Previous studies indicate leukocytes play a role in the pathogenesis of cerebral arterial air embolism. Because doxycycline inhibits numerous leukocyte activities, the authors hypothesized doxycycline would decrease neurologic impairment after cerebral arterial air embolism.
New Zealand White rabbits anesthetized with methohexital received either intravenous saline (n = 7) or 10 mg/kg doxycycline (n = 7) 1 h before administration of 100 microl/kg of air into the internal carotid artery. Somatosensory-evoked potentials (SSEPs) were recorded at 30-min intervals for the next 2 h. After the final recording, the anesthetic was discontinued, and animals recovered. Animals were neurologically evaluated 4 h after air embolism on a scale of 0 (normal) to 99 (coma) points.
At 4 h, doxycycline animals had lesser neurologic impairment (46 +/- 23; median, 41) than animals that received saline (77 +/- 20; median, 81); P = 0.007. SSEP amplitude was greater in the doxycycline group at 60, 90, and 120 min after air embolism; P = 0.001, 0.006, 0.026, respectively. SSEP amplitudes at 30, 60, 90, and 120 min inversely correlated with 4 h neurologic impairment; tau = -0.43, -0.75, -0.85, -0.79, respectively.
Doxycycline decreased electrophysiologic and neurologic abnormalities after cerebral air embolism. Because groups could be distinguished electrophysiologically as soon as 1 h after air embolism and because SSEP amplitude inversely correlated with neurologic impairment, doxycycline appears to inhibit a key early (approximately 1 h) process in the pathophysiology of cerebral air embolism.
Key words: Air embolism, doxycycline, leukocytes, rabbits, somatosensory-evoked potentials.
Celebral arterial air embolism can occur in many settings, including angiography, decompression sickness, and cardiac surgery. [2-4]Although air emboli interrupt cerebral blood flow, the duration of vessel occlusion is usually short, 1-10 min at the arteriolar level. [5-8]Ultrastructural [8-13]and functional [8,14-16]studies indicate air damages endothelium, with development of a thromboinflammatory reaction involving fibrin, platelets, and leukocytes. Accordingly, it has been proposed that neurologic injury from cerebral air embolism may be due, at least in part, to thromboinflammatory events occurring subsequent to the clearance of air.
Using a dog model of cerebral air embolism, Hallenbeck et al. observed radiolabeled granulocytes accumulated in brain within 1 h. Subsequently, Dutka et al. in dogs and Helps et al. in rabbits observed preembolism leukodepletion attenuated reductions in cerebral blood flow and somatosensory-evoked potentials (SSEPs) occurring over the first few h after cerebral air embolism. Thus, leukocytes appear to play an important early ([approximately] 1 h) role in the pathogenesis of injury produced by cerebral arterial air embolism.
Recently, doxycycline, a member of the tetracycline family of antibiotics, has been shown to inhibit a range of leukocyte activities, including chemotaxis, [20,21]adhesion, degranulation, [21,23]superoxide generation, [21,23]and metalloproteinase activity. [24-26]Clark et al. observed rabbits pretreated with doxycycline had reduced in vitro leukocyte adherence and significantly increased tolerance to temporary spinal cord ischemia. Because of the importance of leukocytes in the pathogenesis of cerebral arterial air embolism and the broad inhibitory effect of doxycycline on leukocytes, we hypothesized doxycycline would decrease neurologic impairment after cerebral arterial air embolism. To test this hypothesis, we used our modification of the rabbit model of cerebral air embolism developed by Helps et al. [7,19]
Methods and Materials
Experimental protocols were approved by the Animal Care and Use Committee of the University of Iowa in accordance with the Guide for the Care and Use of Laboratory Animals, National Institutes of Health publication No. 85-23, revised 1985.
