3.13 Development of an Endoluminal Intestinal Lengthening Device: Geometric Intestinal Attachment

Posted on December 22, 2014

F. R. Demehri1, J. Freeman1, Y. Fukatsu1, D. H. Teitelbaum1  1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA

Introduction:   Distraction-mediated enterogenesis, whereby longitudinal force applied to small bowel leads to functional lengthening, may provide a novel therapy for short bowel syndrome (SBS).  Previously described methods have relied upon isolated small bowel segments or transmural fixation, requiring multiple operations and subsequent morbidity.  Our work consisted of two objectives: First, to develop a fully endoluminal distraction-mediated enterogenesis device; Second, to develop a novel approach utilizing geometric coupling between a tapering device and the mesenteric curvature to allow intestinal attachment and trans-stomal distraction-mediated enterogenesis.                          

Methods:   A tapering catheter device was designed to provide endoluminal intestinal attachment via geometric coupling between the rigid device and mesenteric curvature.  This consisted of a high-rigidity 24-Fr malleable catheter with a 30 cm taper to a low-rigidity 8-Fr latex tip to prevent perforation (Fig A).  Yorkshire pigs underwent jejunal Roux limb creation with placement of the device via jejunostomy.  Intestinal attachment and distraction was achieved without significant reduction in bowel perfusion as measured by laser Doppler (0.95 ± 0.03 P.U. vs 1.03 ± 0.04 P.U. with inflation vs deflation; p=0.14).  An external clamp was secured at the stoma to provide external fixation of the device (Fig B).  The device was manually advanced 1cm/day for 7 days before explant.  A second group of pigs underwent extended implant, with explant at day 14.  Results (mean±SEM) were analyzed for significance by t-test.

Results:   After 7 days of trans-stomal, externally-applied endoluminal distraction, the distracted segment achieved a 32.4±4.6% increase in length versus fed, nondistracted bowel, corresponding to an absolute gain of 10.6±1.2cm.  After 14 days, the Roux limb achieved an 80.3±1.4% increase in length versus fed control bowel, corresponding to an absolute gain of 16.7±3.4cm.  The device fixation site demonstrated occasional epithelial sloughing with intact submucosa.  No device-related perforation or stoma-related complication occurred (Fig C). 

Conclusion:  A novel catheter device with tapering rigidity allows for fully endoluminal intestinal attachment via geometric coupling, with successful externally-applied distraction-mediated enterogenesis.  This approach may allow development of clinically applicable technology for the treatment of patients with SBS.

3.14 Safety of a Collagen-Targeted Peptide Amphiphile Nanofiber for Intravascular Use

Posted on December 22, 2014

W. Jiang1,2, E. M. Bahnson1, M. R. Kibbe1  1Feinberg School Of Medicine – Northwestern University,Vascular Surgery,Chicago, IL, USA 2Northwestern University,McCormick School Of Engineering,Chicago, IL, USA

Introduction: We developed, characterized, and validated a collagen-binding peptide amphiphile (PA) nanofiber as a novel targeted delivery vehicle that specifically binds to and releases a therapeutic agent directly at the site of vascular injury and prevents neointimal hyperplasia following systemic administration. However, in order to translate this therapy to the clinical arena, the safety of this nanofiber as a drug delivery tool must be evaluated. Thus, the goal of this current project is to evaluate the safety of this systemically administered targeted nanofiber. We hypothesize that a systemically injected targeted nanofiber will be safe following in vivo administration.

Methods: Male Sprague-Dawley rats received a single dose of a fluorescently labeled collagen-binding PA (2.5mg, N=3). Liver, spleen, and kidney were harvested at different time points (24, 48, and 72h) for H&E staining and immunohistochemical evaluation for macrophage infiltration (ED1 antibody). Blood samples were collected for coagulation studies. C3a levels in plasma was assessed by ELISA. Functional complement activation was assessed by a hemolysis assay following incubation of human serum with PA (1 and 0.5mg/ml) for 1h. Platelet activation was evaluated by assessing adherence ex vivo after incubation of rat platelet rich plasma with PA (1 and 0.5mg/ml) for 30 min.

Result: Assessment of liver, spleen, and kidney histologically revealed the presence of fluorescence at 24, 48, and 72h, suggesting the presence of the targeted nanofiber. However, no change in organ architecture was observed in H&E stained slides and there was no difference in macrophage infiltration in the liver, kidney, and spleen of nanofiber-treated animals compared to control animals. The serum dose-response curves for the complement activation assay showed no difference between treated and untreated serum. At a 1:15 serum dilution, the percentage of residual complement for different concentrations of the PA (1 and 0.5mg/ml) was 99.8 +/- 7.7% and 96.4 +/- 8.8%, respectively vs. 100.7 +/- 9.9% for untreated serum, indicating no significant activation of complement. Similarly, C3a levels in rat plasma did not differ between control and PA-treated rats. Regarding the coagulation studies, the PT was increased to 22 +/- 4s at 24h (p=0.027) and 21 +/- 0.1s at 48h (p=0.044) after PA administration compared to 15 +/- 2s for control, but returned to normal at 72h (14 +/- 0.2s). PTT, fibrinogen and platelet count showed no statistically significant change after treatment. Finally, PA did not cause platelet activation ex vivo.

Conclusion: This data suggested that systemic administration of our novel targeted intravascular nanotherapy is safe, as organ morphology, inflammation, and complement and platelet aggravation were not impacted following administration in vivo. While there was a transient rise in the PTT, this returned to baseline within 3 days. Thus, no major safety concerns were found in our study.

3.15 Pharmacokinetic Model for Supramolecular Nanoscale Carriers Targeted to the Injured Vasculature

Posted on December 22, 2014

E. M. Bahnson1, H. Kassam1, K. T. Nennig3, M. J. Avram2,3, M. R. Kibbe1  1Feinberg School Of Medicine – Northwestern University,Vascular Surgery,Chicago, IL, USA 2Feinberg School Of Medicine – Northwestern University,Anesthesiology,Chicago, IL, USA 3Northwestern University,Donnelley Clinical Pharmacology Core,Chicago, IL, USA

Introduction: Nanoscale carriers that incorporate targeting information are emerging as a promising modality of delivering a therapeutic load with high efficacy and minimal toxicity to the vasculature. However there is little information about the pharmacokinetic properties of such compounds. To prevent neointimal hyperplasia, we developed a novel targeted therapy capable of delivering a therapeutic agent to the site of vascular injury via systemic administration, using a highly customizable peptide amphiphile (PA). The goal of this project is to study the disposition of the collagen-targeted PA and develop a pharmacokinetic model for it. We hypothesize that the targeted nanofibers will exhibit a pharmacokinetic behavior distinct from that of classic small molecule drugs.

Methods: PAs were synthesized using solid-phase peptide synthesis, and purified by preparative reverse-phase HPLC. Male Sprague Dawley rats received a single dose (2.5 mg) of PA via tail vein injection. Blood was collected for analysis at various times over three days (0.1, 0.2, 0.25, 0.5, 0.45, 1, 1.5, 2, 3, 4, 6, 8, 16, 24, 48, and 72 h; n=3/time point). After addition of a PA internal standard, samples were reduced with 5 mM tris(2-carboxyethyl)phosphine. Proteins were precipitated with acetonitrile before HPLC injection.  Concentrations of the PA were determined by HPLC with mass spectrometry detection (Lower Limit of Quantification = 1 ng/ml). A multicompartmental pharmacokinetic model was then developed for the PA nanofibers.

