Influence of segmental spinal cord perfusion on intrathecal oxygen tension during experimental thoracic aortic crossclamping

Transcription

1 BASIC RESEARCH STUDIES Influence of segmental spinal cord perfusion on intrathecal oxygen tension during experimental thoracic aortic crossclamping Anders Hellberg, MD, Itaru Koga, MD, Lennart Christiansson, MD, Hans Stiernström, MD, PhD, Lars Wiklund, MD, PhD, David Bergqvist, MD, PhD, and Sadettin Karacagil, MD, PhD, Uppsala, Sweden Purpose: The purpose of this study was to evaluate the possibility of identifying alterations in blood supply to the spinal cord during thoracic aortic crossclamping. Methods: In 17 pigs, a multiparameter PO 2, PCO 2, and ph sensor was introduced into the intrathecal space for continuous monitoring of cerebrospinal fluid (CSF) oxygenation during aortic crossclamping. An epidural laser Doppler probe was used to measure spinal cord flux. After insertion of an aortic shunt from the left subclavian to the left iliac artery and interruption of the right subclavian and lumbar arteries (L2-L5), the thoracic aorta just distal to the left subclavian artery was clamped for 60 minutes. By placement of the distal aortic crossclamping below the level of L1 in group A (n = 9 animals), perfusion of only the abdominal visceral arteries was maintained. In group B (n = 8 animals), the distal aortic crossclamping was above the level of T12, and thus some spinal cord perfusion was maintained through the aortic shunt. Results: The significant decrease in CSF PO 2 was observed within 3 minutes after the placement of the proximal aortic crossclamping and was normalized in all animals after establishment of the shunt flow. In group A, distal aortic crossclamping caused a decrease in CSF PO 2 with at least 50% of the preclamping values within 3 minutes. The mean CSF PO 2 of 2.99 ± 0.70 kpa at 60 minutes of distal aortic crossclamping in group B was significantly higher than in group A (0.11 ± 0.11 kpa; P <.001). In group A, PCO 2 measurements showed no significant changes in 3 minutes after distal aortic crossclamping but revealed significantly higher values at 30 and 60 minutes compared with group B. Spinal cord flux values showed similar changes as CSF PO 2 during the whole experiment in both groups. Conclusion: In this experimental model of aortic crossclamping, continuous CSF oxygen tension monitoring allows rapid detection of alterations in spinal cord circulation. (J Vasc Surg 2000;31: ) From the Department of Surgery (Drs Hellberg, Bergqvist, and Karacagil) and the Department of Anesthesiology (Drs Koga, Christiansson, Stiernström, and Wiklund), University Hospital. Supported in part by the Swedish Medical Research Council Competition of interest: nil. Reprint requests: Anders Hellberg, MD, Department of Surgery, University Hospital, Uppsala, S , Sweden. Copyright 2000 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter /2000/$ /1/ Clinical and experimental data indicate that the cause of paraplegia after thoracoabdominal aortic replacement is multifactorial. Temporary or permanent interruption of spinal cord circulation is believed to be one of the most important causative factors in the development of neuromuscular deficit. Wadouh et al 1 directly measured PO 2 of the spinal cord surface in pigs after simple thoracic aortic crossclamping and reported that the primary cause of spinal cord injury was oxygen deficiency in the area distal to the aortic occlusion and supplied by critical intercostal arteries. In a dog model of graded spinal cord ischemia, Ishizaki et al 2 measured intrathecal, epidural, and spinal cord oxygen tension with a mass spectrometer and showed a significant correlation between intrathecal and spinal cord oxygen tension. Recently, we have demonstrated that 164