New Zealand White rabbits of either sex (weight, 2.8-3.5 kg) were randomly preassigned to receive either doxycycline (10 mg/kg intravenously, n = 7) or an equivalent volume of saline (1 ml/kg, n = 7) approximately 1 h before cerebral arterial air embolism. Anesthesia was induced in nonfasted animals by inhalation of 5% isoflurane in oxygen. After cannulation of an ear vein with a 22-gauge catheter and orotracheal intubation with a 3.0-mm cuffed endotracheal tube, animals were briefly paralyzed with a single dose of intravenous succinylcholine (1 mg/kg). Thereafter, no muscle relaxants were administered. Animals were ventilated with 2.0% isoflurane in 30% oxygen/balance nitrogen to achieve normocarbia, monitored with a calibrated anesthetic agent analyzer (Datex, Puritan-Bennett, Helsinki, Finland). Throughout surgery, normal saline was infused intravenously at 4 ml [center dot] kg sup -1 [center dot] h sup -1. Rectal temperature was maintained at 37 [degree sign] Celsius to 38 [degree sign] Celsius with a servocontrolled heating pad, but epidural temperature was monitored as the independent variable.
Animals were placed prone in a stereotaxic frame (Kopf Instruments, Tugunda, CA), and the scalp was shaved. The skin was washed with povidone-iodine solution (Purdue Fredrick Co., Norwalk, CT), and all subsequent procedures were performed in aseptic fashion. After skin incision, a 2-mm burr hole was drilled over the left frontoparietal cortex to expose dura. A 1-mm thermocouple (K-type, L-08419-02, Cole Parmer, Chicago, IL) was placed between the cranium and dura to monitor epidural temperature. The bone defect was filled with bone wax. Stainless steel screws for recording SSEPs were placed into the skull with the active electrode located over the left parietal region, 6-mm lateral to the midline and 1-mm anterior to the coronal suture. The reference electrode was placed in midline into maxillary bone. Animals were then turned supine, and through the left femoral artery, a saline-filled polyethylene catheter (PE-90, Intramedic, Parsippany, NJ) was advanced into the abdominal aorta for arterial pressure monitoring and intermittent blood sampling. At this point, either doxycycline (10 mg/kg, 10 mg/ml in saline) or an equivalent volume of saline (1 ml/kg) was administered intravenously over 5 min. The time between doxycycline administration and subsequent air embolism ranged between 53 and 75 min. After this, isoflurane was discontinued (35-58 min before air embolism), and methohexital was administered as a bolus of 10 mg/kg intravenously, followed by a continuous intravenous infusion of 15 ml [center dot] kg sup -1 [center dot] h sup -1. Methohexital anesthesia was continued throughout the remainder of the experiment to avoid isoflurane inhibition of the SSEP and to allow for rapid anesthetic recovery required for neurologic assessment. The methohexital loading dose and infusion rate was based on the pharmacokinetic and pharmacodynamic data of Redke et al. No animal exhibited withdrawal responses or purposeful movement during or after surgery. End-tidal isoflurane concentration was <or= to 0.2% at the time of baseline physiologic and SSEP measurements.
Cerebral Arterial Air Embolism
Through a midline neck incision, the left external, internal, and common carotid arteries were isolated, and a branch of the external carotid (usually the facial) was selected for cannulation. Other branches of the external carotid were ligated with 4-0 silk, and all bleeding points were cauterized. At this point, the following baseline physiologic measurements were obtained: mean arterial pressure, epidural temperature, arterial pH, PO2, PCO2(IL1304, Instrumentation Laboratory, Lexington, MA), hemoglobin concentration (OSM3 [rabbit absorption coefficients], Radiometer, Copenhagen, Denmark), and plasma glucose concentration (YSI model 27, Yellow Springs Instrument Co., Yellow Springs, OH). Baseline SSEP was then obtained by median nerve stimulation by needle electrodes inserted into the right forepaw. After isolation of the carotid arterial system, a temporary aneurysm clip was placed across the left common carotid just proximal to its bifurcation. A saline-filled PE-50 catheter was introduced retrograde through the facial branch of the external carotid and directed into the proximal 1 or 2 mm of the internal carotid. Care was taken to avoid air entrapment in either the facial or internal carotid arteries.
In a previous study using this model, we established a dose-response relationship between volume of injected air and electrophysiologic and neurologic outcomes. Based on this previous study, an air dose of 100 micro liter/kg was chosen to produce unequivocal electro-physiologic and neurologic abnormalities. After baseline data collection, 100 micro liter/kg of air was injected into the internal carotid followed by 0.5 ml of normal saline at a constant rate of 3 micro liter/s by infusion pump. Immediately thereafter, the aneurysm clip was removed from the common carotid, and the injection catheter was withdrawn, reestablishing continuity between the internal and common carotid arteries. The air injection catheter was subsequently removed from the facial artery, and the external carotid artery was ligated at its origin. SSEPs and physiologic data were recorded at 5, 30, 60, 90, and 120 min after completion of air injection.
