Acta Anaesthesiol Scand 2014; 58: 52–60 Printed in Singapore. All rights reserved

© 2013 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd ACTA ANAESTHESIOLOGICA SCANDINAVICA

doi: 10.1111/aas.12217

Pumpless extracorporeal CO2 removal restores normocapnia and is associated with less regional perfusion in experimental acute lung injury S. Kreyer1*, T. Muders1*, H. Luepschen1, C. Kricklies1, K. Linden1, R. Tolba2,3, D. Varelmann4, J. Zinserling1, C. Putensen1 and H. Wrigge5

1 Department of Anesthesiology and Intensive Care Medicine, University of Bonn, Bonn, Germany, 2House of Experimental Therapy, University of Bonn, Bonn, Germany, 3Institute for Laboratory Animal Science and Experimental Surgery, RWTH Aachen University, Aachen, Germany, 4Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, Boston, MA, USA and 5 Department of Anesthesiology and Intensive Care Medicine, University of Leipzig, Leipzig, Germany

Background: Lung protective ventilation may lead to hypoventilation with subsequent hypercapnic acidosis (HA). If HA cannot be tolerated or occurs despite increasing respiratory rate or buffering, extracorporeal CO2-removal using a percutaneous extracorporeal lung assist (pECLA) is an option. We hypothesised that compensation of HA using pECLA impairs regional perfusion. To test this hypothesis we determined organ blood flows in a lung-injury model with combined hypercapnic and metabolic acidosis. Methods: After induction of lung injury using hydrochloric acid (HCl) aspiration and metabolic acidosis by intravenous HCl infusion in nine pigs, an arterial-venous pECLA device was inserted. In randomised order, four treatments were tested: pECLA shunt (1) with and (2) without HA, and clamped pECLA shunt (3) with and (4) without HA. Regional blood flows were measured with the coloured microsphere technique. Results: HA resulted in higher perfusion in adrenal glands, spleen and parts of splanchnic area (P < 0.05) compared with

normocapnia. During CO2-removal with pECLA, regional perfusion decreased to levels comparable with those without pECLA and normocapnia. Cardiac output (CO) increased during HA without a pECLA shunt and was highest during HA with a pECLA shunt compared with normocapnia. During CO2-removal with pECLA, this variable decreased but stayed higher than during normocapnia with clamped pECLA shunt (P < 0.05). Conclusion: In our lung-injury model, HA was associated with increased systemic and regional blood flow in several organs. pECLA provides effective CO2 removal, requiring a higher CO for perfusion of the pECLA device without improvement of regional organ perfusion.

M

Protective mechanical ventilation using low tidal volumes (VT) in patients with ARDS aims at minimising lung strain4–6 and improving survival rate7 but may lead to hypoventilation with subsequent hypercapnic acidosis (HA)8,9 which can be aggravated by metabolic acidosis because of organ failure.10 Extracorporeal CO2 removal can be an alternative if either acidosis is severe, e.g., if HA is combined with metabolic acidosis despite optimisation of alveolar ventilation, or to further decrease VT below 6 ml/kg in severe ARDS,11 or in patients intolerant to HA, e.g., because of increased intracranial pressure or pulmonary arterial hypertension. The percutaneous extracorporeal lung assist (pECLA) is a

ultiple organ failure is the main cause of mortality in acute respiratory distress syndrome (ARDS).1,2 Mechanical ventilation to improve gas exchange may aggravate pulmonary inflammation, thereby initiating and propagating systemic inflammatory responses and consequentially, increasing the risk of organ failure.3 In addition, mechanical ventilation associated deterioration in cardiac output (CO) and regional blood flow may contribute to organ dysfunction.

*Authors contributed equally to this paper. Part of these data has been presented in form of abstracts at the ATS International Conference 2007 (San Francisco) and 2008 (Toronto), and at the ESICM annual congress 2007 (Berlin).

52 bs_bs_banner

Accepted for publication 17 September 2013 © 2013 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

pECLA and organ perfusion in ARDS

pumpless, arteriovenous shunt-driven device for CO2 removal via a membrane oxygenator.12 Despite the established effectiveness of pECLA in reversing HA, the effects of pECLA on regional blood flow in various organs remain ambiguous. In addition, the physiological regulation of organ perfusion may be impaired by HA and by systemic inflammation because of acute lung injury. A reduction in VT and hence, intrathoracic pressure, has been found to increase CO and oxygen delivery (DO2).13 HA induces an increase in sympathetic activity that may simultaneously augment CO and blood flow in some organs.14,15 Conversely, arteriovenous perfusion of the pECLA device reduces organ blood flow when CO does not increase or when organ blood flow is not equally redistributed. Distinct direct and indirect effects of pECLA on the distribution of systemic blood flow to different organs in acute lung injury have never been studied and require use of a large animal model. We hypothesised that in a combined model of respiratory and metabolic acidosis, reversal of HA by the pECLA impairs specific organ perfusion. To test this hypothesis, we determined organ blood flow with the coloured microspheres technique in a porcine lung-injury model of combined hypercapnic and metabolic acidosis.

