infusions, both the 1T and 3T groups showed a second phase of PaCO 2 increase, whereas NT changed very little after 120 minutes. There was a large difference in PaCO 2 at 360 minutes ( showed an increase similar to that of BE. Arterial oxygen tension (PaO 2 ) trended toward decrease and was lower at 360 minutes in the 3T group (p <0.001) compared with controls and the 1T group. The 3T group had a higher shunt fraction than the 1T group at 360 minutes (p =0.03).

Hemodynamics
Hemodynamic data are reported in Table 2 and graphed in Fig. 2.
The MPAP was lower in the 3T and 1T groups from the 30-and 60-minute time points, respectively, onward; at 360 minutes, however, only the 3T group was different from the control animals (25±5 mmHg vs 21±2 mmHg, p =0.17; 18±2 mmHg, p =0.008). PVR was similar to MPAP, and the 3T group was still different at 360 minutes (450±141 dyn•s/m 5 vs 329±77 dyn•s/m 5 , p =0.11; 255±43 dyn•s/m 5 , p =0.0081). CI exhibited a rising trend over time in all the groups, with 3T separating from the rest at 360 minutes (3.5±1.0 L/min/m 2 vs 3.8±0.75 L/ min/m 2 , p =0.42; 5.0±0.57 L/min/m 2 , p <0.001). MAP and systemic vascular resistance (SVR) did not differ between the groups at the examined time points, but SVR showed a decreasing trend over time in all three groups.

Inflammatory markers
Inflammatory marker data are depicted in Fig. 3.
Because many values were at levels too low to be detected with the ELISA, they are not presented in a table. Most of the missing (i.e., below the detection limit) values were in the 1T and 3T groups, indicating that they might be generally lower than in the controls.
In the tissue samples, there was an effect on the IL-6 concentration (detected by two-way ANOVA) in the THAM strata, and post hoc analysis showed a p-value of 0.014 compared with controls, with the 3T group being higher (11.8±11.7 pg/mg protein ) than the 1T group (4.5±2.0 pg/mg protein ) and non-significant values found for the rest (p =0.086 for 3T vs NT and p =0.423 for 1T vs NT). Two-way ANOVA of the TNF-α and IL-1β concentrations was not done, owing to many missing values. Kruskal-Wallis one-way ANOVA on ranks was performed for every tissue strata, with the missing values set below the minimum measured values. No differences were detected.

Lung mechanics
Data regarding lung mechanics are provided in Table 3.
FRC and compliance of the respiratory system decreased with the lavages (p <0.001) but did not differ between the groups.

Electrolytes
Electrolyte data are reported in Table 4.
Arterial sodium concentration fell during the THAM infusion in both the 1T and 3T groups, and arterial potassium concentration increased.