Results: A 3-compartment pharmacokinetic model fit yielded the parameters shown in Figure 1. Estimate of the central volume is consistent with the volume of plasma and interstitial space in rats. The concentration of the nanofibers in plasma decreased two orders of magnitude in the first hour. The first fast exponential phase generally corresponds to distribution clearance of rapidly equilibrating tissues. However, for the PA nanofibers, this first exponential phase corresponded to elimination clearance, that is a process that removes the nanofibers from plasma irreversibly within the time frame of the study. The ex-vivo stability of the nanofibers in plasma revealed that this rapid process is not due to rapid degradation by plasma components. The very fast elimination clearance could be due to organ sequestration or irreversible conjugation.

Conclusions: We describe an unusual 3-compartment pharmacokinetic model of a targeted nanofiber that clearly differs from classic small molecule pharmacokinetics. This information becomes essential when trying to understand the differential properties of nanocarrier-based drugs. Further work is required in order to elucidate the mechanism responsible for the very fast elimination clearance.

3.16 Implantable Hemofilter: An Overview of the Pre-clinical Canine Surgical Model

Posted on December 22, 2014

C. Kensinger1, J. Groszek2, S. Karp1, D. Laneve1, P. Williams1, R. Kant3, T. Yeager3, S. Roy3, W. Fissell2  1Vanderbilt University Medical Center,Department Of General Surgery,Nashville, TN, USA 2Vanderbilt University Medical Center,Department Of Medicine,Nashville, TN, USA 3University Of California – San Francisco,Department Of Bioengineering And Therapeutic Sciences,San Francisco, CA, USA

Introduction:  Patents with end-stage renal disease (ESRD) have high mortality and morbidity rates on dialysis.  Transplantation offers the best treatment option but is limited by organ availability.  An implantable artificial kidney using silicon nanoporous membranes is in development to address this problem.   A key challenge is the long-term blood patency of the hemofilter with respect to thrombosis formation and membrane fouling from plasma protein adsorption. These experiments assess the long-term biocompatibility of an implantable artificial kidney.

Methods:  Single-channel, polycarbonate parallel-plate hemofilters (membrane area 0.72 cm2) are designed using computational fluid dynamics to minimize turbulence and stasis through the device flow path.   Two ultrathin membranes with highly controlled pores made by sacrificial silicon oxide techniques, which allows size selective sieving, define the parallel plates along the flow path.  Hemofiltration results from the hydrostatic pressure gradient across the membrane. 

Seven millimeter polytetrafluoroethylene grafts are anastomosed to the common iliac artery and vein to be used as inflow and outflow conduits to the hemofiltration device.  Heparin is administered intra-operatively.  Effluent collection reservoirs are placed in the upper abdomen.   Access ports are connected to the reservoirs and brought through the abdominal wall into the subcutaneous tissue.  This allows post-operative monitoring of hemofilter function and filtration rates.  Following the operation, the animals are housed without restrictions.   The grafts are serially assessed post-operatively with a doppler ultrasound to ensure patent flow through the devise.   

Results: Fourteen pre-clinical canine surgeries have been performed. Hemofilter patency rates were 67% (8/12) prior the initiation of post-operative Warfarin treatment. With adequate systemic anticoagulation, patency rates have been 100% (2/2). Intra-operative filtration rates have averaged 0.41 microliters/minute with an expected filtration rate of 0.42 microliters/min based on membrane characteristics.  Blood flow through the device has averaged 300cc/minute at the time of hemofilter implant and explant. 

Conclusion: These experiments highlight successful surgical technique and manufacturing biocompatibility providing the foundation for further preclinical experiments aimed towards the future realization of an implantable artificial kidney.
 

3.17 Changes in Liver Lobe Function After Portal Vein Ligation Determined by Selective Biliary Drainage

Posted on December 22, 2014

A. Szijarto1, A. Fulop1, A. Budai1, G. Lotz2, A. Kiss2, L. Harsanyi1  1Semmelweis University,1st Department Of Surgery,Budapest, BUDAPEST, Hungary 2Semmelweis University,2nd Department Of Pathology,Budapest, BUDAPEST, Hungary

Introduction: Portal vein ligation (PVL) and embolization (PVE) are techniques used before extended hepatic resections to prevent posthepatectomy liver failure. These therapies redirect portal blood to liver lobes that will remain after surgery resulting in hypertrophy, while the portal deprived lobes undergo atrophy. Although, the effect of PVL on liver volume is well-documented, the parallel alterations in lobar liver function are still the subject of controversy. Therefore, the aim of the present study was to evaluate the morphological, hemodynamic and functional alterations caused by the selective occlusion of the portal vein in a well-established rat model.

 

Methods:  Male Wistar rats (n=84) underwent PVL by the ligation of the portal veins feeding the median, left lateral and caudate lobes (approximately 80% of total liver mass). Before PVL, as well as 24-, 48-, 72-, 120-, 168 hours after PVL, liver morphology (liver weight; mitotic activity; necrotic-, apoptotic cell death and lobular area), hepatic microcirculation (laser Doppler flowmetry), global liver function (laboratory blood test; total hepatic bile flow; plasma disappearance rate of indocyanine-green (PDR); the percentage of biliary ICG excretion to the administered ICG during the first 20 minutes (ICG%20min)) as well as hepatic lobar function (lobar bile flow and lobar biliary ICG excretion) were examined.

Results: PVL induced atrophy of ligated lobes (from 0.35±0.037 to 0.09±0.018g/BWkg) and hypertrophy of non-ligated lobes (from 0.1±0.102 to 0.31±0.019g/BWkg), while the total liver weight remained unchanged. The microcirculation of ligated lobes impaired, while the microcirculatory blood flow of non-ligated lobes significantly increased with the peak response at 48th postoperative hours. Serum albumin-, total bilirubin levels and total hepatic bile flow did not changed significantly throughout the entire experiment. PDR and ICG%20min significantly decreased after PVL, with the lowest value at postoperative 48th hours and returned near to the baseline at 168 hours after the operation. The bile flow and biliary ICG excretion of ligated lobes decreased significantly after PVL (from 81.99±8.6 to 17.62±3.9g/BWkg and from 1.16±0.03 to 0.1±0.03g/BWkg, respectively), while bile flow and biliary ICG excretion of non-ligated lobes showed a significant increase (from 23.65±2.75 to 88.8±4.67g/BWkg and from 0.34±0.03 to 1.4±0.03g/BWkg, respectively).

 

Conclusion: PVL induced a temporary impairment in global liver function, followed by a rapid recovery mainly caused by the increase in the function of non-ligated liver lobes. In the non-ligated lobes, the functional increase was more pronounced than suggested by the degree of hypertrophy. Consequently, the functional capacity of the liver was shifted towards the regenerating lobes in a greater extent than would be expected according to the volumetric alterations.