Material and methods Animal care and use All experiments were performed in accordance with German legislation governing animal studies and the Guide for the Care and Use of Laboratory Animals, 7th edition, 1996, National Academy Press, 125 p, ISBN 0-309-05377-3. The study was reviewed by a governmental ethical committee and official permission was granted from the governmental animal care office (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Recklinghausen, Germany). Fifteen German domestic pigs (‘Landrasse’ country breed) from a disease-free barrier breeding facility at the University of Bonn were used for studies. Three pigs died during instrumentation or before conclusion of the study, an additional three pigs were excluded for inconsistent data. This resulted in nine animals (mean body weight 41.2 ± 5.3 kg) included into the analysis. Pigs were fasted for 24 h prior to experimentation while having free access to water. Before instrumentation, the animals were premedicated with intramuscular xylazinhydro-

chlorid (2 mg/kg), Esketaminhydrochlorid (15 mg/ kg) and Atropin (0.5 mg), and placed supine on a heating pad to maintain core temperature at 38°C. Anaesthesia was induced by the ear vein with intravenous midazolam (bolus of 15 mg followed by infusion of 0.2 mg/kg/h), fentanyl-citrat (bolus of 250 μg followed by infusion 10 μg/kg/h) and s-ketaminehydrochlorid 6 mg/kg/h. To prevent spontaneous breathing, pigs received pancuroniumbromid (bolus of 4 mg followed by infusion of 0.2 mg/kg/h). To ensure adequate hydration, 500 ml of Ringer solution was rapidly infused followed by a continuous infusion of 250 ml/kg/h. Animals were intubated and ventilated in a volume-controlled mode (Evita 4, Dräger Medical GmbH. Lübeck, Germany) with a VT of 8 ml/kg and a respiratory rate (RR) to achieve normocapnia (NC) throughout preparation.

Instrumentation A detailed description of instrumentation, blood gas analysis, ventilatory and cardiovascular measurements is presented in a recently published work.16 Despite a similar study protocol, animals from the present study did not participate in the previous work, and all data were separated between the two studies. For harvesting of reference blood, an 8 Fr. double-lumen catheter (Arrow Deutschland GmbH, Erding, Germany) advanced through the left femoral artery into the abdominal aorta. A pECLA was connected between an arterial 15 Fr catheter and a venous 17 Fr catheter (Novalung®, Hechingen, Germany) placed in right femoral vessels. Before connection, the pECLA was flushed with saline solution and 5000 IU of heparin followed by a continuous infusion of 2500 IU/h. Correct positioning of all catheters was verified by necropsy at the end of each experiment.

Tissue blood flow measurements The coloured microsphere technique was applied to measure regional blood flow as previously described.17 The amount of polystyrene microspheres (Dye Track; Triton Technology, San Diego, CA, USA) injected into the left cardiac ventricle and trapped in organs correlates with regional perfusion. Coloured microspheres in each tissue and blood sample were quantified by spectrophotometric analysis (Tecan, Safire,2 Männedorf, Switzerland), and tissue blood flow was calculated as mean of multiple samples from the same organ.18 Adequate mixing and distribution of injected microspheres to examined organs was guaranteed by

53

S. Kreyer et al. Table 1 Target pH, target PaCO2 and target VT during the different settings in the experiment. Setting 1 Setting 2 Setting 3 Setting 4 NC/no shunt HA/no shunt HA/shunt NC/shunt Target pH 7.2 Target BE −10 PaCO2 (torr) 40 VT (ml/kg) 8

7.0 −10 100 6

7.0 −10 100 6

7.2 −10 40 6

After preparation sequence of modes was randomised. Target pH and target BE were achieved with intravenous HCl infusion, target PaCO2 was achieved with respiratory frequency and decarboxylation via pECLA. Target VT was calculated. BE, base excess; PaCO2, arterial carbon dioxide tension; VT, tidal volume; NC, normocapnia; HA, hypercapnic acidosis; pECLA, percutaneous extracorporeal lung assist.

highly significant correlations between the trapped number of microspheres in the two reference blood samples [28,060.1 ± 11,607.8 vs 28,897.1 ± 11,547.5, r2 = 0.77 (Bland–Altman: Bias: −838; 95% limits of agreement from −12,145 to 10,469)] and the blood flow to the right and left adrenal gland [2.79 ± 1.73 ml/g/min vs 2.75 ± 1.84 ml/g/min, r2 = 0.9 (Bland–Altman: Bias: −0.0339; 95% limits of agreement from −1.100 to 1.168)]. For statistic analyses, Statistica 6.0 (StatSoft, Tulsa, OK, USA) and Graph Pad Prism® 5.0 (GraphPad Software, Inc., La Jolla, CA, USA) was used.

