cancer support services - both physical and via the Net. We hypoth-esized that most rural patients have Internet access and would beinterested in lung cancer support services, including a website.Methods: All patients presenting to the cardiovascular and thoracicsurgery service at a single university institution were asked to com-plete an anonymous survey regarding Internet use. Survey questionsincluded age, gender, zip-code, and reason for the clinic visit. Pa-tients were asked if they had a home computer and Internet access.They were asked how many hours per week they used the Net and ifthey used it to research health questions. The survey informedpatients of the availability of support groups for patients with lungcancer and asked those with lung cancer to indicate if they would beinterested in attending such a group. They were also asked to indi-cate their interest in a website designed to support lung cancerpatients. Travel distances to the clinic visits were estimated, basedon zip-code data. The percentages of affirmative responses to surveyquestions were calculated based on the number of responders to eachrespective question. Results: Of 528 patient visits during the twomonth study period, 67 were identified as duplicate visits, leaving atotal of 461 potential first-time survey respondents. A total of 413surveys were returned, yielding a response rate of 89.6%. The meanage was 64.4 � 13.6 years (n�412). Two hundred forty patients weremen (58.1%). Fifty-eight patients (14.04%) reported their clinic visitwas for known or suspected lung cancer. Other reasons includednon-cancer lung disease (9.2%), cardiac (28.6%), esophageal (3.4%),vascular (6.8%), and other thoracic (4.8%) disease. An additional 137patients (33.2%) were not categorized due to vague survey responses.Excluding long-distance outliers of greater than 200 miles, the meandistance traveled was 48.1 � 44.0 miles (n�406). Two-hundredthirty eight patients (59.2%) had home computers, and 205 patientshad home Internet access (72.2%). The average Net use was 8.3 �11.0 hours/week (n�210). One-hundred sixty patients (70.2%) re-ported using the Internet to access health information. Eleven pa-tients (11.3%) reported an interest in attending a lung cancer sup-port group while 38 patients (38.4%) expressed interest in anInternet lung cancer support website. Of note, the number of surveyresponders to the questions regarding lung cancer group support andInternet support exceeded the number of survey patients indicatingknown or suspected lung cancer. Conclusions: Based on the abovesurvey data, rural patients seeking cardiovascular and thoracic caredrive moderate distances to receive care. A majority of patientsappear to have Internet access and use it to research health carequestions. More patients appear interested in Internet-based lungcancer support services than attending a support group in person;however, the overall interest is modest.

QS14. EMBRYONIC STEM CELLS ATTENUATE MYOCAR-DIAL DYSFUNCTION AND INFLAMMATION AFTERISCHEMIA VIA PARACRINE ACTIONS. Paul R. Crisos-tomo, Troy A. Markel, Meijing Wang, Yue Wang, ChristineHerring, Keith D. Lillemoe, Daniel R. Meldrum; IndianaUniversity, Indianapolis, IN

Introduction: Transplanted embryonic stem cells (ESCs) may im-prove cardiac function several weeks after infarct; however, theirmechanisms of protection and their usefulness during cardiac surgi-cal procedures remains unknown. The postulated paracrine mecha-nisms include mechanisms involving growth factor production. Wehypothesized that acute ESC treatment after ischemia would atten-uate myocardial dysfunction, reduce myocardial proinflammatorycytokines, and increase myocardial tissue growth factor levels.Methods: Adult Sprague-Dawley rat hearts (n�24) were isolatedand perfused via Langendorff model. Hearts were subjected to 25min warm global ischemia, 40 min reperfusion and randomly as-signed into one of three groups: (1) vehicle treated; (2) ESC treated;(3) ESC conditioned media treated (without cells). Myocardial func-tion was recorded, and hearts analyzed for expression of tissue IL-1b,IL-6, TNF, and VEGF (ELISA). ESCs were also plated at 5�10^6

