The genomic landscape of response to EGFR blockade in colorectal cancer

Abstract

Colorectal cancer is the third most common cancer worldwide, with 1.2 million patients diagnosed annually. In late-stage colorectal cancer, the most commonly used targeted therapies are the monoclonal antibodies cetuximab and panitumumab, which prevent epidermal growth factor receptor (EGFR) activation. Recent studies have identified alterations in KRAS and other genes as likely mechanisms of primary and secondary resistance to anti-EGFR antibody therapy. Despite these efforts, additional mechanisms of resistance to EGFR blockade are thought to be present in colorectal cancer and little is known about determinants of sensitivity to this therapy. To examine the effect of somatic genetic changes in colorectal cancer on response to anti-EGFR antibody therapy, here we perform complete exome sequence and copy number analyses of 129 patient-derived tumour grafts and targeted genomic analyses of 55 patient tumours, all of which were KRAS wild-type. We analysed the response of tumours to anti-EGFR antibody blockade in tumour graft models and in clinical settings and functionally linked therapeutic responses to mutational data. In addition to previously identified genes, we detected mutations in ERBB2, EGFR, FGFR1, PDGFRA, and MAP2K1 as potential mechanisms of primary resistance to this therapy. Novel alterations in the ectodomain of EGFR were identified in patients with acquired resistance to EGFR blockade. Amplifications and sequence changes in the tyrosine kinase receptor adaptor gene IRS2 were identified in tumours with increased sensitivity to anti-EGFR therapy. Therapeutic resistance to EGFR blockade could be overcome in tumour graft models through combinatorial therapies targeting actionable genes. These analyses provide a systematic approach to evaluating response to targeted therapies in human cancer, highlight new mechanisms of responsiveness to anti-EGFR therapies, and delineate new avenues for intervention in managing colorectal cancer.

Extended Data Figure 1. EGFR signaling pathway genes involved in cetuximab resistance or sensitivity

Altered cell surface receptors or members of RAS or PI3K pathways identified in this study are indicated. Somatic alterations related to resistance or sensitivity are highlighted in red or green boxes, respectively. The percentages indicate the fraction of KRAS WT tumors containing the somatic alterations in the specified genes. For the following genes a subset of alterations are indicated: PDGFRA kinase domain mutations; EGFR ecto- and kinase domain mutations and amplifications.

Extended Data Figure 2. Pan-HER monoclonal antibody mixture binds epitopes different from those recognized by cetuximab

A) The H383 (green) residue and the S484/G485 residues (light blue) in EGFR domain III are critical for the binding of Pan-HER anti-EGFR antibodies 1277 and 1565, respectively. Antibodies 1277 and 1565 bind to an epitope distinct from that of cetuximab, which may contribute to the superior tumor growth inhibition in the presence of mutations at residue 465. Mutations identified in this study affecting G465 (red) and the S492 amino acid (yellow) previously reported to confer cetuximab resistance are shown for reference. Similarly to mutations affecting S492, the alterations at 465 that we identified in this study (G465R and G465E) involve changes from a nonpolar uncharged side chain to large electrically charged arginine or glutamic acid residues, respectively, and predict resistance to cetuximab. B) Critical EGFR amino acids selectively recognized by both cetuximab and panitumumab as determined by phage screening are shown in blue and include P373, K467, P411, K489, D379, F376. Residue G465 is in close proximity to K467 and other residues that have been shown to influence the binding of both cetuximab and panitumumab.

Extended Data Figure 3. Expression of IRS2 according to response categories in tumorgraft models

Results were obtained using Illumina-based oligonucleotide microarrays in 100 tumorgrafts that had no mutations in the KRAS, NRAS, BRAF or PIK3CA genes. Response categories are defined in the main text. OR, objective response; SD, stable disease; PD, progressive disease. P < 0.001 for OR compared to PD and SD compared to PD by one-way ANOVA and Bonferroni’s multiple comparison test. IRS2 expression values are shown in Supplementary Table 10.

Extended Data Figure 4. Functional studies of genetic alterations associated with cetuximab response

a, b, Ectopic expression of mutations that correlated with resistance to EGFR blockade prevented responsiveness to cetuximab. NCI-H508 cells expressing EGFR G465E (a, left panel) or DDK-tagged MAP2K1 K57N (b, left panel) were refractory to cetuximab in dose-dependent viability assays after 6 days of treatment. Results are the means ± SD of two independent experiments performed in biological triplicates (n = 6) for EGFR G465E and three independent experiments performed in biological triplicates (n = 9) for MAP2K1 K57N compared to mock vector controls. Biochemical responses of NCI-H508 EGFR G465E (a, right panel) and NCI-H508 MAP2K1 K57N (b, right panel) treated with cetuximab for 24h were documented by western blot analyses. c, Genetic silencing of IRS2 (IRS2 shRNA) in NCI-H508 cells reduced sensitivity to cetuximab in dose-dependent viability assays (left panel). Results are the means ± SD of two independent experiments performed in biological triplicates (n = 6). In biochemical studies using western blot analyses (right panel), IRS2 knockdown attenuated EGF-dependent activation of AKT (P-AKT) and ERK (P-ERK). Cells were treated for 10 min with the indicated concentrations of EGF. Tubulin was used as a loading control. Western blots for total EGFR, ERK and AKT proteins were run with the same lysates as those used for anti-phosphoprotein detection but on different gels.

Extended Data Figure 5. Signaling consequences of FGFR inhibition in FGFR1-amplified CRC477

Immunohistochemistry with the indicated antibodies and morphometric quantitations of representative tumors at the end of treatment. Results are the means ± SD of 5 fields (40×) from two tumors for each experimental point (n = 10). Scale bar, 300 μm. P-ERK, phospho-ERK; P-S6, phospho-S6. NT, not treated (vehicle); CET, cetuximab; BGJ, BGJ398. * P < 0.05; ** P < 0.01 by two-tailed Student’s t-test.

Extended Data Figure 6. Signaling consequences of EGFR inhibition in EGFR mutant (V843I) CRC334

Immunohistochemistry with the indicated antibodies and morphometric quantitations of representative tumors at the end of treatment. Results are the means ± SD of 5 fields (40×) from two tumors for each experimental point (n = 10). Scale bar, 300 μm. P-ERK, phospho-ERK; P-S6, phospho-S6. NT, not treated (vehicle); CET, cetuximab; AFA, afatinib. ** P < 0.01; *** P < 0.001 by two-tailed Student’s t-test.

Extended Data Figure 7. Signaling consequences of PDGFR inhibition in PDGFRA mutant (R981H) CRC525

Immunohistochemistry with the indicated antibodies and morphometric quantitations of representative tumors after acute treatment (4 hours after imatinib and 24 hours after cetuximab administration). Results are the means ± SD of 5 fields (40×) from two tumors for each experimental point (n = 10). Scale bar, 300 μm. P-ERK, phospho-ERK; P-S6, phospho-S6. NT, not treated (vehicle); CET, cetuximab. ** P < 0.01 by two-tailed Student’s t-test.

Extended Data Figure 8. Signaling consequences of MEK1 inhibition in MAP2K1 mutant (K57KN) CRC343

Immunohistochemistry with the indicated antibodies and morphometric quantitations of representative tumors at the end of treatment. Results are the means ± SD of 5 fields (40×) from two tumors for each experimental point (n = 10). Scale bar, 300 μm. P-ERK, phospho-ERK; P-S6, phospho-S6. NT, not treated (vehicle); CET, cetuximab; AZD, AZD6244; SCH, SCH772984. *** P < 0.001 by two-tailed Student’s t-test.

Extended Data Figure 9. Signaling consequences of EGFR inhibition in EGFR mutant (G465E) CRC104

Immunohistochemistry with the indicated antibodies and morphometric quantitations of representative tumors at the end of treatment. Results are the means ± SD of 5 fields (40×) from two tumors for each experimental point (n = 10). Scale bar, 300 μm. P-ERK, phospho-ERK; P-S6, phospho-S6. NT, not treated (vehicle); CET, cetuximab; AFA, afatinib; PAN, panitumumab. n.s., not significant; ** P < 0.01 by two-tailed Student’s t-test.

Extended Data Figure 10. Signaling consequences of EGFR inhibition in EGFR mutant (G465R) CRC177

Immunohistochemistry with the indicated antibodies and morphometric quantitations of representative tumors at the end of treatment. Results are the means ± SD of 5 fields (40×) from two tumors for each experimental point (n = 10). Scale bar, 300 μm. P-ERK, phospho-ERK; P-S6, phospho-S6. NT, not treated (vehicle); CET, cetuximab; AFA, afatinib; PAN, panitumumab. * P < 0.05; *** P < 0.001 by two-tailed Student’s t-test.

Figure 1. Schematic diagram of integrated genomic and therapeutic analyses

To examine the effect of genomic alterations on sensitivity to anti-EGFR blockade, we performed whole exome and copy number analyses of 129 early passage tumorgrafts and targeted analyses of 55 patient tumors, all of which were KRAS wild-type (top box). Twenty-two of tumorgrafts were from patients that had been previously treated with anti-EGFR therapy. 116 of these tumorgrafts were evaluated for response to cetuximab in preclinical therapeutic trials (bottom left box). Integration of genomic and therapeutic information was used to identify candidate resistance and response genes, and to design preclinical trials using novel compounds to overcome resistance to EGFR blockade (bottom right box).

Figure 2. Effect of cetuximab treatment on growth of colorectal tumors with different somatic alterations

Waterfall plot of tumor volume changes after cetuximab treatment, compared with baseline, in 116 KRAS wild-type tumorgrafts. Alterations related to therapeutic resistance or sensitivity are shown in the indicated colors (complete list of alterations are in Tables S3, S4 and S6). For the following genes a subset of alterations are indicated: MET amplification; FGFR1 amplification; PDGFRA kinase domain mutations; BRAF V600 hotspot mutations; PTEN homozygous deletion or truncating mutations; PIK3CA exon 20 mutations; EGFR ecto- and kinase domain mutations and amplifications. The maximum threshold for tumor growth was set at 200%.

Figure 3. Genetic alterations involved in secondary resistance to anti EGFR therapy

a, The location of mutations in EGFR ectodomain are shown including G465 (red) and the S492 residue known to confer cetuximab resistance (yellow). b, Evolution of EGFR mutations in two CRCs with acquired resistance to cetuximab. Cetuximab-naïve samples were sequenced to investigate the presence of EGFR G465 mutations (red) prior to treatment. For each sample, the fraction of mutant tags is indicated. c, As a control for tumor cellularity, for each lesion the fraction of TP53 mutant reads (vertical axis) was plotted against the fraction of reads with EGFR ectodomain mutations (horizontal axis).

Figure 4. Therapeutic intervention in preclinical trials to overcome resistance to anti-EGFR antibody blockade

Tumor growth curves in tumorgraft cohorts from individual patients with a, FGFR1 amplification (CRC477) b, EGFR kinase mutation (CRC334) c, PDGFRA R981H mutation (CRC525) d, MAP2K1 K57N mutation (CRC343) and e, f, EGFR ectodomain mutations (e, CRC104 and f, CRC177) treated with placebo or targeted treatments. Mean tumor volumes ± standard error of the mean are shown (n = 5 mice per group for CRC525 and CRC177 and n = 6 mice per group for all other models). a, b, combo versus cetuximab, P < 0.01; c, combo versus cetuximab, not significant; d, SCH772984+AZD6244 versus either monotherapy, P < 0.01; e, f, afatinib, Pan-HER or panitumumab+afatinib versus panitumumab, P < 0.01. Statistical analyses was performed by two-way ANOVA.