3.18 Remote Ischemic Preconditioning's effect on Fat Graft volume

Posted on December 22, 2014

A. A. Gassman1, M. Lewis2, J. C. Lee1  1University Of California – Los Angeles,Surgery/ Plastic Surgery,Los Angeles, CA, USA 2West Los Angeles VA,Pathology,Los Angeles, CA, USA

Introduction:
Fat grafting has become a useful adjunct in the reconstructive surgeon’s treatment armamentarium. Inconsistencies in transfer and local ischemia prior to the development of recipient circulation all contribute to highly variable long-term results associated with fat grafting. Remote Ischemic Preconditioning (RIPC) is a cheap non-invasive technique that has been used in several animal models and multicenter clinical trials to protect several organ systems. The specific aim of this project was to analyze the volume retention of lipoaspirate transferred in the setting of either donor or recipient RIPC.  

Methods:
We obtained subcutaneous adipose tissue from FVB mice transgenically engineered to express eGFP and Luciferase. These samples were obtained either with or without the use of temporary hindlimb tourniquet time prior to harvest. The samples were excised and passed to through serially smaller lipoaspiration cannulas (16 to 19 gauge), centrifuged (500g for 2 min), and decanted. The treatment and control fat was injected into the dorsal skin folds of genetically identical FVB mice that did not express GFP or Luciferase. The viability and volume of the transferred tissue was examined over a 28-day time period by bioluminescence after intraperitoneal injection of Luciferin using a Maestro IVIS optical small animal scanner. Additionally, after experimental completion the tissue transferred was explanted and examined histologically. The specimens were stained with H&E, CD31, CD34, and GFP.

Results:
Bioluminescence was able to non-invasively track the presence of transferred lipoaspirate tissue over a 28 day time period.  There was a significant difference in bioluminescence and calculated graft volume at Day 0 and 28.  The RIPC group demonstrated approximately 700% and 400% increase over control at each time point, respectively.  Histological analysis at 28 days confirmed the presence of donor adipocytes, and that they were gradually replaced by recipient inflammation and scar tissue.  However the amount of interstitial fibrosis was substantially less in the RIPC group.  Additionally, the RIPC group retained a substantially greater amount of GFP suggesting retention of donor cells. The control tissue demonstrated increased CD31 and CD34 suggesting increased vascularity. 

Conclusion:
This work has achieved two goals. Firstly, It demonstrates that the bioluminescence of adipocytes transferred from a luciferase expressing donor may be used to non-invasively monitor tissue viability and volume over a prolonged period of time. Secondly, RIPC has the ability to increase the viability of donor adipocytes when transferred via liposuction cannula, and the transferred tissue is less likely to undergo interstitial fibrosis.
 

3.19 Hemoglobin Based Oxygen Carriers Exacerbate Hyperfibrinolysis Independent of Plasmin

Posted on December 22, 2014

A. P. Morton3, H. B. Moore3, E. Gonzalez3, G. Wiener3, P. Lawson2, M. Chapman3, C. Silliman3, E. Peltz1,3, A. Banerjee3, E. E. Moore2,3  1University Of Colorado Hospital,Surgery,Aurora, CO, USA 2Denver Health Medical Center,Surgery,Denver, CO, USA 3University Of Colorado School Of Medicine,Surgery,Aurora, CO, USA

Introduction: Hyperfibrinolysis plays an integral role in the genesis of trauma induced coagulopathy (TIC). Recent data demonstrates that red blood cell (RBC) lysis promotes fibrinolysis; however, the mechanism is unclear.  Hemoglobin-based oxygen carriers (HBOC) have been developed for resuscitation and have been associated with coagulopathy.  We hypothesize that replacement of whole blood (WB) using an HBOC results in a coagulopathy due to the presence of free hemoglobin.

Methods: Whole blood was sampled from healthy donors (n=6).  The clotting profile of each citrated sample was evaluated using native thromboelastography (TEG).  Serial titrations were performed using both HBOC (Polyheme) and normal saline (NS) (5%, 25%, and 50%) and evaluated both with and without a 75 ng/microliter tissue-plasminogen activator (tPA) challenge.  Tranexamic acid (TXA) was added to inhibit plasmin dependent fibrinolysis.  Fibrinolysis was measured and recorded as LY30, the percentage of clot lysis at 30 minutes after maximal clot strength.  Statistics were calculated using SPSS software.  Dilution of WB with NS or HBOC was correlated using LY30 via Spearman Rho coefficients.  Groups were also compared using a Friedman test and post-hoc analysis with a Bonferroni adjustment.

Results: TPA-provoked fibrinolysis was enhanced by both HBOC (median LY30 at 5%, 25%, 50% titrations: 11%, 21%, 44%; Spearman=0.94; p<0.001) and NS (11%, 28%, 58%; Spearman=0.790; p<0.001). However, HBOC also enhanced fibrinolysis without the addition of tPA (1%,4%,5%; Spearman=0.735; p=0.001) and NS did not (1%,2%,1%; r=0.300; p=0.186; Figure 1).  Moreover, addition of TXA did not alter or inhibit this fibrinolysis (WB vs 50% HBOC: 1.8% vs 5.65%, p=0.04).  There was no significant difference in fibrinolysis of HBOC with or without TXA (50% HBOC vs 50% HBOC +TXA: 5.55% vs 5.65%, p=0.92).  Additionally, the increased fibrinolysis seen with NS was reversed when TXA was present (WB vs 50% NS: 1.8% vs 1.65%, p=1.0).

Conclusion: HBOCs enhance fibrinolysis without the addition of tPA; moreover, this is independent of plasmin as the phenomenon persists in the presence of TXA.  Our findings implicate the hemoglobin molecule or its components in stimulating fibrinolysis

 

3.20 Self-assembly Nanoparticles of PLGA-polyethylenimine (PLGA-PEI) Copolymer for Gene Delivery

Posted on December 22, 2014

J. LU1, Z. Liang1, Q. Yao1, C. Chen1  1Baylor College Of Medicine,Surgical Research/Surgery,Houston, TX, USA

Introduction: Although gene therapy holds great promise, the progress has been slow because the current gene delivery systems are less successful for clinical applications due to their low efficiency and high toxicity. The objective of this study was to develop a better gene delivery system from two biocompatible polymers, poly(lactic-co-glycolic acid) (PLGA) and polyethylenimine (PEI), potentially for clinical applications.

Methods: The PLGA-PEI copolymer was prepared by directly mixing PLGA and PEI in organic solvent. Self-assembly nanoparticles (NPs) of PLGA-PEI copolymer and DNA were prepared by adding the DNA solution to the water solution of PLGA-PEI copolymer. The size and morphology of PLGA-PEI/DNA NPs were determined with dynamic light scattering and scanning electronic microscopy. The cytotoxicity of PLGA-PEI/DNA NPs at different copolymer to DNA ratios was performed in pancreatic cancer cell line (PANC-1) with an MTT assay. For the gene transfection assay in PANC-1 cells, plasmid DNA containing a green or red fluorescence protein (GFP or RFP) gene was used. In vivo toxicity and transfection efficiency of PLGA-PEI PLGA-PEI/DNA NPs were carried in mouse models with the tail vein administration.