cells/ml/well and subjected to TNF, LPS, or hypoxia. After 24-hourincubation, supernatants were collected and assayed for VEGF(ELISA). Experiments were repeated on three separate occassions(n�6-9 /group). Results: TNF, LPS, or hypoxia exposure signifi-cantly (p�0.05, T-test) increased ESC VEGF production in compar-ison to controls in cell culture. ESC treated and ESC conditionedmedia treated hearts demonstrated significantly (P�0.05, ANOVAand Tukey’s) greater post-ischemic recovery of left ventricular devel-oped pressure (75.1 �/� 8.4%) than controls (35.7 �/� 3.5%) at endreperfusion. Other functional indices (�dP/dT, �dP/dT, end diastolicpressure) were also significantly improved. ESC infused hearts (bothcells or media alone) demonstrated increased VEGF (130.6 �/�9.0pg/mg) and decreased TNF, IL-1b, and IL-6 production compared tocontrols. Conclusions: Embryonic stem cells in culture demonstrateincreased VEGF production in response to stress. Cell free mediaderived from these stressed ESCs as well as ESCs themselves atten-uate myocardial dysfunction and inflammation after ischemia. Em-bryonic stem cells may have protective paracrine effects that proveimportant in developing clinical therapeutic stem cell modalities.

QS15. THORACOSCOPIC LOBECTOMY IS ASSOCIATEDWITH DECREASED MORBIDITY AND DECREASEDLENGTH OF HOSPITAL STAY COMPARED TO OPENLOBECTOMY. Sean Kwon, Amy Kelsey, Valerie Kuderer,Royce Calhoun II; UC Davis Medical Center, Sacramento, CA

Introduction: Thoracoscopic lobectomy provides an anatomic resec-tion that adheres to oncologic principles while utilizing multiple portsites and a video camera, obviating the need to spread or divide ribsfor exposure. In effect, it offers a minimally invasive alternative tolobectomy via a standard thoracotomy. The purpose of this study wasto compare thoracoscopic lobectomy to open lobectomy with respectto short-term morbidity, mortality and other perioperative variables.Methods: Thoracoscopic lobectomies were performed utilizing 4 portsites and a 5-mm 30-degree camera. The lobe was removed via theaxillary extended port site after placing it in a bag. Standard ana-tomic lymph node dissection was performed. Standard posterolateralthoracotomies were performed in the open group. Utilizing our tho-racic surgical database, we compared our first (and most recent) 22consecutive thoracoscopic lobectomies to our most recent 22 consec-utive open lobectomies with respect to multiple perioperative vari-ables and endpoints. Groups were compared using a Student’s t-test.Results: The indication for surgery was cancer for the majority ofpatients in both groups. There were no conversions from thoraco-scopic to open in the study group. The first 4/5 patients in thethoracoscopic group received an epidural catheter with a mean du-ration of 2 days compared to 21/22 in the open group with a meanepidural duration of 2.7 days. There was no significant differencebetween the thoracoscopic and open groups with respect to age (66.2vs. 69.5 respectively), gender, pre-operative comorbidities and oper-ative times. There was no operative mortality in either group. How-ever, there were significant differences between the thoracoscopicand open lobectomy groups with respect to post-operative time on theventilator, length of ICU stay, operative blood transfusions, majorcomplications and hospital length of stay (Table 1). Conclusion:Thoracoscopic lobectomy offers patients a minimally invasive ap-

TABLE 1

Post-operative endpoints in thoracoscopic vs. openlobectomies.

Ventilator(days)

ICU(days)

ORTransfusions

(units) Complications

Lengthof Stay(days)

Thoracoscopic 0.0 0.3 0.0 2 (9%) 4.7Open 0.2 4.3 0.5 11 (50%) 9.7

276 ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS

proach to a standard cancer operation and is associated with signif-icantly decreased morbidity, ICU stay, and hospital length of staycompared to open lobectomy.