Discussion
This study shows the following results in a porcine lung lavage model: 1. Permissive hypercapnia obtained by reducing the effective V T from 6 ml/kg to 3 ml/kg decreased pH with a slow metabolic compensation during the study period and increased PVR, but it did not deteriorate CO or have other severe effects. Thus, these findings indicate that hypercapnia in this model did not compromise right heart function. 2. Buffering with THAM infused for 1 or 3 h intravenously immediately normalized pH both by reducing the increase in PaCO 2 and by a fast  Abbreviations: BE arterial base excess, HCO 3 − arterial bicarbonate concentration, NT control animals that did not receive THAM, PaCO 2 arterial carbon dioxide tension, ΔPaCO 2 arteriovenous difference in carbon dioxide tension, PaO 2 arterial oxygen tension, Q s /Q t venous admixture obtained at inspired oxygen fraction 1.0, SvO 2 mixed venous oxygen saturation, 1T group that received THAM infusion for 1 h, 3T group that received THAM infusion for 3 h, THAM Tris(hydroxymethyl) aminomethane, VO 2 body oxygen consumption The time points refer to time after start of THAM infusion or corresponding time points in the controls (NT) a Value is different from NT group b Value is different from 1T group c Only 5 subjects represented owing to technical error increase in BE. It was also associated with a decreased PVR. 3. After the end of the continuous infusion of THAM, there was a rebound increase in PaCO 2 , and, in spite of a continued high BE, pH decreased to levels similar to those of the controls at the end of the experimental period. 4. Notwithstanding the decrease in pH and marked increase in PaCO 2 , PVR remained low. 5. The inflammatory response in the lungs as well as the lung volumes and lung mechanics were not markedly different between the THAM groups and the controls.
We previously showed, in a porcine apneic model, that a continuous THAM infusion maintained pH at physiologically acceptable levels for at least 3 h. PVR was only slightly increased, despite a marked increase in PaCO 2 to a maximum of 30 kPa [23]. In that study, where we investigated whether THAM could totally replace ventilation of the lungs for a longer period, we did not discontinue the THAM infusion during the experiment. In the present study, where we studied whether THAM could be an adjunct to ventilation, pH decreased after the THAM infusion was stopped. This was due to a rebound increase in PaCO 2 that was severe after the prolonged THAM infusion. Despite this fact, PVR did not increase, MPAP remained low, and CO increased.
Most studies of the pulmonary circulatory effects of hypercapnia have been performed in models with hypoxia-induced pulmonary vasoconstriction (HPV) in either intact animals or isolated lung preparations where vasoconstriction was induced by lowering the inspired oxygenation concentration [29][30][31][32][33]. This is in contrast to our lavage model, where we kept PaO 2 high by FiO 2 1.0 and a lung recruitment maneuver followed by PEEP 10 cmH 2 O. The other important differences are that, in most of the other previous studies, hypercapnia was induced by administering CO 2 in the breathing gas and manipulations of the metabolic component were done by infusion of NaHCO 3 or HCl [17,34,35]. In the majority of these studies, researchers found that both metabolic and hypercapnic acidosis increased PVR and that alkalosis, independent of whether it was metabolic or hypercapnic, reduced or normalized pulmonary vascular tone. Thus, the pH change in itself was interpreted to be the main mechanism of this effect [29,36,37]. Our results challenge this interpretation because PVR in the THAM groups was low despite development of a moderate to severe respiratory acidosis after the THAM infusion was stopped. Instead, BE and HCO 3 − were maintained at high levels, indicating that the metabolic component might be more important than pH alone in the regulation of pulmonary vascular tone. That THAM reduces increases in MPAP caused by hypercapnia has previously been demonstrated by Weber et al. [5]. However, they did not find any significant change in PVR induced by THAM, probably owing to simultaneous administration of cardiovascular active agents. Furthermore, they did not report any remaining metabolic effect induced by THAM after the infusion was ended.
In contrast to the pulmonary circulation, we did not find any relation between arterial BE or pH and vascular resistance in the systemic circulation. However, the reduction in SVR mirrored the increase in PaCO 2 , suggesting that, in the systemic circulation, PaCO 2 per se influences the vascular tone more than BE or pH does, confirming results in previous studies [34,38]. Although the underlying mechanism of the effect on the vasculature by respiratory or metabolic acid-base changes have not been fully elucidated, it has been suggested that intracellular pH changes may regulate the voltage-gated potassium channels and that this effect is different between the pulmonary and systemic vessels [39]. Thus, a decrease in the intracellular pH dilates systemic vessels and constricts pulmonary vessels. In addition, on the basis of our present study, we cannot exclude a direct pharmacological effect of THAM on the vascular smooth muscles. In agreement with clinical studies, the decrease in SVR caused by THAM was associated with an increase in CO and mixed venous oxygen saturation [5].
