WEDNESDAY MORNING, APRIL 27, 1994
7:00 a.m. FORUM SESSION II - Grand Ballroom
Moderators: Frederick L. Grover, M.D.
Robert A. Guyton, M.D.
F10. CAN DONOR HEARTS BE GENETICALLY ALTERED
PRIOR TO TRANSPLANTATION?
Abbas Ardehali, M.D.*, Hillel Laks, M.D., Alistair
I. Fyfe, M.D.*, Davis C. Drinkwater, M.D.*, Jian-Hua Qiao, M.D.* and Aldons J.
Lusis, Ph.D.*
Los Angeles, California
Access to the
donor heart at the time of harvest provides a unique opportunity for genetic
manipulation of this organ prior to transplantation. We sought to determine a)
if donor mouse hearts express a foreign gene administered at harvest, and b) if
so, what is the most effective route of gene delivery. At harvest, 30 ug of
pCMV-Luciferase DNA plasmid in cationic liposomes was injected I)directly into
myocardial apex, or II) into the right atrium, or III) into the coronary
arteries. The donor hearts (n=4 in each group) were then transplanted into the
abdomen of the recipient mice of the same strain. The transplanted hearts were
removed in 96 hours and luciferase expression was assayed by
immunohistochemistry. In group I, luciferase activity was localized to the
apex. In group II, where plasmid was delivered into the right atrium,
luciferase expression was detected solely in the right ventricle endocardium.
In group III, where plasmid was injected into the coronary arteries, the
transplanted hearts demonstrated luciferase expression in a) perivascular areas
surrounding coronary arteries and veins, b) coronary capillaries, and c) the
endocardium of both ventricles. Conclusions: This study suggests that a)
mouse hearts can be genetically modified at the time of harvest, and b)
intracoronary infusion of plasmid yields the most effective method of delivery.
Administration of plasmid in the coronary arteries localizes the expression to
the endocardium and the coronary vasculature, both sites of immunologic
interactions after heart transplantation. Intracoronary infusion of plasmids at
the time of harvest may allow genetic modification of the donor hearts in the
near future.
*By invitation
F11. LUNG TRANSPLANTATION USING
CARDIO-PULMONARY BYPASS EXAGGERATES PULMONARY VASOMOTOR DYSFUNCTION IN THE
TRANSPLANTED LUNG
David A. Fullerton, M.D.*, Robert
C. Mclntyre, M.D.*, Max B. Mitchell, M.D.*, David N. Campbell, M.D.* and
Frederick L. Grover, M.D.
Denver, Colorado
BACKGROUND:
Pulmonary vascular resistance (PVR) is significantly increased in the
transplanted lung. If cardiopulmonary bypass (CPB) is required, the
transplanted lung is reperfused with activated blood elements which might
exacerbate the reperfusion injury. Therefore, we postulated that pulmonary
vasomotor dysfunction is exaggerated when cardiopulmonary bypass is used for
lung transplantation. The purpose of this study was to examine the influence of
cardiopulmonary bypass on the following mechanisms of pulmonary vasomotor
control in a dog model of autologous lung transplantation: (1)
Endothelial-dependent cGMP-mediated relaxation (response to Acetylcholine, ACh)
(2) Endothelial-independent cGMP-mediated relaxation (response to
Nitroprusside, NP) and (3) β-adrenergic cAMP-mediated relaxation (response
to Isoproterenol, ISO).
METHODS:
Autologous lung transplants were performed in dogs using (n=4) and without
(n=5) CPB. After infusing PGE, (10u/kg), modified Euro-Collins solution (4°C,
30cc/kg) was infused into the right pulmonary artery. The right lung was
removed, stored in saline (4°C) for 3 hours, then reimplanted and reperfused
for one hour. 2 third-order pulmonary arteries were dissected from each lung at
each of two times: immediately post harvest (controls) and after one hour of
reperfusion. The vasorelaxing effects of ACh 10-6M, NP 10-6M
and ISO 10-6M were studied in isolated pulmonary arterial rings,
suspended on fine wire tensiometers in individual organ chambers. Vasomotor
function was compared in 3 groups: (1) Control (2) Lung Transplant Without CPB
and (3) Lung Transplant Using CPB. Statistical analysis was by ANOVA (Sheffe's
F-test).
RESULTS:
As shown below, lung transplantation using CPB produced significantly
greater endothelial-dependent as well as -independent pulmonary vasomotor
dysfunction than without CPB. Values are Mean ± SD.
CONCLUSION:
Lung transplantation using CPB greatly exaggerates pulmonary vasomotor
dysfunction. This may result in significantly higher PVR in the transplanted
lung when cardiopulmonary bypass is used.
*By invitation
F12. SUPPLEMENTAL L-ARGININE DURING
CARDIOPLEGIC ARREST AVOIDS REGIONAL POSTISCHEMIC INJURY VIA THE
L-ARGININE-NITRIC OXIDE PATHWAY
Hiroki Sato, M.D.*, Zhi-Qing
Zhao, M.D., Ph.D.*, Jakob Vinten-Johansen, Ph.D.*, D. Scott McGee, B.S.* and
John W. Hammon Jr., M.D.
Winston-Salem, North Carolina
Ischemia and
reperfusion impair contractile function and the generation of cytoprotective
nitric oxide (NO) by the vascular endothelium. This study tested the hypotheses
that blood cardioplegia (BCP) supplemented with the NO precursor L-arginine
(L-Arg) would 1) preserve endothelial function, 2) reduce infarct size, and 3)
reverse postcardioplegia regional dysfunction by the L-Arg-NO pathway. In 16
anesthetized dogs, the left anterior descending coronary artery (LAD) was
ligated for 90 minutes, after which bypass was established for surgical
"revascularization." In 9 dogs, unsupplemented multidose hypothermic BCP was
administered for a total of 60 minutes of cardioplegic arrest. In 7 dogs, L-Arg
was given intravenously (4 mg/kg/min) and in BCP (10 mM) during arrest. Infarct
size (TTC) as percent of the area at risk, was less in L-Arg compared to BCP (28
± 4% versus 41 ± 4%, p=0.04). Postischemic regional segmental work
(sonomicrometry) was significantly better in L-Arg (91 ± 15 mmHg-mm) versus BCP
(28 ± 3 mmHg-mm, p<0.001). Segmental diastolic stiffness was preserved in
L-Arg (0.46 ± 0.06) compared to BCP (1.12 ± 0.11, p=0.0001). In postischemic
LAD vascular rings taken from the above experiments, depressed maximum
relaxation responses to the receptor-dependent stimulator of NO in BCP group
(70 ± 5%) was reversed by L-Arg (92 ± 3%, p=0.002). Smooth muscle function was
unaffected in either group. The beneficial effects of L-Arg in vivo and in
vitro were reversed by the NO-synthase inhibitor L-NA )10-3 M).
We conclude that L-Arg supplementation during cardioplegia reduces infarct
size, preserves acute postischemic systolic and diastolic regional function,
and prevents endothelial dysfunction via the L-arginine-NO pathway.
