WEDNESDAY MORNING, MAY 1, 1996

7:00 a.m. FORUM SESSION II - GENERAL THORACIC SURGERY

Room 6C/F, San Diego Convention Center

Moderators: Keith S. Naunheim, M.D.

Larry R. Kaiser, M.D.

F9. REGULATORY EFFECTS OF INTERLEUKIN-10 ON LUNG ISCHEMIA-REPERFUSION INJURY.

Michael J. Eppinger, M.D.*, Peter A. Ward, M.D.*, Steven F. Bolling, M.D. and G. Michael Deeb, M.D.

Ann Arbor, Michigan

Lung reperfusion injury may predispose the transplanted lung to poor function and early rejection. Interleukin-10 (IL-10), a cytokine with primarily anti-inflammatory effects, was studied to determine its effects on the development of early lung reperfusion injury.

Adult male rats underwent clamping of the left bronchus, pulmonary artery, and pulmonary vein for 90 minutes of ischemia, followed by 4 hours of reperfusion (controls, n=6). Time-matched shams underwent hilar dissection but not lung ischemia (n=4). Lung injury was measured by vascular permeability to ¹²⁵I-BSA (cpm/g lung tissue/ml blood). To evaluate the effect of exogenous IL-10, additional animals (n=4) received 10 μg IL-10 intravenously prior to ischemia. To assess the role of endogenous IL-10, animals received 200 μg of either rabbit anti-mouse IL-10 IgG or pre-immune IgG (n=5 each) prior to ischemia.

Compared to shams, controls demonstrated significantly more lung injury (permeability index 0.358±0.035 vs. 0.102±0.009, p<0.01). Animals receiving IL-10 had significantly less lung injury compared to controls (0.167±0.028, p<0.01). Animals receiving IgG against IL-10 had significantly more lung injury than animals receiving pre-immune IgG (0.411±0.056 vs 0.237±0.044, p<0.05). Alveolar macrophages from animals after 90 minutes of lung ischemia produced more TNF-a in culture than unstimulated macrophages; this production was reduced significantly by the addition of IL-10 to the culture medium. Northern blot analysis of whole lung RNA demonstrated that the reduction in TNF-a occurred at the mRNA level.

We conclude that endogenous IL-10 has a protective effect against lung reperfusion injury during this early phase; and that IL-10 administration can reduce lung reperfusion injury, at least in part through its ability to reduce production of TNF-a by alveolar macrophages.

*By invitation

F10. CHANGES IN LUNG COMPLIANCE AFTER VOLUME REDUCTION SURGERY IN A RABBIT MODEL OF BULLOUS EMPHYSEMA.

Fernando E. Kafie, M.D.*, Matthew Brenner, M.D.*, John C. Chen, M.D.*, Edward A. Stemmer, M.D., Michael Budd, B.S.* and Michael W. Berns, M.D.*

Orange, California

Purpose: Staple lung volume reduction surgery has recently been described for treatment of emphysema resulting in improvement in Forced Expiratory Volume (FEV₁). Little is known regarding physiologic mechanisms of response in surgically treated emphysema patients. We hypothesized that volume reduction surgery in animals with pulmonary bullous emphysema would result in decreased lung compliance. Reduction in compliance may decrease airway resistance and improve Forced Expiratory Volume.

Methods: Seventeen New Zealand (NZ) white rabbits were induced with emphysema according to our previously published model with sephadex beads and carrageenan. This animal model has been previously used to study surgical treatments for emphysema. Pressure-volume relationships were measured at 60, 40 and 20 cc inflation pre- and post-operatively in anesthetized animals. Thoracoscopy was performed prior to thoracotomy to document bullae formation. A mini-thoracotomy was performed on the side of bullae formation. Resection of areas with emphysema was accomplished with a standard pediatric multirow surgical stapler.

Results: Comparison of pressure-volume curves pre- and post-op demonstrate significant decrease in static compliance.

Pressure-Volume Data
Vol (cc H20)	PRE-OP (mmHg)*	POST-OP (mmHg)*	p-VALUE (chi-test)
20	6.8	6.0	<0.01
40	18.1	15.7	<0.01
60	21.9	19.4	<0.01
* average of 17 experiments

Conclusion: Lung compliance is decreased following lung volume reduction surgery in New Zealand White rabbits. This finding suggests that increased elastic recoil and airway support may contribute to the mechanism of improved function following lung volume reduction surgery.

*By invitation

F11. GENETICALLY ENGINEERED PORCINE LUNGS IN A HUMAN XENOTRANSPLANTATION MODEL.

Casey W. Daggett, M.D.*, Mark Yeatman, FRCS*, Andrew J. Lodge, M.D.*, JeffH. Lawson, M.D.*, Edward P. Chen, M.D.*, Meera Srinivasan, B.A.*, Peter Van Trigt, M.D., Gerry Byrne, Ph.D.*, John Logan, Ph.D.*, Jeff L. Platt, M.D.* and Robert D. Davis, M.D.*

