TUESDAY AFTERNOON, APRIL 25, 1995
1:45 am SIMULTANEOUS
SCIENTIFIC SESSION A ADULT CARDIAC SURGERY
Auditorium,
Hynes Convention Center
Moderators: Tirone E. David, M.D.
Bruce A. Reitz, M.D.
20. HOMOGRAFT MITRAL VALVE REPLACEMENT:
MEASUREMENT TECHNIQUES FOR GRAFT SELECTION AND IMPLANTATION
Christophe Acar, M.D.*, Jullien
Gaer, M.D.*, Alain Berrebi, M.D.*, Denis Tixier, M.D.*, Christian Brizard,
M.D.*, Arnaud Farge, M.D.*, Alain Deloche, M.D.* and Alain F. Carpentier, M.D.,
Ph.D.
Paris, France and London, England
Homograft
replacement of the mitral valve (HRMV) has experienced a renewal of interest
since the introduction of improved methods of homograft preservation. However,
the reliability of this technique remains a major concern due to the complexity
and the frequent variations of the subvalvular apparatus. Our experience of 40
HRMV (19 total, 21 partial) carried out in the past 3 years made it possible to
define precise measurement criteria which proved to be useful to achieve
predictable results. The indications for HRMV were bacterial endocarditis in 16
patients and rheumatic valvular disease in 24 patients.
Measurement on the
homograft were the following: 1) Circumference (assessed by prosthetic ring
sizers), 2) Height of the anterior leaflet (average 23 ± 3 mm), 3) Distance
from annulus to apex of the anterior papillary muscle (PM) (average 23 + 4 mm).
In addition, the morphology of each papillary muscle was classified as follows:
Type I, simple head (n=20); Type II, bifid head (n=52); Type III, complex PM
(n=8).
Measurements on
the recipient were obtained by transesophageal echocardiography (TEE): 1)
Circumference of the annulus, 2) Distance from annulus to apex of the anterior
PM which averaged 20 ± 3 mm in the bacterial endocarditis group and 8 ± 2 mm in
the rheumatic valve group.
The homograft
valve was selected to match as much as possible these measurements. Further
adjustments were made by systematic use of a Carpentier-Edwards ring selected
from the size of the anterior leaflet of the homograft and by side-to-side
implantation of the graft PM to the recipient PM. Associated procedures
comprised aortic valve repair (n=3), or replacement using homograft (n=2) and
tricuspid valve repair (n=1), or replacement using homograft (n=1).
There has been
1 hospital death due to low cardiac output and 1 reopera-tion 10 days following
the operation due to annulus dehiscence. The follow-up extended from 3 months
to 34 months (average 8 months). There has been 1 late death due to lung
cancer. All the remaining 37 patients were reviewed for this study and a TEE
was performed. All patients were in NYHA class I (n=33) or II (n=4). TEE showed
no mitral valve regurgitation (MVR) in 18 pts, trivial MVR in 15 and moderate
MVR in 4. The valve surface area was 2.0 ± 0.3 cm2 for partial HRMV and 2.4 ±
0.2 cm2 for total HRMV with an average transvalvular gradient of 3 ±
2 mmHg.
This study
shows that reliable results can be obtained with HRMV, a technique particularly
useful for bacterial endocarditis in the adult and rheumatic valvular disease
in children.
21. RECONSTRUCTION OF THE MITRAL ANNULUS -
A TEN-YEAR EXPERIENCE
Christopher M. Feindel, M.D.*,
Tirone E. David, M.D., Susan Armstrong, M.Sc.* and Zhao Sun, B.A.*
Toronto, Ontario, Canada
The mitral
annulus (MA) may be damaged by extensive calcification, infective endocarditis
or repeated mitral valve (MV) replacement making MV surgery hazardous. It is
possible to reconstruct the MA with either autologous or bovine pericardium.
All calcific or infected tissue is excised and a properly tailored pericardia!
patch is sutured to the endocardium of the left ventricle posteriorly and to
fibrous tissue underneath the aortic valve superiorly. If this fibrous tissue
is also diseased, it can be replaced with pericardium, too. Thus, the MA can be
repaired segmentally, circumferentially or along the entire base of the left
ventricle.
Seventy-six pts
have undergone reconstruction of the MA since 1984. There were 35 men and 41
women, ages 17 to 86 years, mean 58. The indications for reconstruction of the
MA were: calcification in 19 pts (22%), MA abscess in 23 (30%), redo MV surgery
in 30 (39%), and type I rupture of the left ventricle after MV replacement in 4
(5%). Three-fourths of the pts were in NYHA functional class IV and 18 were in
septic and/or cardiogenic shock when operated on; 11 pts had coronary artery
disease; and 48 pts had had at least one previous MV operation. The
reconstruction of the MA was segmen-tal in 40 pts (52%), circumferential in 21
(27%), and involved both aortic and mitral annuli in 15 (20%). The MV was
replaced in 70 pts and repaired in 6. There were 4 early reoperations in 3 pts:
1 for dehiscence of the patch, 1 for thrombosis of the valve, and 2 for new
onset endocarditis. The operative mortality was 9% (7 pts). Pts have been
followed up from 3 to 114 months, mean of 34. Every pt has had Doppler
echocardiography annually. There has been no late patch dehiscence. Four pts
have required reoperation: 2 for endocarditis, and 2 for failed bioprosthesis.
There have been 13 late deaths, 8 cardiac and 3 valve-related. The actuarial
survival at 5 years was 74% ± 4%.
Reconstruction of the MA
with pericardium is an extremely useful operative technique in pts with damaged
annulus, and provides excellent late results.
*By invitation
22. ATRIOVENTRICULAR VALVE REPAIR USING
EXTERNALLY ADJUSTABLE FLEXIBLE RINGS
João Q. Melo, M.D.*, José S.
Neves, M.D.*, Regina Ribeiras, M.D.*, Miguel Abecasis, M.D.*, Manuel Canada,
M.D.*, Maria J. Rebocho, M.D.*, Narciso C. Andrade, M.D.* and Manuel M. Macedo,
M.D.
Lisbon, Portugal
Mitral and
tricuspid valve repair using annuloplasty rings have a significant incidence of
residual regurgitation which leads either to suboptimal results or to the need
of a reoperation. We have developed a new technique for partial annuloplasty
using a flexible externally adjustable ring. If required, it is adjusted from
outside the heart after hemodynamic stabilization is obtained following
interruption of extracorporeal circulation. Adjustment of the ring is controlled
using transesophageal echocardiography. Twenty patients (pts), with a mean age
of 45 years, ranging from 7 to 70 years, had 25 rings implanted, 14 in mitral
position and 11 in tricuspid position. Most pts (12) were rheumatic in origin.
Eight pts had pure mitral insufficiency and the other mixed lesions. All
tricuspid lesions were regurgitation. Eight pts had complex mitral repairs and
8 pts had associated procedures: aortic repair (1 pt), aortic replacement (3
pts), maze 3 operation (2 pts), left atrial isolation (2 pts), coronary artery
bypass (1 pt). Mitral regurgitation before surgery was grade 4 in 5 pts, grade
3 in 6 pts and grade 2 in 3 pts. After surgery it reduced to grade 2 (1 pt),
grade 1 (5 pts) or none (8 pts). Tricuspid regurgitation reduced from grade 4
(4 pts), grade 3 (3 pts) and grade 2 (3 pts) to grade 1 (3 pts) and no
regurgitation (7 pts). Mean pulmonary artery pressure was reduced from 67 to 40
mmHg.