*By invitation
F13. EFFECTS OF L-ARGININE AND L-NAME ON
RECOVERY OF NEONATAL LAMB HEARTS AFTER COLD CARDIOPLEGIC ISCHEMIA: EVIDENCE FOR
AN IMPORTANT ROLE OF ENDOTHELIAL PRODUCTION OF NITRIC OXIDE
Takeshi Hiramatsu, M.D.*, Joseph
M. Forbess, M.D.*, Takuya Miura, M.D.* and John E. Mayer, Jr., M.D.
Boston, Massachusetts
Myocardial ischemia and reperfusion result in both
ventricular and endothelial dysfunction. We have found that the endothelial
(Endo) defect is a reduced response to acetylcholine (ACh), likely due to
reduced nitric oxide (NO) release, but others report that NO is deleterious
after ischemia. To investigate the role of NO in recovery after hypothermic
ischemia, we examined the effects of infusions of 3mM L-arginine (L-ARG), a NO
precursor, 1mM L-nitro-arginine methyl ester (L-NAME), a NO synthase inhibitor,
1mM L-NAME plus 3mM L-ARG and 3mM D-arginine (D-ARG), a stereoisomer of L-ARG
vs controls (C) in isolated blood-perfused neonatal lamb hearts subjected to 2
hrs of cold cardioplegic ischemia. L-NAME was given before reperfusion, and
L-ARG and D-ARG were infused during the first 20 min of reperfusion. At 30 min
of reperfusion, maximal (Max) and volume-normalized (V10) LV developed pressure
(DP), +dP/dt, -dP/dt, coronary blood flow (CBF), myocardial oxygen consumption
(MVO2) were measured. Endo function was assessed by the coronary vascular
resistance (CVR) response to 10-7M ACh and 3x10-5M nitroglycerin
(NTG). Results are given as % recovery of preischemic values. (*=p<.05 vs C
by Student-Newman-Keuls test)
|
|
Max
|
V10
|
|
CVR response
|
|
Group
|
n
|
DP
|
+dP/dt
|
-dP/dt
|
DP
|
dP/dt
|
CBF
|
MVO2
|
ACh
|
NTG
|
|
C
|
8
|
75.5
|
68.3
|
60.7
|
75.1
|
66.8
|
133.0
|
76.1
|
39.2
|
42.4
|
|
L-ARG
|
8
|
94.5*
|
88.8*
|
74.1*
|
90.3*
|
87.5*
|
203.2*
|
96.8
|
61.0*
|
36.4
|
|
ARG+NAME
|
8
|
79.6
|
69.4
|
59.6
|
80.4
|
71.1
|
120.7
|
74.4
|
29.0
|
42.1
|
|
L-NAME
|
8
|
66.4*
|
54.9*
|
46.4*
|
66.1*
|
55.5*
|
86.0*
|
63.0
|
14.1*
|
50.7
|
|
D-ARG
|
8
|
79.4
|
70.8
|
62.6
|
80.8
|
73.1
|
140.0
|
80.7
|
40.1
|
38.2
|
|
Recovery
of all variables in L-ARG was significantly better (p<.05) compared to
L-NAME except NTG.
|
L-ARG, but not
D-ARG improved recovery of ventricular function, MVO2, CBF and
Endo-mediated response to ACh. L-NAME reduced recovery of ventricular function,
CBF and ACh response, and these effects of L-NAME were reversed to equal
control values by adding L-ARG to L-NAME treated hearts. These results confirm
an important salutary role for Endo production of NO in recovery after
hypothermic ischemia inneonatal lamb hearts.
*By invitation
F14. PATTERNS OF CHANGES IN NEUTROPHIL
ADHESION MOLECULES DURING NORMOTHERMIC CARDIOPULMONARY BYPASS - A CLINICAL
STUDY
Philippe Menasche, M.D., Ph.D.*,
Françoise Le Deist, M.D.*, François Tronc, M.D.*, Jacques Larivière, M.D.*,
Armand Piwnica, M.D. and Gerard Bloch, M.D.* Paris, France
Background: Adhesion
of activated neutrophils to endothelial cells and their subsequent migration
represent key features of the inflammatory response to cardiopulmonary bypass
(CPB) involved in postbypass organ dysfunction. However, the molecular
mechanisms of this adhesion during clinical CPB are still poorly
characterized.
Objective: To
assess the patterns of changes of the two sets of neutrophil adhesion molecules
involved in endothelial binding and migration during CPB. One set of these
molecules comprises the three activation-triggered β2 integrins
referred to as the CD 18 complex (GD11a/CD 18, GD11b/CD18, GD11c/CD18) and the
second set is made up by L-selectin, the peripheral lymph node homing receptor
that is normally expressed on unactivated leukocytes.
Methods: We
studied 8 adult patients who underwent coronary (N=6) or valvular (N=2)
operations with the use of normothermic (33°-37°C) CPB and warm blood
cardioplegia. A membrane oxygenator was used in all cases. The mean duration of
CPB was 91 min (range : 70-132). Arterial blood samples were taken after
anesthetic induction, at 5, 10 and 15 min on CPB and 30 min after the end of
CPB. Adhesion molecules were detected by direct immunofluorescence evaluated by
flow cytometry using a FACScan in log scale. Data (mean ± SD) are expressed as
mean linear fluorescence.
Results: They
can be summarized as follows: (1) There was no significant change in GD11a
expression throughout the study period; (2) This contrasted with a drastic rise
in CD11b expression which started early on bypass (at 10 and 15 min : 134 ± 72
and 167 ± 61 vs 66 ± 37 at baseline, p<0.05 and p<0.01 by t tests,
respectively) and persisted 30 min after bypass (154 ± 128); (3) Changes in CD11c
were less consistent but grossly featured a trend toward an early increase on
bypass (at 15 min : 53 ± 18 vs 36 ± 13 before CPB: followed by a return to
baseline levels 30 min after CPB (42 ± 22); (4) L-selectin expression decreased
during bypass but to a lesser extent than that of GD11b increased (at 10 and 15
min : 828 ± 109 and 809 ±119, respectively, vs 994 ± 185 at baseline, p=NS);
(5) Finally, postbypass fluorescence histograms disclosed two populations of
cells : one expressed CD11b and CD11c above preCPB levels whereas the second
had a low expression of the respective epitopes and, therefore, probably
corresponds to cells newly released from the bone marrow.
Conclusion: These
data provide direct molecular evidence that, in human beings, neutrophil activation
during CPB causes upregulation of CD11b/CD18 to a greater extent than it causes
downregulation of L-selectin, thereby resulting in an enhanced adhesiveness of
neutrophils with the potential for neutrophil adhesion-triggered cytotoxic
events. From a clinical standpoint, CD11b/CD18 and L-selectin could thus
represent elective targets for therapeutic interventions designed to reduce the
inflammatory component of post-bypass morbidity.
*By invitation
F15. IMPACT OF INITIAL FLUSH POTASSIUM
CONCENTRATION ON THE ADEQUACY OF LUNG PRESERVATION
Shigeyuki Sasaki, M.D, Ph.D.*,
James D. McCully, Ph.D.*, Francesca Alessandrini, B.S.* and Joseph LoCicero,
III, M.D.
Boston, Massachusetts
Potassium (K+)
is a classic pulmonary artery vasoconstrictor, yet most preservation solutions
have high K+ concentrations. We wished to determine the K+
threshold below which adequate lung preservation could be maintained. We
evaluated flush solutions containing physiologic to high dose potassium
concentrations. Excised Sprague-Dawley rat lungs (n=32) were flushed first with
one of following solution: (1) University of Wisconsin solution (UWS; K+=140mM)
(2) Low-potassium UWS (mUWS;K+=20mM) (3) Dulbecco's Phosphate-Buffered Saline
(DPBS; K+=3.9mM) (4) modified PBS (mPBS; K+=20mM) (5) Euro-Collins
solution (ECS;K+=115mM), then stored in 4°C UWS for 24 hrs. All
lungs were reperfused in the isolated blood perfused working lung system for
2hrs or until lung failure, determined by either blood gas (Sa02<90%) or
appearance of bronchial fluid. Aerodynamic values were measured on airway
volume-pressure loop.
Results: (Values at 30min (upper) and 90min (lower)
perfusion; mean ± SEM)
|
|
UWS
(n=6)
|
mUWS
(n=6)
|
DPBS
(n=7)
|
mPBS
(n=7)
|
ECS
(n=6)
|
|
pO2
|
56.1 ± 4.21
|
72.7 ± 9.09
77.6 ± 9.90
|
87.7 ± 6.93*
88.4 ± 1.54
|
108 ± 9.61**
71.0 ± 11.7
|
53.5 ± 6.01
|
|
SF
|
40.6 ± 8.30
|
16.4 ± 4.86*
13.0 ± 3.77
|
12.8 ± 2.39*
12.7 ± 2.58
|
8.01 ± 2.17**
16.4 ± 3.07
|
36.8 ± 8.69
|
|
RA
|
5.76 ± 0.93
|
2.82 ± 0.16*
3.24 ± 0.50
|
3.06 ± 0.17*
3.16 ± 0.19
|
2.78 ± 0.23**t
2.85 ± 0.18
|
4.95 ± 0.52
|
|
DC
|
0.13 ± 0.02
|
0.23 ± 0.01
0.23 ± 0.03
|
0.25 ± 0.02*
0.24 ± 0.02
|
0.24 ± 0.02
0.24 ± 0.01
|
0.17 ± 0.02
|
|
Wel
|
37.7 ± 6.84
|
19.8 ± 1.49
21.4 ± 3.56
|
18.7 ± 1.24
18.9 ± 1.12
|
19.2 ± 1.33
19.5 ± 1.18
|
27.2 ± 2.55
|
|
(p02 (mmHg), SF:shunt
fraction (%), RA:lung airway resistance (cmH2O/ml/sec),
DC:lung compliance
(ml/cmH2O), Wel:elastic lung work (g-cm);
ANOVA with post hoc pairwise
comparisons (Tukey). *p<0.05, **p<0.01 vs ECS, p<0.05, p<0.01
vs UWS)
|
All lungs
flushed with ECS or UWS demonstrated failure within 1 hr of reperfusion. These
results indicate the superiority of low-potassium flushing solution, either in
crystalloid or colloid flushing. We conclude that lung storage with UWS should
be preceded by low potassium (K+<20mM) flushing which can contain
colloid.