Results: PLGA-PEI copolymer was produced in one-step by mixing PEI and PLGA in the tetrahydrofuran solution and PLGA to PEI ratio (0.5:1). Through the analysis of primary amines of PEI before and after its chemical reaction with PLGA, the chemical structure of the PLGA-PEI copolymer was demonstrated.  PLGA was broken down to LGA single units, which were covalently linked to the primary amine groups of PEI; while PEI was intact. PLGA-PEI copolymers spontaneously formed NPs (~100 nm in diameter) with plasmid DNA at the 1.5:1 ratio due to change of the surface charge, achieving 100% DNA loading. The particle size can be controlled by justifying copolymer and DNA ratios.  PLGA-PEI significantly reduced the toxicity of PEI in both PANC-1 cells and mouse models. PLGA-PEI copolymer more efficiently delivered GFP or RFP plasmid into PANC-1 cells compared with commercially available transfection reagents with additional advantages of less toxicity, serum independency and long duration of transgene expression. More importantly, PLGA-PEI/DNA NPs were tested in the mouse model and showed an effective gene delivery to liver, spleen, and pancreas. Direct intratumor administration of PLGA-PEI/DNA NPs also showed a high transfection rate in the nude mouse model.

Conclusions: PLGA-PEI copolymer is a new gene delivery material, which has high DNA loading capacity and low toxicity in vitro and in vivo; and it condenses DNA into small sized NPs. The PLGA-PEI/DNA NPs have a high transfection efficiency in cell cultures and mouse models. The current study demostrates a better gene delivery system, which may have board clinical applications.

3.01 Intraperitoneal Application of Intestinal Alkaline Phosphatase Decreases Post-Operative Adhesions in Mice

Posted on December 22, 2014

S. K. Hyoju1, S. Morrison1, M. Gharedaghi1, M. Mohamed1, S. S. Gul1, M. Najibi1, T. Phupitakphol1, A. Osmani1, K. Economopoulos1, S. Hamarneh1, R. A. Hodin1  1Massachusetts General Hospital,Department Of Surgery,Boston, MA, USA

Introduction: Damage to the peritoneum initiates an inflammatory response leading to the formation of adhesions which subsequently cause significant morbidity in some patients.  Intestinal alkaline phosphatase (IAP) is a gut enzyme capable of detoxifying various inflammatory mediators such as lipopolysaccharide, lipoteichoic acid, CpG DNA and nucleotide triphosphates, and could therefore play a role in decreasing adhesion formation. This study aimed to test the anti-inflammatory effects of IAP on the development of post-operative adhesions in mice.

Methods: C57BL/6 mice were used in this study and underwent midline laparotomy. Six peritoneal buttons were created by pinching and ligating the peritoneum with 5-0 silk sutures. The buttons were subsequently cut and cauterized. The cecum was exteriorized and abraded with sterile dry gauze and returned to the abdominal cavity. At designated time points during the procedure, a total of, 500µl of either IAP (5000U) or vehicle was applied over the peritoneal cavity. On the 21st postoperative day, mice were euthanized and adhesions from abdominal viscera to each button were scored based on number, tenacity, and extent of the adhesions. The final adhesion score for each mouse was calculated by the summation of these scores. The investigators were blinded to the treatment groups. In a second series of experiments, this procedure was repeated, and the animals were instead euthanized on the 1st post operative day. Peritoneal buttons were then excised and homogenized, and the resulting supernatant was collected for cytokine analysis.Student’s t test was used to analyze the data

Results:The mice treated with IAP had significantly lower adhesion scores compared to vehicle-treated animals (16.4 ± 6.8 vs. 27.8 ± 8.7, P = 0.04).Peritoneal button homogenates from IAP-treated mice had significantly lower IL-1β level and TNF-α levels compared to vehicle-treated animals (64.57±7.4 vs. 111.31±18.9 pg/mg tissue protein, P=0.0006; 6.76±1.69 vs. 10.5±3.07 pg/mg tissue protein, P=0.04) respectively.

Conclusion:Intraperitoneal application of IAP at the time of laparotomy could represent a novel approach to the prevention of post-operative adhesions.

 

3.02 PRP Enhances Posterior Lumbar Spinal Fusion Using a Composite Engineered Scaffold In a Rabbit Model

Posted on December 22, 2014

J. L. Van Eps1,4, J. S. Fernandez-Moure1,4, F. J. Cabrera4, S. J. Minardi4,5, B. Aghdasi2,4, A. Tampieri4,5, E. Tasciotti4, B. K. Weiner2,3,4  1Houston Methodist Hospital,Department Of Surgery,Houston, TX, USA 2Houston Methodist Hospital,Department Of Orthopedics & Sports Medicine,Houston, TX, USA 3Weill Cornell Medical College,New York, NY, USA 4Houston Methodist Research Institute,Department Of Nanomedicine, Surgical Advanced Technology Lab,Houston, TX, USA 5Istituto Di Scienza E Tecnologia Dei Materiali Ceramici (ISTEC),Department Of Bio-Ceramic And Bio-Hybrid Composites,Faenza, FAENZA, Italy

Introduction:

With increased implementation of spinal instrumentation and fusion, an emphasis to develop biologic alternatives to the morbidity of autografting or prosthetic implants has emerged. Additionally, existing platforms utilizing bone morphogenic proteins (BMPs) are limited by potential side effects and alternatives are needed. Autologous platelet-rich plasma (PRP) is a universally-available source of growth factors with currently untapped potential for bone regeneration.

Methods:

A proprietary magnesium-hydroxyapatite/collagen 70/30 wt% composite scaffold was implanted in sixteen New Zealand White rabbits randomly assigned to two experimental groups undergoing L5-L6 intertransverse posterior spinal fusion – scaffold alone (S1) or scaffold + PRP (S2). Autologous PRP was isolated preoperatively for S2 rabbits via double-centrifugation of whole blood (8mL), standardized platelet concentration to 5×106/mL and activation with a solution of 10%CaCl and 300U thrombin, and used to soak scaffolds pre-implantation. The left anatomic side received decortication alone, serving as an internal control (IC). DynaCT imaging and post hoc bone volume quantification was performed at 2, 4, and 6 weeks postoperatively, and differences in experimental side/IC bone volume were compared using thresholds of 200 and 500 Hounsfield units (HU) – known correlates to trabecular and cortical bone respectively. Postoperative specimens were harvested, plastic embedded and sectioned, H&E stained, and visualized under microscopy to confirm fusion of new and native bone.

Results:

Both groups displayed significantly more osteogenesis on the experimental side compared to IC (p<0.05), confirming scaffold osteoconductivity, and earlier, more robust osteogenesis was seen in S2 vs. S1 rabbits by dynaCT imaging and volumetric analysis (Figure 1). This trend for trabecular bone formation was witnessed as early as 2 weeks (p<0.05) and became more pronounced at 4-6 weeks (p<0.01). The near 2-fold greater trabecular bone formation translated into more mature bone formation at 6 weeks as well, as Group S2 rabbits displayed significantly more cortical bone at 4 and 6 weeks postoperatively (p<0.05). Histological analysis confirmed more robust fusion of native and new bone and bridging intertransverse osteogenesis in S2 rabbits.

