Analysis of spinal cord blood supply combining vascular corrosion casting and fluorescence microsphere technique: A feasibility study in an aortic surgical large animal model.

INTRODUCTION: Spinal cord ischemia after cardiovascular interventions continues to be a devastating problem in modern surgery. The role of intraspinal vascular networks and anterior radiculomedullary arteries (ARMA) in preventing spinal cord ischemia is poorly understood. MATERIALS AND METHODS: Landrace pigs (n = 30, 35.1 ± 3.9 kg) underwent a lateral thoracotomy. Fluorescent microspheres were injected into the left atrium and a reference sample was aspirated from the descending aorta. Repeated measurements of spinal cord and renal cortical blood flow from the left and right kidneys with three different microsphere colors in five pigs were taken to validate reproducibility. Spinal cord blood flow to the upper thoracic (T1-T4), mid-thoracic (T5-T8), lower thoracic (T9-T13), and lumbar (L1-L3) levels were determined. After euthanasia, we carried out selective vascular corrosion cast and counted the left and right ARMAs from levels T1-T13. RESULTS: Blood flow analysis of the left and right kidneys revealed a strong correlation (r = .94, p < .001). We detected more left than right ARMAs, with the highest prevalence at T4 (p < .05). The mean number of ARMAs was 8 ± 2. Their number in the upper thoracic region ranged from 2 to 7 (mean of 5 ± 1), while in the lower thoracic region they ranged from 0 to 5 (mean of 3 ± 1 [p < .001]). CONCLUSIONS: This study shows that combining fluorescence microsphere technique and vascular corrosion cast is well suited for assessing the blood flow and visualizing the arteries at the same time.

Spinal ischaemia after thoracic endovascular aortic repair with left subclavian artery sacrifice: is there a critical stent graft length?

OBJECTIVES: Thoracic endovascular aortic repair (TEVAR) is used for treatment of thoracic aortic pathologies, but the covered stent graft can induce spinal ischaemia depending on the length used. The left subclavian artery contributes to spinal cord collateralization and is frequently occluded by the stent graft. Our objective was to investigate the impact of covered stent graft length on the risk of spinal ischaemia in the setting of left subclavian artery sacrifice. METHODS: Twenty-six pigs (German country race, mean body weight 36 ± 4 kg) underwent simulated descending aortic TEVAR via left lateral thoracotomy, with left subclavian artery and thoracic segmental artery occlusion in normothermia. Animals were assigned to treatment groups according to simulated stent graft length: TEVAR to T8 (n = 4), TEVAR to T9 (n = 4), TEVAR to T10 (n = 4), TEVAR to T11 (n = 7) and TEVAR to T12 (n = 1) and a sham group (n = 6). End points included spinal cord perfusion pressure, cerebrospinal fluid pressure and spinal cord blood flow using fluorescent microspheres. RESULTS: There were no group differences in spinal cord perfusion pressure drop or in spinal cord perfusion pressure regeneration potential at 3 h after the procedure: from a baseline average of 75 mmHg (95% confidence interval 71-83 mmHg) to 73 mmHg (67-75 mmHg) at 3 h in Group T10 versus from a baseline average of 67 mmHg (95% CI 50-81 mmHg) to 65 mmHg (95% confidence interval 48-81 mmHg) in Group T8. There were no differences in the spinal cord blood flow courses over time in the different groups nor was there any difference in cerebrospinal fluid pressure levels and cerebrospinal fluid pressure dynamics between groups. However, we did observe local blood flow distribution to the spinal cord that was inhomogeneous depending on the distance between the simulated stent graft end and the first thoracic anterior radiculomedullary artery. CONCLUSIONS: The risk of spinal ischaemia after serial segmental artery occlusion does not depend on the distal extent of the aortic repair alone. Future attempts to allow patient risk stratification for spinal ischaemia need to focus on anterior radiculomedullary artery anatomy together with the extent of planned aortic repair.

Twenty minutes of normothermic cardiac arrest in a pig model: the role of short-term hypothermia for neurological outcome.

