Molecular and Clinical Determinants of Acquired Resistance and Treatment Duration for Targeted Therapies in Colorectal Cancer

Abstract

Purpose: Targeted therapies have improved outcomes for patients with metastatic colorectal cancer, but their impact is limited by rapid emergence of resistance. We hypothesized that an understanding of the underlying genetic mechanisms and intrinsic tumor features that mediate resistance to therapy will guide new therapeutic strategies and ultimately allow the prevention of resistance.

Experimental design: We assembled a series of 52 patients with paired pretreatment and progression samples who received therapy targeting EGFR (n = 17), BRAF V600E (n = 17), KRAS G12C (n = 15), or amplified HER2 (n = 3) to identify molecular and clinical factors associated with time on treatment (TOT).

Results: All patients stopped treatment for progression and TOT did not vary by oncogenic driver (P = 0.5). Baseline disease burden (≥3 vs. <3 sites, P = 0.02), the presence of hepatic metastases (P = 0.02), and gene amplification on baseline tissue (P = 0.03) were each associated with shorter TOT. We found evidence of chromosomal instability (CIN) at progression in patients with baseline MAPK pathway amplifications and those with acquired gene amplifications. At resistance, copy-number changes (P = 0.008) and high number (≥5) of acquired alterations (P = 0.04) were associated with shorter TOT. Patients with hepatic metastases demonstrated both higher number of emergent alterations at resistance and enrichment of mutations involving receptor tyrosine kinases.

Conclusions: Our genomic analysis suggests that high baseline CIN or effective induction of enhanced mutagenesis on targeted therapy underlies rapid progression. Longer response appears to result from a progressive acquisition of genomic or chromosomal instability in the underlying cancer or from the chance event of a new resistance alteration.

Figure 1.

Emergent alterations on treatment. Oncoprint showing genes with emergent alterations. Alterations in these genes at baseline are indicated in the top, and emergent alterations in these genes at resistance are indicated at the bottom. Also shown are the most recurrently altered genes at baseline: TP53 and APC.

Figure 2.

Baseline extent of disease and metastatic sites and TOT. A, Box plot showing TOT by extent of disease (two or fewer sites of disease vs. three or more sites of disease). B, Box plot showing TOT by presence of liver metastases. C, Number of emergent oncogenic alterations (as filtered by oncoKB) per progression sample by presence of liver metastases at baseline. D, Frequency of oncogenic alterations by gene (left) and pathway (right) by the presence of liver metastases at baseline. E, Number of emergent VUS per progression sample by presence of liver metastases at baseline. F, Frequency of VUS by pathway by the presence of liver metastases at baseline. G, Lollipop plots showing location of non-oncogenic alterations in EGFR and KRAS. H, Frequency of VUS in the EGFR and KRAS genes by liver metastases with hashed shading to indicate the proportion of alterations that have been associated with interference with drug binding. I, Diagram of sites selected for sequencing (left), sequencing results (center), and putative resistance changes (right) detected in a patient who underwent rapid autopsy after progression on the triplet regimen of the RAF inhibitor encorafenib + anti-EGFR antibody cetuximab + PI3K inhibitor alpelisib. met, metastasis.

Figure 3.

Baseline amplifications and TOT. A, Box plot showing TOT by presence of amplification (Amp.) detected in baseline tissue by NGS. B, Emergent amplifications in patients with baseline MAPK amplifications in their tumors. C, Microscopy image of a slide from a progression biopsy subjected to FISH. Marked areas show tissue sections with KRAS or MET amplifications. D, Representative images from FISH for KRAS (top) and MET (bottom) DNA. Probes for the genes of interest are in red color, and probes for the corresponding chromosomal centromeres are in green color. E, Representative image from immunofluorescence for cGAS. Imaged small red dots correspond to micronuclei.

Figure 4.

Molecular changes at resistance and associations with TOT. A, Box plot showing TOT by number of emergent putative resistance alterations (four or fewer vs. five or more emergent alterations). B, Box plot showing TOT by type of emergent putative resistance alteration (mutation only vs. copy-number alteration). C, Plot showing number and type of emergent putative resistance alterations detected over TOT in each of the 52 study patients. D, Plot showing emergent alterations in patients with repeated cfDNA sampling. E, Representative images of cGAS immunofluorescence in baseline and progression samples. CNA, copy-number alteration; MUT, mutation.