Durham, North Carolina

Pulmonary xenotransplantation is currently limited by an abrupt rise in pulmonary vascular resistance, capillary leak, loss of compliance, and poor gas exchange. Complement activation is believed to be a central event in this process. The human complement regulatory proteins, decay accelerating factor (hDAF) and CD59 (hCD59), inhibit both the classical and the alternative pathways. Using an ex vivo model has made it possible to study specific aspects of acute pulmonary dysfunction in an heterologous combination. The pulmonary function of swine expressing hDAF/hCD59 (n=7) was compared to that of the lungs from farm bred animals (n=6) while the lungs were perfused with human fresh frozen plasma (FFP). Lungs from adult swine were isolated and preserved with Euro-Collins solution. The perfusate consisted of freshly thawed, heparinized, pooled O+FFP reconstituted in 40% Lactated Ringer's solution. Perfusion fluid was delivered to the pulmonary artery at 37°C via a gravity reservoir and recirculated by a roller pump. The lungs were ventilated with 60% oxygen and the tidal volume was controlled to keep the peak airway pressure between 35-40 cm H₂O. After two hours of perfusion the control lungs had lost an average of 74 ± 16% of their static pulmonary compliance versus a 6 ± 17% loss by the transgenic lungs (p<0.001). The controls had an average transalveolar capillary leak of 561.7 ml compared to 5.9 ml by transgenic lungs (p<0.05). The control lungs achieved an oxygen concentration in the perfusate of 259 ± 42 mmHg compared to 383 ± 42 mmHg in transgenic lungs (p<0.001). The pulmonary vascular resistance was 20.3 ± 12.6 mmHg/L/min in controls and 10.9 ±2.0 mmHg/L/min in transgenic lungs (p = 0.17).

In conclusion, the lungs from swine expressing hDAF/hCD59 demonstrated superior pulmonary function compared to lungs from farm bred swine when perfused with human plasma. The compliance, capillary leak, oxygenation, and pulmonary vascular resistance were all significantly improved in the transgenic lungs as compared to controls. These data indicate that complement activation is in part responsible for acute pulmonary dysfunction in xenotransplantation and that inhibiting complement function with hDAF and hCD59 can improve several aspects of pulmonary function in porcine-to-human pulmonary transplantation.

*By invitation

F12. ISOLATED LUNG PERFUSION WITH MELPHALAN FOR TREATMENT OF METASTATIC PULMONARY SARCOMA.

Sumihiko Nawata, M.D.*, Howard M. Ross, M.D.*, Nuno Abecasis, M.D.*, Komal S. Sachar, B.S.*, Huiming Cheng, M.A.* and Michael E. Hurt, M.D., Ph.D.

New York, New York

Introduction: Metastatic pulmonary sarcoma remains a significant clinical problem with systemic chemotherapy offering little hope for cure. Isolated lung perfusion (ILP) avoids systemic chemotherapeutic toxicity, but the most efficacious agent remains unknown. Melphalan (MN) is active in the treatment of extremity sarcoma via isolated limb perfusion and therefore MN activity in a pulmonary sarcoma metastases model was investigated.

Methods: Toxicity Study: Nineteen F344 rats underwent left ILP with MN at total doses of 20 mg (n=2), 5 mg (n=6), 2 mg (n=6), or buffered hespan (BH) (n=5). Rats underwent contralateral pneumonectomy on day 21 post-perfusion to evaluate left lung toxicity. Efficacy Study: On day 0, 41 F344 rats were injected with 5x10⁶ MCA sarcoma cells via the external jugular vein. On day 7, rats received either 2 mg MN i.v. (n=10), 1 mg MN i.v. (n=8), or underwent ILP with MN (2 mg) (n=12) or BH (n=11). On day 14, rats were sacrificed and lung sarcoma nodules were counted. Statistical analysis was performed with ANOVA and Student T test.

Results: Toxicity: All rats perfused with 20 mg or 5 mg of MN died perioperatively. Rats perfused with 2 mg of MN or BH survived contralateral pneumonectomy at rates of 67% and 80%, respectively. Efficacy: The number of left lung lesions decreased significantly in the animals receiving MN via ILP as compared to all the other groups (p<0.05). In addition, MN ILP resulted in significant reduction of tumor nodules in treated lung as compared to right lung (p<0.02). All rats that received MN 2 mg i.v. died within 5 days of injection.

Group		number of lesions in left lung	right lung
MN 1 mg i.v.	(n=8)	60 ± 21	66 ± 23
MNILP	(n=12)	7 ± 10	185 ± 70
BHILP	(n=ll)	84 ± 52	201 ± 51

Conclusion: Melphalan ILP is well-tolerated at a dose that leads to 100% mortality intravenously. Melphalan ILP significantly decreased the number of metastatic pulmonary nodules compared to i.v. treatment. Melphalan can be considered an effective agent for ILP of metastatic sarcoma and studies evaluating Melphalan by ILP in man are warranted.

*By invitation

F13. GENE THERAPY FOR LUNG CANCER: ENHANCEMENT OF TUMOR SUPPRESSION BY A COMBINATION OF SYSTEMIC CISPLATINUM AND ADENOVIRUS-MEDIATED P53 GENE TRANSFER.

Dao M. Nguyen, M.D., FRCSC*, Sandra A. Weihle, B.Sc.*, Patricia E. Koch, M.Sc.*, Richard J. Cristiano, Ph.D.* and Jack A. Roth, M.D.

