Genomic and transcriptomic determinants of therapy resistance and immune landscape evolution during anti-EGFR treatment in colorectal cancer

Woolston, A.; Khan, K.; Spain, G.; Barber, L. J.; Griffiths, B.; Gonzalez Exposito, R.; Patil, Y.; Mansukhani, S.; Davies, M. N.; Furness, A.; Sclafani, F.; Peckitt, C.; Kouvelakis, K.; Ranftl, R.; Begum, R.; Rana, I.; Thomas, J.; Bryant, A.; Quezada, S.; Wotherspoon, A.; Khan, N.; Fotiadis, N.; Marafioti, T.; Powles, T.; Calvo, F.; Guettler, S.; von Loga, K.; Rao, S.; Watkins, D.; Starling, N.; Chau, I.; Sadanandam, A.; Cunningham, D.; Gerlinger, M.

doi:10.1101/448076

Home/Archive/bioRxiv

bioRxiv · 10.1101/448076

Genomic and transcriptomic determinants of therapy resistance and immune landscape evolution during anti-EGFR treatment in colorectal cancer

View original source ↗Export metadata ↓

Abstract

Anti-epidermal growth factor receptor (EGFR) antibodies (anti-EGFR-Ab) are effective in a subgroup of patients with metastatic colorectal cancer (CRC). We applied genomic and transcriptomic analyses to biopsies from 35 RAS wild-type CRCs treated with the anti-EGFR-Ab cetuximab in a prospective trial to interrogate the molecular resistance landscape. This validated transcriptomic CRC-subtypes as predictors of cetuximab benefit; identified novel associations of NF1-inactivation and non-canonical RAS/RAF-aberrations with primary progression; and of FGF10- and non-canonical BRAF-aberrations with AR. No genetic resistance drivers were detected in 64% of AR biopsies. The majority of these had switched from the cetuximab-sensitive CMS2-subtype pretreatment to the fibroblast- and growth factor-rich CMS4-subtype at progression. Fibroblast supernatant conferred cetuximab resistance in vitro, together supporting subtype-switching as a novel mechanism of AR. Cytotoxic immune infiltrates and immune-checkpoint expression increased following cetuximab responses, potentially providing opportunities to treat CRCs with molecularly heterogeneous AR with immunotherapy.

Source connections

Published article ↗

Explore related subjects

cancer biology Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Woolston, A., Khan, K., Spain, G., Barber, L. J., Griffiths, B., Gonzalez Exposito, R., Patil, Y., Mansukhani, S., Davies, M. N., Furness, A., Sclafani, F., Peckitt, C., Kouvelakis, K., Ranftl, R., Begum, R., Rana, I., Thomas, J., Bryant, A., Quezada, S., Wotherspoon, A., Khan, N., Fotiadis, N., Marafioti, T., Powles, T., Calvo, F., Guettler, S., von Loga, K., Rao, S., Watkins, D., Starling, N., Chau, I., Sadanandam, A., Cunningham, D., Gerlinger, M.. 2018-10-19. Genomic and transcriptomic determinants of therapy resistance and immune landscape evolution during anti-EGFR treatment in colorectal cancer. https://doi.org/10.1101/448076

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Polycomb Marked Promoter Methylation Gains in Aging and Cancer are Independent of Replication

