Pancreatic Cancer Research Initiative

Pancreatic cancer, one of the most serious types of cancer, spreads rapidly and is rarely detected in its early stages. More than half of all patients diagnosed with pancreatic cancer worldwide have a five-year survival rate of less than two percent.

It is estimated that 44,030 men and women will be diagnosed and 37,660 will die from pancreatic cancer in 2011, based on the National Cancer Institute’s SEER Cancer Statistics Review.

AICF made two Pancreatic Cancer Research Initiative grants in 2009-10 adn 2010-11 to Giulio F. Draetta, MD, PhD and Mark Bloomston, MD. Information on their work is included on the following pages.

We would like to extend special thanks to the donors whose generous support has funded the Pancreatic Cancer Research Initiative:

Named Grants

The Alexander Bodini Foundation

The Del Vecchio Family Foundation

The Amy & Joseph Perella Charitable Fund

Additional Donors

Mr. and Mrs. Tarek Abdel-Meguid

Kent Backlund

Gian Andrea Botta

Anna Bulgari

Robert Buxton, CFA

Vivian Cardia

Joy Germont

Mr. & Mrs. Ben Krupinski

Thomas Maloney

Franco Muggia, MD

George Pavia

Richard Perkins

Ottavio Serena di Lapigio

Jonathan Schulhof

Jane Cowles Smith, PhD

Konstantinos Tsouvelekakis

Mario d’Urso

Roy Weiner, MD

Anonymous

The Alexander Bodini Foundation Pancreatic Cancer Research Initiative Grant

Giulio F. Draetta, MD, PhD

Presidential Scholar and Deputy Director for Science

Center for Applied Cancer Science

Dana-Farber Cancer Institute

Identification of Novel Drug Targets in Pancreatic Cancer

Collaborators include William C. Hahn, MD, PhD and Massimo Loda, MD, Dana-Farber Cancer Institute; Giovanni Tonon, MD, PhD, San Raffaele del Monte Tabor Scientific Institute, Milan; and Chiara Conti, PhD, Belfer Institute.

Year One Progress Report

Summary. Pancreatic ducal adenocarcinoma (PDAC) is one of the leading causes of cancer mortality in the United States and Europe. The goal of this project is to perform RTK antibody profiling in PDAC cell lines and in primary tumor samples to identify driver RTK for potential therapeutic application. We are very excited to report that we have made tremendous progress in this area and have identified at least one RTK for further analysis in year two.

Project Aims

Cell line and primary PDAC profiling. Given the recent observations of multiple receptor tyrosine kinase being co-activated in glioblastoma and other cancers, we performed RTK antibody array profiling of 21 PDAC cell lines, most of which harbor KRAS mutation. We found that most PDAC cell lines show multiple RTK co-activation as expected and indepdent of KRAS status (03.27 and BxPC-3 are KRAS WT). Interestingly, a few of the PDAC cell lines (CFPAC-1, HPAF-II and MiaPaca2) do not display RTK co-activation suggesting yet other unidentified RTKs might be involved since only 42 of the 54 RTKs were included in the antibody array experiment. We have also performed RTK array profiling on primary PDAC tumors (n=3, 10-50% stromal tissues). Importantly, co-activation of RTKs (up to 7) can also be observed in primary PDAC or xenograft tumors with EGFR activation being the universal feature. Although only a small number of primary tumors and xenograft tumors were profiled, we estimate that a majority of primary PDACs display RTK co-activation extrapolated from our PDAC cell lines data.

