PamGene's Posters at the AACR 2015
PamGene and it's collaborators presented key posters at the Annual Meeting of the American Association of Cancer Research in Philadelphia, Pennsylvania (April 18th to 22nd, 2015). 
You can directly download PDF copies of the Posters:

4322: Kinase activity profiles distinguish papillary thyroid cancers with and without BRAF V600E mutations
Tuesday, Apr 21, 2015, 1:00 PM 5:00 PM, Section 25, Poster Board 25
Maria H. Hilhorst1, Adrienne van den Berg1, Tom van Wezel2, Tim Kievits3, Piet J.
Boender1, Rik de Wijn1, Rob Ruijtenbeek1, Wim Corver2, Hans Morreau2. PamGene International BV, 'sHertogenbosch, Netherlands; LUMC, Leiden, Netherlands; VitrOmics BV, 'sHertogenbosch, Netherlands

Background: Most differentiated non medullary thyroid cancers (DTC) are curatively treated
by surgery and radioactive iodine ablation therapy. A subset of patients shows recurrence due to a loss of iodine transport. Two main subgroups of recurrent DTC are seen: papillary thyroid cancers (PTC) with somatic BRAF mutations (V600E) and oncocytic follicular cancer. Recurrent DTC are clinically treated by multikinase inhibitors such as sorafenib, with a low affinity for BRAF V600E. Vemurafenib and dabrafenib were specifically designed against this mutant. The aim of this study was twofold: Can benign and malign DTC be classified based on kinase activity profiles? Do sorafenib (and regorafenib) show different inhibition profiles than dabrafenib?
Methods: Tissue cryosections from fresh frozen thyroid tumors were lysed. All tumor specimens were analyzed for BRAF mutations. Serine/threonine kinase (STK) activity profiles of the lysates (0.5 μg protein per array) were generated on PamChip® peptide microarrays, comprising peptide sequences from known human phosphorylation sites. The ex vivo effect of BRAF inhibitors sorafenib, regorafenib and dabrafenib on kinase activity profiles of 14 PTC’s was determined as well. Data were analysed with Bionavigator software.
Results: A classifier built on the STK kinase activity profiles of 57 thyroid cancer samples was able to classify malignant and benign tumors with a limited error rate. Leave One Out Cross Validation classified 26/35 of malignant and 17/22 of benign samples correctly. Kinase inhibition profiles of PTC’s with sorafenib and regorafenib did not discrimate V600E mutants from wild type tumors whereas with dabrafenib 34/144 peptides were identified that potentially differentiated the groups.
Conclusions: Serine/threonine kinase activity profiling appears to be able to differentiate benign and malignant thyroid tumors. Ex vivo spiking in of kinase inhibitors shows differential inhibition in tumors with a somatic BRAF mutation. Potentially, an industrial prediction platform can be envisioned for testing of novel drugs in tumor tissue. Whether individual patient responses against registered kinase inhibitors can be predicted must be investigated. 

2419: Predicting clinical response based on ex vivo drug response in renal cell carcinoma using kinase activity profiling.
Monday, Apr 20, 2015, 1:00 PM 5:00 PM, Section 22, Poster Board 4
Rob Ruijtenbeek1, Liesbeth Houkesvan Kerkhoff1, Maria Hilhorst1, Peter Mulders2,
Jeannette OosterwijkWakka2, Lambertus Kiemeney2, Egbert Oosterwijk2. PamGene International B.V., 'sHertogenbosch, Netherlands; Radboud University Medical Center, Nijmegen, Netherlands

