 |
|
| Wednesday Morning, May 6, 1998 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Back to Annual Meeting Program
WEDNESDAY MORNING, MAY 6, 1998
7:00 a.m. GENERAL THORACIC SURGERY FORUM SESSION Ballroom C, Hynes Convention Center
Moderators: Larry R. Kaiser, M.D. David J. Sugarbaker, M.D.
F9. HUMAN PLEURAL MESOTHELIOMAS CONTAIN SIMIAN VIRUS-40 (SV40) REGULATORY AND LARGE TUMOR (T) ANTIGEN DNA SEQUENCES. Harvey I. Pass, M.D., Jessica S. Donington, M.D.*, Peter Wu, M.D.*, Paola Rizzo, Ph.D.*, Ronald Kennedy, Ph.D.* and Michele Carbone, M.D, Ph.D.* Detroit, Michigan; Washington, DC; Bethesda, MD; Chicago, Illinois and Oklahoma City, Oklahoma
A cohort (20%) of mesothelioma (meso) patients will not have an exposure to asbestos. Recently, a DNA tumor virus (SV40) has been shown to cause hamster mesos, and we previously described SV40-like DNA amino terminus sequences in 29 of 48 mesos (Oncogene 9:1781-90, 1994). We analyzed an additional 42 mesos to determine (1) whether our initial observations were durable (2) if other regions of the SV40 genome were present and (3) whether meso patients exhibited an immune response to SV40. METHODS: Genomic DNA was extracted from snap frozen meso tumor samples and from the SV40-induced hamster meso tumor H9A. PCR primers were used to amplify various SV40 large T antigen (T-ag) regions including a 105 bp amino terminus fragment, a 281 bp carboxy terminus fragment, and a 310 bp fragment of the enhancer promoter region. Endonuclease digestions were used to verify the expected product. SV40 T-ag specificity and liter of human serum antibodies were examined in 31 mesothelioma patients using an enzyme-linked immunosorbent assay (ELISA). RESULTS: 30 of the 42 (71%) samples amplified T-ag amino sequences, and specificity was verified by Southern hybridization. Thirteen of 42 samples (31%) amplified the appropriate size fragment for the carboxy terminus, and digestion with BsaB I matched that of H9A. Twenty of 42 samples (48%) amplified SV40 regulatory sequences and Fok I digestion matched that of the hamster control tumor. Sequence analysis (4 patients) revealed 100% homology with the regulatory region of SV40 strain 776. Compared to non-SV40 exposed controls, the frequency of serum antibodies to T-ag in meso was significantly greater (P2 = 0.045). CONCLUSIONS: These data suggest an association between the SV40 virus and human mesothelioma which could be exploited for diagnostic/therapeutic options including early detection and potential vaccination strategies.
*By invitation F10. EVIDENCE THAT DIFFUSION CAPACITY LIMITS LUNG VOLUME REDUCTION SURGERY IN A RABBIT MODEL OF EMPHYSEMA. John C. Chen, M.D.*, Dan L. Serna, M.D.*, Ledford L. Powell, M.D.*, Henry E. Aryan, B.S.*, Robert J. McKenna, M.D., Arthur Gelb, M.D.* and Matthew Brenner, M.D.* Orange and Irvine, California
Purpose: Lung volume reduction surgery (LVRS) has been suggested to improve lung compliance and expiratory flows in patients with obstructive emphysema. Endpoints for optimal resection volumes are not known. Spirometric improvements may have to be weighed against reduction in gas exchange surfaces as larger volumes of lung tissue are removed. The purpose of this study was to evaluate the effects of LVRS on lung diffusion capacity (DLCO), pulmonary compliance, and airway flow in a rabbit model of emphysema.
Methods: Forty-nine rabbits were induced with 15,000 units of elastase by aerosolization through an endotracheal tube. Emphysema was confirmed histologically 5 weeks following induction. Single breath DLCO, static pressure volume relationships, and forced expiratory flows (FEF 33%) were measured at baseline prior to induction, preoperatively at 4 weeks following induction, and 1 week postoperatively following sham sternotomy (n = 10) or bilateral upper lobe LVRS (n = 39).
Results: Transpleural pressures with 60 cc insufflation above FRC increased with resection of larger volumes of lung tissue (Graph 1, ANOVA p = 0.000). Comparison of changes in DLCO reveal diminishing diffusing capacity with increasing amounts of lung volume resected (Graph 2, p = 0.063). In contrast, early expiratory flow improved most in the rabbits with 2.3-3 grams of lung tissue resected (Graph 3, p = NS).
Conclusions: Static lung
compliance continues to improve with escalating volumes of lung resected.
Decreased compliance and increased airway flow following LVRS in this animal
model parallel findings in clinical studies. These findings support the
hypothesis that mechanisms of improved elastic recoil and airway resistance
contribute to patient improvement. Our findings suggest that there exists an
optimal resection volume range for improvement in
*By invitation F11. ADJUVANT TREATMENT OF REFRACTORY LUNG TRANSPLANT REJECTION WITH EXTRACORPOREAL PHOTOPHORESIS. C.T. Salerno, M.D.*, Soon J. Park, M.D.*, David M. Kulick, M.D.*, Nathan S. Kreykes, B.S.*, Marshall I. Hertz, M.D.* and R. Morton Bolman, III, M.D. Minneapolis, Minnesota
Photophoresis (PPE) is a technique in which a patient's leukocytes, removed by apheresis, are exposed to ultraviolet-A light after pretreatment with 8-methoxypsoralen, a photosensitive drug activated by long-wavelength ultraviolet light. The irradiated mononuclear suspension is subsequently re-infused to the patient. PPE has been used for the treatment of autoimmune diseases, T cell lymphoma and the treatment of acute rejection in heart transplant recipients. PPE appears to induce specific suppression of both cellular and humoral rejection. The altered lymphocytes produce a suppressor response that targets unirradiated T cells of similar clones in a mechanism that is not yet elucidated. To date there have been few published reports detailing the use of PPE in lung transplant recipients. We have used PPE, in addition to standard anti-rejection chemotherapy in 7 lung transplant recipients with refractory rejection since 1992. All 7 patients had progressively worse graft function and were BOS grade 3 prior to the initiation of photophoresis. There were 3 women and 4 men. Two patients had a pre-transplant diagnosis of COPD, one a-1 antitrypsin deficiency, one cystic fibrosis, one bronchiectasis and two primary pulmonary hypertension. Prior to developing refractory rejection all patients had been treated with three drug immunosuppression, and anti T-cell therapy. The mean time from transplant to the initiation on PPE was 16 months (range 7-32). The mean number of treatments was 6 (range 3-12). RESULTS: Five of six patients subjectively improved after PPE therapy. In these 5 patients PPE was associated with a stabilization in the measured FEV1. In two patients there was biopsy proven reversal of rejection following PPE. With a mean follow up of 28 months (range 2-53), all 7 patients are alive and well. Three patients required re-transplant at a mean period of 19 months (range 15-21) after completion of the PPE treatments. Four patients have remained stable post PPE with only subtle changes in their immunosuppressive therapy. There were no PPE related complications. We believe that this treatment is a safe option for patients with refractory lung allograft rejection when increased immunosuppression is contraindicated or ineffective.
*By invitation F12. BILE ACIDS INDUCE CYCLOOXYGENASE-2 AND PROSTAGLANDIN SYNTHESIS: A POTENTIAL MECHANISMOF ESOPHAGEAL CARCINOGENESIS. Fan Zhang, M.D., Ph.D.*, Andrew J. Dannenberg, M.D.* and Nasser K. Altorki, M.D. New York, New York
Bile reflux has been implicated in the genesis of esophageal cancer in animals and humans but the underlying mechanism(s) remains unclear. A large body of data from a variety of experimental systems suggest that cyclooxygenase-2 (Cox-2), the inducible form of Cox, is important in tumor formation because it catalyzes the synthesis of prostaglandins (PCs) from arachidonic acid, inhibits apoptosis and bioactivates chemical carcinogens. We investigated, therefore, the effects of bile acids, known endogenous promoters of colon cancer, on the expression of Cox-2 and synthesis of PCs in a human Barrett's adenocarcinoma cell line. Treatment with the unconjugated dihydroxy bile acids, chenodeoxycholate (CD) and deoxycholate (DC), resulted in about a 10-fold increase in the production of PGE2 Enhanced synthesis of PGE2 was associated with a marked increase in levels of Cox-2 mRNA (Fig. 1) and Cox-2 protein (Fig. 2) with maximal effects at 8-12 h and 12-24 h, respectively. In contrast, neither cholic acid nor conjugated bile acids affected levels of Cox-2 or synthesis of PGE2.
Bile acid-mediated induction of Cox-2 was blocked by inhibitors of protein kinase C activity, including calphostin C and staurosporine. Bile acids were also potent inducers of Cox-2 in esophageal squamous cell carcinoma cell lines. These results may explain, in part, the mechanism by which bile reflux contributes to the pathogenesis of esophageal cancer.
*By invitation F13. LIPOSOME-MEDIATED GENE TRANSFER IN RAT LUNG TRANSPLANTATION: A COMPARISON BETWEEN THE IN VIVO AND EX VIVO APPROACH. Carlos H. Boasquevisque, M.D.*, Bassem Mora, M.D.*, Mariano Boglione, M.D.*, Jonathan S. Bromberg, M.D.*, Ronald K. Scheule, Ph.D.*, Joel D. Cooper, M.D. and G. Alexander Patterson, M.D. St. Louis, Missouri; Ann Arbor, Michigan; Framingham, Massachusetts
Background: We have recently studied the pattern of in vivo and ex vivo liposome-mediated gene transfer to rat lung isografts using the reporter gene chlorarnphenicol acetyl transferase (CAT). We also demonstrated that ex vivo transfection of TGF(3-1 cDNA into rat lungs improved allograft function. In this study we compared the efficacy of in vivo and ex vivo transfection in rat lung transplantation. Methods: Animals were divided into two groups according to the transfected cDNA. (1) CAT: Fischer rats underwent isogeneic left lung transplant and were divided into 6 groups (n = 4). In group I grafts were infused ex vivo with 660 µg of liposome-CAT cDNA and transplanted immediately (1-1/2 hours of exposure at 10°C). In group II animals underwent the same procedure but the exposure time was 10 hrs. In group III (in vivo), donors received an intrajugular injection of 1320 µg of CAT cDNA. Left lungs were harvested after 1-1/2 hours at 10°C and transplanted. In group IV grafts were transfected in vivo, harvested after 10 hours and preserved for 10 hours at 10°C. In groups V and VI, grafts were transfected in vivo and ex vivo respectively with CAT cDNA not complexed to liposomes. The exposure time was 1 1/2 hours in both groups. In all groups recipients were sacrificed at two days post transplant for CAT assay. (2) TGFβ-1 Brown-Norway rats served as donors and Fischer rats as recipients. Animals were divided into 4 groups. In group I (in vivo, n = 7) donors received an intrajugular injection of 1320 ng of TGFβ-1 sense cDNA complexed to liposomes. Left lungs were harvested 3 hours later. In group II (n = 7) grafts were harvested and then infused ex vivo with 660 µg of TGFβ-1 sense cDNA complexed to liposomes. In group 111 (n = 5) grafts were harvested and then infused ex vivo with 660 µg of TGFβ-1 anti-sense complexed liposomes. In group IV (n = 5) lung grafts were harvested and transfected ex vivo with TGFβ-1 sense cDNA not complexed to liposomes. All grafts were preserved for 3 hours at 10°C prior to transplantation. Recipients were sacrificed on postoperative day 5 for arterial oxygenation and histologic assessment of rejection score. Results: (1) CAT -Transfection was clearly superior in the ex vivo subgroup. Transfection was similar for both exposure times. CAT cDNA alone was inefficient either in vivo or ex vivo. (2) TGFβ-1 oxygenation was superior in the ex vivo liposome-TGFβ-1 sense group in comparison to the in vivo liposome TGFβ-1 sense group and ex vivo liposome TGFβ-1 anti-sense group. Allograft function was superior in both groups treated ex vivo and in vivo with TGFβ-1 sense-liposomes compared to TGFβ-1 sense cDNA alone. Rejection scores were not significantly different between the ex vivo liposome-sense versus in vivo liposome-sense subgroups. However rejection scores were superior after liposome-sense transfection compared to liposome-anti-sense and cDNA alone groups, Conclusions: (1) using current liposome technology the ex vivo approach is superior to the in vivo approach in experimental lung transplantation. (2) Infusion of cDNA not complexed to liposomes was sufficient to provide transgene expression but not at levels high enough to produce a functional effect.
*By invitation F14. INDUCTION CHEMOTHERAPY, SURGICAL RESECTION AND RADIOTHERAPY IN PATIENTS WITH MALIGNANT PLEURAL EFFUSION, MEDIASTINO-SCOPY NEGATIVE(STAGE IIIB) NON-SMALL CELL LUNG CANCER. Scott J. Swanson, M.D.*, Michael T. Jaklitsch, M.D.*, Steven J. Mentzer, M.D., Malcolm M. DeCamp, M.D.*, William G. Richards, Ph.D.*, Raphael Bueno, M.D.* and David J. Sugarbaker, M.D. Boston, Massachusetts
Objective:The purpose of this study was to determine the feasibility, morbidity and mortality of multimodality therapy including extrapleural pneumonectomy in patients with malignant pleural effusion and negative mediastinal lymph nodes (by mediastinoscopy) with a diagnosis of non-small cell lung cancer (NSCLC) (stage IIIB).