Conclusion:

This work confirms the feasibility of bone regeneration and spinal fusion using solely biocompatible materials and autologous biologic products. Additionally, the greater overall osteogenesis and earlier mineralization afforded by PRP support its clinical utility and potential abrogation for usage of synthetic moieties such as BMPs – or at minimum offer a safer, diminished required dose – in the future.

3.03 Oncolytic Recombinant Vaccinia Virus GLV-2b372 Efficiently Kills Hepatocellular Carcinoma

Posted on December 22, 2014

J. W. Ady1, C. Johnsen1, K. Mojica1, J. Heffner1, D. Love1, A. Pugalenthi1, J. Belin1, J. R. Aguilar5, N. Chen5, Y. A. Yu5, A. Szalay5, Y. Fong4  1Memorial Sloan-Kettering Cancer Center,Surgery,New York, NY, USA 2University Of California – San Diego,3Department Of Radiation Medicine And Applied Sciences, Rebecca & John Moores Comprehensive Cancer Center,San Diego, CA, USA 3University Of Würzburg,4Department Of Biochemistry, Rudolph Virchow Center For Experimental Biomedicine, And Institute For Molecular Infection Biology,Würzburg, BAVARIA, Germany 4City Of Hope National Medical Center,Surgery,Duarte, CA, USA 5Genelux,Research And Development,San Diego, CALIFORNIA, USA

Introduction:  Hepatocellular carcinoma (HCC) is the most common primary malignancy of the liver with limited treatment options and poor prognosis. Therefore, there is a great need to develop novel therapies for HCC.   Oncolytic viruses that specifically infect, replicate within, and kill cancer cells represent an emerging and promising platform for cancer therapy.   In this study we assess at the ability of GLV-2b372, a novel nonattenuated vaccinia virus derived from the Lister 1.1.1 strain, to kill HCC.

Methods:  The infectivity, viral replication and cytotoxicity of GLV-2b372 was assayed in 4 human HCC cell lines.  Huh-7 flank xenografts were generated in athymic nude mice.  Mice were treated with intratumoral injections of GLV-2b372.  Biodistribution of viral presence in animals treated with GLV-2b372 was assessed.

Results: Infectivity increased in a time and concentration dependent manner in all cell lines. All cell lines supported efficient replication of virus. Flank xenografts treated with GLV-2b372 demonstrated complete erradication of tumor in 75% of animals with significant tumor reduction as compared to controls (p < 0.05).  Biodistribution confirmed sustained intratumoral presence of virus at 2 weeks post infection, with rapid clearance of tumor from all other tissues. 

Conclusion: Our results demonstrate that the novel oncolytic vaccinia virus GLV-2b372 selectively infects and replicates in HCC tissue.  We demonstrate infectivity and killing of HCC both in vitro and in vivo.  These findings indicate that GLV-2b372 is a potent oncolytic vector that could serve as a promising platform for oncolytic cancer immunotherapy.  

 

3.04 Traumatic Brain Injury Alters Circulating Microparticles which then Impact Coagulation

Posted on December 22, 2014

E. F. Midura1, C. C. Caldwell1, M. D. Goodman1  1University Of Cincinnati,Cincinnati, OH, USA

Introduction:  Traumatic brain injury (TBI) is known to result in a sub-acute, post-traumatic, hypercoagulable state, however the pathophysiology behind this alteration is poorly understood. Platelet-based contribution to clot is known to decrease over time following trauma and microparticles have also been shown to be changed after TBI. Whether platelets and microparticles interact to influence coagulation has not been determined. Using a murine model, we hypothesized that microparticle and platelet contributions to clot formation would be altered after TBI. 

Methods:  An established weight-drop model was used to induce TBI in anesthetized mice. Sham mice underwent anesthesia without TBI. Blood samples were collected 24 hours after injury, and circulating microparticle and platelet counts determined. Thromboelastometry (ROTEM) was used to evaluate changes in hemostasis. ROTEM was then repeated on sham and post-injury blood treated with normalized concentrations of microparticles isolated from sham mice (sham microparticles), TBI mice (TBI microparticles), or microparticle-free saline. A microparticle pro-coagulant assay was used to compare activity in sham and TBI mice.

Results: One day after TBI, we observed a decrease in overall microparticle concentrations (3.6 vs. 1.9 x 108 microparticles/ml, p < 0.001). Overall platelet counts did not change after injury. ROTEM demonstrated a reduced platelet contribution to hemostasis 24 hours after TBI compared to sham (62.9% vs. 48.0%, p < 0.01). When TBI microparticles were added to sham blood, a significant decrease in platelet contribution to clot was seen compared to saline (Figure 1). Conversely, when sham microparticles were added to blood from injured mice, there was a normalization of platelet contribution to clot formation when compared to saline (Figure 1). Furthermore, when microparticle pro-coagulant activity was assayed, TBI microparticles had increased activity compared to sham microparticles (3.28 vs. 1.72, p = 0.04).

Conclusion: Our data demonstrate that after TBI, 1) circulating microparticles are decreased, 2) platelet contribution to clot formation is decreased despite unchanged platelet count, and 3) TBI microparticles independently augment post-traumatic clot formation. Post-TBI microparticles may therefore be responsible for the altered platelet role in coagulation and the development of a post-traumatic hypercoagulable state.

 

3.05 Controlled Release of Nitric Oxide Enhances Gemcitabine Cytotoxicity in Pancreatic Adenocarcinoma

Posted on December 22, 2014

J. Fernandez-Moure1,2, D. Kirui1, J. Van Eps1,2, N. Dhanani1,4, F. Cabrera1, M. Ferrari1,3, E. Tasciotti1  1Houston Methodist Research Institute,Nanomedicine,Houston, TX, USA 2Houston Methodist Hospital,Surgery,Houston, TX, USA 3University Of Texas Health Science Center At Houston,Houston, TX, USA 4Texas A & M Health Science Center College Of Medicine,Bryan, TX, USA

Introduction:

Gemcitabine (GEM) is the first-line treatment for pancreatic adenocarcinoma (PAC). Despite such broad use, intrinsic and acquired chemoresistance is common. Loss of the tumor suppressor SMAD4 is present in more than 50% of PAC and is known to play a role in chemoresistant mechanisms. Nitric oxide (NO) is the predominant species responsible for the cytotoxic action of macrophages against cancer cells yet localized delivery is difficult given the short half-life and volatility. Silica (Si) nanoparticle mediated local delivery has been shown to be an effective strategy for NO delivery and controlled release. We sought to study the effects of locally delivered NO on GEM mediated PAC cytotoxicity and the potential role of SMAD4 in this effect. We hypothesized that NO would enhance the cytotoxicity of GEM in a SMAD4 independent manner.

 

Methods:

NO-Silica nanoparticles (NO-Si) were synthesized via a co-condensation of tetraethoxysilane with aminoalkoxysilane under high pressure nitrous oxide for 3 or 4 days. Particle characterization included scanning electron microscopy, dynamic light scattering, and Fourier transform infrared spectroscopy. NO release was measured using a chemiluminescence nitric oxide analyzer. A SMAD4 negative PAC cell line (SMAD4-) was made using lentiviral knockdown of Panc1 PAC cells and confirmed by western blot. Panc1 and SMAD4- cells were then treated with gemcitabine (100nm to 30μm), 30 mg NOSi particles (NOSihi or NOSilo), or both for 72 hours. Cell viability was then quantified by MTS cell proliferation assay.

 

Results:

Half-life of NOSi NO release is 3.5±0.6 hours. NOSilo maximum concentration of NO release was 750ppb/mg and NOSihi was 2250ppb/mg. NOSilo alone reduced cell viability by 30% and 35% in SMAD4 and Panc1 respectively. When combined with GEM, NOSilo led to a significant reduction in cell viability of both cell lines at all concentrations used. The greatest effect was seen at 1μm where cytotoxicity was increased by 50% in Panc1 and 70% in SMAD4-. NOSihi led to >90% reductions in cell viability in Panc1 and no difference was seen when combined with GEM. NOSihi alone led to only 70% reduction in SMAD4- cells. When combined with GEM, cytotoxicity was improved to 90% at GEM concentrations as low as 5μm.

 

Conclusion:

We have demonstrated the in vitro dose dependent cytotoxic effects of Si nanoparticle NO controlled release on PAC. When combined with GEM there is a synergistic effect resulting in improved cytotoxicity seen in both Panc1 and SMAD4- PAC cells with a differential pattern of cell death seen at high concentrations of NO. These findings suggest not only that NOSi is useful chemosensitizing agent but that SMAD4 may play a role in its synergism with GEM. Creation of a novel class of therapeutics where local controlled release of NO is used as a method of chemosensitization may lead to a paradigm shift in not only treating PAC but all solid tumors where chemoresistance is common.

 

2.01 Circulating DNA increases in response to neoadjuvant chemoradiation therapy for pancreatic cancer.

Posted on December 22, 2014

I. A. Naqvi1, R. Gunaratne1, D. Pisetsky2, R. R. White1  1Duke University Medical Center,Surgery,Durham, NC, USA 2Duke University Medical Center,Medicine,Durham, NC, USA

Introduction:  Pancreatic cancer (PC) is the deadliest of the major cancers, and response rates to standard therapies are low.  DNA is released by tumors, and circulating extracellular DNA (exDNA) and nucleosomes (short segments of DNA wrapped around histones) have been evaluated as markers of disease burden for prognosis and response to therapy.  We hypothesized that exDNA and nucleosome (NUC) levels would decrease in patients undergoing neoadjuvant chemoradiation (neoCRT) for localized PC.

Methods:  Blood was collected from 1) patients with radiographically localized PC (N=8) before and 4-6 weeks after neoCRT, 2) patients with advanced disease (N=3) and 3) normal volunteers (N=5) in order to create a spectrum of data for disease burden. ExDNA and NUC were isolated from serum samples and quantified using a fluorescent DNA-specific stain (Picogreen) and ELISA, respectively.  

Results:  There was a strong correlation between exDNA and NUC levels, and therefore only exDNA levels are shown.  Healthy individuals had the lowest levels of exDNA while patients with advanced disease had the highest levels (Figure 1).  Overall, exDNA levels increased in patients who underwent neoCRT to levels similar to those of patients with advanced disease, regardless of whether they had radiographic partial response (PR, N=4), stable disease (SD, N=4), or progressive disease (PD, N=2) on restaging (Figure 2) nor whether they were subsequently able to undergo resection (not shown). 

Conclusion:  Contrary to our hypothesis, exDNA levels increased in response to neoCRT.  These data suggest that exDNA is not simply a marker of disease burden.  We postulate that exDNA is released by tumor cells in response to therapy and also by host immune cells in response to the tumor.  In order to put this into a broader context, we are expanding our dataset to include more patients and time points, including pre- and post-resection.  We are evaluating other specific forms of exDNA as well as DNA-associated proteins as markers of cell death that may be more useful for monitoring response to therapy. 

2.02 Hypoxia Inducible Factor-1α Enhances Hepatic Metastasis in Murine Pancreatic Adenocarcinoma

Posted on December 22, 2014

J. A. Yi1, E. E. Moore1,2, A. Banerjee1, K. El-Kasmi1, C. C. Barnett1,2  1University Of Colorado Denver,Aurora, CO, USA 2Denver Health Medical Center,Aurora, CO, USA

Introduction:  Pathological studies have demonstrated that hypoxia inducible factor-1α (HIF1α) is a marker of advanced malignancy in pancreatic cancer patients.  HIF1α is a transcription factor that regulates multiple cellular processes including oxygen transport, angiogenesis, metabolism, cell adhesion, and inflammation. We hypothesize that HIF1α activity is important for the development of hepatic metastases in pancreatic adenocarcinoma.

 

Methods:  The Pan02 murine pancreas adenocarcinoma cell line was modified using short-hairpin RNA targeting HIF1α via lentiviral transduction to create Pan02-SH+ cells lacking HIF1α activity. Knockdown of HIF1α activity by >80% was confirmed using polymerase chain reaction prior to injection of the tumor cells.  C57/Bl6 mice underwent splenic injection of 200,000 Pan02 or Pan02-SH+ cells followed by hemisplenectomy to create intraabdominal metastatic tumor dissemination. After 4 weeks, mice were euthanized for necropsy performed by blinded observers.  Tumor volume was calculated as (height x width2).  Statistical analysis was performed using Student’s t-test and Fisher’s exact test, with significance determined to be α<0.05.

 

Results:  Abrogating HIF1α activity of the tumor cells resulted in a decreased number of hepatic metastases (p=0.0397) and decreased hepatic tumor volume (Pan02 mean volume=1275 mm3 ± 236.08, Pan02-SH+ mean volume=16.4 mm3 ± 13.98; p=0.0007). There was no significant difference in non-hepatic intraabdominal metastases by count (p=0.1347) or by tumor volume (Pan02 mean volume=1193.8 mm3 ± 264.97, Pan02-SH+ mean volume=358.6 mm3 ± 278.55; p=0.0616) with HIF1α shRNA knockdown. 

 

Conclusion:  HIF1α activity is important for the development of hepatic metastases in pancreatic adenocarcinoma.  This was reflected by both reduced number of malignant lesions and tumor volume in the liver.  Non-hepatic metastases were not significantly decreased by reduction of HIF1α activity.  HIF1α may represent an important target in hepatic directed therapy for pancreatic adenocarcinoma.