INTRODUCTION: Cardiopulmonary resuscitation restores circulation, but with inconsistent blood-flow and pressures. Our recent approach using an extracorporeal life support system, named "controlled integrated resuscitation device" (CIRD), may lead to improved survival and neurological recovery after cardiac arrest (CA). The basic idea is to provide a reperfusion tailored to the individual patient by control of the conditions of reperfusion and the composition of the reperfusate. Hypothermia is one aspect of this concept. Here, we investigated the role of immediate short-term blood cooling after experimental CA and its influence on survival and neurological recovery. METHODS: Twenty-one pigs were exposed to 20 minutes of normothermic CA. Afterwards, CIRD was immediately started for 60 minutes in all animals and the heart was converted to a sinus rhythm. The pigs either received normothermic reperfusion (37°C, n=11) or the temperature was maintained at 32°C for the first 30 minutes (n=10). Thermometric, hemodynamic and serologic data were collected during the experiment. After weaning from CIRD, neurological recovery was assessed daily by a species-specific neurological deficit score (NDS; 0: normal; 500: brain death). RESULTS: One pig in each group could not be successfully resuscitated. Due to severe neurological deficits, only 6/11 animals in the normothermic group finished the observation time of seven days with an NDS of 37±34. In the hypothermic group, all nine surviving animals reached day seven with an NDS of 16±13. Analogous to the lower NDS, animals in the hypothermic group also showed lower neuron-specific enolase end values as a marker of brain injury. CONCLUSIONS: Within this experimental setting, immediate moderate and short-term hypothermia after CA improves survival and seems to result in statistically non-significant better neurological recovery.

Spinal Ischemia in Thoracic Aortic Procedures: Impact of Radiculomedullary Artery Distribution.

BACKGROUND: The aim of this study was to assess the influence of thoracic anterior radiculomedullary artery (tARMA) distribution on spinal cord perfusion in a thoracic aortic surgical model. METHODS: Twenty-six pigs (34 ± 3 kg; study group, n = 20; sham group, n = 6) underwent ligation of the left subclavian artery and thoracic segmental arteries. End points were spinal cord perfusion pressure (SCPP), regional spinal cord blood flow (SCBF), and neurologic outcome with an observation time of 3 hours. tARMA distribution patterns tested for an effect on end points included (1) maximum distance between any 2 tARMAs within the treated aortic segment (0 or 1 segment = small-distance group; >1 segment = large-distance group) and (2) distance between the end of the treated aortic segment and the first distal tARMA (at the level of the distal simulated stent-graft end = group 0; gap of 1 or more segments = group ≥1). RESULTS: The number of tARMA ranged from 3 to 13 (mean, 8). In the large-distance group, SCBF dropped from 0.48 ± 0.16 mL/g/min to 0.3 ± 0.08 mL/g/min (p < 0.001). We observed no detectable SCBF drop in the small-distance group: 0.2 ± 0.05 mL/g/min at baseline to 0.23 ± 0.05 mL/g/min immediately after clamping (p = 0.147). SCBF increased from 0.201 ± 0.055 mL/g/min at baseline to 0.443 ± 0.051 mL/g/min at 3 hours postoperatively (p < 0.001) only in the small-distance group. CONCLUSIONS: We demonstrate experimental data showing that distribution patterns of tARMAs correlate with the degree of SCBF drop and insufficient reactive parenchymal hyperemia in aortic procedures. Individual ARMA distribution patterns along the treated aortic segment could help us predict the individual risk of spinal ischemia.

Preserved brain morphology after controlled automated reperfusion of the whole body following normothermic circulatory arrest time of up to 20 minutes.