*By invitation
F16. SERUM CYTOKINES FOLLOWING PEDIATRIC
CARDIAC SURGERY: THE ROLE OF INTERLEUKIN-8 IN THE POSTOPERATIVE INFLAMMATORY
RESPONSE
Frank W. Mocek, M.D.*, Laman A. Gray, Jr., M.D.,
Michael J. Edwards, M.D.* and Erie H. Austin, III, M.D.*
Louisville, Kentucky
Sequelae related to cardiopulmonary bypass (CPB)
following pediatric cardiac surgery is an important source of early
postoperative morbidity. The purpose of this study was to determine the change
in serum cytokine levels over time and to correlate these changes with known
parameters of the inflammatory response in the pulmonary system.
METHODS: Serum
cytokine levels including Tumor Necrosis Factor (TNF), Interleukin (IL)-l,
IL-6, and IL-8 were measured by ELISA prior to, and at intervals following
induction of CPB in 18 consecutive patients undergoing correction of congenital
anomalies.
|
RESULTS
|
Baseline
|
5 min
|
1 hr
|
24hrs
|
48hrs
|
|
TNF
|
1.2 ± 0.70
|
1.5 ± 1.2
|
3.0 ± 2. 5
|
2. 7 ± 2. 5
|
3.1 ± 2.5
|
|
IL-1
|
0.6 ± 0.3
|
0.4 ± 0.2
|
0.4 ± 0.3
|
0.4 ± 0.2
|
0.4 ± 0.1
|
|
IL-6
|
2.6 ± 0.8
|
4.4 ± 2.1
|
65.2 ± 19.0*
|
50.0 ± 11.7*
|
25.2 ± 5.3
|
|
IL-8
|
30.0 ± 5.6
|
81.6 ± 23.8
|
227.0 ± 46.4*
|
75.7 ± 31.0
|
112.3 ± 62.7
|
|
(All data expressed mean ±
S.E.M., pg/ml; groups compared by ANOVA, correlations by linear regression
analysis, (*) p<.05 considered significant.)
|
IL-6 reached
peak serum levels at 1 hr (p<.05) and remained significantly elevated at 24
hrs (p<.05) but then decreased significantly from peak levels by 48 hrs
(p<.05). Serum IL-8 was significantly increased at 1 hr (p<.05) but
declined significantly by 24 hrs (p<.05) following CPB. Furthermore serum
IL-8 levels were directly correlated with length of CPB time (p<.001),
pulmonary neutrophil sequestration at 1 hr (p<.001) and the presence of
pulmonary edema at 24 hrs (p<.001) and 48 hrs (p<.001). Serum IL-8 was
also closely correlated with intraoperative fluid requirements (p<.001) and
fluid requirements at 24 (p<.001) and 48 hrs (p<.001).
CONCLUSION:
These data demonstrate that production of IL-6 and IL-8, but not TNF or IL-1,
is initiated shortly after the induction of CPB. Serum IL-6 levels remain
elevated for at least 24 hrs suggesting production of this cytokine continues
throughout this period. In contrast, serum IL-8 levels decreased significantly
by 24 hrs following the peak at 1 hr. The strong correlation between IL-8 and
pulmonary neutrophil sequestration, pulmonary edema, and increased fluid
requirements suggest IL-8 may be an integral mediator in the inflammatory
response resulting in pulmonary dysfunction following CPB. Options for
protection should be considered.
*By invitation
F17. ENDOTHELIAL-LINED SKELETAL MUSCLE
VENTRICLES IN CIRCULATION
Gregory A. Thomas, M.D.*, Peter I. Lelkes, Ph.D.*,
Susumu Isoda, M.D., Ph.D.*, Dawn Chick, B.S.*, Huiping Lu, M.D.*, Robert L.
Hammond, B.S.*, Hidehiro Nakajima, M.D., Ph.D.*, Hisako Nakajima, M.D.*, Henry
L. Walters, III, M.D.* and Larry W. Stephenson, M.D.
Detroit, Michigan and
Milwaukee, Wisconsin
Skeletal muscle
ventricles (SMVs) are muscular pumping chambers connected to the circulation
for cardiac assist. In our lab, SMVs have functioned routinely in the
circulation for over 6 months; in 1 dog an SMV pumped blood continuously for
836 days. However, during chronic, in-circulation studies, thrombus has formed
inside many of the SMVs with subsequent thromboembolism. In order to decrease
the incidence of thrombosis, we have been lining SMVs with autogenous
endothelial cells; in a group of 7 dogs seeded with autogenous endothelial
cells, an 80-100% complete monolayer of endothelium was obtained. For this
study, SMVs were constructed from the latissimus dorsi in 6 dogs by wrapping
the muscle around a polypropylene mandrel. Jugular vein endothelial cells were
enzymatically harvested and grown in tissue culture. After 3 weeks vascular
delay and 4 weeks electrical conditioning, 5 SMVs were seeded with 5-8x106
autologous endothelial cells by percutaneous injection of a cellular suspension
in 5 ml of culture media into the SMV lumen; one SMV was injected with culture
media alone as an unseeded control. The autologous endothelial cells were all
prelabeled with a lipid-bound cellular marker, PKH-26. After an additional 4
weeks of electrical conditioning, the mandrels were removed and the SMVs
connected to the descending thoracic aorta and activated to contract during
cardiac diastole at a 1:2 ratio with the heart. After 3 hours of continuous
pumping, mean diastolic pressure was increased by 35% (58 ± 7 vs 78 ± 6 mmHg,
p<0.05). The SMVs were excised for histologic examination. Hematoxylin and
eosin (H&E) stained sections revealed a continuous cellular layer lining
50-80% of the SMVs; there were no cells present on the lumen of the control
SMV. All seeded SMVs exhibited fluorescence secondary to the PKH-26 cellular
marker. Immunofluorescent staining with antibodies to von Willebrand factor and
ultrastructural analysis with electron microscopy confirmed the endothelial
character of these cells lining the lumen of the SMVs. To our knowledge, this
is the first time that an endothelial cell seeded, cardiac assist device has
retained endothelium while functioning effectively in the circulation. This is
not only another major step towards the clinical application of SMVs, but it
also holds enormous implications for the resolution of thrombotic events in
mechanical cardiac assist devices as well.
*By invitation
F18. THE ALLOGRAFT VALVE IN AORTIC VALVE
REPLACEMENT AND IN TRANSPLANTED PATIENTS
Joao Q. Melo, M.D., Ph.D.*, Jose
P. Neves, M.D.*, Sancia Ramos, M.D.*, Ana P. Martins, M.D.*, Narciso C. Andrade,
M.D.* and Manuel M. Macedo, M.D., Ph.D.
Carnaxide and Lisbon, Portugal
Clinical behaviour of aortic
allograft valves is different in aortic valve replacement (AYR) and in heart
transplanted (HT) patients. The study of post-mortem and explanted specimens is
crucial for the understanding of its biology. We have used histology,
immunohistochemistry and DNA individual profiles to assess 10 aortic valves
after AYR (5) or HT (5). Genetic profiles were characterized using
di-nucleotide repeats. DNA extracted from different portions of each cusp was
PCR amplified using fluochrome labelled primers and analyzed in a AB1 gene
scanner.
In the AYR group, 4 allograft
valves were cryopreserved and 1 was fresh. The cryopreserved valves had no warm
ischemic time and the cold ischemic time was 36 to 48 hours. Three allografts
explanted before 6 months had a 66% reduction of fibroblasts and the 2 with
more than 6 months of implantation were acellular. In 2 allografts T
lymphocytes were identified in the leaflets.
The 5 aortic valves of HT
patients were obtained from patients who died 1 to 47 months after heart
transplantation. These aortic valves had normal morphology and the only
histologic abnormal feature was a slightly reduced number of fibroblasts on the
free edge of the cusps. In the HT valves, the fibroblasts in the central
portion of each leaflet had DNA profile from the donor and at the base of the
cusp there were cells from the donor and from the receiver.