Rationale:Stage IIIB NSCLC on the basis of malignant pleural involvement has a guarded prognosis with average survival of 3-6 months in most series regardless of therapy. A subset of these patients at presentation do not manifest hematogenous or mediastinal lymphatic metastases but only local regional pleural spread. Clinically, these patients can be confused with patients with malignant pleural mesothelioma for whom multimodality therapy may extend survival. Induction chemotherapy followed by surgery has been shown to have a survival benefit in several recent series of patients with stage IIIA disease. Surgical resection in early stage malignant mesothelioma has been shown to be feasible. Radiotherapy has been demonstrated to decrease local recurrence following surgery in NSCLC. A combination of the strategies used to treat malignant mesothelioma and stage IIIA NSCLC might be applicable to a select group of patients with stage IIIB (malignant pleural involvement) NSCLC.
Methods:Between 1994 and 1997, 12 consecutive patients with stage IIIB NSCLC on the basis of malignant pleural involvement were selected to undergo a multimodality treatment regimen. Patients who demonstrated adequate functional reserve (ECOG ps 0-2), who had no significant medical comorbidites and had adequate pulmonary reserve to tolerate a pneumonectomy (predicted postoperative FEV 1 ≥ 1) were screened for signs of metastatic disease. Each patient had a bone scan, head ct scan and chest ct scan that included the liver and adrenal glands. Patients without signs of hematogenous metastases underwent mediastinoscopy to rule out mediastinal nodal involvement. In those without N2 or N3 disease, induction chemotherapy was carried out with 3 cycles of a platinum-based regimen. Twelve patients without clinical or radiographic signs of progression went on to have an extrapleural pneumonectomy including prosthetic reconstruction of the hemidiaphragm and pericardium. Seven patients had post-operative external beam radiotherapy. Perioperative data and followup information was gathered from our prospective thoracic database
Results: This cohort was made up of 6 males and 6 females with a median age of 61.5 years (range 41-69), median follow-up 8.5 months (range 3-25). Histological evaluation of the specimens revealed 9 adenocarcinomas and 3 squamous cell carcinomas. Median length of stay was 7.5 days (range 6-80). Seven patients (58%) received red blood cell transfusions (median 1.5 units, range 0-6). There were no perioperative deaths and 7 (58%) patients had complications (6- atrial fibrillation, 1- aspiration pneumonia, 1- vocal cord paresis). Median survival in the N1 node positive subgroup is 24 months (n = 5). The median survival has not been reached in the node-negative subgroup (n = 7), (p = NS).
Conclusion: Multimodality therapy including surgical resection for stage IIIB NSCLC is feasible in selected patients. This pilot study would suggest that aggressive local regional strategies may be the subject of larger clinical trials in patients with stage IIIB NSCLC.
*By invitation F15. PIG LUNG PRE-PERFUSION PREVENTS LUNG HYPERACUTE REJECTION IN THE PIG-TO-HUMAN COMBINATION. Paolo Macchjarini, M.D., Ph.D.*, Rafael Oriol, M.D.*, Robert Rieben, Ph.D.*, Nicolao Bovin, Ph.D.* and Philipe Dartevelle, M.D. Le Plessis and Villejuif, France; Berne, Switzerland; Moscow, Russia
Background: The clinical use of pig lungs is currently limited by hyperacute rejection (HAR), triggered by the binding of human anti-αGal xenoreactive natural antibodies (hXNA) to endothelial xenograft cells. Objective: To test the relative effects of pig organ pre-perfusion and specific depletion of anti-αGal hXNA in preventing HAR in the pig-to-human lung combination. Methods: Large White pig (20-25 kg) left lungs were harvested and ex-vivo continuously ventilated and reperfused, using a neonatal oxygenating system, with either whole human blood pre-perfused trough donor pigs right lung (group I), liver (group II), spleen (group III) or human plasma filtrated on an immunoabsorbent column with α-galactose (Galαl-3Galβ(CH2)3NH2; Bdi) (group IV). Each study group included 6 animals. Results: Pre-perfusion of human blood through pig organs or the αGal column achieved a similar 89 ± 4% reduction of anti-αGal hXNA in the four groups. All pre-perfusing organs displayed HAR after 15 ± 3 min. Following ex-vivo reperfusion, group I lung xenografts had a significantly (p<0.001) longer functional survival than groups II, III and IV xenografts, and were the only histologically normal 5 hours upon reperfusion.
Human blood reperfusing group I xenografts had a significantly (p<0.05) lower: i) fall in white blood cells, clotting factors, total circulating immuno-globulins; ii) total (CH100) and membrane attack complex (C5b-9) complement activation; and in) hemolysis. Conclusions: We provide evidence that specific depletion of anti-αGal hXNA alone incompletely protects pig lungs from HAR. It is speculated that the complete protection afforded by lung pre-perfusion relates to a better removal of other critical humoral or cellular elements provoking HAR.
*By invitation F16. ADENOVIRAL-MEDIATED P53 GENE TRANSFERS TO NON-SMALL CELL LUNG CANCERS. David Weill, M.D.*, Allan N. Shulkin, M.D.*, Juliette L. Wait, M.D.*, Milissa Tuzzolino, R.N.*, John A. Osborne, M.D, Ph.D.*, John J. Nemunaitis, M.D.* and Michael J. Mack, M.D. Dallas, Texas
Introduction The p53 gene, a tumor suppressor gene, is often rendered nonfunctional in human non-small cell lung cancer (NSCLC) by mutation or deletion. By restoring the expression of p53, it is postulated that tumor growth could be suppressed. We performed a Phase I study using an adenoviral vector expressing wildtype p53 (Ad-p53) to demonstrate the feasibility of safely delivering the gene and to document any observed antitumor activity.
Materials and Methods A replication-defective adenoviral vector containing wildtype p53 gene was constructed for intralesional injection into NSCLC that had documented p53 mutations. All patients were refractory to conventional treatment including cisplatin chemotherapy (n = 8) or external beam radiation (n = 3). Patients were enrolled into one of two treatment arms: treatment with Ad-p53 alone or treatment with Ad-p53 plus cisplatin. The Ad-p53 injections were performed monthly in patients with bronchoscopically accessible tumors. The study treatments continued as long as there was no tumor progression or unacceptable adverse effects. A maximum of 6 treatments was given. Results The Ad-p53 was injected into 11 patients with primary NSCLC. Doses of Ad-p53 ranged from 1 x 106 to 1 x 1011 plaque forming units. Following the injections, p53 expression was documented in 8 of the 9 evaluable patients. Toxicity after injection of Ad-p53 was minimal and included minor complications associated with the endobronchial biopsy procedure. During the course of treatment, 2 patients developed pneumonia which was likely secondary to the obstructing tumors.
In 6 of the 9 evaluable patients, endobronchial obstruction was relieved as determined by serial bronchoscopic imaging. Disease remained stable in 2 patients and progressive in 1. The average survival of patients enrolled in the study was 139.3 + days (range 16-292+).
Conclusion We demonstrated that Ad-p53 can be safely delivered intralesionally to patients with NSCLC. Toxicity was minimal, and transgenic expression was achieved. Furthermore, apoptosis as measured by relief of endobronchial obstruction was observed in most patients.
*By invitation F17. A SELECTIVELY REPLICATING ADENOVIRUS (ONYX-015) LYSES NON-SMALL CELL LUNG CANCER CELLS THAT LACK FUNCTIONAL p53. David M. Jablons, M.D.*, Liang You, M.D, Ph.D.*, Adam Sampson-Johannes, M.S.*, David Kirn, M.D.* and Frank McCormick, Ph.D.* San Francisco and Richmond, California Sponsored by: Frank Hanley, M.D., San Francisco, California
Loss or mutation of p53 tumor suppressor gene is a common genetic abnormality in many human tumors including lung cancer. p53 mutations and loss of heterozygosity have been detected in more than 50% of lung cancers. p53 protein prevents replication of damaged DNA in normal cells and promotes apoptosis of cells with abnormal DNA. p53 mutations are frequently associated with poor prognosis in tumor patients. ONYX Pharmaceuticals has genetically designed a tumor-targeting adenovirus (a replication competent E1B-deleted) which only replicates in cells that lack functional p53 gene and therefore kills tumor cells specifically. In an in vitro study, this mutant adenovirus ONYX-015 has been shown to kill cervical carcinoma cells, colon carcinoma cells, glioblastoma ceils and pancreatic adenocarcinoma ceils lacking functional p53. It was also demonstrated that this virus caused a significant reduction in tumor size and caused complete regression of 60% of the tumors induced by p53-defi'cient human cervical carcinoma cells in nude mice. In this study, we tested the cytotoxicity of this mutant adenovirus against two non-small cell lung cancer (NSCLC) cell lines that lack functional p53 gene (NCI-H522 and NCI-HI703) using cytopathic effect (CPE) assays. NCI-H522 is a lung adenocarcinoma cell line with a missense mutation at codon 285 (GAG-»AAG) and NCI-HI703 is a squamous lung carcinoma cell with a single base deletion at codon 191 (CCT->CT) of p53 gene. Loss of heterozygosity of the p53 gene was found in both NCI-H522 and NCI-HI703 cells. We have demonstrated that the virus can lyse those cells but not NSCLC cells with functional p53 (NCI-H2304) or a mesothelioma cell line (MS-1). In replicate experiments, ONYX-015 virus lysed, in a dose-dependent fashion, NCI-H522 and NCI-HI703 cells. Five days after infection, 50% of the cells were lysed at multiplicities of infection (MOI) of 1 plaque-forming unit (PFU) per cell; 9 days after refection, at the MOI of 1 and 0.1, 100% and 50% of the cells were lysed respectively. No signs of cell lysis were noticed in NCI-H2304 and MS-1 at MOI of 0.1 and 1. PFU per cell 9 days after infection. Wild-Type adenovirus served as control for effective infection. We are currently in the process of testing this virus in combination with chemotherapy in these NSCLC cell lines as well as in fresh human lung tumor cells. It is hoped that with subsequent pre-clinical studies ONYX-015 can be advanced to phase I clinical trials for the treatment of patients with advanced lung cancer.
*By invitation F18. MDM-2 EXPRESSION: AN ALTERNATIVE MECHANISM FOR p53 INACTIVATION IN ESOPHAGEAL ADENOCARCINOMA. Robert Soslow, M.D.*, Liang Ying, M.D.* and Nasser K. Altorki, M.D. New York, New York
Several immunohistochemical studies have shown that the p53 protein is expressed in 60-70% of esophageal adenocarcinomas. Since mutations of this tumor suppressor gene are present in approximately 40% of tumors, it is presumed that p53 stabilization and expression may develop as a result of mechanisms other than gene mutation. Amplification and expression of the mdm-2 gene, a known regulator of p53 activity, can result in inactivation of the p53 gene with stabilization of its protein product and loss of its tumor suppressor function.
In this study we evaluated the incidence of p53 abnormalities as well as the expression of the mdm-2 protein product in 16 esophageal adenocarcinomas. Routine immunohistochemical studies were performed following standard antigen retrieval methods and interpreted using a previously described immunoreactive score. Tumor DNA was obtained by microdissection, amplified and sequenced for mutations in the p53 hot spot regions (exons 5-8). P53 expression was observed in 12 of 16 (75%) cases while p53 mutations were detected in 7 of 16 (43%). This does not exclude the possibility of mutations in exons other than 5-8. Overall, mdm-2 expression was present in 8 of 16 tumors. Moderate or strong expression of mdm-2 was observed in 4 tumors none of which had detectable p53 mutations.
We conclude that in esophageal adenocarcinoma (1.) Discordance of p53 immunohistochemistry and mutations occurs in 20-30% of cases and (2.) mdm-2 expression may be responsible for loss of p53 tumor suppressor function in 25% of esophageal adenocarcinomas.
*By invitation 7:00 a.m. CARDIAC SURGERY FORUM SESSION Ballroom B, Hynes Convention Center
Moderators: Fred A. Crawford, Jr., M.D. Edward D. Verrier, M.D.
F19. AN EXPERIMENTAL MODEL OF SMALL INTESTINE SUBMUCOSA AS A GROWING VASCULAR GRAFT FOR CONGENITAL CARDIAC SURGERY. MonicaC. Robotin-Johnson, M.D., F.R.A.C.S.*, Paul E. Swanson, M.D.*, David C. Johnson, M.D., F.R.A.C.S.* and James L. Cox, M.D. St. Louis, Missouri and Washington, DC
The ideal vascular graft for use in children with congenital heart defects should not only be biocompatible, nonthrombogenic and present no infectious risk, but should grow at the same rate as the recipient.
We have tested autologous small intestine submucosa (SIS) as a superior vena cava (SVC) interposition graft in 11 piglets, with a mean weight of 10.7 kg. The grafts were prepared from segments of autologous jejunum, rendered nonthrombogenic by bonding with heparin. The SVC from the level of the azygous vein to the SVC-right atrial junction, measuring a mean length of 9.9 mm and circumference of 25.9 mm was replaced. There was one early and one late death, not related to the SVC replacement. At 90 days the 9 long term survivors had a mean weight increase of 630% and the grafts increased in length by 147% [p<0.03] and in circumference by 184% [p<0.001], paralleling the growth of the native cava. All 11 grafts were patent and free of thrombus at the time of explantation. Cavograms showed no anastomotic stricture or aneurysm formation in 7 of 9 cases. The luminal surface of all grafts was smooth, shiny and indistinguishable from that of the native cava. The light microscopy showed a loosely textured cellular collagen framework, with a dense capillary network and complete luminal coverage by cells resembling endothelial cells. Electron microscopy confirmed that a complete endothelial cell layer was present.
