Experimental Two-Stage Simulated Repair of Extensive Thoracoabdominal Aneurysms Reduces Paraplegia Risk

Transcription

1 Experimental Two-Stage Simulated Repair of Extensive Thoracoabdominal Aneurysms Reduces Paraplegia Risk Stefano Zoli, MD, Christian D. Etz, MD, PhD, Fabian Roder, MS, Robert M. Brenner, MS, Carol A. Bodian, DrPH, George Kleinman, MD, Gabriele Di Luozzo, MD, and Randall B. Griepp, MD Departments of Cardiothoracic Surgery, Anesthesiology, and Neuropathology, Mount Sinai School of Medicine, New York, New York Background. In a pig model, we compared spinal cord injury after extensive segmental artery (SA) sacrifice in a single stage with recovery after a two-stage procedure: lumbar artery followed by thoracic SA sacrifice. Methods. Twenty juvenile Yorkshire pigs were randomly assigned to undergo extensive SA sacrifice at 32 C in a single operation (group 1, n 10), or thoracic SA ligation 7 days after lumbar artery sacrifice (group 2, n 10). Spinal cord perfusion pressure (SCPP) was monitored using a catheter placed in the distal stump of L1. Hind limb function was evaluated intraoperatively using motor-evoked potentials and for 5 days postoperatively using a modified Tarlov score. Results. Motor-evoked potentials were intact in all pigs until 1 hour after surgery. All pigs in group 2 fully recovered hind limb function, whereas 40% in group 1 experienced paraplegia (median Tarlov scores 9 versus 7; p 0.004). Group 1 SCPP fell to 28 6 mm Hg after SA sacrifice, compared with 44 8 mm Hg in group 2 (p < ). After sacrifice of all residual SAs, SCPP in group 2 remained consistently greater than 85% of baseline, significantly higher than group 1 SCPP from end clamping until 72 hours (p ). Histopathologic analysis showed more severe ischemic damage to the lower thoracic (p < 0.001) and lumbar spinal cord (p 0.01) in group 1. Conclusions. In contrast with the single-stage approach, a two-stage procedure, starting with ligation of six or fewer lumbar SAs, leads to only a mild drop in SCPP and stimulates vascular remodeling, minimizing the impact of subsequent SA sacrifice on spinal cord function. The greater safety of extensive SA sacrifice when undertaken in two stages has important implications for endovascular and hybrid aneurysm repair. (Ann Thorac Surg 2010;90:722 30) 2010 by The Society of Thoracic Surgeons Accepted for publication April 12, Presented at the Poster Session of the Forty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 25 27, Address correspondence to Dr Zoli, Department of Cardiothoracic Surgery, Mount Sinai School of Medicine, One Gustave L. Levy Pl, New York, NY 10029; stefanozoli@gmail.com. During the past few decades, considerable effort has been expended in searching for strategies to prevent spinal cord ischemic damage after surgical and endovascular repair of extensive thoracoabdominal (TAAA) aneurysms. A core body temperature of 32 C has been found [1] in experimental settings to be effective in increasing spinal cord ischemic tolerance by reducing metabolic requirements, and almost all aortic surgeons now perform TAAA resection under at least moderate hypothermia. A catheter for cerebrospinal fluid drainage is usually placed during aortic surgery to enhance spinal cord perfusion [2, 3] and, with increasing frequency, even when extensive aortic coverage is carried out with stent grafts [4, 5]. Several perfusion strategies have been reported to be effective in reducing the risk of intraoperative ischemia and hence paraplegia [6], including use of profound hypothermia [7]. However, the belief that reimplantation of intercostal arteries may be the key factor in preventing spinal cord infarction continues to be widespread [8, 9]. We have devoted our efforts to achieving a better understanding of the collateral circulation to optimize its response not only after surgical repair, but also in conjunction with endovascular strategies, in which revascularization of segmental arteries (SAs) is essentially precluded. Our laboratory and clinical observations support the existence of multiple inputs into the anterior spinal artery. This impressive collateral network is able to increase blood flow to the spinal cord when one or another source is reduced, and usually allows serial sacrifice of segmental vessels without significant spinal cord ischemia either during aortic crossclamping or thereafter. The ability of the collateral circulation to compensate for loss of input is, however, limited. In consequence, acceptably low paraplegia rates are expected for limited aortic resection, but when sacrifice of a large number of intercostal and lumbar arteries is undertaken, the incidence of spinal cord ischemia is still far from ideal in clinical [10] as well as in experimental [11] situations. Our studies thus far have suggested that the vascular remodeling necessary to restore an adequate spinal cord 2010 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg ZOLI ET AL 2010;90: STAGED APPROACH FOR TAAA REPAIR blood flow after extensive sacrifice of SAs may be a time-dependent process, requiring an interval of 3 to 5 days to fully develop. Thus, what may be precarious compensation by