These findings favor the concept
that allografts are antigenic and immunosupression may play a role on its long
term function.
*By invitation
WEDNESDAY MORNING, APRIL 27, 1994
9:00 a.m. SIMULTANEOUS SCIENTIFIC SESSION D - ADULT
CARDIAC SURGERY
East Ballroom
Moderators: John A. Waldhausen, M.D.
Mortimer J. Buckley, M.D.
39. PATHOGENESIS OF ACUTE ISCHEMIC MITRAL
REGURGITATION IN THREE-DIMENSIONS
Robert C. Gorman, M.D.*, James S. McCaughan, M.D.*,
Mark B. Ratcliffe, M.D.*, Krishanu B. Gupta, Ph.D.*, James T. Streicher,
Ph.D.*, Victor A. Ferarri, M.D.*, Martin St. John-Sutton, M.D.*, Daniel K.
Bogen, Ph.D., M.D.* and L. Henry Edmunds, Jr., M.D.
Philadelphia, Pennsylvania and
San Francisco, California
Repair of acute
ischemic mitral regurgitation (MR) is compromised by ignorance of the mechanism
by which a structurally normal valve becomes incompetent. We postulate that
acute changes in annular circumference and cross-sectional area and in the
dynamics and distances between mitral subunits and infarcted ventricle cause
acute MR.
In 22 anesthetized sheep, sonomicrometry
transducers were placed on the tips and bases of each papillary muscle (n=4)
and five transducers were placed around the mitral annulus. These nine
transducers continuously generated a total of 36 inter-transducer distances
throughout the cardiac cycle, and defined the interrelationships of
transducer-tagged mitral valve subunits in 3-D space.
Two weeks
later, eight surviving sheep were re-anesthetized and a Millar catheter was
inserted retrograde into the LV. Color flow Doppler echo-cardiograms confirmed
the absence of MR. Multiple sets of sono-micrometry inter-transducer distances
were obtained of the normal mitral valve. Previously placed snares
around the second, third and PDA branches of the circumflex coronary artery
were tightened to produce a 40.4±7.2% infarction of the LV mass and acute 3 or
4+ MR by Doppler echocardiogram. One to four repeat sets of inter-transducer
distances were obtained for the regurgitant valve at LV end-diastolic pressures
between 11 and 29 mmHg.
Some new
observations of the normal valve indicate that annular circumference
decreases 4.6 ± 1.8 mm (4.5 ± 1.7%) with atrial contraction and
decreases an additional 2.2 ± 1.4 mm (2.1 ± .9%) during early LV systole. The
posterior papillary muscle (PPM) lengthens 0.2±0.14 mm (1.2 ± .5%)
during isovolumic systole and then shortens 2.3 ± 1.3 mm (12.1 ± 4.1%)
during ejection. After infarction, the PPM paradoxically lengthens instead of
shortens throughout systole, and the transcavity distance at ES increases 5.3 ±
1.8 mm (24.3±8%). End-systolic (ES) annular circumference and area increase
12.3 ± 2.5 mm (13.6 ± 3.1%) and 150 ± 49 mm2 (31+9%).
We conclude that acute annular
dilatation, loss of PPM shortening and LV dilatation after massive posterior
infarction distort coaption of the mitral leaflets to produce MR of the
structurally normal valve. These data provide a basis for developing improved
reparative operations.
*By invitation
40. AORTIC ROOT REPLACEMENT: RISK FACTOR
ANALYSIS OF A SEVENTEEN YEAR EXPERIENCE WITH 259 PATIENTS
Vincent L. Gott, M.D., A. Marc
Gillinov, M.D.*, Reed E. Pyeritz, M.D.*, Duke E. Cameron, M.D.*, Robert L.
Ferris, B.A.*, Bruce A. Reitz, M.D., Peter S. Greene, M.D.*, Christopher D.
Stone, M.D.* and Victor A. McKusick, M.D.*
Baltimore, Maryland;
Allegheny, Pennsylvania and Palo Alto, California
The results of aortic root replacement were
examined to determine operative risk and late results of treatment of ascending
aortic disease. Between 1976 and 1993, 259 patients (pts) underwent 265 aortic
root replacements. The most common indications for operation were aneurysm (232
pts; 90%) and dissection (26; 10%). Forty three procedures (16%) were performed
emergently. Marfan syndrome was present in 188 (73%), reflecting a referral
pattern of our institution. Compared to other pts, Marfan pts were less likely
to have dissection (p<.009) and emergent operation (p<.1). Mean pt age
was 38 years (range 4-77) and mean aortic diameter was 7 cm (range 4-16). Root
prostheses used were St. Jude composite grafts in 174 pts (66%), Bjork-Shiley
in 79 (30%), and cryopreserved homografts in 12 (4%). Hospital mortality
occurred in 13 pts (5%); mortality was 1.1% in 188 Marfan pts. Risk factors for
death were older age (p<.02), preoperative NYHA class III or IV (p<.001),
acute dissection (p<.02), emergent operation (p<.001), and use of
hypothermic circulatory arrest (p<.008). Late follow-up data were available
in 95% of hospital survivors. At mean follow-up of 47 months, there have been
22 late deaths; 4 were due to rupture of the distal aorta. Five and 10 year
actuarial survival were 90% and 79%, respectively. Fourteen pts (6%) developed
endocarditis; 8 died and 4 underwent root re-replacement with aortic homografts
without mortality or recurrent endocarditis. Actuarial freedom from reoperation
on the aortic root was 96% at 5 years and 88% at 10 years, while freedom from
distal aortic surgery was 94% at 5 years and 87% at 10 years. These results
demonstrate that aortic root replacement in the current era carries low
operative risk, low late mortality and morbidity, and good freedom from
reoperation. Endocarditis was the most frequent late complication and was
optimally treated by antibiotic therapy and root re-replacement using
cryopreserved homograft.
*By invitation
41. LONG TERM FOLLOW-UP OF PULMONARY VALVE
INSERTION: COMPARISON OF ISOLATED PORCINE VALVES, HOMOGRAFTS AND PORCINE VALVED
CONDUITS
Alon S. Aharon, M.D.*, Hillel Laks, M.D., Davis C.
Drinkwater, M.D.*, Peter Grant, M.D.*, Greg Fontana, M.D.*, Ehud Rudis, M.D.*
and Mohammed Qureshi, M.D.*
Los Angeles, California
There continues to be uncertainty regarding the
optimal valve type to be placed in the right ventricular outflow tract. The
results of pulmonary valve replacement using isolated porcine valves (PV),
homografts (H), and porcine valved conduits (PC) were compared. Ninety-six
children ranging in age from 10 days to 17 years (mean 6 years) and 35 adults
ranging in age from 18 years to 74 years (mean 36 years) underwent 163
pulmonary valve replacement operations from July 1978 to November 1993. In
children 77 H (57 aortic homografts and 20 pulmonic homografts), 28 PV, and 23
PC were placed. Insertion of a pulmonary valve was part of the primary
procedure in 98/128 (77%) operations in children with the following diagnoses:
TOP (27), PA-VSD (27), truncus arteriosus (18), TGA-VSD-PS (13), PA-IVS (4),
corrected TGA-VSD-PS (4), PS (3), and others (2). Thirty children underwent
redo procedures for obstructed pulmonary valves with the diagnoses of: truncus
arteriosus (11), TGA-VSD-PS (8), PA-VSD (3), TOP (3), PA-IVS (3) and others
(2). In adults 17 H, 16 PV and 2 PC were placed. Insertion of a pulmonary valve
was a primary procedure in 30/35 (86%) with the following diagnoses: TOP (13),
PA-VSD (6), PS (6), PI (3) and others (2). There was no early mortality, 1 late
death and no valve failures in the adult patients with a mean follow-up of 4
years (range 1 month to 15 years). In the pediatric group there were 5 early
(4%) and 3 late deaths (2%) and none of the deaths were related to valve failure.
Follow-up was available from 1-14 years (mean 4 years) and 14% of children were
reoperated on for valve failure at 4 years. In children, actuarial freedom from
valve failure using Cox regression analysis (fig. 1) revealed increasing valve
size to be the only significant factor predictive of improved freedom from
valve failure (p<0.001). For a given age a significantly larger (4 mm) PV
was used when compared to PC and H (p<0.01). Cox regression analysis (fig.