External
adjustment was performed in 10 mitral rings and in 8 tricuspid rings, improving
the regurgitation by 1 or 2 grades. There was no hospital or late mortality. At
6 months all pts are in functional Class I. One pt had a change of its residual
mitral regurgitation from grade 1 to grade 2. The other had the same mitral or
tricuspid evaluation as at hospital discharge.
We conclude
that flexible, externally adjustable rings are safe and provide excellent early
results. The external adjustment feature provides the possibility of a
significant reduction of the incidence of residual regurgitation after mitral
and/or tricuspid annuloplasty repair.
2:45 pm INTERMISSION - VISIT EXHIBITS
*By invitation
3:15 pm SIMULTANEOUS SCIENTIFIC SESSION
A ADULT CARDIAC SURGERY
Auditorium, Hynes Convention Center
Moderators: Tirone E. David, M.D.
Bruce A. Reitz, M.D.
23. 1026 PATIENTS UNDERGOING COMBINED
CORONARY AND VALVULAR SURGERY: EARLY AND LATE PHASE
EVENTS
Glenn W. Laub, M.D.*, Bridget M.
Bailey, B.S.N.*, Chao L. Chen, Ph.D.*, Robert S. Clancy, B.S.N.*, Javier
Fernandez, M.D., William A. Anderson, M.D.*, Joseph T. Costic, D.O.* and Lynn
B. McGrath, M.D.
Browns Mills, New Jersey
Between January
1981 and December 1991, combined coronary and valve surgery was perfomed in
1,026 patients. Coronary revascularization was combined with aortic valve (AV)
surgery in 551 (53.7%), mitral valve (MV) in 354 (34.5%), multiple valve in 119
(11.6%) and isolated pulmonic or tricuspid in 1 (0.2%) each, with valve repair
being performed in 21% of procedures. Mean age was 66.7 years (range 31-90
yrs), 61% were male, and 56% were in CHF Class III and IV. There were 117
(11.4%) hospital deaths, with surgery on the MV being the most important risk
factor, increasing mortality to 16% as compared to 7% without (p=0.001). CHF,
cross clamp time, female gender, and previous surgery were also predictors of
mortality, while extent of coronary disease, NYHA Class, and ejection fraction
were not. Total follow-up was 5,021 pt-years (mean 5.5, range 1-13.3 yrs) and
99.9% complete. Actuarial survival at 5 and 10 yrs was 70 ± 2% and 42 ± 2% (70%
CL), with survival of AV (74 and 48%) significantly longer than MV (67 and 32%,
p=0.002). Risk factors for late death included MV surgery, preop NYHA and CHF
Class, and liver disease. Survival with a mechanical valve was better than with
a tissue valve at 10 yrs for AV (61 vs 34%, p=.0001), but not different for MV.
The use of the internal mammary artery to the left anterior descending
conferred a survival advantage at 10 yrs (59 vs 42%, p=0.02), while the number
of vessels bypassed showed no effect. Of 415 late deaths, 33 (7.9%) were valve
related: thromboembolic (TE) 10, prosthetic valve endocarditis 10,
anticoagulant bleed (ACRH) 6, structural valve dysfunction (SVD) 6 and
paravalvular leak (PVL) 1. Cardiac, non-valve related deaths (229) were comprised
predominately of progressive heart failure or myocardial ischemia in 203
patients (89%). Postoperative angina occured in 85 (9.4%) and graft occlusion
in 29 (3%) patients. Late survivors fared well as 453 (92%) were in CHF Class I
and II. Overall actuarial freedom from complications at 10 years was 86.8% for
TE, 84.2% from ACRH, 94.5% for PVL, 88.5% for SVD, 99.6% for valve thrombosis,
and 84.1% for MI. Freedom for SVD at 8 yrs was far better with mechanical than
tissue (98 v 71%, p=.01) as was freedom for re-operation at 10 yrs (AV:91 v
72%, p=.002; MV:79 v 64%, p=.04). Tissue valves had a superior freedom from
ACRH at 5 yrs for AV (94 v 89%, p=.002) and MV (94 v 84%, p=.007), but TE rates
were similar. Valve repair was not associated with a reduction in early or late
phase events.
We conclude that although
late survival was markedly lower than in an age, sex, race matched population
(42 v 65%, 10 yr), the survivors' functional results are good. In addition, the
use of the IMA and mechanical valve in the aortic position confer substantial
long term survival advantages and freedom from reoperation in combined
procedures.
*By invitation
24. REVASCULARIZATION OF THE CIRCUMFLEX
ARTERY WITH THE PEDICLED RIGHT INTERNAL MAMMARY ARTERY: CLINICAL, FUNCTIONAL
AND ANGIOGRAPHIC MIDTERM RESULTS
Michel Buche, M.D.*, Erwin
Schroeder, M.D.*, Patrick Chenu, M.D.*, Olivier Gurne, M.D.*, Giulio Pompilio,
M.D.* Baudouin Marchandise, M.D.*, Philippe Eucher, M.D.*, Yves Louagie, M.D.*,
Robert Dion, M.D.* and Jean-Claude Schoevaerdts, M.D.*
Yvoir, Belgium
Sponsored by: Albert Starr,
M.D., Portland, Oregon
Retroaortic
crossing of the pedicled Right Internal Mammary Artery (RIMA) for
revascularization of the Circumflex Artery (Cx) combined with a pedicled Left
Internal Mammary Artery (LIMA) to the Left Anterior Descending Artery (LAD) and
its branches is an attractive technique to achieve a complete arterial
revascularization of the left ventricle. However there is a suspicion that
pulling the RIMA through the transverse sinus could compromise its blood flow
capacity and patency.
Material:
Between January 1990 and July 1994, this technique was applied in 256 patients
(202 male; 54 female; average age 62 years; range 31 to 80 years). 61 patients
had a 2-vessel disease and 195 had a 3-vessel disease. 17 patients were
undergoing a reoperation. 22 had a left ventricular ejection fraction <40%.
30 were diabetecs. The RIMA was directed to the Cx artery through the
transverse sinus (259 anastomoses) and the LIMA was anastomosed to the LAD and
its branches (375 anastomoses) in all patients. 195 patients with a 3-vessel
disease received additional coronary artery bypasses to the Right Coronary
Artery (93 Saphenous Vein grafts; 89 free Inferior Epigastric Artery; 12
pedicled Right Gastroepiploic Artery). In total, the patients received an
average of 3.2 distal anastomoses (2.4 IMA anastomoses).
Clinical
results: 3 patients died early and 8 had a nonfatal myocardial infarction
(MI). 7 patients required postoperative intraaortic balloon pumping. 6 patients
underwent early reoperation for excessive bleeding. Sternal wound complications
occurred in 4 patients. One of these 4 patients died of the complication 10
months after the operation. No patient was lost for follow-up (average 28
months). During follow-up, one sudden death and 4 noncardiac deaths occurred. 2
patients experienced a nonfatal MI and 12 had recurrence of angina. There was
no late reoperation.
Graft
patency and functional assessment: 73 patients enrolled in a prospective
angiographic study agreed to undergo a postoperative angiogram (average 12.6
months, range 6 to 24 months). 72/73 of the RIMA were patent. In comparison
73/73 of the LIMA (104/105 LIMA anastomoses) were patent. Stress Thallium
Scintigraphy obtained in 25 of those patients did not reveal ischemia in the Cx
area.