In conclusion, SIS provides a collagen framework that becomes remodelled, keeps up with the growth of the recipient and acquires a nonthrombogenic endothelial surface. This makes it potentially very well suited as material for pulmonary artery reconstruction or Fontan operations in small children.
1996-97 AATS Graham Fellow *By invitation F20. MEMANTIN IS A POTENT EXCITATORY AMINO ACID BLOCKER FOR PREVENTING SPINAL CORD INJURY IN A RABBIT MODEL. Marek P. Ehrlich, M.D.*, Erich Knolle, M.D.*, Ruxandra Ciovica, M.D.*, Peter Boeck, M.D.*, Martin Grubenwoger, M.D.*, Ricarda Konetschny, M.D.*, Fabiola Cartews-Zumelzu, M.D.*, Edvin R. Turkof, M.D.*, Ernst Wolner, M.D. and Michael Havel, M.D.* Vienna, Austria
Introduction: This study was conducted to investigate the effect of memantin, a noncompetitive N-methyl-D-aspartate receptor antagonist, on the neurological outcome of spinal cord ischemia and reperfusion after aortic occlusion.
Materials and methods: New Zealand white rabbits (3 - 4.5 kg) were anesthetized and spinal cord ischemia was then induced for 40 minutes by infrarenal aortic occlusion. Animals were randomly assigned into three groups. Group A animals (n = 8) received intraaortic memantin infusion (20 mg/kg) after aortic clamping. Group B animals (n = 8) were pretreated with memantin infusion (20 mg/kg) 45 minutes prior to aortic occlusion. Group C (n = 8) was the control group, in which no pharmacological intervention was applied. Neurological status was assessed at 12, 24, 36 and 48 hours after operation and scored using the Tarlov system (4 normal, 0 paraplegia). Lumbar spinal root stimulation potentials (SRS) as well as electrical transcranial motor evoked potentials (MEP) from lower limb muscles were monitored pre-, intra- and postoperatively. After sacrifice at 48 hours, the spinal cord was fixed and studied histopathologically.
Results: All measured potentials disappeared shortly after aortic cross-clamping. MEP did not correlate with clinical findings. Quantitative analysis of SRS showed, that potentials in both memantin treated groups regained activity earlier compared to the placebo group. Histologic examination of spinal cords from group A and B rabbits revealed only a few abnormal motorneurons, whereas spinal cords from the control group had evidence of extensive cord injury with prominent lysis of Nissl substance, destruction of nuclear chromatin and vacuolization of anterior horn motorneurons. Median values for neurological observations from each group are reported below:
Conclusion: Memantin significantly reduced neurological injury secondary to spinal cord ischemia and reperfusion after aortic occlusion at 40 minutes in the rabbit model. F21. MODULATION OF MYOCARDIAL PERFUSION AND VASCULAR REACTIVITY BY PERICARDIAL bFGF: IMPLICATIONS IN THE TREATMENT OF INOPERABLE CORONARY ARTERY DISEASE. Roger J. Laham, M.D.*, Motohisa Tofukuji, M.D., Ph.D.*, Michael Simons, M.D.* and Frank W. Sellke, M.D. Boston, Massachusetts
Endothelium-dependent relaxation and perfusion are reduced in the collateral-dependant myocardium. To examine the modulating effects of pericardial fluid basic-fibroblast growth factor (bFGF) on endothelium-dependent responses and expression of inducible (iNOS) and endothelial (eNOS) nitric oxide synthases in the collateral-dependant myocardium, ameroid occluders were placed around the left circumflex (LCx) artery of pigs. After 3 weeks, LCx occlusion was confirmed with angiography. Saline containing 30 µg (n = 6) or 2 mg (n = 6) bFGF or no bFGF (control, n = 6) was injected into the pericardial space. Four weeks later myocardial blood flow was determined with colored microspheres, endothelium-dependent coronary microvascular responses to ADP (10 µM) and endothelium-independent responses to nipride (100 µM) were examined, and protein and gene expressions of iNOS and eNOS were determined (Western analysis, PCR) in the ischemic LCx and non-ischemic LAD regions. Vascular response = % relaxation of precontract diameter.
Both iNOS protein and mRNA were increased 3.3 ± 1 fold in the LCx compared to the LAD territory, whereas eNOS expression was similar in both regions. This suggests that the decreased endothelium-dependent relaxation in the collateral-dependent circulation may be due to increased iNOS expression and NO-induced inhibition of eNOS. bFGF improves endothelium- dependent relaxation and blood flow in the collateral-dependent myocardium. These findings may have implication regarding the cause of altered blood flow regulation in chronically ischemic myocardium and in the treatment of patients with inoperable coronary disease with the injection of bFGF into the pericardial space.
*By invitation F22. ANGIOGENESIS ACCOMPANIES IMPROVED PERFUSION IN REGIONS OF HIBERNATING MYOCARDIUM FOLLOWING TRANSMYOCARDIAL LASER REVASCULARIZATION. G. Chad Hughes, M.D.*, Alan P. Kypson, M.D.*, James D. St. Louis, M.D.*, Brian H. Annex, M.D.*, Timothy R. DeGrado, Ph.D.*, R. Edward Coleman, M.D.*, James E. Lowe, M.D. and Kevin P. Landolfo, M.D.* Durham, North Carolina
Background. The mechanism for clinical improvement following transmyocardial laser revascularization (TMR) is unknown, although preliminary work in normal myocardium suggests that angiogenesis may play a role. The purpose of this study was to describe the quantity and nature of the neovascularization accompanying improved myocardial perfusion following TMR in a model of hibernating myocardium.
Methods. Five adult mini-swine (n = 5) underwent left circumflex coronary artery (LCx) occlusion to reduce resting blood flow by 90% as documented by ultrasonic flow probe measurement. Two weeks post-occlusion, the animals underwent positron emission tomography (PET) to document the presence of ischemic, viable myocardium in the LCx distribution. TMR was then performed using a holmium:YAG (n = 3) or CO2 (n = 2) laser with 20 channels placed at 1 cm intervals in the ischemic region. Six months post-TMR, repeat PET was performed. Animals were then sacrificed, and their hearts harvested for histology as well as immunohistochemical analysis to identify the presence of endothelial cells.
Results. Mean myocardial blood flow by PET in the ischemic LCx distribution increased from 0.37 ± 0.16 to 0.60 ± 0.13 ml/g/min (p<0.05) at 6 months post-TMR. There was no change in control septal flow from pre- to post-TMR (0.73 ± 0.38 to 0.64 ± 0.15 ml/g/min). Histologic examination of the lased LCx area revealed many neovessels located predominantly at the periphery of the TMR channels. These vessels were present in a highly disorganized pattern consistent with angiogenesis. The presence of endothelial ceils within the neovessels was confirmed with the endothelial cell specific antibodies anti-von Willebrand factor and anti-human tie-2 (TEK). Quantitative analysis revealed the lased ischemic regions contained a mean of 26.0 ± 8.8 microvessels per 200x field, compared with a mean of 5.1 ± 2.0 in control non-lased regions (p = 0.0003).
Conclusions. Angiogenesis occurs in the channel regions following TMR in hibernating myocardium and accompanies improved regional perfusion. These findings strongly suggest that angiogenesis is the mechanism of increased blood flow following TMR in ischemic myocardium.
*By invitation F23. TIME-COURSE OF FUNCTIONAL RECOVERY AFTER CORONARY BYPASS SURGERY IN PATIENTS WITH CHRONIC LEFT VENTRICULAR ISCHEMIC DYSFUNCTION. Jean-Louis J. Vanoverschelde, M.D., Ph.D.*, Christophe Depre, M.D., Ph.D.*, Bernhard L. Gerber, M.D.*, Marcel Borgers, Ph.D.*, William Wijns, M.D.*, Jacques A. Melin, M.D., Ph.D.* and Robert A. Dion, M.D.* Brussels, Belgium Sponsored by: Bruce W. Lytle, M.D., Cleveland, Ohio
Background Chronic left ventricular (LV) ischemic dysfunction, a condition often referred to as myocardial hibernation, is associated in humans with ultrastructural alterations of the myocytes, including the loss of myofilarnants and the accumulation of glycogen. Given the severity of these structural changes, contractile function is unlikely to resume immediately upon revascularization.
Methods and results We studied 32 patients with coronary disease and chronic LV ischemic dysfunction undergoing bypass surgery. Dynamic Positron Emission Tomography with 13N-ammonia and 18F-deoxyglucose to assess myocardial perfusion and glucose metabolism was performed in 29 patients. Coronary bypass surgery was subsequently performed with the use of the left internal mammary artery to graft the left anterior descending coronary artery. All other co-diseased vessels were also revascularized. On average, each patient received 2.9 anastomoses, of which 1.3 were constructed with arterial conduits. In every patient, a peroperative transmural biopsy was harvested from the center of the dysfunctional area, to quantify the increase in extracellular matrix and the presence of structurally altered cardiomyocytes. LV function was serially measured by digitized 2D-echocardiography before and again 10 days, 2 months and 6 months after revascularization. The time-course of recovery of regional function was estimated from the monoexponential decrease in dysfunctional wall motion score. At the 6 months followup, 19 patients had improved LV function while 13 patients showed persistent dysfunction. Before revascularization, reversibly dysfunctional segments had higher myocardial blood flow (83 ± 29 versus 60 ±21 ml·(min·100g)-1, p<0.01), higher glucose uptake (40 ±14 versus 21 ±9 µmol·(min·100g)-1, p<0.05) and less increase in extracellular matrix (25 ± 17 versus 46 ± 17%, p = 0.01) than segments with persistent dysfunction. The extent of functional recovery correlated with myocardial blood flow (r = 0.84) and the increase in extracellular matrix (r = -0.60). In patients with reversible dysfunction, the return of segmental function was progressive and followed a monoexponential time-course with a time constant of 23 days (range 6-78 days). The rate of recovery correlated best with the proportion of altered cardiomyocytes in the biopsy (r = 0.83).
Conclusions. The present study indicates that the recovery of regional and global LV function after successful revascularization is progressive and follows a monoexponential time-course which is influenced by the extent and severity of the structural changes affecting the cardiomyocytes.
*By invitation F24. MYOCYTE ENDOTHELIN EXPOSURE DURING CARDIOPLEGIC ARREST EXACERBATES CONTRACTILE DYSFUNCTION WITH REPERFUSION. R. Brent New, M.D.*, Jeffrey S. Mandel, M.D.*, Angela C. Sampson, B.A.*, Christopher A. Kerr, B.S.*, Rupak Mukherjee, Ph.D.*, Fred A. Crawford, Jr., M.D. and Francis G. Spinale, M.D., Ph.D.* Charleston, South Carolina
Background: While transient left ventricular (LV) dysfunction can occur following cardioplegic arrest (CA), the contributory mechanisms for this phenomenon remain incompletely understood. Institution of cardiopulmonary bypass and CA results in neurohormonal system activation which includes increased release of the vasoactive peptide, endothelin (ET). Past studies have suggested that ET can influence LV myocyte contractility. Therefore, this project tested the hypothesis that exposure of LV myocytes to ET during simulated CA, would have direct effects on contractile processes with subsequent reperfusion.
Methods: LV porcine myocytes were randomly assigned to 3 groups: (1) Control: Normothermic (37°C) cell media (n=156); (2) Cardioplegia: simulated CA (K+ 24 mEq/L, 4°C x 2hrs) followed by reperfusion and rewarming with cell media (n = 73); (3) Cardioplegia and ET: simulated CA in the presence of ET (200 pM) followed by reperfusion with cell media and ET (n = 54). Myocyte contractility was measured following reperfusion by videomicroscopy.
Results: Myocyte shortening velocity was reduced following simulated CA compared to controls (62 ±3 vs 80 ± 3 µm/s, respectively p<0.05) and was further reduced with CA and ET exposure (49 ± 3 µm/s, p<0.05). Myocyte velocity of relengthening, which reflects sarcomere cross-bridge release rates and calcium resequestration, was reduced after CA compared to controls (51 ± 3 vs 77 ± 3 µm/s, respectively, p<0.05) and was reduced to a greater degree with CA and ET exposure (41 ± 3 µm/s, p<0.05).
Conclusions: The unique findings of the present study demonstrated that exposure of the LV myocyte to ET during simulated CA, directly contributed to contractile dysfunction following reperfusion. A contributory molecular mechanism for the effect of ET with CA may be alterations in myocyte active relaxation processes. These findings suggest that increased ET levels which occur in the setting of cardiac surgery directly influence myocyte contractility, which in turn contributes to the transient LV dysfunction following cardioplegic arrest.
