PROJECT SUMMARY/ABSTRACT
Colorectal cancer (CRC) is a major cause of cancer deaths which is curable if detected early. Unfortunately,
many CRC tumors are diagnosed at more advanced stages where cure rates remain low with 20% five-year
survival. The EGFR/RAS signaling pathway is a strong driver of CRC when constitutively activated via mutations
(KRAS ~40%, BRAF ~10%, NRAS ~5%), but targeted molecular inhibitors, when used alone (EGFRi, BRAFi,
MEKi or ERKi), have yielded disappointing survival benefit. To date, it remains difficult to identify the subset of
patients who will meaningfully benefit from EGFR/RAS pathway-targeted therapies, likely due to complex
resistance mechanisms. A widely held view is that resistance to targeted therapies is due to pre-existence of
rare, drug-resistant mutant cells (subclones) that expand and repopulate the tumor following initial successful
treatment leading to permanent, clonal resistance. Recently, however, adaptive mutability (AM) of CRC in
response to targeted therapies has been reported as a novel, early resistance mechanism. This hypothesis
proposes that blocking critical, hard-wired signaling pathways causes a “stress” reaction characterized by the of
genotoxic, reactive oxygen species (ROS) and the suppression of DNA mismatch repair enzymes. These early
drug-induced changes effectively cause a sub-clonal, hypermutable state that facilitates the eventual emergence
of permanently drug-resistant clones. Microsatellite instability (MSI) and other mutator states (i.e. POLE-mutated
tumors) are now known to provide a greater probability of checkpoint inhibitor responses. Thus, we hypothesize
that the AM induced by EGFR/RAS pathway inhibitors may lead to rapid, sub-clonal, enhanced mutability, and
a higher load of neoantigens producing an “MSI-like” state. This hypermutability, if detectable, would portend
the higher likelihood of response to checkpoint inhibitors. Traditional analysis of tumor mutation burden (TMB)
by NGS assessment of a cancer gene panel would not likely detect genome-wide sub-clonal TMB (scTMB).
Standard TMB assays detect bountiful, clonally-expanded mutations fixed in the tumor from the time of its
founding. Conversely, we propose that the mutations resulting from a rapid, drug-induced mutator phenotype
will likely be present at low variant allele frequencies (i.e. subclonal) that are below the limit of detection of
standard NGS approaches. In this proposal, we will test the hypothesis that scTMB occurs in response to
targeted therapies, and can be detected using an ultrasensitive and highly specific genome-wide Duplex DNA
sequencing assay as an indicator (biomarker) of a mutator state induced by targeted therapies. We propose that
the presence of scTMB would predict responses to checkpoint inhibitors, and suggest a benefit to targeted
therapies combined with checkpoint inhibitors. These studies will lead to a better molecular understanding of
the newly identified “adaptive mutability” resistance mechanism, and potentially provide a solid rationale for the
use of combination targeted- and immune-therapies in biomarker-selected patient subpopulations to improve
response.