Conclusions:
Complete arterial revascularization of the left ventricle by means of both
pedicled IMA can be performed with acceptable mortality and morbidity.
The midterm
patency rate of the pedicled RIMA passed through the transverse sinus and
anastomosed to the Cx artery is excellent and does not differ of the patency
rate of the LIMA anastomosed to the LAD.
*By invitation
25. RISK FACTORS FOR POSTOPERATIVE
MORBIDITY
Victor A. Ferraris, M.D., Ph.D.*
and Suellen P. Ferraris, Ph.D.*
Albany, New York
Sponsored by: L. Henry
Edmunds, Jr., M.D., Philadelphia, Pennsylvania
Background:
Analysis of outcomes after coronary artery bypass grafting (CABG) has
focused on risk factors for operative motality. Non-fatal peri-operative
morbidity is far more costly and more common after CABG and the factors that
predispose to morbid events are not well documented. In order to identify the
risk factors that lead to postoperative morbidity, we evaluated 938 patients
undergoing CABG at our institution during 1993.
Methods:
Univariate statistical analysis was performed on 46 patient variables in
order to identify risk factors for either serious postoperative morbidity
(stroke, perioperative Q-wave myocardial infarction, deep sternal infection,
life-threatening postoperative hemorrhage) or increased hospital length-of-stay
(LOS). Variables were considered both individually and in combination. For
example, age was considered individually or in combination with other variables
including parameters of blood volume (i.e., age divided by red blood cell
volume or Age/RBCVOL), renal function, cardiac function, etc. Univariate
analysis of these 46 variables revealed 16 that were significant risk factors
for serious postoperative morbidity or increased LOS. These 16 variables were
entered into multivariate analyses to determine independent risk factors for
serious postoperative morbidity (logistic regression) or increased LOS (Cox
proportional hazards regression). Similar multivariate analysis was performed
to identify independent risk factors for hospital mortality (logistic
regression).
Results:
Hospital length-of-stay (LOS) was correlated with serious postoperative
morbidity (Spearman rank correlation coefficient = -0.31, p<0.001) but not mortality
(correlation coefficient = 0.03, p=0.38). In order of decreasing importance,
the following patient variables were significantly associated with increased
LOS by stepwise Cox proportional hazards analysis: Age/RBCVOL (p<0.001),
history of CHF (p<0.001), CPB time (p<0.001), hypertension (p=0.004),
chronic obstructive lung disease (p=0.002), and previous stroke (p=0.041). In
order of decreasing importance the following patient variables were
significantly associated with the occurrence of any serious postoperative
morbidity by logistic regression: history of CHF (p<0.001), Age/RBCVOL:
(p=0.001), hypertension (p=0.004), CPB time (p=0.003), and previous stroke
(p=0.03). The ratio of age divided by the RBCVOL (blood volume multiplied by
hematocrit) was an important risk factor for both increased LOS and serious
postoperative morbidity. Interestingly, variables that were significant
independent predictors of increased mortality such as, preoperative shock,
previous open heart operation, renal failure, ejection fraction less than 20%,
and MI within 6 hours of operation, were not risk factors for either serious
morbidity or increased LOS. Using the variables and relative risk values
identified in this analysis, it is possible to develop a scoring system that assigns
morbidity risk scores to patients before operation.
Conclusions:
We conclude that: 1) hospital length-of-stay is a continuous variable that
correlates with postoperative morbidity but not with mortality, 2) advanced age
and low red blood cell volume are very important risk factors for serious
postoperative morbidity and increased LOS, and 3) risk factors for
postoperative morbidity are different than those for postoperative mortality.
These results suggest that older patients with preoperative anemia and low
blood volume, who also have other co-morbidities (CHF, stroke, COPD or
hypertension), are at increased risk for postoperative complications. These
studies allow identification of a high-risk cohort of patients who are likely
candidates for interventions to lessen postoperative morbidity.
*By invitation
26. IS BODY SIZE THE CAUSE FOR POOR
OUTCOMES OF BYPASS SURGERY IN FEMALES?
George T. Christakis, M.D., Karen
Buth, M.Sc.*, Richard D. Weisel, M.D., Vivek Rao, M.D.*, Kostas P.
Panagiotopoulos, B.Sc.*, Joan Ivanov, R.N.* and Stephen E. Fremes, M.D.,
Toronto, Ontario, Canada
Females have a higher
incidence of operative mortality (OM), low output syndrome (LOS), myocardial
infarction (MI), postoperative bleeding (BLD), and infection (INF) than males (M)
following CABG. The poor outcomes have been attributed to the smaller size of
females and (by extension) to smaller coronary diameter. In order to assess
risk factor differences between males and females, and the influence of body
size on postoperative outcome, data were gathered prospectively on 7,025
patients (M=5,694, F=l,331) proceeding to isolated CABG between January 1990
and June 1994. Univariate and multivariate statistics were used to determine
risk factors for CABG in M and F separately. Body size was assessed by both
body mass index (BMI) and body surface area (BSA).
|
|
|
%
OM
|
%
LOS
|
%
MI
|
%
BLD
|
%
INF
|
|
BMI
(kg/m2)
|
"pts"
|
M
|
F
|
M
|
F
|
M
|
F
|
M
|
F
|
M
|
F
|
|
<20
|
125
|
3.5
|
7.5
|
13.0
|
20.0
|
3.5
|
5.1
|
3.5
|
2.5
|
2.3
|
15.0
|
|
20-25
|
1700
|
2.5
|
3.8
|
8.0
|
15.7*
|
3.2
|
5.8*
|
2.2
|
2.0
|
2.3
|
3.5
|
|
25-30
|
3169
|
1.8
|
3.9*
|
6.3
|
14.3*
|
2.8
|
5.5*
|
1.3
|
2.4*
|
3.5
|
5.1
|
|
>30
|
2031
|
1.6
|
3.1*
|
5.1
|
12.1*
|
2.8
|
5.3*
|
1.0
|
2.4
|
6.3
|
9.9*
|
*=M diff than F p<0.05 =<20 diff than >30,
p<0.05 for M
Small body size
(BMI<20) resulted in increased morbidity postoperatively for both sexes.
However, within each classification of body size females continued to have a
higher morbidity and mortality. In patients with heights less than 164 cm, the
incidence of OM (M=2.2%, F=3.7%) and LOS (M=7.9%, F=14.2%) was higher
(p<0.05) for females. In patients with BSA<1.7 m2, the
incidence of OM (M=3.8%, F=3.8%) and LOS (M=l 1.4%, F=15%) were similar for
both sexes. BSA<1.7m2 accounted for 47% of F but only 8% of M.
For patients with BSA>1.7m2, OM (M=1.7%, F=3.8%) and LOS (M=6.3%, F=14%)
were higher (p<0.0001) in females. The multivariate risk factors (odds
ratios) for OM in M were: age>70 (2.7), LVEF<40% (2.6), diabetes mellitus
(DM) (1.6), peripheral vascular disease (PVD) (2.3), smoking (1.9), renal
failure (2.5), preoperative MI (PMI) (2.0) and for F the predictors were:
age>70 (2.0), urgent surgery (US) (2.5), LVEF<40% (3.2), PVD (2.8). The
predictors of LOS in M were: age>70 (1.6), LVEF<40% (2.2), reoperative
surgery (REDO) (5.6), left main stenosis (LMS) (1.3), endarterectomy (EA) (1.8),
DM (1.6), congestive failure (2.1), PMI (2.0). LOS predictors for F were: US
(1.6), LVEF<40% (2.5), LMS (1.7), REDO (4.2), EA (3.2).
