36Cancer genome sequencing has uncovered substantial complexity in the mutational 37 landscape of tumors. Given this complexity, experimental approaches are necessary to 38 establish the impact of combinations of genetic alterations on tumor biology and to uncover 39 genotype-dependent effects on drug sensitivity. In lung adenocarcinoma, EGFR mutations co-40 occur with many putative tumor suppressor gene alterations, however the extent to which 41 these alterations contribute to tumor growth and their response to therapy in vivo has not 42 been explored experimentally. By integrating a novel mouse model of oncogenic EGFR-driven 43Trp53-deficient lung adenocarcinoma with multiplexed CRISPR-Cas9-mediated genome editing 44 and tumor barcode sequencing, we quantified the effects of inactivation of ten putative tumor 45 suppressor genes. Inactivation of Apc, Rb1, or Rbm10 most strongly promoted tumor growth. 46Unexpectedly, inactivation of Lkb1 or Setd2 -which were the strongest drivers of tumor growth 47 in an oncogenic Kras-driven model -reduced EGFR-driven tumors growth. These results were 48 consistent with the relative frequency of these tumor suppressor gene alterations in human 49 EGFR-and KRAS-driven lung adenocarcinomas. Furthermore, Keap1 inactivation reduced the 50 sensitivity of tumors to osimertinib in the EGFR L858R ;p53 flox/flox model. Importantly, in human 51 EGFR/TP53 mutant lung adenocarcinomas, mutations in the KEAP1 pathway correlated with 52 decreased time on tyrosine kinase inhibitor treatment. Our study highlights how genetic 53 alterations can have dramatically different biological consequences depending on the 54 oncogenic context and that the fitness landscape can shift upon drug treatment. 55 56 57 58 During tumor evolution, cancer cells accumulate alterations in oncogenes and tumor 59 suppressor genes, which contribute to many of the hallmarks of cancer 1 . Despite their extensive 60 genomic complexity, tumors are frequently classified based on the presence of a single 61 oncogenic driver mutation, while the function of co-incident tumor suppressor gene alterations 62 is largely ignored. There is emerging evidence that the interplay between oncogenic drivers and 63 tumor suppressor genes may influence tumor fitness and impact treatment response 2 . 64However, the combinatorially vast landscape of genomic alterations makes it difficult, except in 65 the most extreme cases, to glean information about the epistatic interactions between tumor 66 suppressor genes and oncogenes 3 . This complexity makes inferring the relationship between 67 genotype and therapy response even more tenuous. 68Recently, high-throughput, tractable systems that combine autochthonous mouse modeling 69 and genome editing have been developed to directly uncover the functional consequences of 70 genetic alterations during tumorigenesis in vivo 4-10 . However, very few studies have 71 investigated the biological consequences of inactivating tumor suppressor genes in the context 72 of different oncogenic drivers in vivo, a...