Predicting clinical response to anticancer drugs remains a major challenge in cancer treatment. Emerging reports indicate that the tumour microenvironment and heterogeneity can limit the predictive power of current biomarker-guided strategies for chemotherapy. Here we report the engineering of personalized tumour ecosystems that contextually conserve the tumour heterogeneity, and phenocopy the tumour microenvironment using tumour explants maintained in defined tumour grade-matched matrix support and autologous patient serum. The functional response of tumour ecosystems, engineered from 109 patients, to anticancer drugs, together with the corresponding clinical outcomes, is used to train a machine learning algorithm; the learned model is then applied to predict the clinical response in an independent validation group of 55 patients, where we achieve 100% sensitivity in predictions while keeping specificity in a desired high range. The tumour ecosystem and algorithm, together termed the CANScript technology, can emerge as a powerful platform for enabling personalized medicine.

KRAS mutation status can distinguish between metastatic colorectal carcinoma (mCRC) patients who may benefit from therapies that target the epidermal growth factor receptor (EGFR), such as cetuximab. However, patients whose tumors harbor mutant KRAS (codons 12/13, 61 and 146) are often excluded from EGFR-targeted regimens, while other patients with wild type KRAS will sometimes respond favorably to these same drugs. These conflicting observations suggest that a more robust approach to individualize therapy may enable greater frequency of positive clinical outcome for mCRC patients. Here, we utilized alive tumor tissues in ex-vivo platform termed CANscript, which preserves the native tumor heterogeneity, in order to interrogate the antitumor effects of EGFRtargeted drugs in mCRC (n = 40). We demonstrated that, irrespective of KRAS status, cetuximab did not induce an antitumor response in a majority of patient tumors. In the subset of non-responsive tumors, data showed that expression levels of EGFR ligands contributed to a mechanism of resistance. Transcriptomic and phosphoproteomic profiling revealed deregulation of multiple pathways, significantly the Notch and Erbb2. Targeting these nodes concurrently resulted in antitumor efficacy in a majority of cetuximab-resistant tumors. These findings highlight the importance of integrating molecular profile and functional testing tools for optimization of alternate strategies in resistant population.Colorectal cancer (CRC) is the third most commonly diagnosed cancer worldwide with a 5-year survival rate of less than 10% 1 . An important molecular target implicated in disease progression is Epidermal Growth Factor Receptor (EGFR) signaling, which after ligand binding triggers two main pathways: the RAS-RAF-MAPK cascade leading to cell proliferation, survival, invasion and metastasis; and the PI3K-PTEN-AKT pathway which controls cell survival, motility and neo-angiogenesis 2 . Notably, EGFR is overexpressed in 60-80% of colorectal tumors 3 . Current chemotherapeutic options include 5FU + leucovorin, XELOX, XELIRI, FOLFOX and FOLFIRI which are combinations of capecitabine, 5-fluorouracil, leucovorin and oxaliplatin or irinotecan. Two classes of anti-EGFR monoclonal antibodies (mAbs) are at present prescribed in combination with conventional chemotherapy for the treatment of CRC. However the underlying problem of using cetuximab (a chimeric-IgG1mAb) is that it has only 8.8% efficacy when used in monotherapy, and 22.9% when used in combination therapy for

The PI3K/AKT/mTOR pathway is an important signaling axis that is perturbed in majority of cancers. Biomarkers such as pS6RP, GLUT1, and tumor FDG uptake are being evaluated in patient stratification for mTOR pathway inhibitors. In the absence of a clear understanding of the underlying mechanisms in tumor signaling, the biomarker strategy for patient stratification is of limited use. Here, we show that no discernible correlation exists between FDG uptake and the corresponding Ki67, GLUT1, pS6RP expression in tumor biopsies from patients with head and neck cancer. Correlation between GLUT1 and pS6RP levels in tumors was observed but elevated pS6RP was noticed even in the absence of concomitant AKT activation, suggesting that other downstream molecules of PI3K/AKT and/or other pathways upstream of mTOR are active in these tumors. Using an ex vivo platform, we identified putative responders to rapamycin, an mTOR inhibitor in these tumors. However, rapamycin did not induce antitumor effect in the majority of tumors with activated mTOR, potentially attributable to the observation that rapamycin induces feedback activation of AKT. Accordingly, treatment of these tumors with an AKT inhibitor and rapamycin uniformly resulted in abrogation of mTOR inhibition-induced AKT activation in all tumors but failed to induce antitumor response in a subset. Phosphoproteomic profiling of tumors resistant to dual AKT/mTOR inhibitors revealed differential activation of multiple pathways involved in proliferation and survival. Collectively, our results suggest that, in addition to biomarker-based segregation, functional assessment of a patient's tumor before treatment with mTOR/AKT inhibitors may be useful for patient stratification. Cancer Res; 73(3); 1118-27. Ó2013 AACR.

