The impact of ischaemic preconditioning on spinal cord blood flow and paraplegia

PII: S0967-2109(01)00083-7

Cardiovascular Surgery, Vol. 9, No. 6, pp. 575–579, 2001  2001 The International Society for Cardiovascular Surgery Published by Elsevier Science Ltd. All rights reserved 0967-2109/01 $20.00

www.elsevier.com/locate/cardiosur

The impact of ischaemic preconditioning on spinal cord blood flow and paraplegia Tetsuya Ueno, Zhi-Li Chao, Yukio Okazaki and Tsuyoshi Itoh Department of Thoracic and Cardiovascular Surgery, Saga Medical School, 5-1-1 Nabeshima, Saga City, Saga 849-9501, Japan Purpose: We investigated the effect of ischaemic preconditioning (IP) on ischaemic spinal cord injury in a rabbit model. Methods: Fourteen rabbits were divided into IP and control groups of seven rabbits each. We repeated 3-min clamping of the infrarenal abdominal aorta and 3-min reperfusion twice (preconditioning), followed by 15 min clamping in the IP group. In the control group, the aorta was clamped for 15 min without preconditioning. Lumbar cord blood flow and systemic blood pressure were measured until 3 h of reperfusion. Another 14 rabbits underwent the same procedures with or without IP and neurologic status was assessed on the second postoperative day. Results: The percent change in lumbar cord blood flow after reperfusion was significantly greater (P = 0.013) in the IP group despite lower mean blood pressure. There was no significant difference in overall neurologic status (P = 0.461) but the incidence of spastic paraplegia in the IP group was lower (14%, 1/7) than that of control group (43%, 3/7). Conclusion: IP increased postischaemic lumbar cord blood flow and contributed to lower incidence of spastic paraplegia.  2001 The International Society for Cardiovascular Surgery. Published by Elsevier Science Ltd. All rights reserved Keywords: ischaemic preconditioning, spinal cord ischaemia, paraplegia

Introduction The cause of spastic paraplegia after descending thoracic or thoracoabdominal aorta surgery is thought to be multifactorial and remains one of the most devastating complications. Numerous experimental investigations have been performed to reduce the incidence of ischaemic spinal cord injury and to prevent the occurrence of spastic paraplegia. Ischaemic preconditioning (IP), in which repeated brief episodes of nonlethal ischaemia increase tolerance to subsequent lethal ischemia, was first introduced in 1986 [1]. The fundamental mechanisms of IP have been extensively investigated, mainly in the regional myocardium or global heart, in the past decCorrespondence to: Tetsuya Ueno M.D. Tel.: +81-954-43-1120; Fax: +81-954-42-2452; e-mail: [email protected]

CARDIOVASCULAR SURGERY

DECEMBER 2001 VOL 9 NO 6

ade [2–4]. However, only a limited number of animal studies have investigated the protective effect of IP against ischaemia-reperfusion injury in other organs. The aim of this study was to determine whether IP provides a similar beneficial effect on ischaemic spinal cord injury by evaluating spinal cord blood flow and neurologic status after 15 min of infrarenal aortic occlusion in a rabbit model.

Materials and methods We used 28 Japanese white rabbits (KBT-JW) weighing 2.1–2.4 kg. Animal care and all procedures were performed in compliance with the Saga Medical School Guidelines for Animal Experimentation formulated in 1988 and revised in 1995. Prior to the experiments, all rabbits received subcutaneous ketamine hydrochloride (40 mg/kg) and were anesthet575

Ischaemic preconditioning: T. Ueno et al.