*By invitation F25. ROLE OF ENDOTHELIN-1 AND ENDOTHELIN-1 RECEPTOR BLOCKADE IN PLACENTAL DYSFUNCTION AFTER FETAL CARDIAC BYPASS. Doff B. McElhinney, M.D.*, V. Mohan Reddy, M.D.*, Hiranya A. Rajasinghe, M.D.*, John R. Liddicoat, M.D.*, Karen Hendricks-Munoz, M.D.*, Jeffrey R. Fineman, M.D.* and Frank L. Hanley, M.D. San Francisco, California
Fetal cardiac bypass (FCB) causes placental dysfunction, characterized by increased placental vascular resistance (PVR), decreased placental blood flow (PBF), hypoxia, and acidosis. A variety of factors have been found to contribute to the development of this placental dysfu nction, but its exact mechanisms remain unclear. Vasoactive factors produced by the vascular endothelium, such as nitric oxide and endothelin-1 (ET-1), are important regulators of placental vascular tone. To study the role of the vascular endothelium in placental dysfunction related to FCB, we studied 3 groups of fetal sheep. In the first (n:7), we determined placental hemodynamic responses pre- and post-FCB to an endothelium-dependent vasodilator (acetylcholine), an endothelium-independent vasodilator (sodium nitroprusside), and ET-1, a vasoactive polypeptide produced by the endothelium. Controls (n = 7) received the same vasoactive substances but were not exposed to FCB. In the second group (n = 5), an ET-1 receptor blocker (PDQ123) was administered and placental hemodynamics were measured before and after FCB. Results were compared with control fetuses that did not receive PDQ123 in order to assess the effect of ET-1 receptor blockade. In the third group (n = 5), no medications were given and plasma ET-1 levels were measured before and after FCB, then compared with controls that did not undergo FCB. Before FCB, exogenous ET-1 decreased PBF by 8.7% and increased PVR by 9.3%. After bypass, however, ET-1 decreased PBF by 47% and increased PVR by t06%. In addition, there was a significant attenuation of the placental vascular relaxation response to acetylcholine after FCB, with a decrease in PVR of 14%, compared with 20% pre-FCB. The response to sodium nitroprusside was not significantly altered by FCB. In fetuses that received the ET-1 blocker, PVR increased from 0.32 ± 0.03 U pre-FCB to 0.41 ± 0.07 after bypass, which was significantly less than in control animals (0.31 ± 0.04 pre-FCB, 0.51 ± 0.14 post-FCB). Similarly, PBF decreased significantly more in control animals (from 172 ± 29 ml/min/kg to 116 ± 36) than in fetuses receiving ET-1 receptor blocker (from 168 ± 31 to 140 ± 35). Plasma ET-1 increased significantly in fetuses exposed to FCB, but did not change in control animals. This study demonstrates that FCB causes a significant increase in plasma ET-1 and impairs the placental vascular response to endothelium-dependent modulators of vascular tone. ET-1 receptor blockade substantially reduces post-FCB placental dysfunction. This and other pharmacologic or physiologic interventions aimed preserving endothelial function may be effective means of optimizing fetal outcome after FCB.
1995-97 TSFRE Research Fellow *By invitation F26. PHOSPHODIESTERASE INHIBITORS PREVENT THE SYSTEMIC INFLAMMATORY RESPONSE SYNDROME. Koh Takeuchi, M.D.*, Pedro J. del Nido, M.D, Dimitrios N. Poutias, B.S.* and Francis X. McGowan, Jr., M.D.* Boston, Massachusetts and Bethesda, Maryland
The systemic inflammatory response syndrome (SIRS) is an important cause of multiple organ dysfunction after cardiopulmonary bypass; its magnitude and consequences are particularly great in neonates and infants. We and others have previously found that certain phosphodiesterase inhibitors (PDI) interfere with inflammatory signaling pathways at several points. The present study tested the hypothesis that two clinically used PDIs, amrinone (AMR) and vesnarinone (VES), would decrease the SIRS.
A well-characterized, severe SIRS model of rabbit endotoxemia was used. Rabbits received S. lyphi endotoxin (lipopolysaccharide, LPS; 0.2 mg/kg iv bolus); LPS + AMR (1.0 mg/kg bolus + 10 mg/kg/min infusion); LPS + VES (3 mg/kg bolus only-t½ =~8 hours); or vehicle alone. Systemic effects were assessed by mortality, fever, and acidosis. Indices of the inflammatory response included plasma cytokine and myeloperoxidase (MPO) concentrations. Pulmonary involvement was assessed by changes in pulmonary vascular resistance (PVR) and lung MPO. Myocardial function was quantified in excised, Langendorff-perfused hearts; myocardial calcium cycling and contractile protein calcium sensitivity were measured using fluorescence spectroscopy. Plasma hepatocellular enzyme concentrations (SGPT, SGOT) served as markers of liver injury. Myocardial protein kinase C (PKC) and inducible nitric oxide synthase (iNOS) activity were used as indices of cytokine signaling. Measurements were made 0, 1, 2, and 6 hours after LPS. Results are summarized in the Table as mean ± sem, N = 6-8 each. *P<0.05 vs control; #P<0.05 vs LPS alone.
Thus, VES, which has unique ion channel and gene expression effects unrelated to its PDI properties, was particularly effective in SIRS suppression. VES also protected against LPS-induced increases in cytokines, hepatocellular enzymes, myocardial PKC activity, and PVR; the protective effects of AMR upon these indices were significant, but less than VES. VES prevented LPS-induced reductions in myocardial calcium-activated contractile force, diastolic relaxation, and diastolic calcium removal. Neither VES nor AMR exacerbated LPS-induced hypotension. VES also prevented LPS-induced TNF production in cultured macrophages and cytokine-stimulated iNOS production in cultured neonatal cardiomyocytes. The beneficial effects of VIES and AMR in vivo were not shared by dobutamine, and the in vitro effects of VES were not mimicked by a cell-permeable cyclic AMP analog. These results indicate that certain phosphodiesterase inhibitors have multiple and potent effects upon inflammatory activation and signaling pathways. The mechanism does not appear to be due to elevations in cyclic AMP. These compounds, which are used clinically for their inotropic and vasodilating properties, may be useful to limit inflammatory activation and signaling cascades during pediatric CPB, as well as other states that are associated with inflammatory cytokine production.
*By invitation F27. IN VIVO IMAGING OF APOPTOSIS DURING CARDIAC ALLOGRAFT REJECTION USING RADIOLABELED ANNEXIN V. Patrick W. Vriens, M.D.*, Francis G. Blankenberg, M.D.*, Eric R. Davis, M.D.*, Gerald J. Berry, M.D.*, Bruce A. Reitz, M.D., H. William Strauss, M.D.* and Robert C. Robbins, M.D.* Stanford, California
Introduction We hypothesized that technetium 99m labeled annexin V, a new radioactive tracer of apoptotic cell death, can be utilized to monitor cardiac allograft rejection. Annexin V is a human protein that binds to phosphatidyl serine, a phospholipid that is selectively expressed on the cell surface, during the early stage of apoptosis.
Methods Untreated ACI rats served as recipients of heterotopically placed, allogeneic PVG rat, or syngeneic ACI rat cardiac grafts. Function of grafts was assessed by daily palpation. Annexin V was derivatized with technetium 99m hydrazinonicotinamide (99mTcHYNIC), and injected i.v. one hour prior to imaging at day 3, 4, and 5 after transplantation. Region of interest image analysis was used to quantify uptake of the radiopharmaceutical. Histopathologic analysis and TUNEL staining were performed at each time point after imaging. To investigate whether uptake of 99mTcHYNIC-annexin V decreased after treatment of rejection, recipients of allogeneic grafts were treated daily with Cyclosporin A (CSA, oral, 10 mg/kg), starting at day 4, and imaged at day 4, 7, and 11 after transplantation.
Results Allogeneic PVG
cardiac grafts showed a readily visualizable 50% increase in uptake of
99mTcHYNIC-annexin V at day 3, a 160% increase at day 4 and a 274% increase at
day 5 after transplantation in untreated ACI rats (Fig. 1) (n = 6 for each time
point), compared to syngeneic hearts (Fig. 2)
Conclusion Uptake of radiolabeled annexin V correlates with histopathological grades of acute rejection in cardiac allografts. When rejection is reversed by CSA treatment, uptake decreases to levels comparable to syngeneie grafts. Imaging of apoptosis using radiolabeled annexin V may enable detection of acute cardiac allograft rejection, and may provide a new tool to monitor rejection.
*By invitation F28. SKELETAL MUSCLE VENTRICLES FROM LEFT VENTRICULAR APEX TO AORTA, EXPERIENCE UP TO 37WEEKS: STEP TOWARDS CLINICAL APPLICATION. Frank A. Baciewicz, Jr., M.D.*, Gregory A. Thomas, M.D.*, Kevin A. Greer, M.D.*, Robert L. Hammond, Ph.D.*, Huiren Lu, M.D.*, Steven Bastian, M.S.* and Larry W. Stephenson, M.D. Detroit, Michigan
Skeletal muscle ventricles (SMVs) were constructed from the latissimus muscle and lined with autogenous pericardium in 6 dogs. After 3 weeks of vascular delay and 6 weeks of electrical conditioning, SMVs were connected from the left ventricular apex with a valved conduit and then from the SMV to the descending aorta with a second-valved conduit. The SMV was stimulated during diastole at a 1:2 ratio with the heart. SMV contraction increased femoral pressure by 23% at 33 Hz and 28% at 50 Hz, resulting in 32% and 33% of the total cardiac output being pumped by the SMV. Data at implant (n = 6) is shown below:
The dogs survived 6, 17, 30, 72 and 263 days in circulation; one dog is alive at 160 days. Three deaths, including that at 263 days were related to complications of indwelling monitoring devices and not directly to the SMV. In the two longest surviving dogs, SMV function remained stable over time. During propranolol induced heart failure, SMV contraction augmented diastolic pressure 89% at 33Hz and 96% at 50 Hz. At a 1:2 contraction ratio with the heart, SMV assist increased systemic perfusion (Qsmv QAV) 23% at 33 Hz and 64% at 50 Hz. At a 1:1 contraction ratio, systemic perfusion was increased further by 25% at 33 Hz and 114% at 50Hz. This model of skeletal muscle assist is the most hemodynamically effective that we have tested, and now appears capable of functioning long-term.
*By invitation 9:30 a.m. SIMULTANEOUS SCIENTIFIC SESSION A-2 CONGENITAL HEART DISEASE Ballroom A, Hynes Convention Center
Moderators: John E. Mayer, Jr., M.D. John A. Waldhausen, M.D.
40. VENO-VENOUS MODIFIED ULTRAFILTRATION AND CIRCULATING CYTOKINES: A PROSPECTIVE RANDOMIZED STUDY. Ugursay Kiziltepe, M.D.*, Azhar Hossain, M.D.*, Daniel Remick, M.D.*, Samuel Barst, M.D.*, Jeffrey P. Gold, M.D. and Hani A. Hennein, M.D.* Atlanta, Georgia; New York, New Hyde Park and Bronx, New York; Ann Arbor, Michigan Discussant: J. William Gaynor, M.D., Philadelphia, Pennsylvania
Background: Cardiopulmonary bypass (CPB) is associated with the production of both proinflammatory (IL-6, IL-8 and TNFcc) and antiinflammatory (IL-10) cytokines, and a resultant systemic inflammatory response. Arterio-venous modified ultrafiltration has been shown to be associated with a reduction in total body water and improved hemodynamic and hemostatic functions. Veno-venous modified ultrafiltration (VVMUF) is a further modification of the technique with the potential advantage of not reducing effective cardiac output. We tested the hypothesis that VVMUF reduces extracellular body water and removes circulating cytokines following CPB.
Methods: Thirty-seven children undergoing open surgical correction of congenital heart defects were randomly assigned to VVMUF or controls. Inferior and superior vena cava cannulas were used as the inflow and outflow of the VVMUF circuit. VVMUF was performed for 10 min after weaning from CPB. IL-1P, IL-6, IL-8, IL-10 and TNF-a plasma levels measured at seven time points before, during and after CPB.
Results: Both groups were similar in terms of age (6.48 ± 5.38 years in VVMUF vs. 6.27 ± 5.37 years in control, p: not significant (n.s.)). Mean cross-clamp times were 604-35 minutes (rain) in VVMUF and 41 ± 23 rain in control group, p:n.s. Mean bypass duration was 98 ± 56 min in VVMUF and 92 ± 50 min in control group, p:n.s. Minimum temperatures were 28.8 ±2.1 C in VVMUF and 28.4 ±4.4 C in control group, p:n.s. VVMUF resulted in a significant removal of extracellular fluid (980 ± 601 cc) and rise in hematocrit levels (from 28.27 ± 3.64% to 34 ± 3.64%, p<0.05 ). There was no mortality in either group. All 5 cytokines measured rose during and following bypass in both groups. A significant reduction of IL-lβ levels (from 11.3 ± 39.7 pg/ml to 4.4 ± 20.9 pg/ml) followed VVMUF. Changes in cytokines other than IL-lβ could not be demonstrated.
Conclusions: Veno-venous modified ultrafiltration is a safe and effective method of removing extracellular volume following CPB. VVMUF is associated with a significant reduction in IL-lβ levels and may therefore reduce the deleterious effects of CPB by diminishing systemic inflammatory response.
*By invitation 41. TOTAL REPAIR OF PULMONARY ATRESIA WITH VENTRICULAR SEPTAL DEFECT AND MAJOR AORTOPULMONARY COLLATERALS: AN INTEGRATED APPROACH. Adriano Carotti, M.D.*, Roberto M. Di Donate, M.D.*, Cosimo Squitieri, M.D.*, Paolo Guccione, M.D.* and Glauco Catena, M.D.* Rome, Italy Sponsored by: Aldo R. Castaneda, M.D., Guatemala City, Guatemala Discussant: Frank L. Hanley, M.D., San Francisco, California
Background: Predicting postrepair ratio of the peak systolic pressure in the right ventricle (pRV) to that in the left ventricle (pLV) may be of absolute prognostic value for patients undergoing total repair of pulmonary atresia, ventricular septal defect and major aortopulmonary collaterals (PA.VSD.MAPCAS). To this purpose, we currently rely on 2 novel parameters: 1) preoperative total neopulmonary index (TNPAI = CAI [total MAPCA index] + PAI [pulmonary artery index]); 2) mean pulmonary artery pressure changes during an intraoperative flow study, according to Reddv M. et al.