Houston, Texas

Mutations of the p53 tumor suppressor gene occur in up to 70% of human non-small cell lung cancers. Restoration of the normal p53 function by gene replacement therapy in cancer cells with an abnormal p53 gene leads to G₁cell cycle arrest or apoptosis (programmed cell death). We observed that brief exposure of H1299 lung cancer cells (deleted p53) to low doses of cisplatinum (CDDP) prior to gene transfer resulted in a 2-fold elevation of reporter gene expression. To determine if such treatment would potentiate the tumor suppression effect of the AdV-CMV-p53 (recombinant adenovirus carrying the p53 gene driven by the cytomegalovirus [CMV] enhancer/promoter), H1299 cells were treated with CDDP (0.062 μg/ml x 24 hrs) 2 days prior to transfection with AdV-CMV-p53 at multiplicities of infection (MOI) of 1 and 5 viral particles per cell (n=6 per group). Prior exposure to CDDP resulted in a 35% (MOI = 1) to 61% (MOI = 5) enhanced inhibition of tumor cell proliferation 3 and 5 days after AdV-CMV-p53 transfection as compared to that of similarly treated cells without prior CDDP exposure. In vitro transfection of CDDP-treated cells with AdV-CMV-p53 led to earlier, higher levels of p53 gene expression as well as increased apoptosis. Subcutaneous H1299 tumors were created in irradiated nude mice. A combination of sequential intraperitoneal CDDP (5 μg/g of body weight) and injections of AdV-CMV-p53 (5x10⁹ viral particles/injection) into H1299 tumors (200 mm³) 2, 4, 6 days following CDDP administration resulted in a profound and prolonged inhibition of tumor growth of H1299 tumors in nude mice (n=5 per group). While systemic administration of CDDP had a small effect on H1299 tumor growth (3000±218 mm³) compared to saline-injected tumors (3550±240 mm³, 20 days after injection), tumors treated by a combination of CDDP and AdV-CMV-p53 were significantly smaller (1570±140 mm³) than those treated with AdV-CMV-p53 alone (3100±260 mm³, 32 days after treatment, p<0.0001). The timing of systemic CDDP administration relative to gene transfer was identified to be critical as simultaneous intraperitoneal CDDP and intratumoral AdV-CMV-p53 injections were less effective than sequential treatment (2300±196 mm³ vs 1570±140 mm³, p<0.001). A second cycle of combined CDDP and gene therapy given 10 days after completion of the first one led to further suppression of tumor growth (679±89 mm³ vs 1570±140 mm³, p<0.001). In conclusion, the combination of sequential systemic CDDP and intratumoral injection of AdV-CMV-p53 results in a superior tumor suppression effect. Inhibition of tumor growth can be maintained by repeated cycles of gene therapy. This gene therapy strategy has been incorporated into a phase I clinical trial for the treatment of lung cancer and provides the basis for the development of improved therapeutic protocols.

*By invitation

F14. THE DUAL FACES OF INHALED NITRIC OXIDE: IMPROVED LUNG PRESERVATION WITH EXOGENOUS NITRIC OXIDE GIVEN AT THE TIME OF HARVEST BUT NOT WHEN GIVEN DURING REPERFUSION.

Yoshifumi Naka, M.D., Ph.D.*, Dilip K. Roy, M.D.*, Hui Liao, M.D.*, David M. Stern, M.D.*, Arthur J. Smerling, M.D.*, Robert E. Michler, M.D., David J. Pinsky, M.D.* and †Mehmet C. Oz, M.D.*

New York, New York

Although inhaled nitric oxide (NO) lowers pulmonary vascular resistance in ARDS, its usefulness in the setting of lung transplantation remains controversial. We hypothesized that NO may have either beneficial or harmful effects depending upon the circumstances in which it is given. If NO given to the pulmonary donor raises endogenous (tissue) cGMP levels, this should benefit lung preservation by promoting vascular function, as cGMP analog supplementation is known to do. NO administered during reperfusion may rapidly combined with superoxide to become either ineffective or toxic (forming peroxynitrite and hydroxyl radical). Using an orthotopic rat left lung transplant model in which hemodynamics and functional parameters can be measured independent of the native lung [following ligation-of the right pulmonary artery (PA)], 4 experimental groups were established using male Lewis rats: (1) no supplemental gas given (No NO); (2) NO given at the time of harvest (65 ppm measured by chemiluminescence, Harvest NO); (3) supplementation of the preservation solution with a membrane permeable cGMP analog 8-Bromo-cGMP under No NO conditions (500 nM, cGMP); and (4) NO given during reperfusion (65 ppm, Reperfusion NO). For all groups, lungs were preserved for 6 hours at 4°C in Euro-Collins solution. Thirty minutes following ligation of the native PA, PA flow (ml/min), arterial oxygenation (pO₂, mmHg), graft neutrophil infiltration (myeloperoxidase activity, MPO, Δabsorbance/min at 460 nm), and recipient survival were determined.

Condition	PA flow	pO2	MPO	Survival
No NO (n=25)	3.9 ± 1.5	94 ± 13	2.8 ± 0.1	20%
Harvest NO (n=9)	14.1 ± 3.6*†	165 ± 33*†	2.3 ± 0.3*†	67%*
CGMP (n=11)	17.9 ± 4.0*†	165 ± 31*†	1.9 ± 0.1*†	73%**†
Reperfusion NO (n=11)	4.9 ± 2.2	94 ± 20	2.8 ± 0.2	27%
(Means±SEMS are shown: *=p<0.05, and **=p<0.01 vs No NO, and †=P<0.05 vs Reperfusion NO)

To explore potential mechanisms underlying these beneficial effects of Harvest NO, and knowing that NO stimulates the soluble guanylate cyclase to produce endogenous cGMP, we determined (by ELISA) that Harvest NO increases endogenous pulmonary cGMP (by 38% vs No NO, p<0.05). These data suggest that stimulating the NO/cGMP pathway (such as by Harvest NO or by supplementing the preservation solution with a cGMP analog) is beneficial. We conclude that inhaled NO can be either beneficial or neutral, depending upon the circumstances in which it is given.