Ex vivo human tumor models have emerged as promising, yet complex tools to study cancer immunotherapy response dynamics. Here, we present a strategy that integrates empirical data from an ex vivo human system with computational models to interpret the response dynamics of a clinically prescribed PD-1 inhibitor, nivolumab, in head and neck squamous cell carcinoma (HNSCC) biopsies (N = 50). Using biological assays, we show that drug-induced variance stratifies samples by T helper type 1 (Th1)-related pathways. We then built a systems biology network and mathematical framework of local and global sensitivity analyses to simulate and estimate antitumor phenotypes, which implicate a dynamic role for the induction of Th1-related cytokines and T cell proliferation patterns. Together, we describe a multi-disciplinary strategy to analyze and interpret the response dynamics of PD-1 blockade using heterogeneous ex vivo data and in silico simulations, which could provide researchers a powerful toolset to interrogate immune checkpoint inhibitors.

Background: Emerging clinical evidence using immunotherapy in recent years has demonstrated its power to suppress tumor growth by releasing the brakes on the immune system. For example, blockade of immune checkpoints, such as PD-1, has revolutionized treatment options for patients with aggressive cancers such as head and neck squamous cell carcinoma (HNSCC). However, clinical responses to PD-1 inhibition vary widely among patients while majority of them do not show any anti-tumor response. Multiple FDA-approved drugs against the same immune checkpoints have resulted in globally distinct outcomes in the clinic. There is a huge unmet need to understand these disparities at the individual patient level and to maximize the clinical benefits of these agents. Methods: Here, we employed a patient-derived ex vivo model, CANScript™ (Majumder B et al. Nature Commun 2015 Feb 27;6:6169 and Goldman A et al. Nature Commun 2015 Feb 11;6:6139), which recreates the native 3D tumor microenvironment, autocrine-paracrine dynamic and response to therapy by incorporating fresh tumor tissue and autologous immune cells with immunotherapy agents. Utilizing late stage HNSCC (N=50) we interrogated phenotypic response to two FDA-approved PD-1 inhibitors, Pembrolizumab (KEYTRUDA) and Nivolumab (OPDIVO). To do this, we used a comprehensive panel of immunological assays to evaluate changes in the immune compartments by flowcytometry and immunohistochemistry (primarily CD8, CD45, FOXP3, CXCR4, CD68, PDL1, PD1), multiplex cytokine profiling (IL6, IL8, IFN-g, IL10, IL12, Perforin, GranzymeB), along with functional/phenotypic effects including tumor proliferation, histological changes and cell death. Results: The data demonstrated that CANScript™ preserves the tumor-immune contexture and native heterogeneity across different clinical stages and patients. Importantly, we observed that PD-1 blockade resulted in patient-specific therapeutic response, which was characterized by differential distribution and maintenance of infiltrating CD8+ and CD4+ lymphocytes, distinct patterning of cytokines linked to functional dysregulation, and changes in tumor proliferation and apoptosis. Interestingly, data suggest that both Pembrolizumab and Nivolumab act on the same immune network axis but trigger functionally diverse phenotypes in the tumor immune compartment and distinct antitumor effects within an individual patient tumor. Conclusion: Together, these findings demonstrate the utility of CANScript™ as an ex vivo platform to predict therapeutic response of immune checkpoint inhibitors at the individual patient level. It also highlights mechanistic variations that could impact clinical outcome of these agents having the same molecular target. Such information can re-shape our understanding of patient selection and rational combinations for novel immune checkpoint inhibitors. Citation Format: Padhma Radhakrishnan, Vasanthakumar Sekar, Nilesh Brijwani, Priyanka Chevour, Babu Balakrishnan, Dency D Pinto, Muthusami Oliyarasi, Debapriya G. Mehrotra, Manjusha Biswas, Sabitha K S, Kodaganur S. Gopinath, Arkasubhra Ghosh, M s Ganesh, Ashok M. Shenoy, Saravanan Thiyagarajan, Biswanath Majumder, Aaron Goldman. A patient derived ex vivo platform CANScript™ predicts distinct therapeutic outcomes to multiple PD-1 checkpoint inhibitors in single tumor biopsies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3681. doi:10.1158/1538-7445.AM2017-3681