Patients and methods: Since January 1994, 15 patients (age mean±SD: 64 ± 54 months) with PA.VSD.MAPCAS were managed according to TNPAI: a preoperative value of ≥ 150 mm2/m2 was indication for total repair. Seven patients with hypoplastic pulmonary arteries and TNPAI < 150 mm2/m2 first underwent palliative conduit right ventricular outflow tract reconstruction (RVOTR), followed by secondary one-stage, midline total unifocalization and VSD closure. The other 8 patients with preoperative TNPAI > 150 mm2/m2 (absent pulmonary arteries m 2 cases) underwent single-stage unifocalization and repair. The VSD was closed in all cases. In 9, the decision to close the VSD was based on an intraoperative pulmonary flow study. In one case the VSD had to be reopened due to hypersystemic pRV.
Results: The 7 patients who initially underwent RVOTR had a significant increase of PAI (from 46 ± 26 to 194 ± 74 mm2/m2 p<0.0001) within 22 ± 6 months from the palliation. Of the total group of 15 patients, repair was successful in 14, with a postrepair pRV/pLV ratio of 0.47 ± 0.1. There was one hospital death due to hypersystemic pRV, despite a reassuring intraoperative pulmonary flow study. Accuracy of this test in predicting the postrepair mean pulmonary artery pressure was 89% (70% CL : 68 - 98%). At a mean follow-up of 12 ± 11 months, all patients arc in NYHA I, on no medications. Their mean pRV assessed by 2DD-ECHO is 40 ± 8 mmHg.
Conclusion: The integrated approach to total repair of PA.VSD.MAPCAS by preoperative calculation of TNPAI, RVOTR (when required) and intraoperative flow study, has led to optimal one-year results with low pRV/pLV ratios.
*By invitation 42. LONG-TERM OUTCOME OF INFANTS WITH SINGLE VENTRICLE AND TOTAL ANOMALOUS PULMONARY VENOUS CONNECTION. J. Williams Gaynor, M.D.*, Margaret H. Collins, M.D.*, Jack Rychik, M.D.* and Thomas L. Spray, M.D. Philadelphia, Pennsylvania Discussant: William G. Williams, M.D., Toronto, Ontario, Canada
Since 1984, 50 infants (28 male/22 female) with functional single ventricle (SV) and total anomalous pulmonary connection (TAPVC) have undergone palliative surgery at our institution. Heterotaxy syndrome was present in 33 patients and Hypoplastic Left Heart Syndrome (HLHS) in 9. TAPVC was supracardiac in 28 patients, cardiac in 7, infracardiac in 10 and mixed in 5. Obstructed TAPVC was present in 15 patients. Initial palliation (median age 5 days, range 1 day-2.5 years) included aortopulmonary shunt (22), the Norwood procedure (17), cavopulmonary connection (5), pulmonary artery band (2), pulmonary valvotomy (1) and the Fontan procedure (1). Repair of TAPVC alone (2) or with other procedures was performed at the initial operation in 29 patients. Overall hospital mortality for initial palliation was 44% (22/50). Mortality was 59% (17/29) in patients undergoing TAPVC repair compared to 24% (5/21) in patients whom TAPVC repair was not performed, p<0.02. Sixteen of the 28 survivors have subsequently died and 1 has been lost to follow-up. TAPVC repair was performed at a subsequent operation in 3 patients. Only 8 patients have successfully undergone the Fontan procedure. Overall actuarial survival was 58 ± 7% at 6 months, 47 ± 7% at 1 year, 36 ± 7% at 2 years and 26 ± 7% at 5 years. Cardiac TAPVC was associated with improved survival compared to other forms of TAPVC (86 ± 13% vs. 42 ± 8% at 1 year and 64 ± 21% vs. 21 ± 6% at 5 years, p = 0.01). Actuarial survival for survivors of initial palliation was 81 ± 8% at 1 year, 6 ± 10% at 2 years, and 47 ± 10% at 5 years. By univariate analysis using Cox's proportional hazard model; repair of TAPVC, younger age at the time of initial surgery and non-cardiac TAPVC were predictors of mortality, p<0.05. Heterotaxy syndrome, HLHS, obstructed TAPVC, and performance of the Norwood procedure were not predictors of mortality, p>0.1. Time of TAPVC repair (initial vs. subsequent operation) did not alter survival, p>0.1. Lung tissue from 12 patients (median age 20 days, range 2 days-8.8 months) was available for histologic examination. TAPVC was supracardiac in 7 of these patients, infracardiac in 3 and mixed in 2. Obstructed TAPVC was present in only 3 patients. In all 12 patients the pulmonary veins were dilated with increased wall thickness for age. Increased muscularization of the arteries was present in 10 patients.
The long-term outcome for infants with SV and TAPVC is poor except for infants with cardiac TAPVC. Mortality at the initial palliative procedure is high and there is a continuing risk of late death. Repair of TAPVC is associated with a worse outcome. Development of the pulmonary vasculature, particularly the pulmonary veins, is abnormal even in infants without clinical evidence of obstructed TAPVC.
*By invitation 43. SURGICAL TREATMENT OF SUBAORTIC STENOSIS: A 17-YEAR EXPERIENCE. Alain Serraf, M.D.*, Joy Zoghby, M.D.*, François Lacour-Gayet, M.D., Remi Houel, M.D.*, Emre Belli, M.D.*, Lorenzo Galletti, M.D.* and Claude Planche, M.D. Le Plessis Robinson, FranceDiscussant: Michel N. Ilbawi, M.D., Chicago, Illinois
Left ventricular outflow tract (LVOT) obstruction due to subaortic stenosis (SAoS) covers a wide range of anatomic features from the localized discrete fibrous stenosis to the more complex multiple left-sided obstructive lesions. Several surgical techniques have been proposed to reestablish a widely patent LVOT, each of which being adapted to the anatomical form.
From 1980 to 1997, 160 pts with SAoS underwent surgical repair. Pts with isolated aortic valve (AoV) stenosis, supravalvar stenosis, those with VSD and SAoS or who developed SAoS after VSD closure were excluded. Preoperatively, 126 pts were in NYHA classes I-II and 34 in classes 1II-IV. Morphological and hemodynamic assessments were performed by angiography in 109 pts and echocardiography alone in 51. Localized discrete fibrous stenosis was present in 28 pts, fibromuscular stenosis in 90, localized hypertrophic obstructive cardiomyopathy in 12, diffuse subaortic stenosis in 23 and in 7 the SAoS was due to accessory mitral valve (MV) tissue or abnormal insertion of mitral papillary muscle. AoV stenosis was associated in 24 and 16 had MV stenosis. In 13, there were multiple left-sided obstructive lesions (Shone's complex). Aortic coarctation was present in 25 pts. The mean preoperative peak systolic gradient across the LVOT was 80 ± 34.8 mmHg. Thirty two pts had previous surgery: 25 coarctation repair, 5 AoV commissurotomy and 2 repair of partial AV canal. The median age at repair was 10 years (Ranges: 0.1-30 years). Subaortic membrane excision was performed in 134 pts, associated to septal myotomy in 101, to septal myectomy in 24. Concomitant AoV commissurotomy was performed in 17 pts. Two had a Konno procedure, 5 had resection of abnormal MV tissue or papillary muscle and 2 had apical conduit insertion. AoV replacement was associated in 6 pts, MV replacement in 2 and mitral valvuloplasty in 4.
There were 5 early deaths (3.1%; 70% CL: 1.5-5.2%): in 2 Shone's complex, 2 diffuse SAoS and 1 with preoperative NYHA class IV. Permanent pace-maker insertion was mandatory in 5 pts. Twenty-five reoperations with 2 post operative deaths were performed in 19 pts in a mean delay of 67.1 ±55.7 months for recurrent LVOT obstruction in 18 and AoV endocarditis in 1. Ten of these pts initially presented with a diffuse SAoS. It was surgically addressed by new septal myectomy in 10, Ross procedure in 2, Konno operation in 2, an apical conduit in 3 and 2 had AoV replacements which was associated in 1 with a Manougian procedure. Statistical analysis revealed that diffuse SAoS with or without AoV stenosis was associated to higher risk for mortality and reoperation (p = 0.002). A median followup of 13.3 years (Ranges: 1.2-17.9 years) was achieved in 90% of survivors. There were 2 late deaths, all other patients were asymptomatic with a mean gradient across the LVOT of 20 ± 13 mmHg. Actuarial survival and freedom from reoperation rates at 15 years were 94.25 ± 1.34% and 85.3 ± 6%, respectively. In conclusion, SAoS covers a wide range of anatomical forms in which diffuse SAoS with or without AoV involvement carries the higher operative mortality and reoperation rates. A more aggressive initial operation should be performed in those patients.
10:50 a.m. INTERMISSION
1993-94 AATS Graham Fellow *By invitation 11:10 a.m. SIMULTANEOUS SCIENTIFIC SESSION A-2 CONGENITAL HEART DISEASE Ballroom A, Hynes Convention Center
Moderators: John E. Mayer, Jr., M.D. John A. Waldhausen, M.D.
44. LONG-TERM CLINICAL AND HEMODYNAMIC EVALUATION OF PORCINE VALVED CONDUITS IMPLANTED FROM THE RIGHT VENTRICLE TO THE PULMONARY ARTERY. Gerard L. Champsaur, M.D., Jacques A. Robin, M.D.*, Francois Tronc, M.D.*, Alain Curtil, M.D.*, Catherine Vedrinne, M.D.*, Francois Sassolas, M.D.* and Jean Ninet, M.D.* Lyon, France Discussant: Ivan M. Rebeyka, M.D., Edmonton, Alberta, Canada
To evaluate the long-term results of valvular prosthetic conduits implanted as a new right ventricular outflow tract (RVOT), a retrospective study was conducted in a consecutive series of 127 patients presenting with complex ventriculo-pulmonary discontinuity and operated on between 1973 and 1996. The mean age was 5 ± 3.48 years (range 2 months-42 years). Initial pathology was tetralogy of Fallot in 42 cases, D-transposition of the great arteries in 25, truncus arteriosus in 21, double outlet right ventricle in 20, L-transposition of the great arteries in 18, and double outlet left ventricle in 1 case. Conduit sizes ranged from 14 to 25 mm, with a 20 mm size used in 38.5% (70% CL: 31-45%) and a 22 mm size used in 22.04% (70% CL: 14-29%) of the cases. Early mortality rate was 19% (24 patients). A post-operative evaluation was performed in 103 operative survivors who were followed up from 1 to 21.6 years, with a mean followup of 8.38 ± 6.2 years. The followup was over 10 years in 33 patients, (32%, 70% CL: 24-40%) and over 15 years in 28 (27%, 70% CL: 19-35%). There were 16 late deaths (RV failure and/or pulmonary hypertension in 7 cases, at various types of redo surgery in 3, sudden death in 1, acute conduit obstruction in 1, neurological in 1, infectious in 1, and miscellaneous in 2). The actuarial survival, including early mortality, was 72.91 ± 0.04% at 5 years, 63.16 ± 0.05% at 10 years, and 58.17 ± 0.05% at 15 years, where 20 patients were still exposed. A total of 74 hemodynamic studies were performed post-operatively, 50 patients having undergone at least one cardiac catheterization during their followup period. The mean peak systolic gradient across the RVOT was plotted as a function of time showing a gradual increase and a significant step-up after the 8th year.
Reoperation was required in 25 cases (24%, 70% CL: 15-33%) for progressive conduit obstruction between 1.1 and 17.7 years after implantation (mean 7.4 ± 4.8 years) in patients with generally very few symptoms or residual VSD in 3 cases. Conduits were replaced by either a RVOT patch (10 cases), a new porcine conduit (7 cases) or a non-valved conduit (5 cases). Freedom from reoperation was 79.52 ± 0.05% at 10 years and 65.81 ± 0.07% at 15 years. More readily available, porcine conduits may represent a valuable alternative to biological substitutes with similar long-term results in large (18 to 20 mm) sizes. Given the few symptoms, progressive conduit stenosis after the 8th postoperative year imposes a yearly noninvasive evaluation during their followup.
*By invitation 45. SURGICAL MANAGEMENT OF PROGRESSIVE PULMONARY VENOUS OBSTRUCTION FOLLOWING REPAIR OF TOTAL ANOMALOUS PULMONARY VENOUS DRAINAGE. François Lacour-Gayet, M.D., Joy Zoghby, M.D.*, Alain Serraf, M.D.*, Emre Belli, M.D.*, Lorenzo Galletti, M.D.*, Jacqueline Bruniaux, M.D.* and Claude Planché, M.D. Le Plessis Robinson, France Discussant: Thomas L. Spray, M.D., Philadelphia, Pennsylvania
Out of 158 patients who underwent correction of total anomalous pulmonary venous drainage (TAPVD) over the past fifteen years, 13 patients developed a progressive pulmonary venous (PV) obstruction. All included patients had satisfactory initial repair, documented by postoperative echo-Doppler, with biphasic flow and maximal velocity less than 1.5 m/s. The anatomical types were: 7 supra cardiac, 4 infra cardiac, 1 intra cardiac and 1 miscellaneous. At initial correction, the common pulmonary vein trunk was anastomosed to the left atrium using nonresorbable suture. A progressive pulmonary venous obstruction occurred in a mean interval of 4 months ranging from 2 weeks to 12 years. Eleven patients required 15 reoperations; 1 patient is awaiting surgery and the last patient was judged inoperable. Two different anatomic lesions were frequently associated: - an anastomotic stenosis with inflammatory retraction of the suture line and - a PV ostial stenosis made of inflammatory intimal hyperplasia developing along the extra cardiac segment of the PV, involving one or several PV on either side. Two patients had isolated anastomotic stenosis, 3 patients had isolated PV ostial stenosis and 8 patients had both. PV ostial stenosis involved: 4 veins in 2 pts, 3 veins in 3 pts, 2 veins in 4 pts, 1 vein in 2 pts; the lesions were bilateral in 7 pts and unilateral in 4. The reoperations consisted in: 10 patch enlargement of the anastomotic stenosis, 5 patch enlargement of PV, 1 intra-operative stenting of PV and 6 intra pericardia! tunnelisation of PV. This last technique consisted in resection of the right pulmonary vein stenosis. The proximal normal PV are incised up to the pericardial reflection. The PV drain passively into the left atrium through a pericardial pouch made of a vascularized pericardial flap. No suture material is used on the PV walls. A similar technique is used on the left PV. There were 5 deaths (38%), all occurring in patients with bilateral PV stenosis. There was no death among the patients with isolated anastomotic stenosis. In the 11 patients who had PV ostial stenosis, 4 of the 6 survivors were successfully treated by the intra pericardial tunnel technique with normal pulmonary artery pressure and no residual gradient at a mean follow up of 15 months.
