Decoding Inter-tumor Heterogeneity: Dissecting Patient-specific Signaling Signatures Towards Personalized Cancer Therapy

Introduction

World-wide efforts are invested, aiming to develop effective strategies for personalized cancer medicine. Clinics today often utilize lists of altered biomarkers to assign a treatment for each patient. Yet these may lack critical information about the protein-protein relationships in each tumor. We aim to explore the data space of cancer patients and find a way to accurately classify tumors according to their signaling signatures, such that every single tumor can be mapped precisely and unambiguously according to the molecular aberrations that it harbors.

Methods

We study cancer from an information-theoretic point of view. Using a large dataset obtained from ~3500 patients of 11 types we decipher the altered protein network structure in each tumor, namely the patient-specific signaling signature. Each signature can harbor several altered protein subnetworks. We suggest that simultaneous targeting of central proteins from every altered subnetwork is essential to efficiently disturb the altered signaling flow in each tumor. In contrast to common statistical analyses that divide the patient population into categories and then assign new patients to these pre-defined groups, our approach addresses individual patients unbiasedly. We experimentally validate our ability to dissect sample-specific signaling signatures with high resolution by rationally designing combinations of targeted drugs against cancer cell lines.

Results and Discussion

We unraveled a surprisingly simple order that underlies the extreme apparent complexity of tumor tissues, demonstrating that only 17 protein altered protein subnetworks characterize ~3500 patients of 11 types. Each tumor was described by a specific subset of 1-4 subnetworks out of 17. We show that the majority of tumor-specific sets, named signaling signatures, are extremely rare, and are shared by only 5 tumors or less, supporting a personalized, comprehensive study of tumors in order to design the optimal combination therapy for every patient.

We experimentally validated our approach using 10 different cancer cell lines. Using information-theoretic surprisal analysis each cell line was assigned a signaling signature and a combined drug therapy. We show in-vitro and in-vivo that therapies predicted by us for triple negative breast cancer models achieved higher rates of killing than the clinically prescribed chemotherapy.

Conclusion

We present a novel approach to deal with the inter-tumor heterogeneity and to break down the high complexity of cancer systems into simple, easy to crack, patient-specific signaling signatures, that guide the rational design of patient-tailored drug therapies.