Clinically Based Cetuximab Re-Challenge in Patients with RAS Wild-Type Metastatic Colorectal Cancer and Retrospective Analysis of Liquid Biopsies—Preliminary Data

Abstract

Background: Anti-EGFR therapy, combined with chemotherapy, represents the standard therapeutic approach for triple wild-type (KRAS/NRAS and BRAF), left-sided, microsatellite stable (MSS) metastatic colorectal cancer (mCRC). However, acquired resistance develops in approximately 50% of patients. This study evaluated the efficacy of anti-EGFR therapy re-challenge and analyzed circulating tumor DNA (ctDNA) for potential resistance mechanisms. Methods: Eleven patients with triple wild-type, MSS, HER2-negative, left-sided mCRC were included. All patients received Cetuximab with chemotherapy as the first-line treatment, with three patients subsequently receiving Cetuximab re-challenge. Twenty-one plasma samples were collected at baseline and at each response assessment for retrospective ctDNA analysis using next-generation sequencing with a 16-gene panel. Results: Genetic alterations were detected in only 14.2% of ctDNA samples. In one re-challenge patient, the KRAS: c.35G>A mutation appeared during progression. No RAS mutations were identified in four patients who progressed on first-line Cetuximab treatment. Conclusions: This preliminary study suggests that clinically based anti-EGFR re-challenge may benefit selected mCRC patients. The low detection rate of resistance-conferring mutations indicates potential alternative resistance mechanisms beyond RAS pathway alterations. Our findings, while limited by sample size and the retrospective design of ctDNA testing, contribute to the growing evidence supporting anti-EGFR re-challenge strategies in mCRC management.

1. Introduction

In patients with colorectal cancer (CRC), KRAS-activating mutations account for approximately 35–40% and are the cause of epidermal growth factor receptor (EGFR)-independent intracellular signal transduction activation, which render EGFR inhibitors like Cetuximab ineffective [1]. Thus, KRAS mutations were the first negative predictive marker to be discovered [2]. In addition, NRAS mutations account for another 3–10%. Therefore, approximately 50% of patients with CRC harbor RAS mutations [3,4].

Targeted anti-EGFR therapy, combined with chemotherapy, represents the leading therapeutic approach for triple wild-type (NRAS/KRAS and BRAF), microsatellite stable (MSS) and HER2-negative (human epidermal growth factor receptor 2) metastatic colorectal cancer (mCRC). However, the development of secondary acquired resistance in more than 50% of patients receiving anti-EGFR targeted therapy presents a significant clinical challenge [5].

Following treatment with anti-EGFR therapy, there is persistent activation of KRAS and NRAS signaling pathways as a result of emerging mutations, as well as the emergence of mutations in other genes that confer resistance [6,7,8]. These mutations render initially sensitive tumors resistant, leading to the discontinuation of anti-EGFR therapy as a result of disease progression. After the onset of acquired resistance, the response rate to anti-EGFR therapy gradually decreases. In order to explore the hypothesis that RAS and EGFR clones decay over time from the last anti-EGFR therapy application, Parseghian et al. (2019) developed a mathematical model to describe the clonal dynamics [6]. The authors found an inverse relationship between the rMAF (defined as relative mutant allele frequency) of RAS and EGFR and time since the last treatment with anti-EGFR. Their analysis showed that the decline in the RAS and EGFR rMAF after anti-EGFR withdrawal is best described by an exponential decay model; the clones exponentially decayed with half-lives of 3.4 and 6.9, and a cumulative half-life of 4.4 months, validated in an external cohort [6]. This gradual reduction in RAS-mutated cells is the biological reason for the possibility of re-challenge anti-EGFR therapy in patients who have been treated with, and responded to, anti-EGFR therapy, then progressed on anti-EGFR therapy and received a second-line anti-EGFR-free therapy such as anti-VEGF (vascular endothelial growth factor) targeted therapy in combination with different backbone chemotherapy regimens [8,9].

