 |
|
| Wednesday Morning, May 7, 1997 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Back to Annual Meeting Program
WEDNESDAY MORNING, MAY 7, 1997
7:00 a.m. FORUM SESSION II - GENERAL THORACIC SURGERY South Sheraton Ballroom
Moderators: Nasser K. Altorki, M.D. Larry R. Kaiser, M.D.
F9. TOTAL RESPIRATORY SUPPORT FROM SWINE PULMONARY XENOGRAFTS IN PRIMATES. Casey W. Daggett, M.D.*, Mark Yeatman, F.R.C.S.*, Edward P. Chen, M.D.*, Andrew J. Lodge, M.D.*, Carmelo Gulotto, B.B.A.*, Shu S. Lin, M.D.*, Jeffery L. Platt, M.D.* and Robert D. Davis, M.D.* Durham, North Carolina Sponsored by: Ross M. Underleider, M.D., Durham, North Carolina
Background. The use of nonhuman lung donors, such as swine, has the potential to provide an unlimited supply of organs. However, a major barrier to pulmonary xenotransplantation is hyperacute rejection. Objective. To test the hypothesis that pre-transplant swine lung perfusion will deplete xenoreactive antibody, prevent hyperacute swine-to-primate pulmonary xenograft rejection and allow for a functional swine pulmonary xenograft. Methods. Six baboons (12-15 kg) underwent left pneumonectomy followed by left orthotopic swine lung transplantation. Three baboons (group I) received antibody depletion by perfusion with swine lungs prior to transplantation, and three received no treatment prior to transplantation (group II). Results. Perfusion of baboon blood through swine lungs for 120 minutes achieved an 85 ± 5% reduction in xenoreative IgM levels. Following transplantation, group I pulmonary xenografts had a blood flow of 613.3 ± 122.1 ml/min and a pulmonary vascular resistance (PVR) of 29.4 ± 14.4 mm Hg/L/min at 60 minutes of reperfusion, compared to group II, which had a pulmonary blood flow of 26.7 ± 12.1 ml/min (p<0.05) and a PVR of 440.4 ±196.1 mm Hg/L/min (p<0.05). While group II lungs lost all pulmonary blood flow by three hours of reperfusion, group I pulmonary xenografts continued to function well for the duration of the study and at eleven hours of reperfusion maintained a pulmonary blood flow of 440.0 ± 40.7 ml/min with a PVR of 51.5 ± 23.7 mm Hg/L/min. After 60 minutes of reperfusion, two of three group I animals also tolerated complete occlusion of the right pulmonary artery, having the baboon rely completely on the swine pulmonary xenograft for respiratory function over eleven hours. Group II animals did not tolerate occlusion of the right pulmonary artery and displayed cardiovascular failure within 20 seconds of occlusion. Pathologic analysis of group I lungs displayed little histologic evidence of injury in biopsies taken at one and eleven hours of reperfusion. However, group II lung biopsies taken at one hour of reperfusion showed alveolar edema and hemorrhage with small vessel thrombosis. The function of group I pulmonary xenografts during occlusion of the right pulmonary artery were measured by arterial blood gas analysis, illustrated in the table below. (Data are given as the mean (± SEM).
Conclusion. Pre-transplant perfusion with swine lungs is effective in removing xenoreactive antibodies and prevents hyperacute pulmonary xenograft rejection. Swine pulmonary xenografts can provide complete respiratory support in primate recipients.
*By invitation F10. CONTROLLED REPERFUSION AFTER LUNG ISCHEMIA: IMPLICATIONS FOR IMPROVED FUNCTION AFTER LUNG TRANSPLANTATION. Bradley S. Allen, M.D.*, Ari Halldorsson, M.D.*, Michael Kronen, M.D.*, Kirk S. Boiling, M.D., M.P.H.*, Tingrong Wang, M.D.*, Shaikh Ramon, M.S.* and Harold Feinberg, Ph.D.* Chicago, Illinois Sponsored by: Renee S. Hartz, M.D., Chicago, Illinois
Despite improvements in tissue preservation, reperfusion injury remains a major source of morbidity and mortality following lung transplantation. Although controlling the conditions of reperfusion and the composition of the reperfusate have been shown to modify the reperfusion injury in the myocardium, these principals have not been investigated following lung ischemia. Twenty adult pigs underwent 2 hours of warm left lung ischemia by cross clamping the left bronchus and pulmonary artery. In 5 (Group 1) the cross clamp was simply removed (unmodified reperfusion). Fifteen other pigs underwent modified reperfusion using blood from the femoral artery to perfuse the lung via the pulmonary artery (pressure <50mm Hg) for 10 minutes prior to removing the clamps. In 5 (Group 2) blood was mixed with crystalloid using a BCD resulting in a substrate enriched, hypocalcemic, hyperosmolar, alkaline solution, in 5 (Group 3) the blood was circulated through a leukocyte depleting filter, and 5 (Group 4) underwent reperfusion with a WBC filter and modified solution. Left lung function was assessed 60 minutes after reperfusion and expressed as percentage of control, and a biopsy taken for lung water and myloperoxidase (us/me protein).
In conclusion, after 2 hours of pulmonary ischemia: 1) a severe lung injury occurs following uncontrolled (unmodified blood) reperfusion, 2) controlled reperfusion with either a modified reperfusion solution or WBC filter limits, but does not avoid, a lung reperfusion injury, 3) reperfusion using both a modified reperfusate and WBC filter results in complete preservation of pulmonary function. We therefore believe surgeons should control the reperfusate following lung transplantation to improve postoperative pulmonary function.
*By invitation F11. PROGNOSTIC SIGNIFICANCE OF p27 AND p53 IN PATIENTS WITH ADENOCARCINOMA IN BARRETT'S ESOPHAGUS. Surendra P. Singh, M.D.*, Jennifer Lipman, M.D.*, M. Giulia Cangi, Ph.D.*, Laura Aizenman*, F. Henry Ellis, Jr., M.D. and Massimo Loda, M.D.* Boston, Massachusetts
BACKGROUND. Loss of cell cycle control is believed to be an important step in human tumor development. The Cyclin-dependent kinase inhibitors (Ckis) p21 and p27 mediate cell cycle arrest by preventing cells from entering S phase. p21 is induced by tumor suppressor p53 in response to DNA damage, while p27 is induced in quiescent/terminally differentiated cells.
HYPOTHESIS. We hypothesize that loss of p27 in invasive carcinomas may be associated with disease progression. Finally, overexpression of p53, indicative of mutation of this gene with subsequent failure to induce the Cki p21, may also be associated with cancer progression.
METHODS. We previously evaluated the expression of p27 in formalin-fixed and paraffin-embedded sections from 49 cases of invasive adenocarcinoma, 4 carcinoma in situ, in Barrett's esophagus (BE) by immunohistochemistry with a monoclonal anti-p27kip1 antibody. In this study we also examined the expression of p53 as another cell cycle regulator. Twenty-three of these cases had BE-associated dysplasia. As previously described, cases with ≥50% and ≥10% positive nuclear staining were considered positive for p27 and p53, respectively.
RESULTS. As expected, normal mucosa and non-dysplastic BE showed no p53 immunoreactivity while p27 was expressed in the nuclei of superficial differentiated cells. In contrast, p27 was expressed in the base of the pits of all cases of dysplasia, presumably to counteract enhanced proliferative activity. p53 was also overexpressed in 74% of dysplasias. Tumor progression was associated with loss of p27 and p53 overexpression: 82% of invasive cancers had low p27 expression while p53 was overexpressed in 55%. p27 expression correlated with patient survival (p=0.0007), presence of lymph node metastasis (p=0.0014) and histopathologic differentiation (p=0.0001) whereas p53 correlated only with histopathologic differentiation (p=0.0035). There was no correlation between p27 and p53.
CONCLUSIONS. 1) p27 overexpression in dysplastic cells of BE may be a physiological response to genetically damaged cells. 2) p27 and p53 may be used as immunohistochemical markers of dysplasia in patients with BE. 3) Loss of p27 expression, but not p53 overexpression, is a negative prognostic factor in patients with adenocarcinoma in BE.
*By invitation F12. INHALED NITRIC OXIDE ATTENUATES ISCHEMIA-REPERFUSION INJURY AFTER NON-HEART-BEATING-DONOR LUNG TRANSPLANTATION. Emile A. Bacha, M.D.*, Shinya Murakami, M.D.*, Paolo Machiarini, M.D.*, Guy-Michel Mazmanian, M.D.*, Alain R. Chapelier, M.D.*, Philippe Herve, M.D.* and Philippe G. Dartevelle, M.D. Boston, Massachusetts and Le Plessis-Robinson, France
Nitric oxide inhibits polymorphonuclear neutrophils (PMN) activation and attenuates pulmonary IR injury. We studied the effect of inhaled NO on IR injury after NHBD lung transplantation by measuring lung function, recipient survival, graft PMN sequestration as well as adherence of recipient circulating PMN to cultured pulmonary artery endothelial cells (PAEC).
Methods: Pigs were assigned to a NO (30 ppm) vs control group (n = 9). Cadavers were ventilated. After 3 hours of postmortem in situ warm ischemia, and 2 hours of cold ischemia, left allotransplantation was performed. The right PA was ligated one hour after reperfusion. Hemodynamic and gas exchange data were recorded hourly for 9 hours. Circulating PNM adherence to tumor necrosis factor-alpha (TNF)- and calcium ionophore (Cal)-stimulated PAEC was measured after reperfusion. Lung PMN sequestration was determined by measuring myeloperoxidase activity.
Results: After PA ligation, NO-treated animals exhibited significantly (two-way analysis of variance) lowered pulmonary vascular resistance (p<0.01), improved oxygenation (p<0.01), and survival (p<0.05). Adhesion of PMN to PAEC was significantly inhibited in the NO group (22 ± 3 vs 33 ± 3% after Cal stimulation; 20 ± 3 vs 52 + 3% after TNF stimulation, p<0.0001). PNM sequestration was significantly reduced by NO (0.12 ± 0.06 vs 0.25 ± 0.04 U/100mg tissue, p<0.05).
Conclusions: Inhaled NO attenuates IR injury after NHBD lung transplantation. This is likely the result of a dual action by inhaled NO: 1) prevention of IR-induced pulmonary vasoconstriction, and 2) direct action on PMN resulting in inhibition of adherence to endothelium.
*By invitation F13. IN VIVO AND EX VIVO GENE TRANSFER IN RAT LUNG ISOGRAFTS. Carlos H. Boasquevisque, M.D.*, Bassem N. Mora, M.D.*, Teng C. Lee, B.S.*, Ronald K. Scheule, Ph.D.*, Joel D. Cooper, M.D, Mitchell D. Botney, M.D.* and G. Alec Patterson, M.D. St. Louis, Missouri and Framingham, Massachusetts
Background: In transplantation, gene transfer to the donor organ prior to transplant offers potential for targeted therapy directed at specific post-operative complications, such as ischemia-reperfusion injury and rejection. It is important that the vector be non-toxic and the transferred gene be expressed at the time of implantation. The aim of this study was to achieve transgene expression in transplanted lung grafts using a cationic lipid complexed to a reporter gene.
Methods: cDNA encoding for chloramphenicol acetyl transferase (CAT) was complexed to a cationic lipid, Lipid #67 (Genzyme Corporation, Framingham, MA), and was injected into Fischer rats. Successful transfection was assessed by the CAT assay. The distribution and type of transfected cells were evaluated by in situ hybridization. Lung toxicity was assessed by measuring arterial oxygenation (PaO2), the host inflammatory response (by H&E staining and EDI immunohistochemistry) and TNF-α levels. Animals were divided into three major groups. In Group 1 (non-transplant setting), the CAT-Lipid #67 complex was injected intravenously via the left external jugular vein. Lungs were harvested at various time points later: 2 hours, 6 hours, 12 hours, 1, 2, 3, 5, 8 and 21 days (n = 3). In Group 2 (transplant setting - in vivo graft transfection). rats were divided into 3 sub-groups (n = 5). In sub-group 1, animals were intravenously injected with the CAT-Lipid #67 complex 4 hours prior to left lung harvest and orthotopic implantation in recipient animals, which were sacrificed 44 hours later. In sub-groups 2 and 3, lungs were harvested 4 hours and 48 hours after intravenous injection and served as controls. In Group 3 (transplant setting - ex vivo graft transfection) the CAT-Lipid #67 complex was infused retrograde via the left pulmonary vein after graft harvest and flush (n = 6). Grafts were then kept at room temperature for 4 hours prior to implantation. Recipients were sacrificed 44 hours later.
Results: Gene expression was detected as early as 2 hours. High levels of gene expression were present from 6 hours to 8 days. By 21 days, gene expression was greatly attenuated. Transgene expression was observed in all treated animals and was homogeneously distributed throughout the lung. In situ hybridization localized CAT mRNA to endothelial cells, macrophages and interstitial cells. Lung gas exchange was not significantly different in treated and untreated animals (PaO2, mmHg: 501.18 ± 40.89, 537.17 ± 71.24, and 523.56 ± 18.4 for transplanted treated animals, non-transplanted treated animals and untreated normal rats, respectively, p=0.8). Inflammatory infiltrate was minimal, although TNF-a levels increased seven-fold in treated animals.