ized with intravenous thiamylal sodium (20 mg/kg). After tracheostomy and endotracheal intubation, pancronium bromide (0.5 mg/kg) was administered and the rabbits were placed on a Harvard-type ventilator (inspired oxygen fraction, 1.0) with a tidal volume of 30 ml set at a rate of 20 cycles/min. Anesthesia was maintained by additional subcutaneous injection of ketamine hydrochloride and pancronium bromide. Lactated Ringer’s solution was infused at 20 ml/h until final measurement in each experiment was completed. In experiment A, 14 rabbits were divided into the IP group and control group, consisting of seven rabbits each. We performed laminectomy at the level of L4 and L5 to expose the lumbar spinal cord. We inserted an electrode for measurement of tissue blood flow (UH meter electrodes, type UHE-114, Unique Medical Co., Tokyo, Japan) to the lumbar spinal cord, immobilized the electrode, and then closed the wound. We measured the tissue blood flow of the lumbar spinal cord using the hydrogen clearance method with a tissue flow meter (Digital UH meter, model MHG-Dl; Unique Medical Co.). We infused heparin (1000 units) and inserted a 20gauge catheter into the right carotid artery to monitor systemic blood pressure. A midline laparotomy was performed, and the abdominal aorta just below the origin of the left renal artery was exposed and encircled with a suture. Baseline measurements of systemic blood pressure and lumbar cord blood flow were made. In the IP group, we clamped the abdominal aorta with a bulldog forceps for 3 min and then released the clamp for 3 min. We repeated these procedures twice and clamped the abdominal aorta for 15 min. Then we released aortic clamp and closed the wound. We measured the systemic blood pressure and lumbar cord blood flow during IP and 15-min aortic clamping and after release of aortic clamp at 2, 5, 10, 15, 20, 30, 45, 60, 120 and 180 min. In the control group, the abdominal aorta was clamped for 15 min without IP and the abdomen was closed. In experiment B, another 14 rabbits were divided into the IP group and control group, consisting of seven rabbits each. We performed exactly the same procedures in experiment A except for laminectomy and insertion of an electrode for the measurement of lumbar cord blood flow. Neurologic status was assessed on the second postoperative day by an observer blinded to rabbit group using Tarlov’s criteria: grade 0, spastic paraplegia with no movement of the hind limbs; grade I, spastic paraplegia with slight movement of the hind limbs; grade II, good movement of the hind limbs, but unable to stand; grade III, able to stand, but unable to walk normally; and grade IV, complete recovery.

576

Statistical analysis The computer program package Statview 4.01 (Abacus Concepts, Berkeley, CA, USA) for Macintosh (Apple Computer, Inc., Cupertino, CA, USA) was used for statistical analysis. Data are presented as mean±standard error of the mean. Repeated twoway analysis of variance (ANOVA) with the post-hoc Fisher PLSD test was used to analyze continuous data. The unpaired Student t-test was used to analyze other data. The Mann-Whitney U-test was used to compare neurologic status. A P value of less than 0.05 was considered statistically significant.

Results Experiment A Lumbar cord blood flow and the percent change from the baseline value are presented in Figures 1 and 2. Lumbar cord blood flow was reduced almost to zero by clamping the infrarenal abdominal aorta during repeated 3-min preconditioning and 15 min clamping. The blood flow in the IP group tended to be greater than that in control group during 3-h reperfusion and its percent change was significantly different (P = 0.013). Mean systemic blood pressure during 15 min aortic clamping in the IP group was lower, although not significantly (P = 0.377), than that in control group (Figure 3). During reperfusion, mean blood pressure in the IP group tended to be lower, especially during the first 15 min, than that in control group (Figure 3).

Figure 1 Lumbar spinal cord blood flow. The lumbar spinal cord blood flow in the IP group tended to be greater after unclamping of the infrarenal aorta than that in the control group (P = 0.13), especially within 15 min after unclamping. The cross-clamping of the infrarenal aorta reduced the lumbar spinal cord blood flow almost to zero. P1=first clamping of the infrarenal aorta for 3 min.; P2=second clamping of the infrarenal aorta for 3 min

CARDIOVASCULAR SURGERY

DECEMBER 2001 VOL 9 NO 6

Ischaemic preconditioning: T. Ueno et al. Table 1 Neurologic status on the second postoperative daya Neurologic status

IP group

Control group

Grade Grade Grade Grade Grade

0 1 4 1 1

1 2 2 1 1

0 I II III IV

a

No significant difference existed in overall neurologic status between the IP group and the control group (P = 0.461), although the incidence of spastic paraplegia (grades 0 and I) was reduced from 43 to 14% by IP.

Discussion Figure 2 Percent change in lumbar spinal cord blood flow. The percent change in lumbar spinal cord blood flow in the IP group was significantly greater after unclamping of the infrarenal aorta than that in the control group (P = 0.013). P1=first clamping of the infrarenal aorta for 3 min; P2=second clamping of the infrarenal aorta for 3 min

Figure 3 Mean blood pressure. The mean blood pressure in the IP group tended to be lower after unclamping of the infrarenal aorta than that in the control group (P = 0.774). During 15 min clamping, the mean blood pressure in the IP group was lower, but not significantly, than that in the control group. P1=first clamping of the infrarenal aorta for 3 min; P2=second clamping of the infrarenal aorta for 3 min

Experiment B The results of neurologic assessment are presented in Table 1. There was no statistically significant difference between the two groups in motor function of the hind limbs on the second postoperative day (P = 0.461), but the incidence of spastic paraplegia (grades 0 and I) was less in the IP group (14%) when compared with that in the control group (43%). CARDIOVASCULAR SURGERY