the collateral system in the immediate postoperative period may become a stable basis for perfusion within a remarkably short time. On the basis of these findings, we sought to exploit the capacity of the collateral network to mount a response after an ischemic insult to reduce the risk of paraplegia after extensive TAAA repair. In a chronic porcine model, we evaluated clinical recovery, spinal cord perfusion pressure (SCPP), and histopathologic findings after complete SA sacrifice performed in a staged fashion: lumbar artery sacrifice, followed after 7 days by intercostal artery ligation. These results were compared with a similar group of animals that underwent the same extent of SA resection in a single stage. Material and Methods Study Design Twenty female juvenile Yorkshire pigs (about 3 months old, Animal Biotech Industries Inc, Danboro, PA), weighing kg, were randomly allocated into two groups of 10 animals each. Group 1 underwent complete stepwise craniocaudal SA sacrifice under moderate hypothermia in a single operation, and group 2 underwent the same SA resection in two stages. The first step of the two-stage operation involved the sacrifice of the lumbar arteries, followed by ligation of thoracic SAs after a 7-day recovery period. Intraoperative spinal cord monitoring was performed with motor-evoked potentials (MEPs) and with direct measurement of what we consider the SCPP through a catheter placed in the distal stump of an SA. The rectal target temperature of 32 C was reached by means of cooling blankets and operating room temperature reduction; MEP-monitored segmental vessel sacrifice with a 3-minute interval between successive arteries was then undertaken. Spinal cord perfusion pressure, as well as arterial blood gas data, was collected at baseline, immediately after SA sacrifice, at 1 and 5 hours after the procedure, and then once a day for 5 days. Intermittent postoperative hind limb function evaluation using a modified Tarlov score was also continued for 5 days. All animals received humane care in accordance with the guidelines from the Principles of Laboratory Animal Care formulated by the National Society for Medical Research, and the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. The Mount Sinai Institutional Animal Care and Use Committee reviewed and approved the protocol for this research. Perioperative Management and Anesthesia After placement of a 20-gauge catheter in a suitable ear vein and insertion of an endotracheal tube, the animals were transferred to the operating room. Mechanical ventilation was initiated with an inspired fraction of oxygen of 0.7 and a minute volume adequate to maintain 723 partial pressure of carbon dioxide values between 30 and 40 mm Hg. To obtain comparable variables among animals and groups, crystalloid infusion was started and eventually increased to reach a mean arterial pressure (MAP) of 85 to 90 mm Hg. Anesthesia was induced by the bolus intravenous administration of propofol (1 mg/kg) and fentanyl (50 g/kg), and was maintained with continuous infusions of ketamine (15 mg kg 1 h 1 ), propofol (7 mg kg 1 h 1 ), and fentanyl (5 g kg 1 h 1 ). To avoid interference with MEP monitoring, use of pancuronium was limited to a small dose for anesthesia induction. An 8F Foley catheter was placed in the bladder for continuous monitoring of urine output, and temperature probes were placed in the rectum and esophagus. Electrocardiographic monitoring was continuous. Systemic arterial pressure was monitored by direct placement into the descending or abdominal aorta of an arterial catheter. Heart rate, ph, partial pressure of oxygen, partial pressure of carbon dioxide, hemoglobin, and glucose concentrations (Ciba Corning 865; Chiron Diagnostics, Norwood, MA) were carefully monitored and recorded, and maximal care was taken to maintain these values within the normal range. Motor-Evoked Potential Monitoring A 3-cm midline scalp incision was made to reveal the sagittal and coronal sutures. Four stainless-steel electrodes were screwed into the skull: two were placed 1 cm to the left and two 1 cm to the right of the sagittal suture; the anterior one of each pair was positioned 1 cm in front of the coronal suture; the other 1 cm behind the suture. The electrodes were connected to an electrical stimulator (Digitimer Stimulator Model D 180A, Welwyn Garden City, UK). Electromyographic recordings were made from stainless-steel needle electrodes placed into the anterior and posterior muscle compartments of both fore and hind limbs. A stimulation train (3 pulses: 200 to 300 V, 100-ms pulse duration with a 2-ms interstimulus pause) was delivered to the skull electrodes, and the MEPs were recorded in the limbs. The MEP signal was amplified (gain 2,000), bandpass filtered (10 to 1,000 Hz), and evaluated using a Spectrum 32 neurophysiologic recording system (Cadwell Laboratories Inc, Kennewick, WA). Operative Technique and Spinal Cord Perfusion Pressure Monitoring All animals in group 1 underwent a concomitant left posterolateral thoracotomy and a linear left flank incision from the costovertebral angle to the posterior aspect of the iliac crest. The thoracic aorta was mobilized from the left brachiocephalic artery down to the diaphragm, and through a retroperitoneal approach, the abdominal aorta was exposed from the distal extent of the thoracic dissection down to the vessel s trifurcation to reveal the SAs at their origins. These vessels arise on the left side dorsolaterally, usually as a single vessel that later divides into right and left branches. The first lumbar artery was selected for SCPP catheter placement for its location in the anticipated center of SA sacrifice, and therefore the