2) also demonstrated that PV showed a trend toward improved freedom from valve
failure when compared to H and PC (p<0.06). Freedom from reoperation at 4
years was 100% for PV, 85% for PC and 82% for H.
Conclusion: (1) Larger pulmonary valves offer
greater freedom from reoperation. (2) Larger PV can be placed for a given age
when compared to PC and H thus conferring greater freedom from reoperation.
*By invitation
42. TRICUSPID VALVE REPLACEMENT: FIFTEEN
YEARS OF EXPERIENCE WITH MECHANICAL AND BIOPROSTHESES
Hugh E. Scully, M.D., Cathy Tong, H.R.A.* and Susan
Armstrong, M.Sc.*
Toronto, Ontario, Canada
Tricuspid valve replacement is not a common
operation. The purpose of this study is to examine the early and late results
in 61 patients undergoing 33 (54%) mechanical and 28 (46%) bioprosthetic
tricuspid valve replacements. All operations took place between January 1978
and June 1993 when a total of 4741 patients underwent valve replacement
surgery.
Mean patient
age was 50 ± 17 (18-75) years. 42 pts (69%) were female; 19 pts (31%) were
male. 49 pts (80%) were in NYHA Class III or IV pre-operatively. 43 pts (70%)
were undergoing redo cardiac valve surgery. 13 pts (21%) had complex congenital
cardiac problems. Surgery was urgent in 16 pts (26%). Hospital mortality was
26% (16 pts) - all NYHA Class III or IV, redos and/or complex congenital cases.
Low output syndrome was observed in 21 pts (41%). Re-operation for bleeding was
required in 8 pts (13%). 19 pts (31%) required permanent (epicardial lead)
pacemaker implantation.
Mean follow-up
is 69 months (max. 170) and is 100% complete for the 45 patients who left
hospital. There have been 13 late deaths (21%). 9 of these pts (69%) had
mechanical valves and 4 pts (31%) had bioprostheses. Of the 9 cardiac deaths, 2
were valve-related (bioprostheses). 3 pts (7%) of the 32 still surviving
required re-operation because of TV prosthetic failure (one thrombosed
mechanical, 2 failed porcine). Of the remaining 29 patients, 10 (66%) are in
NYHA Class I or II. 16 pts have mechanical and 13 pts bioprostheses. 25 pts
(86%) are on coumadin. Thromboembolism (transient TIA) has occurred in 1 pt
with a mechanical valve who also had a previous CVA. There has been no
haemorrhage, endocarditis or new pacemaker requirement. Actuarial survival for
the series is 39 ± 8% at 15 years. Freedom from valve-related complications
among the 45 hospital survivors is 76 ± 9%.
Tricuspid valve
replacement is a beneficial procedure for patients with structural tricuspid
valve disease, many of whom have other valvular or congenital disease. Mechanical
and bioprostheses are equally effective in the tricuspid position.
10:20 a.m. INTERMISSION
*By invitation
11:05 a.m. SIMULTANEOUS SCIENTIFIC SESSION D - ADULT
CARDIAC SURGERY
East Ballroom
Moderators: John A. Waldhausen, M.D.
Mortimer J. Buckley, M.D.
43. MYECTOMY FOR HOCM: EARLY AND LATE
RESULTS
Elaine Heric, M.D.*, Bruce W. Lytle, M.D, Eliot R.
Rosenkranz, M.D.*, Harry M. Lever, M.D.* and Delos M. Cosgrove, M.D.
Tacoma, Washington and
Cleveland, Ohio
From 1975 through 1993 178 patients underwent
surgical management of hypertrophic obstructive cardiomyopathy (HOCM).
Operations included isolated septal myectomy (SM), 96, SM and coronary artery
bypass grafting (CABG), 41, SM plus a valve procedure, 24, SM, valve and CABG,
14, and mitral valve replacement (MVR) without SM, 3. Recent myectomy results
were monitored by transesophageal echocardiography. After initial myectomy 32
patients (20%) underwent a second pump run for more extensive myectomy only
(22), MVR only (7) or both (2). In-hospital mortality was 11 (6%), 6 (4%)
patients undergoing SM or SM plus CABG. Heart block occurred in 18 patients
(10%). Left ventricular outflow tract systolic gradients decreased from a mean
of 93 mmHg to 21 mmHg post myectomy.
Late survival
was 86% and 70% at 5 and 10 postoperative years, respectively, 93% and 79% for
patients undergoing SM alone or SM plus CABG. Only 3 of 131 in-hospital
survivors of SM or SM plus CABG died late cardiac deaths, for a yearly
mortality of .6%. However, the 5-year late survival of patients undergoing
valve operation plus SM was 51% and multivariate testing confirmed that adverse
influence on late survival (p=0.01), as well as adverse influences of
increasing age (p=0.023) and return to cardiopulmonary bypass for further
procedures (p=0.027). At follow-up 94% (136) of patients had NYHA I or II
symptoms.
For patients with HOCM, SM alone
or in combination with CABG produces effective symptom relief, excellent
long-term survival and a low risk of late cardiac death.
*By invitation
44. EXPERIENCE WITH THE WEARABLE NOVACOR
LEFT VENTRICULAR ASSIST SYSTEM AS A BRIDGE TO CARDIAC TRANSPLANTATION
Herbert O. Vetter, M.D.*, Hans G. Kaulbach, M.D.*,
Eckart Kreuzer, M.D.*, Christoph Schmitz, M.D.*, Hermann Reichenspurner, M.D.,
Ph.D.* and Bruno Reichart, M.D.*
Munich, Germany
Sponsored by: R.W.M. Prater,
M.D., Bronx, New York
The three components of the
Novacor left ventricular assist system (LVAS) - compact controller, battery,
and back-up battery - have been miniaturised in the development of the wearable
system. Therefore, patients (pts) can be mobilized almost completely during
mechanical circulatory support (MCS) while awaiting heart transplantation
(HTx).
Between 2/92
and 10/93 a total of 8 pts suffering from decompensated heart failure (4 dilated
cardiomyopathy, 3 ischemic heart disease, 1 acute myocarditis, 1
postcardiotomy) were treated with the Novacor LVAS; the last 4 cases using the
wearable system N100P. The pts age ranged from 17 to 59 years. In 6 pts severe
right heart failure was present at the time of implantation. Heparin IV was
used for anticoagulation treatment followed by oral phenprocoumon combined with
low dose acetylsalicyl acid.
Hemodynamic
stabilisation could be achieved in all pts during the 2 to 50 days (mean 15 ±
19 days) of MCS. One pt is supported at present and treated for legionellosis
of the lungs since more than 4 weeks. The following parameters were measured
before and 24 hrs after LVAS implantation: mean arterial pressure 67 ± 7 vs. 90
± 13 mmHg (p<0.05), cardiac index 1.69 ± .43 vs 3.17 ± .80 1/min/m2
(p<0.01), mean pulmonary artery pressure 40 ± 9 vs 25 ± 8 mmHg (p<0.05),
pulmonary capillary wedge pressure 25.1 ± 4.4 vs 9.6 ± 5.8 mmHg (p<0.05).
Right ventricular ejection fraction (RVEF), measured by a rapid response
thermodilution catheter, improved during left ventricular support in pts with
global heart failure; RVEF before LVAS implantation 13.2 ± 10.7% to 24.3 ±
13.9% at the time of HTx. The postcardiotomy pt died from multi-organ failure.
In 6 pts a HTx could be performed; 1 pt died due to unspecific graft failure. 5
pts are rehabilitated and 4 of them have returned to work. Pts on the wearable
system were able to manage their own power supply and power care during MCS
allowing them to walk in the hospital garden to go to the shopping area.
The new wearable Novacor LVAS provides major
advantages regarding "quality of life" of pts during MCS.
*By invitation
45. RETROGRADE CEREBRAL PERFUSION DURING
HYPOTHERMIC CIRCULATORY ARREST REDUCES NEUROLOGIC MORBIDITY
G. Michael Deeb, M.D., Steven F.
Bolling, M.D., Louis A. Brunsting, M.D.*, David M. Williams, M.D.*, Leslie E.
Quint, M.D.* and Nancy D. Deeb, R.N.*
Ann Arbor, Michigan
Hypothermia circulatory arrest (HCA) has become an
accepted technique for a variety of cardiac and complex aortic operations.
