Differential Perfusion During Cardiopulmonary Bypass: Maintaining aCool Head During Warm Heart Surgery

TJ Jones, FRCS, DD Deal, BS, JC Vernon, BD, JW Hammon, MD, andDA Stump, PhD. Departments of Anesthesiology and Cardiac Surgery,Wake Forest University School of Medicine, Winston-Salem, North Carolina

Introduction. The choice of perfusion temperature during cardiopulmo-nary bypass (CPB) remains a contentious issue. The reduction in cerebralmetabolic rate and cerebral blood flow (CBF) associated with moderatehypothermia is neuroprotective by reducing the brain embolic load andischemic lesion size. In contrast, normothermic CPB is associated withimproved hemodynamic, respiratory, and hematological outcomes. Theoptimal temperature management during CPB would appear to be ahypothermic brain with a normothermic body. Differential perfusion ofthe aorta using a dual-lumen catheter provides independent temperaturecontrol of the head and body. The purpose of this study was to establishthe ability to maintain independent perfusion and to compare CBF valueswith those during conventional hypothermic and normothermic CPB.Methods. Following Animal Care and Use Committee approval, 12 dogsunderwent a stat CPB with a roller pump and membrane oxygenator.After anesthesia with fentanyl and diazepam, temperature probes wereplaced in the brain, nasopharynx, and bladder. The animals were ran-domized to normothermic CPB (N; 37°C, n 5 4), hypothermic CPB (H;28°C, n 5 4), or differential perfusion (DP; brain 28°C, bladder 37°C, n 54). Differential perfusion was achieved by dividing the output line fromthe combined heat exchanger oxygenator with one limb passing to asecond heat exchanger and arterial filter before connecting to the proxi-mal lumen (aortic arch) of a dual-lumen aortic catheter. The other limbpassed via an arterial filter to the distal lumen (descending aorta) of theaortic catheter. All animals underwent 15 minutes of stable normothermicCPB before groups H and DP were cooled over 30 minutes to a braintemperature of 28°C, which was maintained for a further 30 minutes.Colored microspheres were used to measure CBF.

Table 1. Results

Group

CPBTime(min)

BrainTemperature

(°C)

BladderTemperature

(°C)MAP

(mm Hg)CBF

(mL/g/min)H 15 36.9 6 0.2 38.3 6 0.2 78.8 6 13.4 0.65 6 0.10

75 27.8 6 0.4 29.4 6 1.0 80.0 6 14.5 0.29 6 0.07N 15 37.4 6 0.5 38.9 6 0.2 75.3 6 7.1 0.69 6 0.14

75 37.0 6 0.2 38.7 6 0.1 89.8 6 8.4 0.44 6 0.15DP 15 36.9 6 0.3 38.7 6 0.4 69.0 6 6.7 0.59 6 0.06

75 28.3 6 0.3 38.3 6 0.6 67.5 6 7.0 0.29 6 0.04

Data are mean 6 SD.

Conclusions. Differential perfusion using a dual-lumen aortic catheterenables independent temperature control of the brain while maintainingbody normothermia. The technique provides adequate CBF as comparedwith values obtained during standard hypothermic and normothermic CPB.Issues. Differential aortic perfusion during CPB facilitates cerebral hypo-thermia in combination with body normothermia.

Cerebrospinal Fluid Drainage During Thoracoabdominal Aortic Aneu-rysm Repair Prevents Spinal Cord Ischemia

SA LeMaire, MD, C Koksoy, MD, ZC Schmittling, MD, CC Miller III, PhD,PJ Oberwalder, MD, PE Curling, MD, and JS Coselli, MD. Baylor Collegeof Medicine and The Methodist Hospital, Houston, Texas

Introduction. Despite various strategies for preventing spinal cord isch-emia, paraplegia and paraparesis continue to complicate thoracoabdomi-nal aortic aneurysm (TAAA) repair. Although cerebrospinal fluid drain-age (CSFD) is often used as an adjunct for spinal cord protection, itsbenefit remains unproven. The purpose of this prospective randomizedtrial was to evaluate the impact of CSFD on the incidence of spinal cordinjury after extensive TAAA repair.Methods. After randomization, 145 patients underwent Crawford extent Ior II TAAA repairs with a consistent strategy of moderate heparinization,permissive mild hypothermia, left heart bypass, and reattachment ofpatent critical intercostal arteries; the repairs were performed with (n 576) or without CSFD (n 5 69). In the former group, CSFD was initiatedduring the operation and continued for 48 hours postoperatively. Thetarget CSF pressure was 10 mm Hg or less.Results. The prevalence of risk factors for paraplegia (aneurysm extent,acute presentation, and dissection) was similar in both groups. Aorticclamp time, left heart bypass time, and the number of reattachedintercostal arteries were also similar in both groups. Nine patients (13.0%)in the control group developed paraplegia or paraparesis; in contrast,only 2 patients in the CSFD group (2.6%) developed deficits ( p 5 0.026).This difference was due to the marked impact of CSFD in reducingimmediate paraplegia. Overall, CSFD resulted in an 80% reduction in therelative risk of postoperative deficits.Conclusions. Perioperative CSFD reduces the incidence of immediatepostoperative paraplegia after repair of extent I and II TAAAs.Issues. Why does cerebrospinal fluid drainage protect the spinal cordagainst ischemic injury? Should it be used routinely in patients with lessextensive aneurysms?