Cell proliferation and survival assays in PDAC cell lines and tumor explants. Ax1 is frequently co-activated in PDAC and is a novel oncology target for potential drug development. We therefore focused on the Ax1 protein for further study. We have analyzed 12 PDAC cell lines by immunoprecipitation of Ax1 followed by phosphor-tyrosine western blotting. Interestingly, 11 of the 12 PDAC cell lines display high levels of phosphor-AXL. To determine the significance of this near universal activation of AXL in PDAC cell lines, we performed functional validation assays to determine whether AXL is a suitable PDAC target. By shRNA knockdown of AXL expression in PDAC cell lines, we found these cell lines displayed reduced soft agar growth, invasion/migration as well as reduced proliferation in culture. We also performed clinicopathological validation by IHC-TMA. We found that AXL was highly expressed in over 50% of the PDAC core samples (n=306). Collectively, these results support a role of AXL in the pathogenesis of PDAC development.

Technology Development

We have recently developed an ex vivo organotypic culture (EVOC) system which is a re reproducible, rapid and customizable culture method allowing the investigation of tumors similar to the original cancer microenvironment (recently published in Vaira et. al., 2010). To demonstrate clinical utility, the PI3K signaling using the pharmacologic inhibitor LY294002 can diminish pAKT and pS6RP in ex vivo, primary human lung tumor specimens, compared to immediately adjacent control sections in a dose-dependent manner. We find that this technique can visually demonstrate differential dependence among different cell types of signaling pathways which converge on activation of ribosomal protein S6 by phosphorylation. Whereas in cultures treated with U0126, staining of tumor cells (T) for p-S6RP is greatly attenuated but staining of stromal elements persists. In the same xenograft, LY294002 did not inhibit tumor cell staining but inhibited stromal staining. Treatment with both inhibitors results in loss of p-S6RP staining in all compartments.

Thus, the EVO culture technique will permit rapid assessment of co-inhibition of multiple RTKs using pharmacological inhibitors in combination and to monitor treatment response using well defined biomarkers.

In parallel with this, we have continued to accrue human primary PDAC samples and have established short-term tissue cultures: out of 59 tumors passaged in vitro, we have established 18expandable in vitro cultures and have snap-frozen 39 of them. We have been able to efficiently grow cells out of these tumors upon thawing, demonstrating banking possibility, which will enormously facilitate our efforts in drug testing and molecular profiling. This has occurred while continuing to grow the collection of FFPE PDACs to more than 1200 cases, one third of which are also available as frozen tumor tissue.

Preclinical model of organotypic culture for pharmacodynamic profiling of human tumors.

Vaira V, Fedele G, Pyne S, Fasoli E, Zadra G*, Baily D, Snyder E, Faversani A, Coggi G, Flavin R, Bosari S, Loda M**.

PNAS. 2010 May 4;107(18):8352-6.

*Amy & Joseph R. Perella Fellow, 2008-09.

**AICF Fellowship mentor.

The Amy & Joseph R. Perella and Debra & Claudio Del Vecchio Grant

Pancreatic Cancer Research Initiative Grant

Mark Bloomston, MD

Assistant Professor of Surgery

The Ohio State University Comprehensive Cancer Center

MicroRNAs as predictors of (pre)malignant phenotype in cystic neoplasms of the pancreas

Collaborators include Aldo Scarpa, MD, PhD, University of Verona and Carlo M. Croce, MD, The Ohio State University Comprehensive Cancer Center.

Year One Progress Report

Summary. This project focuses on the use of microRNAs harvested from pancreatic juice in patients undergoing resection for pancreatic cancer and from the fluid of patients with cystic neoplasms of the pancreas. As such, the specific aims as described in the original proposal have not changed.