†Robert E. Gross Research Scholar

*By invitation

F15. MITIGATION OF INJURY IN CANINE LUNG GRAFTS BY EXOGENOUS SURFACTANT THERAPY.

Ken E. Gehman, M.D.*, Richard J. Novick, M.D., Andrea A. Gilpin, HBSc.*, Imtiaz S. AH, M.D.*, Ruud A.W. Veldhuizen, Ph.D.*, Jenifer Duplan, AHT*, Lynn Denning, AHT*, Fred Possmayer, Ph.D.* and James F. Lewis, M.D.*

London, Ontario and Edmonton, Alberta, Canada

We have previously demonstrated alterations in endogenous surfactant after lung transplantation and improved pulmonary function after 36 hour preservation of normal canine lung grafts using donor bovine lipid extract surfactant (bLES) therapy. The objective of the current study was to determine whether exogenous bLES can mitigate the damage in lung grafts induced by high volume ventilation before procurement. Five control donor dogs were subjected to 8 hours of mechanical ventilation using a tidal volume of 45 ml/kg. This produced a significant decrease in PO₂ values (p<0.01) and significant increases in bronchoalveolar lavage (BAL) neutrophil count (p = 0.05), BAL protein concentration (p<0.01) and the ratio of poorly-functioning small surfactant aggregates (SA) to superior-functioning large aggregates (LA, p = 0.02) [see Table 1]. Animals (n=5) given instilled bLES (100 mg/kg) and subsequently ventilated with a tidal volume of 45 ml/kg demonstrated no significant change in PO₂ values over 8 hours and a decrease in BAL protein concentration (p = 0.04 versus control) and SA/LA ratio (p = 0.01 versus control).

Experimental Group	Time (hours)	PO2/FiO2 (mmHg)	BAL Neutrophil Count (x106/L)	BAL Protein (mg/kg)	BAL SA/LA
Control	0	476 ± 26	43 ± 39	0.42 ± 0.13	0.86 ± 0.1 5
	8	337 ± 27	1387 ± 502	2.54 ± 0.31	2.36 ± 0.41
Instilled bLES	0	512 ± 28	11± 6	0.31 ± 0.08	0.66 ± 0.36
	8	518 ± 23	665 ± 554	1.06 ± 0.53	0.91 ± 0.22

All 10 lung grafts were then flushed with 60 ml/kg modified Euro-Collins solution and stored for 18 hours at 4°C. Left lungs were transplanted into recipient dogs and reperfused for 6 hours. No additional bLES therapy was used. Results after 6 hours of reperfusion, including SA/LA in whole lung lavages from transplanted grafts, are shown in Table 2 (Recipient Animal data):

Experimental Group	PO2/FiO2 (mmHg)	PCO2 (mmHg)	Peak Inspired Pressure (cm H2O)	Transplanted Lung SA/LA
Control	73 ± 14	47.8 ± 1.4	34 ± 3.3	0.77 ± 0.17
Instilled bLES	307 ± 63	38.2 ± 4.1	24 ± 2.1	0.17 ± 0.04
p value	0.007	0.058	0.03	0.009

We conclude that instillation of exogenous bLES prior to 8 hours of high volume ventilation decreased protein leak, decreased surfactant SA/LA ratio and prevented the subsequent deterioration of PO₂ values in donor animals. Moreover, when these lungs were transplanted into recipients, bLES-treated grafts had superior PO₂ values, improved ventilation efficiency and a higher proportion of superior-functioning surfactant aggregate forms in the alveolar space than control grafts. bLES therapy can thus protect lung grafts from ventilation-induced injury and may offer a promising means to expand the donor pool.

*By invitation

F16. EXOGENOUS SURFACTANT TREATMENT BEFORE AND AFTER 16-HOUR ISCHEMIA IN EXPERIMENTAL LUNG TRANSPLANTATION.

Bernard Hausen, M.D.*, Wolfgang Bernhard*, Charles Hewitt, M.D., Ph.D.*, Frank Schroder*, Maike Beuke* and Hans-Joachim Schafers, M.D.*

Hannover, Germany and Camden, New Jersey

Sponsored by: Hans-George Borst, M.D., Hannover, Germany

Severe alterations in surfactant content of lung grafts occur following extended ischemia. A syngeneic, acute, in situ transplant model in the rat was used to determine the impact of exogenous surfactant treatment. Double lung blocs were flush perfused and preserved for 16 hours and then reperfused for 120 minutes. Group I received intratracheal surfactant (200 mg/kg; Curosurf) before perfusion and donor harvesting (n=6), group II after ischemia and before reperfusion (n=6). Untreated lungs served as controls (group III). Serial measurements of graft pulmonary vascular resistance (PVR), alveolar arterial oxygen difference (AADO₂), compliance and resistance were obtained. Final graft assessment included weight gain and histological analysis. Data is listed as mean ± standard error (*p<0.05 by ANOVA).

The mean survival after reperfusion in group I was 120 min. versus 113 ± 3 in group II and 117 ± 3 in group III. The weight increase was 12 ± 4% in group I, 105 ± 15% in group II and 87 ± 17% in group III.