3 724 ZOLI ET AL Ann Thorac Surg STAGED APPROACH FOR TAAA REPAIR 2010;90: Fig 1. Functional recovery after extensive segmental artery sacrifice in group 1 (n 10) versus group 2 (n 10). The modified Tarlov score used for neurobehavioral evaluation is represented on the y axis. Animals with a score greater than 6 (dashed line) at the end of the 5-day protocol were considered to have recovered. site in which the most significant SCPP drop was anticipated to occur. After careful dissection, the catheter (3F pressure catheter; Medtronic Inc, Minneapolis, MN) for SCPP monitoring was placed into the distal end of the proximally ligated L1 and connected for intraoperative and postoperative pressure monitoring. The distal end of the catheter was then passed through the muscles of the back, and was secured in place with silk sutures; patency of the catheters was maintained with frequent intermittent flushes of heparin. During the preparations described, the pigs were actively cooled using heating and cooling blankets, aiming for mild hypothermia (32 C). Once the target temperature had been reached, baseline MEPs were assessed and SCPP measurement was obtained. The SAs were sacrificed in a craniocaudal sequence one at a time, every 3 minutes. The scalp electrodes were then detached, the incisions were closed, and the pig was weaned from the ventilator. An average of SAs were sacrificed in group 1. Animals allocated to group 2 underwent SA sacrifice in two steps. During the first operation, only the abdominal aorta was accessed through the left flank incision; an SCPP catheter was placed into the distal stump of the first divided lumbar artery, and an average of vessels were subsequently taken. The second procedure required only a left thoracotomy, and after assessment of the patency of the SCPP catheter inserted 1 week before, sacrifice of the residual SAs was performed in the usual manner; SAs were taken during the second stage, for a total of interrupted overall in group 2. Postoperative Neurologic Outcome Hind limb function was evaluated after each operation once a day for 5 days using a modification of the Tarlov score we previously conceived to obtain a more sophisticated neurologic evaluation. The modified Tarlov score is scaled as follows: no voluntary movements (0); perceptible movements at joints (1); good movements at joints but inability to stand (2); ability to get up and stand with assistance less than 1 minute (3); ability to get up with assistance and stand unassisted less than 1 minute (4); ability to get up with assistance and stand unassisted more than 1 minute (5); ability to get up and stand unassisted more than 1 minute (6); ability to walk less than 1 minute (7); ability to walk more than 1 minute (8); complete recovery (9). Each animal was videotaped, and blind evaluation of the modified Tarlov score was carried out. Histopathologic Evaluation of Spinal Cord Damage After the animals were sacrificed, the spinal cord was surgically harvested, fixed in 10% formalin solution for 3 days, and then sectioned transverse to the craniocaudal axis, with samples including each spinal nerve root from C1 all the way down to S1. Sections 6 m in thickness were stained with hematoxylin and eosin and examined and scored blindly by an experienced neuropathologist (G.K.) according to a schematic grading system. This grading system was developed and published earlier by our group [11], and allows a refined assessment of both gray and white matter, with a score ranging from 0 (no damage) to 8: 1 necrosis of single (motor) neurons only; 2 necrosis of one posterior horn only; 3 necrosis of both posterior horns only; 4 necrosis of both posterior horns plus surrounding white matter; 5 necrosis of both anterior horns only; 6 central necrosis involving posterior and anterior horns plus parts of white matter; 7 complete necrosis of gray matter only; 8 complete necrosis of the whole section. Statistical Methods The animals were randomly assigned to one of the two surgical procedures in accordance with a schedule prepared by the study statistician (C.B.). The group assignment was revealed just before the induction of anesthesia. Data were entered in Excel (Microsoft Corp, Redmond, WA) spreadsheets and transferred to an SAS (SAS Institute, Cary, NC) file for data description and analysis. Variables are described as mean ( standard deviation), median (range) or percents. 2 tests or Wilcoxon rank-sum tests were used to compare groups with regard to values at individual time points or averages over specified ranges of times. The mixed model for repeated measures was used for comparisons during a time period.