2 Volume 31, Number 1, Part 1 Hellberg et al 165 cerebrospinal fluid (CSF) oxygen tension measured by a multiparameter sensor is sensitive to detect spinal cord ischemia in a pig model induced by a simple thoracic aortic crossclamping without a shunt. 3 In the present study, we have developed a new experimental model in pigs by interruption of collateral circulation from subclavian arteries and insertion of an aortic shunt in which spinal cord blood flow is dependent on segmental arteries from T12 to L1. The alterations in spinal cord blood flow during preservation or exclusion of the segmental arteries were studied by on-line intrathecal oxygen tension monitoring. MATERIAL AND METHODS Eighteen Swedish domestic pigs (Sus scrofa domestica), nine males and nine females that weighed from 20.8 to 35.0 kg (mean, 26.2 kg) were premedicated with tiletamine 3 mg/kg and zolazepam 3 mg/kg (Zoletil 100) and xylazine 2.2 mg/kg (Rompun vet). Anesthesia was induced with morphine 20 mg and pancuronium bromide 8 mg (Pavulon) intravenously. A tracheostomy was performed, and ventilation was maintained with an oxygen and air mixture that provided 40% oxygen (to assure oxygenation with PaO 2 of kpa). The animals were connected to a respirator (900D Servoventilator; Siemens Elema, Solna, Sweden) with a tidal volume of approximately 10 ml/kg. Arterial blood gas analysis was used to set the respiratory rate to provide normoventilation at baseline (PCO kpa). A moderate positive end-expiratory pressure of 4 cm water was used. Anesthesia was then continued with intravenously administered infusion consisting of glucose 25 mg/ml, 5 g Ketamine hydrochloride (100 mg/ml), 120 mg morphine, and 60 mg pancuronium bromide (Pavulon) at a rate of 4 ml/kg. A continuous infusion of isotonic saline solution (154 mmol/l) was given for basal fluid requirements at 10 ml/kg during the whole procedure. Normothermia was maintained with warm fluids and the use of a thermostatically controlled heating pad. Right carotid and femoral arteries were catheterized (Ohmeda 18 gauge; Medical Devices Division, Beeton Dickinson Medical System, Franklin Lakes, NJ) for arterial pressure monitoring. A flow-directed Swan Ganz catheter (Ohmeda 7F; Medical Devices Division) was introduced through the right internal jugular vein for measurements such as cardiac output, pulmonary artery pressure, pulmonary capillary wedge pressure, mixed venous saturation, and core temperature. Central venous pressure was monitored through the right jugular vein. All hemodynamic pressures were record- ed with Ohmeda transducers (Medical Devices Division) connected to a Siemens Sirecust 1280/1281 surveillance monitor and a printer (Siemens Medical Electronics Inc, Danvers, Mass). The electrocardiogram was continuously recorded. A urinary catheter was inserted by direct puncture of the bladder for urine output monitoring. A median sternotomy was performed, and the dissection was extended to the abdominal aorta and the left common iliac artery through a median retroperitoneal incision. Circumferential dissection of the left diaphragm and anterior mobilization of the left kidney was performed. The left subclavian artery just distal to the aortic arch, the right subclavian artery immediately after its origin from the brachiocephalic trunk, and the entire descending thoracic and abdominal aorta together with the iliac arteries were dissected. The abdominal visceral, intercostal (T4-13), and lumbar arteries were also dissected. After these preparations, a limited distal thoracic laminectomy was performed; over an arterial needle introducer, a multiparameter PO 2, PCO 2, ph, and temperature sensor (Paratrend 7; Biomedical Sensors, High Wycombe, UK) was introduced into the intrathecal space for continuous monitoring of PO 2, PCO 2, and ph in CSF. The principles and the use of this method for detection of spinal cord ischemia during experimental and preliminary clinical aortic crossclamping has previously been reported. 3,4 This probe consists of a combined electrode fiber-optic system that simultaneously measures PO 2, PCO 2, ph, and temperature after a 30-minute calibration period. A laser Doppler probe (probe T201; Perimed, Järfälla, Sweden) was placed posterior to the dura (cranial to the level of Paratrend catheter insertion) and connected to a laser Doppler flowmeter (Pf2B; Perimed) for measurement of the spinal cord flux (SCF). After a stabilization period of 30 minutes and baseline hemodynamic measurements, Paratrend and laser Doppler recordings were obtained. Paratrend-derived measurements were always obtained 3 minutes after each clamp manipulation to allow equilibrium between spinal cord and CSF. A nonheparinized plastic tube was inserted as an aortic shunt through the proximal left subclavian artery close to the aortic arch and to the distal abdominal aorta through the proximal left common iliac artery. The thoracic aorta just distal to the left subclavian artery was clamped, and the distal aortic perfusion was restored after 3 minutes through the aortic shunt. Volume flow through the shunt was continuously monitored by transit-time ultrasonography (probe 6RB32; Transonic Systems Inc, Ithaca, NY). Shunt flow was recorded on a printer (BBC Goertz metrawatt, SE 120; Goertz, Wien, Austria). The aver-

