Conference Coverage

Reattaching intercostals fails to squelch spinal cord ischemia in TAAA repairs

Key clinical point: Intercostal artery reimplantation (ICAR) did not produce a significant reduction in spinal cord ischemia following thoracoabdominal aortic aneurysm repair, even in the highest risk patients.

Major finding: ICAR was not a significant predictor of spinal cord ischemia (OR, 0.78; P = .55).

Data source: Retrospective analysis of 805 patients undergoing TAAA with or without ICAR.

Disclosures: The authors reported having no conflicts of interest.

View on the News

Counting on collaterals

Spinal cord ischemia is a rare but devastating complication of thoracoabdominal aneurysm repair. Crawford and his colleagues documented in 1993 an incidence of spinal cord ischemia (SCI) as high as 30% for extensive thoracoabdominal repairs. Efforts to diminish the risk of SCI were concentrated in identifying and preserving the direct arterial perfusion to the spinal cord from segmental arteries but continued experimental and clinical experience have suggested that multiple factors contribute to SCI.

Dr. Luis A. Sanchez

Some generally accepted principles for minimizing SCI include hypothermia, distal aortic perfusion with atriofemoral bypass or partial cardiopulmonary bypass, cerebrospinal fluid drainage, and avoidance of hemodynamic instability. Reimplantation of intercostal branches has been suggested as an adjunct to these techniques by some investigators with limited data to support its generalized application. More recently, a growing body of evidence supports the concept of a collateral network that can support the perfusion to the spinal cord after interruption of multiple intercostal arteries and the importance of the hypogastric and subclavian arteries as critical branches that perfuse the spinal collateral network.

The retrospective review of the extensive experience at the University of Wisconsin in Madison supports the concept that “physiologic interventions that increase spinal cord tolerance and collateral network perfusion during and after surgery” are more important than the reimplantation of intercostal vessels during this complex procedure, even in patients considered at the highest risk for SCI. Intercostal artery reimplantation failed to achieve significance as an SCI predictor when comparing two large cohorts of patients (540 vs. 265) treated with intercostal ligation vs. reimplantation. Increasingly, available data support the concept of a collateral network that maintains perfusion to the spinal cord after intercostal artery occlusion.

Additional new concepts and techniques including a two-stage approach for extensive thoracoabdominal repair, preliminary occlusion of some segmental arteries, and the use of hybrid and endovascular techniques may further decrease the incidence of SCI by taking advantage of the collateral network and allow some preconditioning of the spinal cord. Fortunately for these challenging patients, significant advances continue to be made to better understand and prevent spinal cord ischemia.

Dr. Luis A. Sanchez is Chief, Section of Vascular Surgery and the Gregorio A. Sicard Distinguished Professor of Surgery and Radiology, Department of Surgery, Washington University in St. Louis.




CHICAGO – Intercostal artery reimplantation fails to significantly reduce spinal cord injury following thoracoabdominal aortic aneurysm surgery, results of a large retrospective study show.

“Although there was a small decrease in spinal cord ischemia with ICAR, reattaching the intercostals did not produce a statistically significant reduction in spinal cord ischemia, even in the highest risk patients,” Dr. Charles W. Acher of the University of Wisconsin–Madison, said at the annual meeting of the Midwestern Vascular Surgical Society.

Intercostal artery reimplantation (ICAR) is one of several strategies that have been used to prevent spinal cord ischemia (SCI), paraplegia, and paraparesis that occurs from the interruption of the blood supply to intercostal arteries (ICAs) during thoracoabdominal aortic aneurysm (TAAA) repair.

Surgeons at UW–Madison adopted the ICAR strategy in 2005and now reimplant open ICAs located at T7-L2 in all Type I, II, and III TAAAs, using a previously published technique (J Surg Res. 2009;154:99-104).

Dr. Charles W. Acher
Dr. Charles W. Acher

Using a prospectively maintained database, the current analysis sought to compare outcomes between 540 patients who had TAAA surgery during 1989-2004 when open ICAs were ligated and 265 patients who had surgery during 2005-2013 with ICAR.The surgical technique for both groups was cross clamp without assisted circulation. The anesthetic technique was also uniform during the study period and included moderate systemic hypothermia (32° - 33° C); spinal fluid drainage (spinal fluid pressure less than 5 mm Hg); naloxone 1 mcg/kg per hour; use of mannitol, methylprednisolone, and barbiturate burst suppression; goal-directed therapy for a mean arterial pressure of 90-100 mm Hg and cardiac index of 2.5 L per minute/meter2; and proactive component blood therapy to avoid anemia, hypovolemia, and hypertension.

Aneurysm extent, acuity, mortality, renal failure, and pulmonary failure were the same in both groups.

The incidence of SCI was similar in all TAAAs at 5.25% without ICAR and 3.4% with ICAR (P = .23) and in the subset of patients with Type I, II, and III aneurysms (8.8% vs. 5.1%; P = .152), Dr. Acher reported on behalf of lead author and his colleague, Dr. Martha M. Wynn.

Interestingly, ICAR patients had more dissections than did the open ICA ligation patients (18% vs. 15%; P = .0016), more previous aortic surgery (47% vs. 31%; P = .0004), and longer renal ischemia time (61 minutes vs. 53 minutes; P = .0001), but had a shorter length of stay (14 days vs. 22 days; P = .0001) and were younger (mean age, 66 years vs. 70 years; P = .0001).

In a multivariate model of all TAAAs, significant predictors of spinal cord ischemia/injury were type II TAAA (odds ratio, 7.59; P = .0001), dissection (OR, 4.25; P = .0015), age as a continuous variable (P = .0085), and acute TAAA (OR, 2.1; P = .0525), Dr. Acher said. Time period of surgery, and therefore ICAR, was not significant (OR, 0.78; P = .55).

ICAR also failed to achieve significance as an SCI predictor in a subanalysis restricted to the highest-risk patients, defined as those having Type II TAAA, dissection, and acute surgery (OR, 0.67; P = .3387).

“Interrupting blood supply to the spinal cord causes spinal cord ischemia that can be mitigated almost entirely by physiologic interventions that increase spinal cord ischemic tolerance and collateral network perfusion during and after surgery,” Dr. Acher said. “Although the cause of SCI in TAAA surgery is anatomic, prevention of the injury is largely physiologic.”

During a discussion of the study, Dr. Acher surprised the audience by saying the findings have not changed current practice at the university. He cited several reasons, observing that there were more dissections in the ICAR group, and most of the ischemia in the ICAR group was delayed, suggesting that more patients could be rescued. In addition, there was a slight downward trend in spinal cord injury and immediate paraplegia with ICAR, however, these were not statistically significant.

“Because of those things, I still think it’s valuable, particularly in patients that are at highest risk, which are the dissections, with lots of open intercostals, but the emphasis should still be on physiologic parameters,” he said. “If you want to salvage patients, that’s the most important thing.

“Even if ICAR were ever shown to be statistically significant in a larger patient population, any role it has in reducing spinal cord injury would be extremely small,” he added in an interview.

The authors reported having no conflicts of interest.

Next Article:

   Comments ()