Conclusion: In vivo and ex vivo cationic-lipid-mediated gene transfer to lungs isografts is possible and allows significant transgene expression without impairment in graft function.
*By invitation F14. AEROSOL CYCLOSPORINE PREVENTS ACUTE ALLOGRAFT REJECTION IN EXPERIMENTAL PULMONARY TRANSPLANTATION. Surindra N. Mitruka, M.D.*, Si M. Pham, M.D.*, Adrianna Zeevi, Ph.D.*, Sen Li, M.D.*, Jane Cai, M.D.*, Gilbert J. Burckart, Pharm.D.*, Samuel A. Yousem, M.D.*, Robert J. Keenan, M.D.* and Bartley P. Griffith, M.D. Pittsburgh, Pennsylvania
BACKGROUND: The incidence of acute rejection and the morbidity associated with systemic cyclosporine (CsA) following pulmonary transplantation is significant. Recent evidence suggests that the lung allograft locally initiates and modulates the immune mechanisms involved in acute rejection. The purpose of this study was to determine if regional immunosuppression with aerosolized cyclosporine would prevent acute lung rejection, achieve high intra-graft concentration with low systemic delivery, and effect production of the pro-inflammatory cytokines involved in the acute rejection response.
METHODS: Unilateral orthotopic left lung transplantation was performed in 18 rats (ACI to Lewis) across major and minor histocompatibility barriers. The rats were divided into two groups : allogeneic control (n = 6) and aerosolized (3 mg/kg/day) cyclosporine (n = 12). Rats were sacrificed on POD 2, 4, and 6, and the transplanted lung, native lung, spleen, and blood collected. Histology, HPLC for CsA concentrations, and RT-PCR for cytokine gene expression was performed. Low dose (2 mg/kg/day) and high dose (10 mg/kg/day) systemic CsA groups (previous data) were used for comparison.
Aerosol CsA controlled rejection with a significantly less blood concentration (725 ng/ml vs. 2046 ng/ml) and a similar tissue concentration (12, 824 ng/mg vs. 14, 519 ng/mg) compared to the high dose systemic group. The pro-inflammatory cytokines increased continuously in untreated animals from POD 2 to POD 6 at which time rejection was complete. Aerosol CsA treated animals initially expressed IL-6 and IFN-g on POD 2 but none thereafter, and iNOS production was completely attenuated; similar to results obtained in the high dose systemic CsA group.
CONCLUSION: Local delivery of CsA by aerosol inhalation effectively prevented acute rejection of the rat lung allograft. Moderate dose aerosolized CsA achieved high graft concentrations with low systemic delivery. The gene expression of pro-inflammatory cytokines involved in acute rejection was suppressed by aerosol CsA therapy.
*By invitation F15. EXPRESSION OF ACIDIC FIBROBLAST GROWTH FACTOR CONTRIBUTES TO MALIGNANT TRANSFORMATION IN BARRETT'S ESOPHAGUS. Robert Soslow, M.D.*, Liang Ying, M.D.* and Nasser K. Altorki, M.D. New York, New York
The process of tumorigenesis involves loss of function of tumor suppressor genes or activation of oncogenes many of which encode for various growth factors. Acidic fibroblast growth factor (aFGF) is a potent mitogen whose RNA transcripts were shown to be overexpressed in Barrett's adenocarcinoma. In this study we investigated aFGF protein expression in 17 esophagectomy specimens from patients with Barrett's adenocarcinoma. Immunostaining was performed on paraffin embedded tissue using a streptavidin-biotin technique with monoclonal antibody against aFGF. In nine cases, the examined sections contained residual Barrett's epithelium (metaplasia 5, low grade dysplasia 3, high grade dysplasia 9). Epithelial cells were considered positive for aFGF if greater than 10% were immunostained. Results are shown below:
Immunostaining was very intense (3+) in all carcinoma cases and 8/9 cases with high grade dysplasia. We conclude that increased expression of aFGF plays an important role in tumorigenesis in patients with Barrett's esophagus and further studies should be conducted to evaluate its use as a clinical blomarker.
*By invitation F16. SUCCESSFUL IN VIVO AND EX VIVO TRANSFECTION OF PULMONARY ARTERY SEGMENT IN LUNG ISOGRAFTS. Motoki Yano, M.D.*, Carlos H. Boasquevisque, M.D.*, Itaru Nagahiro, M.D.*, Masafiimi Hiratsuka, M.D.*, Bassem N. Mora, M.D.*, Joel D. Cooper, M.D. and G. Alec Patterson, M.D. St. Louis, Missouri
Background: Gene transfer into donor lung grafts is feasible and may be useful in reducing reperfusion injury and rejection. However using conventional viral vectors whole organ transfection remains inefficient. Yet focal proximal pulmonary artery endothehal transfection may provide satisfactory downstream effects on the whole graft. The aim of this study was to achieve transfection of proximal pulmonary artery segments in left lung isografts.
Methods: Fisher rats (250-280 g) were divided into three groups. In group I (n = 4) and group II (n = 7) intact donor rats were subjected to occlusion of the proximal left pulmonary artery segment for twenty minutes after which flow was restored. In group I, 0.03 ml saline was injected into the pulmonary artery segment via a catheter in the right ventricle. In group II, the pulmonary artery segments were injected with 2-4x1010 pfu/ml replication deficient adenovirus type V with LacZ gene encoded β-galactosidase. In group III (n = 5) donor lungs were flushed with 20 ml LPDG solution and extracted. The same construct as for group II was instilled ex vivo into the occluded left pulmonary artery segment. After three hours storage (10°C), the grafts were implanted. In all groups, seventy-two hours after reperfusion, heart-lung blocks were flushed with PBS and Bluo-gal and immersed in Bluo-gal for three hours.
Results: The survival rates were 50% (group I), 43% (group II) and 100% (group III). Macroscopically, in all survival animals of group II and III, multiple blue spots were observed on the endothelial surface of the proximal left pulmonary artery indicating successful gene transfection. Microscopically, blue stained endothelial and smooth muscle ells were observed.
Conclusion: A high rate of focal gene transduction was observed in proximal pulmonary artery segments following in vivo and ex vivo exposure. Direct gene transfer to pulmonary artery segments is feasible and may avoid potential complications of systemic transfection strategies.
*By invitation F17. CHANGES IN PULMONARY PHYSIOLOGY AFTER LUNG VOLUME REDUCTION SURGERY IN A RABBIT MODEL OF OBSTRUCTIVE DIFFUSE EMPHYSEMA. Joseph Huh, M.D.*, Matthew Brenner, M.D.*, John C. Chen, M.D.*, Edward A. Stemmer, M.D., Benedict Yoong, B.S.*, David Mukai, B.S.* and Jeffrey C. Milliken, M.D.* Orange, California
Purpose: While surgical treatment of emphysema has recently gained popularity, the mechanism by which lung volume reduction surgery (LVRS) improves respiratory physiology is still incompletely understood. Using an elastase induced purely obstructive emphysema model in New Zealand White rabbit, we studied the effects of LVRS on puhnonary compliance, airway flow, measured lung volume, and diffusion capacity.
Methods: Emphysema was induced in 14 New Zealand white rabbits by aerosolizing 15, 000 units of porcine elastase through an endotracheal tube under general anesthesia. Transpleural pressures were measured at 60, 50, 40, 30, and 20 cc's inflation above functional residual capacity (FRC). Measurements were taken at baseline prior to induction of emphysema, preoperatively at 4 weeks following induction of emphysema, and 1 week postoperatively following LVRS. FEV1, helium dilution lung volume, and single breath DLco were also measured concurrently. Stapled resection of bilateral upper lobes was performed through a midline sternotomy with a standard multirow surgical stapler (Ethicon). Histologic examination was obtained one week postoperatively.
Results: Comparison of compliance curves showed an increase in compliance following induction of emphysema and a decrease in response to LVRS (graph). In like fashion, FEV1 showed improvement in airway flow postoperatively, although this did not reach statistical significance, while FRC decreased following LVRS. DLco did not show a significant change (Table). Histologic examination confirmed presence of severe diffuse emphysema in each animal at necropsy.
Conclusion: We have developed an animal model of elastase induced diffuse emphysema applicable for LVRS studies. Decreased compliance and increased airway flow following volume reduction surgery parallels findings in human studies and suggests that similar mechanisms of increased elastic recoil and airway support contribute to improvement. Furthermore, helium dilution volumes show a decrease in lung volume postoperatively without significant decrease in diffusion capacity. This model may be useful in assessing surgical techniques in LVRS, and may help identify optimal location and quantity of lung tissue excision in the surgical treatment of emphysema.
Supported by NIH Grant #RR-011-92, DOE Grant #DE-FG03-91ER61, and DOD Grant #N00014-91-C-0134.
*By invitation F18. THE RELATIONSHIP OF ISCHEMIA REPERFUSION INJURY AND THE EXPRESSION OF MAJOR HISTOCOMPATIBILITY COMPLEX ON HOST LYMPHOCYTES. Karim A. Qayumi, M.D., Ph.D.*, David V. Godin, Ph.D.*, Maryam Nikbakht-Sangari, B.Sc.*, John C. English, M.D.*, Kathleen J. Horley, Ph.D.*, Seung P. Lim, M.D., Ph.D.*, Shahid Gul, B.Sc.* and Michael S. Koehle, B.Sc.H.* Vancouver, British Columbia and Toronto, Ontario, Canada; Chung-Ku, Korea Sponsored by: G. Frank O. Tyers, M.D., Vancouver, British Columbia, Canada
This study was designed to examine the effect of ex vivo preservation time on the release of specific inflammatory mediators, and the levels of plasma and tissue antioxidants related to the rejection of the transplanted organ (measured by the expression of MHC HLA-DR-β on host lymphocytes). Single lung transplantation was performed on three gropus of domestic swine. Group A (n = 7) and Group B (n = 6) had ex vivo preservation times of 4 and 15 hours respectively at 4°C hypothermia. Group C (n = 6) underwent 2 hours of warm ischemia with the left pulmonary artery, vein, and bronchus cross-clamped without explantation. Methods of assessment included: the release of inflammatory mediators-thromboxane B2 (TxB), interleukin-2 (IL-2), IL-4, IL-10, tumour necrosis factor α (TNFα) quantitated by radioimmunoassay and/or enzyme linked immunosorbent assay; the levels of plasma and tissue antioxidants determined by enzyme bioassay, the expression of MHC HLA-DR-P on host lymphocytes by fluorescence intensity; and the mechanics of lung function by measurement of lung compliance, a/A ratio, and lung weight. The results demonstrated increases (p<0.05) TxB, IL-2, IL-4, lung weight, O2 gradient, and HLA-DR-β expression on host lymphocytes directly proportional to ischemic time. IL-10, TNFα, lung compliance, a/A ratio, and both plasma and tissue antioxidants were inversely proportional to ischemic time (p<0.05). Similar results were observed in Group C which experienced ischemia reperfusion injury without tissue incompability. These results suggest that ischemia reperfusion injury alone without the complication of tissue incompatibility. These results suggest that ischemia reperfusion injury alone without the complication of tissue incompatibility is enough to initiate an acute post-transplantation response. Thus the severity of ischemia reperfusion injury, as measured by the release of inflammatory mediators and the levels of antioxidants, could be directly related to the intensity of rejection of the transplanted organ, as measured by the expression of HLA-DR-β on host lymphocytes.
*By invitation 7:00 a.m. FORUM SESSION III - CARDIAC SURGERY North Sheraton Ballroom
Moderators: D. Glenn Pennington, M.D. Tirone E. David, M.D.
F19. LONG-TERM GENE EXPRESSION AFTER VIRAL TRANSDUCTION OF CARDIAC ISOGRAFTS USING DNA VIRAL VECTORS. Boulos Asfour, M.D.*, Paul D. Kessler, M.D.*, Ralph H. Hruban, M.D.*, Duke E. Cameron, M.D. and Barry J. Byrne, M.D., Ph.D.* Baltimore, Maryland
Objectives: Viral transduction of cardiac allografts provides the opportunity to genetically modify graft vasculature and myocardium as well as to achieve graft-specific immunosuppression or possibly tolerance. However, successful gene therapy in cardiac transplantation will require both gene delivery in a clinically applicable manner and long-term gene expression. Adenovirus has been used successfully to achieve short-term transduction in cardiovascular tissue. Adeno-associated virus (AAV) is a parvovirus which we have previously shown to be effective in long-term transduction of skeletal muscle, vascular smooth muscle, and cardiomyocytes in vitro and in vivo.
Methods: Female adult Sprague-Dawley rats (200-250 gms, n = 20) served as isograft donors and recipients for vascularized heterotopic cardiac transplants. Following cannulation of the right carotid artery, the graft was perfused with oxygenated cold (15°C) Krebs-Henselite solution and the donor organ harvested. A viral solution containing either, AAV-IacZ (5 x 109 particles) or Ad-lacZ (1 x 109 pfu), encoding the marker protein, bacterial β-galactosidase was then delivered to the coronary vasculature. Perfusion was discontinued to allow for intracardiac recirculation of virus via creation of an atrial septal defect or pulmonary bypass (PA-LA or PA-Ao shunt).
Results: Two weeks to two months following transplantation, grafts were removed for histological analysis of β-galactosidase activity. Uniform staining of cardiac and vascular smooth muscle was observed at the highest dose with viral recirculation. The transgene was expressed for up to two months in hearts transduced by AAV-lacZ.