The Role of Aortic Clamp Manipulation as a Source of Particulate EmboliGeneration: Risk Factors and Outcomes of the ICEM Study Group

W van Boven, on behalf of the ICEM Study Group. St. Antonius Hospital,Niewagen, NetherlandsObjective. Prior studies have shown that atheroemboli are associatedwith neurologic complications after cardiac surgery. Aortic clamping hasalso been implicated as a cause of atheromatous embolization in previousstudies. A prospective consecutive enrollment registry has been initiatedto investigate particulate capture with intraaortic filtration by the ICEMStudy Group.Methods. Data were collected on patient selection, procedural details,particulate capture, and clinical outcomes on 398 patients in whomintraaortic filtration was used. Before removal of the aortic cross-clamp,the intraaortic filter (EMBOL-X Inc., Mountain View, CA) was deployedand left in place until the patient was weaned from extracorporealcirculation. Upon removal, filters were fixed in formalin and shipped to acore lab (Stanford University) for examination.Results. Patients enrolled in the registry had multiple risk factors forneurologic events including calcification of the ascending aorta (42.1%),prior stroke (6.9%), hypertension (64.4%), peripheral vascular disease(19.6%), congestive heart failure (18.6%), and low cardiac output (10.0%).Average age was 67.4 years (range 25 to 88 years). Coronary artery bypassgrafting (CABG) was performed in 59.6% of the patients, valve repair orreplacement in 18.8%, combination CABG/valve procedures in 12.7%, andother procedures in 8.9%. Mean filter dwell time was 33 minutes (range 2to 95 minutes). Partial cross-clamps were used in 80% of the CABGprocedures. Aortic cross-clamps were repositioned in 75 cases an averageof 1.15 times per case. Partial cross-clamps were repositioned in 77 casesan average of 1.52 times per case. Particulate matter was found in 276 of280 filters. Fibrous atheroma was found in 65% of the filters, 4% containedsurgical debris, 54% contained platelets and fibrin, and 20% containedthrombus. Mean number of particles found in each filter was 8.2 andmean surface area was 7.7 mm2. Findings of fibrous atheroma werestatistically significant in cases where the aortic cross-clamp was reposi-tioned ( p 5 0.07) and in cases where the partial cross-clamp wasrepositioned ( p 5 0.02). Three patients suffered a stroke (1.2%), 1 patienta transient ischemic attack (0.4%), and 1 patient a coma (0.4%).Conclusions. These findings from the ICEM registry confirm that ather-omatous emboli are released during aortic and partial cross-clamprepositioning. The surprising low incidence of adverse neurologic eventsmay substantiate previous studies linking particulate embolization andneurologic sequelae. Continued observational and randomized studiesare necessary to confirm the clinical relevance of particulate extraction.

Elevated S100b Levels Are Associated With Neurologic ComplicationsAfter Thoracic Aortic Surgery Requiring Hypothermic Circulatory Arrest

JK Bhama, MD, SA LeMaire, MD, ZC Schmittling, MD, PJ Oberwalder,MD, C Koksoy, MD, SA Raskin, CCP, PE Curling, MD, and JS Coselli,MD. Baylor College of Medicine and The Methodist Hospital, Houston,Texas

Background. Ischemic cerebral injury frequently complicates aortic sur-gery performed with hypothermic circulatory arrest (HCA). The aim ofthis study was to determine if elevated serum S100b (a marker forcerebral injury) correlates with neurologic complications after thoracicaortic surgery with HCA.Methods. After institutional approval and informed consent were ob-tained, blood samples were collected before induction of anesthesia in 37patients undergoing HCA during repair of thoracic aortic pathology.Additional samples were obtained 30 minutes after cardiopulmonarybypass (CPB) and 24 hours postoperatively. Serum S100b levels wasmeasured using a standard immunoradiometric assay.Results. Compared with baseline (0.09 6 0.11 mg/L), serum S100b levelswere significantly elevated after CPB (3.91 6 2.43 mg/L, p , 0.001) and 24hours postoperatively (0.73 6 1.40 mg/L, p 5 0.009). Neurologic compli-cations occurred in 3 patients (8.1%). Patients who developed neurologiccomplications had a higher mean S100b level (7.17 6 1.01 mg/L) after HCAthan those without neurologic complications (3.85 6 2.36 mg/L, p 50.013). Patients with serum S100b levels $ 6.0 mg/L had a higherincidence of neurologic complications (3/7, 42.9%) compared with thosewith levels , 6.0 mg/L (0/30, p 5 0.005).Conclusions. Post-CPB serum S100b levels $ 6.0 mg/L are associated withneurologic complications after thoracic aortic surgery requiring HCA.Issues. Should presumptive treatment for stroke be initiated based solelyon an elevated S100b level ($ 6.0 mg/L) after hypothermic circulatoryarrest?

1790 OUTCOMES 2000 Ann Thorac SurgSCIENTIFIC ABSTRACTS 2000;70:1786–97