OBJECTIVES: Clinical outcomes following cardiac arrest (CA) and resuscitation remain a cause for concern. The use of Controlled Automated Reperfusion of the whoLe body (CARL) confers superior neurological outcome even after extended periods of CA. We aimed at investigating clinical outcome and brain morphology preservation when employing CARL following CA periods of 20 min. METHODS: Twenty-eight pigs were allocated to four extracorporeal circulation treatment strategies; seven others served as magnetic resonance imaging (MRI) controls. In prompt cardiopulmonary resuscitation (CPR; n = 6), induced circulatory arrest was followed immediately by open cardiac massage of 15 min, thereafter by CARL for 60 min. In delayed CPR (n = 6), induced CA was maintained for 15 min, after that open cardiac massage of 10 min duration was performed prior to extracorporeal CPR (ECPR) of 60 min. Induced CA times of 15 min in the ECPR 15' group (n = 6) and CA of 20 min in the CARL 20' group (n = 10) were followed by ECPR of 60 min and CARL of 60 min, respectively, without prior CPR. Daily neurological deficit scoring (NDS) up to the seventh day, markers of cellular injury [alanine transaminase (ALT), aspartate transaminase (AST) and neuron-specific enolase (NSE)] and brain MRI were performed. RESULTS: 100% survival and normal NDSs were achieved in all animals in the prompt CPR and ECPR 15' groups. In CARL 20', nine animals survived. In contrast, only one animal in the delayed CPR group survived; three animals died within 24 h with a further two dying on Days 4 and 5, respectively. All markers of cellular injury were elevated in the delayed CPR group, ALT [38 (20.3) to 206 U/l (158.2); P = 0.0095], AST [26 (18.8) to 97 U/l (1965.8); P = 0.0095] and NSE [0.45 (0.25) to 7.95 µg/l (24.03); P = 0.0095]. In the ECPR 15' group, only NSE [0.45 (0.15) to 1.20 µg/l (2.40); P = 0.0065] remained elevated. In the CARL 20' group, differences in ALT [36 (10) to 53 U/l (20); P = 0.0005] and NSE [0.50 (0.40) to 1.5 µg/l (0.40); P < 0.0001] values were evident. T2-weighted MR images of the cerebellum [454 (28) to 495 mm2/s (55); U = 11; P = 0.0311], caudate nucleus [400 (59) to 467 mm2/s (42); U = 9; P = 0.0156], lentiform nucleus [377 (89) to 416 mm2/s (55); U = 11; P = 0.0311] and hippocampus [421 (109) to 511 mm2/s (58); U = 9; P = 0.0164] in the CARL 20' group showed higher signal intensities compared with controls. In delayed CPR, corresponding regions of interest on early apparent diffusion coefficient images showed a restricted diffusion. CONCLUSIONS: In our experimental animal model of CA, CARL results in satisfactory survival at CA periods of 20 min despite detected enzyme and morphological changes. These changes did not translate to clinical neurological deficits.

Immediate Spinal Cord Collateral Blood Flow During Thoracic Aortic Procedures: The Role of Epidural Arcades.

The objective of this study was to investigate the functional differences between paraspinal and intraspinal compartments of the spinal collateral network and the importance of circular epidural arcades in thoracic aortic surgery. N = 33 pigs (mean body weight: 34 ± 3kg) were included. A single-inlet-model of spinal collateral flow was created: paraspinal inflow into the collateral network was isolated by cephalad and caudal interruption of inflow into epidural arcades using laminectomies. Animals were assigned to treatment groups (Treatment "open" [patent epidural arcades, n = 10] and Treatment "closed" [closed epidural arcades, n = 10]) and Sham groups (Sham "open" n = 8 and Sham "closed" n = 5). Treatment was a simulated Frozen Elephant Trunk procedure with occlusion of left subclavian and thoracic segmental arteries under mild permissive hypothermia. Observation time was 3 hours. Endpoints were motor and somatosensory evoked potentials (motor evoked potentials and sensory evoked potentials), spinal cord perfusion pressure, cerebrospinal fluid pressure, regional spinal cord blood flow, and neurologic outcome. Animals with interrupted inflow into epidural arcades (Group Treatment "closed") had higher cerebrospinal fluid pressure levels (P < 0.05), were not able to maintain sufficient spinal cord perfusion pressure during Frozen Elephant Trunk procedure (P < 0.001) and did not generate reactive hyperemia as did group Treatment "open." spinal cord blood flow was strongly decreased in group Treatment "closed" (P < 0.001) at 0 hour, did not recover out to 3 hours of observation and 90% of the animals suffered flaccid paraplegia (P < 0.05). Immediate spinal cord backup blood flow is almost exclusively delivered using the system of epidural arcades in the immediate setting, serving as an immediate backup system. Intraspinal arcades are responsible for generating sufficient intraspinal perfusion pressures, reactive hyperemia, and spinal cord integrity. Paraspinal collaterals might need to undergo arteriogenesis, and thus serve as a long-term backup system.