3 166 Hellberg et al January 2000 were killed with 10 to 20 ml of potassium chloride (Addex potassium chloride) after the last measurements. All animals were treated in compliance with the Principles of Laboratory Animal Care and the Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication No , revised 1985). Study approval was obtained from the local Ethics Committee for animals. The Mann-Whitney U test was used for a comparison between groups, and Wilcoxon signed rank test was used for paired data in statistical calculations. All values are expressed as mean ± SEM. A P value of less than.05 was considered statistically significant. Fig 1. Schematic illustration of the experiment and the placement of aortic crossclamps in groups A and B. age volume flow through the shunt was calculated as the mean flow at each minute. The right subclavian artery was also clamped, followed by interruption of all the lumbar arteries with ligatures except for L1; 3 minutes were allowed for CSF recordings before placement of the distal aortic clamp below the level of T12 or L1. The animals were divided randomly into two groups (Fig 1) depending on the level of the distal aortic crossclamping: group A (n = 9 animals), distal aortic crossclamping caudal to the level of L1 and perfusion of only the visceral arteries through the aortic shunt, Group B (n = 8): distal aortic crossclamping cranial to the level of T12 and perfusion of the spinal cord (T12-L1) together with the visceral arteries via the aortic shunt. Double aortic crossclamping was kept for 60 minutes followed by a 60-minute reperfusion period. Paratrend and laser Doppler recordings were continuously obtained during aortic crossclamping and reperfusion. One pig was excluded because of bleeding problems, giving a hypotensive state (hemorrhagic shock) and thereby occlusion of the shunt. In the present study, laser Doppler flowmetry was used in 12 experiments (7 experiments in group A and 5 experiments in group B) because of lack of equipment in 5 cases. Hemodynamic measurements were obtained before aortic crossclamping, before declamping, and 30 and 60 minutes after declamping. The pigs RESULTS One pig was excluded, and 17 animals were analyzed. In all animals, during proximal aortic crossclamping without shunt flow, CSF PO 2 measurement demonstrated a decrease from a mean value of 3.87 ± 0.54 kpa (median, 2.9 kpa; range, kpa) to a mean of 1.45 ± 0.24 kpa (median, 1 kpa; range, kpa) in 3 minutes (P <.001). Restoration of the distal aortic perfusion through the shunt before placement of the distal aortic clamp almost normalized the CSF oxygen tension in all animals within 3 minutes. In group A, after the placement of the distal aortic clamp, CSF PO 2 declined rapidly (in 3 minutes) from 3.88 ± 0.75 kpa to 1.18 ± 0.42 kpa (P =.008) and remained low during clamping period (Fig 2). All animals in group A demonstrated at least 50% (range, 50%-100%) decrease of CSF oxygen tension after the application of the distal aortic crossclamping compared with preclamping values. The mean intrathecal oxygen tension of 2.99 ± 0.70 kpa (range, kpa) at 60 minutes of double aortic crossclamping in group B was significantly higher than in group A (0.11 ± 0.11 kpa; P <.001; Fig 3). The CSF PCO 2 measurements in all animals showed a similar but an inverted trend compared with the oxygen tension (Figs 4 and 5). It was significantly higher in group A 30 and 60 minutes after distal aortic crossclamping compared with group B (P <.01). On the other hand, contrary to the observations with PO 2, there was no significant increase in PCO 2 values after 3 minutes of distal aortic crossclamping in group A. CSF ph measurements were significantly lower in group A at 30 and 60 minutes after distal aortic crossclamping compared with group B. They were 6.48 and 6.38 in group A, 6.91 and 6.89 in group B, respectively (P <.05). Paratrend-derived parameters demonstrated no significant changes after the application of double aortic crossclamping in group B.