Experimental protocol Lung injury was induced by intratracheal injection of 3 ml/kg body weight 0.1 M hydrochloric acid (HCl) followed by repetitive administrations of 1 ml/kg body weight necessary to achieve an arterial oxygen tension/inspired oxygen fraction (PaO2/FiO2) below 300 after 60 min of stabilisation. In addition, pigs received 0.2 M HCl intravenously until the base excess amounted to −10. The sequence of experimental parameters for the animals was randomised using sealed envelopes. Setting 1: clamped pECLA shunt, normoventilation (NC/no shunt) Setting 2: clamped pECLA shunt, hypoventilation (HA/no shunt) Setting 3: open pECLA shunt, hypoventilation, no CO2 elimination, no gas flow (HA/shunt) Setting 4: open pECLA shunt, hypoventilation, CO2 elimination with 12 l/min sweep gas flow (NC/ shunt) Targets for pH, BE, VT and resulting arterial carbon dioxide tension are shown in Table 1.

54

Before randomisation, ventilation parameters were defined for treatment 1 and treatments 2–4. During setting 1, upper airway pressure was set to achieve a VT of 8 ml/kg, and RR was set to achieve NC. During settings 2–4, VT was set to 6 ml/kg, and RR was set to achieve hypercapnia. Positive endexspiratory pressure (PEEP) was set to 5 cm H2O and the inspiratory to expiratory ratio to 1 : 1. After 30 min of equilibration following each change in treatment, ventilator, gas exchange, regional blood flow and hemodynamic data were recorded for 10 min.

Statistical analysis

Data are expressed as mean ± standard deviation. A normal distribution was confirmed with a Shapiro– Wilks W test. To verify adequate mixing and even distribution of microspheres and blood flow to various organs, linear regressions were calculated comparing numbers of microspheres in two reference blood samples as well as right and left adrenal glands. Ventilator, lung mechanic, gas exchange, organ blood flow, and systemic hemodynamic data obtained during the different treatments were compared using one-way analysis of variance tests. When a significant F ratio was obtained, differences between the means were analysed by post-hoc tests (Newman–Keuls). Differences were considered statistically significant if P was less than 0.05. For statistic analyses, Statistica 6.0 was used.

Results Regional perfusion Blood flow data of selected organs are given in Figs 1 and 2, and Table 2. HA resulted in significantly higher perfusion in the adrenal glands, kidney marrow, ileum, and jejunum (P < 0.05). This fact was enforced by the artificial shunt, resulting in significantly higher perfusion in spleen, kidney cortex and colon too. During pECLA treatment with CO2 removal (NC/shunt), regional perfusion decreased to levels comparable with NC/no shunt.

CO Ventilatory and hemodynamic variables are shown in Tables 3 and 4. CO and DO2 were lowest during the NC/no shunt, increased during the HA/no shunt, and were highest during the HA/shunt. During CO2 removal with pECLA, these variables decreased but remained elevated compared with NC/no shunt (P < 0.05).

pECLA and organ perfusion in ARDS

Ventilator parameters Reduction of ventilation resulted in a significant lower VT, PAW mean, VE, RR and PAW max levels during all treatments compared with the NC/no shunt. There were no statistical changes between the HA/no shunt, the HA/shunt and the NC/shunt.

pECLA flow Fig. 1. Regional perfusion of adrenal glands, spleen and kidney cortex in ml × g/wet tissue × /min during the different settings in the experiment. Regional organ blood flow in the four modes normocapnia/no shunt, hypercapnic acidosis/no shunt, hypercapnic acidosis/shunt, normocapnia/shunt. Randomised order of ventilatory settings. Organ blood flow is shown in ml × g/wet tissue × / min. Values are mean ± standard deviation, repeated measurement analysis of variance; post-hoc: Newman–Keuls test. *P < 0.05 compared with NC/no shunt; #P < 0.05 compared with HA/no shunt; $P < 0.05 compared with HA/shunt. NC, normocapnia; HA, hypercapnic acidosis.

Blood flow through the pECLA and the shunt fractions was 30.5 ± 5.7 ml/kg/min for the HA/shunt (shunt fraction 19 ± 4%) and 30.4 ± 6.7 ml/kg/min for the NC/shunt (shunt fraction 23 ± 7%). As mentioned earlier, we excluded three animals from our study. One animal was severe hypoxemic (SaO2 under 88%). One animal had unrealistic high regional blood flow in several organs during setting 4: open pECLA shunt, hypoventilation, CO2elimination with 12 l/min sweep gas flow (NC/ shunt). Regional perfusion was 5–10 times higher compared with other settings. In one animal, we were unable to retrieve arterial blood samples after injection of the microspheres and had therefore to exclude the animal.