	20 min. of reperfusion			120 min. of reperfusion
Group	I	II	III Controls	I	II	III Controls
AAD02	85 ± 16*	108 ± 18	147 ± 36	244 ± 60'	403 ± 99	487 ± 56	mmHg
PVR	104 ± 18*	411 ± 102	161 ± 46	72 ± 11*	411 ± 130	349 ± 59	mmHg/ml/min
Compliance	56 ± 5*	32 ± 4	42 ± 4	52 ± 3'	28 ± 6	28 ± 4	ml/cmH2O
Resistance	307 ± 17	695 ± 95	435 ± 45	287 ± 10*	720 ± 130	672 ± 120	cmH2O/1/sec

There was no significant difference in the histological analysis regarding interstitial and intra-alveolar edema or pulmonary hemorrhage.

Graft pretreatment before perfusion resulted in significantly improved oxygenation and compliance as well as decreased vascular resistance when compared to controls or treatment before reperfusion. It is therefore concluded that donor surfactant pretreatment is advantageous for preservation of overall graft function after 16 hours of ischemia.

*By invitation

F17. BOTH BLOOD AND CRYSTALLOID BASED EXTRACELLULAR SOLUTIONS ARE SUPERIOR TO INTRACELLULAR SOLUTIONS IN LUNG PRESERVATION.

Oliver A.R. Binns, M.D.*, Nuno F. DeLima, M.D.*, Scott A. Buchanan, M.D.*, Jeff T. Cope, M.D.*, Robert C. King, M.D.*, Chris A. Marek, B.S.*, Curtis G. Tribble, M.D. and Irving L. Kron, M.D.

Charlottesville, Virginia

Lung transplantation remains limited by donor ischemic time, inadequate graft preservation, and reperfusion injury. We evaluated the effects of an extracellular preservation solution, with or without the addition of blood, as compared to the standard intracellular solution Euro-Collins. Using an isolated, whole blood perfused/ventilated rabbit lung model, we studied three groups of animals. Lungs were flushed with either Euro-Collins (EC), low-potassium-dextran (LPD), or a 20% blood/low-potassium-dextran solution (BLPD). All lungs were harvested en bloc, stored inflated at 4°C for 18 hrs, and then reperfused at 60 ml/min with whole blood. Continuous measurements of pulmonary artery pressure (PAP), pulmonary vascular resistance (PVR), left atrial pressure, dynamic airway compliance (CPL), and weight gain were obtained. Fresh, non-recirculated venous blood was used to determine the single pass pulmonary venous-arterial O₂gradient (V-A O₂). All data are reported as means ± SEM after 30 minutes of reperfusion and analyzed by ANOVA.

Group	PAP (mmHg)	PVR (Dynes.sec.cm-5)	% change CPL	Wet/Dry Ratio	V-A O2 (mmHg)
EC (n=8)	40.8 ± 2.2*	46.0 ± 3.1*	-21.9 ± 4.7*	7.4 ± 0.3*	37.2 ± 4.6*
LPD (n=8)	28.9 ± 2.4	29.0 ± 4.2	1.8 ± 3.3	5.6 ± 0.1	296.3 ± 54.6
BLPD(n=7)	28.3 ± 1.5	28.8± 2.3	1.4 ± 6.2	5.7 ± 0.3	290.2 ± 66.4
ANOVA results:	*p=0.0003	*p=0.0005	*p=0.002	*p=0.0001	*p=0.001
	vs. LPD & BLPD	vs. LPD & BLPD	vs. LPD & BLPD	vs. LPD & BLPD	vs. LPD & BLPD

We conclude that extracellular solutions provide superior preservation of pulmonary function as demonstrated by increased oxygenation, decreased pulmonary artery pressure, decreased pulmonary vascular resistance, improved airway compliance, and decreased edema formation as measured by wet-to-dry ratios. However, the addition of blood does not confer any demonstrable advantage over LPD alone in this model of 18 hour cold ischemia. A potential mechanism of injury by intracellular solutions may involve endothelial damage of the pulmonary vasculature.

*By invitation

F18. NICORANDIL, K⁺ CHANNEL OPENER, AMELIORATES LUNG REPERFUSION INJURY.

Motohiro Yamashita, M.D.*, Ralph A. Schmid, M.D.*, Shozo Fujino, M.D.*, Koei Ando, M.D.*, Joel D. Cooper, M.D. and G. Alexander Patterson, M.D.

St. Louis, Missouri

Adenosine triphosphate-sensitive K⁺ (K_ATP) channels are a class of ionic channels recently found important in ischemic injury. Nicorandil (Nic) acts as a K_ATPchannel opener. Nic also acts as a nitric oxide donor and through that mechanism may reduce lung allograft reperfusion injury. In this study, we examined the effect of Nic on post-transplant function of preserved lung allografts. Donor lungs were flushed with modified Euro-Collins solution and stored for 21 hours at 1°C. Immediately following transplantation, the contralateral right main pulmonary artery and bronchus were ligated to assess isolated allograft function. Hemodynamics and arterial blood gas analysis (FiO₂ 1.0) were assessed for 6 hours prior to sacrifice. Allograft myeloperoxidase (MPO) activity was assessed as an index of leukocyte sequestration. Group I (n=5) animals received no Nic. In group II (n=5), Nic (24 mg/L) was added to the flush solution, recipient animals received Nic (0.5 mg/kg, IV) just prior to reperfusion and a continous infusion of Nic (0.74 ± 0.03 mg/kg/hr) during the assessment period. In group III (n=4), Nic was administered as in group II. In addition, group III animals received glibenclamide, a potent K_ATP channel antagonist (3 mg/kg) 15 minutes before Nic administration. Superior gas exchange (Fig. 1), hemodynamics and MPO data (Table 1) were noted in group II. The improvement of gas exchange and hemodynamics was suppressed by glibenclamide. These findings suggest Nic administration in the flush solution and during the reperfusion period ameliorates allograft function, improves cardiac output, and reduces pulmonary vascular resistance (PVR) and MPO activity in the transplanted lung. Lung allograft reperfusion injury is reduced by Nic likely as a result of its effect on K_ATP channels.