Conclusions: We have shown that AAV-lacZ and Ad-lacZ delivered via the coronary vasculature of cardiac allografts are able to transduce vascular and cardiac tissue under conditions of hypothermic perfusion and storage currently used in clinical cardiac transplantation. This approach will be useful for genetic modification of cardiac allografts in the management of graft vasculo-pathy and rejection.
*By invitation F20. HEARTS AND KIDNEYS FROM TRANSGENIC SWINE EXPRESSING HUMAN COMPLEMENT REGULATORY PROTEINS ARE PROTECTED FROM HYPERACUTE REJECTION IN PRIMATES. Sepehre Naficy, M.D.*, Lisa E. Diamond, Ph.D.*, John S. Logan, Ph.D.* and David H. Adams, M.D.* Boston, Massachusetts and Princeton, New Jersey Sponsored by: Lawrence H. Cohn, M.D., Boston, Massachusetts
Unmodified vascularized porcine organs undergo destruction of tissue architecture and organ function by hyperacute rejection within minutes to hours of transplantation into baboon recipients. Recipient complement activation and deposition onto donor endothelium is pivotal in the pathogenesis of hyperacute rejection. Decay accelerating factor (DAF), membrane cofactor protein (MCP), and CD59 are important human complement regulatory proteins (CRP). In order to clarify the role donor CRP play in pig to baboon hyperacute rejection, transgenic swine expressing different human CRP combinations were used as donors. One heart expressing MCP, two hearts expressing CD59/DAF, and two kidneys expressing CD59/DAF were transplanted into five baboon recipients. Cardiac grafts were placed heterotopically in the neck and followed by periodic inspection, echocardiography, and biopsy. Renal grafts were placed heterotopically in the abdomen following bilateral recipient nephrectomies. Recipient blood was tested serially for trough Cyclosporine A levels, CBCs, creatinine, and serum xenoreactive antibody liters. Xenograft samples were examined by light microscopy and immunostained for detection of antibody deposition and complement activation.
Transgenic organs were protected against hyperacute rejection. The MCP heart was explanted with normal function after 1.8 days following sudden death of the baboon. The CD59/DAF hearts survived 3.6 and 5.4 days. One CD59/DAF kidney was explanted after 8.3 days with normal function (creatinine 2.1) from a terminally anemic recipient. The second transplanted kidney is functioning well (creatinine 1.3) after 11 days. Biopsies from functioning grafts were noted to have preserved histologic architecture. Rejected hearts demonstrated regional hemorrhage and interstitial edema, with fibrin plugs in thrombosed vessels. Immunostaining revealed specific and prominent IgM and C4 endothelial deposition and inhibition of terminal membrane attack complex formation. Kidney samples showed normal histology with positive immunostaining for IgM and C4. Our results demonstrate transgenic swine organs expressing human CRP are protected from hyperacute rejection. Further understanding of ongoing immune events resulting in graft failure is required before pig xenotransplantation can be applied clinically.
*By invitation F21. EFFICIENCY OF A HIGH-TITRE RETROVIRAL VECTOR IN GENE TRANSFER INTO SKELETAL MYOBLASTS. Reida M. El Oakley, F.R.C.S., M.D.*, Mark A. Poznansky, M.R.C.P., Ph.D.*, Madeliene C. McMullen*, Gregor M. Adams, B.Sc.*, Nigel J. Brand, Ph.D.*, Paul J.R. Barton, Ph.D.* and Magdi H. Yacoub, F.R.C.S., Ph.D. London, England
Background: Grafting genetically modified skeletal myoblasts for myocardial repair is dependent on an efficient gene transfer system that integrates the gene(s) of interest into the chromosome of the target cell and its progeny. The aim of this investigation is to evaluate the use of a new retroviral-based gene transfer system for this purpose.
Methods & Results: The retroviral vector MFG, carrying the β galactosidase gene (lacZ) with nuclear localisation signal, was used to transduce the skeletal myoblast cell line L6. The MFG-lacZ construct was packaged in a fourth generation, high-litre, split-genome packaging cell line (FLYA4). This cell line produced 105-106 infectious units per ml. L6 cells were cultured in tissue culture flasks and transduced with MFG-lacZ using filtered supernatant from the packaging cells. Transduced L6 cells were divided into 4 groups. Group I, cells were fixed as myoblasts 3 days after transduction. Group II, cells were allowed to differentiate into myotubes. Group III, cells were split every 3 days for 4 months. Group IV, cells were split as in group III, and allowed to differentiate into myotubes. For each group, un-transduced L6 cells acted as "control". All samples were fixed with 3.7% formaldehyde and stained for lacZ activity. The percentage of cells with successful transgene expression are presented in the following table:
Conclusion: Recombinant MFG retroviral agents packaged in a high-litre, split-genome packaging cells are efficient in gene-transfer into skeletal myoblasts and result in stable transgene expression even after repealed cell division and/or differentiation.
*By invitation F22. EX-VIVO ADENOVIRAL-MEDIATED GENE TRANSFER TO THE TRANSPLANTED ADULT RAT HEART. Alan P. Kypson, M.D.*, Karsten Peppel, Ph.D.*, Shahab A. Akhter, M.D.*, R. Eric Lilly, M.D.*, Donald D. Glower, M.D.*, Robert J. Lefkowitz, M.D.* and Walter J. Koch, Ph.D.* Durham, North Carolina Sponsored by: Robert W. Anderson, M.D., Durham, North Carolina
The ability to transfer functional genes to adult myocardium represents an area of study with potentially significant therapeutic implications. We investigated the feasibility of adenoviral-mediated transfer of both marker genes LacZ and Luciferase, as well as the potentially functional gene of the human P2 adrenergic receptor (P2-AR) in a heterotopic heart transplant model using adult male Long Evans rats. Donor hearts were arrested with warm cardioplegia and after removal flushed via the aortic root with one milliliter of solution containing 1012 total viral particles of recombinant adenovirus encoding one of the three transgenes. Hearts were transplanted into the recipient's abdomen and harvested at five days for hearts injected with the marker genes and at various time points for hearts injected with the β2-AR. lacZ-treatcd hearts were assessed by histochemical staining (X-gal). Luciferase-treated hearts were assayed for luciferase activity. β2-AR-treated hearts underwent radioligand binding assays and immunohistochemistry using an antibody specific for the human P2-AR. LacZ hearts (n = 6) revealed diffuse myocyte staining of both the right and left ventricles, as opposed to no staining within control hearts which received empty adenovirus. Luciferase hearts (n = 6) demonstrated a mean activity of 970, 000 ± 220, 000* arbitrary luciferase units as compared to controls which had a mean activity of 500 ± 200 arbitrary luciferase units (*p<0.05 vs. controls). Total P-AR densities (fmol/mg membrane protein) for hearts that received the β2-AR transgene, at 3, 5, 7, 10, and 14 days after transfection were as follows; right ventricle - 488.5 ± 126.8, 519.4 ± 81.8*, 477.1 ± 51.8*, 183.0 ± 6.5*, 82.7 ± 19.1; left ventricle - 511.0 ± 167.6, 1206.4 ± 321.8*, 525.3 ± 188.7, 183.5 ± 18.6*, 75.9 ± 15.2 (n = 3 for each group; *p<0.05 as compared to control value of 75.6 ± 6.4). Immunohistochemical analysis with anti-β2-AR antibodies revealed diffuse staining of varying intensity within myocardial sarcolemmal membranes. We conclude that global overexpression of two different adenoviral-mediated reporter genes and, a potentially functional gene, the human β2-AR, is possible during cardiac transplantation. Furthermore, β2-AR overexpression increases with time, peaking at five days, followed by a gradual decline returning to native levels at two weeks. Ultimately, gene transfer during cardiac transplantation may provide a unique opportunity for genetic manipulation of the donor organ, potentially enhancing the function of the heart.
*By invitation F23. ESTROGEN INHIBITS THE DEVELOPMENT OF TRANSPLANT ARTERIOSCLEROSIS BY PREVENTING INDUCIBLE MHC CLASS II ANTIGEN IN THE EARLY PHASE FOLLOWING THE TRANSPLANTATION. Satoshi Saito, M.D.*, Noboru Motomura, M.D., Ph.D.*, Hong Lou, M.D.* and Marie L. Foegh, M.D., D.Sc.* Washington, DC Sponsored by: Edward A. Lefrak, M.D., Annandale, Virginia
Background: The development of transplant arteriosclerosis (TA) is a major limiting factor for long time survival of cardiac transplants. We find chronic estradiol (E2) treatment inhibited TA. Very recently, we discovered that E2 inhibition of coronary TA in a chronic cardiac transplant model is associated with complete abolition of MHC class II expression. This model employs cyclosporin A (10 mg/kg/day) immunosuppression. We hypothesize that E2 independent of cyclosporin A inhibits inducible MHC class II antigen expression in professional and non-professional antigen presenting cells from the early phase following transplantation. The objective of this study is to investigate in noncyclosporin requiring TA model that E2 treatment abolish MHC class II antigen expression in the allograft in the early phase following transplantation.
Methods: Orthotopic abdominal aorta allograft transplantation was performed using Brown-Norway rats as donors and Lewis rat as recipients . All recipients were treated with either 20µg/kg/day of estradiol 17 (J (n = 20) or placebo (n = 20) continuously s.c. from 2 days prior to transplantation until sacrifice using an osmotic minipump. The animals were sacrificed on postoperative days 1, 3, 7, and 14 and the grafts were harvested following perfusion fixation and then embedding in paraffin. Cross sections of the allografts were used for computerized morphometric analysis of medial area (M) and intimal thickening. Intimal thickening (I/I + M) was quantitated as the ratio of intimal area (I) over total vascular area (intima + media). Following immunohisto-staining, the expression of MHC class II antigen and macrophage was graded semiquantitatively on a scale from 0 to +3.
Results: Intimal thickening was measureable at day 14 and I/I + M in the allograft from E2 treated recipients was significantly lower than placebo treated recipients. (9.2 + 2.2 % vs. 2.3 + 3.6%, p<0.01).
Summary: 1. Chronic estrogen treatment inhibits development of transplant arteriosclerosis.
2. Estrogen abolishes inducible MHC class II antigen expression on SMC in the media in the early phase following the transplantation.
3. Estrogen suppresses inducible macrophag MHC class II antigen expression in the allograft vessel wall.
Conclusion: Estrogen may inhibit the development of transplant arteriosclerosis by abolition of inducible MHC class II antigen in the early phase following the transplantation.
*By invitation F24. AORTIC VALVE GRAFTS IN THE RAT MODEL: EVIDENCE FOR REJECTION. Ahmad Moustapha, M.D.*, David B. Ross, M.D.*, Bindu Bittira, B.Sc.*, Dick Van Velzen, Ph.D.*, Vivian C. McAlister, M.B.*, Christopher L. Lannon, B.Sc.* and Timothy D. Lee, Ph.D.* Halifax, Nova Scotia, Canada Sponsored by: David A. Murphy, M.D., Halifax, Nova Scotia, Canada
Background: The role of immune mediated rejection in the failure of allograft heart valves in man is uncertain. Serial sampling of human allografts is not feasible and grafts are seldom removed shortly after implantation when any such damage may be occurring. The use of heart valve transplants between syngeneic and allogeneic strains of rats permits investigations into the role of immune mediated rejection of these grafts.
Aims: To describe the pathological changes in rats, over time, following transplantation of allogeneic versus syngeneic aortic valve grafts and to identify changes indicative of cell mediated rejection.
Methods: Transplantations were performed placing the donor ascending aorta, valve and rim of subvalvular myocardium in the abdominal aorta of the recipient. Recipients were Lewis (Le) rats; syngeneic donors were Lewis and allogeneic donors were Brown Norway (BN). At sacrifice, the grafts were removed, fixed and serially transacted at right angles. Samples of the aortic graft valve ring and a sample of the distal graft aorta, 2 mm above the valve, were routinely processed in paraffin for microscopical assessment using 4 micron sections, H/E and van Giesson/Elastica stains. Histopathological assessment was performed by two pathologists blinded to the original protocol and study design.