4 Ann Thorac Surg ZOLI ET AL 2010;90: STAGED APPROACH FOR TAAA REPAIR 725 Recovery of Spinal Cord Function As seen in Figure 1, 6 animals (60%) in group 1 regained normal spinal cord function, and by day 5 were able to get up without assistance and to walk for more than 1 minute with an average score of (median, 8; range, 7 to 9). Despite intact intraoperative MEPs, 4 animals (40%) in group 1 exhibited paraplegia (average Tarlov score, ; median, 1; range, 1 to 4). The average score for group 1 as a whole was , with a median of 7, and a range between 1 and 9. All pigs in group 2 fully regained hind limb function; they quickly recovered from the first procedure, reaching the maximal behavioral score within 48 hours. After sacrifice of all residual SAs was completed, they were all able to get up and to walk without assistance. The average modified Tarlov score after the 5-day protocol was (median, 9; range, 8 to 9), significantly higher (p 0.004) than for group 1. Hemodynamic and Metabolic Variables In view of its key role in providing adequate spinal cord perfusion, similar MAP values (group 1, 89 6mmHg; group 2, 90 9 mm Hg; not significant) were obtained at baseline in both groups, and MAP was maintained greater than 90 mm Hg throughout the entire procedure as well as during the recovery period. The same care was taken to obtain comparable values for arterial oxygen and carbon dioxide tensions. Statistical analysis found no significant differences between groups (Fig 2). Other relevant intraoperative and postoperative variables are detailed in Table 1. The expected target temperature of 32 C was easily reached and maintained in both groups ( C in group 1; C in group 2) without any significant difference. Fig 2. Intraoperative and postoperative mean arterial pressure (MAP; A), partial pressure of oxygen (po 2 ; B), and partial pressure of carbon dioxide (pco 2 ; C) in group 1 (solid symbols; n 10) versus the second stage of group 2 (open symbols; n 10). Repeated measures analysis of variance detected no statistically significant differences (n.s.) between groups. Note the mean arterial pressure peak at 5 hours in conjunction with the awakening process. Results All animals in group 2 survived the first surgical procedure, the recovery period, and the subsequent second stage as planned in the study design. Data reported in the study refer to the second procedure, when all the SAs had been sacrificed. One animal in group 1 died on postoperative day 1 with severe pleural effusions and respiratory insufficiency, and was therefore replaced. Motor-evoked potential monitoring was carried out until 1 hour after surgery, and no significant variations from baseline were observed in any pig. Animals with a modified Tarlov score less than 7 at the end of the protocol were considered to have sustained significant spinal cord damage (paraplegia or paraparesis) [12]. Number of Segmental Vessels Sacrificed The number of segmental vessels arising from the aorta that could be identified ranged from 14 to 15, with a Fig 3. Intraoperative and postoperative spinal cord perfusion pressure (SCPP) measured at the level of L1 after complete segmental artery (SA) sacrifice in group 1 (solid symbols; n 10) versus lumbar artery sacrifice in group 2 (open symbols; n 10). After sacrifice of an average of lumbar arteries, end clamp spinal cord perfusion pressure in group 2 showed only a mild drop, whereas in group 1 complete segmental artery sacrifice caused a dramatic reduction in spinal cord perfusion (p ).