DNA methylation alterations are hallmark features of aging and cancer, occurring early during tumor initiation. Yet how these changes evolve during early tumorigenesis, and whether oncogenic signaling actively drives them, remains poorly understood due to limitations in resolving this at the longitudinal and clonal level. Here we developed a genome-wide DNA methylation profiling method for low DNA amounts and applied it to single stem cell-derived clones from a BRAFV600E -inducible mouse colon organoid model, tracking nucleotide-level methylation heterogeneity over time. BRAFV600E induction produced stepwise methylation divergence consistent with progressive clonal takeover, in contrast to gradual epigenetic drift in uninduced controls. Strikingly, while the specific CpG sites altered differed between BRAFV600E and control lineages, they converged on the same genomic elements. Particularly, Polycomb-marked promoter CpG-island methylation gains were not induced or accelerated by BRAFV600E, challenging models in which oncogenic signaling or cell proliferation actively instructs de novo methylation of Polycomb-targeted promoters. Instead, these gains tracked with "biological aging" independent of replication, which was validated across independent aging, cancer, and cell-line datasets. BRAFV600E -induced clones additionally showed progressive loss of intestinal lineage-specification transcription factor programs, consistent with selection for cells adapted to oncogenic stress. Together, these findings show that epigenetic trajectories during early tumor development, are rooted in baseline, age-dependent drivers of epigenetic alterations, and that oncogenic signaling acts primarily by selecting and expanding clones bearing pre-existing, age-associated epigenetic drift rather than instructing new methylation. Our findings have implications for how promoter methylation is interpreted as a cancer biomarker and for identifying upstream aging-associated processes as targets for early intervention.

cancer biology↗

bioRxiv2026

ORTHOTOPIC TUMOUR INDUCTION PROTOCOL SHAPES TUMOUR MICROENVIRONMENT

Robust preclinical cancer models are essential for studying tumour progression and evaluating therapeutic strategies. Orthotopic implantation of tumour cells into the mammary gland is widely used to model triple-negative breast cancer (TNBC), yet substantial variation exists in the number of cells used to initiate tumours, often without explicit consideration of how this parameter shapes tumour biology. Here, we systematically assess how tumour cell number at implantation influences the tumour microenvironment using the 4T1 orthotopic TNBC model. Mammary tumours initiated with different cell numbers exhibited comparable early growth kinetics and gross size, yet diverged markedly in cellular density, hypoxia, necrosis, proliferation, vascular integrity, and immune cell infiltration. High-density tumours displayed increased HIF-1 staining and necrosis, alongside differences in vascular-associated markers and increased macrophage recruitment. Conversely, low-density tumours remained highly proliferative, less necrotic and exhibited higher PDGFR{beta} staining. Importantly, tumour cells disseminated systemically within days of implantation across all conditions, although differences in colony recovery and ex vivo expansion were not statistically significant. Together, these findings demonstrate that tumour initiating protocols are powerful determinants of tumour state, with potential implications for dissemination frequency and tumour cell fate at distant sites. Rather than representing a source of experimental variability alone, this understanding provides an opportunity to intentionally tailor orthotopic models to investigate specific stages of disease progression and therapeutic response, thereby enhancing cross-study reproducibility and translational relevance.

cancer biology↗

bioRxiv2026

Single-cell characterization of 3D CLL spheroid model that recapitulates lymph node microenvironment-driven phenotypes and in vivo drug responses to ibrutinib

Lymph node (LN) microenvironmental signals promote chronic lymphocytic leukemia (CLL) cell survival, proliferation, and treatment resistance. We previously developed a 3D culture model that recapitulates key features of this protective LN niche. Using single-cell RNA sequencing of ex vivo paired peripheral blood and LN samples, together with matched patient-derived CLL spheroids, we investigated how microenvironmental cues shape CLL phenotypes and their response to ibrutinib. A total of 3 functional CLL states (resting, activated, proliferating) were identified, with expanded populations of activated and proliferating subsets in 3D cultures. After 5 days of in vitro ibrutinib treatment, CLL cells shifted toward a resting phenotype, reflecting published responses in patients at 6 months. Activated CLL cells showed relative resistance to ibrutinib, associated with TLR9-MALT1-MYC pathway activation. The 3D model recapitulated tissue-specific T cell phenotypes observed in patients, and computational analyses predicted several T-to-CLL cell-mediated interactions supporting the LN niche. Collectively, these data reveal distinct ibrutinib responses among CLL subsets and establish the 3D model as a physiologically relevant platform.

cancer biology↗