Recovery and Neurologic Evaluation
Two hours after air embolism, the arterial catheter, cranial screws, and cortical thermocouple were removed. Incisions were closed and infiltrated with 3 ml of 0.5% lidocaine. The methohexital infusion was then discontinued. Animals were extubated when they regained spontaneous ventilation and protective airway reflexes. After extubation, animals received 50% oxygen by mask. Neurologic status was assessed 4 h after air embolism. The neurologic scoring system used was a modification of that described by Baker et al. (Table 1). Best possible neurologic score equaled 0. Worst possible neurologic score equaled 99. Baker et al. found rabbits undergoing 10 min of global cerebral ischemia during hypothermic conditions (29 [degree sign] Celsius) had less neurologic impairment, less histopathologic change, and less glutamate release than rabbits made ischemic while normothermic. Hence, this neurologic scoring system corresponds to histopathologic and biochemical indices of neurologic injury in this species. The neurologic examiner was aware animals had undergone air embolism but was unaware of group assignment.
If by 5 h an animal did not spontaneously ventilate, it was considered to be nonviable and was killed by pentobarbital overdose (150 mg/kg, intravenously). Four animals met this criteria (one doxycycline, three control animals), each having a neurologic impairment score of 96. Five hours after air embolism, extubated animals were returned to the Animal Care Unit where food and water were available. Animals underwent a final neurologic evaluation 24 h after air embolism. Animals were then reanesthetized with 5% isoflurane in oxygen and killed by pentobarbital overdose.
SSEPs were derived by stimulation of the right median nerve. Using subcutaneous needle electrodes, supra-maximal square-wave direct current (DC) pulses of 0.25-ms duration were delivered at 1.4 Hz. Sixty-four cortical responses were averaged with Grass Model 10 evoked response system with bandpass filters of 0.3 and 10,000 Hz (Grass Instruments, Quincy, MA). High amplitude electrical artifact was automatically rejected. The analog signal was converted to digital data by an A-D board interfaced with an IBM AT computer for subsequent analysis. The amplitude of the primary cortical deflection was measured from the trough of the first major negative deflection (N1, occurring at 13 +/- 1 ms) to the peak of the next positive deflection (P sub 1, occurring at 28 +/- 3 ms). N1and P1latencies were recorded for each animal before air embolism; postembolism amplitude measurements were made from those points. The amplitude of the N1-P1complex was expressed as a percentage of the baseline value for each animal.
Data are reported as mean +/- SD. The Mann-Whitney-Wilcoxon test was used to compare median neurologic impairment scores between groups 4 h after air embolism and SSEP amplitudes between groups at 30, 60, 90, and 120 min after air embolism. To test whether SSEP amplitudes correlated with neurologic scores, we used Kendall's nonparametric correlation coefficient (tau) between SSEP amplitude at 30, 60, 90, and 120 min after air embolism and 4 h neurologic impairment score, using data from both groups. We used this correlation coefficient because we had no a priori reason to specify a linear relationship between these measurements. In addition, we wanted to assure that the results of no one animal had a disproportionate effect on the P value. All P values are two-sided and exact (StatXact 3 for Windows, Cytel Software Corp., Cambridge, MA).
Doxycycline resulted in no observable change in hemodynamics. There were no important differences between groups in any systemic variable at any point during the experiment (Table 2). Although epidural temperature appeared to be slightly less in saline control animals compared with doxycycline animals, temperature differences among animals did not affect the results.*
Four hours after air embolism, rabbits that had received doxycycline had lesser neurologic impairment scores (46 +/- 23; median, 41) than animals that had received saline (77 +/- 20; median, 81); P = 0.007. Five hours after air embolism, six of seven doxycycline animals had airway reflexes and spontaneous ventilation sufficient to permit return to the Animal Care Unit versus four of seven saline animals. In the doxycycline group, six of seven animals survived to 24 h (24 h neurologic scores: 5, 5, 6, 26.5, 32, 41.5) compared with three of seven in the control group (24 h scores: 5, 21, 43).