However, prolonged periods (>45 min) of HCA in older patients is associated
with marginal cerebral protection and an increased incidence of adverse
neurologic events. In an effort to minimize such morbidity, we employed a
technique of retrograde cerebral perfusion (RCP) during HCA in 16 patients who
underwent thoracic aortic surgery or resection of intracardiac tumor. There
were 12 males and 4 females (mean age 62, range 36-76). Six patients presented
with acute dissection, 8 had thoracic aortic aneurysms, and 2 with
hypernephromas extending into the heart. Ten patients underwent root and arch
replacement utilizing composite grafts (3 with simultaneous coronary artery
bypass grafting), 4 had arch replacement, and 2 resection of tumor in the heart
and retrohepatic vena cava. Ten cases were elective and 6 emergent, 4 (25%)
were reoperations.
Operative
technique included a median sternotomy, cardiopulmonary bypass (CPB) using
common femoral arterial perfusion (CFAP) and right atrial venous drainage,
retrograde cardioplegia, and venting of the left ventricle. RCP was
accomplished during HCA using a right angled cannula in the superior vena cava
(SVC), Y-connected into the arterial perfusion line. When the core temperature
was <20°C, CFAP was discontinued and RCP was instituted to maintain a
perfusion pressure in the SVC<15 mm/Hg. RCP flows varied from 350-800
ml/min. CPB suckers were used to return RCP flow. After completion of the
surgical procedure, RCP was halted and CFAP resumed.
The mean RCP
time was 57 ± 4 min (range 35-96 min), with 13 (81%) patients >45 min. There
was one operative death and 1 late death, both due to pre-op myocardial
infarction from aortic dissection. One patient had a stroke secondary to acute
dissection of the left carotid artery. There were no other neurologic events,
re-operations for bleeding or adverse outcomes. The average length of stay for
elective cases was 9 days and for emergent cases 28 days. At a mean follow up
of 5 months all surviving patients are well.
HCA is a
relatively simple technique which provides a bloodless field and good
visualization without the need for aortic cross clamps. Moreover, RCP extends
the "safe" time for HCA allowing ample opportunity to perform complicated
cardiac and aortic surgery with reduced risk of adverse neurologic events.
12:00 p.m. ADJOURN
*By invitation
WEDNESDAY MORNING, APRIL 27, 1994
9:00 a.m. SIMULTANEOUS SCIENTIFIC SESSION E -
GENERAL THORACIC SURGERY
Trianon Ballroom
Moderators: Andre C.H. Duranceau, M.D.
Willard A. Fry, M.D.
46. PULMONARY RESECTION FOR INVASIVE
ASPERGILLUS INFECTIONS IN IMMUNOCOMPROMISED PATIENTS
Lary A. Robinson, M.D.*,
Elizabeth C. Reed, M.D.*, Timothy A. Galbraith, M.D.*, Anselmo Alonso, M.D.*,
Anthony L. Moulton, M.D. and William H. Fleming, M.D.
Omaha, Nebraska and
Providence, Rhode Island
Standard medical therapy of invasive pulmonary
aspergillus infections occurring after bone marrow transplantation (BMT) or
liver transplantation (LT) results in less than a 5% survival. Therefore, we
adopted an aggressive approach with surgical resection of the involved area
along with systemic antifungal therapy when a localized pulmonary aspergillus
infection develops in these severely immunocompromised patients.
METHODS: from May 1987 TO October 1993, 12 (BMT)
and 3 (LT) patients underwent resection of localized acute pulmonary masses
suspicious for invasive aspergillus, as suggested by chest CT scans and
clinical signs and symptoms. Operative procedures performed include 2
pneumonectomies, 1 bilobectomy with limited thoracoplasty, 8 lobectomies, and 5
wedge resections (one BMT patient had two procedures). All patients were also
treated with systemic antifungal agents. The diagnosis of invasive pulmonary
aspergillus infection was confirmed by characteristic histopathology and/or
positive cultures.
RESULTS: 8 of
12 (67%) BMT patients and 2 of 3 (67%) LT patients survived the peri-operative
period with no evidence of recurrent aspergillus infection. Despite the severe
pyrexia present in most patients, none had positive blood cultures for
aspergillus nor any other evidence of disseminated disease. Five hospital
deaths (4 BMT, 1 LT) occurred a mean 22.4 ± 12.4 days after surgery from
continuing systemic toxicity and multiple organ failure. Both ventilator-dependent
patients in the series died. In addition, 3 of the 4 deaths in BMT patients
occurred in allogeneic BMT recipients.
Of survivors,
no surgical complications occurred post-operatively except for renal
insufficiency that was felt related to amphotericin B. Surgery on BMT patients
was performed a mean 37.4 ± 16.0 days after starting initial high dose
chemotherapy. Two LT patients underwent their resection 9 and 48 days after
transplant. Surgery was performed in the other LT patient 7 days after starting
medical treatment for an acute rejection episode. The 10 initial survivors (8
BMT and 2 LT) were followed for a mean 16.3 ± 24.8 months (range 0.25-78)
months. Four BMT patients have subsequently died from progression of their
malignancy, but none developed a recurrent aspergillus infection.
CONCLUSIONS: 1)
Immunocompromised BMT and LT patients who develop the uniformly fatal
complication of invasive pulmonary aspergillus infections may benefit from
early surgical resection of the involved area, with 2/3 of such patients in
this series cured of their fungal infection long term. 2) Ventilator dependent
patients, allogeneic BMT recipients, and patients with multiple loci of
infection are less likely to benefit from this approach. 3) Early pulmonary resection
with lobectomy or occasionally wedge resection for smaller lesions should be
performed when the characteristic clinical and radiographic pictures appear,
and should not await positive cultures or a trial of antifungal chemotherapy.
*By invitation
47. OUTCOME OF ADENOCARCINOMA ARISING IN
BARRETT'S ESOPHAGUS IN ENDOSCOPICALLY SURVEYED AND NON-SURVEYED PATIENTS
Geoffrey W.B. Clark,
F.R.C.S.(Ed)*, Adrian P. Ireland, F.R.C.S.(I)*, Jeffrey H. Peters, M.D.*,
Thomas C. Smyrk, M.D.*, Para Chandrasoma, M.D.* and Tom R. DeMeester, M.D.
Los Angeles, California and
Omaha, Nebraska
The value of endoscopic
surveillance of Barrett's esophagus and the appropriate management of high
grade dysplasia (HGD) remains unclear. Recent reports have proposed that
endoscopy and biopsy can accurately differentiate HGD from intramucosal
adenocarcinoma, thus continued surveillance of HGD is safe. The method of
surveillance and outcome of treatment in patients with severe dysplasia or
adenocarcinoma arising in Barrett's esophagus was evaluated to address these
uncertainties.
The study population consisted
of 16 patients referred from endoscopic surveillance programs, 11 with HGD and
5 with adenocarcinoma, and 33 patients with a newly recognized adenocarcinoma
who were not under surveillance. The flow chart shows the subsequent outcome of
those patients referred from surveillance. Following repeat endoscopy with
extensive biopsy, 2 were diagnosed as adenocarcinoma while 9 were operated with
a diagnosis of HGD. In 7 of these 9 patients no mucosal abnormality was seen
endoscopically while one had stricture and one a small superficial ulcer.
The median number of biopsies
taken was 15 (range 7-34). The median number of biopsies per cm of Barrett's
was 3.9 (range 0.8-7.5). Tumors were staged post operatively by the WNM
classification based on the degree of wall penetration and the number of lymph
node metastasis. Despite the high number of biopsies per cm 5 patients with
adenocarcinoma were missed. Twelve patients in the screened group had
adenocarcinoma: 11 early and 1 intermediate. Patients not under surveillance
presented with more advanced tumors: 11 early, 9 intermediate and 13 late
(X2=12.2, p<0.01). Sixteen patients referred from surveillance, despite the
presence of adenocarcinoma in 12, enjoyed a significantly improved survival
compared to the non surveyed group (X2=4.2, p<0.05).
Patients referred from
surveillance programs for Barrett's esophagus have a better outcome and earlier
stage tumors compared to non surveyed patients. Multiple biopsies do not
exclude the presence of adenocarcinoma making continued surveillance of high
grade dysplasia dangerous and potentially destructive to surveillance efforts.