Table 1
	C.O. l/min	PAP mmHg	PVR dynes.sec.m2/cm5	MPO ΔOD/mg/min
Group I	1.44 ± 0.17*	27 ± 2	1000 ± 80*	0.40 ± 0.01*
Group II	2.51 ± 0.17	28 ± 4	620 ± 120	0.30 ± 0.03
Group III	1.34 ± 0.17*	26 ± 5	1200 ± 130*	0.38 ± 0.05
C.O.: Cardiac output; PAP: Pulmonary artery pressure; (mean ± SE) *p<0.05 (ANOVA) vs Group II

*By invitation

7:00 a.m. FORUM SESSION III - CARDIAC SURGERY

Room 6A/B, San Diego Convention Center

Moderators: D. Craig Miller, M.D.

Randall B. Griepp, M.D.

F19. MECHANISMS UNDERLYING DEGENERATION OF CRYOPRESERVED HOMOGRAFTS.

José P. Neves, M.D.*, Sérgio Gulbenkian, MSc., Ph.D.*, Ana P. Martins, M.D.*, Antònio M. Ferreira, Pharm.D., Ph.D.,* Ramiro Mascarenhas, Vet.D., Ph.D.*, Ricardo N. Santos, MSc.* and João Q. Melo, M.D, Ph.D.*

Lisbon and Santarem, Portugal

Sponsored by: Manuel E.M. Macedo, Lisbon, Portugal

Recent studies comparing heart valve (HV) homografts and valves of transplanted hearts showed that while the latter contained fibroblasts of both donor and recipient origin, the former were mostly acellular. These differences could be either due to the occurrence of an immune response in HV recipients that was prevented or abrogated by the immunosuppressive therapy administered to transplanted patients but not to the HV recipients, or to the cryopreservation process to which the HV homografts were subjected. To distinguish between these two alternatives, an experimental model was designed in which the behavior of cryopreserved autografts (CA) and homo-grafts (CH), implanted in the same animal, were compared. Fresh autografts (FA) were used to analyze the role of denervation and devascularization.

Cryopreserved aortic conduit homografts were implanted in the descending thoracic aorta of 15 sheep (6 males), aged 2 to 18 months. The excised aortic segment was then subjected to the same cryopreservation process used for the treatment of the homograft. One to eight weeks later, the CA was implanted, 1 to 2 cm below the CH. The intermediate segment of the native aorta was, at this point, dissected to be used as an FA control. Animals were sacrificed at different intervals (2 weeks, 1, 3, 6, 12, and 24 months) and the implanted segments harvested together with a portion of native aorta. Histological and immunohistochemical analysis as well as cell viability assessment were then performed on each of the explanted segments. Similar studies were also conducted on fragments of CA and CH collected before implantation.

With the exception of a partial loss of the endothelial cells, cryopreserved specimens had preserved cell viability and histology prior to implantation. Explanted CH, however, showed profound histological changes that affected all strata, as well as a decline in cell viability. Thus, after an initial period of non-specific inflammatory reaction which in most cases subsided after one
month, progressive neuronal and smooth muscle degeneration was observed, which led, in later stages, to the disappearance of axons and Schwann cells, fibrosis, hyalinization and calcification. Most likely due to this process, one CH ruptured after 17 months. Lymphocyte infiltrates were found up to 12 months after implantation. Endothelial cells were absent in all cases. In contrast, re-endothelization occurred in CA. After an initial inflammatory reaction as in all other segments, CA showed immunohistochemical signs of nerve degeneration with loss of Schwann cells and axons. After 1 month, however, progressive re-innervation occurred with re-establishment of the normal nervous tissue pattern being achieved 6 months after surgery. Histologically, a single alteration was present in these explants, consisting of an intimal thickening. Cell viability was similar to that of native aorta. Histological and immunohistochemical findings with regard to the FA were similar to those of the cryopreserved autografts, with the exception of the thickening of the intima, which did not occur.

In conclusion, it appears that the immunological reaction rather than the cryopreservation process is responsible for the degeneration that occurs in CH. Of particular interest were the findings that re-innervation, re-endothelization and regeneration of vasa vasorum occurred both on CA and FA. These conclusions are also important to the knowledge of the long-term behavior of aortic root replacement with CH or FA (Ross operation) in patients.

*By invitation

F20. BOTH PAPILLARY TIPS KEEP CONSTANT DISTANCE FROM THE MITRAL ANNULAR PLANE UNDER VARIOUS CONDITIONS.