5 726 ZOLI ET AL Ann Thorac Surg STAGED APPROACH FOR TAAA REPAIR 2010;90: Table 1. Intraoperative and Postoperative Metabolic Indicators for Group 1 (n 10) and Group 2 (n 10) Variable/Group Baseline End Clamp 1 Hour 5 Hours 24 Hours 48 Hours 76 Hours 96 Hours 120 Hours ph (at 37 C) Group Group HCO 3 (meq/l) Group Group Hemoglobin (g/dl) Group Group Hematocrit (%) Group Group Glucose (mg/dl) Group Group Lactate (mg/dl) Group Group HCO 3 bicarbonate. median of 14 for both groups. In group 1 (single stage), the average number of SAs sacrificed was ; in group 2 (two-stage) after both procedures, it was There was no statistically significant difference between groups with regard to the overall number of SAs sacrificed. Spinal Cord Perfusion Pressure in Group 1 The baseline SCPP in group 1 was 66 6 mm Hg, 75% of the MAP measured concurrently. Immediately after complete SA sacrifice, SCPP decreased to 28 6 mm Hg, and the decrease continued until it reached its lowest point at 5 hours (26 6 mm Hg; 29% of contemporaneous MAP) with values less than 40% of baseline. A slow but steady recovery period followed, and SCCP was back in the range of baseline (65 10 mm Hg) after the 5-day postoperative interval. Spinal Cord Perfusion Pressure in Group 2 (1st Stage) Spinal cord perfusion pressure at baseline in group 2 was very close to the one recorded in group 1 (69 6mmHg; no difference between the groups). When expressed as a percentage of baseline MAP, SCPP was identical in both groups (75% of baseline MAP). In contrast with the marked drop observed in group 1, SCPP in group 2 decreased only to 44 8mmHg (65% of baseline) after sacrifice of a relatively small number of SAs (Fig 3). The difference in SCPP decrease between the two groups after the initial operation was highly significant (p ). The time needed for the collateral network to restore SCPP to baseline after lumbar artery sacrifice was only 72 hours (67 6mm Hg at 3 days). Spinal Cord Perfusion Pressure in Group 2 (2nd Stage) Seven days after the first procedure, the pigs in group 2 were brought back to the operating room, a left thoracotomy was performed as described earlier, and an updated intraoperative SCPP baseline was obtained. One week after complete lumbar SA flow interruption, the SCPP recorded at the level of L1 was 66 5 mm Hg (72% of baseline MAP), the same value that was recorded at baseline in group 1. Sacrifice of all residual SAs was undertaken, and only a negligible decrement in SCPP was observed, with its lowest value of 57 4mmHg1 hour after clamping (64% of contemporaneous MAP). Spinal cord perfusion pressure after the second stage of complete SA sacrifice remained consistently greater than 85% of baseline, which was significantly higher than group 1 during the period from end clamping until 72 hours (p ; Fig 4). Histopathologic Findings Adequate spinal cord samples for histopathologic analysis were obtained in all animals. The entire cord was divided for descriptive purposes and for statistical analysis into three distinct segments: cervical and upper thoracic (spanning from the first cervical spinal nerve root to the level of T7), lower thoracic (from T8 to T13), and abdominal portion (including levels from L1 down to S1). The cervical and upper thoracic area, receiving the major part of their blood supplies from proximal inputs that were not interrupted during the current study, were almost completely free of damage (only 1 animal showed punctuate necrosis of a few motor neurons in a single segment), and were not the subject of further analysis. Although 6 animals were considered to have recovered functionally, being able to walk after the procedure, histopathologic evidence of ischemic

6 Ann Thorac Surg ZOLI ET AL 2010;90: STAGED APPROACH FOR TAAA REPAIR 727 some evidence that prior abdominal aortic aneurysm resection is an important risk factor for neurologic deficit after subsequent TAA or TAAA repair [14], we suspect that the reason this occurs is because of hypogastric artery rather than lumbar artery sacrifice at the time of the initial operation. Other surgeons have noted a protective effect of previous aortic surgery [15]; in 723 patients who underwent open TAAA repair, Coselli and associates [16] observed a trend suggesting that previous TAA repair may help prevent spinal cord injury. Fig 4. Intraoperative and postoperative spinal cord perfusion pressure (SCPP) measured at the level of L1 after complete segmental artery (SA) sacrifice. Spinal cord perfusion pressure in group 2 (open symbols) is quite stable, showing only a mild drop ( 14% from baseline) after sacrifice of all the residual segmental arteries. In contrast, ligation of a comparable number of segmental arteries in a single stage led to a dramatic drop in spinal cord perfusion pressure in group 1 (solid symbols) that reached its lowest point (26 mm Hg) at 5 hours. injury was detected in all but 1 pig in group 1 (Fig 5). The lower thoracic and the abdominal segments presented severe and diffuse signs of spinal cord injury, often including complete necrosis of an entire section and involving several adjacent segments. The average score was (median, 0.5; range, 0 to 8) for the lower thoracic segments, and (median, 0; range, 0 to 8) for the abdominal segments. As previously detailed, all animals in group 2 fully recovered hind limb function, and only 4 animals showed traces of ischemic spinal cord damage: lesions were of minimal extent and without involvement of white matter. The degree of necrosis presented a statistically significant difference from what was seen in group 1 in both the lower thoracic (average score in group 2, ; median, 0; range, 0 to 3; p 0.001) and abdominal areas (average score in group 2, ; median, 0; range, 0 to 3; p 0.01). Comment The impetus for the present study came both from clinical experience and experimental findings, which during the years have consistently supported our theory of the major role played by a large muscular and epidural collateral network in providing blood flow to the spinal cord after descending thoracic aortic aneurysm (TAA) and TAAA surgery. In 2009, we reported our experience in a group of 60 patients who underwent TAA or TAAA repair through a redo lateral thoracotomy. We made the observation that even if the second operation resulted in total TAAA replacement, the risk of spinal cord injury was apparently not enhanced [13]. The average interval between the two procedures was more than 7 years, a period greatly exceeding what we believe is the minimum necessary to allow compensatory vascular remodeling after SA sacrifice. Although a recent meta-analysis reported Fig 5. Spinal cord ischemic damage according to the Kleinman scoring system (shown on the x axis along the top with a corresponding picture of the hematoxylin and eosin stained spinal cord) after complete sacrifice under moderate hypothermia of all segmental arteries in group 1 (black bars; single stage, n 10) and group 2 (striped bars; two-stage, n 10). The y axis depicts the position from which the cord sections were taken from C1 (upper left corner) down to S1 (lower left corner). Although 6 animals retained spinal cord function, signs of ischemic injury were detected in all but 1 pig in group 1. In group 2, only 4 animals showed mild traces of spinal cord damage, apparently not serious enough to jeopardize hind limb function. Statistical analysis detected more severe ischemic damage to the lower thoracic (p 0.001) and lumbar spinal cord (p 0.01) in group 1.