CONCLUSION:
Risk factors for OM and LOS were similar for males and females. Small body size
increases morbidity following isolated CABG. Independent of body size, females
still have a higher operative mortality and morbidity. Increased risk of
surgery for F may be due to differences in atherogenesis.
4:35 pm EXECUTIVE SESSION (Limited to
Members)
Republic Ballroom, Sheraton Boston Hotel
6:30 pm MEMBER RECEPTION
Museum of Fine Arts
*By invitation
1:45 pm SIMULTANEOUS SCIENTIFIC SESSION
B GENERAL THORACIC SURGERY
Republic Ballroom, Sheraton Boston Hotel
Moderators: J. Kent Trinkle, M.D.
Valerie W. Rusch, M.D.
27. MUTATION IN THE P53 TUMOR SUPPRESSOR
GENE IN BARRETT'S ESOPHAGUS IS A MARKER FOR DEVELOPMENT OF ADENOCARCINOMA:
RESULTS OF AN INTERNATIONAL MULTI-INSTITUTIONAL PROSPECTIVE STUDY
Paul M. Schneider, M.D.*, Alan G.
Casson, M.D.*, Harinder S. Garewal, M.D.*, Bernard Levin, M.D.*, A.H.
Hoelscher, M.D.*, J. Rudinger Stewart, M.D.* and Jack A. Roth, M.D.
Munich, Germany, Toronto,
Ontario, Canada,
Houston, Texas and Tucson, Arizona
Current histologic
criteria are not adequate to predict which patients with Barrett's esophagus
(BE) are at high risk of developing invasive cancer and thus are candidates for
esophagectomy. The purpose of this study was to determine the value of p53
mutation in BE as a marker for the development of adenocarcinoma. The p53 gene,
a tumor suppressor gene whose function is critical to cell cycle control and
DNA repair, is the most commonly mutated gene in human cancers. We had
previously identified p53 mutations in BE and therefore began a
multi-institutional study to determine their significance as a marker for
malignancy. Ninety-seven patients from four institutions were studied.
Forty-seven (37 males and 10 females, mean age 55) had BE with metaplasia or
dysplasia but no evidence of malignancy at a mean follow-up of 2.2 ± 0.2 years.
The BE was classified as metaplasia in 30, low-grade dysplasia in 12, and
high-grade dysplasia in 5. The other 50 patients (46 males and four females,
mean age 63) had adenocarcinoma arising in BE (BC). Tissue for DNA analysis,
including normal stomach or esophagus, tumor, and BE, was obtained by
endoscopic biopsy in patients with BE or BC or during surgery in some patients
with BC. Exons 5-9 of the p53 gene were studied for mutations by single-strand
conformational polymorphism analysis (SSCP) after polymerase chain reactions.
Mutations detected by SSCP were confirmed by direct DNA sequencing. None of the
tissue samples from patients with BE alone with metaplasia or low-grade
dysplasia had a p53 mutation, and only one of five patients with high-grade
dysplasia but no evidence of invasive malignancy had a p53 mutation. However,
23 of the BC patients (46%) had a p53 mutation in Barrett's epithelium, tumor,
or both. Twenty of these had a p53 mutation in the tumor only (N=16) or in both
(N=4), suggesting a direct role for the mutation in carcinogenesis. In the
other three, a mutation was present in the Barrett's epithelium but not the
tumor, suggesting that p53 mutation may be a marker for genetic instability
predisposing to tumor formation. Biopsies of BE were classified as metaplasia or
low-grade dysplasia for 16 patients with a p53 mutation and cancer, indicating
that p53 mutation is associated with cancer regardless of the degree of
clinically detectable dysplasia. We conclude that mutation in the p53 gene
identifies a subset of BE patients at increased risk for developing cancer.
*By invitation
28. ERYTHROMYCIN STIMULATES GASTRIC
EMPTYING STATUS POST-ESOPHAGECTOMY: A RANDOMIZED CLINICAL TRIAL
Michael Burt, M.D., Ph.D.,
William Williard, M.D.*, Andrew Scott, M.D.*, Samuel Yeh, M.D.*, Manjit Bains,
M.D., Allan Turnbull, M.D.*, Joseph Former, M.D.*, Patricia McCormack, M.D. and
Robert Ginsberg, M.D.
New York, New York
Delayed gastric
emptying following esophagogastrectomy can pose a significant early
post-operative problem. Since erythromycin, which stimulates the gastric antral
and duodenal motilin receptor, has been demonstrated to significantly increase
gastric emptying in patients with diabetic gastroparesis, we decided to
evaluate its effect on gastric emptying after esophagogastrectomy.
Methods:
24 patients (age: 41-79 yr, median 66; 18 men/6 women) were randomized to
receive either erythromycin lactobionate (200 mg in 50ml NSIV, n=13) or placebo
(50 ml NS IV, n=11) eleven days after esophagogastrectomy (with pyloric
drainage procedure). After infusing erythromycin or placebo over 15 min,
patients ate a solid meal (scrambled egg with bread) labelled with [99mTc]
sulfur colloid (500 |aCi) over approximately 15 min. Dynamic images of the
post-surgical stomach were then acquired over 90 min in the supine position by
gamma counter. Results were expressed as percentage of counts retained in the
stomach (% gastric retention) over time.
Results:
There were no side effects of eryhromycin. The figure displays the gastric
emptying over time for both groups.
Analysis of covariance demonstrated that the rate of
gastric emptying (slope of the line) was significantly greater in the
erythromycin treated group compared to the placebo group (p<0.0001).
Conclusion:
Early satiety following esophagogastrectomy is secondary to decreased
gastric motility and not secondary to a decrease in the gastric reservoir.
Intravenous erythromycin significantly improves gastric emptying in patients
following esophagogastrectomy by stimulating gastric motility.
*By invitation
29. THE ROLE OF FUNDOPLICATION IN THE
TREATMENT OF TYPE II PARAESOPHAGEAL HERNIA
Clark B. Fuller, M.D.*, Jeffrey
H. Peters, M.D.*, Manfred Ritter, M.D.*, Cedric G. Bremner, M.D.* and Tom R.
DeMeester, M.D.
Los Angeles, California
The role of fundoplication
in patients with pure Type II paraesophageal hiatal hernia remains
controversial. Conventional thinking suggests that because the lower esophageal
sphincter (LES) is located within the abdomen, it is competent, and
fundoplication unnecessary. Few studies have used objective evaluation to guide
the addition of an antireflux procedure.
Patients and
Methods: Sixteen consecutive patients with Type II paraesophageal hernia
were treated from May 1991 to July 1994. All demonstrated radiographic criteria
of pure Type II hernias. Evaluation before surgery included upper endoscopy,
esophageal manometry, and 24-hour ambulatory pH monitoring. The lower
esophageal sphincter was considered incompetent if any of the following
criteria were present; a resting pressure <6 mmHg, abdominal length <1 cm
or total length <2 cm. Primary symptoms responsible for surgery were related
to the hernia in 81% of patients postprandial pain (n=7), abdominal distension
(n=5), vomiting (n=1), and anemia (n=0). Symptoms typical of gastroesophageal
reflux disease were noted in 3 patients heartburn in 2 and regurgitation in 1.