Results:1) Early Post Transplant: 9 allogeneic (BN to Le) were transplanted with 3 sacrificed at 2, 5 and 7 days with 3 Le transplanted as syngeneic controls. In both syn- and allogeneic cases a progressive, massive infiltration of neutrophils at the valve base and proximal aorta occurred increasing from day 2 to day 7. There was no difference seen between syngeneic and allogeneic animals. 2)4 Weeks Post Transplant: 8 syngeneic (Le to Le) and 8 allogeneic (BN to Le) were performed. After initial blinded review, only two types of morphology were found. In one, virtually normal aortic grafts were seen, with minimal reactive changes to valves and minimal irregularity of the endothelium. In the other, severe chronic peri-aortic inflammation was associated with continuous cusp obliterating thrombosis, secondary valve degeneration and multi-focal, often completely full thickness, medial cell necrosis. All cases were attributed to either of the two types, reproducibility was complete. After unblinding, normal morphology was seen to be associated with the syngeneic transplants, chronic inflammation and destructive thrombosis was associated with allogeneic transplants. 3) 8 Weeks Post Transplant, Effects of Cryopreservation: 12 Le to Le and 12 BN to Le were transplanted with half of the donors cryopreserved using standard tissue bank techniques. Syngeneic transplants showed intact valves with no signs of attachment to aortic wall, degeneration or calcification. Only one (1/12) showed any thrombus in a sinus; the elastin layers were preserved and there was no perivascular inflammation. The allogeneic grafts showed complete obliteration of the sinuses of Valsalva by organized thrombus with early calcification in some. Valve tissue remains were seen in the organised thrombus. Development of a neo-intima with medial cell necrosis was present. A moderate, mainly lymphocytic reactive infiltrate was present in the perivascular tissues (11/12) and occasionally in the confines of the aorta (3/12). The pathology did not differ between the cryopreserved and fresh grafts. Conclusions: The pathology of aortic valve transplantation in the rat model is dominated in the first week by a massive infiltration of neutrophils, probably induced by the necrotic muscle transplanted and is similar for both allogeneic and syngeneic grafts. Thereafter, syngeneic grafts remain intact while allogeneic grafts show progressive obliterative thrombosis of the valve leaflets and lymphocytic infiltration of the aorta, unaffected by cryopreservation. These changes are characteristic of cell-mediated rejection. Investigations into therapies to modify this response appear justified.
*By invitation F25. THE RELATIONSHIP BETWEEN CALCIUM AND MAGNESIUM IN PEDIATRIC MYOCARDIAL PROTECTION. Bradley S. Allen, M.D.*, Michael Kronen, M.D.*, Kirk S. Boiling, M.D.*, Shaikh Ramon, M.S.*, Tingrong Wang, M.D.* and Harold Feinberg, Ph.D.* Chicago, Illinois Sponsored by: Renee S. Hartz, M.D., Chicago, Illinois
We have shown that adding magnesium to normocalcemic cardioplegic solutions offsets the detrimental effects of calcium in neonatal hearts by competing with calcium entry. However, it is not known whether magnesium offers any benefit when added to low calcium cardioplegic solutions. Fifteen 5-18 day old neonatal piglets underwent sixty minutes of ventilator hypoxia (FiO2 8-10%) followed by reoxygenation using cardiopulmonary bypass (FiO2 100%) for 5 minutes, and 20 minutes of normothermic ischemia by cross clamping the aorta. This produces a severe injury that combines ischemia with hypoxia and reoxygenation. The hearts were then protected for seventy minutes with hypocalcemic (Ca2+ 0.2-0.4 mM/1) multidose blood cardioplegia. In five (Group 1) no magnesium was added to the blood cardioplegia, in 5 (Group 2) magnesium was added to the cardioplegia to produce a concentration of 5 meq/1, and in the last 5 (Group 3) magnesium was added at a concentration of 10 meq/1. Function was assessed using pressure volume loops and expressed as percentage of control. Coronary vascular resistance (CVR) was measured during each cardioplegic infusion. Despite the use of a hypocalcemic cardioplegia solution, in the absence of magnesium supplementation (Group 1) there was marked post bypass depression of systolic (Ees 38 + 1%) and global myocardial function (40 ± 1%), and a marked rise in diastolic stiffness (238 ± 3%). Conversely, even low dose (5 meq/l, Group 2) magnesium supplemented cardioplegia resulted in complete return of systolic (101% vs 38%)* and global myocardial function (102% vs 40%)*, preserved diastolic compliance (154% vs 238%)*, reduced myocardial edema (79.7% vs 80.6%), maintained ATP levels (15.8 vs 12.2 ug/gm dry wt)* and preserved CVR*, compared to a hypocalcemic cardioplegic solution without magnesium (Group 1). The use of a higher dose of magnesium (10 meq/1) did not result in any further improvement. In conclusion: 1) there is complete functional preservation, even in severely stressed neonatal hearts, when cardioplegia solutions are supplemented with magnesium, 2) this occurs despite even when a hypocalcemic cardioplegic solution is used, and 3) doses as low as 5 meq/1 of magnesium are effective. This study therefore strongly supports adding magnesium to all blood cardioplegia solutions regardless of calcium concentration. *p<0.05
*By invitation F26. EFFECT OF THE COX-MAZE PROCEDURE ON THE SECRETION OF ATRIAL NATRIURETIC PEPTIDE. Ki-Bong Kim, M.D.*, Chang-Ha Lee, M.D.*, Young-Joo Cha, M.D.* and Cheol-Ho Kim, M.D.* Seoul, Korea Sponsored by: James L. Cox, M.D., St. Louis, Missouri
The Cox-Maze procedure (CMP) has been confirmed to be effective in curing atrial fibrillation (AF). Some authors reported severe fluid retention after CMP and suggested decreased secretion of atrial natriuretic peptide (ANP) as a possible mechanism. This study was designed 1) to follow the serial changes in ANP after CMP as compared to after coronary artery bypass grafting (CABG), and 2) to elucidate any differences between ANP levels in patients with transient recurrence of AF after CMP and those without recurrence. Blood samples were drawn from the right atrium (RA) and left atrium (LA) in patients undergoing CMP (n = 19) and from the RA in patients undergoing CABG (n = 6) before and 1, 2, and 3 days after surgery. The plasma samples were prepared by refrigerated centrifugation and stored til radioimmunoassay. In the CMP group, ANP levels in the RA were 629 ± 366, 153 ± 112, 162 ± 112, and 183 ± 97, and in the LA were 276 ± 168, 152 ± 91, 162 ± 111, and 145 ± 80 (pg/ml, mean ± SD) before and 1, 2, and 3 days after surgery, showing a marked decrease in ANP levels after CMP (p<0.01). In the CABG group, ANP levels in the RA were 115 ±37, 124 ± 48, 154 ± 54, 156 ± 36 (pg/ml mean±SD) before and 1, 2, and 3 days after surgery, showing no change after surgery. There were no differences in ANP levels between patients with transient recurrence of AF (n = 6) and those without recurrence (n = 13) after CMP. There was no significant correlation between ANP levels and LA or RA pressure after CMP, which suggests that the secretion of ANP from atria was impaired. In summary, we observed a significant decrease in ANP levels after CMP and this might be one of the possible causes of fluid retention after CMP. The decreased ANP levels after CMP may result from the multiple atriotomy incisions of the CMP rather than from the conversion of AF to sinus rhythm.
*By invitation F27. EXTENDING THE CONCEPT OF AUTOGRAFT FOR COMPLETE REPAIR OF TRANSPOSITION OF THE GREAT ARTERIES WITH VSD AND LEFT VENTRICULAR OUTFLOW TRACT OBSTRUCTION. A REPORT OF 10 CASES OF A MODIFIED PROCEDURE. Dominique R. Metras, M.D., Bernard Kreitmann, M.D.*, Alberto Riberi, M.D.*, John Yao, M.D.*, Elie El Khoury, M.D.*, François Wernert, M.D.* and Adrienne Pannetier, M.D.* Marseille, France
BACKGROUND: Although in most cases of TGA with VSD and LVOTO, a REV (Lecompte) procedure is possible without interposition of a conduit between RV and PA, the anterior location of the PAs after the Lecompte maneuver may be a potential cause for RV outflow obstruction that remains reported in 5 to 10% of cases. We have used a tubular segment of aortic autograft to connect the pulmonary artery, left in orthotopic posterior position (without Lecompte maneuver) to the RV in 10 consecutive cases of TGA VSD LVOTO. Late results up to 3, 4 years, show no obstruction in the RV reconstruction, low RV pressure, and no calcifications of the autograft.
METHODS: Ten consecutive patients aged 9 months to 11 years (mean 32 months) have been corrected with a modified REV-Lecompte operation. Eight had a severe pulmonary stenosis, 2 had a pulmonary atresia, 4 had a restrictive VSD at the time of surgery, One had multiple VSDs. Seven had undergone one (5) or two (2) previous modified Blalock-Taussig shunts. All patients underwent a total correction with LV - Ao intra-ventricular connection (4 needed a VSD enlargement), connection between RV and PAs with a tubular segment of autograft aorta, without Lecompte maneuver (anterior location of the bifurcation of Pas) on the right (5) or the left (4) of the aorta. No valvular device was used for the RV outflow repair.
RESULTS: There was no early or late death. One patient with multiple VSDs needed an early (one month) reoperation for a residual muscular VSD. All patients are currently in NYHA class I, without medications, in sinus rhythm, at a mean follow-up of 2 years. There is no calcification on the chest X-ray, and at the most recent echocardiogram, RV pressures were low (25-40, mean 33 mmHg) and no significant gradient (over 10 mmHg) was found between RV and PA. Left and right ventricular functions were satisfactory.
CONCLUSION: This modification of the REV operation using a segment of autograft allows an excellent early and late result, with no danger of compression of anteriorly placed PAs, no significant RV outflow obstruction, normal aspect of the tubular autograft. In view of laboratory and clinical evidence, normal growth of the autograft can be anticipated. It allows an elective correction of TGA VSD LVOTO without previous BT shunt (3 cases) and correction at a young age (three patients less than one year).
*By invitation F28. FIRST EXPERIENCE WITH A MODIFIED REPAIR TECHNIQUE FOR TRICUSPID INCOMPETENCE IN EBSTEIN'S MALFORMATION. Roland Hetzer, M.D., Ph.D., Nicole Nagdymann, M.D.*, Peter Ewert, M.D.*, Vladimir Alexi-Meskhisvili, M.D, Ph.D.*, Yu-Guo Weng, M.D.* Felix Berger, M.D.* and Peter E. Lange, M.D, Ph.D.* Berlin, Germany
Tricuspid incompetence (TI) in Ebstein's malformation is preferably treated by tricuspid repair (TR) rather than replacement (TVR) because of the characteristic complications of valve prosthesis and the higher incidence of heart block after TVR. Most repair techniques involve plication of the "atrialized" chamber which in some cases of small functional right ventricle may be difficult to achieve or, if attempted, may jeopardize the function of both ventricles.
In October 1988 a concept was instituted to restore tricuspid competence by a technique which leaves the atrialized chamber unplicated, reduces tricuspid orifice circumference at the level of the true annulus and uses the individually most mobile leaflet for closing mechanism. Following this concept among 19 highly symptomatic patients, ages 2 to 52 years (mean 21 years) the following procedures were performed: creation of a double orifice tricuspid valve in 1, closure of posterior part of tricuspid orifice in 10, closure of anterior part in 5 and bilateral annulus plication in 3 cases.
There was no early death. Late death occurred in one patient with recurrent sepsis, after operation in active endocarditis. Intraoperative echocardiography revealed residual TI of grade 0-1 in 11, grade I-II in 8 patients. TI progressed in 2 patients and required repeat TR after 19 and 24 months. At a mean follow-up of 27 months (6 to 96 months) clinical status remains improved to NYHA I in 3, II in 13 and III in 2 patients. There were 2 cases each of transient and permanent heart block. The atrialized chamber has not gained size in any case so far.
It is concluded that the proposed repair principle may offer an acceptable alternative to other repair techniques for TI in Ebstein's malformation.
*By invitation 9:30 a.m. SIMULTANEOUS SCIENTIFIC SESSION D - ADULT CARDIAC SURGERY South Sheraton Ballroom
Moderators: Robert L. Hardesty, M.D. Fred A. Crawford, Jr., M.D.
40. CORONARY ARTERY BYPASS WITHOUT ROUTINE PULMONARY ARTERY CATHETER USE - ASSESSMENT OF CRITERIA TO REDUCE COST AND PRESERVE QUALITY. Robert D. Stewart, M.D.*, Christian T. Campos, . M.D.*, Triffin Psyhojos, M.D.*, Stephen J. Lahey, M.D.* and Sidney Levitsky, M.D. Boston, Massachusetts Discussant: Jack M. Matloff, M.D.
To reduce the cost and resource utilization of coronary artery bypass surgery (CABG) at our institution, criteria to limit pulmonary artery (PA) catheter use were developed. Patients undergoing isolated, non-emergent, primary CABG with a left ventricular ejection fraction ≥ 40%, a creatinine < 2.0 mg/dl, without documented pulmonary hypertension or severe chronic obstructive pulmonary disease, and without unstable angina within 24 hours of surgery requiring the institution of intravenous (IV) heparin or nitroglycerin (TNG) or an increase in the dose of IV TNG or the placement of an intraaortic balloon pump were eligible for CABG with only central venous pressure (CVP) monitoring. The impact of these criteria from their implementation on April 22, 1996 until July 31, 1996 in 77 patients (CVP group) were compared with results in 36 patients who met CVP criteria but had a PA catheter placed due to surgeon or anesthesiologist preference (PA group). These 113 patients represented 65% of the 175 isolated, primary CABG operations performed at our institution during this three-month period.
The CVP and PA patients were well matched in demographics (age, gender, body mass index), severity of disease (ejection fraction, coronary vessel disease extent, priority of operation, and pre-op heparin or TNG use), and the prevalence of co-morbid conditions known to influence CABG outcome including hypertension, peripheral vascular disease, previous MI, and diabetes. Mean graft number, internal thoracic artery use, total cardiopulmonary bypass time, and cross clamp time were identical. Significantly more PA patients left the OR with inotropic support (14% vs. 1%; p=0.01), however, no significant increases in in-hospital mortality, mediastinitis, stroke, re-operation for bleeding, need for re-intubation, renal insufficiency, or need to institute post-op inotropic support were seen with CVP use. One patient had a CVP changed to a PA catheter.