Discussion

Fig. 2. Regional perfusion of kidney marrow, ileum and jejunum in ml × g/wet tissue × /min during the different settings in the experiment. Regional organ blood flow in the four modes normocapnia/no shunt, hypercapnic acidosis/no shunt, hypercapnic acidosis/shunt, normocapnia/shunt. Randomised order of ventilatory settings. Organ blood flow is shown in ml × g/wet tissue × / min. Values are mean ± standard deviation, repeated measurement analysis of variance; post-hoc: Newman–Keuls test. *P < 0.05 compared with NC/no shunt; #P < 0.05 compared with HA/no shunt; $P < 0.05 compared with HA/shunt. NC, normocapnia; HA, hypercapnic acidosis.

Hemodynamic parameters Pulmonary arterial occlusion pressure, global enddiastolic volume (GEDV) and central venous pressure remained constant in all four treatments, while mean arterial pressure was significantly lower during the NC/shunt compared with the NC/no shunt (P < 0.05). Arterial haemoglobin was lower during NC/shunt compared with the HC/no shunt (P < 0.05). Arterial oxygen saturation (SaO2) decreased in both hypercapnic phases compared with NC (P < 0.05).

This experimental study showed that during extracorporeal CO2 removal by pECLA, regional blood flow was comparable with normocapnic control conditions. Although CO with pECLA remained as high as during HA, the higher systemic perfusion was obviously required to compensate for the arteriovenous shunt associated with pECLA since our examined regional perfusion did not improve. Furthermore, HA resulted in higher systemic perfusion and in a significant increase in splanchnic perfusion, as well as improvement in perfusion of other organs.

Regional perfusion of adrenal glands and spleen In our porcine model, we found significantly higher regional perfusion during HA in the adrenal glands and spleen (significant only during HA/shunt). Parallel to these increases, we observed significantly higher CO associated with HA. The elevation of CO caused by HA has been reported in previous studies and has been mainly attributed to sympathetic activation and catecholamine release.15,19,20 Although we did not measure catecholamine levels, sympathetic activation by HA is indirectly supported by the observed higher perfusion of the adrenal glands (production of catecholamines) and spleen (release of erythrocytes21) in the state of HA.

55

S. Kreyer et al. Table 2 Regional perfusion in ml × g/wet tissue × /min during the different settings in the experiment. Colon Duodenum Liver Pancreas Heart right Heart left

NC/no shunt

HA/no shunt

HA/shunt

NC/shunt

0.36 ± 0.16 0.33 ± 0.16 0.38 ± 0.19 0.33 ± 0.29 0.86 ± 0.32 1.23 ± 0.39

0.45 ± 0.20 0.46 ± 0.39 0.48 ± 0.33 0.28 ± 0.19 1.32 ± 1.26 1.40 ± 0.78

0.58 ± 0.30* 0.55 ± 0.32 0.58 ± 0.26 0.34 ± 0.16 1.52 ± 1.08 1.72 ± 0.67

0.43 ± 0.15 0.35 ± 0.08 0.40 ± 0.20 0.37 ± 0.21 1.39 ± 0.95 1.53 ± 0.68

*P < 0.05 compared with NC/no shunt. Regional organ blood flow in the four settings: normocapnia/no shunt, hypercapnic acidosis/no shunt, hypercapnic acidosis/shunt, normocapnia/shunt. Randomised order of ventilatory settings. Organ blood flow is shown in ml × g/wet tissue × /min. Values are mean ± standard deviation, repeated measurement analysis of variance, post-hoc: Newman–Keuls test. NC, normocapnia; HA, hypercapnic acidosis.

Table 3 Cardiovascular parameters during the different settings in the experiment. HR (/min) MAP (mmHg) MPAP (mmHg) PAOP (mmHg) GEDV (ml/kg) CO/kg (ml/kg/min) SV/kg (ml/beat/kg) SVR (dyne.s/cm5) CVP (mmHg) DO2 (ml/min) caO2 (ml/dl) cvO2 (ml/dl)

NC/no shunt

HA/no shunt

HA/shunt

NC/shunt

87.8 ± 25.2 115.9 ± 19.5 30.5 ± 7.1 6.6 ± 4.2 818.2 ± 99 101 ± 23.6 1.2 ± 0.3 2133.6 ± 579.8 6.8 ± 4.2 626.2 ± 158.6 14.8 ± 1.4 8.4 ± 1.8

104.3 ± 19.7 107.7 ± 17 32.5 ± 6.1 6.1 ± 4.6 841.9 ± 143.4 140.1 ± 32.2* 1.4 ± 0.3* 1439.8 ± 363.2* 6 ± 3.5 857.5 ± 300.6* 14.2 ± 1.4 9.5 ± 2.1

117.8 ± 28.7* 105.2 ± 14.2 34.3 ± 6.1 7.1 ± 4 817.7 ± 86.7 165.6 ± 18.2*† 1.5 ± 0.3* 1156.2 ± 238.2* 6.2 ± 4.1 931.8 ± 180.6* 13.3 ± 1* 9.7 ± 1.3

