MONDAY AFTERNOON, APRIL 30, 1979

2:00 P.M. Scientific Session - Ballroom

7. A Comparison of Bubble and Membrane Oxygenators in Short and Long Perfusions

RICHARD E. CLARK, THOMAS B. FERGUSON and

CLARENCE S. WELDON, St. Louis, Missouri

This study was conducted over a 3-year period to ascertain the physiologic and biochemical alterations invoked by the use of bubble (N=20) and membrane oxygenators (N=20) in 40 selected high risk patients where perfusion times were anticipated to be ≥3 hrs. Similar matched sets of 40 patients each who had perfusion times of 2 hours or less were chosen to examine dependency of the results on perfusion time. The membrane oxygenator was a folded system using microporous expanded polytetrafluoroethylene. Independent automated transducer systems autoregulated venous and arterial roller pumps. Data were obtained before, at 30 min. intervals during, and at 1,4, and 24 hrs. after bypass. The following were measured: Hgb, Hct, WBC, platelets (direct), IgA, IgM, IgG, C'3, C'4, total protein, plasma Hgb, fibrin split products (FSP), protamine sulfate precipitations (PSP), electrolytes, gases, pH, glucose, BUN, urinary output, CVP, MAP, HR, temperature, blood loss and volume replacement. All cases used hemodilution, systemic and topical cardiac hypothermia and ischemia arrest. The average perfusion time in the long cases was 188 ± 14 min., range 121 - 316 min. In the short cases the average time of perfusion was 109 ± 11 min. The results demonstrated statistically similar changes in Hgb, Hct, and total protein in all cases reflecting similar degrees of hemodilution. Significant changes favorable to membrane oxygenators occurred only in the long cases where less WBC, platelet, C'3 decrease, less postoperative blood loss, and markedly lower hemolysis occurred. FSP and PSP were less frequently elevated or present with membrane oxygenators in long cases. Pump flows and urinary output were greater and total peripheral resistance during perfusion was decreased compared to bubblers. C'4 losses were greater with membranes in long perfusions. It is concluded that the theoretic advantages of folded membrane oxygenators are only evident in long perfusions and may be offset by greater cost, complexity and less margin of safety.

*By invitation

8. Pulsatile Perfusion Vs. Conventional High Flow Non-pulsatile Perfusion for Rapid Core Cooling and Rewarming of Infants for Circulatory Arrest Cardiac Surgery

G. DOYNE WILLIAMS*, ASTRIDEB. SEIFEN*, JIMB. NORTON*,

RICHARD I. READINGER*, THOMAS W. DUNCAN*, and

JANE K. CALLA WA Y, Little Rock, Arkansas

Sponsored by: Gilbert S. Campbell, Little Rock, Arkansas

Thirty consecutive infants undergoing hypothermia and circulatory arrest for repair of VSD, TGV, or A/V canal defects were alternately selected for conventional high flow non-pulsatile perfusion (Group A) or pulsatile perfusion (Group B) during core cooling and rewarming. All received morphine anesthesia, 30 mg/kg solu-medrol and 10-15 micrograms/kg phentolamine. Group A was perfused at 160-180 cc/kg/min with a roller pump and oxygenator with arterial pressure of 50-55 mm. Hg. Group B utilized a roller pump and oxygenator and an especially constructed pulsatile assist device (Datascope Corp.) was interposed in the arterial line to provide pulsatile perfusion with 75/40 mm. Hg. pressure at slightly reduced flow (150 cc/kg/min). The average rectal, esophageal, and tympanic membrane temperatures were reduced to approximately 16° C prior to circulatory arrest. Following repair, perfusion was resumed until these temperatures returned to 37° C. Results were as follows:

Cooling and rewarming were enhanced by pulsatile perfusion with over 30% reduction in total pump time. Additionally, the larger patients in the pulsatile group cooled almost as rapidly as the smaller. The rate of decline and subsequent rise of rectal, esophageal,, and tympanic membrane temperatures were equal in the pulsatile group but, the rectal lagged far behind in the non-pulsatile infants. The pulsatile flow infants awakened more quickly, were more alert, and required less post-opera-time mechanical ventilation. We suggest that pulsatile perfusion for core cooling and rewarming of infants is safe and is more rapid and physiologic than conventional high flow non-pulsatile perfusion.