A significant reduction in the volume infused in the first 12 hours following operation was observed in the CVP group. Definite trends toward reductions in all other resource outcome indicators were seen in the CVP group as well. These clinically important trends approached statistical significance (Type II error).
We conclude that these criteria for CVP catheter use, applicable to the majority of patients undergoing isolated primary CABG, resulted in clinically significant reductions in all measures of CABG resource utilization with no increase in morbidity or in-hospital mortality.
*By invitation 41. A "PRIMELESS PUMP" FOR CARDIOPULMONARY BYPASS ENHANCES BLOOD CONSERVATION BY REDUCING THE NEED FOR PERIOPERATIVE BLOOD TRANSFUSIONS IN CORONARY BYPASS OPERATIONS. John A. Rousou, M.D., Richard M. Engelman, M.D., Joseph E. Flack, III, M.D.*, David W. Deaton, M.D.*, Jane L. Garb, M.S.* and Susannah G. Owen, B.A.* Springfield, Massachusetts Discussant: Kenneth M. Taylor, M.D.
Severe hemodilution during cardiopulmonary bypass (CPB) often leads to significant drop in hematocrit (Hct) and coagulation factor (CF) levels requiring blood product transfusions (Tx). A method of removing pump prime prior to CPB, initiated and clinically used at our institution, was noted to limit hemodilution and reduce the need for perioperative Tx. A prospective evaluation of two consecutive series of patients, 52 with prime (control) and 95 without prime (primeless) undergoing coronary bypass operation (CABG) was undertaken to objectively study the method's effectiveness in reducing Tx. Baseline characteristics between the control and the primeless groups such as body surface area (BSA), pre-CPB Hct, number of redos, use of hemocon-centrator and antifibrinolytics, fluid balance in OR, operative and perfusion technique and transfusion criteria were the same in both series and strictly adhered to in all patients as per protocol. Patients within each group were consecutive and non-selected, regardless of their BSA, Hct or any other factors. Group differences in Hct during operation were analyzed using repeated measures analysis of covariance. The proportion of patients requiring transfusions of red blood cells (RBCs) and/or CF were compared using multiple logistic regression controlling for other contributing factors. The drop in Hct during operation (over time) was significantly less for the "primeless" group (p<<0.0001) as shown in table below. Also shown are group differences in RBCs and CF given. There was no mortality in either group and no significant differences in complications. Although not statistically significant in a smaller subgroup of this population, there was a trend for more pronounced effect of "primeless" pump on intraoperative Hct for patients with BSA < 2.
Conclusions: The "primeless pump" leads to: 1) Significantly higher intraoperative Hct; 2) Significantly fewer patients requiring RBCs during CABG and (possibly) CF transfusion factors postop; 3) The technique is simple and applicable to all cardiac operations leading to significant cost and risk reductions.
*By invitation 42. WHAT FACTORS TRULY AFFECT WAITING TIME TO TRANSPLANTATION? A MULTIVARIABLE ANALYSIS OF THE ISHLT/UNOS THORACIC REGISTRY. Jonathan M. Chen, M.D.*, Keith D. Aaronson, M.D.*, Alan D. Weinberg, M.S.*, Berkeley M. Keck, R.N., M.P.H..*, Leah E. Bennett, Ph.D.*, Jeffrey D. Hosenpud, M.D.* and Robert E. Michler, M.D. New York, New York, Richmond, Virginia and Milwaukee, Wisconsin Discussant: Margaret D. Allen, M.D.
Because of the present donor organ crisis, the equity of organ distribution continues to be scrutinized. Although patients are currently listed and subsequently matched for cardiac transplantation (CT) according to clinical urgency, blood type, and weight, other variables may have an important impact on overall waiting time. We applied the Cox proportional hazards model to all 7791 CT candidates listed with UNOS from 4/94 to 4/96 in an effort to assess the simultaneous effect of multiple variables on the waiting times of CT candidates. The mean waiting time to CT was 151.2 days (median time 175 days), the mean age was 45.6 years, and the mean time spent waiting as a status I candidate was 20.5 days. At the termination of the study period, 4058 (52.1%) patients had undergone transplantation, 1954 (25.1%) were still waiting, 1216 (15.6%) had died and 563 (7.2%) were removed from the list for other reasons. Variables that were not significantly associated with waiting time included: race, education level, arrhythmias, AICD and amiodarone use. Results of a multivariable analysis are depicted below:
Although preoperative predictors of patients at highest risk for death on the waiting list remain elusive, we describe herein factors that have an impact on the waiting time to transplantation. It is our hope that these data may help to identify certain patients at high risk for longer waiting times for whom other surgical alternatives to transplantation may be indicated.
*By invitation §43. VIDEO-ASSISTED MINIMALLY INVASIVE MITRAL VALVE SURGERY. W. Randolph Chitwood, Jr. M.D., Christopher L. Wixon, M.D.*, Joseph R. Elbeery, M.D.* and Jon F. Moran, M.D.* Greenville, North Carolina Discussant: Aubrey C. Galloway, Jr., M.D.
Video-assisted minimally invasive mitral valve surgery (VMIMS) may have advantages over conventional operative approaches. Since May of 1996, we have used a 2.5 inch right thoracic incision and video-assisted techniques in repairing (Rep N = 8) or replacing (Rpl N = 5) mitral valves. Ejection fractions ranged between .35 and .60 (.54 ± 2.2 SEM) with ages being 18 to 77 years (55.5 ± 5.2 SEM). Other clinical characteristics were similar. Cardiac arrest was induced either by retrograde (N = 5) or antegrade (N = 6) blood cardioplegia using a new transthoracic cross-clamp to occlude the ascending aorta. In two patients cold ventricular fibrillation was used. Systemic perfusion was maintained at 28°C either by central or peripheral arterio/venous cannulation. Newly designed instruments enabled video-assisted suture placement and knot tying. Superb illumination and visualization were provided by a 10 mm port-access 30° thoracoscopic camera. Mean cardiopulmonary perfusion times were 197 ± 7.8 SEM minutes and arrest times averaged 138 ± 12.8 minutes. Postoperative transesophageal echocardio-graphic studies showed excellent valve function with minimal insufficiency, and all patients had little postoperative pain. There were no operative deaths. One patient developed lower extremity deep venous thrombosis as the only major complication. Other patients were discharged between postoperative day three and five (4.6 ± .4 days SEM). The mean hospital stay for the previous 111 conventional mitral operations (N = 67 Rep, N = 44 Rpl) was 9.2 ± 0.9 days SEM. In the VMIMS patients atrial fibrillation (8% vs 28%), reoperation for bleeding (0% vs 5%), ICU length of stay (1.1 vs 2.0 days), and hospital charges (↓ > 30%) also were significantly less than the conventional cohort. Thus, despite long operative times, these early results are encouraging and suggest that video-assisted minimally invasive mitral valve operations are safe and may benefit patients by minimizing postoperative pain, allowing earlier discharge, and decreasing hospital expenses.
10:50 a.m. INTERMISSION
§Authors have a relationship with Scanlan International, Inc. *By invitation 11:10 a.m. SIMULTANEOUS SCIENTIFIC SESSION D - ADULT CARDIAC SURGERY South Sheraton Ballroom
Moderators: Robert L. Hardesty, M.D. Fred A. Crawford, Jr., M.D.
44. MITRAL VALVE REPAIR WITH AND WITHOUT CHORDAL REPLACEMENT WITH EXPANDED POLYTETRAFLUOROETHYLENE SUTURES -A 10-YEAR EXPERIENCE. Tirone E. David, M.D., Susan Armstrong, M.Sc.*, Zhao Sun, Ph.D.* and Ahmed Omran, M.D.* Toronto, Ontario, Canada Discussant: Lawrence H. Cohn, M.D.
Replacement of chordae tendineae with expanded polytetrafluoroethylene sutures (ePTFE) increases the probability of mitral valve (MV) repair in pts with mitral regurgitation (MR) due to myxomatous disease of the MV. This study compares the results of MV repair with and without chordal replacement with ePTFE. From 1985 to 1995, 324 consecutive pts underwent MV repair: 165 with ePTFE chordae and 159 without. There were no statistical differences between these two groups as far as age (58 ± 14), functional class, left ventricular function and the incidence of coronary artery disease (16%). The differences were in the mechanism of mitral regurgitation and in the degree of myxomatous changes in the leaflets.
There were 2 operative deaths, both in the ePTFE group, neither one related to the MV repair. Operative morbidity was similar in both groups. Pts have been followed for a mean of 35 ± 30 months. No pt was lost to follow-up. Every pt had an annual Doppler echocardiographic study. Eight pts developed severe recurrent MR and 2 other developed hemolysis and were reoperated on; 6 of them were from the ePTFE group. All reoperations occurred in the first three years of follow-up. A logistic regression analysis identified that only the combination of prolapse of both leaflets and advanced myxomatous changes was predictive of reoperation in all pts. The latest echocardiographic study showed moderate MR in 9 pts (5 pts ePTFE group), and mild or none in 284. The actuarial freedom from morbid events showed no statistical differences between the two groups.
This experience indicates that replacement of chordae tendineae with ePTFE sutures during MV repair for myxomatous disease of the MV provides as good long-term results as MV repair without ePTFE.
*By invitation 45. TEN-YEAR RESULTS OF CORONARY ARTERY BYPASS GRAFTING IN PATIENTS WITH ADVANCED LEFT VENTRICULAR DYSFUNCTION. Gregory D. Trachiotis, M.D.*, William S. Weintraub, M.D.*, Thomas S. Johnston, M.D.*, Ellis L. Jones, M.D., Robert A. Guyton, M.D. and Joseph M. Graver, M.D. Atlanta, Georgia Discussant: Lynda L. Mickleborough, M.D.
The combination of severe coronary artery disease (CAD) and advanced left ventricular dysfunction carries a poor outlook with medical therapy. Coronary artery bypass grafting (CABG) in this group has often been regarded as high risk. We reviewed patients with CAD and sequentially decreased ejection fraction (EF) compared to those with an EF>50% with 10 year follow-up to determine if CABG can provide long-term symptomatic improvement and survival in patients with severe left ventricular dysfunction (EF<25%). Between 1971-1994, 156 (1.3%) patients with an EF<25% [Group I], 588 (5%) patients with an EF=25-34% [Group II], 2438 (20.6%) patients with an EF=35-49% [Group III], and 8648 (73.1%) patients with an EF>50% [Group IV] underwent CABG. The EF was estimated from the contrast left ventriculogram (either uniplaner or biplaner). For all groups mean age was 60 ± 10 years. Groups l-III compared to Group IV had a higher percentage of patients with men (<.0004); diabetes mellitus (p<.0001); class III-IV angina (p<.0001); heart failure (p<.0001); prior MI (p<.0001); 3 vessel disease (p<.0001); and left main disease (p<.0001). Group I had the highest percentage of patients with men (88%); heart failure (34%); and left main disease (24%). The mean EF's were 19 ±4 in Group I, 29 ± 3 in Group II, 42 ± 4in Group III, and vs 64 ± 9 in Group IV. The results were as follows (<.05 significant by ANOVA for Groups I, II, or III vs IV):
Despite having a higher percentage of risk factors, poorer functional status, and more complex coronary anatomy, patients with compromised left ventricular function have comparable in-hospital outcome to patients with a normal EF. While long-term there is much higher mortality in patients with comprised LV function, over 60% of patients with EF <25% were alive at 5 years. In addition, patients after CABG with EF <25% have excellent long-term control of angina despite the lower use of IMA grafts and less complete revascularization. These results suggest that in selected patients with ischemic cardiomyopathy, CABG may preserve remaining viable myocardium, provide relief of symptoms and offer survival >60% at >5 years.
*By invitation 46. MYOCARDIAL PERFUSION AND FUNCTION FOLLOWING ADENOVIRUS-MEDIATED TRANSFER OF THE VEGF121, cDNA TO ISCHEMIC PORCINE MYOCARDIUM. Charles A. Mack, M.D.*, Shailen R. Patel, M.D.*, Eric A. Schwarz, M.D.*, Pat Zanzonico, Ph.D.*, Rebecca T. Hahn, M.D.*, Arzu Ilercil, M.D.*, Richard B. Devereux, M.D.*, Stanley J. Goldsmith, M.D.*, Timothy F. Christian, M.D.*, Timothy A. Sanborn, M.D.*, Imre Kovesdi, Ph.D.*, O. Wayne Isom, M.D., Ronald G. Crystal, M.D.* and Todd K. Rosengart, M.D.* New York, New York and Rockville, Maryland Discussant: Andrew S. Wechsler, M.D.
Background: A replication-deficient adenovirus vector expressing the cDNA for the angiogenic protein vascular endothelial growth factor121 (AdVEGF121) induces prolonged VEGF121 expression in vivo. We therefore hypothesized that direct myocardial injection of AdVEGF121 would induce collateral vessel formation and enhance myocardial perfusion and function in ischemic myocardial territories.