In all cases, SSEP was totally abolished 5 min after air injection. As shown in Figure 1, recovery SSEP amplitude was greater in the doxycycline group at 60, 90, and 120 min after air embolism; P = 0.001, 0.006, 0.026, respectively. SSEP recovery was essentially complete by 60 min. SSEP amplitudes at 30, 60, 90, and 120 min inversely correlated with 4-h neurologic impairment score; tau = -0.43, -0.75, -0.85, -0.79, respectively. The correlation between SSEP amplitude and 4-h neurologic impairment score did not improve after 60 min.
Doxycycline's protective effect was evident early. One hour after air embolism, animals receiving doxycycline had significantly greater recovery of SSEP amplitude than control animals. Over the next h, SSEP remained stable in both groups. Four hours after air embolism, neurologic impairment was significantly less in doxycycline animals than in control animals. The correlation between 1 h SSEP amplitude and 4 h neurologic status suggests the neurologic benefit of doxycycline was conferred by 1 h after air embolism. In a previous dose-response study in this model, we observed that 2 h SSEP amplitude correlated with 3 h neurologic outcome and predicted 24 h survival. The current study demonstrates that the correlation between SSEP amplitude and early neurologic outcome also holds in the presence of a beneficial intervention, specifically, doxycycline. This observation provides further support for early SSEP as a surrogate for neurologic outcome in this model. The mechanisms by which doxycycline decreased early electrophysiologic and neurologic impairment after cerebral arterial air embolism were not tested in this experiment. Nevertheless, given doxycycline's inhibition of leukocyte activities [20-26]and previous studies [17-19]that indicate leukocytes play a key early role in the pathogenesis of cerebral air embolism, we postulate doxycycline's protective effect was possibly mediated by inhibition of leukocyte activity.
Role of Leukocytes in Models of Temporary Cerebral Ischemia
The possibility of early involvement of leukocytes after cerebral air embolism differs greatly from what is observed in studies using temporary middle cerebral artery occlusion (MCAO). After 1 h of MCAO in rats, Matsuo et al. reported brain granulocyte accumulation did not occur before 6-12 h after reperfusion. [31,32]Subsequent MCAO studies showed endothelial expression of leukocyte adhesion molecules (e.g., intercellular adhesion molecule-1 [ICAM-1]) does not begin sooner than 2-4 h after reperfusion. [32,33]It is perhaps for this reason that a recent MCAO study found antileukocyte therapy (antiCD11b) to be equally effective when given at onset of reperfusion or when given 2 h later. Like MCAO, air embolism causes temporary cessation of cerebral blood flow followed by reperfusion. However, unlike MCAO, air emboli also cause endothelial injury. We propose endothelial injury differentiates cerebral air embolism from other models of central nervous system (CNS) ischemia/reperfusion and may initiate early leukocyte involvement.
After air embolism, morphologic endothelial damage ranges from seemingly mild (endothelial flattening) to obviously severe (denudation, exposure of basement membrane, and collagen). [12,13]However, even when appearing near normal, cerebrovascular endothelium is functionally impaired after air embolism. [8,9]Using a cranial window, Haller et al. observed cat cerebral arterioles that had undergone transient air embolism had reduced vasodilatory responses to carbachol, an acetylcholine receptor agonist. This suggests air-damaged endothelium may have reduced nitric oxide (NO) production. NO synthase inhibitors have been shown to induce rapid (10-30 min) leukocyte adhesion in mesenteric postcapillary venules. [35-38]Although it has been suggested that NO's primary role is to scavenge superoxide produced during reperfusion, [35,37,39]the proadhesive effect of NO synthase inhibition can be abolished by platelet-activating factor (PAF) and leukotriene B4(LTB4) antagonists, [36,38]as well as antibodies against P-selectin and leukocyte adhesion molecule CD18. [35,38]Thus, after air embolism, impaired endothelial NO production would seem likely to promote leukocyte adherence induced by any number of inflammatory mediators. [40,41]
With exposure of basement membrane, thrombin is activated. Thrombin causes rapid coexpression of endothelial P-selectin and PAF. In addition, transient hypoxia, oxygen radicals, and superoxide also induce endothelial expression of PAF or P-selectin. In vitro and in vivo [46,48-50]PAF causes rapid (10-30 min) high-affinity leukocyte adhesion to endothelium. Superoxide dismutase and NO donors inhibit PAF-induced leukocyte adherence. Thus, we propose in the setting of cerebral arterial air embolism, thrombin- or superoxide-induced PAF induction, in concert with decreased endothelial NO production, could result in a self-sustaining inflammatory process leading to rapid leukocyte adhesion. Two studies are consistent with this hypothesis. Kochanek et al. reported dogs pretreated with a PAF receptor antagonist had greater return of SSEP amplitude than control dogs after cerebral air embolism. In a previous experiment with this model, we found that heparin, which inhibits thrombin activity and leukocyte adhesion, reduced 5 h and 24 h neurologic impairment after cerebral arterial air embolism.