The significant increase in PaCO 2 after THAM infusion was completed in the THAM-treated animals compared with the controls was unexpected and, to our knowledge, has not been described before. The two THAM dosages were calculated to increase BE to balance the increased CO 2 to keep pH normal. We assumed that PaCO 2 would level out at a new high steady state similar to the controls when the CO 2 excretion via the lungs would again equalize the CO 2 production in the tissues. In the control animals, this new, higher steady state (at PaCO 2 around 15 kPa) occurred after about 120 minutes. However, in the THAM-treated animals, PaCO 2 showed a two-phase course. During the infusion, the increase in PaCO 2 followed, but, as expected, below the increase in PaCO 2 of the controls because THAM reduces the CO 2 content in the blood by catching protons, moving the Henderson-Hasselbalch equation to the right. However, after the Value is different from 1T group infusion was stopped, PaCO 2 increased and leveled off at a higher level than in the controls. In the animals treated with THAM for 3 h, the PaCO 2 increase was substantial, from a mean of 15 to 23 kPa with no clear steady state at the end of the experiment. Three theoretical mechanisms might explain the increased PaCO 2 after the THAM infusion was stopped: 1. Reduced alveolar ventilation: However, the tidal volumes were kept constant throughout the study, and we do not have any indications in any of the groups that physiological dead space increased or that a marked increase in pulmonary shunt occurred after the end of the infusion. 2. Increased metabolism: However, body temperature was kept constant, and calculated oxygen consumption was unchanged; therefore, it is not plausible that CO 2 production increased. 3. Increased CO 2 accumulation in the tissues due to reduced pulmonary excretion during the THAM infusion: This is indicated by the significantly lower pulmonary mixed venous-arterial PaCO 2 gradient value during the THAM infusion in the 3-h THAM group. Thus, THAM might have sequestered CO 2 that thereafter accumulated in the tissues. Because the tension of CO 2 cannot be higher in the tissues than in the blood, the CO 2 must have been stored as HCO 3 − , probably in the intracellular compartment. However, THAM penetrates slowly into the intracellular compartment, and during the first hour most of the protonated THAM was probably excreted via the kidneys [19]. This can explain why we found only a minor (non-significant) increase in PaCO 2 after the infusion in the 1-h THAM infusion group, in contrast to the marked increase in the 3-h infusion group. Thus, one could speculate that the intracellular concentration of CO 2 , stored as HCO 3 − together with protonated THAM, might have been very high at the end of the infusion in the 3-h group. When the infusion was stopped, the intracellular HCO 3 − was, via the effect of carbonic anhydrase, transformed into CO 2 , which penetrated the cell membranes THAM can have important side effects, such as hypoglycemia, hyperkalemia, and hypotension [19]. In this study, although potassium levels increased, glucose levels were normal and blood pressure was maintained.
We did not find any major differences in inflammatory markers between the groups. Only IL-6 showed a significant increase in lung tissue, whereas we could not find any difference in TNF-α or IL-1β. The time with acidosis was significantly less in the THAM groups, which indicates that, when very low V T ventilation in addition to adequate PEEP level are used, the potential specific anti-inflammatory effect of hypercapnic acidosis is minor and that THAM is probably safe to administer in this regard. This notion is further strengthened by the fact that lung mechanics and FRC recovered to a similar degree after lavage in all groups. Our findings are in accord with those reported by Caples et al., who found that buffering with THAM ameliorated  ventilation-induced lung cell injuries in isolated rat lungs [21]. The possible clinical implication of our study is that THAM might be useful in cases of high PVR complicating permissive hypercapnia by reducing the pulmonary vascular tone. However, this has to be done very cautiously, particularly because the mechanism of the rebound effect of PaCO 2 is unknown. Thus, the dose and infusion time should be restricted, and PaCO 2 and potassium levels should be carefully monitored, as should PaO 2 , which might decrease owing to inhibition of HPV.
This study has the following inherent limitations of all animal models, and the results cannot be assumed to be fully valid in patients: 1. Although pigs have physiology very similar to that of as humans, their immunologic responses, as well as their response to THAM, might be different from those of humans. 2. The effect on the pulmonary vasculature might be more prominent in pigs because pigs have a strong HPV [40]. 3. We used an open lung approach, the animals were never hypoxemic, and inhibition of HPV may deteriorate oxygenation importantly in ARDS [37]. 4. We used FiO 2 1.0, which might have augmented the pulmonary vasodilatory effect associated with THAM. 5. Although THAM was infused in a large vein, its high osmolality may have caused endothelial damage. However, the osmolality is less than the commonly used concentration of hypertonic saline and amino acid solutions used for parenteral nutrition. 6. Possible toxic side effects were not addressed. 7. The number of animals used, as well as the observation period, was limited.

Conclusions
This experimental study of permissive hypercapnia in a porcine lung lavage model shows that intravenous infusion of THAM increased BE and bicarbonate concentration, normalized pH, and decreased PaCO 2 during the infusion. After a prolonged infusion, however, pH decreased to values similar to those in controls owing to a rebound PaCO 2 increase. Despite a similar low pH and a higher PaCO 2 compared with controls, the PVR remained low in the THAM group. No major signs of augmentation of lung injury by THAM were found. These findings suggest that CO 2 removal from the lungs is hampered during THAM infusion and that the metabolic component (i.e., BE, bicarbonate) has an important influence on the pulmonary vascular tone, indicating the potential use of THAM in situations with hypercapniainduced pulmonary hypertension and increased PVR.

Key messages
Increased pulmonary vascular resistance caused by respiratory acidosis was countered with THAM. After a high-dose infusion of THAM, PaCO 2 rebounded to a higher-than-expected level. THAM did not importantly affect the inflammatory response.