7 728 ZOLI ET AL Ann Thorac Surg STAGED APPROACH FOR TAAA REPAIR 2010;90: Experimental studies in rabbit and pig models suggest that the main triggers stimulating the growth of functional collateral arteries are physical forces, such as altered blood pressure. As a result of a complex cascade of events involving vascular cell proliferation and wall remodeling, an overall increment in blood flow may be achieved through the structural enlargement of preexisting collateral arterioles [17, 18]. To properly assess the process of collateral circulation adaptation, we undertook direct measurement of what we believe is the true SCPP by means of a catheter inserted into the distal end of the first lumbar artery (L1). The position of the catheter was chosen because L1 is approximately in the middle of SA resection, in a location as remote as possible from subclavian and hypogastric artery inflow. We wanted to increase the possibility of identifying even minimal pressure changes caused by SA sacrifice and its subsequent recovery mediated by vascular remodeling by minimizing the possible confounding impact of proximity to these known and well-established proximal and distal inputs into the collateral system. Our group has used direct SCPP monitoring [11] previously, and the results obtained from that first experience served as a guide for the planning of the current study. From observing SCPP in a group of 10 animals undergoing extensive SA sacrifice, we had already established that complete restoration of SCPP is achievable within a relatively short interval. The valuable insight obtained from this earlier experimental study prompted us to extend the use of direct SCPP monitoring to clinical practice, and a preliminary report in which a catheter was placed for intraoperative and postoperative analysis of pressure within the collateral network in a small group of patients has recently been published [19]. Spinal cord perfusion pressure monitoring has shown excellent sensitivity, allowing recording with good accuracy of even minimal pressure changes, and thus confirmed its value as a potentially powerful and not very invasive new tool for assessing adequacy of spinal cord perfusion. Baseline SCPP was similar between the two groups, both when expressed as an absolute value and as percentage of MAP. The finding that baseline SCPP is about 70% of MAP is consistent with the results obtained from other stump pressure measurements routinely used, eg, for monitoring during surgical carotid endarterectomy [20, 21]. In group 2, sacrifice of a limited number of SAs, even in a location considered particularly relevant to spinal cord ischemia, led to only a mild drop in SCPP, which consistently remained well above 65% of baseline, and recovered fully within a short period. In contrast, complete SA sacrifice in group 1 triggered a tremendous decrement in SCPP (to 26 mm Hg at 5 hours). Despite this dramatic intraoperative fall, subsequent vascular remodeling was able to restore SCPP values comparable to baseline after a 5-day period. The precariousness of compensation during the immediate postoperative period after all segmental blood flow sources are interrupted is highlighted by the failure of recovery of spinal cord function observed in 4 animals in group 1 despite intact intraoperative MEPs. We were not surprised that all the animals emerged from complete SA sacrifice with intact MEPs, which remained unchanged for as long as this monitoring could be carried out. Careful intraoperative hemodynamic management that guarantees a stable MAP of approximately 90 mm Hg has yielded similar results in previous experiments [11]. We have previously observed that the interval of rewarming and awakening from anesthesia is the time when ischemia and spinal cord injury are likeliest to occur [22]. As intraoperative management has become more sophisticated, an increasing proportion of clinical cases of paraplegia are delayed rather than immediate, underscoring the importance of postoperative perfusion. A possible predominant role of a delayed ischemic insult in the development of irreversible paraplegia has increasingly also been suggested by others [23]. One observation suggested by our findings is that less reduction of segmental inflow leads to a smaller decrease in SCPP, posing not as much of a challenge requiring intraoperative and postoperative compensation: this