*By invitation
48. p53 IMMUNOREACTIVITY IN BARRETT'S
METAPLASIA, DYSPLASIA AND CARCINOMA
Thomas W. Rice, M.D., John R. Goldblum, M.D.*, Gary
W. Falk, M.D.*, Raymond R. Tubbs, M.D.*, Thomas J. Kirby, M.D.* and Graham
Casey, M.D.*
Cleveland, Ohio
Mutations of the tumor
suppressor gene p53 frequently result in intranuclear protein accumulation and
are implicated in the loss of regulation of normal cell growth and
differentiation. Barrett's esophagus is a metaplastic condition with an
unpredictable potential for neoplasia. The study was undertaken to determine 1)
if the expression of p53 in Barrett's esophagus is a marker for neoplasia and
2) when in the metaplasia-dysplasia-carcinoma sequence p53 expression occurs.
Twenty-eight
esophageal resection specimens were studied. In each specimen (28) Barrett's
mucosa (BM) was present. Low grade dysplasia (LGD) was seen in 27 specimens,
high grade dysplasia (HGD) in 26, intramucosal cancer (IMC) in 18, and
submucosal cancer (SMC) in 5. Immunohistochemical staining with the monoclonal
antibody PAblSOl was used to detect mutated intranuclear p53.
No p53
immunoreactivity was seen in BM or LGD. p53 was seen only in HGD, IMC, and SMC.
Sixty-nine percent (18/26) of these specimens expressed p53. In those specimens
in which p53 staining was observed, the spectrum of neoplastic changes (HGD,
IMC, SMC) within the specimen was positive for p53. Of the 8 specimens where
the most severe neoplastic change was HGD, 6 (75%) expressed p53
immunoreactivity. Similarly, 10 of 13 IMC (77%) and 2 of 5 SMC (40%) expressed
p53 immunoreactivity.
We conclude
that p53 immunoreactivity in Barrett's esophagus 1) is a frequent, but not
exclusive, marker for HGD, IMC, and SMC and 2) occurs late in the
metaplasia-dysplasia-carcinoma sequence when HGD transformation has occurred.
*By invitation
49. CARCINOMA OF THE ESOPHAGUS: PROGNOSTIC
SIGNIFICANCE OF HISTOLOGY
Michael D. Lieberman, M.D.*. Craig D. Shriver,
M.D.*, Steven Bleckner, B.S.*, Michael Burt, M.D., Ph.D. and the members of the
Thoracic and Gastric and Mixed Tumor Services
New York, New York
Introduction: Previous investigators have suggested
that adenocarcinoma of the esophagus when compared to squamous cell carcinoma
adversely affects prognosis. A review of 258 patients, from 1985-1991,
undergoing esoph-agectomy for adenocarcinoma or squamous cell carcinoma was
performed to test the hypothesis that histology is an independent prognostic
variable, and to identify other predictors of survival following esophagectomy.
Methods:
Seventeen demographic, clinical, treatment and pathologic variables were
analyzed for effect on overall and disease free survival following curative
esophageal resection. The primary operative procedures were transthoracic
esophagectomy (79%), transabdominal esophago-gastrectomy (9%), transhiatal
esophagectomy (7%) and pharyngo-esophagectomy (5%). The median followup was 18
mos. The in-hospital mortality was 5%. Univariate analysis of survival for each
variable was determined by the log-rank test. Multivariate analysis was
performed using the Cox proportional hazards model.
Results: The median overall survival was 27 mos.
for adenocarcinoma compared to 22 mos . for squamous cell, p=0.16 (figure).
Univariate analysis identified T stage, N stage, number of positive nodes,
tumor differentiation, tumor site and blood transfusion as significant (p<0.05)
variables in predicting overall survival. Age, sex, dysphagia, tobacco use,
alcohol use, weight loss, Barrett's esophagus, operative procedure, margins,
splenectomy, and anastomotic leak were not significant prognostic factors. The
table indicates prognostic factors derived from the Cox model, and the figure
demonstrates overall survival according to histology.
|
Cox Multivariate Analysis
|
|
Characteristic
|
Overall Survival (p)
|
3-Yr Overall Surv (%)
|
|
TNM: T stage
|
0.006
|
T1=64, T2=56, T3=30
|
|
TNM: N stage
|
0.01
|
N0=56, N1=27
|
|
No. of Pos. Nodes
|
0.03
|
0=56, 1-3=37, >3=18
|
|
Blood Transfusion
|
0.07
|
No=47, Yes=31
|
Histology, tumor location, tumor differentiation,
microscopic margin, splenectomy, preoperative weight loss and age were not
significant variables in the Cox model.
Conclusion:
Multivariate analysis demonstrated that histology is not an independent
variable for overall survival in patients undergoing curative resection of
esophageal carcinoma. Outcome following curative esophagectomy is most strongly
influenced by extent of disease defined by T and N stage.
10:20 a.m. INTERMISSION
*By invitation
11:05 a.m. SIMULTANEOUS SCIENTIFIC SESSION E - GENERAL
THORACIC SURGERY
Trianon Ballroom
Moderators: Andre C.H. Duranceau
Willard A. Fry, M.D.
50. TREATMENT STRATEGIES FOR
BRONCHOPLEURAL FISTULA
John D. Puskas, M.D.*, Douglas J. Mathisen, M.D.,
Hermes C. Grille, M.D., John C. Wain, M.D.*, Cameron D. Wright, M.D.* and Ashby
C. Moncure, M.D.
Boston, Massachusetts
Successful management of chronic bronchopleural
fistual (BPF) remains a challenge for thoracic surgeons. Forty-seven patients
were treated for postoperative BPF since 1978. Patients had undergone an
average of 3.1 surgical procedures to correct their BPFs during a mean interval
of 24 months prior to our treatment. BPFs were located in the right main
bronchial stump (22), right lobar bronchial stumps (12), left main bronchial
stump (8), left lobar stump (1), trachea (2) and parenchyma (2). Patients were
treated by suture closure of the bronchial stump in 37, buttressed with
vascularized pedicle flaps of omentum (19), muscle (14), or pleura (3). In 10
cases, direct suture closure was not possible, and omental (7) or muscle (2)
flaps were sutured over the BPF. Suture closure without pedicle coverage and
simple drainage were performed successfully in one case each. Initial repair of
BPF was successful in 21 of 26 patients treated with omentum, in 11 of 16
patients treated with muscle and in 1 of 3 patients treated with pleural flaps.
In 12 patients with persistent or recurrent BPF after our initial repair, 7
underwent a second procedure (5 successful) and 5 were managed with drainage
only. BPF was successfully closed in 11 of 13 patients who had received
high-dose radiation therapy (9 with omentum). Overall, successful closure of
BPF was achieved in 40 of 47 patients (85%). Four in-hospital deaths resulted
from pneumonia and sepsis, 2 in patients with recurrent BPF after pleural flap
closure. In 25 patients, the empyema cavity was obliterated during definitive
repair of the BPF. The cavity resolved with drainage in 6 others, while 10
required a total of 17 Clagett procedures and 1 had a delayed myoplasty. Direct
surgical repair of chronic BPF may be achieved in most patients after adequate
pleural drainage by suture closure and aggressive transposition of vascularized
pedicle flaps. Omentum is particularly effective in buttressing the closure of
BPF.
*By invitation
51. PROGNOSIS AND MANAGEMENT OF MELANOMA
PATIENTS WITH PULMONARY METASTASES
Lorraine Tafra, M.D.*, Paul S.
Dale, M.D.*, Leslie A. Wanek, Dr.P.H.* and Donald L. Morton, M.D.
Santa Monica, California
Although melanoma that
metastasizes to distant sites is generally associated with a median survival of
only 6 to 8 months, certain metastatic sites including the lung may carry a
better prognosis than others. Surgical therapy of pulmonary metastases remains
controversial because of the variable survival rates reported for previous
small series. To determine the prognosis and optimal management of melanoma
patients with pulmonary metastases, we reviewed our 21-year melanoma database
of over 6, 000 patients. Of 953 AJCC stage IV patients with metastatic melanoma
involving the lung, 100 underwent surgical resection by unilateral/bilateral
thoracotomy or median sternotomy. Operative mortality was zero and median
follow-up was 61 months. The remaining 853 patients were managed nonsurgically
by immunotherapy, chemotherapy and/or radiation therapy. In both treatment
groups the male:female ratio was similar (approximately 2:1). The primary
lesion's Clark level of invasion and Breslow thickness, and the patient's age
at initial diagnosis and diagnosis of stage IV disease were not significantly
different between the two groups. Fifty-nine percent of the nonsurgical group
developed extrathoracic metastases, compared to 38% of the surgical group
(p<.001). The 1-year, 3-year and 5-year survival rates of surgical patients
were 77%, 34% and 28%, respectively, compared to 32%, 7% and 3% in nonsurgical
patients; these differences were highly significant (p<.001). Sixty-eight
percent of the surgical patients received some form of immunotherapy, compared
to 39% of the nonsurgical patients. Multivariate analysis revealed that both
surgery and immunotherapy were independent predictors of survival (p<.0001).