DECEMBER 2001 VOL 9 NO 6

The proposed mechanisms by which IP improves myocardial preservation and functional recovery after lethal ischaemia are preservation of adenosine triphosphate (ATP) by slowing energy metabolism [1], activation of AI adenosine receptors [5], activation of ATP-sensitive potassium channels [6] and induction of heat shock proteins (HSP) [3]. In addition, recent investigations demonstrated a delayed anti-infarct effect of IP, termed the ‘second window of protection’, which reappeared more than 24 h after the initial sublethal ischaemia [3, 7]. The delayed protective effect of IP is suggested to be related to the new synthesis of proteins with cytoprotective effects, including manganese-superoxide dismutase [8] and HSP 72 [3]. We assessed the neurologic status of the hind limbs on the second postoperative day so that the second window of protection would not be overlooked, if it actually exists, and to evaluate the sum of effects possibly induced by immediate and delayed protection against ischaemia-reperfusion injury on the spinal cord. Zvara and associates reported that 3 min of aortic occlusion followed by 30 min of reperfusion and 12 min of occlusion resulted in significantly better neurologic function and survival, associated with less histologic damage, in a rat model [9]. Our study also investigated this immediate effect of IP. Munyao et al. showed significantly improved hindlimb function when short-term ischemia was induced 12 h, but not 48 h before the second ischemia in a rabbit model similar to ours [10]. On the other hand, Matsuyama and coworkers demonstrated a delayed (48 h later) benefit of IP on improved neurologic outcomes associated with HSP induction in a canine model [11]. From these experimental results, it is speculated that the effective time window of IP may differ among species. It has been controversial how long sublethal ischaemic stress should be induced and how many times it should be repeated. We set the duration of IP at 3 min because it required about 2.5 min to obtain the results of lumbar spinal cord blood flow using the hydrogen-clearance method. The frequency of IP in most studies was once or twice. Ili577

Ischaemic preconditioning: T. Ueno et al.

odromitis and associates showed that 6–8 episodes of brief 5-min ischaemia resulted in the loss of efficacy of IP in rabbit hearts [12]. Considering the methodological limitations in our present study and the need to avoid potential adverse effects, we repeated 3-min ischaemia followed by 3-min reperfusion twice as the conditions for IP. In the present study, the lumbar spinal cord blood flow after the release of aortic clamping in the IP group was markedly greater than that in the control group despite lower mean systemic aortic pressure (Figure 1). Hyperaemia is a transient and remarkable increase in tissue blood flow which occurs immediately after reperfusion. Hyperaemia after ischaemia can be either beneficial or harmful to recovery, depending on the conditions where hyperaemia occurs [13]. Hyperaemia present in slightly impaired regions may work to accelerate recovery from ischaemic injury. However, if it occurs in irreversibly damaged regions, it may exacerbate endothelial injury and permeability edema, causing secondary damage to neuronal tissue [14]. Since 15-min duration is thought to be shorter than the critical aortic clamping time to induce acute permanent or delayed-onset paraplegia in rabbits [15], various degrees of ischaemia-reperfusion damage might have occurred and coexisted after reperfusion in the lumbar spinal cord. The wide data distribution of neurologic status in the control group supports this hypothesis. Despite the lack of significant improvement in overall neurologic function, IP and consequent hyperaemia might have worked to promote the recovery from ischaemia-reperfusion injury at the borderline region between reversible and irreversible injury in the lumbar spinal cord because IP led to a marked reduction in the rate of spastic paraplegia (grades 0 and I) from 43 to 14%. One possible etiology for hyperaemia in the IP group is the attenuation of postischaemic capillary no-reflow. Jerome and associates suggested that activation of ATP-sensitive potassium channels induced by IP may attenuate the development of postischaemic capillary no-reflow by preventing arteriolar vasospasm and endothelial cell swelling [16]. This hypothesis is supported by results showing that IP attenuates leucocyte adhesion and migration by activation of ATP-sensitive potassium channels and adenosine receptors [17]. Increased lumbar spinal cord blood flow in the IP group suggests the presence of an IP-induced vasodilating effect through these mechanisms. In addition, the mean systemic blood pressure during 15-min clamping in the IP group was considerably lower than that in control group, which suggests the presence of IP-induced vasodilating action not only on lumbar spinal cord alone, but also on systemic vascular beds. Zvara and associates showed a similar reduction in systemic arterial pressure in their IP group [9]. 578

Based on the increased lumbar spinal cord blood flow, lower systemic blood pressure during aortic cross-clamping as well as after reperfusion and marked reduction in the incidence of spastic paraplegia in the IP group, it is speculated that the increase in spinal cord blood flow through both attenuation of postischaemic capillary no-reflow and of vasodilating action might have contributed to limit the irreversibly injured region in the spinal cord and, consequently, to reduce the incidence of spastic paraplegia in the present study.

