A framework for target discovery in rare cancers.

While large-scale functional genetic screens have uncovered numerous cancer dependencies, rare cancers are poorly represented in such efforts and the landscape of dependencies in many rare cancers remains obscure. We performed genome-scale CRISPR knockout screens in an exemplar rare cancer, TFE3- translocation renal cell carcinoma (tRCC), revealing previously unknown tRCC-selective dependencies in pathways related to mitochondrial biogenesis, oxidative metabolism, and kidney lineage specification. To generalize to other rare cancers in which experimental models may not be readily available, we employed machine learning to infer gene dependencies in a tumor or cell line based on its transcriptional profile. By applying dependency prediction to alveolar soft part sarcoma (ASPS), a distinct rare cancer also driven by TFE3 translocations, we discovered and validated that MCL1 represents a dependency in ASPS but not tRCC. Finally, we applied our model to predict gene dependencies in tumors from the TCGA (11,373 tumors; 28 lineages) and multiple additional rare cancers (958 tumors across 16 types, including 13 distinct subtypes of kidney cancer), nominating potentially actionable vulnerabilities in several poorly-characterized cancer types. Our results couple unbiased functional genetic screening with a predictive model to establish a landscape of candidate vulnerabilities across cancers, including several rare cancers currently lacking in potential targets.

Effects of immunosuppressive treatment on patient outcomes after immune checkpoint inhibitor-related gastrointestinal toxicity.

PURPOSE: Immune checkpoint inhibitors (ICIs) are increasingly used in the treatment of certain cancers but cause immune-related adverse events (irAEs). Gastrointestinal irAEs may necessitate extended periods of steroid use and the initiation of selective immunosuppressive therapy (SIT) which could theoretically counteract the effect of ICIs. In this study, we aim to explore the impact of immunosuppression use and duration on cancer progression and progression-free survival (PFS). METHODS: This is a single-center retrospective review exploring cancer outcomes in patients taking ICIs who developed gastrointestinal irAEs within 1 year of ICI initiation. Cancer outcome and progression free survival (PFS) were measured and compared by using IBM SPSS Statistics 26. RESULTS: Of the 116 patients included in this study, 69 received immunosuppression to treat irAEs. The occurrence of colitis and use of immunosuppression for colitis were associated with less cancer progression by later assessment (p < 0.05). Shorter durations of steroids with or without SIT for colitis were associated with less cancer progression within the study window than no immunosuppression (p < 0.05). Immunosuppression has no effect on PFS (p < 0.05). CONCLUSION: Our study reported shorter duration of steroid treatment for colitis may be associated with less cancer progression. Though the use of immunosuppression was not found to impact PFS, this may be confounded by the presence of colitis, which is known to improve cancer outcomes and could mask any negative impact of immunosuppression on survival. It may be preferable to limit long-term immunosuppression in the treatment of immune-mediated colitis to minimize potential complications. Prospective studies are needed to clarify this relationship, and treatments that abrogate the need for immunosuppression in these patients such as fecal microbiota transplantation.