Conclusion: progressive PV stenosis following TAPVD repair is generally considered as an unpredictable and often lethal complication. It has been successfully treated in 4 patients in this series, using passive drainage of the pulmonary veins into the left atrium through an intra pericardial vascularized tunnel. 1993-94 AATS Graham Fellow *By invitation 46. SUPERIOR OUTCOME FOLLOWING SURGERY FOR PULMONARY ATRESIA AND INTACT VENTRICULAR SEPTUM. Jack Rychik, M.D.*, Hara Levy, M.D.*, J. William Gaynor, M.D.*, William M. DeCampli, M.D.* and Thomas L. Spray, M.D. Philadelphia, Pennsylvania Discussant: Mohan V. Reddy, M.D., San Francisco, California
Pulmonary atresia and intact ventricular septum (PA/IVS) is an anatomically heterogeneous anomaly with a variety of possible surgical management strategies. In a recent multi-institutional study, survival was found to be only 64% at 48 months. Since 1981, 67 patients with PA/IVS have had surgery at our center with 51 survivors (median followup 51 months, range 1 month to 25 years). Surgical strategy was determined, not by formal protocol, but by subjective assessment of right ventricular (RV) size and coronary anatomy in each individual patient. Overall actuarial survival was 80 ± 5% at 1 year, 75 ± 6% at 5 years and 75 ± 6% at 8 years. Initial procedure in infancy consisted of an aorto-pulmonary shunt alone in 33 (group I), shunt with RV decompression in 32 (group H), and primary heart transplant in 2 (group III). In group I, 3 patients (9%) died after shunt placement without further intervention; Fontan operation was performed in 21 (64%) with 2 early deaths (no late deaths), late RV recruitment in 4 (12%) with no deaths, and transplant in 3 (9%) with 1 early death (no late deaths). Two patients have had superior cavopulmonary connection and are awaiting Fontan completion. In group II, 7 infants (22%) died early after RV decompression, 2 of these had dysplastic tricuspid valves with massively dilated RV and 2 had associated aortic/subaortic stenosis; 16 (50%) completed separation into a biventricular circulation with 1 death, 3 (9%) underwent superior cavo-pulmonary connection with maintenance of antegrade RV flow ("½ ventricle repair") with 0 deaths, 2 (6%) had transplant with 1 early death, and 1 (3%) had successful Fontan operation. Three patients (9%) are pending separation into a biventricular circulation. In group III, 1 patient is alive after transplant and 1 died early after organ graft failure. One and 5 year actuarial survivals for group I (91 ± 5%, 80 ± 7%) versus group II (71 ± 8%, 71 ± 8%) were not statistically different. Coronary angiographic data was available in 60 patients; 35 (58%) had no abnormalities, 17 (28%) had RV-to-coronary fistulas alone, and 8 (13%) had fistulas with important coronary stenoses/interruption. Actuarial survival at 1 year was similar for these groups: 83 ± 6%, 93 ± 7%, 75 ± 15%, respectively. Superior long-term outcome is achieved for PA/IVS when the surgical strategy is targeted to the specific anatomical substrate present.
12:10 p.m. ADJOURN
*By invitation 9:30 a.m. SIMULTANEOUS SCIENTIFIC SESSION B-2 ADULT CARDIAC SURGERY Ballroom B, Hynes Convention Center
Moderators: Karl H. Krieger, M.D. Hartzel V. Schaff, M.D.
47. CORONARY BYPASS SURGERY IN PATIENTS WITH PREVIOUS MEDIASTINAL RADIATION THERAPY. Nobuhiro Handa, M.D.*, Christopher G.A. McGregor, M.B., F.R.C.S., Gordon K. Danielson, M.D., Thomas A. Orszulak, M.D., Charles J. Mullany, M.D.*, Richard C. Daly, M.D.*, Joseph A. Dearani, M.D.*, Betty J. Anderson, R.N.* and Francisco J. Puga, M.D. Rochester, Minnesota Discussant: Bruce W. Lytle, M.D., Cleveland, Ohio
Between January 1976 and December 1996, 72 patients with previous mediastinal radiation therapy (MRT) underwent combined cardiac surgery. Forty-seven of these 72 patients who had CABG without valve surgery were reviewed for this study. The mean age was 63.5 ± 12.8 years (range 31 to 82 years). Indications for MRT were breast cancer (n = 26), lymphoma (n = 18) and lung cancer (n = 3). The mean interval between MRT and CABG was 15.1 ± 9.8 yrs (range 1.1 to 37.8 yrs). Preoperative NYHA class III or IV and CAC class III or IV were 93.6 and 89.4, respectively. Fifty-one percent of patients had a history of previous MI. The number of distal anastomosis performed was 2.9 ± 0.9. The left internal mammary artery was used as a LAD graft in 29.8% of patients. Associated cardiac procedures included coronary endarterectomy (2), LV aneurysmectomy (1), and pericardiectomy (2). Operative mortality was 4 cases (8.5%). Complications included prolonged ventilatory support 4 (8.5%), sternal wound infection or mal-union 4 (8.5%), IABP insertion 2 (4.3%), bleeding 2 (4.3%), renal insufficiency 1, gastric perforation 1, and infective endocarditis 1. Total followup was 293.7 patient-years (mean 6.2 ±5.1 yrs). There were 17 late deaths (malignancy 9, heart failure 5, stroke 1, unknown 2). Twelve of these 43 discharged patients developed valvular disease echocardiographically during follow-up (MR+ TR 4, AS+MS 5, AS+AR 2, mild AS 1). Nine patients required late reoperation (CABG 3, AVR+CABG 1, MV repair 1, heart transplant 1, wound repair 2, AICD/pacemaker placement 2). The mean interval between the original CABG and the six open cardiac procedures was 4.4 ± 2.8 years (range 1.1 to 8.6 yrs). Actuarial survival of the whole group (n = 47) was 71.6% (95% confidence interval 58.7%, 86.9%) at five years. These results indicate that increased early mortality, sternal wound complications and reoperation can be expected in this patient population. As progressive valvular disease developed in 27.9% of discharged patients at followup, and 25 of the original 72 patients required concomitant valve and coronary surgery, careful assessment of valvular lesions is important at the time of initial CABG. Careful followup, including echo, is recommended.
*By invitation 48. VALUE OF DOBUTAMINE ECHOCARDIOGRAPHY FOR PREDICTION OF FUNCTIONAL RECOVERY, SYMPTOMATIC IMPROVEMENT AND LONG-TERM SURVIVAL IN PATIENTS WITH CHRONIC LEFT VENTRICULAR ISCHEMIC DYSFUNCTION UNDERGOING CORONARY BYPASS SURGERY. Jean-Louis J. Vanoverschelde, M.D., Ph.D.*, Agnes Pasquet, M.D.*, Jacques A. Melin, M.D., Ph.D.*, Philippe Noirhomme, M.D.*, Gebrine El Khoury, M.D.*, Robert Verhelst, M.D.* and Robert A. Dion, M.D.* Brussels, Belgium Sponsored by: Bruce W. Lytle, M.D., Cleveland, Ohio Discussant: Hillel Laks, M.D., Los Angeles, California
Background Most control studies comparing bypass surgery (CABG) to medical treatment in patients with poor left ventricular (LV) function have demonstrated the superiority of CABG in alleviating symptoms, preventing reinfarction and prolonging survival. Yet, not every patient with poor LV function does benefit from CABG, perhaps because it continues to entail significant immediate risks. In this study, we examined the possible contribution of dobutamine echocardiography (DbE) to the risk stratification of these patients.
Methods. Seventy-six consecutive patients (age, 60 ± 10 years) with coronary disease and chronic LV ischemic dysfunction (ejection fraction: 35 ± 11%) underwent DbE prior to CABG. CABG was performed with the use of the left internal mammary artery to graft the left anterior descending coronary artery. All other co-diseased vessels were also revascularized. On average, each patient received 2.9 anastomoses, of which 1.4 were constructed with arterial conduits. After CABG, patients were followed for 3 years, starting from the date of CABG and ending with a cardiac fatal event or on the most recent date in survivors. In each patient, the recovery of LV function was evaluated by echocardiography 5.3 ± 2.4 months after CABG. Survival and symptomatic status at follow-up were obtained from review of visit and hospital records or by telephone contact.
Results. Functional recovery was evaluated on both a segmental and an individual patient basis. On a segmental basis, pre-operative DbE correctly identified 74% of the segments that improved and 86% of those that remained dysfunctional after CABG. Overall accuracy was 81%. In individual patients, assessment of residual contractile reserve, defined as a decrease in wall motion score by > 3.5 during low-dose DbE, correctly identified 84% of those who improved LV ejection fraction by > 5% after CABG (n = 37) and 73% who failed to do so (n = 33). Overall accuracy was 79%. During followup, 11 patients (15%) died of cardiac causes, 4 in hospital (5%) and 7 later on (6 sudden deaths, 1 progressive pump failure). Kaplan-Meier survival curves indicated that 3-year survival was significantly better in the 40 patients who exhibited residual contractile reserve pre-operatively (95%) than in the 36 patients who did not (75%, P<0.01). Among survivors, symptoms of heart failure (as reflected by the NYHA functional class) improved only in those with residual contractile reserve (from 1.9 ± 0.9 to 1.2 ± 0.4, P<0.01) but not in those without (from 1.9 ± 0.9 to 1.8 ± 0.9, P = ns).
Conclusions. The present study indicates that DbE allows for a comprehensive evaluation of patients with chronic LV ischemic dysfunction and permits accurate prediction of functional recovery and symptomatic improvement, as well as both short- and long-term survival.
*By invitation 49. TWELVE YEAR EXPERIENCE WITH THE RIGHT GASTROEPIPLOIC ARTERY GRAFT. Hisayoshi Suma, M.D., Taiko Horii, M.D.*, Tadashi Isomura, M.D.* and Tetsuya Ichihara, M.D.* Kamakura, Japan Discussant: Robert A. Dion, M.D., Brussels, Belgium
Since 1986, the right gastroepiploic artery (GEA) graft has been used in 800 CABGs. Clinical and angiographic late outcome is reported.
Material and method: There were 682 men and 118 women with a mean age of 61 years. Single, double , triple and left main diseases were noted in 6 , 134, 522 and 138 patients, respectively. Previous MI was noted in 474 (60%) patients and 68 (9%) patients had previous CABG.
Sites of GEA grafting were 68 LAD, 7 diagonal, 131 CX and 616 RCAs with 22 sequential and 26 free grafts. Combined grafts were 768 (96%) ITAs (123 double ITAs ), 45 (6%) inferior epigastric and 65 (8%) radial arteries. Saphenous vein was combined in 389(49%) patients.
Angiographic restudy was performed in 618 patients within one year, 682 patients after I to 5 years and 23 patients from 5 to 10 years.
Consecutive 189 patients (163 men and 26 women, with a mean age of 59 years) who were operated between 1986 to 1992 with single ITA+GEA ± SV graft were followed for more than 5 years (mean 7.7 years) and their late outcomes were evaluated.
Results: Operative death, new Q wave and postoperative IABP were noted in 16 (2%), 12 (1.5%) and 19 (2.4%) patients, respectively.
Angiographic restudy showed the patency rates of GEA graft at early (2 months), mid-term (2 years ) and late ( 7 years ) periods were 94%, 89% and 83% respectively. Late closure of GEA was mainly noted in case of less critically stenosed RCA and focal stenosis was rare in late period.
In the late followup group, 5 and 10 year survival rates were 88.4% and 85.0%, cardiac death-free rate was 94.5% and 93.4%, and cardiac event-free survival was 85.6% and 80.1%, respectively.
Conclusion: GEA graft can be used safely and effectively. Angiographic late result was favorable and the late outcome of single ITA + GEA is comparable with the reported double ITA grafts.
*By invitation 50. INTRAVENOUS T3 IMPROVES MYOCARDIAL FUNCTION AND REDUCES MORBIDITY AFTER CORONARY BYPASS SURGERY: RESULTS OF A DOUBLE-BLIND RANDOMIZED TRIAL. Samantha L. Mullis-Jansson, M.D.*, Michael Argenziano, M.D.*, Steven J. Corwin, M.D.*, Sunichi Homma, M.D.*, Eric A. Rose, M.D. and Craig R. Smith, M.D. New York, New York Discussant: Andrew S. Wechsler, M.D., Philadelphia, Pennsylvania
Background: Thyroid hormone is known to exert profound effects on myocardial function. Although triiodothyronine (T3) deficiency has been described after cardiopulmonary bypass, preliminary studies examining the role of T3 in the management of cardiovascular performance following CABG surgery have yielded conflicting results. In order to further define the clinical role of T3 in cardiac surgery, a double-blind, randomized, placebo-controlled study was undertaken.