QS16. EFFECTS OF HDAC INHIBITOR LBH589 IN 36 DIF-FERENT LUNG CANCER CELL LINES. Maria C.Crisanti, A. S. Wallace, K. Coleman, V. Kapoor, M. Dowling,G. D. Kao, S. A. Albelda; University of Pennsylvania, Phila-delphia, PA

Background: Lung cancer is the leading cause of cancer deaths inthe US, with as many as 162,420 deaths in 2006. Mesotheliomarepresents 1.5% of these cases, and small cell lung cancer (SCLC)comprises approximately 20%, with non small cell lung cancer(NSCLC) being responsible for the rest. Lung cancer is characterizedby rapid growth and early metastasis. With little progress beingmade in the development of effective treatments and a poor five-yearsurvival rate, there is a need for new therapeutic approaches. His-tone deacetylase inhibitors (HDACI) are a recently developed class ofanticancer agents. The acetylated state of histones is associated withtranscriptional activity, and active histone acetylation plays a role inre-expression of silenced tumor suppressor genes. Specific HDACI,like LBH589, in vitro induce hyperacetylation of histones, cell cyclearrest and apoptosis. This agent is currently being studied in clinicaltrials (hematologic malignancies). Our goal was to expand the studyof LBH589 to lung cancer. Materials and Methods: In vitro: MTTcell proliferation assays were carried out for 36 cell lines (11 NSCLCcell lines �7 human (h), 4 murine (m)-, 13 SCLC cell lines (h), 12mesothelioma cell lines �8 h, 4 m- and one human thymoma cellline). IC50 and LD50 were determined and compared to Suberoyla-nilide hydroxamic acid (SAHA). Also the combination with standardchemotherapy was analyzed. Cell cycle analysis was performed withFlow Cytometry. Conditions tested were LBH589, etoposide andirradiation (IR) alone and in combination. In vivo: Wild type miceand/or SCID mice were injected with tumor cells in the flank (H69,M30, TC1, A549, AE17). When tumors reached 200mm3, LBH589was administered with daily intraperitoneal (IP) injections (10 and20mg/kg). Tumors were measured twice weekly. After 2 weeks oftherapy tumors were harvested and pharmacokinetic studies werecarried out. In addition, SCID mice with H69 tumors were adminis-tered etoposide as standard SCLC therapy and combination therapywith LBH. Results: MTT Assay. LBH589 IC50s are significantlylower than SAHA, with values in the low nM range (ranging from 4to 470nM, median of 20), significantly lower than SAHA. SCLC celllines were the most sensitive ones, so we decided to pursue the studyon this type of lung cancer cell lines. Cell cycle analysis. SCLC celllines express time-dependent efficacy of LBH589, expressed by anincrease in the sub-G1 amount of cells according to the exposuretimes. Etoposide and IR alone both show modest efficacy in inducingapoptosis, which was increased when adding LBH589. In vivo.LBH589 significantly decreases tumor growth when compared tocontrol in the different cell lines tested. Repeating the experimentsin SCID mice yielded the same results, this indicates that there is noimmunologic effect related to the effect of LBH589. The effect ofHDACI followed by the administration of etoposide was synergisticand statistically significant (p�0.05). Conclusion: HDAI are a re-cently developed class of anticancer agents, which seem to havepromising results on lung cancer cell lines. SCLC cell lines seem tobe particularly sensitive to LBH589, an HDAI, with IC50 in the lownM range and efficacy in decreasing tumor size. LBH589 also showedsynergy when combined with etoposide. The promising resultsagainst SCLC should encourage the design of clinical trials in thesearch for new therapeutic approaches for this devastating disease.

