Colorectal cancer (CRC) remains the third leading cause of cancer-related deaths in the United States. Despite important improvements in therapeutic strategies, the overall 5y survival remains around 65%. Thus, a large number of patients receive negligible benefit from current treatments. Consequently, there is a dire need to establish novel prognostic markers that can identify patients who will fail to benefit from the standard of care; 5- fluorouracil (5FU)-based chemotherapy. Here we seek to validate our exciting preliminary findings indicating that the ribosomal protein, Rpl22-Like1 (Like1), prognosticates poor survival in patients with stage III CRC. Alterations in ribosomal proteins have long been associated with increased cancer risk, but neither the mechanism by which they confer risk nor their utility as prognostic markers has been assessed. We have determined that Like1 promotes development and progression of T-acute lymphoblastic leukemia, when its expression is induced by loss of its paralog and antagonist, Rpl22. Rpl22 loss causes Like1 induction because Rpl22 normally acts to post-transcriptionally repress Like1 expression. Like1 induction upon Rpl22 loss not only facilitates cancer progression in leukemia, but also appears to do so in CRC. Indeed, in a small pilot study of 23 patients, we found that 20% of patients with stage III CRC express high levels of Like1 protein. Importantly, these patients exhibit low overall 5y survival (25%), compared to patients with low Like1 expression (95%). Preclinical studies conducted in parallel indicate that Rpl22-deficient (Like1-high) mice are highly resistant to weekly, myeloablative treatments with 5FU. We seek to test our hypothesis that patients with Like1-high CRC tumors exhibit reduced overall 5y survival, due to resistance to the 5FU-based chemotherapy used in treating CRC. Our rationale for positing 5FU-resistance as a possible mechanism of Like1 action is that Rpl22-deficient (Like1-high) mice are highly resistant to weekly, myeloablative treatments with 5FU. We will test our hypothesis in two Aims. In the first Aim, we will assess the level of Like1 expression in samples from a large cohort (338) of stage III CRC patient samples and then evaluate its relationship to overall survival (OS), disease free survival (DFS), and other variables such as stage (IIIa, IIIb, IIIc), histological grade, and microsatellite instability (MSI) status.
In Aim 2, we will assess the molecular basis for Like1 induction and then identify the pathways (e.g., NFkB/Lin28b) that are dysregulated in association with Like1 induction, as these are likely to facilitate the aggressiveness of Like1-high CRC. Identification of these pathways is critical, as they may serve as novel therapeutic vulnerabilities in Like1-high patients, for whom standard 5FU-based chemotherapy regimens appear to be ineffective. Through these efforts, we expect to determine if Like1 induction is prognostic for reduced OS and DFS in stage III CRC and identify the pathways coregulated with Like1 that may be responsible for the chemoresistant behavior of Like1-high tumors. This is particularly critical for stage III CRC, where there is a dearth of prognostic biomarkers to guide clinical care.
Prognostic markers in CRC that predict outcome and guide therapy decisions are desperately needed. The identification of such markers would increase therapeutic efficacy by identifying responders, while sparing those patients who are unlikely to benefit from the harmful side effects of 5FU-based chemotherapy.