Results: Objective
evidence of gastroesophageal reflux disease was present in the majority of
patients. Five patients (31%) had evidence of esophageal mucosal injury:
esophagitis in 3, stricture and esophageal ulceration in 1 patient each. Eleven
of 15 patients (69%) were found to have pathologic esophageal acid exposure on
24-hour pH monitoring. Twelve (75%) patients had a defective LES usually
secondary to an inadequate intra-abdominal length (8/12, 66%). Hernia
reduction, crural closure and Nissen fundoplication was performed in 14/16
patients (2 patients await definitive surgery). Symptomatic relief has been
excellent/good in all patients. No patients has developed hernia recurrence at
an average follow-up of 8.8 months (range 2-28 months).
Conclusion: Objective
evaluation reveals that gastroesophageal reflux accompanies type II
paraesophageal hernia in a high proportion of patients, usually secondary to an
incompetent LES. Appropriate treatment includes reduction of the hernia, crural
closure and fundoplication in most if not all patients.
2:45 pm INTERMISSION - VISIT EXHIBITS
*By invitation
3:15 pm SIMULTANEOUS SCIENTIFIC SESSION
B GENERAL THORACIC SURGERY
Republic Ballroom, Sheraton
Boston Hotel
Moderators: J. Kent Trinkle, M.D.
Valerie W. Rusch, M.D.
30. OBJECTIVE ASSESSMENT OF
GASTROESOPHAGEAL REFLUX AFTER SHORT ESOPHAGOMYOTOMY FOR ACHALASIA USING
MANOMETRY AND PH MONITORING
John M. Streitz, Jr., M.D.*, F.
Henry Ellis, Jr., M.D., Ph.D., Warren A. Williamson, M.D.*, Michael E. Glick,
M.D.*, Johannes A. Aas, M.D.* and Robert L. Tilden, Dr.P.H.*
Duluth, Minnesota and
Burlington, Massachusetts
The role of an
antireflux procedure as an adjunct to esophagomyotomy for achalasia remains
controversial. In our experience, a short myotomy alone is followed by a low
incidence of severe reflux symptoms (4-5%). However, little objective
documentation exists of this operation's effect upon sphincteric competence and
the degree of postoperative gastroesophageal reflux. This report provides such
documentation.
We performed
esophageal manometry and 24 hour pH monitoring on 14 patients with esophageal
achalasia who had previously undergone a short esophagomyotomy without antireflux
procedure by us. Eight of the 14 were tested to evaluate the postoperative
symptom of heartburn, and the remaining six asymptomatic patients were studied
to assess postoperative sphincteric function.
Esophagomyotomy
reduced lower esophageal sphincter (LES) pressure by 12 to 71% (mean 41%) from
a preoperative mean of 27 mmHg to 14 mmHg postoperatively. The number of
postoperative acid reflux episodes per patient in 24 hours was below 29 (normal
<47) in 12 patients, with a median of 12 episodes for the entire group.
Esophageal acid exposure, measured as % total time with pH <4.0 (normal
<4.5%), was below 4.5% in 9 patients, 6 of whom had values less than 1%. Of
the five patients with values greater than 4.5%, a temporal correlation of
symptoms with an episode of acid reflux occurred in only one. Multivariate
analysis showed that esophageal acid exposure time correlated only with the
level of residual LES pressure during the relaxation phase of deglutition. A
pressure less than 8 mmHg was predictive of normal acid contact time
(p<.018). Age greater than 42 was a covariant and predictive of normal acid
exposure by predicting low residual pressure (p<.001), irrespective of
follow-up duration. Mean LES pressure, percent reduction in LES amplitude,
postoperative vector volume and length of the LES did not significantly
correlate with amount of esophageal acid exposure.
We conclude that a short
esophagomyotomy without an antireflux procedure results in a competent LES in
most patients, and that symptomatic gastro-esophageal reflux is unusual.
Increased esophageal acid exposure, when it occurs, is due to slow esophageal
acid clearance of relatively few reflux episodes, not from too extensive a
myotomy, and is more likely to occur when the myotomy results in a high
residual pressure during deglutition, a finding more common in young patients.
In view of these findings, the addition of an antireflux procedure to a short
esophagomyotomy would not be expected to improve results.
*By invitation
31. COMBINED THORACOSCOPIC AND
LAPAROSCOPIC LYMPH NODE STAGING IN ESOPHAGEAL CANCER: THE UNIVERSITY OF
MARYLAND EXPERIENCE
Mark J. Krasna, M.D.*, John L.
Flowers, M.D.*, Safuh Attar, M.D. and Joseph S. McLaughlin, M.D.
Baltimore, Maryland
Unlike
mediastinoscopy in lung cancer, there exists no standard minimally invasive
test to stage esophageal cancer. If it were possible to obtain exact
preoperative staging in esophageal cancer, patients could be separated
prospectively to receive adjuvant therapy appropriately. We studied the
feasibility and efficacy of thoracoscopic lymph node staging (TSLN) and
laparoscopic lymph node staging (LSLN) in esophageal cancer. TSLN was performed
in 45 patients with biopsy proven carcinoma of the esophagus. LSLN was done in
the last 19 pts (after missing celiac LN in 3 pts). TSLN was aborted in 3 pts
due to adhesions. LN stage was NO in 39 pts, N1 in 3 and N3 (celiac LN) in 6
patients. Esophageal resection was performed in 30 patients after TSLN; 17 of
these underwent LSLN. These patients form the basis for subsequent data
analysis.
TSLN staging
showed NO lymph node status in 28 patients and N1 in 2 patients. Two of the 28
NO patients (7%) were found at resection to have para-esophageal lymph node
involvement (N1) and were thus understaged by TSLN. One patient with Nl disease
at thoracoscopy had no lymph node involvement (NO) after esophagectomy. The
other patient remained N1 at the time of resection. Thus TSLN was accurate in
detecting the presence of thoracic LN in 28/30 cases (93% accuracy).
LSLN staging
found NO disease in 12 pts and N3 in 5 pts. After esophagectomy, final
pathology of the 12 NO patients was NO in 11 and N3 in one patient. LSLN missed
one pt with N3 disease. Final pathologic staging of the 5 N3 patients revealed
N3 status in 2 patients and NO status in 3 patients. Thus, LSLN was accurate in
detecting lymph node metastases in 16/17 patients (94%).
TSLN/LSLN
staging are more accurate than existing staging methods. 6/18 LSLN pts had
unsuspected celiac axis LN involvement missed by standard noninvasive
techniques. 3% of thoracic LN and 17% of celiac LN were downstaged after
preoperative chemo/radiotherapy. LN near the diagphragmatic hiatus may be
difficult to sample without extensive dissection. The role of TSLN/LSLN in
staging esophageal cancer should be further evaluated in a multi-institutional
trial.
*By invitation
32. CATASTROPHIC COMPLICATIONS ASSOCIATED
WITH CERVICAL ESOPHAGOGASTRIC ANASTOMOSIS
Mark D. Iannettoni, M.D.*,
Richard I. Whyte, M.D.* and Mark B. Orringer, M.D.