4 Volume 31, Number 1, Part 1 Hellberg et al 167 Fig 2. CSF PO 2 measurements in group A throughout Fig 3. CSF PCO 2 measurements in group A throughout The calculated average volume flow through the shunt was 1012 ml/min (minimum flow range, ml/min; mean, 821 ml/min; SE, 79.9; maximum flow range, ml/min; mean, ; SE, 70.0). The laser Doppler measurements (SCF) during the whole procedure in both groups are summarized in Figs 6 and 7. A dramatic decrease of SCF (100%-8%; P <.001) was observed in all animals after proximal aortic crossclamping and then almost normalization (84%; P <.05) when distal reperfusion through the shunt was obtained in all 12 animals. There was no further significant decrease in SCF during distal aortic occlusion in group B. On the other hand, SCF values were around 24% during distal aortic occlusion in group A, which was similar to the CSF PO 2 decline in the same group. DISCUSSION Despite the use of various surgical methods and pharmacologic adjunctive measures, irreversible spinal cord ischemia leading to paraplegia continues to be the most dreaded complication after thoracoabdominal aortic replacement. A number of intraoperative monitoring methods such as somatosensory, motorevoked potentials, spinal-evoked potentials, and intrathecal hydrogen-induced current impulse techniques have been developed to detect spinal cord ischemia Somatosensory-evoked potentials have been criticized by many authors because they may fail to monitor directly the function of anterior spinal cord motor tracts during aortic crossclamping and seem to be too sensitive to body temperature changes, anesthetic drugs, hemodilution, hypotension, and hypoxia. 11 Monitoring spinal cord function with

5 168 Hellberg et al January 2000 Fig 4. CSF PO 2 measurements in group B throughout Fig 5. CSF PCO 2 measurements in group B throughout motor-evoked potentials is a relatively new technique, and recording myogenic motor responses evoked by either transcranial or spinal electrical stimulation has been shown to be effective in assessing spinal cord ischemia during interruption of critical blood supply; this technique is not yet universally accepted. 7-9 Intraoperative identification of those arteries that supply the spinal cord by use of an intrathecal platinum electrode to detect hydrogen in solution that has been injected into the aortic ostia has been successfully used only in a small pilot study. 10 We have recently demonstrated that CSF oxygen tension monitoring with the Paratrend catheter in a pig model with single proximal thoracic aortic clamping without shunt for 30 minutes correlates with changes in spinal cord microcirculation measured by laser Doppler flowmetry. 3 Laser Doppler flowmetry has previously been validated in experimental spinal cord ischemia models, and a significant correlation was found also between percent changes of laser Doppler flowmetry and microsphere technique Laser Doppler flowmetry has also been shown to predict neurologic and histopathologic outcome in a spinal cord ischemia model that has been induced by thoracic aortic occlusion in cats. 15 Disadvantages of this technique are the difficulty of obtaining absolute values and the sensitivity to artifacts. It also requires laminectomy for placement of the probe in the epidural space. The present experimental model differs from our previous study in the application of double aortic crossclamping and an aortic shunt that resembles the common surgical technique used for aortic replacement. Interruption of the subclavian and lumbar arter-

6 Volume 31, Number 1, Part 1 Hellberg et al 169 Fig 6. Epidural laser Doppler measurements in group A throughout the whole experiment. Values are presented as mean ± SEM. AXC, Aortic crossclamping; DAXC, double aortic Fig 7. Epidural laser Doppler measurements in Group B throughout the whole experiment. Values are presented as mean ± SEM. AXC, Aortic crossclamping; DAXC, double aortic ies converts the pig spinal cord circulation to be solely dependent on the blood supply through segmental arteries from T12 to L1 and provides the opportunity to study the alterations in intrathecal oxygen tension during and after crossclamping for 60 minutes below the level of L1 (group A) and above the level of T12 (group B). This animal model with aortic shunt also eliminates the possible deleterious effects of distal organ ischemia-reperfusion and avoids the use of various drugs for the prevention of proximal hypertension that may influence spinal cord perfusion. The results of this study demonstrate that continuous on-line monitoring of intrathecal oxygenation is sensitive for detection of alterations in spinal cord blood flow during exclusion or preservation of distal thoracic and proximal lumbar segmental arteries. Despite the considerable range in baseline intrathecal PO 2 values, similar changes in response to aortic crossclamping were demonstrated in all animals. The reason for lower baseline CSF PO 2 values in some animals might be due to direct contact of the catheter with the spinal cord because of limited amount of CSF in pigs. Spinal cord PO 2 values have been reported to be considerably lower than the intrathecal PO 2 values. 