Conclusion In conclusion, repeated 3-min occlusion of the infrarenal abdominal aorta induced a significantly greater increase in lumbar spinal cord blood flow compared with the control group. IP did not significantly improve overall neurologic status of the hind limbs, but contributed to lower incidence of spastic paraplegia.

References 1. Murry, C. E., Jennings, R. B. and Reimer, K. A., Preconditioning with ischemia: a delay in lethal cell injury in ischemic myocardium. Circulation, 1986, 74, 1124–1136. 2. Schott, R. J., Rohmann, E. R., Braun, E. R. et al., Ischemic preconditioning reduces infarct size in swine myocardium. Circulation Research, 1990, 66, 1133–1142. 3. Marber, M. S., Latchman, D. S., Walker, J. M. et al., Cardiac stress protein elevation 24 hours after brief ischemia or heat stress is associated with resistance to myocardial infarction. Circulation, 1993, 88, 1264–1272. 4. Illes, R. W. and Swoyer, K. D., Prospective, randomized clinical study of ischemic preconditioning as an adjunct to intermittent cold blood cardioplegia. Annals of Thoracic Surgery, 1998, 65, 748–753. 5. Liu, G. S., Thornton, J., Van Winkle, D. M. et al., Protection against infarction afforded by preconditioning is mediated by A1adenosine receptors in rabbit heart. Circulation, 1991, 84, 350– 356. 6. Speechly Dick, M. E., Grover, G. J. and Yellon, D. M., Does ischemic preconditioning in the human involve protein kinase C and the ATP-dependent K+ channel? Studies of contractile function after ischemia in an atria1 in vitro model. Circulation Research, 1995, 77, 1030–1035. 7. Baxter, G. F., Goma, F. M. and Yellon, D. M., Characterisation of the infarct-limiting effect of delayed preconditioning: timecourse and dose-dependency studies in rabbit myocardium. Basic Research in Cardiology, 1997, 92, 159–167. 8. Kuzuya, T., Hoshida, S., Yamashita, N. et al., Delayed effects of sublethal ischemia on the acquisition of tolerance to ischemia. Circulation Research, 1993, 72, 1293–1299. 9. Zvara, D. A., Colonna, D. M., Deal, D. D. et al., Ischemic preconditioning reduces neurologic injury in a rat model of spinal cord ischemia. Annals of Thoracic Surgery, 1999, 68, 874–880. 10. Munyao, N., Kaste, M. and Lindsberg, P. J., Tolerization against loss of neuronal function after ischemia-reperfusion injury. NeuroReport, 1998, 9, 321–325. 11. Matsuyama, K., Chiba, Y., Ihara, A. et al., Effect of spinal cord preconditioning on paraplegia during cross-clamping of the thoracic aorta. Annals of Thoracic Surgery, 1997, 63, 1315–1320. 12. Iliodromitis, E. K., Kremastinos, D. T., Katritis, D. G. et al., Multiple cycles of preconditioning cause loss of protection in open-chest rabbits. Journal of Molecular and Cellular Cardiology, 1997, 29, 915–920.

CARDIOVASCULAR SURGERY

DECEMBER 2001 VOL 9 NO 6

Ischaemic preconditioning: T. Ueno et al. 13. Barone, G. W., Joob, A. W., Flanagan, T. L. et al., The effect of hyperemia on spinal cord function after temporary thoracic aortic occlusion. Journal of Vascular Surgery, 1988, 8, 535–540. 14. Jacobs, T. P., Kempski, O., McKinley, D. et al., Blood flow and vascular permeability during motor dysfunction in a rabbit model of spinal cord ischemia. Stroke, 1992, 23, 367–373. 15. More, W. Jr. and Hollier, L. H., The influence of severity of spinal cord ischemia in the etiology of delayed-onset paraplegia. Annals of Surgery, 1991, 213, 427–432. 16. Jerome, S. N., Akimitsu, T., Gute, D. C. et al., Ischemic precon-

CARDIOVASCULAR SURGERY

DECEMBER 2001 VOL 9 NO 6

ditioning attenuates capillary no-reflow induced by prolonged ischemia and reperfusion. American Journal of Physiology, 1995, 268, H2063–H2067. 17. Akimitsu, T., Gute, D. C. and Korthuis, R. J., Ischemic preconditioning attenuates postischemic leukocyte adhesion and emigration: role of adenosine and ATP-sensitive potassium channels. Circulation, 1994, 90(Suppl I), 476. Paper accepted 1 May 2001

579