Project Aims

Optimize microRNA extraction from pancreatic juice. The challenge of this project is to obtain adequate quality RNA from pancreatic cystic fluid to allow reliable detection of small microRNAs, to be able to harvest RNA in sufficient quantities for microRNA detection, and to be able to standardize small RNA for microRNA experiments. We analyzed RNA obtained in the operating room or by endoscopic ultrasound, snap frozen immediately or approximately 30 minutes after collection. RNA quality was assessed using the Agilent Bioanalyzer small RNA chip. We have learned that the small RNA chip does not provide an RNA integrity number (RIN). However it does provide a RNA concentration reading and a chromatograph in which the integrity of the RNA may be qualitatively defined. We used this in the following manner: an equal volume of RNA solution that was isolated from 250 μl of the FNA specimens and/or 250 μl of pancreas fluid that was immediately frozen after the cyst was surgically removed was subjected to Agilent bioanalysis. There were no large ribosomal RNA peaks present in any of the samples except the tissue sample that served as the positive control. From this we conclude that (a) there are no noticeable large ribosomal RNAs present in pancreatic cyst fluid and therefore a RIN may not be determined from pancreas cyst fluid and (b) the RNA in the cyst fluid samples likely represents intact, smaller RNAs and not the by-product of RNA degradation since large ribosomal RNAs are not present in this fluid even though if it was frozen immediately upon removal from the patient. It is possible that the large RNAs are degraded in cystic fluid when the fluid is in the patient.

Discovery of microRNAs as potential biomarkers of pancreatic cancer in pancreatic juice. Fluid and tissue acquisition has begun under an IRB-approved protocol. Fluid samples are obtained in the operating room at the time of surgical resection and in the endoscopy unit at the time of pancreatic cyst fluid sampling. This allows us to develop a bank of RNA from various cancer stages as well as cysts with varying malignant risks (i.e. adenoma, borderline, and carcinoma in situ) to be utilized for delineation of microRNA profiles. Determination of these profiles is planned for year two.

To determine risk of premalignancy or occult malignancy in pancreatic cystic neoplasms using microRNA profiles. Cursory analysis has been done on our first few collected cyst fluid samples to optimize our system. We elected to first look at expression of miR-21 since it is one of the most dysregulated microRNAs in all cancer, particularly pancreatic cancer. Total RNA ws isolated using the Qiagen kit. RNA was subjected to the small RNA chip (Agilent Bioanalysis). Only RNA samples wit ha visible peak on the chromatograph were deemed acceptable. The small RNA concentration determined from the Agilent small RNA chip was used to normalize the cDNA. Eight ng of RNA was reverse transcribed using the MegaPlex pool A (Applied Biosystems). The resulting cDNA ws subjected to 10 cycles of pre-amplification (Applied Biosystems) and then diluted and profiled against 385 miRNAs and reference genes (Applied Biosystems). Through a series of trila and error, we learned that the small RNA concentration is good only if the RNA is in the linear range of detection. If they are off scale (y-axis > 150) then the RNA solution needs to be diluted and re-assayed. We worked with 9 samples of cyst fluid. The RNA was diluated and re-assayed until it fell into the range of < 150 on y-axis. Eight ng of small RNA from each sample was added to the RT and the primers for several reference genes as well as miR-21 were pre-amplified. Pre-amped cDNA was subjected to qPCR for the reference genes as well as the miR-21. The expression of snoRNA U43 showed the least amount of variation among the 9 samples and was therefore used as the reference gene. The relative expression of miR-21 normalized to snoRNA U43 was calculated. Comparing the data on the samples, the expression of miR-21 increased in carcinoma in situ (CIS) and cancer compared to the borderline adenomas.

Significance of Findings

Our early work has allowed us, for the first time, to optimize protocols for obtaining RNA of sufficient quality and quantity for microRNA analysis from pancreatic juice and cystic fluid. As well, we have identified a small housekeeping gene (snoRNA U43) to normalize future qRT-PCR reactions. Finally, preliminary evaluation of a single microRNA (miR-21) shows promise in being able to discriminate between low-risk (i.e. adenoma) and high-risk (i.e. CIS and cancer) cystic lesions.

The Pancreatic Cancer Research Initiative Grants of Drs. Draetta and Bloomston were renewed for 2010-11. We would like to extend special thanks to the following donors, whose generous support is funding Pancreatic Cancer Research Initiative grants this year:

The Alexander Bodini Foundation

Amy & Joseph R. Perella

Debra & Claudio Del Vecchio