Methods: Yorkshire swine (28-30 kg) underwent left thoracotomy and ameroid constrictor placement on the left circumflex coronary artery. Three weeks later, AdVEGF121 or the control vector, AdNull (each 108 pfu in 100 ml) was injected in the myocardium at 10 sites in the circumflex distribution. Ischemia was assessed by echocardiography and 99mTc-sestamibi imaging during rest and rapid atrial pacing (≥ 200 beats/min) at the time of gene transfer and after 4 weeks, and collateral vessel formation was assessed by ex vivo an angiography. In a separate group of animals, VEGF121 expression was quantified in the injected myocardium and serum by ELISA following gene transfer.
Results: AdVEGF121-treated animals demonstrated significant VEGF121 expression (9 ng/mg protein) in the myocardium 3 days following vector administration. VEGF121 was undetectable in the serum of AdVEGF121 -treated animals. An improvement in ventricular perfusion and function in the circumflex territory was suggested by rest versus stress 99mTc-sestamibi scans and echocardiographic assessment of segmental wall thickening, respectively, in AdVEGF121 versus AdNull-treated animals. Angiography in the AdVEGF121-treated animals demonstrated a collateral network with apparent reconstitution of the distal circumflex artery.
Conclusions: An adenovirus vector can be used to transfer the VEGF121 cDNA to the myocardium. This strategy may be useful in inducing therapeutic angiogenesis and improving perfusion and function in the ischemic myocardium.
12:10 p.m. ADJOURN
*By invitation 9:30 a.m. SIMULTANEOUS SCIENTIFIC SESSION E - GENERAL THORACIC SURGERY North Sheraton Ballroom
Moderators: Mark K. Ferguson, M.D. Andre C.H. Duranceau, M.D.
47. ANGIOGENESIS AS A PREDICTOR OF SURVIVAL FOLLOWING SURGICAL RESECTION FOR STAGE I NON-SMALL CELL LUNG CANCER. Ignacio G. Duarte, M.D.*, Bradley L. Bufkin, M.D.*, Marion F. Pennington, M.D.*, Anthony A. Gal, M.D.*, Cynthia Cohen, M.D.*, Kamal A. Mansour, M.D. and Joseph I. Miller, M.D. Atlanta, Georgia Discussant: Valerie W. Rusch, M.D.
A subset of surgically resected Stage I non-small-cell lung cancer (NSCLC) patients will later develop metastatic disease. A histologic marker of metastatic potential and diminished survival for Stage I NSCLC may help identify this subset of patients. This study evaluates the degree of angiogenic activity as a predictor of cancer-related mortality in patients having undergone surgical resection for Stage I NSCLC. Demographic, surgical, and histopathologic data were reviewed for 107 patients with Stage I NSCLC from 1985-1990. Visual quantitation of Factor VHI-related antigen (FVIII) and CDS 1 immunostained microvessels, 0.74 mm2 area (200x magnification), in 5µ sections from paraffin blocks defined tumor angiogenesis. Mean microvessel count was 20.7 ± 11.2 for FVIII, and 29.6 ± 18.1 for CD31. Mean follow-up was 5.2 ±3.0 years (8 days - 11.06 years) and 95.3% complete. Lung cancer-related mortality was 23% at 5 years. Kaplan-Meier survival curves revealed FVIII count >20 (p=0.028) and the presence of blood vessel invasion (p=0.034) to be significant predictors of disease-related mortality. Logistic regression analysis identified FVIII quantitation > 20 as the single most significant independent correlate of lung cancer mortality (p=0.021, hazard ratio 2.64). CD31 quantitation did not predict survival in univariate and multivariate analyses and did not correlate with FVIII quantitation (Spearman's rank correlation, r = 0.20). This analysis displays a significant association between tumor neovascularization and cancer-related mortality in patients with Stage I NSCLC. FVIII microvessel quantitation, as an indicator of tumor angiogenesis and metastatic potential, may help identify a subset of patients with Stage I NSCLC who may benefit from adjuvant therapy following surgical resection.
*By invitation 48. THYMIC CARCINOMA: CURRENT STAGING DOES NOT PREDICT PROGNOSIS. David B. Blumberg, M.D.*, Juan Rosai, M.D.*, Manjit S. Bains, M.D., Robert J. Downey, M.D., Robert J. Ginsberg, M.D., Nael Martini, M.D., Patricia M. McCormack, M.D, Valerie W. Rusch, M.D. and Michael E. Burt, M.D. New York, New York Discussant: Paul A. Kirschner, M.D.
Thymic carcinomas are currently staged by Masaoka classification, a staging system for thymomas. We retrospectively evaluated surgical patients with thymic carcinoma to determine factors predicting survival.
Methods: Our computerized tumor registry yielded 118 patients with thymoma. Review of pathologic material revealed 43 cases of thymic carcinoma. Medical charts were reviewed. Follow-up was performed by physician charts and telephone. Analysis by Kaplan-Meier method and Cox proportional hazards.
Results: Between 1949 and 1993, 43 patients underwent surgery for thymic carcinoma. Overall survival was 65% at 5 years with a median survival of 6.7 years. Survival was not dependent on Masaoka stage (p=0.3). There were 3 stage I patients alive at 3.6, 4.1 and 8.2 years. Five year survivals were 58% for stage II (n = 15), 55% for stage III (n = 20) and 100% for stage IV (n = 5) patients. Five year survivals of patients with complete resection (n = 29) and patients with partial resection (n=14) were 68% and 62%, respectively, (p=0.2). Survival of completely resected patients (n = 29) was not dependent on age (p=0.1), sex (p=0.7), tumor size (p=0.7), or Masaoka stage (p=0.4). Six patients had tumors invading the innominate vessels, 4 which were low grade (well/moderately differentiated) and one indeterminate. Tumor invasion of the innominate vessels (n = 6) was associated with a worse survival (p=0.01) with only 40% alive at 2 years, compared to 75% alive at 6 years with no invasion of the innominate vessels (n = 23). Of patients with no invasion of the innominate vessels, there were 19 low grade tumors and 4 high grade tumors (poorly differentiated). Survival of patients with low grade tumors was superior (p=0.03) with 95% alive at 6 yrs. as compared to a 50% 4 yr. survival for high grade tumors. By Cox proportional hazards model, survival was predicted only by grade (p=0.07) and innominate vessel invasion (p=0.05).
Conclusions: 1. Masaoka staging does not predict prognosis of patients with thymic carcinoma. 2. Patients with low grade tumors without invasion of the innominate vessels have early stage disease and long term survival may be achieved with surgical resection. 3. Despite complete resection, patients with high grade tumors and tumors invading the innominate vessels have a poor survival and these patients should be considered as having advanced disease. 4. These results have important implications for design of future adjuvant trials of patients with thymic carcinoma.
*By invitation 49. SAFETY AND EFFICACY OF BRONCHO VASCULAR RECONSTRUCTION AFTER INDUCTION CHEMOTHERAPY FOR LUNG CANCER. Erino A. Rendina, M.D.*, Federico Venuta, M.D.*, Tiziano De Giacomo, M.D.*, Isac Flaishman, M.D.* and Costante Ricci, M.D.* Rome, Italy Sponsored by: Valerie W. Rusch, M.D., New York, New York Discussant: Jean DesLauriers, M.D.
Desmoplastic reactions secondary to induction chemotherapy and/or residual tumor can make lung resection extremely difficult. In these patients, increased postoperative complications and mortality are reported, owing also to the high incidence of pneumonectomy. Between 1990 and July 1996 we have operated on 68 patients who had received 3 cycles of cisplatin-based induction chemotherapy. In 27 of these we have performed a lobectomy (# 25) or bilobectomy (# 2) associated with reconstruction of the bronchus and/or the Pulmonary Artery (PA). In only 5 additional patients pneumonectomy had to be carried out. Before chemotherapy 14 patients were at stage HIA and 13 at stage IIIB. At thoracotomy, 1 patient had no evidence of tumor, 6 were at stage I, 13 at stage II, 6 at stage IIIA, and 1 at stage IIIB. Fourteen patients had epidermoid carcinoma and 11 had adenocarcinoma. The type of reconstruction is tabulated below.
In 26 patients resection was radical with histologically negative margins. Bronchial anastomoses were wrapped by the intercostal pedicle flap. No bronchial complications nor mortality occurred. One patient had empyema and 2 had wound infection. Mean chest tube stay was 6 days (3 to 15 days). Twelve patients had additional adjuvant therapy with no problems. After a postoperative follow-up of 5 to 73 months (mean 25 months), 14 patients are alive disease-free, 1 is alive with disease and 12 died. No local recurrence occurred. One and four-year survival is 69% and 38%. Although technically demanding, lobectomy associated with bronchovascular reconstruction is feasible with good immediate and long term results after induction chemotherapy.
*By invitation 50. PREOPERATIVE TUMOR VOLUME PREDICTS OUTCOME IN MALIGNANT PLEURAL MESOTHELIOMA. Harvey I. Pass, M.D., Karen C. Kranda, R.N.*, Seth M. Steinberg, Ph.D.*, Barbara K. Temeck, M.D.* and Irwin R. Feuerstein, M.D.* Bethesda, Maryland Discussant: Larry R. Kaiser, M.D.
The staging systems for malignant pleural mesothelioma (MPM) rely upon postoperative pathologic findings for prognostic determination. Since MPM surgical cytoreduction remains controversial, it would be desirable to predict outcomes from quantitative preoperative data. We prospectively analyzed the impact of preoperative and postresection solid tumor volumes on prognostic variables in 47 of 48 consecutively resected MPM patients. Methods: From 7/93 to 6/96, 48 MPM patients had cytoreductive debulking to 5 mm or less residual tumor via extrapleural pneumonectomy (EEP-25) or pleurectomy/ decortication (P/D-23). Three dimensional CT reconstructions of pre- and postresection solid tumor using the Voxell Q were prospectively performed. All patients received the same postoperative adjuvant therapy and were staged postoperatively according to the new International Mesothelioma Interest Group (IMIG) staging. Patients were followed by chest and abdomen CT scans every 3 months until death. Prognostic factors were examined by Cox proportional hazards model. Results: With a median potential follow-up of 23.1 months, median survival for all patients is 14.4 months (EPP-11 mos, P/D-22 mos, p2 = 0.066). Median survival for preop volume <100 cc was 22 months vs 11 months if >100cc, p2 = 0.027. Median survival for postop volume <9 cc was 25 months vs 9 months if >9 cc, p2 = 0.0002. Thirty-two of 47 (68%) had positive Nl or N2 nodes. Tumor volumes associated with negative node patients were significantly smaller (51 cc) than those with positive nodes (166cc, p2 = 0.0099). Progressively higher stage was associated with higher median preoperative volume: 1-4 cc, II-94 cc, III-143 cc, IV-505 cc, p2 = 0.0070 for I vs II vs III vs IV. Patients with preoperative sizes >52 cc had shorter progression-free intervals (8 mos) than those ≤51 cc (11 mos), p2 = 0.021. By the Cox model, male sex, preoperative platelet count >314K, preoperative volume >100cc and postresection volume >9cc were associated with decreased survival. Conclusions: Preresection tumor volume is representative of T status in MPM, and CT volumetrics can predict overall and progression-free survival, as well as postoperative IMIG stage. Large volumes are associated with nodal spread, and postresection residual tumor burden may predict outcome. Future trials should develop a uniform, simple method to quantify pretreatment MPM volume, and to verify its prognostic and therapeutic implications.
10:50 a.m. INTERMISSION
*By invitation 11:10 a.m. SIMULTANEOUS SCIENTIFIC SESSION E - GENERAL THORACIC SURGERY North Sheraton Ballroom
Moderators: Mark K. Ferguson, M.D. Andre C.H. Duranceau, M.D.
51. EFFECT OF VOLUME REDUCTION ON LUNG TRANSPLANT TIMING AND SELECTION FOR COPD. Joseph E. Bavaria, M.D.*, Alberto Pochettino, M.D.*, Robert Kotloff, M.D.*, Bruce R. Rosengard, M.D.*, Peter M. Wahl, B.A.*, Harold Palevsky, M.D.* and Larry R. Kaiser, M.D. Philadelphia, Pennsylvania Discussant: Joel D. Cooper, M.D.
Introduction: End-stage COPD has traditionally been treated with lung transplantation (LTX). For two years, our LTX program has placed patients with appropriate criteria for LTX and Volume Reduction (LVR) into a prospective management algorithm. These patients are offered the LVR option as a means to "bridge" or extend the eventual time to LTX. These data examine the results of this pilot program.
Methods: From 7/7/94 to 10/25/96, 33 patients were evaluated for LTX who also had physiological criteria for LVR (FEV1 ≤25%; RV >200%; significant V/Q heterogeneity). These patients were divided into two groups: 26 patients (Group I) underwent LVR as a "bridge" and were simultaneously listed for LTX. Seven patients (Group II), for various reasons, were offered LVR alone and not listed. All patients completed 6 weeks of pulmonary rehab, and then had baseline pulmonary function (PFTs) and Six Minute Walk (6MW) tests. LVR was performed via video thoracic in 81.8% or sternotomy in 18.2% of the patients. Patients were followed postop with repeat PFTs and 6MW at 3 month intervals.
Results: Nineteen of 26 pts (73.1%) in Group I had satisfactory clinical improvement after LVR. These 19 patients (Group I A) were subsequently delisted (Status 7). The remaining 7 patients in Group I (26.9%) had unsatisfactory results (Group IB) and one died perioperatively. Three of 6 survivors were subsequently transplanted with good outcomes. The other 3 patients are presently awaiting organs.