*By invitation

9. Intramyocardial pH as an Index of Myocardial Metabolism During Cardiac Surgery

FRANCIS J.M. WALTERS*, GREGORYJ. WILSON*,

DAVID J. STEWARD*, RAUL J. DOMENECH* and

DAVID C. MacGREGOR, Toronto, Ontario, Canada

A practical method for continuously monitoring the state of tissue metabolism in the individual patient's heart during cardiac surgery is not yet available. We have examined the use of microelectrode measurements of myocardial interstitial pH to provide this monitoring capability, making comparisons with intracellular pH in left ventricular (LV) biopsy specimens and with tissue pC0₂ measured by mass spectrometry. An H+ sensitive glass microelectrode, housed in the bevelled end of a 21 gauge (0.8 mm diameter) needle, plus a 2 mm diameter reference electrode, with an internal Ag/AgCl electrode coupled to tissue through a saline bridge, were used. Microelectrode pH measurements in blood at 37°C were compared with conventional blood gas analysis over a pH range from 7.4 to 6.4. Linear regression analysis (n=26) revealed a high correlation (r=0.997) and a negligible difference in paired observations of only 0.01 ± .004 (mean ± SEM) pH units.

In 14 dogs on cardiopulmonary bypass, the pH needle and reference electrodes were inserted into the anterior wall of the LV to a mean depth of 6 mm. Under control conditions (37°C, arterial pH 7.36 ± .02, pCO₂>2 39 ± 1.3 mm Hg, p0₂ >100 mm Hg) the tissue microelectrode pH was 7.32 ± .02. Ischemic arrest of the heart at 37°C was then used to alter myocardial pH. In Group I (n=8), intracellular pH was estimated from LV biopsy specimens (400 mg each) taken at intervals over a microelectrode pH range of 7.37 to 6.37. The microelectrode and biopsy pH measurements were significantly correlated (r=0.905, n=41, P<.001). Microelectrode (interstitial) pH exceeded biopsy (intracellular) pH under control conditions by 0.28 ± .025 pH units (P<.001), but below a microelectrode pH of 6.8 the results of the two techniques did not differ significantly. In Group II (n=6), tissue pCO₂ in the anterior wall of the LV was determined by mass spectrometry. The tissue pCO₂ rose from 69 ± 2 mm Hg under control conditions to a final plateau during ischemia of 419 ± 25 mm Hg, the latter not being significantly different from a level of 422 ± 28 mm Hg theoretically calculated from the pH change (7.37 ± .014 to 6.01 ± .07), providing a further independent check on the pH microelectrode technique.

These results indicate that continuous microelectrode intramyo-cardial pH measurements do reflect the evolving state of intracellular metabolism during elective arrest of the heart and have a great potential for clinical application.

*By invitation

10. Hypothemic Circulatory Arrest: 31-P Nuclear Magnetic Resonance (NMR) of Isolated Perfused Neonatal Rat Brain

WILLIAM I. NORWOOD*, CAROL R. NORWOOD*,

JOANNE S. INGWALL *, ALDO R. CASTANEDA and

ERIC T. FOSSEL*, Boston, Massachusetts

Deep hypothermic circulatory arrest facilitates repair of congenital cardiac anomalies in infants. However, in spite of its widespread use, little ~is known of the fundamental cellular and molecular changes induced. An isolated perfused brain model was developed in part to study high energy phosphate metabolism with 31-P NMR at 109.3 MHz. Neonatal Sprague-Dawley rats cannulated through the ascending aorta were perfused with modified Krebs-Henseleit buffer. Soft tissues surrounding the calvarium, and cervical and thoracic spine were excised and the preparation lowered into a 15 mm. OD NMR tube. 200 to 400 free induction decays were averaged and transformed to produce each spectrum. The 31-P NMR spectra of well perfused brain show six major resonances representing a, b, andgphosphates of ATP, sugar phosphate, inorganic phosphate (Pi), and creatine phosphate (CrP). Preparations equilibrated to 37°C and 20°C were subjected to 20 minutes of ischemia with 15-20 minutes of reperfusion. Levels of CrP and ATP fell coordi-nately to 13 ± 1% and 28 ± 4% of control levels, respectively by 15 minutes of normothermic ischemia. This is distinctly different from rat heart where CrP significantly precedes the decrease in ATP. At 20°C CrP fell rapidly while ATP remained unchanged suggesting isolation of creatine kinase from the ATP pool in brain. On reflow at 37°C, ATP and CrP recovered substantially, but failed to return to control levels (51 ± 16% and 52 ± 8%, respectively). Following ischemia at 20°C ATP and CrP returned to control levels by 10 minutes. Intracellular pH determinations by chemical shift of P_i revealed a decrease from 7.2 to 6.7 during ischemia at 37°C, while pH at 20°C remained unchanged at 7.4. NMR proved a valuable tool for studying high energy phosphate metabolism in brain. This study suggests that permanent changes in ATP and CrP pools induced by 20 minutes of normothermic ischemia are attenuated while intracellular pH changes are abated by hypothermia.