Ann Arbor, Michigan
Recent
enthusiasm for the cervical esophagogastric anastomosis (CEGA) has arisen
because of its perceived low morbidity. While catastrophic complications of a
CEGA are unusual, they can and do occur, and prevention is possible if the
potential for them is recognized. In 866 patients undergoing a CEGA after
transhiatal esophagectomy, catastrophic complications associated with
anastomotic disruption have occured in 12 patients (1.4%): vertebral body
osteomyelitis (1); epidural abscess with neurologic impairment (2); pulmonary
microabscesses from internal jugular vein abscess (1); tracheogastric fistula
(1); and major dehiscence requiring anastomotic takedown (7). These
complications became manifest from 8 to 85 days after the esophageal resection
and reconstruction (mean 23 days). Leakage from a suspension stitch placed in
the anterior spinal ligament over the vertebral bodies resulted in a posterior
gastric leak and either osteomyelitis or an epidural abscess in three patients,
none of whom had evidence of extravasation on their routine 10 day
postoperative barium swallow. Cervical exploration for a presumed anastomotic
leak lead to the unexpected discovery of an abscess formed by the stomach and
the adjacent wall of the internal jugular vein which was ligated and resected.
One asymptomatic patient who was discharged with a contained anastomotic leak
on his postoperative barium swallow was readmitted 7 days later with a cervical
tracheogastric fistula from which he ultimately died. In 7 patients, 9% of all
who have experienced anastomotic leaks, there was sufficient gastric ischemia
and/or necrosis to require takedown of the anastomosis and intrathoracic
stomach with a plan for later reconstruction. As a result of this experience,
it is recommended that cervical gastric suspension sutures either be omitted
entirely or placed in the fascia over the longus coli muscles anterior to the
spine but not directly into the prevertebral fascia overlying the vertebral
bodies or cervival discs. All cervical anastomotic leaks, even if apparently
contained, are best drained rather then treated expectantly. Patients who
remain febrile and ill after bedside drainage of a leak should undergo cervical
reexploration in the operating room; major gastric ischemia and/or necrosis may
warrant takedown of the anastomosis and intrathoracic stomach. Refinement of
any operation frequently evolves from complications associated with previous
experience. With an awareness of their possibility and with minor technical
modifications, the incidence of these catastrophic complications and their
morbidity can be reduced.
*By invitation
33. DELAYED PRIMARY REPAIR OF ESOPHAGEAL
PERFORATION: IS IT SAFE?
Nan Wang, M.D.*, Ali Safavi,
M.D.*, Anees J. Razzouk, M.D.*, Karen Gan, M.D.*, Arthur C. Hill, M.D.*, Bryan
Fandrich, M.D.*, Michael J. Wood, M.D.*, Edwin E. Vyhmeister, M.D.*, Changwoo
Ahn, M.D.* and Steven R. Gundry, M.D.
Loma Linda and Fontana,
California
The management of
esophageal perforation with delayed diagnosis (>24 hrs) is controversial.
Because of the obvious advantages of primary repair as a simple single-stage
operation, this technique was preferentially used to treat 17 of 21 consecutive
esophageal perforations. Patients were stratified into three groups according
to the time elapsed between perforation and operation:
|
|
n
|
Age
|
Delay
|
Iatrogenic
|
Preop
|
|
|
|
(yrs)
|
(hrs)
|
Perforation
|
Sepsis
|
|
Grp A - <6 hrs
|
5
|
77 ± 12
|
5.2 ± 0.8
|
4 (80%)
|
0 (0%)
|
|
Grp B - <24 hrs
|
6
|
59 ± 12
|
13.8 ± 2.9
|
2 (33%)
|
2 (33%)
|
|
Grp C - >24 hrs
|
6
|
62 ± 12
|
46.0 ± 12.8
|
3 (50%)
|
3 (50%)
|
In addition to primary
repair in these pts, additional coverage of the sutured site was performed
using a fundic wrap in 7 pts (41%), pericardial fat in 2 pts (12%),
diaphragmatic flap in 2 pts (12%), pleural patch in 1 pt (6%), and intercostal
pedicle in 1 pt (6%). Results:
|
|
n
|
Leaks
|
Sepsis
|
MOF
|
Mortality
|
|
Grp A
|
5
|
0 (0%)*
|
2 (40%)
|
0 (0%)
|
1 (20%)
|
|
Grp B
|
6
|
4 (67%)
|
4 (67%)
|
1 (17%)
|
1 (17%)
|
|
Grp C
|
6
|
4 (67%)
|
5 (83%)
|
2 (33%)
|
1 (17%)
|
|
*p<0.03
|
In Group C, the only death
occurred in a profoundly septic pt whose diagnosis was delayed by more than 2
weeks. Esophageal continuity was maintained in all of the 14 (82%) survivors.
We conclude that in the era of advanced intensive-care capabilities, delayed
primary repair of esophageal perforation in the properly selected pt can lead
to a satisfactory outcome. Although an esophageal leak is common unless
immediate repair of the perforation is performed, the leak does not necessarily
lead to an adverse outcome.
4:35 pm EXECUTIVE SESSION (Limited to
Members)
Republic Ballroom, Sheraton Boston Hotel
6:30 pm MEMBER RECEPTION
Museum of Fine Arts
*By invitation
1:45 pm SIMULTANEOUS SCIENTIFIC SESSION
C CONGENITAL HEART DISEASE
Independence Ballroom, Sheraton Boston Hotel
Moderators: Richard A. Hopkins, M.D.
Richard A. Jonas, M.D.
34. APPLICATION OF COMPUTATIONAL FLUID
DYNAMICS TO THE STUDY OF COMPETITIVE FLOWS IN
CAVOPULMONARY CONNECTIONS
Marc Roger de Leval, M.D., FRCS,
Gabriele Dubini, Ph.D.*, Francesco Migliavacca, Ph.D.*, Homayoun Jalali, M.D.*,
Riccardo Pietrabissa, Ph.D.* and Andrew Redington, M.D.*
London, England
Computational
fluid dynamics (CFD) based on a finite element method were applied to the study
of (i) competition of inferior (IVC) and superior vena caval flows (SVC) in the
total cavopulmonary connection (TCPC) and (ii) competition between SVC flow and
forward flow from a stenosed pulmonary artery (PA) in the bidirectional
cavopulmonary anastomosis (BCPA). 3-D parametric models of the connections were
created from angiocardiograms and the fluid dynamics (pressure and velocity)
were computed by a fluid dynamic computational code.
1. TCPC:
12 models corresponding to various degrees of offsetting and shape of the
IVC-PA anastomosis were simulated to evaluate energy dissipation and flow
distribution between the two lungs. A minimal energy loss was found with an
offset value close to zero and the enlargement of the IVC anastomosis towards
the right PA was shown to divert more IVC flow to the right lung, thus
distributing more caval flow to the bigger lung (QLPA/QRPA = 0.75).
2. BCPA:
The wisdom of leaving forward flow from a stenosed PA while constructing a BCPA
remains controversial. 98 such operations were performed between January 1988
and September 1994. Nine patients (9%) had postoperative SVC hypertension (18
mmHg or more), of whom five were reoperated. A computational model of the
operation was developed in an attempt to predict postoperative haemodynamics.
In tight PA stenosis (75% or more) the non-pulsatile forward QP is primarily
directed to the LPA with little influence on the SVC pressure. A pulsatile
forward flow through the PA corresponding to 15, 30, 50 and 75% of the systemic
cardiac output increases the mean PA and SVC pressure respectively by 1.0, 1.7,
2.4 and 3.6 mmHg.
Inferences:
1. CFD has been applied
clinically to refine the design of the TCPC. A small triangular patch is now
placed on the right lateral aspect of the IVC pathway to reduce energy loss and
achieve a more physiological distribution of flow between the two lungs.