Methods: 130 consecutive patients undergoing elective CABG were enrolled. Emergency surgery, age greater than 85 years, and a history of thyroid disease or amiodarone therapy were exclusionary criteria. Upon removal of the aortic crossclamp, patients were randomized to intravenous T3 (0.4 mcg/kg bolus + 0.1 mcg/kg infusion for 6 hours, n = 66) or placebo (n = 64). Outcome measures included perioperative hemodynamics and inotrope/pressor requirements at several time points (mean ± SEM), perioperative morbidity (arrhythmia/infarction), and mortality.
Results: Despite similar baseline characteristics, patients randomized to T3 had a higher cardiac index and lower inotropic requirements postoperatively (table). In addition, patients randomized to T3 demonstrated a significantly lower incidence of postoperative myocardial infarction (4.5% vs. 15.6%, p < 0.05) and pacemaker dependence (15.2% vs. 31.3%, p < 0.05). T3 and placebo patients had equivalent rates of atrial fibrillation. Five patients in the placebo group required postoperative mechanical assistance (3 IABP, 2 LVAD), compared to none in the T3 group (p = 0.02). There were 2 deaths in the placebo group and no deaths in the T3 group.
Conclusions: Parenteral T3 given after crossclamp removal during elective CABG significantly improved postoperative ventricular function, reduced dependence on inotropic agents and mechanical devices, and decreased the incidence of myocardial infarction. While the incidence of atrial fibrillation was not affected, T3 reduced the requirement for postoperative pacemaker support.
10:50 a.m. INTERMISSION
*By invitation 11:10 a.m. SIMULTANEOUS SCIENTIFIC SESSION B-2 ADULT CARDIAC SURGERY Ballroom B, Hynes Convention Center
Moderators: Karl H. Krieger, M.D. Hartzel V. Schaff, M.D.
51. LONG TERM IMPLANTED LEFT VENTRICULAR ASSIST DEVICES FUNCTION AS IMMUNE-INFLAMMATORY ORGANS. Talia B. Spanier, M.D.*, Mehmet C. Oz, M.D., David M. Stern, M.D.*, Silviu Itescu, M.D.* and Ann Marie Schmidt, M.D.* New York, New York Discussant: D. Glenn Pennington, M.D., Winston-Salem, North Carolina
Successful long-term use of implantable left ventricular assist devices (LVAD) is dependent on improved understanding of host-device interactions. The textured surface LVAD was designed to promote the time-dependent cellular population of the device and creation of a biologic lining. We examined the procoagulant and proinflammatory consequences of implantation of this mechanical device and found evidence of sustained thrombin generation (elevated thrombin-antithrombin III complex and prothrombin fragment 1+2) and fibrinolysis (d-dimers and fibrinogen degradation products) as well as clinically-significant immunological alterations with polyclonal B cell activation and autoantibody production. Significant elevations in anti-HLA antibodies were found to directly correlate with difficulty in identification of a negative cross match for cardiac transplantation (anti-MHC class I antibodies), and increased frequency/severity of cellular rejections post transplantation (anti-MHC class II antibodies). In addition, an increased incidence of anti-phospholipid antibodies was recognized in the LVAD population, often associated with clinically-apparent thrombocytopenia. A qualitative T cell defect with an overall increased susceptibility to opportunistic infections (25% incidence of candidal infections over first 100 days awaiting cardiac transplantation vs 2% in UNOS I heart failure controls, p = 0.026) supports the overall immunoregulatory dysfunction associated with LVAD implantation. We believe that circulating blood cells which are selectively adsorbed to the LVAD surface became activated causing the LVAD to become an immune-inflammatory organ with these significant proinflammatory/procoagulant consequences. Analysis of the cells populating the LVAD surface (35 days postop and beyond) support this contention revealing dendritic-type CD34+ cells, as well as cells bearing monocyte (CD14)/macrophage (CD68) markers, and T (CD4/CD25/CD3) and B (CD20) cell markers. Dendritic-type cells also strongly express Vascular Cell Adhesion Molecule-1 (VCAM-1) and Intercellular Adhesion Molecule-1 (ICAM-1), consistent with their capacity to further recruit other circulating cells. RT-PCR confirms cellular activation on the LVAD surface, with transcripts for proinflammatory Interleukin (IL)-la, IL-2 and Tumor Necrosis Factor-a, in addition to VCAM-1 and ICAM-1. RT-PCR also shows that the activated T cells on the LVAD have a TH2 cytokine profile (mRNA was detected for IL-10 and CD40 Ligand but not for gamma interferon, IL-4 or IL-5), consistent with their promotion of an autoimmune phenotype. Furthermore, LVAD recipients demonstrate a reduced proliferative response to candidal antigens compared with heart failure controls ( mean stimulation index 1.4 vs 2.0), and are anergic on intradermal challenge with candidal and control antigens. Together, these observations suggest that selective adsorption and activation of circulating dendritic-type cells and monocytic cells by the LVAD surface contribute to the development of an overall immunoregulatory dysfunction in LVAD recipients. The LVAD therefore emerges as an implanted immune-inflammatory organ providing a sustained pro-inflammatory and pro-thrombotic stimulus with significant impact on the host beyond its mechanical function as a pump.
1994-96 AATS Robert E. Gross Research Scholar *By invitation 52. THE COX-MAZE III PROCEDURE: PARALLEL NORMALIZATION OF SINUS NODE DYSFUNCTION, IMPROVEMENT OF ATRIAL FUNCTION AND RECOVERY OF THE CARDIAC NERVOUS SYSTEM. Miralem Pasic, M.D., Ph.D.*, Onnen Grauhan, M.D.*, Michele Musci, M.D.*, Takeo Teodoriya, M.D.*, Barbara Edelmann, M.S.*, Roland Hetzer, M.D., Ph.D. and Hendryk Siniawski, M.D.* Berlin, Germany Discussant: James L. Cox, M.D., Washington, DC
OBJECTIVE: The Cox-maze III procedure, the technique of choice for the management of atrial fibrillation, includes isolation of the pulmonary veins and multiple incisions in both atria in what corresponds to partial autotransplantation and partial denervation of both parasympathetic and sympathetic systems of the heart. In contrast, transplanted heart is completely denervated and, therefore, without efferent parasympathetic or sympathetic innervation. The aim of this prospective longitudinal study was to identify physiological effects of reinnervation on changes in heart rate at rest and in response to various stimulations, and atrial function following the Cox-maze procedure and orthotopic heart transplantation.
PATIENTS AND METHODS: Thirty adult patients with combined the Cox-maze III procedure and mitral valve surgery (PD-partially denervated maze group) and 15 heart-transplant recipients (CD-completely denervated Tx group) were prospectively followedup at 1, 3, 6, and 12 months after surgery. The results were compared to normal probands (control group, n = 12). Power spectral analysis of heart rate variability, exercise testing, rapid positional changes, Valsalva maneuver, 24-hour Holler monitoring and standard electrocardiogram were used to assessed dysfunction of the autonomic nervous system, ability of the sinus node to accelerate in response to internal physiological chronotropic stimuli, and heart rate variability. The atrial function was assessed by transesophageal echocardiography.
RESULTS: At 1 and 3 months, both groups exhibited the physiological effects of denervation with no differences in cardiac autonomic activity between the two groups as demonstrated by very low values in power spectral analysis of heart rate variability, inhibited cardiac autonomic activity, and no response after sympathetic stress (LF, HF, LF/HF: at 0 degree 12.5, 32.5, 0.38; at 60 degrees 1.2, 16.4; and 0.07, respectively). At 6 months, the maze group but not the Tx group demonstrated evidence of autonomic reinnervation. Recovery of cardiac autonomic activity was documented at one year in the maze group (LF, HF, LF/HF: at 0 degree 206.7, 403.4, 051; at 60 degress 9.6, 25.6, 0.37, respectively) but not in the Tx group. Inappropriate heart rate responses during physical exercise were clearly evident in both groups after 1 and 3 months, with progressive improvement seen between 6 and 12 months only in the maze group. Atrial function after the maze procedure improved parallel to the recovery of the sinus node function.
CONCLUSION: Progredient improvement of sinus node function and atrial contractions with functional normalization one year after the Cox-maze procedure corresponded to functional reinnervation and recovery of autonomic nervous system.
*By invitation 53. MANAGEMENT OF VASODILATORY SHOCK AFTER HIGH- RISK CARDIAC SURGERY: INDENTIFICATION OF PREDISPOSING FACTORS AND USE OF A NOVEL PRESSOR AGENT. Michael Argenziano, M.D.*, Jonathan M. Chen, M.D.*, Asim F. Choudhri, B.S.*, Donald W. Landry, M.D.*, Alan D. Weinberg, M.S.*, Craig R. Smith, M.D., Eric A. Rose, M.D. and Mehmet C. Oz, M.D. New York, New York Discussant: Richard D. Weisel, M.D., Toronto, Ontario, Canada
Background: Cardiopulmonary bypass (CPB) may induce significant peripheral vasodilatation after cardiac surgery, often requiring perioperative catecholamine pressor support. Although arginine vasopressin (AVP) normally has no vasoactive properties in doses as high as 0.26 U/min, pressor effects have been described for lower doses in hypotensive patients unresponsive to traditional pressors. We investigated the incidence and clinical predictors of vasodilatory shock (VS) in high-risk cardiac surgical patients and the effects of low-dose AVP in the treatment of this syndrome.
Methods: In Phase I, 122 patients undergoing CPB for cardiac surgery (CABG, AYR, MVR, LVAD, heart transplant) were studied. Preoperative ejection fraction (EF), medications, and perioperative hemodynamics were recorded, and post-bypass serum AVP levels measured. VS was defined as mean arterial pressure (MAP) <70 mmHg, cardiac index (CI) >2.5 L/min/m2 and norepinephrine (NE) dependence. In Phase II, patients who developed VS after heart transplantation (HTx) or left ventricular assist device (LVAD) placement were identified and treated with AVP infusions (0.1 U/min) and hemodynamic responses recorded.
Results: In Phase I, 14 of 122 patients (11%) met criteria for post-bypass VS (MAP 65 ± 2 mmHg and NE dose 7.4 ± 1.8 mcg/min). By multivariate analysis, EF < 0.35 and diuretic use were identified as independent predictors of post-bypass VS (relative risk of 7.4 and 3.8, respectively). VS was associated with inappropriately low serum AVP concentrations (12.9 ± 3.1 pg/ml), and the severity of vasodilatation was correlated to the degree of AVP deficiency (r = 0.81). In Phase II, 25 patients (14 HTx, 11 LVAD) developed post-bypass VS (MAP 59.5 ± 2.0 mmHg and NE dose 19.8 ± 4.1 mcg/min) and responded to low dose AVP infusions, with rapid increases in MAP (to 82.2 ± 3.2 mmHg, p<0.0001) and reductions in NE dependence (to 8.6 ± 3.2 mcg/min, p = 0.02). Ten patients (40%) were weaned completely off NE within 30 minutes.
Conclusions: Heart failure and diuretic dependence are risk factors for post-bypass vasodilatory shock, which may be associated with vasopressin deficiency. In patients exhibiting this syndrome after high-risk cardiac surgery, physiologic replacement infusions of AVP stabilize hemodynamics by improving blood pressure and reducing or eliminating catecholamine pressor requirements.
12:10 p.m. ADJOURN
1994-96 AATS Robert E. Gross Research Scholar *By invitation 9:30 a.m. SIMULTANEOUS SCIENTIFIC SESSION C-2 GENERAL THORACIC SURGERY Ballroom C, Hynes Convention Center
Moderators: Mark K. Ferguson, M.D. G. Alexander Patterson, M.D.
54. LONG-TERM FOLLOWUP AFTER SURGICAL TREATMENT OF PRIMARY SOFT TISSUE SARCOMAS OF THE LUNG. John Langenfeld, M.D.*, Nael Martini, M.D., Michael Burt, M.D., Ph.D., Manjit S. Bains, M.D., Robert J. Downey, M.D.*, Valerie W. Rusch, M.D. and Robert Ginsberg, M.D. New York, New York Discussant: Cameron D. Wright, M.D., Boston, Massachusetts
Objective: Primary soft tissue sarcomas of the lung are rare, representing 0.4% of all lung malignancies, and data evaluating the results of treatment are sparse. Because of this, we reviewed our experience with this uncommon entity.
Methods: Retrospective review of medical records. Resectability, tumor size, and histological cell type were analyzed as predictors of survival. Survival was calculated by Kaplan-Meier method; survival differences compared by log rank.
Results: There were 42 patients; age ranged from 2 - 78 yr (median 53). There were leiomyosarcomas, 13 spindle cell sarcomas, 5 rhabdomyosarcomas, 4 malignant fibrous histiocytomas, 2 fibrosarcomas, and 1 malignant hemangiopericytoma. Of the 42 patients, 39 were explored with 19 complete resections (CR), 12 incomplete resections (IR), and 8 biopsy (Bx) only. There were 2 perioperative deaths (1 IR, 1 Bx). The overall actuarial five year survival was 25% (median 17 mos). The 1, 3, and 5 year survival for those having CR was 89%, 72% and 58%, respectively, which was significantly (p<0.0001) greater than those having an IR (1, 3, 5 yr survival = 42%, 8%, 0%; median 8 mos) or those having a Bx (1, 3 yr survival = 18%, 0%; median 7 mos). Eighteen patients (43%) developed distant metastases (11 to the contralateral lung, 7 bone, 3 brain and 1 liver). Regional metastases to mediastinal or hilar lymph nodes were identified in 12 patients (28%). Of the 19 patients having a CR, 10 (53%) developed a recurrence; 3 recurred distantly, 4 locoregionally, and 3 both distantly and locoregionally. The median time to the development of a recurrence following complete resection was 10 months. Tumor size did not impact on survival.