114.6 ± 28.6* 98.8 ± 14.5* 31.9 ± 5.2 5.3 ± 4.5 754.2 ± 102.2 138.9 ± 28.8*‡ 1.3 ± 0.3‡ 1298.3 ± 255.3* 5.8 ± 4.4 774.7 ± 129.2* 13.5 ± 1.8* 9.8 ± 2

*P < 0.05 compared with NC/no shunt. †P < 0.05 compared with HA/no shunt. ‡P < 0.05 compared with HA/shunt. Systemic parameters in the four settings normocapnia/no shunt, hypercapnic acidosis/no shunt, hypercapnic acidosis/shunt, normocapnia/shunt. Randomised order of ventilatory settings. Values are mean ± standard deviation, repeated measurement analysis of variance; post-hoc: Newman–Keuls test. NC, normocapnia; HA, hypercapnic acidosis; HR, heart rate; MAP, mean arterial pressure; MPAP, mean pulmonary arterial pressure; PAOP, pulmonary arterial occlusion pressure; GEDV, global end-diastolic volume; CO, cardiac output; SV, stroke volume; SVR, systemic vascular resistance; CVP, central venous pressure; DO2, oxygen delivery; caO2, arterial oxygen content; cvO2, pulmonary arterial oxygen content.

During CO2-elimination by the pECLA, regional perfusion of nearly every studied organ decreased to levels comparable with the NC/no shunt treatment, while CO remained significantly higher. We can therefore conclude that a CO elevation is needed to compensate for the shunt, and that regional perfusion does not benefit from the higher CO (Fig. 1).

Regional perfusion of splanchnic organs The effect of HA on splanchnic perfusion is ambiguous. While some studies found higher splanchnic perfusion during HA,22,23 others could not.24 Patients from Kiefer et al. showed only mild CO2 increases,24 while in our model, the enforced combined meta-

56

bolic and HA resulted in statistically significantly higher regional perfusion in the ileum, jejunum and kidney marrow.

Regional perfusion affected by pECLA The effects of extracorporeal CO2 removal on regional perfusion has been examined by Brunston et al. in which organ blood flow distribution was only mildly affected and the reduction in flow was proportional to the degree of arteriovenous shunt.25 At a pECLA-associated shunt fraction of 20%, which is comparable with our pECLA-associated shunt, the decrease in regional perfusion was 10–20% depending on the organ system. However, these

pECLA and organ perfusion in ARDS Table 4 Ventilatory parameters during the different settings in the experiment. Hb (g/dl) SaO2 (%) SvO2 (%) PaCO2 (torr) pHarterial BEarterial (mM) VT/kg−1 (ml/kg) RR (/min) VE (l/min) PAW mean (cm H2O) PAW max (cm H2O) PaO2/FiO2 (torr)

NC/no shunt

HA/no shunt

HA/shunt

NC/shunt

11.2 ± 1.3 95 ± 5.9 54.3 ± 10.6 44.9 ± 5.3 7.2 ± 0.04 −10.2 ± 1.6 7.8 ± 0.6 31.3 ± 4.5 10.2 ± 2 13.1 ± 4.3 29 ± 9.5 184.1 ± 76.8

11.8 ± 1.3 86.7 ± 4.5* 56.9 ± 7.4 102.6 ± 13.8* 6.97 ± 0.03* −11.4 ± 1.6 5.8 ± 0.6* 11 ± 3.1* 2.6 ± 0.8* 11.2 ± 2.8* 23.8 ± 8.1* 135 ± 35.9*

11 ± 1 86.7 ± 4.3* 62 ± 5.8* 106.3 ± 7.7* 6.95 ± 0.03* −11.6 ± 1.5 5.5 ± 0.9* 10.7 ± 3.3* 2.4 ± 0.9* 11.7 ± 3.5* 23.9 ± 8.7* 132.7 ± 31.8*

10.2 ± 1.4† 95.1 ± 3.2†‡ 68.3 ± 7.3*†‡ 43.9 ± 4.1†‡ 7.23 ± 0.03†‡ −9 ± 1.4†‡ 5.7 ± 1.2* 11.3 ± 3.7* 2.7 ± 1.3* 12 ± 3.7* 25.3 ± 10* 152.4 ± 52.7*

*P < 0.05 compared with NC/no shunt. †P < 0.05 compared with HA/no shunt. ‡P < 0.05 compared with HA/shunt. Systemic and ventilator parameters in the four settings normocapnia/no shunt, hypercapnic acidosis/no shunt, hypercapnic acidosis/ shunt, normocapnia/shunt. Randomised order of ventilatory settings. Values are mean ± standard deviation, repeated measurement analysis of variance; post-hoc: Newman–Keuls test. NC, normocapnia; HA, hypercapnic acidosis; Hb, arterial haemoglobin; SaO2, arterial oxygen saturation; SvO2, pulmonary oxygen saturation; PaO2, arterial oxygen tension; FiO2, inspired oxygen fraction; PaCO2, arterial carbon dioxide tension; pHarterial, arterial pH; BE, base excess; VT, tidal volume; RR, respiratory rate; VE, minute ventilation; PAW mean, mean airway pressure; PAW max, maximum airway pressure.