Interestingly, 2 of the 33 patients had A-l antitrypsin deficiency both of which had poor LVR outcome.
Conclusions: LVR in these low FEV1 patients is safe. LVR has substantially impacted the practice, timing, and selection of patients for LTX. Our wait list presently has a reduced percentage of patients with a COPD diagnosis compared to 2 years ago. Seventy-three percent of otherwise suitable LTX candidates achieved good LVR results, especially reduction in RV and were deactivated from the list. The majority of patients entering our prospective management algorithm have either significantly delayed or completely avoided LTX after LVR. Our experience suggests that LVR may be limited as a "bridge" in alpha-1 antitrypsin patients.
*By invitation 52. PLEURAL TENTING DURING UPPER LOBECTOMY DECREASES CHEST TUBE TIME AND TOTAL HOSPITALIZATION DAYS. Lary A. Robinson, M.D. and Dianne Preksto, PA-C* Tampa, Florida Discussant: Joseph I. Miller, Jr., M.D.
Purpose: A prolonged air leak following an upper lobectomy is one of the major determinants of postoperative morbidity and hospital stay. The use of stapling devices during pulmonary resections has greatly decreased air leaks, but the problem of sealing small persistent leaks to allow early chest tube removal is still present. Creation of a generous pleural tent following upper lobectomy was employed to investigate whether bringing the parietal pleura down to the lung to obliterate the usual post-op apical space would help seal the air leak and shorten chest tube time.
Methods: From August, 1994 to September, 1996, the records of 43 consecutive patients undergoing an upper lobectomy for a malignancy were reviewed. Twenty-three patients had creation of a pleural tent and 20 patients (undergoing surgery in the first year of the study period) did not. Mean patient age: tented 65.6 ± 1.7 years; non-tented 62.8 ± 3.0 years. Demographic and operative profiles of both groups were not significantly different. Patients excluded from the study were those undergoing concomitant chest wall resection (7), patients requiring post-operative mechanical ventilation (1), and those developing the alcohol withdrawal syndrome (2). All resections were performed by the same surgeon through a muscle-sparing thoracotomy and included a mediastinal lymphadenectomy. Chest tubes were removed when there was no air leak for 48 hours and when the total chest tube drainage was less than 75 ml per 8 hours.
Results: The tented patients had significantly shorter chest tube times and total hospitalizations compared to the non-tented patients, as shown below:
There were no deaths. Morbidity was minimal: purulent bronchitis (17% tented, 5% non-tented, p=0.25); wound infections 0%; empyema 0%; reoperation for bleeding 0%; cardiovascular events 0%; venous thrombosis 0%; and transfusion 0%. Time required to create the pleural tent averaged 4 minutes.
Conclusions: 1. Creation of a pleural tent at the time of upper lobectomy significantly reduces the time postoperatively that chest tubes remained in place, resulting in shorter hospital stays. 2. There was no morbidity or mortality associated with this simple, quick procedure. 3. Surgeons should consider routine creation of a pleural tent at the time of upper lobectomy.
*By invitation 53. AGGRESSIVE SURGICAL MANAGEMENT IN LOCALIZED PULMONARY MYCOTIC AND NON-MYCOTIC INFECTIONS FOR NEUTROPENIC PATIENTS WITH ACUTE LEUKEMIA: REPORT OF 18 CASES. Olivier Baron, M.D.*, Betty Guillaume, M.D.*, Philippe Despins, M.D.*, Patrick Germaud, M.D.*, Philippe Moreau, M.D.*, Anne-Yvonne De Lajartre, M.D.* and Jean Luc Michaud, M.D.* Nantes, France Sponsored by: Williard A. Fry, M.D., Evanston, Illinois Discussant: Marvin Pomerantz, M.D.
Patients treated by chemotherapy or bone marrow transplants for hematologic malignancies are at risk for a variety of infectious complications. During a 8-year period (1988-1996), 18 patients (10 women, 8 men; median age 47 years) were referred to our institution for the surgical management of a suspected localized invasive pulmonary aspergillosis (IPA). Only four times the association of aspergillus at the bronchoscopy and the air crescent sign at the chest CT scan was obtained. In the other cases, the diagnosis was based on clinical features, acute localized pulmonary mass at the CT scan, failure to respond to antibiotic therapy and retrieval of fungi by bronchoalveolar lavage. Five patients had haemoptysis. No patient was known to have active fungal or bacterial infection at the time chemotherapy was performed. The diagnosis of IPA was suspected 28 ± 6 days after the beginning of the chemotherapy. Seventeen patients had an antifungal medical treatment before surgery for 32 ± 6 days. The infection was localized in the upper lobe (n = 15), in the lower lobe (n = 5) and the middle lobe (n = 3) (p<0.001). Operative procedures included one pneumonectomy, four bilobectomies, seven lobectomies, six wedge resections and one lobectomy with wedge resection (one patient had two procedures). Twice surgery was performed urgently because the mass was located close to the main pulmonary artery. Sixteen patients were treated with antifungal agents after the operation. There were no perioperative deaths and no complications. The histologic examination of the resected specimens confirmed the diagnosis of IPA in 12 cases where invasion of blood vessels by the fungus leaded to ball pulmonary infarction. This infarcted piece of tissue was often separated from the surrounding lung by phagocytes. In the six other cases, the diagnosis was: one classical aspergilloma, one pneumonia, one pulmonary abcess and three pulmonary abcesses colonized with aspergillus without typical invasion of blood vessels by aspergillus that defines IPA. With univariate analysis, in the non-invasive pulmonary aspercillus group (NIPA) there were less thoracic pain (1/6) than in the IPA group (8/12) (p<0.05), a tendency to find less air crescent sign at the CT scan (1/6 in the NIPA group versus 6/12 in the IPA group) and aspergillus was more rarely retrieved by bronchoalveolar lavage (1/6 in the NIPA groupversus 7/12 in the IPA group). Sixteen patients required subsequent hematological therapies. Sixty-six percent of the patients are alive with a mean follow-up of 29.1 ± 27.8 months (range 2 to 103 months) without any statistical difference between the IPA and the NIPA group. Five patients died with a recurrence of their malignancy at a mean of 17.2 ± 12.5 months (range 2 to 30 months) and one had a cerebral recurrence of aspergillus infection during a bone marrow transplantation three months later. Those good results of operation may be attributed to the relative young age of patients, their good pulmonary function and the brief evolution of the disease before surgery that allows limited operation. Those results encourage an aggressive policy in the management of resistant to medical treatment localized infectious pulmonary mass to prevent life threatening haemoptysis and to allow patients to proceed with further chemotherapy and bone marrow transplantation.
12:10 p.m. ADJOURN
*By invitation 9:30 a.m. SIMULTANEOUS SCIENTIFIC SESSION F - CONGENITAL HEART DISEASE Washington Ballroom
Moderators: Frank L. Hanley, M.D. John A. Waldhausen, M.D.
54. DILUTIONAL AND MODIFIED ULTRAFILTRATION REDUCES PULMONARY HYPERTENSION AFTER OPERATIONS FOR CONGENITAL HEART DISEASE: A PROSPECTIVE RANDOMIZED STUDY. Ko Bando, M.D.*, Palaniswamy Vijay, Ph.D.*, Mark W. Turrentine, M.D.*, Scott Purvines, B.S.*, Brian J. LaLone, C.C.P.*, Thomas G. Sharp, M.D.*, Yasuo Sekine, M.D.*, Lynn Means, M.D.*, Eri Sekine, B.S., MPH*, John W. Brown, M.D. Indianapolis, Indiana Discussant: Ross Ungerleider, M.D.
Background and Purpose: Pulmonary hypertension (PH) is an important cause of morbidity and mortality after congenital heart surgery (CHS). Studies have shown that a potent endothelium-derived vasoconstrictor, endothelin-1 (ET-1) may initiate the development of PH after CHS. This prospective, randomized clinical study tested the hypothesis that removal of plasma ET-1 using ultrafiltration techniques will reduce PH after CHS with cardio-pulmonary bypass (CPB).
Method: Twenty-four patients with pre-op PH (systolic pulmonary arterial pressure/systemic pressure ratio: Pp/Ps > 0.6) undergoing CHS with CPB were randomized into 2 groups: a control group (n = 12) who had conventional ultrafiltration and an experimental group (n = 12) who underwent dilutional ultrafiltration (DUF) during CPB and modified ultrafiltration (MUF) after CPB (DUF/MUFgroup). DUF was designed to actively reduce liters of ET-1 during CPB. Veno-venous MUF was performed to further minimize ET-1 and remove excess fluid after CPB. Plasma ET-1, nitric oxide metabolites (NO) and cyclic GMP (c-GMP) levels were determined: immediately before CPB, 10 minutes into CPB, immediately after CPB, and 0, 3, 6 and 12 hours after surgery in both groups and immediately after MUF and in the ultrafiltrates for the DUF/MUF group. Both groups received prophylactic a-blockers (Chlorpromazine and/or Prazosin) after CPB based on the same protocol. Perioperative changes in hemodynamics, ET-1, NO, and c-GMP levels as well as the incidence of pulmonary hypertensive crisis (PHC) and the duration of ventilatory support were compared between the groups.
Results: DUF and MUF significantly removed plasma ET-1 (1.81 ± 0.86 pg/ml in the DUF ultrafiltrate, 6.44 ± 1.82 pg/ml in the MUF ultrafiltrate). Post-op plasma ET-1 levels and Pp/Ps ratio were significantly lower in the DUF/MUF groups compared to controls. NO and c-GMP increased in both groups up to 12 hrs post-op, with no significant differences between the groups. Three of 12 controls (25%), but none of the DUF/MUF patients had PHC after CPB (p=0.07). Patients treated with DUF/MUF required significantly shorter durations of ventilatory support (68 ± 47 hr vs 178 ±139 hr for controls, p=0.048).
Conclusions: Higher levels of ET-1 may predispose patients to PH after CHS. DUF/MUF reduces plasma ET-1 and Pp/Ps after CPB and thus may represent an important adjunct for prevention of PH early after operations for congenital heart disease in high risk patients.
1991-92 AATS Graham Fellow *By invitation 55. BENEFIT OF NEUROMONITORING FOR PEDIATRIC CARDIAC SURGERY. Erie H. Austin, III, M.D., Harvey L. Edmonds, Ph.D.*, Vedad Seremet, M.D.*, Gregg Niznik, M.S.*, Aida Sehic, M.D.*, Steve M. Audenaert, M.D.* and Michael K. Sowell, M.D.* Louisville, Kentucky Discussant: Richard A. Jonas, M.D.
Purpose: The incidence of neurologic sequelae after PCS may reach 25% (Ferry PC, American Journal Diseases of Childhood 1990; 144:369-73). Therefore, we prospectively examined the potential benefit of interventions based on intraoperative neuromonitoring in decreasing both neurologic events and length of stay as a cost proxy.
Methods: With IRB-approved informed parental consent, 232 PCS patients received intraoperative neuromonitoring which consisted of 4-channel quantitative EEG/evoked potentials (EP), transcranial Doppler (TCD) ultrasonic measurement of middle cerebral artery blood flow velocity, and transcranial near-infrared spectroscopic determination of frontal lobe cerebral venous oxygen saturation (CVOS). Surgeon and anesthesiologist were notified if there were signs of seizure activity, a near-loss of EEG/EP or TCD signal, or a >25% CVOS decline from the prebypass baseline. Monitoring-based interventions consisted of 1) perfusion cannula or clamp repositioning, 2) arterial blood pressure increase, 3) cooling or anesthetic-induced decrease in brain metabolism, 4) resumption of cardiopulmonary bypass, 5) correction of perfusion system malfunction and/or 6) neuroprotection with dexamethasone and phenytoin.
Results: During the first year, 155 patients were monitored. Noteworthy changes in brain function were observed in 64/155 (41%) cases. The changes included two patients with a sudden total loss of the TCD signal and 24 cases with very low (<10cm/s) flow velocities despite normal systemic hemodynamics and oxygenation. Interventions were deemed appropriate in 39/64 (61%) cases. Repair complexity was unrelated to the likelihood of a monitored change or the decision to intervene. Neurologic sequelae, ranging from prolonged delerium or transient EEG-detected seizures to radiographically confirmed cerebral infarcts, occurred in 4/91 (4%) cases without noteworthy change, 4/39 (10%) cases with intervention and 17/25 (68%) without intervention (P<.001). Survivors' median length of stay was 7 days in the no change group, 6 days with intervention and 9 days without intervention. Interim analysis of these results led to more comprehensive monitoring and increased responsiveness during the second year. In the 77 cases monitored thus far in year two, the notification rate increased to 62% and interventions were made in 92% of these cases. Only 3 patients had neurologic sequelae (i.e. prolonged confusion, transient visual neglect, and choreiform movements following a 54 minute period of deep hypothermic circulatory arrest).
Conclusions: Timely detection and correction of cerebral ischemia/hypoxia through multimodality neuromonitoring appears to improve outcome and decrease the cost of PCS. Although additional studies are needed to confirm and expand these findings, the use of randomized designs incorporating an unmonitorcd control group may raise ethical questions.