Limitations of this Experiment
This experiment has several important limitations. First, as previously noted, it does not address the mechanism by which doxycycline improved SSEP and early neurologic recovery. It is entirely possible doxycycline's protective effect was not mediated by inhibition of leukocyte activities [20-26]but instead through other unrecognized properties of the drug. Additional studies are needed to characterize the mechanisms of doxycycline protection in this and other models of CNS and non-CNS ischemia and reperfusion. [52,53]
Second, this experiment did not address the effect of doxycycline on long-term neurologic outcome. It seems likely delayed inflammatory processes, as observed after temporary MCAO, [31-34]could significantly affect long-term outcome after air embolism. Clark et al. recently reported doxycycline inhibited leukocyte infiltration into brain 24 h after forebrain ischemia as effectively as anti-ICAM-1. Thus, it seems possible doxycycline could reduce long-term and short-term inflammatory injury. Nevertheless, in this experiment, support was withdrawn from animals that failed to have spontaneous ventilation by 5 h. In previous studies with this model, neurologic impairment scores improved only an average of 10 points between 5 and 24 h after air embolism. [16,28]Hence, in this model, neurologic outcome appears to largely determined in the first few hours. We also have demonstrated that 2 h SSEP amplitude is predictive of 24 h outcome, with animals failing to have any SSEP recovery uniformly dying within 24 h. For these reasons, we considered it highly unlikely that those animals so neurologically impaired as to lack spontaneous ventilation 5 h after air embolism (all of whom had 4 h neurologic impairment scores of 96 and 0 SSEP recovery) would make any substantive neurologic recovery. Although doxycycline decreased electrophysiologic and neurologic impairment, it was only partially effective. Had we used a smaller dose of air, resulting in lesser neurologic impairment and greater survivability in control animals, the protective effect of doxycycline might not be demonstrable.
Another limitation of the current experiment is it cannot address how variation in doxycycline dose or timing of administration might alter its protective effect. The selected dose (10 mg/kg) was based on the work of Clark et al. These investigators showed this doxycycline dose decreased in vitro leukocyte adherence and conferred protection in a rabbit spinal cord ischemia model when given before reperfusion. Plasma doxycycline concentration 1 h after administration was reported as 1.7 micro gram/ml. This concentration is equivalent to that achieved after standard clinical doses of doxycycline. Thus, we speculate doxycycline could potentially serve as a readily available, inexpensive, and safe prophylactic therapy in settings at risk of cerebral arterial air embolism. Because of the limitations of our study, our findings should be considered preliminary. Additional mechanistic, dose-response, and long-term outcome studies are needed before clinical investigation can be considered.
In summary, doxycycline pretreatment reduced early electrophysiologic and neurologic abnormalities in a rabbit model of cerebral arterial air embolism. Because doxycycline resulted in greater SSEP recovery by 1 h, and SSEP amplitude inversely correlated with 4 h neurologic impairment, we conclude doxycycline inhibited a key early mediator of injury.
*There was no statistically significant correlation (n = 14) between temperature and 2 h SSEP amplitude (tau = -0.07) or 4 h neurologic impairment (tau = 0.30). To test for interaction between group assignment and temperature, for each animal, we calculated the difference between outcome and corresponding group median. There was no correlation between temperature and differences in 2 h SSEP amplitude (tau = -0.08) or 4 h neurologic impairment (tau = 0.14).