results are hardly comparable with our findings as Brunston et al. used healthy, spontaneously breathing sheep, whereas here, we used controlled mechanically ventilated pigs in experimental lung injury. CO2 levels were constant, as the sheep reduced spontaneous breathing during extracorporeal CO2 removal.25 During NC, we did not observe any significant differences in regional perfusion associated with the pECLA-associated shunt. In a previous work, we could show the effect of pECLA on regional perfusion of the brain.16 Using a pECLA device enables effective CO2 removal in a lung injury model without metabolic acidification but requires a higher CO for the perfusion of the pECLA device with no improvement of regional cerebral blood flow.16

Effect of pECLA on CO2 and CO By using the pECLA device, even in extreme conditions of combined metabolic and HA, an effective CO2 removal is possible. Analogous with previous findings,12,26 we were able to eliminate CO2 in experimental acute lung injury to levels of NC while using a VT in the range of 6 ml/kg. Furthermore, in accordance with previous reports,27 our data show that CO is elevated due to HA and an additional elevation occurred under the pECLA-associated shunt.12 Higher CO with the pECLA is associated with an artificial shunt and

does not result in higher regional perfusion. The increased myocardial workload imposed by the pECLA-induced shunt is potentially harmful in patients with compromised heart function, who are unable to meet the necessary increase in CO. Although a pressure gradient is the driving force for CO2 removal through pECLA,28 reduced CO leads to additional reductions in shunt flow through the pECLA, impairing its function.12,29 Interestingly, in our animal model, hypercapnia and the pECLAshunt increased CO despite severe acidosis with a pH of 7.0. Our data from the adrenal glands suggest that this is an effect of sympathoadrenergical activation because of hypercapnia. This mechanism might be impaired in patients with sepsis and hemodynamic instability. The exact cause of elevation in CO can unfortunately not be explained by our results because this study was not designed to measure catecholamines or inflammation.

Clinical implications In therapy of ARDS, mechanical ventilation plays a major role in achieving adequate oxygenation and decarboxylation. In some patients, conventional ventilation must be very invasive to achieve adequate gas exchange, resulting in side effects of mechanical ventilation. Using a pECLA allows adequate CO2 elimination in patients where conventional ventilation fails. Recent data may also suggest

57

S. Kreyer et al.

that further reducing VT to about 3 ml/kg facilitated by pECLA use may shorten time on mechanical ventilation in patients with more severe ARDS.11 We show here that the pECLA device is effective in reducing CO2 while keeping airway pressures low. This phenomenon is also occurring in pigs where a combination of hypercapnic and metabolic acidosis was present, a common clinical finding in critical ill patients. In our opinion, the use of a lung-injury model is, in this study, superior to just changing PaCO2 because the pECLA was designed to be used in the state of lung injury, allowing a reduction in invasiveness of mechanical ventilation. In contrast, because of the shunt, we measured a higher CO without higher end organ perfusion here. Physicians should be aware of this artificial elevated cardiac workload and the possible benefits of establishing a pECLA to reduce CO2 must be counterbalanced with potential negative consequences. If the use of a veno-venous extracorporeal CO2 removal device has similar effects on regional organ perfusion with less influence on systemic parameters must be target of additional studies.

Methodology Direct measurement of microperfusion in splanchnic organs in critically ill patients is very difficult. Therefore, we used a porcine model of ARDS and the coloured microsphere method for blood flow measurements,18,30 which we have established previously.16,17 Using this method makes quantification of microcirculation even in different regions in organs possible. With this method, we could avoid surgical stress for the animal, such as implantation of blood flow sensors.17 For the same reason, we avoided a thoracotomy, usually necessary for placement of a catheter in the left atrium, by advancing a catheter via a carotid artery back into the left ventricle. A prerequisite for exact measurement is ideal mixing of microspheres before and after injection. The latter is usually ensured by turbulent flow, especially over the atrial-ventricular valve. In larger animals, however, using the left ventricle as injection site, as was done here, there was no significant difference microsphere mixing compared with injection into the left atrium.31 In the present study, adequate mixing of microspheres was verified by a high correlation between the microspheres trapped in the reference blood samples obtained from different sites of the aorta as well as the two adrenal glands. Because of the arteriovenous shunt, the pECLA system is more effective in reducing CO2 than in oxygenation of the blood.32 In case of hypoxemia, a

58

small amount of oxygenation will occur. In our examination, the pulmonary arterial saturation was higher when the pECLA with sweep gas flow was used. In contrast, arterial (caO2) and pulmonary arterial oxygen content (cvO2) did not differ or were lower under shunt conditions. Significant changes in DO2 were finally dependent on CO.