The levels of RAS-mutated clones may change over time, making a “re-challenge” approach to anti-EGFR therapy possible as a third and/or subsequent line of treatment. Thus, RAS-mutated tumor clones have a “pulse behavior”. Their numbers increase as a result of EGFR blockade and then gradually decrease upon cessation of anti-EGFR monoclonal antibody treatment, allowing for the tumor to become sensitive again to targeted therapy [6,8]. The reduction in mutant tumor clones has never been demonstrated in tumor biopsies; only liquid biopsies allow for this mechanism to be demonstrated [9,10]. The advantage of a liquid biopsy is that it allows for the analysis of circulating tumor DNA, which provides information on the suitability of a given patient for a given line of anti-EGFR therapy, as well as for the subsequent re-challenge of anti-EGFR therapy. The analysis of circulating tumor DNA provides insights into the genetic alterations present at both the primary tumor site and in metastatic lesions [11].

The methodologies for circulating tumor DNA detection and analysis rely on polymerase chain reaction techniques. Given the substantial cost of anti-EGFR therapy, implementing next-generation sequencing (NGS) for circulating tumor DNA analysis has proven to be cost-effective, particularly for monitoring patients via liquid biopsy during and after anti-EGFR therapy [12,13].

The aim of this preliminary study was to evaluate the potential benefit of a clinically based anti-EGFR re-challenge approach in selected patients with RAS wild-type mCRC and to conduct a retrospective liquid biopsy analysis for an initial exploration of resistance mechanisms in a small cohort of patients.

2. Results

The study included 11 patients with triple-negative (KRAS/NRAS, MSS and HER2-negative) left-sided mCRC. The cohort had a mean age of 62.5 years with a slight male predominance (55%). All patients except one had a radical resection of the primary tumor. Most patients (73%) presented with synchronous metastatic disease and most patients (72.7%) also exhibited liver-only metastases. In the remaining patients, more than one metastatic site was present, with metastatic involvement of the bones, lungs, non-regional lymph nodes, and peritoneum, in addition to the liver (

Table 1. Demographic and clinical characteristics of patients with triple-negative (KRAS/NRAS, MSS and HER2-) mCRC treated with anti-EGFR therapy *.

Patient ID	Age	Sex	Tumor Location	Metastatic Pattern	Primary Tumor Resection	Metastatic Presentation
1	48	F	Left-sided	Liver only	Yes	Synchronous
2	50	F	Left-sided	Liver only	Yes	Synchronous
3	75	F	Left-sided	Liver only	Yes	Synchronous
4	59	M	Left-sided	Liver only	Yes	Synchronous
5	73	M	Left-sided	Liver only	Yes	Synchronous
6	51	M	Left-sided	Liver only	Yes	Synchronous
7	53	M	Left-sided	Liver, Lung	Yes	Metachronous
8	68	F	Left-sided	Liver, Bone	Yes	Metachronous
9	73	M	Left-sided	Liver, Lymph nodes, Peritoneum	No	Synchronous
10	68	M	Left-sided	Liver only	Yes	Metachronous
11	70	F	Left-sided	Liver only	Yes	Synchronous

* Cetuximab was administered in combination with chemotherapy in all cases. F—female; M—male.

). All patients were treated with first-line treatment and re-administration of Cetuximab as anti-EGFR therapy in combination with chemotherapy. In our study, 2 out of 11 patients (18%) received the Cetuximab clinically based re-challenge in combination with chemotherapy. These patients (patients 1 and 2,

Table 2. Treatment response and ctDNA findings in mCRC patients receiving anti-EGFR therapy and re-challenge.