Masashi Komeda, M.D., Ph.D.*, Julie R. Glasson, M.D.*, Ann F. Bolger, M.D.*, George T. Daughters, M.S.*, Neil B. Ingels, Ph.D.* and D. Craig Miller, M.D.

Stanford and Palo Alto, California

Mitral valve homografts are drawing more attention because they may preserve normal mechanics of the mitral subvalvular apparatus and improve postoperative LV performance, similar to reparative valve surgery. The dynamic nature of the LV, however, complicates precise preoperative and intraoperative assessment of LV geometry essential for homograft placement or complex valve repairs. To study various effects on 3-D mitral geometry, we investigated eight closed-chest dogs using implanted radiopaque markers under 4 conditions: 1) Baseline: automatic blockade (esmolol at 50-100 ug/kg/min), 2) Caval Occlusion: reduced preload (EDV fell from 143 ± 16 to 104 ± 13 ml [p<0.001], 3) Tachycardia: atrial pacing (heart rate increased 108 ± 11 to 131 ± 5 min^-1 [p<0.001], and 4) Nitroprusside: decreased afterload (2-5 μg/kg/min) (maximum LV pressure decreased from 132 ± 23 to 108 ± 29 mmHg [p<0.001]. Using cylindrical coordinates with the origin at the midpoint of the line connecting the anterior and posterior commissures and the LV long axis (z-axis) defined by the origin and the LV apex, D_TIP-MA(the distance along the z-axis between the papillary muscle [PM] tip and mitral annular plane) was measured at end-diastole and end-systole (mm, mean ± 1SD; n=8 for posterior PM, and n=7 for anterior PM):

(DTIP-MA)		Baseline	Caval Occlusion	Nitroprusside	Tachycardia
Posterior PM:	End-Diastole	25.8 ± 4.8	25.1 ± 5.2	25.6 ± 4.8	25.1 ± 4.7
	End-Systole	25.5 ± 4.5	25.5 ± 4.5	25.5 ± 4.4	25.0 ± 4.5
Anterior PM:	End-Diastole	20.7 ± 2.7	21.1 ± 2.6	20.9 ± 2.6	20.5 ± 2.5
	End-Systole	20.8 ± 2.8	20.7 ± 2.8	20.8 ± 2.8	20.4 ± 2.7

There were no significant differences in any dimensions by ANOVA. The distance between each PM tip and mitral annular plane was constant regardless of time during the cardiac cycle, or changes in preload, afterload, and heart rate. The mechanisms of maintaining this fixed tip-annulus distance are not known; however, these findings raise the possibility that the PM tip-annular distance might be a useful parameter to determine mitral homograft chordal length or help create more precise intraoperative strategies for complex valve repairs. Further investigations in dilated LV models with MR and the clinical setting are obviously necessary to define mechanisms and confirm these observations.

*By invitation

F21. PORT-ACCESS MITRAL VALVE REPLACEMENT IN DOGS.

Mario F. Pompili, M.D.*, John H. Stevens, M.D.*, Thomas A. Burdon, M.D.*, William S. Peters, M.B., Ch.B.*, Lawrence C. Siegel, M.D.*, Greg H. Ribakove, M.D.* and Bruce A. Reitz, M.D.

Palo Alto and Stanford, California; New York, New York

Introduction: Minimally invasive techniques have been elusive in cardiac surgery. We describe a method of MVR using an endovascular CPB system and one 35 mm by 17 mm oval port and two 10 mm lateral thoracic ports in dogs.

Methods: Fifteen dogs, 28 ± 3 kg (mean ± SD), were studied using the port-access MVR system (Heartport, Redwood City). Eleven dogs underwent acute studies and were euthanized immediately following the procedure. Four dogs were recovered and euthanized 4 weeks after surgery. CPB was conducted via femoral cannulae using an endovascular balloon catheter for aortic occlusion, root venting and delivery of antegrade cardioplegia. Catheters were inserted in the jugular veins for pulmonary artery venting and retrograde cardioplegia delivery. Through the oval port, a prosthesis (St. Jude or Carbomedics) was inserted via the left atrial appendage and secured to the annulus with 8 to 12 sutures. De-airing was performed.

Results: All animals were weaned from CPB in sinus rhythm. There was no MR by left ventriculography or PAOP v-wave in all but 2 dogs. In these 2 dogs, there was interference with prosthetic valve closure by residual native anterior leaflet tissue. Pathologic examination otherwise showed normal healing without peri-valvular discontinuity. Microscopic and SEM studies showed no damage to the valve surfaces. Cardiac output and PAOP were unchanged (2.8 ± 0.7 1/min and 7 ± 3 mmHg preop vs. 2.6±0.6 and 9 ± 4 postop). CPB duration was 113 ± 23 minutes and aortic clamp duration was 71 ± 15 minutes. Transthoracic echo of the 4 chronic dogs showed normal ventricular function and prosthetic valve function four weeks postoperatively.

Discussion: Mitral valve replacement with a minimally invasive method has been demonstrated in dogs. A clinical trial is appropriate.

*By invitation

F22. THE INDUCTION OF TOLERANCE TO AN EXPERIMENTAL CARDIAC ALLOGRAFT REQUIRES INTRATHYMIC INOCULATION OF CLASS IIMHC DISPARATE ANTIGENS.

Zhenya Shen, M.D.*, Muhammad Mohiuddin, M.D.*, Hitoshi Yokoyama, M.D., Ph.D.*, G. Russell Reiss, M.D.* and Verdi J. DiSesa, M.D.

Philadelphia, Pennsylvania

Indefinite donor-specific tolerance to a cardiac allograft disparate in both Class I and Class II major histocompatibility (MHC) antigens has been achieved in our laboratory and others by the pre-transplant intrathymic (IT) injection of donor spleen cells and a single intraperitoneal (IP) injection of anti-lymphocyte serum (ALS). This study was designed to determine whether this phenomenon was reproducible with either Class I MHC only or Class II MHC only disparate grafts. Three strains of inbred rats were studied in these experiments. Donors of cells and hearts in all experiments were RP rats which are rat MHC RT1 (AuBlDl). Class I MHC disparate grafts were performed by placing an RP heart into a Lewis recipient (RT1 A1B1D1C1) and Class II disparate grafts were performed with RP donors and Wistar Furth (WF) recipients (RT1 AuBuDuCu). Lewis (n=10) and WF (n=10) recipients underwent intra-peritoneal injection of 1 ml ALS and intrathymic injection of 5x10⁷ RP spleen cells. Three weeks later heterotopic cardiac transplantation was done using a heart from an RP rat. Control rats had no pretreatment or ALS alone. Without any pretreatment, RP hearts survive 7-9 (mean 8) days in Lewis recipients (n=5) and 9-14 (mean 12) days in WF recipients (n=5). ALS alone produces slight prolongation of graft survival (12 days in Lewis recipients [n=5] and 14 days in WF recipients [n=5]). Lewis rats pretreated with Class I disparate RP splenocytes and ALS had graft survivals of 8 - 27 (mean 14) days not significantly different from the effect of ALS alone. Class II disparate RP grafts placed in pretreated WF rats had significant prolongation of graft survival with 4 out of 5 grafts surviving more than 60 days (p<0.01 vs ALS alone). These results suggest that a disparity at the Class II locus of the MHC is critical for the induction of cardiac allograft tolerance after by intrathymic inoculation of allogeneic cells. This implies that the specific requirements for antigen presentation in the thymus are quite stringent even in this rodent model.

*By invitation

F23. INCREASED GRAFT AND SYSTEMIC VASCULAR PERMEABILITY DURING EARLY CARDIAC ALLOGRAFT REJECTION IS MEDIATED BY GRAFT AND SYSTEMIC EXPRESSION OF INDUCIBLE NITRIC OXIDE SYNTHASE.

Neil K. Worrall, M.D.*, Kathy Chang, Ph.D.*, Patrick M. Sullivan, B.A.*, Thomas P. Misko, Ph.D.*, Jia-Ji Hui, M.D.*, Joseph R. Williamson, M.D.* and T. Bruce Ferguson, Jr., M.D.

St. Louis, Missouri

We recently demonstrated that inducible nitric oxide synthase (iNOS) expression results in increased nitric oxide (NO) production during cardiac allograft rejection. In contrast to the physiologically protective role of NO in decreasing leukocyte adherence to endothelium, NO has also been implicated in mediating increased vascular permeability caused by various pathophysiological mediators, including LPS, TNF, and histamine. The present study examined whether NO contributes to increased vascular permeability to macromolecules during the early stages of graft rejection. Given the clinical systemic sequelae of rejection, we also examined whether early allograft rejection was associated with increased systemic vascular permeability. A double tracer permeation and microsphere method was used to examine vascular permeation (VP) during early graft rejection (POD 4) in a rat heterotopic cardiac transplant model: at time 0, ¹²⁵I-albumin was injected iv to measure VP; 8 min. later ¹³¹I-albumin was injected iv (intravascular space marker); and 1 min. later ⁴⁶Sc-labeled microspheres were injected iv (blood flow). One min. later the tissues were excised for γ-spectrometry, weighed, and 1) ¹²⁵I-albumin VP (intravascular tracer corrected), 2) blood flow, and 3) water content (wet/dry weight ratio) determined for each tissue. Allografts (Lewis to ACI) had increased VP and wet/dry weights in the grafted heart, lung, and brain compared to isografts (ACI to ACI) and controls (SEE TABLE). Increased allograft VP was associated with increased NO production (serum nitrite/nitrate levels) and with iNOS mRNA expression (determined by ribonuclease protection assay) in the grafted heart and the lung (not examined in brain; not detectable in the control or isograft tissues). iNOS inhibition with aminoguanidine (AG; 375 mg/kg/d iv) prevented the increased graft and systemic VP and water content, and normalized the serum nitrite/nitrate levels. Blood flow, cardiac output, and BP were not different between groups and were not affected by AG (not shown). AG had no effect on: 1) mild histological rejection score in allografts (1.7 ± 0.4 vs 1.6 ± 0.3; 0-5 scale); and 2) VP in isografts and controls (data not shown). These data demonstrate the novel observations that: 1) early allograft rejection increases vascular permeability and tissue water content in the systemic vasculature; 2) increased allograft heart and systemic vascular permeability is associated with increased NO production and iNOS mRNA expression in the allograft heart and lung; and 3) inhibition of NO production by iNOS prevents allograft heart and systemic vascular barrier dysfunction during early rejection.

Group	Grafted Heart		Lung		Brain		Nitrite/Nitrate
	VP	W/D Wt	VP	W/D Wt	VP	W/D Wt	(μM)
Con (n=15)	NA	NA	1324 ± 443	4.5 ± 0.2	52 ± 15	5.0 ± 0.1	14.6 ± 2.4