Limitations Because this device is designed to be used in lung injury, we chose an acid aspiration model because this is a clinically relevant mechanism of lung injury and has no tendency to spontaneous improvement of lung function. The amount of inflammation associated with experimental lung injury might depend on the model and the species. We did not measure the inflammatory response and cannot estimate its effect on vasoregulation here. This study was not designed to examine the effect of different levels of acidosis or different PaCO2 on the regional perfusion but the effect of the pECLA (with subsequent changes in PaCO2 and pH). The effect of acidosis metabolic or respiratory achieved has to be researched in additional studies. HA was achieved by adjusting the RR but also by reducing VT by about 2 ml/kg. This was associated with slightly lower intrathoracic pressures that could have improved venous return and CO33,34 resulting in a higher microperfusion as well.17 In our study, however, preload measures such as GEDV did not change significantly, suggesting that a direct hemodynamic effect caused by the change in ventilator settings was negligible. In this study, we measured organ perfusion rather than function. We cannot generally conclude that improved organ perfusion will result in improved organ function, thereby reducing the incidence of multiple organ failure in ARDS. This was a pig model of ARDS, and physiological terms in humans may be different. Therefore, one has to be careful to extrapolate our results to humans.

Conclusion HA increased systemic and regional blood flow in several organs in our porcine model of ARDS. An effective CO2 removal by using a pECLA device is possible but requires a higher CO for perfusion of the pECLA device. This higher CO did not improve regional organ perfusion. Thus, the increase of regional perfusion induced by HA is entirely reversed when using pECLA for CO2 removal.

pECLA and organ perfusion in ARDS

Acknowledgement This study was supported by an unrestricted grant from Novalung, Hechingen, Germany. Novalung also provided the pECLA-Device (iLA®). Conflicts of interest: Authors declare no conflicts of interest.

References 1. Bersten AD, Edibam C, Hunt T, Moran J. Incidence and mortality of acute lung injury and the acute respiratory distress syndrome in three Australian States. Am J Respir Crit Care Med 2002; 165: 443–8. 2. Montgomery AB, Stager MA, Carrico CJ, Hudson LD. Causes of mortality in patients with the adult respiratory distress syndrome. Am Rev Respir Dis 1985; 132: 485–9. 3. Ricard J-D, Dreyfuss D, Saumon G. Ventilator-induced lung injury. Eur Respir J 2003; 22 (Suppl. 42): 2s–9s. 4. The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000; 342: 1301–8. 5. Petrucci N, Iacovelli W. Ventilation with lower tidal volumes versus traditional tidal volumes in adults for acute lung injury and acute respiratory distress syndrome. Cochrane Database Syst Rev 2004; CD003844. 6. Petrucci N, Iacovelli W. Ventilation with smaller tidal volumes: a quantitative systematic review of randomized controlled trials. Anesth Analg 2004; 99: 193–200. 7. Kregenow DA, Rubenfeld GD, Hudson LD, Swenson ER. Hypercapnic acidosis and mortality in acute lung injury*. Crit Care Med 2006; 34: 1–7. 8. Laffey JG, Kavanagh BP. Carbon dioxide and the critically ill – too little of a good thing? Lancet 1999; 354: 1283–6. 9. Ni Chonghaile M, Higgins B, Laffey JG. Permissive hypercapnia: role in protective lung ventilatory strategies. Curr Opin Crit Care 2005; 11: 56–62. 10. Fidkowski C, Helstrom J. Diagnosing metabolic acidosis in the critically ill: bridging the anion gap, Stewart, and base excess methods. Can J Anaesth 2009; 56: 247–56. 11. Bein T, Weber-Carstens S, Goldmann A, Müller T, Staudinger T, Brederlau J, Muellenbach R, Dembinski R, Graf B, Wewalka M, Philipp A, Wernecke K-D, Lubnow M, Slutsky AS. Lower tidal volume strategy (≈ 3 ml/kg) combined with extracorporeal CO2 removal versus ‘conventional’ protective ventilation (6 ml/kg) in severe ARDS: the prospective randomized Xtravent-study. Intensive Care Med 2013; 39: 847–56. 12. Bein T, Weber F, Philipp A, Prasser C, Pfeifer M, Schmid F-X, Butz B, Birnbaum D, Taeger K, Schlitt HJ. A new pumpless extracorporeal interventional lung assist in critical hypoxemia/hypercapnia. Crit Care Med 2006; 34: 1372– 7. 13. Downs JB, Douglas ME, Sanfelippo PM, Stanford W, Hodges MR. Ventilatory pattern, intrapleural pressure, and cardiac output. Anesth Analg 1977; 56: 88–96. 14. Cardenas VJ, Zwischenberger JB, Tao W, Nguyen PD, Schroeder T, Traber LD, Traber DL, Bidani A. Correction of blood pH attenuates changes in hemodynamics and organ blood flow during permissive hypercapnia. Crit Care Med 1996; 24: 827–34. 15. Hickling KG, Joyce C. Permissive hypercapnia in ARDS and its effect on tissue oxygenation. Acta Anaesthesiol Scand Suppl 1995; 107: 201–8. 16. Kreyer S, Muders T, Luepschen H, Kricklies C, Linden K, Soehle M, Zinserling J, Putensen C, Wrigge H. The effect of

17.