Conclusion: The best predictor for long-term survival in patients with primary soft tissue lung sarcomas is a complete resection. However, despite complete resection, patients remain at risk for the development of distant metastases, which occur predominately in lung and bone. As with other soft tissue sarcomas, adjuvant therapies should continue to be investigated.
*By invitation 55. BRONCHOALVEOLAR CARCINOMA: CLINICAL, RADIOLOGICAL, PATHOLOGICAL FACTORS AND SURVIVAL. Kenichi Okubo, M.D.*, Eugene J. Mark, M.D.*, Douglas Flieder, M.D.*, John C. Wain, M.D.*, Cameron D. Wright, M.D., Ashby C. Moncure, M.D., Hermes C. Grille, M.D. and Douglas J. Mathisen, M.D. Gifu, Japan and Boston, Massachusetts Discussant: Mark S. Allen, M.D., Rochester, Minnesota
Bronchoalveolar carcinoma (BAG) is the least common of the non-small cell lung carcinoma variants. The tumors typically spread along airways or by aerogenous routes and may be multifocal. Prognostic factors and patterns of survival have not been clearly defined.
We studied 119 patients with pathologically confirmed BAG, of which 114 underwent thoracotomy (107 resection, 7 biopsy only). Clinical, radiological and pathologic factors were examined and factors affecting survival were analyzed.
We have identified favorable prognostic factors that predict long term survival in BAC.
*By invitation 56. LONG-TERM RESULTS OF PROSTHETIC CHEST-WALL RECONSTRUCTION. Claude Deschamps, M.D., Ramin Darbandi-Tonkabon*, Mehmet B. Tirnaksiz, M.D.*, Victor F. Trastek, M.D., Daniel L. Miller, M.D.*, Mark S. Allen, M.D., Phillip G. Arnold, M.D.* and Peter C. Pairolero, M.D. Rochester, Minnesota Discussant: Richard H. Feins, M.D., Rochester, New York
Between January 1977 to December 1992, 137 patients (77 males and 60 females) underwent chest wall (CW) resection and reconstruction with prosthetic material. Median age was 57 years (range, 11-86). Indication for resection was recurrent malignancy in 46 patients (33.6%), contiguous lung or breast carcinoma in 44 (32.1%), primary CW malignancy in 37 (27.0%) and others in 10 (7.3%). Three patients (2.2%) presented with an open draining wound. Fifty patients (36.5%) were smokers, 39 (28.5%) had preoperative chemotherapy and/or radiation, and 19 (14.0%) were on oral corticosteroids. This study covers two time periods. Fifty-three patients (38.7%) were reconstructed with Prolene mesh (PM) during the period from 1977 to 1986. Eighty-four patients (61.3%) were reconstructed with Gore-tex (GT) soft tissue patch from 1984 to 1992. Soft tissue coverage with transposed muscle or omentum was used in 75 patients (54.7%); the remainder had coverage with local tissue only. There were 2 deaths (1.5%) and 48 patients (35%) had at least one complication. Median hospitalization was 19 days (range, 4-142). Followup was complete in all 135 operative survivors and ranged from one to 144 months (median, 11 months). Ninety-five patients (70.4%) had well-healed wounds and were asymptomatic. An additional 36 patients (26.7%) initially had well healed wounds but developed a local recurrence. Local recurrence was greatest in those patients with breast cancer. Four PM and none of the GT had to be removed. Seromas occurred in eight patients; seven were small and resolved while the last (GT) required wound exploration. Late wound infections occurred in seven patients (5 PM and 2 GT; P0.05). Four of the five PM had to be removed. The wounds in the remaining three patients (2 GT and 1 PM) were all opened and cleaned with eventual healing in all three. While late wound infection occurred more commonly in patients with PM, this may reflect our early inexperience of placing prosthetic material in contaminated wounds. We conclude that CW resection and prosthetic reconstruction will yield satisfactory results in most patients. 57. PREVALENCE AND LOCATION OF NODAL METASTASES IN DISTAL ESOPHAGEAL ADENOCARCINOMA CONFINED TO THE WALL. John J. Nigro, M.D.*, Steven R. DeMeester, M.D.*, Jeffrey A. Hagen, M.D.*, Stefan Oberg, M.D.*, Jeffrey H. Peters, M.D.*, Milton Kiyabu, M.D.*, Peter F. Crookes, M.D.*, Cedric G. Bremner, M.D.* and Tom R. DeMeester, M.D. Los Angeles, California Discussant: Thomas W. Rice, M.D., Cleveland, Ohio
Background: Barrett's surveillance programs have identified an increasing number of patients with early esophageal cancer. The purpose of this study was to characterize the prevalence and location of lymph node metastases in patients with invasive adenocarcinoma confined to the esophageal wall.
Methods: Esophagogastrectomy combined with systematic mediastinal and abdominal lymphadenectomy was performed on 34 patients with Barrett's adenocarcinoma confined to the esophageal wall. Patients who had preoperative chemotherapy, radiotherapy, or previous esophageal resection were excluded. Followup was complete in 33 of 34 patients at a mean of 34 months (median 22 months).
Results: The depth of tumor invasion was limited to the mucosa in 15 patients, the submucosa in 9 patients, and the muscularis propria in ten patients. Histologically positive nodes were found in 12 of 34 patients (35.3 %), and the prevalence of nodal metastases increased progressively with depth of invasion.
The location of each positive node is presented in the following table, grouped by the number of involved nodes per patient.
Actuarial survival for the entire group was 63% at 5 years. Recurrence was identified in 5 of the 34 patients (15 %). There have been no recurrences in patients with intramucosal carcinoma. However, systemic recurrence was identified in 2 of 9 patients with submucosal and 3 of 10 patients with muscular invasion.
Conclusions: The prevalence of nodal metastases correlates with depth of wall invasion, and once the muscular wall is penetrated the majority of patients have involved nodes. Although most metastases are found along the lesser curvature, lymph node metastases do occur in areas not routinely removed by trans-hiatal resection. A low recurrence rate and good long term survival can be achieved in these patients emphasizing the importance of total esophagectomy with lymphadenectomy to include all potentially involved nodes.
10:50 a.m. INTERMISSION
*By invitation 11:10 a.m. SIMULTANEOUS SCIENTIFIC SESSION C-2 GENERAL THORACIC SURGERY Ballroom C, Hynes Convention Center
Moderators: Mark K. Ferguson, M.D. G. Alexander Patterson, M.D.
58. HEART-LUNG VERSUS DOUBLE-LUNG TRANSPLANTATION FOR SUPPURATIVE LUNG DISEASE. Clifford W. Barlow, M.D.*, Robert C. Robbins, M.D.*, Marc R. Moon, M.D.*, James Theodore, M.D.* and Bruce A. Reitz, M.D. Stanford, California Discussant: Thomas M. Egan, M.D., Chapel Hill, North Carolina
Heart-lung (HLTx) and double-lung (DLTx) transplantation have both been advocated for patients with end-stage suppurative lung disease. We reviewed our experience of all patients with cystic fibrosis (CF) or bronchiectasis who underwent HLTx or DLTx between 1/88 and 8/97 to obtain medium-term follow-up and comparative results. Twenty-two patients (13 male, 21 CF) had HLTx and 24 patients (8 male, 19 CF) had DLTx. Domino1 heart transplantation took place in all HLTx patients with suitable cardiac function. The mean age of the HLTx group was 26 ± 13 years (range 9 to 49) and of the DLTx group was 28 ± 14 years (range 10 to 53). There were no differences in weight, pre-operative creatinine, bilirubin, CMV status, maintenance immunosuppression, and medical management between the 2 groups. Patients all received induction therapy with monoclonal (OKT3) (9 HLTx, 3 DLTx) or polyclonal (RATG) antibody (13 HLTx, 21 DLTx). Sixteen of 24 patients had DLTx after 1/95 while 13 of 22 patients had HLTx before 1/91. Mean waiting time for DLTx was 3614 ± 47 days and for HLTx was 270 ± 52 days. Actuarial analyses of outcomes for the groups are as follows:
Thirty day survival was 100 % for HLTx and 96 % for DLTx. There were 7 late HLTx deaths (3 OB, 2 infection, 0 GCAD, 2 Other) and 2 late deaths in the DLTx group (2 OB).
These data show that HLTx and DLTx for suppurative lung diseases have comparable short and medium-term survival, OB, rejection and infection rates. While heart rejection episodes and GCAD may be experienced in HLTx, they have not resulted in retransplantation or death. We continue to consider DLTx or HLTx with domino' heart transplantation for patients with end-stage suppurative lung disease.
*By invitation 59. EFFECT OF CMV AND DEVELOPMENT OF HLA ANTIBODIES ON LUNG TRANSPLANT SURVIVAL AND BRONCHIOLITIS ON OBLITERANS SYNDROME. Michael A. Smith, M.D.*, Sudhir Sundaresan, M.D.*, T. Mohanakumar, Ph.D.*, Elbert P. Trulock, M.D.*, John P. Lynch, M.D.*, Donna L. Phelan, B.A, C.H.S.*, Joel D. Cooper, M.D. and G. Alexander Patterson, M.D. St. Louis, Missouri Discussant: R. Morton Bolman, III, M.D., Minneapolis, Minnesota
The long term survival of lung transplant recipients continues to be limited by the development of bronchiolitis obliterans syndrome (BOS). A retrospective analysis was done to identify factors which may effect long-term survival and play a role in the development of BOS.
Methods: Of 295 consecutive lung transplants (LT) performed from July 1988 to December 1995, 264 survived at least 3 months and form the basis for this analysis. There was a minimum followup period of 12 months. BOS was defined as a decline in spirometry (FEV1 < 80% of baseline) and/or histologic presence of obliterative bronchiolitis. Variables analyzed included cytomegalovirus (CMV) antibody status and post-transplant development of lymphocytotoxic antibodies against human leukocyte antigens (HLA antibodies). Recipient sera were tested by microcytotoxicity assays. Those having cytotoxicity to greater than 10% of a known panel of HLA alleles were identified as positive for HLA antibodies. Variables were subjected to Kaplan-Meier analysis for effects on overall survival and freedom from BOS.
Results: Overall survival was 78%, 55%, and 39% at 3, 5, and 7 years, respectively. At 3 and 5 years 39% and 23% of patients, respectively, remained free of BOS. There appeared to be worse survival of CMV seronegative patients who received allografts from a seropositive donor compared to seronegative recipient/donor combinations and seropositive patients receiving allografis from seronegative donors (p = 0.074). There was also a trend toward greater frequency of development of BOS in seronegative patients who received allografts from seropositive donors (p = 0.116). Most strikingly, the development of HLA antibodies after transplantation significantly correlated with the development of BOS (p = 0.048). Further, these patients who developed HLA antibodies had poorer survival (p = 0.063).
Conclusion: These data suggest that BOS is the result of an immune mediated process in which HLA antibodies may play a significant role. The effect of CMV on the development of BOS and long term survival remains uncertain.
1998-99 TSFRE Research Fellow *By invitation 60. BILATERAL SEQUENTIAL LUNG TRANSPLANT WITHOUT STERNAL DIVISION ELIMINATES POST TRANSPLANT STERNAL COMPLICATIONS. Bryan Meyers, M.D.*, Sudhir Sundaresan, M.D.*, Joel D. Cooper, M.D. and G. Alexander Patterson, M.D. St. Louis, Missouri
Bilateral sequential single lung transplantation (BLT) is the procedure of choice for bilateral lung replacement. This procedure is usually performed through a bilateral anterolateral thoracosternotomy ("clam-shell") incision. Unfortunately the "clam-shell" is associated with complications such as sternal malunion, dehiscence, osteomyelitis and migrating hardware in as many as 50% of patients. From 1989 to the present we have performed 228 adult BLT. During the past 16 months, 35 of 47 BLT were conducted without sternal division. Thirty-one were performed through bilateral anterolateral fourth interspace thoracotomies, three through left posterolateral and right anterolateral thoracotomy and one through sequential posterolateral thoracotomies. The recipient diagnoses were COPD in 11 patients, alpha-1 antitrypsin deficiency emphysema in seven patients, cystic fibrosis in seven patients, bronchiectasis in five patients, pulmonary fibrosis in three patients, bronchiolitis obliterans (retransplant) in one patient and primary pulmonary hypertension in one patient. Three patients were placed on cardiopulmonary bypass electively to permit transplantation. The intact sternum did not impede ascending aortic and right atrial cannulation. Early in our experience with this technique two patients underwent sternal division for institution of emergent cardiopulmonary bypass to repair an atrial laceration in one patient and a left atrial suture line defect in another patient. These two patients were the only deaths in this experience (5.7%). There were no incisional complications among operative survivors. With experience BLT can be safely conducted without sternal division thereby avoiding the frequent complications associated with transverse sternotomy.
12:10 p.m. ADJOURN
*By invitation |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
We Model Excellence |
|
Read the Privacy Policy.
IMPORTANT REMINDER: The preceding information is intended only to provide
general guidance and not as a definitive basis for diagnosis or treatment in any particular case.
It is very important that you consult a doctor about any specific medical problem or question.