2. In BCPA it is safe to
maintain forward flow so as to have an overall systemic to pulmonary flow ratio
of 1, thus improving the systemic arterial saturation without excessive SVC
hypertension.
1973-74 Graham Fellow
*By invitation
35. ATRIAL FLUTTER INDUCED BY LATERAL
TUNNEL CONSTRUCTION IN THE MODIFIED FONT AN OPERATION
Mark D. Rodefeld, M.D.*, Hurt I.
Bromberg, M.D.*, Charles B. Huddleston, M.D.*, Richard B. Schuessler, Ph.D.*,
John P. Boineau, M.D.* and James L. Cox, M.D.
St. Louis, Missouri
The cavopulmonary
connection, or modified Fontan operation, utilizes a lateral tunnel through the
right atrium. Intra-atrial reentry or atrial flutter (AFL) is a relatively
common post-operative problem. We postulate that the lateral tunnel suture line
establishes the anatomic substrate for AFL in these patients. The purposes of
this study were 1) to determine if the lateral tunnel suture line alone is
sufficient to permit initiation of AFL and 2) to map any resultant arrhythmias.
After induction
of general anesthesia, 30-35 kg canines (n=6) underwent median sternotomy,
cradling of the pericardium, and placement of a pacing electrode on the right
atrial appendage. Normothermic cardiopulmonary bypass was initiated. Through a
right atriotomy, a sham modified Fontan operation was performed. A suture line
(no baffle) was placed beginning at the fossa ovalis, around the IVC, up the
crista terminalis, around the SVC, and back to the fossa ovalis, simulating
lateral tunnel construction. After closure of the atriotomy, bi-atrial 256
point form-fitting endocardia! electrodes were transannularly positioned in the
atria via bilateral ventriculotomies. Atrial burst pacing and programmed
extrastimulation were performed before and after placement of the lateral
tunnel suture line.
AFL was
non-inducible in 6/6 dogs prior to suture line placement. After suture line
placement, sustained AFL was repeatedly inducible in all cases (100%), although
in one case isoproterenol was required. The mean flutter cycle length was
196±77 msec. Activation sequence mapping revealed that the suture line was
integral to the AFL pathway in all cases (100%). In each case, the AFL circuit
revolved in the right atrium where slow conduction and unidirectional block
occurred at the suture line. In no instance did the flutter pathway revolve
around the atriotomy site alone.
We conclude
that the baffle suture line used in the modified Fontan operation, without any
alteration in circulatory physiology, is sufficient to create the substrate for
AFL. A more complete understanding of the reentrant circuit may permit use of
prophylactic measures such as incisional modification or placement of
cryolesions that create bi-directional block, thereby inhibiting AFL in
patients undergoing the modified Fontan operation.
*By invitation
36. STAGED OPERATION TO FONT AN INCREASES
THE INCIDENCE OF SINO-ATRIAL NODE DYSFUNCTION
Peter B. Manning, M.D.*, John E.
Mayer, Jr., M.D., Gil Wernovsky, M.D.*, Steven B. Fishberger, M.D.* and Edward
P. Walsh, M.D.*
Kansas City, Missouri and Boston,
Massachusetts
Morbidity and
mortality of total cavo-pulmonary connection (modified Fontan procedure) is
decreased in many high-risk single ventricle patients by performing the
operations in a staged fashion. Some have advocated routine use of the bidirectional
Glenn or "hemi-Fontan" procedure in all patients with single ventricle
physiology. Each operative intervention, however, exposes the sino-atrial (SA)
node region to risk of injury. A multi-stage approach may increase the risk of
SA node dysfunction in patients that may not tolerate such a dysrhythmia or
atrio-ventricular dysynchrony.
All patients
completing a Fontan procedure between January 1988 and December 1992 were
reviewed. Of 324 total patients, 227 completed the Fontan in a single operation,
and 97 followed a two-stage approach. The mean age at Fontan completion was not
different between groups (5.3 vs 4.6 yrs, p=.21). Overall survival was the same
for both groups (91%).
The incidence
of transient and fixed SA node dysfunction was similar for both groups
following the first operative intervention despite a heterogeneous patient
population (1 stage: 14%/7%, 2 stage: 12%/3%, p=28). Second operative
intervention resulted in a higher incidence of dysrhythmia (transient: 29%,
fixed: 23%), and more frequent SA node dysfunction on follow-up (1 stage: 11%,
2 stage: 46%, p<.002). In the two-stage group, 49% of patients without
arrhythmia following first intervention experienced an arrhythmia after the
second intervention, while of those with an arrhythmia following first
operation, 67% experienced one at second intervention (p<.01). Despite these
findings, most patients with SA node dysfunction tolerated it well clinically,
with few patients requiring permanent pacing for atrial arrhythmias (1 stage:
3%, 2 stage: 7%, p=18).
In conclusion,
a multi-staged operative pathway to Fontan reconstruction is associated with a
higher early risk of SA node dysfunction. Longer follow-up is needed to assess
the full impact of this finding. The possibility of this complication must be
weighed carefully against factors increasing the risk of a single-staged
approach.
2:45 pm INTERMISSION - VISIT EXHIBITS
*By invitation
3:15 pm SIMULTANEOUS SCIENTIFIC SESSION
C CONGENITAL HEART DISEASE
Independence Ballroom, Sheraton Boston Hotel
Moderators: Richard A. Hopkins, M.D.
Richard A. Jonas, M.D.
37. CLINICAL EXPERIENCE WITH REPAIR OF
CONGENITAL HEART DEFECTS USING
ADJUNCTIVE ENDOVASCULAR DEVICES
John G. Coles, M.D., Jeanne M.
Lukanich, M.D.*, Jean Perron, M.D.*, Greg J. Wilson, M.D.*, Marlene
Rabinovitch, M.D.*, David G. Nykanen, M.D.*, Lee N. Benson, M.D.*, Ivan M.
Rebeyka, M.D.*, George A. Trusler, M.D., Robert M. Freedom, M.D.* and William
G. Williams, M.D.
Toronto, Ontario, Canada
The use of intravascular
devices as an adjunct in the repair of congenital heart anomalies represents a
novel but unproven therapeutic approach. Intra-operative occlusion of an apical
muscular ventricular septal defect (VSD) using a clamshell device from the
right atrial approach was accomplished in 4 patients. One patient died
following associated aortic arch reconstruction as a result of a hypoplastic
left ventricle (LV). The results in the remaining 3 patients were favourable
based on absence of late residual shunting, LV dysfunction or arrhythmia.
Intraoperative implantation of pulmonary arterial stent (5-15 mm expanded
diameter) was performed in 15 patients (unilateral 8; bilateral 7) during
repair of neonatal pulmonary atresia VSD (n=4), obstructed pulmonary arterial
confluence following truncus repair (n=3), bidirectional cavopulmonary shunt
(n=2), Fontan procedure (n=2), and miscellaneous pulmonary arterial stenoses
(n=5). The endovascular stents were effective at achieving immediate patency in
all patients. There were 2 early deaths. Early reoperation was required in each
of the 3 survivors of neonatal stent implantation due to bilateral, obstructive
neointimal hyperplasia at 3, 10 and 11 months postoperatively. Intraoperative
stent implantation was used at the site of obstructed pulmonary venous drainage
in 4 patients. Lethal recurrent intraluminal obstruction occurred in all 4
patients, evident histologically as smooth muscle cell proliferation and
extracellular matrix deposition. This initial experience supports continued
application of intraoperative deployment of endovascular devices for closure of
muscular VSD's otherwise inaccessible from the right atrial approach, and for
cases of pulmonary arterial obstruction within larger calibre pulmonary
arteries. Recurrent obstruction due to a pro-liferative healing response
appears to be an eventual certainty in currently designed small diameter
endovascular stents.