18.

19. 20. 21.

22. 23. 24. 25. 26.

27. 28.

29. 30. 31.

32.

33.

34.

pumpless extracorporeal CO2 removal on regional perfusion of the brain in experimental acute lung injury. J Neurosurg Anesthesiol 2013; 25: 324–9. Hering R, Viehöfer A, Zinserling J, Wrigge H, Kreyer S, Berg A, Minor T, Putensen C. Effects of spontaneous breathing during airway pressure release ventilation on intestinal blood flow in experimental lung injury. Anesthesiology 2003; 114: 1137–44. Kowallik P, Schulz R, Guth BD, Schade A, Paffhausen W, Gross R, Heusch G. Measurement of regional myocardial blood flow with multiple colored microspheres. Circulation 1991; 83: 974–82. Akça O. Optimizing the intraoperative management of carbon dioxide concentration. Curr Opin Anaesthesiol 2006; 19: 19–25. Cullen DJ, Eger EI. Cardiovascular effects of carbon dioxide in man. Anesthesiology 1974; 41: 345–9. Hartwig H, Hartwig HG. Structural characteristics of the mammalian spleen indicating storage and release of red blood cells. Aspects of evolutionary and environmental demands. Experientia 1985; 41: 159–63. McGinn FP, Mendel D, Perry PM. The effects of alteration of CO2 and pH on intestinal blood flow in the cat. J Physiol 1967; 192: 669–80. Semb BK, Hysing E, Mørkrid L. Effect of CO2 on peripheral flow and central hemodynamics. Eur Surg Res 1984; 16 (Suppl. 2): 133–9. Kiefer P, Nunes S, Kosonen P, Takala J. Effect of an acute increase in PCO2 on splanchnic perfusion and metabolism. Intensive Care Med 2001; 27: 775–8. Brunston RL, Tao W, Bidani A, Traber DL, Zwischenberger JB. Organ blood flow during arteriovenous carbon dioxide removal. ASAIO J 1997; 43: M821–4. Zimmermann M, Bein T, Arlt M, Philipp A, Rupprecht L, Mueller T, Lubnow M, Graf BM, Schlitt HJ. Pumpless extracorporeal interventional lung assist in patients with acute respiratory distress syndrome: a prospective pilot study. Crit Care 2009; 13: 1–7. Akça O, Doufas AG, Morioka N, Iscoe S, Fisher J, Sessler DI. Hypercapnia improves tissue oxygenation. Anesthesiology 2002; 97: 801–6. Jayroe JB, Wang D, Deyo DJ, Alpard SK, Bidani A, Zwischenberger JB. The effect of augmented hemodynamics on blood flow during arteriovenous carbon dioxide removal. ASAIO J 2003; 49: 30–4. Bein T, Müller T, Weber-Carstens S. [Extracorporeal lung support procedures (ECMO/iLA)]. Pneumologie 2008; 62: 137–42. Levine BA, Sirinek KR, Gaskill HV. The radiolabeled microsphere technique in gut blood flow measurement – current practice. J Surg Res 1984; 37: 241–55. Buckberg GD, Luck JC, Payne DB, Hoffman JI, Archie JP, Fixler DE. Some sources of error in measuring regional blood flow with radioactive microspheres. J Appl Phys 1971; 31: 598–604. Kopp R, Bensberg R, Wardeh M, Rossaint R, Kuhlen R, Henzler D. Pumpless arterio-venous extracorporeal lung assist compared with veno-venous extracorporeal membrane oxygenation during experimental lung injury. Br J Anaesth 2012; 108: 745–53. Putensen C, Zech S, Wrigge H, Zinserling J, Stueber F, von Spiegel T, Mutz N. Long-term effects of spontaneous breathing during ventilatory support in patients with acute lung injury. Am J Respir Crit Care Med 2001; 164: 43–9. Theres H, Binkau J, Laule M, Heinze R, Hundertmek J, Blobner M, Erhardt W, Baumann G, Stangl K. Phase-related changes in right ventricular cardiac output under volume-

59

S. Kreyer et al. controlled mechanical ventilation with positive endexpiratory pressure. Crit Care Med 1999; 27: 953–8.

Address: Stefan Kreyer Department of Anesthesiology and Intensive Care Medicine University Hospital Sigmund-Freud-Str. 25 53105 Bonn Germany e-mail: [email protected]

60

Pumpless extracorporeal CO(2) removal restores normocapnia and is associated with less regional perfusion in experimental acute lung injury.

Lung protective ventilation may lead to hypoventilation with subsequent hypercapnic acidosis (HA). If HA cannot be tolerated or occurs despite increas...
251KB Sizes 0 Downloads 0 Views