*By invitation 56. CLINICAL TRIAL OF pH MANAGEMENT STRATEGY IN INFANTS: PERIOPERATIVE RESULTS. Richard A. Jonas, M.D., Adre J. du Plessis, M.D.*, David Wypij, Ph.D.*, Christine Plumb, R.N.*, David Farrell, M.A., C.C.P.*, Pedro J. del Nido, M.D.*, John E. Mayer, M.D., and Jane W. Newburger, M.D.* Boston, Massachusetts Discussant: Julie A. Swain, M.D.
In a randomized, single-center trial, we compared perioperative outcome in infants undergoing reparative open heart surgery after use of the a stat vs pH stat strategy during deep hypothermic (< 18°C) cardiopulmonary bypass. Admission criteria included (1) reparative open heart surgery; (2) age ≤9 months; (3) birth weight > 2.25 kg; and (4) absence of associated congenital or acquired extracardiac disorder.
Among the 182 study infants, diagnoses included dTGA (n = 92), TOP (n = 50), TOF/PA (n = 6), VSD (n = 20), truncus arteriosus (n = 8), CAVC (n = 4, not Downs), and TAPVR (n = 2). In total, 90 pts were assigned to a stat and 92, to pH stat, with randomization balanced within diagnosis, surgeon, and age group (< 1 mo, 1 -5 mo, and 5-9 mo). There were no differences between the a stat versus pH stat groups in the duration of circulatory arrest (22 ±16 vs. 21±17 min, mean ± S.D.) or total support time (129 + 49 vs. 124 ± 39 min). Early mortality (< 30 days) occurred in 4 infants (2%), all in the a stat group.
Perfusion strategy was not associated with differences in cardiac index measured in 123 patients at 3 hour intervals in the first 24 hours postoperatively; however, within the TGA subgroup, there was a tendency for those assigned to pH stat to have higher cardiac index at 12 (P=.14), 15 (p=.11), and 18 hours (p=.12). Also in the TGA subgroup, patients assigned to the pH stat strategy had significantly shorter duration of mechanical ventilation (P=.01) and stay in the intensive care unit (P=.01); however, there were no significant differences in these variables among patients in the other diagnostic groups. Pts assigned to a stat tended to have a greater incidence of postoperative hypocalcemia (p=.056) and coagulopathy (p=.056).
Continuous EEG was monitored during the first 48 hours postoperatively in 108 infants; ictal (seizure) activity was present in 5/51 pts (9.8%) assigned to a stat and 1/57 pts (1.8%) assigned to pH stat (p=.098). Clinical postoperative seizures were observed in 4 infants in the a stat group (4.4%) and 2 infants (2.2%) in the pH stat group (one later diagnosed with DiGeorge syndrome) (P=NS). First EEG activity tended to return sooner among infants randomized to pH stat (p=.068).
CONCLUSION: Use of the pH stat strategy in infants undergoing deep hypothermic bypass with or without circulatory arrest was associated with a tendency toward fewer EEG seizures and shorter recovery time to first EEG activity, and, in patients with TGA, shorter duration of intubation and ICU stay.
*By invitation 57. THE DAMUS PROCEDURE IN NEONATES AND INFANTS WITH SINGLE VENTRICLE, SUBAORTIC STENOSIS, AND ARCH OBSTRUCTION: REVISITED. Doff B. McElhinney, M.S.*, V. Mohan Reddy, M.D.*, Norman H. Silverman, M.D.* and Frank L. Hanley, M.D. San Francisco, California Discussant: Thomas L. Spray, M.D.
Background. The Damus procedure (DKS), originally proposed for biven-tricular repair of transposed great arteries, is now routinely used for palliation of functional univentricular heart. In the presence of significant arch obstruction, DKS is typically performed with periods of total circulatory arrest (CA), which may contribute to impaired neurological development. In addition, potential for semilunar valve insufficiency is a concern.
Methods. Since 1990, we have performed DKS in 16 infants (median age 12 d; 5 d to 7 mo) with functional single ventricle and subaortic stenosis, 10 of whom were neonates. Significant arch obstruction was present in 11 pts. Diagnoses were {S, L, L} double-inlet left ventricle (n = 10), {S, D, D} tricuspid atresia (n = 2), and other forms of hypoplastic left ventricle (n = 4). All pts were documented to have (or potential for) subaortic obstruction by either a bulboventricular foramen (BVF) to aortic valve diameter ratio of < 1, a pressure gradient across the ventricular septal defect or BVF, or a left ventricle outflow tract gradient at the subvalvar level. In 14 pts, DKS was performed as a primary palliation, at a median age of 9 days. Various techniques were used for the DKS anastomosis of the pulmonary trunk to the ascending aorta, with emphasis on avoiding any distortion of the semilunar valves. In the most recent 5 pts with significant arch obstruction, arch repair was achieved by performing an end-to-side anastomosis of the descending aorta to the ascending aorta, without CA to the upper body, by cannulating at the base of the innominate artery or the arch with an 8 Fr arterial cannula. In the first 6 pts with arch obstruction, a median of 40 minutes total CA was used. In pts without arch obstruction, whole body perfusion was maintained. In 13 pts a systemic to pulmonary artery shunt was placed, and 1 pt underwent concurrent bidirectional Glenn shunt (BGS).
Results. There were 3 early deaths (19%), all in patients with arch obstruction who underwent periods of complete CA. There were no clinically evident neurologic events. At median follow-up of 24 months (2 to 76 months), there were no late deaths or known neurologic complications. Five pts have undergone subsequent BGS and 2 have undergone Fontan completion. No pt has more than trivial aortic or pulmonic valvar regurgitation at follow-up.
Conclusion. DKS is an effective first stage palliation for pts with univentricular heart and subaortic stenosis, with or without arch obstruction. Arch repair can be achieved without CA to the brain. With proper attention to technical details, semilunar valve insufficiency can be avoided.
10:50 a.m. INTERMISSION
*By invitation 11:10 a.m. SIMULTANEOUS SCIENTIFIC SESSION F - CONGENITAL HEART DISEASE Washington Ballroom
Moderators: Frank L. Hanley, M.D. John A. Waldhausen, M.D.
58. BLOOD CARDIOPLEGIA IS NOT DEMONSTRABLY ADVANTAGEOUS OVER CRYSTALLOID CARDIOPLEGIA IN PEDIATRIC CARDIAC SURGERY. J. Nilas Young, M.D., Isaac O. Choy, M.D.*, Nolli K. Silva, M.D.* and Derek Y. Obayashi* Oakland, California Discussant: Bradley S. Allen, M.D.
The superiority of blood cardioplegia in pediatric cardiac surgery has not previously been demonstrated in a controlled clinical trial. We prospectively randomized 138 pediatric patients (median age 12 mos ± 3.9[se]) to receive either blood (4:1 dilution, KCl 15 mEq/L) or crystalloid (Plegisol®) cardioplegia during a variety of congenital heart operations (excluding atrial septal defects). Cold (4°C), antegrade multidose cardioplegia was administered in addition to topical cooling during surgery. Systemic hypothermia perfusion (30°C) was routinely utilized and total circulatory arrest was used in 40 patients (median circ. arrest time: 29.5 mini 5.1). Myocardial recovery and outcome measures were assessed by the following clinical endpoints: (1) inotropic support in the first 8h post-op, INT (scale 1-10); (2) echocardiographic assessment of ventricular function in the first 24h post-op, VF (scale 1-10); (3) overall complication rate, COMP(%); (4)ICU length of stay (days) and (5) 30-day survival, SV(%). Statistical significance of multivariate associations was evaluated using multiple logistic regression and analysis of variance to investigate which of the following clinical determinants were contributory: (1) cardioplegia, CP (blood, n = 62 vs. crystalloid, n = 76); (2) urgency of operation, URG; (3) aortic cross clamp time, X-time (mean 66.7 min ± 2.8) and (4) AGE of patient. Population data did not differ between the two cardioplegia groups (p>0.05). Results:
There were no statistically significant differences between blood and crystalloid cardioplegia with all measured endpoints. Sub-group cohort analysis of cyanotic lesions (n = 55) also showed no differences between blood and crystalloid cardioplegia. The most important clinical determinant of studied endpoints was the aortic cross clamp time (ischemic interval). Our results suggest no clear clinical advantage of blood cardioplegia during hypothermic cardioplegic arrest in pediatric congenital heart surgery.
*By invitation 59. MODIFIED ULTRAFILTRATION IMPROVES LEFT VENTRICULAR SYSTOLIC FUNCTION IN INFANTS AFTER CARDIOPULMONARY BYPASS. Michael J. Davies, F.R.C.S.*, Khan Nguyen, M.D.*, J. William Gaynor, M.D.* and Martin J. Elliott, M.D.* London, England and Philadelphia, Pennsylvania Sponsored by: Marc R. de Leval, M. D., F. R. C. S., London, England Discussant: Patricia A. Penkoske, M.D.
Cardiopulmonary bypass (CPB) in children is often associated with increased capillary permeability leading to increased total body water (TBW), tissue edema, and organ dysfunction. Modified ultrafiltration (MUF), performed after CPB, has been shown to reduce TBW and reverse hemodilution as well as increase cardiac index and systolic blood pressure. The mechanism of the improvement in hemodynamic parameters following MUF is unclear. This study was designed to test the hypothesis that the use of MUF after CPB improves left ventricular (LV) systolic function. Twenty-one infants undergoing CPB were instrumented with a LV micromanometer and ultrasonic dimension transducers to measure the LV anterior-posterior minor axis diameter. Patients were randomized to MUF (n= 11, age 226 ±355 days, weight 6.7 ± 3.1 kg) or control (n = 10, age 300 ± 240 days, weight 7.0 ± 2.5 kg) (p=NS difference between groups). LV systolic function was assessed using the slope of the preload recruitable stroke work (PRSW) index, a load-insensitive index of LV function. Myocardial cross-sectional area was measured by echocardiography. Data were acquired immediately following separation from CPB, at steady state and during transient vena caval occlusion. In the MUF patients, data acquisition was repeated after 13 ± 5 minutes of MUF. In the control patients, data acquisition was repeated after 12 + 5 minutes (p=NS). Inotropic drug support was the same at both study points. In the MUF group, the filtrate volume was 363 ± 262 ml and the hematocrit increased from 26 ± 2.7% to 37 ± 9.5% after MUF (p=0.018). In the control group, the hematocrit did not change (p=NS). Heart rate, end-diastolic dimension, and end-diastolic pressure were unchanged in both groups (p=NS). In the MUF group, mean ejection pressure increased from 58 ± 25 to 71+23 mmHg after MUF (p=0.005), but did not change in the control group (p=NS). Myocardial cross sectional area decreased from 3.72 ± 0.35 to 3.63 ± 0.36 CM2 after MUF (p=0.01), suggesting a reduction in myocardial edema. Myocardial cross sectional area remained constant in the control group (p=NS). After MUF, the slope of the PRSW index increased from 52.3 ± 52 to 74.2 ± 66 (103 erg/cm)3 (p=0.02), but did not change in the control group (p=NS). One patient from each group died in the postoperative period. Patients in the MUF group received less inotropic drug support in the first 24 hours following surgery (156.62 + 92.31 µg/kg/24hr) than patients in the control group (865.33 ± 1772.26 µg/kg/24hr, p=0.03). The use of MUF after CPB improves intrinsic LV systolic function, increases blood pressure, and decreases inotropic drug utilization in the early postoperative period.
*By invitation 60. ONE-STAGE MIDLINE UNIFOCALIZATION AND COMPLETE REPAIR IN INFANCY VERSUS MULTIPLE STAGE UNIFOCALIZATION FOLLOWED BY REPAIR FOR COMPLEX HEART DISEASE WITH MAJOR AORTOPULMONARY COLLATERALS. Christo I. Tchervenkov, M.D.*, Gary Salasidis, M.D.*, Renzo Cecere, M.D.*, Marie J. Beland, M.D.*, Luc Jutras, M.D.*, Marc Paquet, M.D.* and Anthony R.C. Dobell, M.D. Montreal, Quebec, Canada Discussant: Erie H. Austin, HI, M.D.
Patients with pulmonary atresia (PA), VSD and major aortopulmonary collaterals (MAPCA's) have traditionally required multiple unifocalization staging operations. In the few large series, complete repair was possible on only between 12-60% of all the patients. Recently, the feasability of a single stage unifocalization and repair was demonstrated by Hanley. We would like to share our recent experience with both these approaches. Since 1989, 11 consecutive patients not previously operated with complex heart disease and MAPCA's have undergone corrective surgery. The first 6 pts underwent staged unifocalizations with 5 achieving complete repair (Group I). The last 5 pts since May 1995 have undergone one stage midline unifocalization and complete repair (Group II). Four of these were infants (2 wks to 9 mos) and one was 13 years old. All pts in Group I had Tetralogy (TOP), PA whereas in Group II, three pts had TOP, PA, one had DORV, PA and one CAVC, TGA and severe PS.
Complete repair was achieved in 83% of Gr I and 100% in Gr II. There was one intraoperative death (unrecognized severe mitral stenosis) and one late death in Gr I and all 5 pts in Gr II are alive and well with a mean follow-up of 9 mos (2-17 mos).
We conclude that early intervention with both surgical approaches can lead to complete biventricular repair in almost all patients. Because the single stage midline unifocalization and repair can achieve complete repair and excellent survival in infancy with one operation, it is currently our approach of choice.
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.