INTERMISSION - VISIT EXHIBITS

*By invitation

11. Acute Adrenal Insufficiency Following Cardiac Surgical Procedures

WILLIAM C. ALFORD, JR., CLIFTON K. MEADOR*,

GEORGE R. BURRUS*, DAVID M. GLASSFORD, JR. *,

WILLIAM S. STONEY and CLARENCE S. THOMAS, JR *,

Nashville, Tennessee

Four of 4,064 adult patients undergoing cardiac surgical procedures at this hospital from 1974 to 1978 have experienced bizarre and confusing postoperative courses, ultimately shown to be on the basis of acute adrenal insufficiency. Three were men and the age range was 53 to 70 years (average 60.5 years). None had evidence of preoperative Addison's Disease or endocrine hypofunction.

In each instance, the operation was coronary artery bypass performed without untoward incident. Following uncomplicated postoperative courses ranging from 4 to 7 days, each developed insidious symptoms of flank or abdominal pain, delirium, low grade fever and eventual shock, occasionally preceded by hypertension. Diagnoses considered include leaking abdominal aneurysm, ischemic bowel, retroperitoneal hemorrhage, cholecystitis, pancreatitis, nephrolithiasis, stress ulcer, cecal volvulus, septicemia, brain tumor, cerebral edema and schizophrenia. None had infection. All had abdominal surgical consultation and three underwent laparotomy.

The correct diagnosis was first suspected at postoperative day 11 to 34 (average 20.5) and proven at postoperative day 15 to 41 (average 29.0). Low serum cortisol levels strongly suggested adrenal insufficiency. Confirmation was based on lack of rise in urinary steroid determinations after 3 days of maximal adrenal stimulation with ACTH. The four patients remain well on steroid replacement 7 to 58 months later, except for one who died 16 months postoperatively of aortic dissection. The adrenal glands showed apparent old hemorrhagic destruction, the cause of which is speculative.

The complexities of care of the postoperative cardiac surgical patient make the recognition of new adrenal insufficiency especially difficult. However, the rare person with such a diagnosis can be managed successfully and expect long-term clinical benefit.

*By invitation

12. The Bjork-Shiley Tilting Disc Valve - A Ten Years' Appraisal

VIKING O. BJORK and AXEL HENZE*, Stockholm, Sweden

The Bjbrk-Shiley tilting disc has passed 10 years, clinical use with an excellent performance and with significant improvement within its original design in the three most important aspects:

1. Durability. The original Delrin disc still gives an excellent performance after 10 years in patients. The pyrolytic carbon disc has increased the durability and diminished the regurgitation. The large strut has been made an integral part of the valve ring and thereby three times stronger. No genuine mechanical failure has been encountered in more than 1800 consecutive patients at Karolinska Hospital.

2. Flow-resistance. Haemodynamic evidence at rest and during exercise in 140 cases has been reported from our clinic to confirm the appropriate haemodynamic performance of the Bjork-Shiley prosthesis. No other presently available biological or mechanical heart valve shows such a low flow-resistance or gradient for a given tissue diameter, a fact of particular importance in cases with narrow aortic roots. A 10% further reduction in flow-resistance was obtained by the new convexo-concave version.

3. Thrombo-resistance. The new convexo-concave model uses an opening mechanism that adds a sliding of the disc when it tilts open and simultaneously increases the smaller opening of the prosthesis by 40%. Early experience in 320 operated cases, followed-up to 2^ years, has shown a promising low incidence of thrombo-embolic complications, which is probably due to increased velocity of flow through the minor prosthetic orifice and elimination of the low flow area behind the disc. Single MVR had no thrombo-embolic complications in 701 patient months and 1 thrombosis in AVR+MVR during 216 patient months.