Patient ID	Baseline ctDNA	1st-Line Treatment/ Response	1st ctDNA	2nd-Line Treatment/Response	2nd ctDNA	3rd-Line Treatment/Response	3rd ctDNA	4th-Line Treatment/Response	4th ctDNA	5th-Line Treatment/Response	5th ctDNA
1	PIK3CA mutation	FOLFOX + anti-EGFR (6 cycles), PR	NAF *	FOLFIRI + Bevacizumab (7 cycles), SD **	NAF	Re-challenge: FOLFOX + anti-EGFR (4 cycles), PD	NAF	FOLFOX + anti-EGFR (7 cycles), SD	NAF	PD, NFLT ***	NAF
2	NAF	FOLFOX + anti-EGFR (6 cycles), PR	NAF	FOLFIRI + Bevacizumab (9 cycles), PD	NAF	Re-challenge: FOLFOX + anti-EGFR (6 cycles), PD	NAF	FOLFOX + anti-EGFR (4 cycles), SD	PD	NFLT	KRAS: 35G>A
3	NAF	FOLFOX + anti-EGFR (5 cycles), PD	NAF	FOLFIRI (9 cycles), PD	NAF	NFLT	-	-	-	-	-
4	NAF	FOLFOX + anti-EGFR (13 cycles), SD	NAF	Maintenance: De Gramont + anti-EGFR (4 cycles), SD	NAF	CAPIRI + Bevacizumab (7 cycles), PD	-	-	-	-	-
5	IDH2: c.418C>T	FOLFOX + anti-EGFR (4 cycles), PD	NAF	FOLFIRI + Bevacizumab (12 cycles), SD	NAF	-	-	-	-	-	-
6	NAF	FOLFOX + anti-EGFR (4 cycles), PR	NAF	FOLFOX + anti-EGFR (5 cycles), PR	NAF	FOLFOX + anti-EGFR (5 cycles), PR	NAF	FOLFOX + anti-EGFR (5 cycles), PR	PR	Maintenance: De Gramont + anti-EGFR (6 cycles)	NAF
7	NAF	FOLFOX + anti-EGFR (6 cycles), PR	NAF	FOLFOX + anti-EGFR (5 cycles), PR	NAF	-	-	-	-	-	-
8	NAF	FOLFIRI + anti-EGFR (6 cycles), SD	NAF	FOLFIRI + anti-EGFR (4 cycles), PR	NAF	-	-	-	-	-	-
9	NAF	FOLFOX + anti-EGFR (6 cycles), SD	NAF	FOLFIRI + anti-EGFR (5 cycles), SD	NAF	-	-	-	-	-	-
10	NAF	FOLFOX + anti-EGFR (6 cycles), SD	NAF	-	-	-	-	-	-	-	-
11	NAF	FOLFOX + anti-EGFR (8 cycles, conversion therapy), PR	NAF	-	-	-	-	-	-	-	-

* NAF: No alterations found; ** SD: Change of therapy due to toxicity; *** NFLT: No further lines of therapy. Response abbreviations: PR—Partial response, SD—Stable disease, PD—Progressive disease.

) received re-challenge as a third-line treatment following disease progression on anti-EGFR-free second-line therapy. It is important to note that patient 6, although receiving sequential anti-EGFR-containing regimens, did not undergo true re-challenge but rather received continuous anti-EGFR therapy with modifications of the chemotherapy backbone (from FOLFOX + Cetuximab to maintenance De Gramont + Cetuximab) due to Oxaliplatin-induced neurological toxicity.

A retrospective analysis of circulating tumor DNA (ctDNA) was conducted on 21 plasma samples; three additional samples were excluded due to poor quality. Genetic alterations were detected in only three samples (14.2%).

The PIK3CA: c.3140A>G mutation was present in the baseline study of the first patient, with a VAF of 32.4%. The second genetic alteration was the KRAS: c.35G>A mutation at the last follow-up of progression on re-challenge Cetuximab and FOLFIRI as third-line treatment in patient 2, with a VAF of 1.0%. The third mutation found was IDH2: c.418C>T in the baseline sample of patient 5, with a VAF of 27.2%.

No genetic alterations were found in the remaining 18 samples from the patients studied. Of these 18 samples, 7 were taken during disease progression and 4 were in patients who progressed while on Cetuximab therapy—as the third-line treatment in two patients and as the fourth-line treatment in one. KRAS: c.35G>A mutation was found in a patient who progressed on Cetuximab re-challenge in combination with chemotherapy; however, due to a worsening general condition and liver decompensation, active treatment was discontinued (Table 2).

3. Discussion

The clinical re-challenge approach of anti-EGFR therapy is possible when patients initially respond to anti-EGFR-based therapy (clinical benefits such as complete, partial remission, or stable disease for at least 6 months or more). Then, they progress, switch to another anti-tumor regimen that does not include anti-EGFR therapy and patients are treated long enough with this until a new progression, after which the anti-EGFR-based anti-tumor therapy is re-administered [5,8]. In our study, the approach to re-treatment with anti-EGFR was taken based on clinical factors such as the duration of anti-EGFR therapy over 4 months and the duration of second-line therapy over 4.4 months.

According to most authors, after the exposure of tumor cells to anti-EGFR therapy and a period of sensitivity and response to treatment, acquired resistance gradually arises, most likely as a result of either the presence of initially RAS-mutated tumor cells [14] or as a result of the emergence of RAS mutations in tumor cells during treatment with anti-EGFR therapy, which occurs in 48% of patients [15]. These data are not confirmed in our study.