These results indicate that the prognosis associated with metastatic melanoma
may be less dismal when distant disease involves thoracic sites. We believe
that surgical resection is the treatment of choice for melanoma patients with
pulmonary metastases; when combined with immunotherapy, this regimen offers the
best chance for long-term survival.
*By invitation
52. RESULTS OF CALGB 8935: A
MULTI-INSTITUTIONAL PHASE II TRI-MODALITY TRIAL FOR STAGE IIIA (N2) NON-SMALL
CELL LUNG CANCER (NSCLC)
David J. Sugarbaker, M.D.*, James E. Herndon*,
Leslie J. Kohman, M.D.*, Hani Shennib, M.D.*, William C. Nugent, M.D.*, Mark
Krasna, M.D.*, Jemi Olak, M.D.*, Carolyn M. Dresler, M.D.*, Mark K. Ferguson,
M.D., Carolyn E. Reed, M.D.*, Frederic C. Kass, M.D.*, Parvesh Kumar, M.D.* ,
Michael T. Jaklitsch, M.D. *, Michael A. Maddaus, M.D.*, Nasser K. Altorki,
M.D.* and Mark R. Green, M.D.*
Boston, Massachusetts; Durham,
North Carolina; Syracuse and New York, New York; Montreal, Quebec, Canada;
Dartmouth, New Hampshire; Baltimore, Maryland; Richmond, Virginia; St. Louis,
Missouri; Chicago, Illinois; Charleston, South Carolina; Santa Barbara and San
Diego, California; Memphis, Tennessee and Minneapolis, Minnesota
Thirty
institutions participated in CALGB 8935, a phase II tri-modality trial for
patients (pts) with stage IIIA (N2) NSCLC. Two cycles of induction cisplatin
(P) 100 mg/m2 and vinblastine (V) 5 mg/m2/week preceded
standardized resection. Resected pts received two cycles adjuvant P and V
followed by 55 Gy radiotherapy (RT). From 10/89 to 2/92, 80 pts were accrued.
The age of 74 eligible pts ranged between 45 to 75 yrs. Histology: 43% squam,
35% adeno, 22% undiff. All pts were staged as IIIA by cervical and/or anterior
mediastinoscopy; 92% had more than 1 node biopsied and bronchoscopy identified
endobronchial tumor in 29%. There were no radiographic complete responses to
induction P and V; 65 (88%) had a response or stable disease, 6 (8%) developed
distant metastases, and 3 (5%) were unevaluable. Sixty-two (84%) pts underwent
thoracotomy, of whom 45/62 (73%) were resectable; 24/45 (53%) had lobectomy,
17/45 (38%) pneumonectomy, 2/45 (4%) sleeve resection, 2/45 (4%) lobectomy with
chest wall resection and all had radical lymphadenectomy. Operative mortality
was 3.2% (2/62 pts). Postoperative morbidity: pneumonia (9%), dysrhythmia (7%),
respiratory failure (3%), bronchopleural fistula (3%), and empyema (2%). Ten
pts (22%) were pathologically downstaged at resection; 4/45 (9%) stage II, 6/45
(13%) stage I. Twenty-three of 45 pts underwent complete resection. Twenty-two
pts were incompletely resected (7 positive margins, 11 positive highest node
sampled, and 4 both). Thirty-one pts of 45 resected (69%) completed adjuvant P
and V and RT. Median follow-up of pts currently alive is 23.6 months.
|
|
|
PATTERNS
OF RECURRENCE
|
SURVIVAL
|
|
|
N
|
Local
|
Distant
|
Both
|
None
|
1 yr
|
2 yr
|
3 yr
|
Eligible
|
74
|
|
|
|
|
64%
|
33%
|
19%
|
|
Resected
|
45
|
|
|
|
|
65%
|
43%
|
36%
|
|
Complete
|
23
|
4%
|
39%
|
9%
|
48%
|
68%
|
42%
|
42%
|
|
Partial
|
22
|
0%
|
27%
|
32%
|
41%
|
63%
|
44%
|
29%
|
Unresectable
|
17
|
|
|
|
|
35%
|
20%
|
0%
|
CALGB 8935 has shown in IIIA (N2) NSCLC: 1)
mediastinoscopy, induction chemotherapy, standardized surgical resection, and
adjuvant chemo and RT is feasible in a multi-institutional setting, 2) disease
progression during induction chemo is uncommon (8%), 3) single modality
induction therapy can yield acceptable resectability rates with low operative
mortality, 4) resection is associated with improved survival compared to
previous studies. These data lay the foundation for future randomized phase III
multi-institutional trials to determine the superiority of this approach to
single modality therapy.
12:00 p.m. ADJOURN
*By invitation
WEDNESDAY MORNING, APRIL 27, 1994
9:00 a.m. SIMULTANEOUS SCIENTIFIC SESSION F -
CONGENITAL HEART DISEASE
West Ballroom
Moderators: Richard A. Jonas, M.D.
Julie A. Swain, M.D.
53. THE NORWOOD OPERATION AND SUBSEQUENT
FONTAN OPERATION IN INFANTS WITH COMPLEX CONGENITAL HEART DISEASE
Paul W. Weldner, M.D.* and John
L. Myers, M.D.
Hershey, Pennsylvania
From April 1987 to September
1993, 60 infants underwent a Norwood operation for complex congenital heart
disease including hypoplastic left heart syndrome (HLHS) (n=41), ventricular
septal defect (VSD) and subaortic stenosis (SubAS) with arch interruption/severe
coarctation (IAA) (n=6), complex single right ventricle (SRV) with SubAS (n=8),
critical aortic stenosis with endocardial fibroelastosis (n=3), and malaligned
primum atrial septal defect with coarctation (n=2). Age at operation ranged
from 1 day to 3.9 months (mean 10.5 days, median 3.5 days).
The overall operative mortality
(<30 days) was 30% (20 patients). Late mortality was 17% (10 patients)..
Fifteen of the 20 (75%) operative deaths occurred suddenly during the first two
days postop from sudden hemodynamic instability. All four infants with
premature closure of the foramen ovale had pulmonary lymphangiectasia and died
from pulmonary failure. There have been 7 operative deaths in 36 patients since
1990 (19%) and in the last 2 years there have been no operative deaths in 22
patients.
Overall, there
are 30 late survivors (50%). Nineteen of these 30 infants have undergone a
modified Fontan operation at 9.7 to 27.6 months of age (mean 18.1 months) with
no mortality. Another six patients have undergone a hemi-Fontan at 6.8 to 23.0
months (mean 11.7 months) with no mortality. In our early experience, infants
undergoing the Norwood operation had a high early mortality, related to sudden
hemodynamic instability. Following the institution of a protocol for the addition
of CO2 to the inspired gas during postop mechanical ventilation
there has been no operative deaths. The operative mortality for the Norwood
operation continues to improve. Subsequent Fontan operation can be performed
with excellent clinical results.
*By invitation
54. WHAT AFFECTS VENTRICULAR
CHARACTERISTICS AFTER A FONT AN TYPE OPERATION?
Hideki Uemura, M.D.*, Toshikatsu Yagihara, M.D.*,
Yasunaru Kawashima, M.D., Fumio Yamamoto, M.D.*, Osamu Matsuki, M.D.* and
Robert H. Anderson, M.D.*
London, England and Osaka,
Japan
Postoperative conditions after a Fontan type
operation, particularly as they affect results in the early term, are thought
to depend on factors such as the state of the pulmonary circulation and
ventricular function, as well as the sequels of the operative procedure. In
this study, we have attempted to determine the factors affecting ventricular
characteristics in the middle term after Fontan-type procedures.
Postoperative
catheterization was performed at a mean of 1.4 years after surgery in 50
patients with univentricular atrioventricular connection who underwent the
operation at the age of 1.0-22.6 years. End diastolic volume (EDV: % of
anticipated normal value), ejection fraction (EF) and end diastolic pressure
(EDP) of the systemic ventricle were analyzed in addition to calculation of the
cardiac index (CI).
As shown on the
table, these parameters were affected by presence of residual or recurrent
atriovent