explains the clinical evidence that short aortic resections are at low risk of paraplegia regardless of location. The fact that more than half the pigs that underwent complete SA sacrifice fully regained hind limb function despite the lack of important adjuncts such as cerebrospinal fluid drainage is itself an interesting observation. We speculate that the sequential ligation of SAs with a certain interval between successive vessels may trigger mechanisms of early ischemic preconditioning, proposed to have a role in protecting against spinal cord injury in a similar animal model [24]. Further studies may clarify whether there is indeed a difference between brisk and slow stepwise SA sacrifice with regard to spinal cord viability. This is of particular relevance in relation to possible endovascular strategies. The most fascinating finding of the present study is the stability of SCPP during the second stage of SA sacrifice, reflected by full recovery of hind limb function in 100% of group 2 pigs. The fact that SA sacrifice during the first procedure stimulated vascular remodeling, allowing the stabilized collateral network to efficiently counteract an even greater additional ischemic insult after a few days of recovery, paves the road for possible new tactics for further minimizing paraplegia risk. During the completion of SA ligation in group 2, SCPP exhibited only a negligible decrease from baseline, correlating well with the excellent clinical recovery and the favorable histopathologic findings in that group. The correspondence between the SCPP pressure drop and clinical outcome offers another reason to include direct SCPP monitoring among other valuable means of assessing spinal cord viability. The clinical implications of these findings are exciting. In a recent retrospective study, we observed no instances of paraplegia among 35 patients who underwent complete TAAA replacement in a staged fashion (inadvertently) [25], suggesting the likelihood that direct translation of these experimental results to a clinical environment may prove successful. A possible scenario for a planned two-stage repair might involve preoperative endovascular embolization with platinum coils or second-generation embolic devices [26] of a few SAs to strengthen the collateral network before extensive surgical resection is carried

8 Ann Thorac Surg ZOLI ET AL 2010;90: STAGED APPROACH FOR TAAA REPAIR out. It may also be possible to devise an attractive option for extensive endovascular TAAA repair with some sort of temporary landing zone in the peridiaphragmatic area. This site may not be suitable for a long-lasting anastomosis, but could be used transiently as a bridge to obtaining a more robust collateral network by the time the resection is completed a few days later, either surgically or by stent grafting. It is of note that preliminary experimental observations suggest that the order of SA sacrifice (intercostal first followed by lumbar artery resection, or vice versa) does not seem to alter the impact on spinal cord protection, and therefore creative hybrid solutions may be designed to suit each patient depending on the specific anatomy of each aneurysm. In conclusion, complete elimination of the risk of paraplegia after repair of TAA or TAAA no longer seems an unachievable goal. We think spinal cord injury will become increasingly rare with the routine use of adequate adjuncts, the progressive reduction of surgical attrition, the increased availability and reliability of endovascular strategies, and the ability to exploit the resources offered by the collateral network. References 1. Strauch JT, Lauten A, Spielvogel D, et al. Mild hypothermia protects the spinal cord from ischemic injury in a chronic porcine model. Eur J Cardiothorac Surg 2004;25: Coselli JS, Lemaire SA, Koksoy C, Schmittling ZC, Curling PE. Cerebrospinal fluid drainage reduces paraplegia after thoracoabdominal aortic aneurysm repair: results of a randomized clinical trial. J Vasc Surg 2002;35: Cina CS, Abouzahr L, Arena GO, Lagana A, Devereaux PJ, Farrokhyar F. Cerebrospinal fluid drainage to prevent paraplegia during thoracic and thoracoabdominal aortic aneurysm surgery: a systematic review and meta-analysis. J Vasc Surg 2004;40: Amabile P, Grisoli D, Giorgi R, Bartoli JM, Piquet P. Incidence and determinants of spinal cord ischaemia in stentgraft repair of the thoracic aorta. Eur J Vasc