*By invitation
38. MANAGEMENT OF TETRALOGY OF FALLOT WITH
PULMONARY ATRESIA AND DIMINUTIVE PULMONARY ARTERIES
Francis D. Pagani, M.D, Ph.D.*,
John P. Cheatham, M.D.*, Robert H. Beekman, III, M.D.*, Thomas R. Lloyd, M.D.*,
Ralph S. Mosca, M.D.* and Edward L. Bove, M.D.
Ann Arbor, Michigan and Omaha,
Nebraska
Since September
1991, 14 consecutive patients with tetralogy of Fallot, pulmonary atresia, and
diminutive pulmonary arteries (PA) have undergone staged repair. All patients
had multiple aortopulmonary collateral arteries and the ductus arteriosus was
absent in 12. Mean size of the right and left PA's was 2.3±1.0 mm and 2.0±0.8 mm,
respectively (range, 1-3 mm). Six patients (42%) have gone on to complete
repair (CR). Age at initial procedure (BT shunt [1], direct aorta-PA
anastomosis or conduit [3], RV-PA conduit [2]) was 6.0±7.7 months. The number
of operative procedures to achieve CR was 3.3±1.6 per patient (range, 2-6).
Post bypass RV/LV was 0.5±0.1. Most patients required one or more
interventional catheterizations for balloon dilatation of peripheral stenoses
and/or stents in addition to collateral embolization. Mean follow-up in this
group was 7.4±8.4 months (range, 0.5 to 20.3 months) and was 100% complete.
There were 2 late deaths secondary to neurological complications in one and
following reoperation for RV to PA conduit stenosis in the other. Eight
patients are awaiting further surgery. Age at initial procedure was 22.2±57.2
months and follow-up was 8.5±8.0 months. Initial palliative procedures included
unifocalization (1), direct ascending aorta-PA anastomosis (3), RV to PA
conduit (3), and transannular RV outflow patch (1). One patient has had 2
additional procedures (RV to PA conduit and unifocalization). Operative
mortality was 25% (2 of 8 patients). Six of 8 patients are alive and awaiting
further intervention or repair.
Of the total of
4 deaths in this series, 3 occurred in patients undergoing shunts or
unifocalization as the initial procedure (3/5, 60%). Among these 5 patients,
only 1 is alive with a CR. There was only 1 death among 9 patients (11%)
receiving RV to PA conduits (5) or direct ascending aorto-PA anastomosis (4) as
the intial operative procedure. Three of these 9 patients have gone on to
complete repair with no deaths.
This experience
suggests that CR is feasible even in patients with extremely diminutive PA's
(<3 mm) and may be achieved by early (3-6 month) establishment of central PA
flow by RV to PA conduit (PA's > 1.5 mm) or direct ascending aorta-PA
anastomosis (PA's < 1.5 mm). Subsequent interventional catheterization and
additional operative procedures as required for further PA stenoses and collateral
embolization allow continued recruitment of central PA's and may obviate or
minimize the need for unifocalization procedures.
*By invitation
39. PH-STAT COOLING IMPROVES CEREBRAL
METABOLIC RECOVERY AFTER CIRCULATORY ARREST IN INFANTS WITH AORTOPULMONARY
COLLATERALS
Paul M. Kirshbom, M.D.*, Lynne A.
Skaryak, M.D.*, Louis R. DiBernardo, M.D.*, Frank H. Kern, M.D.*, William J.
Greeley, M.D.*, J. William Gaynor, M.D.* and Ross M. Ungerleider, M.D.
Durham, North Carolina
The presence of
aortopulmonary collateral arteries (APCA's) in infants with congenital heart
disease has been associated with an increased incidence of neurologic injury
following deep hypothermic circulatory arrest (DHCA). This may be due to a
"steal" phenomenon resulting in inadequate cooling of the brain. This study was
designed to examine the effects of different blood gas management strategies
during cooling on cerebral blood flow (CBF) and metabolic recovery following
DHCA in the presence of APCA's. PTFE shunts (4mm) were placed between the left
subclavian and pulmonary arteries in twenty 1-month old piglets. Animals were
randomized as follows:
|
|
Alpha-State (α -S)
|
pH-Stat (pH-S)
|
|
Control (shunts ligated)
|
C- α (n=5)
|
C-pH (n=5)
|
|
Shunt (shunts open)
|
S- α (n=5)
|
S-pH (n=5)
|
All animals were placed on
CPB, cooled to a nasopharyngeal temperature of 18°C with either α-S or
pH-S blood gas strategy, arrested for 90 minutes, and then rewarmed to 37°C.
Global CBF (radioactive microspheres) and cerebral metabolic rate of oxygen
consumption (CMRC>2) were measured: I) Warm on CPB, II) Cold on CPB before
DHCA, and III) Warm on CPB post-DHCA.
|
|
CBF (ml/100gm/min±S.E.)
|
CMROj (ml O2/100gm/min±S.E.)
|
|
|
C-α
|
C-pH
|
S-α
|
S-pH
|
C-α
|
C-pH
|
S-α
|
S-pH
|
|
I
|
68 ± 6
|
65 ± 3
|
46 ± 4
|
47 ± 2
|
3.2 ± 0.2
|
3.2 ± 0.5
|
2.5 ± 0.2
|
2.8 ± 0.4
|
|
II
|
22 ± 4
|
45 ± 12
|
12 ± 2
|
16 ± 1
|
0.5 ± 0.1
|
0.6 ± 0.2
|
0.4 ± 0.1
|
0.4 ± 0.01
|
|
III
|
50 ± 5
|
41 ± 6
|
23 ± 4
|
37 ± 4*
|
2.3 ± 0.2
|
2.0 ± 0.3
|
1.2 ± 0.2
|
2.1 ± 0.2*
|
p<0.05 vs C-α, * p<0.05 vs S-α
Conclusions: In
the absence of APCA's (C-α and C-pH), a-stat and pH-stat cooling
strategies provide equivalent cerebral protection. In the presence of
APCA's (S-α and S-pH), CBF and CMRO2 recovery following DHCA
are significantly decreased when compared to controls if a-stat cooling is
employed. This effect is eliminated through the use of pH-stat cooling. In
infants with APCA's which are not amenable to pre-operative embolization or
intra-operative ligation, pH-stat cooling may reduce the likelihood of
neurologic injury.
*By invitation
40. BYPASS EFFECTS OF DELAYED REWARMING ON
CEREBRAL BLOOD FLOW IN INFANTS FOLLOWING TOTAL CIRCULATORY ARREST
Erie H. Austin, III, M.D.*,
Rosendo A. Rodriguez, M.D.* and Steve M. Audenaert, M.D.*
Louisville, Kentucky
Sponsored by: Laman A. Gray,
Jr., M.D., Louisville, Kentucky
Purpose: A
possible mechanism of brain injury after TCA may be cerebral hypoperfusion,
which results in an unbalanced flow/metabolism ratio. A single report
(Astudillo et al. 1993) that a delay in rewarming modified the flow pattern of
recovery suggested that cold reperfusion could improve this unbalanced ratio.
Our purpose was to detect any possible beneficial effect.
Methods:Transcranial Doppler son