2 The findings from the present study should be interpreted cautiously with regard to the practical clinical implications during thoracoabdominal aortic surgery because the present experimental model lacks clinical, histopathologic, or neurophysiologic outcome. It was designed to assess the sensitivity of intrathecal oxygen tension monitoring during preservation or exclusion of some of the segmental arteries

7 170 Hellberg et al January 2000 by the use of an aortic shunt during aortic crossclamping compared with microcirculatory changes obtained from laser Doppler flowmetry. To further validate CSF PO 2, PCO 2, and ph as indicators of spinal cord ischemia, quantitative studies to determine the critical levels of Paratrend-derived parameters and crossclamping times that correlate with the loss of spinal cord function and neurologic outcome are the subjects of some of our ongoing projects. The return of CSF PO 2 values to baseline after prolonged severe spinal cord ischemia does not necessarily mean the return of spinal cord function. For this reason, neurophysiologic outcome with motor-evoked potentials is necessary for further validation of the method, and PCO 2 or ph may be more valuable than PO 2 in regard to the prediction of the severity of spinal cord ischemia. REFERENCES 1. Wadouh F, Arndt C-F, Metzger H, Hartmann M, Wadouh R, Borst H-G. Direct measurements of oxygen tension on the spinal cord surface of pigs after occlusion of the descending aorta. J Thorac Cardiovasc Surg 1985;89: Ishizaki M, Sugiyama S, Uchida H, Nawa S, Shimizu N. Intrathecal oxygen concentration as a new indicator of spinal cord ischemia. Acta Med Okayama 1997;51: Christiansson L, Hellberg A, Koga I, Thelin S, Bergqvist D, Karacagil S. A new method of intrathecal PO 2, PCO 2 and ph measurements for continuous monitoring of spinal cord ischemia during aortic clamping in pigs. Surgery. In press. 4. Christiansson L, Karacagil S, Thelin S, et al. Continuous monitoring of intrathecal PO 2, PCO 2 and ph during surgical replacement of type II thoracoabdominal aortic aneurysm. Eur J Vasc Endovasc Surg 1998;15: Galla JD, Ergin A, Sadeghi AM, et al. A new technique using somatosensory evoked potential guidance during descending and thoracoabdominal aortic repairs. J Cardiac Surg 1994; 9: Cummingham JN, Lashinger JC, Spencer FC. Monitoring of somatosensory evoked potentials during surgical procedures on the thoracoabdominal aorta: clinical observations and results. J Thorac Cardiovasc Surg 1987;94: Svensson LG, Patel V, Robinson MF, Ueda T, Roehm JOF Jr, Crawford ES. Influence of preservation or perfusion of intraoperatively identified spinal cord blood supply on spinal motor evoked potentials and paraplegia after aortic surgery. J Vasc Surg 1991;13: Mongan PD, Peterson RE, Williams D. Spinal evoked potentials are predictive of neurologic function in a porcine model of aortic occlusion. Anesth Analg 1994;78: Haan P, Kalkman CJ, Mol BA, Ubags LH, Veldman DJ, Jacobs MJHM. Efficacy of transcranial motor-evoked myogenic potentials to detect spinal cord ischemia during operations for thoracoabdominal aneurysms. J Thorac Cardiovasc Surg 1997;113: Svensson LG, Patel V, Coselli JS, Crawford ES. Preliminary report of localization of spinal cord blood supply by hydrogen during aortic operations. Ann Thorac Surg 1990;49: Shenaq SA, Svensson LG. Paraplegia following aortic surgery. J Cardiothorac Vasc Anesth 1993;7: Aadahl P, Saether OD, Stenseth R, Myhre HO. Microcirculation of the spinal cord during proximal aortic crossclamping. Eur J Vasc Surg 1990;4: Lindberg PJ, Jacobs TP, Frerichs KU, Hallenbeck JM, Feuerstein GZ. Laser-Doppler flowmetry in monitoring regulation of rapid microcirculatory changes in spinal cord. Am J Physiol 1992;263:H Lindberg PJ, O Neill JT, Paakkari IA, Hallenbeck JM, Feuerstein G. Validation of laser-doppler flowmetry in measurement of spinal cord blood flow. Am J Physiol 1989;257:H Yamada T, Morimoto T, Nakase H, Hirabayashi H, Hiramatsu K, Sakaki T. Spinal cord blood flow and pathophysiological changes after transient spinal cord ischemia in cats. Neurosurgery 1998;42: Submitted Mar 12, 1999; accepted Aug 9, 1999.