Introduction: Sunitinib, a potent multitargeted receptor tyrosine kinase inhibitor, is the firstline treatment for metastatic renal cell carcinoma (mRCC). Because sunitinib responses and toxicity are highly variable, there is a need for biomarkers predicting sunitinib response or predicting the optimal sequence preference when using alternative tyrosine kinase inhibitors. The aim of this study was to investigate the correlation between ex vivo drug response and clinical response in renal cell carcinoma (RCC) and to explore alternative treatment options. This study has received funding from the European Union’s Seventh Framework Programme (FP7/20072013) under grant agreement no 259939.
Methods: Protein tyrosine kinase activity profiles were generated on PamChip® peptide
microarrays of lysed tumor resection tissues (15mm3) from 22 mRCC patients. The ex vivo effect of kinase inhibitors (sunitinib, axitinib, sorafenib, pazopanib, erlotinib and crenolanib) was determined and analyzed with Bionavigator software. A two-group (sunitinib responders versus nonresponders) comparison applied on the inhibition ratios identified the significantly different peptide phosphorylations. Peptides were clustered according to their correlation with clinical response.
Results: As little as 5 μg protein input (0.05 mm3 tissue) was used per kinase activity profile. The ex vivo sunitinib effect positively correlated with clinical responses especially in the subgroup which received sunitinib as 1st line treatment (4 responders vs. 4 nonresponders). 18 of the 105 peptides were significantly (p<0.05) less phosphorylated upon ex vivo sunitinib treatment in 4 responders compared to 4 nonresponders. Pathway analysis of the peptide phosphorylation patterns revealed sunitinib target involvement. Ex vivo erlotinib, axitinib or crenolanib showed reversed inhibition patterns i.e. more inhibition in the nonresponder than in the responder group. 
Conclusions: In summary, we have shown that predicting clinical response to sunitinib based on the ex vivo response to sunitinib is feasible. This requires further investigation with a larger sample set. Furthermore, the ex vivo response to a drug panel suggests that identification of novel treatment options for nonresponders might be feasible.

312: Profiling drug sensitivity and kinomic pathways utilizing a novel human tumor derived MicroTumor assay
Sunday, Apr 19, 2015, 1:00 PM 5:00 PM, Section 14, Poster Board 11
Christopher D. Willey1, Ashley N. Gilbert1, Rachael Shevin2, Catherine P. Langford1, Raj
Singh2, Joshua C. Anderson1, G. Yancey Gillespie1.
The University of Alabama at Birmingham, Birmingham, AL; Vivo Biosciences Inc., Birmingham, AL

Introduction: At present, drug screening studies are commonly performed using monolayer or spheroid culture and xenograft models of tumor cell lines. However these do not fully replicate the primary tumor’s microenvironment and fail to accurately predict clinical endpoints. Vivo Biosciences has developed a novel MicroTumor 3D matrix based assay system that emulates primary tumor multicellular growth and biology ex vivo, providing an advanced drug screening platform. We postulated that MicroTumors established from patient-derived xenograft (PDX) tumors will allow for accurate analysis of drug response and preserve molecular signaling of parent tumors. Glioblastoma multiforme (GBM), the most common primary brain malignancy, was used to test our hypothesis. MicroTumors were evaluated by comparing kinome activation profiles of GBM Micro Tumors with corresponding parental orthotopically implanted PDX; and determining single and combination treatment effects of small molecule kinase inhibitors (SMI) on GBM MicroTumors. 
Methods: We investigated 6 GBM PDX tumor lines representing the 4 known molecular subtypes: Classical (JX10, X1016, X1046); Proneural (XD456); Mesenchymal (JX22P); and Neural (JX10). Four SMIs (primary kinase target indicated) were studied: WP1066 (JAK2), selumetinib (MEK1/2), crizotinib (cMET, ALK), and cediranib (VEGFR, FLT1, FLT4, cKIT, PDGFR). MTT assays and Calcein AM imaging were used for cytotoxicity assessment and PamStation 12 Kinomic analyses were performed (UAB Kinome Core).
Results: Kinomic analyses of GBM orthotopic PDX and GBMMicroTumors revealed similar kinase signaling profiles based on comparison of commonly shared, most variant phosphopeptides. Upstream kinase analyses identified these peptides as substrates of EGFR, AXL, ZAP70 and MERTK kinases. Initial drug response studies demonstrated dose dependency and PDXspecific responses for each drug used independently informing doses for ongoing combination studies. Interestingly, the least cytotoxic drug across all 6 MicroTumors, selumetinib, did impact MicroTumor morphology observed with Calcein AM imaging.
Conclusions: We identified kinomic alterations that may correlate MicroTumor and patient tumor biology and guide the use of molecularly targeted SMIs. SMI activities towards these targets highlighted that this novel 3D translational model for GBM can provide relevant drug sensitivity information. Future studies with in vivo PDX tumors will examine the most promising SMI combinations based on MicroTumor data. We believe this two-stage approach (Microtumor screening to predict PDX sensitivities) will improve preclinical drug screening in GBM and other cancers.