4. Function control. The disc was equipped with a ring-shaped radi-opaque marker, permitting non-invasive function control of the tilting motion by means of cine-radiography or fluoroscopy.

A summary of the 10-year-experience is given.

*By invitation

13. Long-term Evaluation of the Porcine Xenograft Bioprosthesis

PHILIPE. OYER*, EDWARD B. STINSON*, BRUCE A. REITZ*,

D. CRAIG MILLER*, STEVEN ROSSITER*,

and NORMAN E. SHUMWAY, Stanford, California

The principal advantage of bioprosthetic cardiac valves, as compared to mechanical valve substitutes, consists of relative freedom from embolic complications and anticoagulant-related morbidity and mortality. However, the long-term durability of presently available biopros-theses has not been fully documented. This report provides followup data on 1285 patients [557 aortic (AYR), 561 mitral (MVR),and 167 aortic-mitral (AVR+MRV) replacements] who received Hancock xeno-graft valves at Stanford between 1971 and 1978. The total followup duration is 2740 patient-years, with a maximum followup of 7.3 years. One hundred ninety-six patients have been followed for > 4 years and 66 for > 5 years. Actuarial survival (±SEM) for AVR patients was 77 (± 4)% at 4 years, for MVR patients 72 (± 3)% at 5 years, and for AVR+MVR patients 73 (± 4)% at 4 years. Linearized morbidity and mortality expressed as percent per patient-year, unless otherwise indicated, are tabulated below.

Long-term anticoagulation was maintained in 10% of AVR patients and 31% of MVR patients (usually because of severe atherosclerosis of the ascending aorta, anatomic left-atrial abnormalities, or persistant atrial arrhythmias), accounting for the small incidence of anticoagulant-related morbidity and mortality reported. Actuarial probability (± SEM) of remaining free of thromboembolism for AVR patients is 95 (± 2)% at 4 years, for MVR patients 92 (± 2)% at 5 years, and for AVR+MVR patients 96 (± 2)% at 4 years. The actuarial probability of remaining free of valve dysfunction for AYR patients is 95 (± 2)%, for MVR patients 93 (± 2)%, and for AVR+MVR patients 95 (± 3)% at 4, 5, and 4 years respectively. These current data over an extended followup interval indicate that the Hancock xenograft valve continues to perform satisfactorily in terms of bioprosthesis-related morbidity and mortality, and, in particular, has shown no propensity for late failure due to leaflet tissue disruption.

*By invitation

14. Failure of Porcine Valve Heterografts in Children

ALEXANDER S. GEHA, HILLEL LAKS*,

HORACE C. STANSEL, JR., J. FREDERICK CORNHILL*,

JAMES W. KILMAN, MORTIMER J. BUCKLEY and

WILLIAM C. ROBERTS*, New Haven, Connecticut,

Columbus, Ohio, Boston, Massachusetts, and Bethesda, Maryland

Heterograft porcine valves have gained wide acceptance in replacement of diseased cardiac valves and their clinical performance in adults has been very satisfactory over follow-up periods of up to eight years. Valve replacement in children is relatively infrequent and experience with porcine xenografts is necessarily small. Our combined experience at three university hospitals has been with 25 children, 17 months to 16 years of age, who have been followed for five months to five years (mean follow-up 28 months). Porcine valves were used to replace the aortic valve in 15, the mitral valve in four, the tricuspid valve in one, and the pulmonary valve in five patients. Four (12%) of these valves have failed so far and required replacement because of severe stenosis in mitral (one) or aortic (three) valve prostheses at 18 to 45 months after implantation. Pathologic examination showed extensive fragmentation of collagen with focal heavy calcification and degeneration. In addition we have encountered deterioration and calcification of one porcine valve in 23 valved conduits followed up to five years (mean three years), requiring removal and replacement of the valve five years after implantation.

This experience indicates a disquietingly high incidence of relatively early failure of porcine xenograft valves in children. This is significantly higher than the failure rate observed in adult patients, and may be related to the small size of the implanted valves which become relatively narrow with the growth of the patient, leading to excessive turbulence and trauma to the prosthesis. Other factors which may contribute to these failures should be examined also in order to obtain better long-term results. A satisfactory performance would make heterografts the ideal valvular prostheses in children since anticoagulation is avoided.

*By invitation