Endovasc Surg 2008;35: Hnath JC, Mehta M, Taggert JB, et al. Strategies to improve spinal cord ischemia in endovascular thoracic aortic repair: outcomes of a prospective cerebrospinal fluid drainage protocol. J Vasc Surg 2008;48: Kouchoukos NT, Masetti P, Murphy SF. Hypothermic cardiopulmonary bypass and circulatory arrest in the management of extensive thoracic and thoracoabdominal aortic aneurysms. Semin Thorac Cardiovasc Surg 2003;15: Coselli JS, LeMaire SA. Left heart bypass reduces paraplegia rates after thoracoabdominal aortic aneurysm repair. Ann Thorac Surg 1999;67:1931 4; discussion Coselli JS, Bozinovski J, LeMaire SA. Open surgical repair of 2286 thoracoabdominal aortic aneurysms. Ann Thorac Surg 2007;83(Suppl):S862 4; discussion S Safi HJ, Miller CC 3rd, Carr C, Iliopoulos DC, Dorsay DA, Baldwin JC. Importance of intercostal artery reattachment during thoracoabdominal aortic aneurysm repair. J Vasc Surg 1998;27: Griepp RB, Ergin MA, Galla JD, et al. Looking for the artery of Adamkiewicz: a quest to minimize paraplegia after operations for aneurysms of the descending thoracic and thoracoabdominal aorta. J Thorac Cardiovasc Surg 1996;112: Etz CD, Homann TM, Plestis KA, et al. Spinal cord perfusion after extensive segmental artery sacrifice: can paraplegia be prevented? Eur J Cardiothorac Surg 2007;31: Etz CD, Luehr M, Kari FA, et al. Selective cerebral perfusion at 28 degrees C is the spinal cord safe? Eur J Cardiothorac Surg 2009;36: Etz CD, Zoli S, Kari FA, et al. Redo lateral thoracotomy for reoperative descending and thoracoabdominal aortic repair: a consecutive series of 60 patients. Ann Thorac Surg 2009; 88: Schlosser FJ, Mojibian H, Verhagen HJ, Moll FL, Muhs BE. Open thoracic or thoracoabdominal aortic aneurysm repair after previous abdominal aortic aneurysm surgery. J Vasc Surg 2008;48: Lombardi JV, Carpenter JP, Pochettino A, Sonnad SS, Bavaria JE. Thoracoabdominal aortic aneurysm repair after prior aortic surgery. J Vasc Surg 2003;38: Coselli JS, Poli de Figueiredo LF, LeMaire SA. Impact of previous thoracic aneurysm repair on thoracoabdominal aortic aneurysm management. Ann Thorac Surg 1997;64: Heil M, Eitenmuller I, Schmitz-Rixen T, Schaper W. Arteriogenesis versus angiogenesis: similarities and differences. J Cell Mol Med 2006;10: Pipp F, Boehm S, Cai WJ, et al. Elevated fluid shear stress enhances postocclusive collateral artery growth and gene expression in the pig hind limb. Arterioscler Thromb Vasc Biol 2004;24: Etz CD, Di Luozzo G, Zoli S, et al. Direct spinal cord perfusion pressure monitoring in extensive distal aortic aneurysm repair. Ann Thorac Surg 2009;87: Calligaro KD, Dougherty MJ. Correlation of carotid artery stump pressure and neurologic changes during 474 carotid endarterectomies performed in awake patients. J Vasc Surg 2005;42: Kato K, Tomura N, Takahashi S, et al. Balloon occlusion test of the internal carotid artery: correlation with stump pressure and 99mTc-HMPAO SPECT. Acta Radiol 2006;47: Etz CD, Homann TM, Luehr M, et al. Spinal cord blood flow and ischemic injury after experimental sacrifice of thoracic and abdominal segmental arteries. Eur J Cardiothorac Surg 2008;33: Etz CD, Luehr M, Kari FA, et al. Paraplegia after extensive thoracic and thoracoabdominal aortic aneurysm repair: does critical spinal cord ischemia occur postoperatively? J Thorac Cardiovasc Surg 2008;135: Toumpoulis IK, Anagnostopoulos CE, Drossos GE, Malamou-Mitsi VD, Pappa LS, Katritsis DG. Early ischemic preconditioning without hypotension prevents spinal cord injury caused by descending thoracic aortic occlusion. J Thorac Cardiovasc Surg 2003;125: Etz CD, Zoli S, Mueller CS, et al. Staged repair significantly reduces paraplegia rate after extensive thoracoabdominal aortic aneurysm repair. J Thorac Cardiovasc Surg 2010;139: Killer M, Hauser T, Wenger A, Richling B, Ladurner G. Comparison of experimental aneurysms embolized with second-generation embolic devices and platinum coils. Acta Neurochir (Wien) 2009;151: INVITED COMMENTARY The reasons for the occurrence of paralysis after open descending and thoracoabdominal aneurysm repairs are (1) ischemia during aortic cross clamping, (2) segmental artery related reduction of perfusion of the spinal cord after the operation, (3) postoperative events, including reduced collateral flow from low blood pressure and a 2010 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur