| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2003;126:1288-1294
© 2003 The American Association for Thoracic Surgery
Surgery for acquired cardiovascular disease |
a Department of Cardiothoracic and Vascular Surgery, Memorial Hermann Hospital, Houston, Tex, USA
b The University of Texas at Houston Medical School, Memorial Hermann Hospital, Houston, Tex, USA
Read at the Eighty-second Annual Meeting of The American Association for Thoracic Surgery, Washington, DC, May 5-8, 2002.
Received for publication May 14, 2002; revisions received July 8, 2002; revisions received April 24, 2003; accepted for publication June 3, 2003.
* Address for reprints: Hazim J. Safi, MD, The University of Texas at Houston Medical School, Department of Cardiothoracic and Vascular Surgery, UTH Medical Center, Suite 450, 6410 Fannin Street, Houston, TX 77030, USA
Hazim.J.Safi{at}uth.tmc.edu
 |
Abstract |
|---|
 Top Abstract Material and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
METHODS: We performed 854 thoracoabdominal aortic repairs between February 1991 and May 2001. For this study we excluded 26 patients who died before postoperative neurologic status could be evaluated and 38 who had immediate neurologic deficit on initial postoperative evaluation, leaving 790 consecutive patients. We evaluated a wide range of demographic, preoperative physiological and intraoperative data, using univariate and multivariable statistical analyses.
RESULTS: Twenty-one of 790 (2.7%) patients had delayed neurologic deficit. Significant univariate predictors included preoperative renal dysfunction (odds ratio 5.9; P < .006), acute dissection (odds ratio 3.9; P < .05), extent II thoracoabdominal aorta (odds ratio 3.0; P < .03), and use of adjuncts (cerebrospinal fluid drainage and distal aortic perfusion; odds ratio 7.7; P < .03). The use of the adjuncts dropped from the multivariable model but all other factors remained. No other significant risk factors were identified. Twelve of 21 (57%) patients recovered neurologic function with optimization of blood pressure and cerebrospinal fluid drainage.
CONCLUSION: Preoperative renal dysfunction, acute dissection, and extent II thoracoabdominal aorta are significant predictors of delayed neurologic deficit. Previous studies have demonstrated that the use of adjuncts protects against immediate neurologic deficit. The findings of this study are consistent with the hypothesis that adjuncts reduce ischemic insult enough to prevent immediate neurologic deficit but that a period of increased spinal cord vulnerability persists several days postoperatively.
|
Although different surgical adjuncts, advanced anesthesia, and improved critical care have contributed to improved results, neurologic deficit following repairs of the thoracic and thoracoabdominal aorta (TAAA) remains a devastating complication. The incidence of postoperative neurologic deficit, as high as 16% to 25% in the era of clamp-and-go,1-3 currently ranges between 5% and 16%, depending on the use of adjuncts and the extent of the aneurysm.4-6 Factors that have been identified as part of the multifactorial etiology of neurologic injury include aortic ischemic period, extent of the aneurysm, and use of adjuncts.1,7-10
Neurologic deficit following TAAA repair can be categorized by the degree of injury (paraplegia or paraparesis) and by the time of onset. Immediate neurologic deficit refers to paraplegia or paraparesis upon awakening from anesthesia. Delayed neurologic deficit (DND) occurs after a patient has recovered from anesthesia and has been evaluated as neurologically intact yet develops paraplegia or paraparesis several hours or days later. The distinction between immediate and delayed onset is important because the etiology involved may be different.11 In addition, the prognosis for immediate neurologic deficit is poor with unlikely neurologic recovery and an increased risk of early death.2,8,11,12 However, DND may be reversed if identified early and treated promptly with therapeutic measures such as cerebrospinal fluid (CSF) drainage and optimization of blood pressure and oxygen delivery.8,11,13,14
DND was identified as a postoperative complication in the late 1980s and has been reported to occur in 1% to 12% of patients following TAAA graft replacement.10,14-17 Cases of DND have also been reported after endovascular repair.18-20 We previously reported the positive impact of CSF drainage on recovery from DND.11,12 Few reports, however, have analyzed the risk factors associated with DND. The purpose of this study was to analyze the preoperative and operative predictors of DND following TAAA repair.
|
Material and methods |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
Outcome variables
Paraplegia or paraparesis upon awakening, regardless of the severity, defined immediate neurologic deficit. Paraplegia or paraparesis after a period of normal neurologic function classified DND. Aortic dissection accompanying aortic aneurysms was considered acute if surgery was performed within less than 14 days from the onset of pain and chronic if after 14 days. Cerebrovascular disease was defined by a history of stroke, transient ischemic attack, or intervention for carotid artery disease. A history of chronic bronchitis and emphysema, or <60% of predicted forced expired volume in 1 second, identified chronic obstructive pulmonary disease. A serum creatinine level of greater than 2.0 mg/dL or the need for dialysis characterized renal dysfunction. Surgery was considered emergent when performed for symptomatic, contained, or free ruptured TAAA. Previous aneurysm repair referred to any previous replacement of the ascending, transverse arch, descending thoracic, thoracoabdominal, or abdominal aorta.
Technique
Descending thoracic and thoracoabdominal aortic repair was performed according to our standard surgical protocol, as described previously.5 In September 1992, based on our experience with animal studies and on reports that were beginning to appear in the literature, we changed our surgical protocol to incorporate distal aortic perfusion and CSF drainage. This technique allows increased time to evaluate and manage the intercostal arteries. Prior to September 1992, we utilized the clamp-and-go technique.
All patent lower thoracic intercostal arteries (level T8-T12) were reattached when possible (about 75% of all extent I and II cases, 60% of extent III and V cases, and about 9% of type IV cases). Extent IV and descending thoracic aortic aneurysms of the upper portion of the aorta seldom warrant intercostal artery reattachment. Occluded arteries and aortic dissections also rule out the feasibility of reimplantation. Active visceral and renal cooling was performed in addition to moderate systemic hypothermia.
The patient was anesthetized and intubated using a double lumen endotracheal tube. An arterial line and a pulmonary artery catheter monitored patient hemodynamics through central venous access. A CSF catheter (NMT, Neurosciences, Duluth, Ga) placed in the third or fourth lumbar space provided CSF drainage and monitoring of CSF pressure. The CSF pressure was maintained at less than 10 mm Hg by allowing passive CSF drainage. Mean arterial pressure (MAP) was maintained between 60 and 85 mm Hg by manipulating cardiac preload and afterload. During crossclamping, MAP was controlled by adjusting the flow rate of distal aortic perfusion and the administration of preload as required. Care was taken to prevent the temperature from drifting below 34°C. After completion of the repair the patient was warmed to 36°C core body temperature and distal aortic perfusion was discontinued.
Since September 1992 we have used CSF drainage for all elective and urgent repairs of TAAA and descending thoracic aortic aneurysms. Contraindications for CSF drainage included cases of free rupture, hypotension, sepsis or active bacteremia, recent history of intracerebral hemorrhage, or previous spinal surgery. CSF pressure was recorded at the insertion of the CSF drain, prior to initiation of distal aortic perfusion, at bypass, crossclamping, 10-minute intervals during crossclamping, release of clamp, and at closure when the patient was warmed.
The CSF drain was kept in place postoperatively while the patient was in the intensive care unit and removed after 3 days if no neurologic deficits were encountered. CSF was drained to maintain pressure below 10 mm Hg. Neurologic function was monitored on an hourly basis while the CSF drain was in place. The surveillance was reduced to once every 2 to 4 hours if normal function was documented.
If a new neurologic deficit was encountered after discontinuation of the drain, the CSF drain was reinserted by the anesthesiologist and gravity drainage was initiated. The patient was placed in the recumbent position to improve drainage. Hemodynamics and oxygen delivery were optimized by increasing serum hemoglobin to greater than 10 mg/dL and by improving oxygen saturation and cardiac output. Volume status was optimized and MAP maintained at greater than 90 mm Hg. Magnetic resonance imaging of the spine was obtained to exclude the possibility of spinal hematoma or other causes of cord compression and to identify spinal cord infarction.
An independent neurologist performed a comprehensive neurologic examination to identify the level of deficit. This involved the assessment of the patient's motor skills classified according to a modified Tarlov scale21 (Table 1). Paraplegia was classified by a Tarlov score of 0, 1, or 2; paraparesis by a score of 3 or 4. An increase in the Tarlov score of 2 or more upon discharge or transfer denoted improvement in the neurologic deficit.11
|
|
Results |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
|
Univariate analysis identified extent II TAAA (OR, 2.98; 95% CI, 1.23-7.18; P < .03), preoperative renal insufficiency (OR, 5.87; 95% CI, 2.35-14.62; P < .006), acute dissection (OR, 3.97; 95% CI, 1.09-14.19; P < .05), and adjuncts (OR, 7.71; 95% CI, 1.03-57.82; P < .03) as associated factors for DND (Table 2). Independent predictors of DND as identified by multiple logistic regression were extent II (OR, 3.21; 95% CI, 1.29-7.97; P < .012), acute dissection (OR, 5.25; 95% CI, 1.42-19.47; P < .013), and preoperative renal dysfunction (OR, 18.10; 95% CI, 3.09-106.00; P < .0013; Table 3). Distal aortic perfusion time, aortic crossclamp time, and intercostal artery reattachment were not associated with DND.
|
|
|
Discussion |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
In the clamp-and-go era of TAAA repair, the incidence of neurologic deficit ranged from 15% to 25%, depending on the extent of aneurysm,1-3 with rates as high as 31% in extent II TAAA.2 With adjunctive strategies to protect the spinal cord from ischemia the incidence has improved, falling to between 5% and 10%.4-6 Despite the use of adjuncts, the incidence of DND remains significant, comprising 36% (21/59) of overall neurologic deficits after repair in this series and comparable to a previous prospective study that identified delayed deficits to be one third of all neurologic deficits.17
As little information exists regarding DND, we sought to identify the preoperative and operative factors that may predispose patients to this complication. Using multivariable analysis, we found acute dissection, extent II TAAA, and preoperative renal dysfunction to be independent predictors of DND. These risk factors are similar to the risk factors that are associated with immediate neurologic deficit.5 This suggests that patients at risk for immediate neurologic deficit may also be at risk for DND. Thus, even after successful TAAA repair, patients with these risk factors require close follow-up to guard against the development of DND.
Other researchers as well as ourselves have found that the use of the adjuncts (distal aortic perfusion and CSF drainage) reduces the overall incidence of neurologic deficit.4,5,24 In this study univariate analysis showed us that adjuncts were associated with an increased risk of DND (P < .03). This suggests a patient who would have developed immediate neurologic deficit previous to the use of adjuncts might now recover from repair with intact but vulnerable neurologic function. That is, intraoperative ischemic insult is mitigated but not eliminated by the adjuncts. In the postoperative period this patient then suffers an insult such as an episode of hypotension that leads to further ischemia, resulting in the delayed deficit. So although distal aortic perfusion and CSF drainage may minimize spinal cord ischemia during the operative repair, the adjuncts do not appear to prevent DND.
Postoperative factors that may incite DND include hypotension, systemic inflammatory response syndrome, sepsis, cardiac dysrhythmias or cardiac failure, and diminished oxygen delivery caused by anemia, hypoxia, and low cardiac output. The importance of maintaining adequate blood pressure postoperatively for collateral circulation of the spinal cord has been emphasized previously.2,25 With regard to prognosis, the distinction between immediate and delayed neurologic deficit is important. The prospects for a patient with immediate neurologic deficit are poor since subsequent recovery is rare and early death likely.16 If DND goes unresolved the patient fares no better. However, DND may be reversed if identified early and if immediate countermeasures are undertaken (ie, CSF drainage, increase of blood pressure, and improvement of oxygen delivery).8,11,13,14 This was demonstrated in this study with a 57% (12/21) improvement in neurologic function if the CSF drain was utilized. When the CSF drain was in place at the time of the DND, 75% (9/12) of patients recovered function. When the CSF drain was reinserted at the time of the DND due to previous removal or malfunction, 43% (3/7) recovered neurologic function (Table 4). Reinsertion of the drain takes time. It requires assembly of the anesthesia team, preparation of the area, and so on, which can take an hour or more to accomplish. Our drain maintenance protocol requires that CSF pressure be kept at 10 mm Hg for 3 days postoperatively. When a delayed deficit is recognized in a patient with a drain in place, we open the device to drain freely, which can be done immediately. We think the timing of the intervention is responsible for the apparent difference between the success of CSF compartment decompression in the drain-in and drain-reinsertion groups, but the numbers are too small to test this formally at the present. The 2 patients who did not receive a CSF drain when the deficit occurred demonstrated no recovery at all. The significance of CSF drainage in immediate neurologic deficit has been established,4 but here again demonstrates a possible beneficial effect when used for DND. The improvement in neurologic recovery when the CSF drain was used suggests that DND may be caused, at least in part, by increased CSF pressure.
|
Despite its demonstrable benefits, CSF drainage is not an entirely benign procedure. Since integrating CSF drainage into our surgical protocol we have had 3 major complications in patients with CSF drainage (2 subdural bleeds and 1 case of meningitis) that may have been related to drainage. We have also had 1 case of a spinal hematoma at the insertion site that required a laminectomy, which was certainly related to CSF drainage. Clearly, judicious management of the drain is incumbent on clinicians who choose to use it.
Limitations of this study included its retrospective, nonrandomized design. Although the study group was of ample size (790 cases), the group that suffered DND was small (21) and any conclusions drawn from this group must be taken with caution. A better understanding of DND will need clinical studies analyzing the postoperative events involved as well as the development of an animal model for delayed spinal cord injury.
|
Conclusion |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
|
Discussion |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
There appear to be at least 3 common pathways: first, loss of spinal cord blood flow, particularly with a tenuous and perilous collateral blood supply; second, failure of adequate spinal cord oxygenation; and third, secondary spinal cord injury from the delayed complex biochemical cascade of deleterious pathways following ischemia and reperfusion.
The preoperative risk factors presented by Dr Safi of type II thoracoabdominal aneurysm, acute dissection, and renal disease reflect the problem of maintaining collateral blood supply in some of these patients. Because you only reattach intercostals from T8 to T12, according to your manuscript, do you think reattaching a wider range of intercostals or lumbar arteries may have reduced the risk? What are you now doing differently? We have found postoperative induced hypertension and delayed extubation useful. Do you think in patients with extensive aneurysms and few intercostal or lumbar artery ostia, endarterectomies should be performed for reattaching segmental arteries?
You had more deficits occur with CSF drainage. Do you think delayed CSF hypertension may be a factor? In a previous prospective randomized study in which postoperative hypotension resulted in 32% of the deficits, CSF drainage tended to be protective. In an excellent study recently published by Coselli, there also appeared to be some possible protection against delayed deficits. In 57% of your patients, delayed deficits improved with repeat drainage. Do you think that delayed removal of the catheter may be useful or do you think this would increase the risk of complications from the catheter too much?
Dr Safi. Thank you, Dr Svensson. With regard to our practice with intercostal arteries, we believe that intercostal arteries 8 to T12 are very important, and our work showed that these intercostal arteries are important for spinal cord perfusion. But in a paper from Amsterdam where they felt that if the lower intercostal arteries are not patent or occluded, the upper intercostal arteries are important, and currently, if we open the aneurysm and the intercostal arteries and the lower ones are not open, then we reattach the T5, 6, and 7. That is the only change.
With regard to the postoperative management, this is an intensivist study by Tam Huynh, and we currently believe that keeping a mean pressure of a 100 mm Hg is the order of the day, because hypotension correlates with paraplegia, as you and I have seen.
With regard to endarterectomy, I caution endarterectomy because I helped my late mentor, Dr Crawford, with 2 cases where the patient developed postoperative bleeding, one 2 hours after the endarterectomy and one 24 hours. I currently don't know how to reattach intercostal arteries to a graft, either an acutely dissected aorta or endarterectomy, because of risk of bleeding with a catastrophic event.
With regard to CSF drainage, we believe that CSF pressure, if it is high, may correlate with paraplegia. How long we are going to keep it, according to our median time for the CSF of the development of neurological deficits, is 2 days, so we are keeping it a day extra. I think if you keep it longer than that there is a risk of infection and some other complication.
Dr Ludwig von Segesser (Lausanne, Switzerland). I wish to congratulate Dr Safi for his excellent presentation, and I have a few questions with regard to the extent of repair. As a matter of fact, I did not see the exact proportion of thoracoabdominal repairs versus descending thoracic aortic repairs, and I believe that in acute aortic dissections, most of these cases can be repaired proximally first, and maybe there is a necessity for a distal reentry procedure.
I am asking this because in our environment there is an increasing proportion of endovascular aneurysm repairs, and under these circumstances there is no aortic crossclamping, and of course this may induce a shift or a selection bias in the patients you have referred.
Could you comment on this?
Dr Safi. In our experience, two thirds of our patients are extent I and extent II, so most of them are the more serious aneurysm. With regard to endovascular repair, there are reports in the literature that show delayed paraplegia following the insertion of the stent. With regard to acute dissection, if you remove the upper half of the descending thoracic aorta, risk of death, neurological deficit, or renal failure is minimal, but when we remove the thoracoabdominal aorta completely, from the left subclavian to the iliac bifurcation, the risk in our hands is about 35% to develop a neurologic deficit and the mortality rate is very high.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractMaterial and methodsResultsDiscussionConclusionDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. L. Estrera, R. Sheinbaum, C. C. Miller, A. Azizzadeh, J.-C. Walkes, T.-Y. Lee, L. Kaiser, and H. J. Safi Cerebrospinal fluid drainage during thoracic aortic repair: safety and current management. Ann. Thorac. Surg., July 1, 2009; 88(1): 9 - 15. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. D. Sanders and M. Maze Translational Research: Addressing Problems Facing the Anesthesiologist Anesth. Analg., October 1, 2007; 105(4): 899 - 901. [Full Text] [PDF]
|
|
|
|
|
|
P. E. Achouh, A. L. Estrera, C. C. Miller III, A. Azizzadeh, A. Irani, T. L. Wegryn, and H. J. Safi Role of Somatosensory Evoked Potentials in Predicting Outcome During Thoracoabdominal Aortic Repair Ann. Thorac. Surg., September 1, 2007; 84(3): 782 - 788. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Kawanishi, K. Okada, M. Matsumori, H. Tanaka, T. Yamashita, K. Nakagiri, and Y. Okita Influence of Perioperative Hemodynamics on Spinal Cord Ischemia in Thoracoabdominal Aortic Repair Ann. Thorac. Surg., August 1, 2007; 84(2): 488 - 492. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Winnerkvist, R. E. Anderson, L.-O. Hansson, L. Rosengren, A. E. Estrera, T. T.T. Huynh, E. E. Porat, and H. J. Safi Multilevel somatosensory evoked potentials and cerebrospinal proteins: indicators of spinal cord injury in thoracoabdominal aortic aneurysm surgery Eur. J. Cardiothorac. Surg., April 1, 2007; 31(4): 637 - 642. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Wong, J. S. Coselli, K. Amerman, J. Bozinovski, S. A. Carter, W. K. Vaughn, and S. A. LeMaire Delayed Spinal Cord Deficits After Thoracoabdominal Aortic Aneurysm Repair Ann. Thorac. Surg., April 1, 2007; 83(4): 1345 - 1355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Weigang, M. Hartert, M. P. Siegenthaler, N. A. Beckmann, R. Sircar, G. Szabo, C. D. Etz, M. Luehr, P. von Samson, and F. Beyersdorf Perioperative Management to Improve Neurologic Outcome in Thoracic or Thoracoabdominal Aortic Stent-Grafting Ann. Thorac. Surg., November 1, 2006; 82(5): 1679 - 1687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. L. Estrera, C. C. Miller, A. Azizzadeh, and H. J. Safi Adjuncts during surgery of the thoracoabdominal aorta and their impact on neurologic outcome: distal aortic perfusion and cerebrospinal fluid drainage MMCTS, October 9, 2006; 2006(1009): 1933. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Estrera, C. C. Miller III, E. P. Chen, R. Meada, R. H. Torres, E. E. Porat, T. T. Huynh, A. Azizzadeh, and H. J. Safi Descending Thoracic Aortic Aneurysm Repair: 12-Year Experience Using Distal Aortic Perfusion and Cerebrospinal Fluid Drainage Ann. Thorac. Surg., October 1, 2005; 80(4): 1290 - 1296. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