2.03 A Novel Orthotopic Mouse Model of Pancreatic Cancer with Immunocompetent Milieu and Robust Stroma

Posted on December 22, 2014

K. Majumder1, S. Modi1, N. Arora1, R. Chugh1, A. Nomura1, S. Banerjee1, R. Dawra1, A. Saluja1, V. Dudeja1  1University Of Minnesota,Surgery,Minneapolis, MN, USA

Introduction:  The development of novel therapeutics for pancreatic cancer has been hindered by a lack of relevant preclinical models. The introduction of the KPC (KrasLSL.G12D/+; p53R172H/+; PdxCretg/+) mouse model has shed some light on tumorigeneis and progression in pancreatic cancer and has been a step in the right direction. An accurate tumor model, besides recapitulating the tumorigenic properties, has to recapitulate the immune and stromal microenvironment, the components lacking in commonly used subcutaneous or orthotopic models in immunodeficient mice (SCID, Athymic nude). While these components are present in the KPC model, this genetic model is fraught with inconsistency making this unsuitable for study of various component of tumorigenesis as well as the study of novel therapies. For example, the time to invasive disease in this model varies from 47-355 days of life. The amount of pre-invasive disease and invasive adenocarcinoma varies between litter mates and up to 20% of animals never develop tumor. Furthermore, the entire pancreas in this model has Kras-p53 mutation as opposed to the sporadic mutations generally seen in human pancreatic cancer. We therefore propose a novel orthotopic tumor model with tumors from KPC mice implanted in C57Black6 mice, which recapitulates the tumor microenvironment and circumvents the drawbacks of the above mentioned models.

Methods:  6 month KPC mice with palpable pancreatic tumors were sacrificed and the tumors were cut into ~3mm3 pieces using core biopsy punch. Laparotomy was performed using a midline incision on female C57BL/6 mice and the tumor piece was sewn into a pocket of pancreas using a figure of 8 stitch of 7-0 prolene incorporating the superior and inferior border of the pancreas. Mice were sacrificed at two time points, 4-weeks and 8 weeks after implantation. Tumor volume and weight was measured. Stromal component and immune infiltration were studied by immunohistochemistry.

Results: Tumor take rate was ~90%. Tumor volume was very similar in all mice suggesting similar growth rate (tumor volume. 4 weeks: 331 ± 122 mm3, 8 weeks: 433 ± 77 mm3). Mortality at 8 weeks was ~20%. Immunohistochemistry confirmed presence of pancreatic adenocarcinoma at both time points. Staining for stromal components (collagen and αSMA) showed intense desmoplastic stromal reaction at both 4 and 8 weeks after implantation. CD45 staining demonstrated intense infiltration of tumor and surrounding pancreas with leukocytes.

Conclusion: The proposed model of pancreatic cancer offers all the advantage of state of the art genetic model along with the predictability of a conventional orthotopic model. This model mimics the stromal reaction and immune infiltration observed in human pancreatic ductal adenocarcinoma as well as circumvents the issue of variability seen in previous mouse models. This clinically relevant model can be a valuable tool to evaluate novel therapeutics in pancreatic cancer.

 

2.04 The Tumor Supportive Phenotype of Chemotherapy Conditioned PDAC Fibroblasts Depends on NfkB

Posted on December 22, 2014

S. G. Patel1, L. Li1, A. Nguyen1, P. Toste1, N. Wu1, C. Choi1, J. Smogorzewski1, T. Donahue1  1University Of California – Los Angeles,Surgery,Los Angeles, CA, USA

Introduction:
We have previously shown that exposure of pancreatic cancer (PDAC) tumor associated fibroblasts (TAFs) to cytotoxic chemotherapy unleashes a tumor supportive and pro-inflammatory phenotype, increasing tumor cell (TC) growth and invasion in vitro and in vivo. We hypothesize that Nfkb is the predominant transcription factor driving this TAF response to DNA damage, and therefore genetic knockout of the activator subunit, p65 in TAFs will attenuate the induction of these genes and tumor cell viability.

Methods:
Immortalized PDAC TAFs (Logsden Lab) were homozygously deleted for p65 using CRISPR:Cas9.  RT-PCR and MTT assay were done using standard protocols.  Conditioned media was collected for 24 hours from the TAFs preconditioned with 48 hours of 100 nM gemcitabine or basal media. PANC-1 TCs were cultured in this media for two days and MTT assay used to assess cell viability and proliferation.  CHIP-Seq data and promoter motif enrichment was derived from NIH Epigenome roadmap public data and processed using GREAT. 

Results:
Gemcitabine treatment of PDAC TAFs induced a pro-inflammatory gene expression program including statistically significant (p < 0.05) upregulation of a number of cytokines: IL-6 (9.2 fold), IL-8 (17.8 fold), IL-1a (62.1 fold), and Spp1 (2.7 fold) [Fig 1A].  The genes are part of a DNA damage response collectively known as the senescence associated secretory phenotype (SASP).  Each of these genes have Nfkb binding sites identified from Chip-Seq for p65 or from binding site motif identification.  The expression of these gemcitabine-induced SASP genes followed the cytosolic to nuclear translocation of the activating subunit of the Nfkb complex: p65 [Fig 1B].  Homozygous knockout of p65 in TAFs (TAF p65 -/-) resulted in loss of upregulation of multiple SASP mediators in gemcitabine treated cells as compared to wild-type TAFs with intact p65 (p<0.05).  Additionally, two genes in our panel expressed at low levels under basal conditions in WT TAFs (IL8, IL-6) were silenced in the p65 KO TAFs.  Conditioned media collected from naïve and 48 hour gemcitabine treated WT TAFs or non-target control (RelA NC) increased the proliferation of PDAC TCs (Panc-1, p<0.05); however, conditioned media from gemcitabine treated p65 KO (-/-) TAFs did not permit cancer cell growth in low serum conditions [Fig 1C].

Conclusion:
Gemcitabine treatment of PDAC TCs shows a cytotoxic effect in vitro, but is complicated by the emergence of resistance in vivo.  TAFs influence on TCs is typically not appreciated in the in vitro setting.  We show that culture media conditioned by TAFs is supportive of TC growth, in vitro, and inflammatory cytokines expressed subsequent to induction of DNA damage in TAFs are silenced in the absence of p65.   We conclude that in the presence of gemcitabine, Nfkb signaling in stromal cells can promote cancer cell growth.  
 

2.05 E-Cadherin Expression in Obesity-Associated, Kras-Initiated Pancreatic Ductal Adenocarcinoma

Posted on December 22, 2014

A. P. Stark1, W. Sheppard1, X. Jung1, K. Hertzer1, A. Moro1, H. Chang1, M. Xu1, O. J. Hines1, G. Eibl1  1University Of California – Los Angeles,Department Of Surgery,Los Angeles, CA, USA

Introduction: Epithelial-mesenchymal transition (EMT) is increasingly considered to be a necessary first step in the development of invasive disease. Accordingly, loss of the epithelial marker E-cadherin is associated with decreased survival in human pancreatic ductal adenocarcinoma (PDAC). Alarmingly, EMT has been demonstrated prior to tumor formation in a mouse model of PDAC—and is accelerated by inflammation. Obesity is a known pro-inflammatory state and an established risk factor for PDAC. We have previously demonstrated accelerated pancreatic tumorigenesis as a result of a high fat, high calorie diet; however, the direct effect of diet-induced obesity on EMT and E-cadherin expression is unknown.

Methods: Conditional KrasG12D (KC) mice were fed a control diet (CD; 3726 kcal/kg, 12% fat) or a high fat, high calorie diet (HFCD; 4536 kcal/kg, 40% fat). Mice were weighed weekly. To demonstrate the spectrum of pancreatic epithelial neoplasia (PanIN 1-3), cohorts were sacrificed at 3 and 9 months of age. Pancreata were preserved in formalin and embedded in paraffin. Immunohistochemistry (IHC) was performed using antibody against E-cadherin. Immunoreactivity was characterized by staining intensity, location (cell membrane, cytoplasm, nucleus), and proportion of positive cells. 