QS17. THE VOLUME-OUTCOME RELATIONSHIP AF-FIRMED: HOSPITAL VOLUMES, CABG RISK ANDOUTCOMES OVER 10 YEARS IN NEW YORK STATE.Rosemarie E. Hardin, Giancarlo Cires, Robert C. Lowery,

Michael E. Zenilman, James J. Reilly, Jeremy Weedon,Joshua H. Burack; SUNY Downstate Medical Center, Brook-lyn, NY

Introduction: Volume-outcome relationships have been proposedfor various surgical procedures. We hypothesize that high volumecoronary artery bypass (CABG) centers produce improved risk ad-justed outcomes. Methods: We investigated the relationship be-tween hospital volume and outcomes using the New York StateDepartment of Health’s Cardiac Surgery database from 1993-2003.We examined observed, expected, and risk-adjusted mortality rates(OMR, EMR and RAMR, respectively) of over 180,000 patients in the35 hospitals performing CABG. Longitudinal regression analyseswere performed, using these rates as the dependent variables, andcalendar year and hospital volume (Q� quartiles by volume) as theindependent variables. (Q1� up to 322 cases/year, Q2� 323-487/year, Q3�488-708 /year, Q4� above 708 /year). We then performedpairwise comparisons of the risk adjusted mortality rates comparingQ1 and Q4 hospitals. Absolute risk reduction (ARR), relative riskreduction (RRR) and number needed to treat (NNT) using Q1 and Q4hospital data were calculated. Results: Cardiac surgical risk (EMR)rose over 11 years, with a positive correlation with hospital volume(p�0.035). Observed mortality rates (OMR) were unaffected overtime or by volume. RAMR was significantly higher in Q1 (2.62%)compared to the Q4 hospitals (2.07%). Compared to Q1 hospitals, Q4hospitals demonstrated an ARR of 0.55% and a RRR of 20.9%. Thenumber needed to treat at high volume hospitals to achieve thisbenefit is 181.8. Conclusions: Over the past decade, CABG surgicalrisk, reflected by EMR, increased in NYS, with higher proportion ofpatients being treated at the highest volume Q4 hospitals. Despitethe higher risk of their patients, high volume hospitals produce lowerrisk-adjusted mortality rates than low volume Q1 hospitals.

QS18. HEPATOCYTE GROWTH FACTOR SIGNALING ANDBCL-XL DYSREGULATION IN MALIGNANT PLEU-RAL MESOTHELIOMA. James E. Littlejohn1, Xiabao Cao2,Lidong Zhang2, Charles Rodarte2, Philip Rasco2, JonathanDaniel2, W. R. Smythe2; 1Texas A&M HSC - College of Med-icine, Temple, TX; 2Scott and White Memorial Hospital andClinic, Temple, TX

Introduction: Malignant Pleural Mesothelioma (MPM) is an ag-gressive neoplasm which is highly resistant to conventional chemo-therapeutics. Bcl-xL is a key anti-apoptotic protein that is expressedin many tumor types, and has been shown to contribute to therapeu-tic resistance in MPM via up-regulated expression. Bcl-xL transcrip-tion is responsive to the Ets family of transcription factors such asETS and PU.1. Serum levels of Hepatocyte Growth Factor (HGF) aretwice as high in MPM patients as in the healthy population and itsreceptor, HGFR (c-MET), is overexpressed in tumor. Attenuating theHGFR signal has been shown to inhibit tumor cell growth andmigration, however, the mechanism(s) associated with HGF promo-tion of neoplasia are not well characterized. In this study we exam-ined the role of HGF signaling in controlling apoptosis and Bcl-xLexpression in MPM. Methods: The human I-45 MPM cell line wasutilized for all studies. Promoter activity was assessed using a lucif-erase reporter gene construct. Bcl-xl mRNA level was measured withRT-PCR. Phosphorylated ETS and PU.1 (transcription factor) levelswere obtained using immunoprecipitation-western blot analysis.Subcellular distribution of ETS-2 and PU.1 was analyzed via immu-nostaining and fluorescence microscopy. Results: Over-expression ofEts-2 and PU.1 cDNA in I-45 cells stimulated Bcl-xL promoter ac-tivity (Figure 1.). Human MPM cells demonstrated up-regulation ofBcl-xL mRNA at baseline (250% above serum-starved levels) andgreater up-regulation when exposed to HGF (600% above serum-starved levels). This up-regulation was specific to Bcl-xl, as we ob-served no alteration in the level of other Bcl-2 family members,

277ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS