Although anti-EGFR treatment is recommended for patients with wild-type RAS, patients could acquire RAS mutations with the continuous use of anti-EGFR antibodies. This is hypothesized to be the consequence of sustained stress conditions induced by anti-EGFR antibodies on tumor cells and the tumor microenvironment (Rios-Hoyo et al., 2024). According to previous studies, the prevalence of RAS mutations after anti-EGFR treatment ranges from 7.5% to 84% (Diaz et al., 2012; Morelli et al., 2015; Vidal et al., 2017; Toledo et al., 2017; Normanno et al., 2018; Pietrantonio et al., 2017; Strickler et al., 2018; Kim et al., 2018; Takayama et al., 2018; Vitiello et al., 2019; Yamada et al., 2020; Lim et al., 2021; Rachiglio et al., 2022).

The BRAF (B-Raf) proto-oncogene belongs to the RAF gene family and is a vital serine/threonine protein kinase that is part of the RAS/RAF/MAPK signaling pathway. BRAF-mediated signaling involves the binding of RAF to MAPK kinases (MAPKK/MEK1/2), which regulate cell proliferation. BRAF mutations are present only in RAS wild-type patients, and approximately 9% of patients have BRAF (V600E) mutations (Tol et al., 2009). The BRAF (V600E) mutation results in an abnormally active, upstream signal-independent BRAF protein, leading to sequential activation of downstream MAPK/MEK/ERK pathways, which contributes to acquired anti-EGFR resistance (Wu et al., 2020).

Human epidermal growth factor receptor 2 (HER2) belongs to the epidermal growth factor receptor family. The binding of HER2 with other members of the family results in the formation of heterodimers and enables autophosphorylation, which can further bind to various downstream signaling molecules and activate multiple signaling pathways. In addition, the heterodimeric form of HER2 can more readily return to the cell surface in response to multiple subsequent stimuli, which further contributes to the acquisition of drug resistance (Bronte et al., 2015). Approximately 3%–4% of mCRC patients have HER2 amplification, which is mostly observed in patients with wild-type RAS/BRAF tumors (Loree et al., 2018). Several studies have shown that HER2 gene amplification may be a critical contributor to acquired anti-EGFR resistance in mCRC and may also serve as a negative marker for predicting the response to anti-EGFR therapy in mCRC patients (Barry et al., 2016; Takegawa et al., 2016).

MET is the gene encoding the MET receptor tyrosine kinase, which can be serially activated by gene amplification, mutation or autocrine or paracrine stimuli from ligands. Activated MET further triggers several pathways, such as the phosphatidylinositol-3-kinase/protein kinase B/mammalian target of rapamycin (PI3K/AKT/mTOR), RAS/RAF/MAPK/ERK, and SRC/signal transducer and activator of transcription 3 (STAT3) pathways, to induce cell invasion and inhibit apoptosis (Comoglio et al., 2008). Bardelli et al. suggested that wild-type KRAS CRC cells exhibit high levels of MET gene amplification after the acquisition of anti-EGFR resistance (Bardelli et al., 2013). MET is overexpressed in a number of malignant tumors, including CRC, and is further enhanced in tumors refractory to EGFR signaling inhibition.

The insulin-like growth factor-1 receptor (IGF-1R) is a receptor tyrosine kinase on cell membranes that binds to IGF-1 and IGF-2. The IGF-1/IGF-1R signaling pathway is important for a variety of biological processes, including CRC proliferation, differentiation, and invasion, and its overexpression in cancer cells inhibits apoptosis by activating multiple intracellular signaling pathways (Wang et al., 2023). Hyperactivation of IGF-1R impedes drug-triggered apoptotic signaling by upregulating the PI3K/AKT pathway. This desensitization of CRC cells to anti-EGFR drugs leads to drug resistance (Vigneri et al., 2015). Scartozzi et al. reported that after anti-EGFR treatment of wild-type KRAS mCRC, patients with a poor prognosis had high expression of IGF-1 (Scartozzi et al., 2010).

Phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) is the p110-α catalytic subunit of PI3K, whose activation further initiates the PI3K/AKT/mTOR pathway to regulate various cell biological activities, such as proliferation, metabolism and apoptosis (Janku et al., 2018). Studies have shown that PIK3CA mutations are most commonly found in exons 20 and 9, but only PIK3CA exon 20 mutations are related to the outcome of drug resistance in wild-type KRAS CRC patients receiving anti-EGFR therapy (De Roock et al., 2010; Mao et al., 2012).

A recent study by Johnson RM and colleagues revealed that AT-rich interactive domain-containing protein 1 (ARID1A) gene mutation is significantly related to resistance to anti-EGFR treatment in CRC patients. The ARID1A gene is an epigenetic regulator located on chromosome 1. It works together with the multisubunit chromatin-remodeling complex SWI/SNF chromatin remodeling complex to regulate and remodel the chromatin structure. This modulation of gene expression and cellular function helps maintain cellular homeostasis and prevent excessive signaling pathway activation. A decrease in SWI/SNF activity after ARID1A mutation induces EGFR activation, leading to resistance to anti-EGFR drugs in CRC patients (Johnson et al., 2022).

PRSS is a serine protease secreted by tumor cells that promotes angiogenic and invasive behavior by degrading the extracellular matrix. PRSS is highly expressed in cetuximab-resistant CRC cells (Tan et al., 2020). Three distinct PRSSs were identified in this study, among which PRSS1, a pancreatic protease that is expressed mainly in the pancreas and involved in the digestive process, is significantly associated with cetuximab resistance. The possible mechanism is that PRSS1 impairs the inhibitory effect of cetuximab on the PI3K/AKT and MEK/ERK pathways; PRSS1 is also able to cleave monoclonal antibodies such as cetuximab, trastuzumab, and bevacizumab, thereby reducing their efficacy and ultimately leading to resistance. In addition, the team reported that serine peptidase inhibitor Kazal type 1 (SPINK1) inhibits the cleavage of cetuximab by PRSS1 and that SPINK1 in combination with cetuximab can resensitize cetuximab-resistant tumor cells (Tan et al., 2020). Therefore, pharmacological inhibition of PRSS1 may be an effective solution for patients with CRC who are resistant to anti-EGFR therapy and have high serum PRSS1 levels.

In a recent retrospective cohort study, phospholipase C γ 1 (PLCγ1) was shown to be highly expressed in CRC patients treated with cetuximab (Cruz-Duarte et al., 2022). Additionally, a high level of PLCγ1 contributes to cetuximab resistance in RAS wild-type colorectal cancer cells. Before resistance is acquired, tumor cells cannot maintain the downstream signaling pathway under cetuximab treatment; however, when PLCγ1 is overexpressed, the PLCγ1-SH2-containing protein tyrosine phosphatase 2 (SHP2) tandem structural domain activates ERK and AKT in parallel through the noncatalytic role of SHP2, thus ensuring continued activation of downstream signaling and leading to resistance to cetuximab. The study also revealed that, in combination with an SHP2 inhibitor, cetuximab exerts an antitumor effect on PLCγ1-dependent cetuximab-resistant cancer cells (Cruz-Duarte et al., 2022).

Luo’s research team reported that occult metastatic tissues from CRC patients presented an increase in methyltransferase-like protein 4 (METTL4), an enzyme that catalyzes the N6-methyladenosine (m6A) modification of RNA molecules. Moreover, METTL14-mediated m6A modification stabilizes pleckstrin homology-like domain family member 2 (PHLDB2) mRNA by preventing its degradation, leading to upregulation of the protein level of PHLDB2. The overexpression of PHLDB2 contributes to colorectal cancer cell resistance to cetuximab and CRC cell metastasis through two mechanisms. First, PHLDB2 stably binds to EGFR through the Arg1163 site, thereby promoting its nuclear translocation to continuously activate proliferative signaling. Second, PHLDB2 competes with ubiquitin for binding to EGFR to prevent EGFR degradation by ubiquitin (Luo et al., 2022). Researchers also reported that the replacement of the arginine at position 1,163 in the PHLDB2 sequence with an alanine (R1163A mutation) eliminated the ability of PHLDB2 to promote EGFR nuclear localization, and CRC cells harboring the R1163A mutation in PHLDB2 were no longer resistant to anti-EGFR drugs (Luo et al., 2022).

After long-term anti-EGFR treatment, the microsatellite stability/mismatch repair function of CRC cells gradually decreases, and these cells can gradually exhibit a mismatch repair defective (dMMR) phenotype; in turn, these cells exhibit high microsatellite instability (MSI-H) and generate irreversible resistance to anti-EGFR drugs. The incidence of MSI-H CRC in patients is 12%–15% (Haraldsdottir, 2017), and MSI-H CRC is associated with shorter survival (Innocenti et al., 2019). MSI-H tumors are associated with many gene mutations, which can lead to the coding of antigens with a high degree of immunogenicity, potentially improving the immune response of tumor-infiltrating lymphocytes (Luchini et al., 2019). Therefore, immunotherapy is important for patients with MSI-H disease. Immune checkpoint inhibitors (ICIs), including the PD-1 inhibitors nivolumab and pembrolizumab and the anti-cytotoxic T lymphocyte-associated protein 4 (CTLA-4) monoclonal antibody ipilimumab, have achieved good results in MSI-H CRC patients (Petrelli et al., 2020).

Noncoding RNAs, which include mainly microRNAs (miRNAs), long noncoding RNAs (lncRNAs), and circular RNAs (circRNAs), lack the ability to translate into RNAs but exert vital regulatory effects on diverse biological processes (Sommerauer and Kutter, 2022). In recent years, studies have revealed that noncoding RNAs participate in the acquisition of anti-EGFR resistance in CRC. microRNAs can bind to the 3′ untranslated region (UTR) of mRNAs to modulate the translation of a specific protein, thus exerting their function. Chen et al. reported that after treatment with cetuximab, miRNA-216b was significantly downregulated, whereas Beclin-1 and its associated adaptive autophagy were significantly activated in CRC cells, which led to resistance to cetuximab treatment (Chen et al., 2016). Moreover, a bioinformatics study revealed that miRNA-216b targeted the 3′UTR of Beclin-1 mRNA to negatively regulate the translation of Beclin-1 protein expression, which was further confirmed by a luciferase reporter gene assay (Chen et al., 2016). Zhou reported that miRNA-133b was downregulated in CRC cells and that restoration of miRNA-133b inhibited the proliferation and invasion of CRC cells. Moreover, miRNA-133b targeted the 3′UTR of EGFR and negatively regulated EGFR expression; thus, miRNA-133b insufficiency led to anti-EGFR resistance. The authors further reported that the combination of a miRNA-133b mimic and cetuximab had a synergistic anti-CRC effect (Zhou et al., 2015).

Unlike miRNAs, lncRNAs and circRNAs serve as decoys of miRNAs to further exert their functions. Zhang et al. established cetuximab-resistant Caco-2 cells and reported that the lncRNA cetuximab resistance-associated RNA transcript 16 (CRART16) was upregulated in these cells. The lncRNA CRART16 serves as a decoy of miRNA-371a-5p to further regulate the downstream V-Erb-B2 Erythroblastic Leukemia Viral Oncogene Homolog 3. Moreover, lncRNA CRART16 overexpression was associated with the differential expression of genes enriched in the MAPK signaling pathway (Zhang et al., 2020). Another study reported that the circRNA interferon gamma receptor 2 (IFNGR2) promoted the proliferation, migration, and invasion of CRC cells both in vitro and in vivo. By sponging miRNA-30b, the circRNA IFNGR2 increases the mRNA expression of wild-type and mutant-type KRAS, which further activates the AKT signaling pathway. Moreover, the overexpression of the circRNA IFNGR2 reduces the sensitivity of CRC cells to cetuximab treatment both in vitro and in vivo (Zhang et al., 2023).

The tumor microenvironment has an important impact on the processes of tumorigenesis, development, invasion and metastasis and is also an important target for tumor therapy (Xiao and Yu, 2021). Cancer-associated fibroblasts (CAFs) and inflammatory cytokines are the two main factors that mediate cetuximab resistance in the tumor microenvironment. CAFs are the predominant stromal cell type in the tumor microenvironment, accounting for up to 50%–90% of the total cells. CAFs can promote tumor cell growth, invasion and metastasis by secreting a variety of biologically active molecules, such as growth factors, cytokines and proteases (Biffi and Tuveson, 2021). Garvey CM et al. reported that colorectal cancer patients have increased CAF abundance in the tumor microenvironment after anti-EGFR treatment. Moreover, CAFs secrete exogenous epidermal growth factor (EGF) to maintain the conduction state of the MAPK signaling pathway under anti-EGFR therapy, leading to drug resistance in CRC cells (Garvey et al., 2020).

Inflammatory cytokines in the tumor microenvironment also mediate resistance to anti-EGFR drugs. The inflammatory cytokines interleukin-1α (IL-1α), IL-1β, and IL-8 and the activated inflammatory transcription factor NF-κB are responsible for the activation of signaling pathways that overcome the drug response and lead to tumor cell resistance after anti-EGFR therapy, and these cytokines may reduce the response of other sensitive cells to cetuximab through paracrine or autocrine effects (Gelfo et al., 2016). Inhibition of IL-1 receptor signaling via the use of the recombinant decoy TRAP IL-1 decreased CRC cell proliferation and inactivated the MAPK/AKT pathway, suggesting the potential of inhibiting inflammatory cytokine signaling as a strategy for reversing anti-EGFR resistance (Gelfo et al., 2018).

Exosomes are extracellular vesicles secreted by all cells and have diameters ranging from 40 to 100 nm; these vesicles contain DNA, RNA, proteins, and lipids as cargos. Cells can also absorb exosomes and simultaneously absorb the cargo contained in the exosomes. In recent years, exosomes have received increasing interest, and exosome-mediated intracellular communication has been shown to participate in various biological processes (Tang et al., 2025). Studies have also shown that exosome-mediated intracellular communication plays a vital role in acquired resistance to anti-EGFR treatment in CRC. A study conducted by Mason et al. revealed that under nutrient stress, cetuximab-resistant, KRAS-mutant CRC cells released amphiregulin-carrying exosomes with the monomeric G protein Rab11a. When receiving amphiregulin-carrying exosomes with the monomeric G protein Rab11a, KRAS wild-type CRC cells acquired resistance to cetuximab. However, free amphiregulin had no such effect, suggesting the role of exosome-mediated intracellular communication in acquired anti-EGFR resistance (Mason et al., 2024). Wei et al. reported that exosomes derived from multidrug-resistant CRC cells induced resistance to cetuximab in CRC cells that were previously sensitive to cetuximab through the activation of the PI3K/AKT signaling pathway. Moreover, exosomes derived from multidrug-resistant CRC cells promoted the stemness of cetuximab-sensitive CRC cells. Similar findings were also reported in xenograft models (Wei et al., 2023). A study conducted by Morimura et al. reported that exosomes derived from cetuximab-resistant CRC cells induced drug resistance in cetuximab-sensitive CRC cells. In this study, sera exosomes from CRC patients who did or did not respond to cetuximab and from healthy volunteers were collected. Cetuximab resistance in CRC cells is induced by exosomes collected from patients who do not respond to cetuximab. In contrast, exosomes collected from healthy volunteers promoted the efficacy of cetuximab in CRC cells (Morimura et al., 2018).

Post-transcriptional modification is a process in which proteins undergo chemical modifications, such as phosphorylation, ubiquitination, methylation, and acetylation, to perform specific biological functions. Phosphorylation is the most widely studied post-transcriptional modification, and the common protein phosphorylation involved in acquired anti-EGFR resistance in CRC is the activation of the RAF/RAS/MEK/ERK and PI3K/AKT/mTOR signaling pathways (Li et al., 2020). However, other post-transcriptional modifications associated with acquired anti-EGFR resistance in CRC are rarely reported. Lam et al. reported that glycosylation of EGFR at asparagine residue 361 reduces the sensitivity of CRC cells to the anti-EGFR treatment necitumumab by promoting dimerization of EGFR and activation of the EGFR signaling pathway (Lam et al., 2024). A study by Yang et al. revealed that STAT1 was highly expressed in CRC cells resistant to cetuximab. Moreover, insufficient Smurf1-mediated ubiquitination of STAT1 leads to increased proliferation, migration, and invasion of CRC cells, as well as insensitivity to cetuximab (Yang et al., 2024). Rodrigues et al. reported that terminal ⍺2,6-sialylation of EGFR downregulated EGFR expression and suppressed the activation of EGFR, which further increased resistance to cetuximab in CRC cells (Rodrigues et al., 2021). These studies highlight that post-transcriptional modifications, in addition to widely studied phosphorylation, play important roles in acquired resistance to anti-EGFR treatment in CRC cells, which could serve as a potential strategy to reverse this resistance.

Acquiring tumor tissues via biopsy is a common sampling strategy for investigating biomarkers for predicting acquired anti-EGFR resistance. Using this method, genetic mutations in RAS, BRAF, and PIK3CA; amplifications in HER2 and MET; and rearrangements in neurotrophic tyrosine receptor kinase (NTRK)/ROS proto-oncogene 1 (ROS1)/anaplastic lymphoma kinase (ALK)/RET proto-oncogene have been associated with the acquisition of anti-EGFR resistance (Ciappina et al., 2025).

In recent years, researchers have explored novel biomarkers for predicting acquired anti-EGFR resistance in patients with RAS wild-type CRC. For example, Cardone et al. included 136 patients with RAS wild-type CRC who received first-line anti-EGFR therapy. AXL positivity in tumor tissues was associated with worse PFS (6.2 months in the AXL-positive cohort vs. 12.1 months in the AXL-negative cohort) and OS (23.0 months in the AXL-positive cohort vs. 35.8 months in the AXL-negative cohort) (Cardone et al., 2020). Martini et al. reported that in tumor tissues from 82 patients with RAS wild-type CRC who received FOLFIRI plus cetuximab, the EPHA2 tyrosine kinase receptor was detected in 55 tissue samples, and a high level of EPHA2 was associated with a worse PFS (8.6 vs. 12.3 months) and progression rate (29% vs. 9%) (Martini et al., 2019).

Noncoding RNAs also possess the potential to predict acquired anti-EGFR resistance in patients with advanced CRC. Anandappa et al. collected 91 core biopsies from 45 patients with metastatic CRC before single agent anti-EGFR treatment. In situ hybridization was used to detect miRNA-31-3p in these biopsies; the results revealed that patients with low expression of miRNA-31-3p exhibited an improved response to anti-EGFR antibody, as well as prolonged PFS and overall survival (OS), before and after adjustment for age, sex, and the sidedness of CRC. Moreover, the expression of miRNA-31-3p does not vary after anti-EGFR treatment (Anandappa et al., 2019). However, Boisteau et al. reported that miRNA-31-3p detected in tumor biopsies was not associated with treatment response in patients with right-sided metastatic CRC who received chemotherapy plus anti-EGFR antibody. Nevertheless, prolonged OS was found in patients with low levels of miRNA-31-3p compared with those with high levels of miRNA-31-3p (Boisteau et al., 2022). Since slight differences exist between these two studies, the prognostic value of miRNA-31-3p should be further verified. Fiala et al. retrospectively enrolled 46 patients with metastatic CRC who were treated with cetuximab or panitumumab and detected the levels of miRNA-125b, let-7c, miRNA-99a, miRNA-17, miRNA-143, and miRNA-145 in tumor biopsies. A high level of miRNA-125b was associated with a better ORR, high levels of miRNA-125b and let-7c, as well as low levels of miRNA-17, were associated with a better disease control rate (DCR), and a high level of miRNA-125b was associated with improved PFS and OS (Fiala et al., 2020). Peng et al. used a lncRNA array to identify potential lncRNAs associated with acquired anti-EGFR resistance in patients with advanced CRC. Nine of the 212 lncRNAs were differentially expressed between patients who achieved disease control and those without a treatment response after anti-EGFR treatment. Among the nine lncRNAs, 5 were associated with PFS, and the lncRNA POU5F1P4 was confirmed to be downregulated in CRC cells resistant to cetuximab (Peng et al., 2018).

Liquid biopsy is a noninvasive biospecimen collection method that involves the collection of blood samples for various tumor indicators, including cell-free DNA (cfDNA), circulating tumor DNA (ctDNA) and microRNAs (miRNAs). In particular, cfDNA analysis can detect multiple mechanisms of coexisting drug resistance that may be missed by single-focus tumor tissue biopsies, making it an advantageous assay for assessing tumor heterogeneity (Blakely et al., 2017). The latest studies have utilized liquid biopsy-based ctDNA and cfDNA to predict treatment response and survival in CRC patients receiving anti-EGFR antibody-based treatment, further contributing to the personalized management of CRC. For example, the CAPRI-2 GOIM trial included 192 patients with metastatic CRC harboring wild-type RAS and planned to receive FOLFIRI plus cetuximab. According to the results of baseline liquid biopsy-comprehensive genomic profiling of potential acquired anti-EGFR resistance genes (including RAS, BRAF, EGFR, PIK3CA, MAP2K1, MET, RET, ALK, ROS1, NTRK, NFR, and FGFR mutations, as well as HER2 amplification) in ctDNA with FoundationOne (F1) CDx and F1 Liquid (F1L) CDx (324 genes), patients were stratified into a wild-type group and a mutation group. The ORR was 54.5% for the mutation group and 78.1% for the wild-type group. The median PFS was 8.68 months for the mutation group compared with 12.35% for the wild-type group (Ciardiello et al., 2025). The PLATFORM-B study included 100 patients with metastatic CRC harboring wild-type RAS and who were scheduled to receive first-line chemotherapy plus cetuximab. ctDNA at baseline and early after treatment was analyzed with next-generation sequencing, and mutations in RAS, MEK, and BRAF were negatively correlated with PFS. Moreover, a decrease in mutations in RAS, MEK, and BRAF after treatment was associated with increased response and PFS (Vidal et al., 2023). Toledo et al. enrolled 25 patients with metastatic CRC harboring wild-type RAS and who received first-line FOLFIRI plus cetuximab. A 2-year follow-up was conducted, during which a routine sample was collected for cfDNA detection via the BEAMing technique. Wild-type KRAS, NRAS, PIK3CA, and BRAF are associated with prolonged treatment response, whereas mutations in these genes are associated with acquired resistance (Toledo et al., 2017). These studies highlight the potential of liquid biopsy as a tool for predicting treatment response and survival in patients with CRC harboring wild-type RAS and receiving anti-EGFR antibody-based treatment. When therapeutic regimens are formulated to reverse acquired drug resistance, liquid biopsy may be useful because patients may have multiple mechanisms of resistance, and regular identification of cfDNA is critical for continued treatment after resistance. Second, the ease of accessibility, low invasiveness, and ease of monitoring of liquid biopsies make them an even more attractive testing technique.

Although noncoding RNAs from tumor biopsies have potential predictive value for anti-EGFR resistance, the process of acquiring tumor biopsies could cause harm to patients. On the other hand, circulating RNAs, despite easy and harmless collection of samples, can be degraded by RNAse, resulting in inaccurate prediction. Exosomes, however, can protect RNAs from degradation because of their bilayer structure. Studies have reported that noncoding RNAs derived from exosomes have good predictive value in cancer patients (Zhu et al., 2024; Jing et al., 2024; Linares-Rodriguez et al., 2025). However, relevant studies have rarely been conducted on anti-EGFR resistance in CRC. Yang et al. included 53 patients with advanced CRC who received cetuximab treatment and collected blood samples for exosome separation. Exosomal lncRNA urothelial carcinoma-associated 1 was upregulated in patients with poor treatment response compared with those with objective response (Yang et al., 2018).