Results: For the 3-month cohort, implementation of the diet led to significantly increased weight gain among HFCD mice (n=7) vs. CD mice (n=7), 15.8gm vs. 5.6gm (p<.001). In the 9-month cohort, HFCD mice (n=7) also had higher average weight gain compared to CD mice (n=5), 19.8gm vs. 12.9gm (p = .007). Compared to CD-fed animals, mice fed the HFCD for 3 and 9 months showed more robust inflammatory features in the pancreas as demonstrated by a loss of acinar tissue and formation of desmoplasia. Mice fed the HFCD for 9 months had more severe pathology than mice fed the HFCD for 3 months. No invasive cancer was detected. Overall, E-cadherin expression was ubiquitous, strong, and membranous regardless of diet in both 3- and 9- month cohorts. Staining intensity was intermittently decreased in higher-grade PanIN lesions, but no differences between groups were noticed (See Figure 1).

Conclusions: Despite evidence of more advanced disease, KC mice fed a HFCD did not show evidence of grossly altered E-cadherin expression. If early EMT is part of the mechanism underlying diet-induced acceleration of pancreatic tumorigenesis, it may occur despite E-cadherin expression.

2.06 Increased Stromal Integrin-Linked Kinase Expression in IPMN is Associated with Worse Survival

Posted on December 22, 2014

L. A. Shirley1, B. Swanson2, W. Frankel2, T. Bekaii-Saab3, M. Bloomston1, C. Chen4  1Ohio State University Wexner Medical Center,Surgical Oncology,Columbus, OH, USA 2Ohio State University Wexner Medical Center,Pathology,Columbus, OH, USA 3Ohio State University Wexner Medical Center,Medical Oncology,Columbus, OH, USA 4Ohio State University College Of Pharmacy,Medicinal Chemistry,Columbus, OH, USA

Introduction: Intraductal papillary mucinous neoplasms (IPMNs) are pre-malignant lesions of the pancreas that are typically found incidentally.  The presence of an IPMN presents a clinical dilemma in that it is difficult to predict which patients will go on to succumb to pancreatic cancer.  Integrin-linked kinase (ILK) is a serine-threonine kinase known to be overexpressed in pancreatic cancer, correlating with worse survival.  We examined resected pancreata with IPMN to see if similar correlations exist.

Methods: Intraductal papillary mucinous neoplasms (IPMNs) are pre-malignant lesions of the pancreas that are typically found incidentally.  The presence of an IPMN presents a clinical dilemma in that it is difficult to predict which patients will go on to succumb to pancreatic cancer.  Integrin-linked kinase (ILK) is a serine-threonine kinase known to be overexpressed in pancreatic cancer, correlating with worse survival.  We examined resected pancreata with IPMN to see if similar correlations exist.

Results: In our cohort, 10 samples (35.7%) had an invasive component.  Mean IPMN ILK score was 0.56, while mean stromal ILK score was 1.96.  If an invasive component was present, the mean ILK score in this region was 0.6.  There was no correlation between IPMN (P=0.46) or invasive component (P=0.55) ILK scores and patient survival.   However, high ILK expression (score of 3) in the stroma surrounding the IPMN was associated with a significantly worse overall survival (median survival not reached vs. 33.5 months, 5-year survival 69.8% vs. 22.2%, P=0.016) (Figure 1).  This correlation remained significant when stratified by the presence of an invasive component (non-invasive median survival not reached vs. 39.6 months, 5-year survival 75.2% vs. 50%, invasive median survival 62.6 vs. 19.9 months, 5-year survival 60% vs. 0%, P=0.029).

Conclusion: Stromal ILK over-expression in IPMNs was associated with worse survival, independent of the presence of an invasive component.  Stromal ILK expression could be used in the future as a marker to stratify patients with IPMNs that need more aggressive therapy.  Additionally, targeting ILK overexpression may provide a novel therapeutic strategy. 

 

2.07 Poor PDA Patient Prognosis Predicts Xenograft Tumor Engraftment Likelihood

Posted on December 22, 2014

H. L. Beal1, J. Shea1, M. Firpo1, S. Mulvihill1,2, C. Scaife1,2  1University Of Utah,Department Of Surgery Research,Salt Lake City, UTAH, USA 2Huntsman Cancer Institute,Salt Lake City, UTAH, USA

Introduction:

Pancreatic adenocarcinoma (PDA) is an exceptionally lethal disease. The development of new efficacious therapies could be improved through the development of animal models that more faithfully replicate disease biology.  One such approach is to utilize patient-derived primary mouse xenograft tumors.  Here we describe the results of our xenograft tumor banking program from 2011-2013 and compare the xenograft phenotype to the patients, as well as patient prognosis, and compare characteristics of the xenograft tumor with the patient tumor. We tested the hypothesis that a more aggressive tumor phenotype will grow more frequently in mice.

Methods:

With IRB and IACUC approval, tumor samples from suspected PDA cases were obtained within an hour of resection, placed in media, cut into approximately 1mm3 pieces, mixed with Matrigel, and implanted subcutaneously into SCID hairless female mice. Upon reaching 2cm in diameter, the tumor was processed for histology and re-implanted in a new mouse.  Tumors for histology were fixed in formalin, embedded in paraffin and then stained with hematoxylin and eosin or immunohistochemically stained with epithelial markers (E-cadherin and  β-catenin).  Each patient’s chart was reviewed for patient demographics, diagnosis, recurrence, and overall and disease-free survival.

Results:
In the three year study period, 54 patients with periampullary adenocarcinoma were enrolled.  Twenty-two patients were female (age 68 +/- 7 years) and 32 were male (age 63 +/- 9 years).  We implanted 47 PDA tumors, 4 cholangiocarcinoma tumors, and 3 ampullary tumors. The overall xenograft take rate was 44%.  Take rates for PDA (13/39; 9 under observation), cholangiocarcinoma (4/4), and ampullary (3/3) tumors were 33%, 100%, and 100%, respectively. Once established, 84% of tumors successfully grew in a second mouse.

Histologically, mouse PDA xenograft tumors were remarkably similar to the original patient tumor with all samples demonstrating extensive desmoplastic tissue surrounding tumor cells.  The histological characteristics of the tumor appear to be stable over several passages. The xenograft tumors cells were positive for both E-cadherin and β -catenin. Mean survival in patients with succefully engrafted PDA tumors was 10 +/- 6 months, while patients whose tumor did not engraft had mean survival of 21 +/- 10 months (p=0.005).

Conclusion:

We have successfully established 13 patient derived PDA, 3 ampullary and 4 cholangiocarcimonas. The PDA xenograft tumors maintain histological similarities with the patient tumor and preliminary survival analysis suggests that more aggressive tumors are more likely to establish in the mouse model. The tumors established in our xenograft banking program, are an important resource that can be utilized to evaluate new therapeutics treatments. 

 

« Previous Page
Next Page »