There is evidence of a relationship between the site of metastasis in CRC and ctDNA. Patients with peritoneal metastases have lower ctDNA levels than other metastatic sites due to the plasma–peritoneal barrier [16]. In comparison with patients with peritoneal metastases, CRC patients with liver metastases have higher detectable ctDNA levels [17].

A retrospective analysis of mutational status on circulating DNA did not reveal the presence of KRAS mutations or other genetic alterations in the four patients who progressed on first-line Cetuximab and chemotherapy. The only KRAS: c.35G>A mutation was found in a patient during progression on Cetuximab and chemotherapy as a fourth-line treatment. We implemented strict quality-control measures and excluded low-quality samples to ensure the reliability of our findings. From a biological perspective, this pattern may reflect the relatively homogeneous and small patient cohort in our study. More importantly, the development of resistance to anti-EGFR therapy often results from diverse molecular events beyond RAS mutations, including alterations in the PIK3CA, BRAF and EGFR ectodomain genes, or amplifications of HER2 and MET [13,14,15]. These molecular changes may be present at diagnosis, before treatment initiation, at minimal allele frequencies in a tumor classified as RAS wild-type by standard methods, or they may emerge as subclonal events during treatment [18,19]. The low detection rate of RAS mutations during progression in our study aligns with growing evidence that resistance to anti-EGFR therapy involves multiple simultaneous mechanisms, including epigenetic changes and alterations in signaling pathways that may not be detectable by targeted sequencing panels [12,18]. The biological heterogeneity of resistance mechanisms represents an important consideration in interpreting liquid biopsy results for treatment decision making. In a young woman with unresectable liver metastases as the only site of metastatic disease in our study, we identified a PIK3CA mutation on baseline circulating DNA testing. The frequency of the activating PIK3CA mutation in patients with CRC is about 15–20% [20]. De Roock, W. et al. (2010) found that BRAF, NRAS, and PIK3CA exon 20 mutations are significantly associated with a low response rate to Cetuximab therapy [21]. There is no consensus on the relationship between PIK3CA and clinical and patho-anatomical factors in patients with CRC, with most authors not establishing a relationship between them [20,22]. The meta-analysis, compromising five studies, reported that the CRC prognosis for PIK3CA mutations in exons 9 and 20 separately showed that neither the exon 9 mutation nor the exon 20 mutation in PIK3CA were significantly associated with patient survival. Thus, the PIK3CA mutation has neutral prognostic effects on CRC overall survival (OS) and progression-free survival (PFS) [23]. There is also controversy regarding the prognostic significance of PIK3CA, with the simultaneous presence of a mutation in KRAS and PIK3CA being found to be associated with a worse prognosis and OS in patients, while the presence of PIK3CA alone is associated with a better prognosis [19,20]. These data correlate with the better prognosis in our patient with the PIK3CA mutation20. In her, first-line treatment with Cetuximab and chemotherapy was over 6 months, followed by second-line treatment with chemotherapy and Bevacizumab, also over 6 months. The re-administration of Cetuximab and chemotherapy also lasted for over six months (Table 2).

The last genetic alteration in our study is the IDH2: c.418C>T mutation at baseline in an older man with unresectable liver metastases. He had a short duration of first-line Cetuximab and FOLFOX for less than 6 months. After PD on the first-response evaluation, second-line FOLFIRI and Bevacizumab as a second-line therapy was started. The second line continued more than 16 months with a manageable toxicity profile. The second progression was only hepatic, with the appearance of new, small liver metastases. The patient continued with a third-line therapy with a combination of Lonsurf and Bevacizumab. Reevaluation is pending.

Data on the role of IDH2 mutation in the role and prognosis of patients with CRC are scarce. IDH2 normally plays a crucial role by converting isocitrate to α-ketoglutarate in mitochondria. DH1/2 mutations in CRC are uncommon but are more common in patients with BRAF p.V600E-mutated CRCs and is possibly associated with colitis-correlated CRC. More research on IDH2-mutated CRC is needed to clarify its role in the initiation and progression of CRC as well as on its role as a target for therapy [24,25,26].

4. Materials and Methods

4.1. Patients Selection and Molecular Profiling

All patients with mCRC underwent sequential molecular profiling on tumor tissues: microsatellite instability (MSI) assessment followed by KRAS/NRAS mutation testing using the Real-Time PCR kit EasyPGX^® ready KRAS/NRAS (Diatech Pharmacogenetics, Jesi, Italy). Patients with double wild-type status received additional BRAF mutation testing also performed with the Real-Time PCR kit (EasyPGX^® ready BRAF from Diatech Pharmacogenetics, Jesi, Italy); those with triple wild-type status underwent HER2 immunohistochemistry.

For MSI testing, immunohistochemistry with the following four antibodies was used: PMS2 (clone EP51, DAKO, Glostrup, Denmark, ready-to-use); MLH1 (clone ES05, DAKO, Glostrup, Denmark); MSH6 (clone EP49, DAKO, Glostrup, Denmark); and MSH2 (clone FE11, DAKO, Glostrup, Denmark). All immunostaining procedures were performed on the Automated Link Platform (DAKO, Glostrup, Denmark). Loss of expression of any of these proteins was interpreted as microsatellite instability.

For the HER2 protein overexpression assessment, the HercepTest™ (polyclonal Rabbit Anti-Human HER2 protein, DAKO, Glostrup, Denmark) was used on the DAKO Automated Link Platform according to the manufacturer’s instructions. Cases were evaluated using standard scoring criteria for colorectal cancer (0, 1+, 2+, and 3+), with scores of 3+ or 2+ with confirmed amplification by FISH considered positive for HER2 overexpression.

Patients with triple wild-type RAS and BRAF, microsatellite stable, and HER2-negative mCRC initiated treatment with anti-EGFR targeted therapy plus chemotherapy at the Medical Oncology Department of the Military Medical Academy in Sofia, Bulgaria. Treatment protocols for the first and subsequent lines followed ESMO guidelines for unresectable mCRC [4]. The anti-EGFR re-challenge approach was implemented based on the following published criteria: “Anti-EGFR re-challenge involves retreating mCRC patients with anti-EGFR antibodies if they previously achieved a response or SD during prior anti-EGFR therapy. This strategy requires an EGFR-free interval between two anti-EGFR-containing lines of treatment, ideally lasting 6 months” [17]. Circulating tumor DNA analysis was conducted retrospectively on plasma samples collected at baseline and at each tumor response assessment timepoint.

4.2. Sample Colection

Circulating DNA samples were collected prospectively, at baseline, and at each tumor response assessment timepoint to enable longitudinal molecular monitoring. Twenty-four plasma samples from 11 patients were collected for a period of 2 years (18 February 2022–25 March 2024). All patients signed informed consent forms before each sample collection, and all plasma samples were stored at −80 °C for up to two years.

4.3. DNA Extraction

DNA extraction was performed at the National Genetics Laboratory in Sofia, Bulgaria. DNA was extracted using a TANBead Maelstrom Switch 8 instrument with a TANBead cfDNA extraction kit (TANBEAD, Taoyuan, Taiwan), according to the manufacturer’s instructions. Initial quality and quantity assessment of the DNA was measured using a NanoDrop 2000 spectrophotometer. The concentrations of DNA in all samples ranged between 5 and 15 ng/mL.

4.4. Next-Generation Sequencing Analysis

Library preparation for next-generation sequencing of all samples was performed using the Myriapod^® NGS Cancer panel, which is CE IVD-certified and compatible with Illumina’s MiSeq^® platform, on which the sequencing was conducted. The Myriapod^® NGS Cancer panel (Diatech Pharmacogenetics, Jesi, Italy) encompasses 16 clinically relevant genes: ALK, BRAF, EGFR, ERBB2, FGFR3, HRAS, IDH1-2, NRAS, KIT, KRAS, MET, ROS1, PDGFRA, PIK3CA, RET, and POLE. Subsequent data analysis was performed using Myriapod^® NGS Data Analysis Software, which complies with the European Union General Data Protection Regulations (GDPR).

5. Conclusions

Our small retrospective study examined anti-EGFR re-challenge in 11 patients with RAS and BRAF wild-type metastatic colorectal cancer. All patients had MSS and HER 2 negativity. The homogeneity of the patients in this study is its advantage. The analysis of circulating tumor DNA revealed genetic alterations in only three samples (14.2%), specifically, PIK3CA, KRAS, and IDH2 mutations. We observed clinical benefits in some patients receiving anti-EGFR re-challenge, suggesting its potential utility in selected cases.

Several limitations must be acknowledged, including the small sample size (11 patients), which restricts definitive conclusions. The retrospective nature of ctDNA analysis may not fully capture dynamic molecular changes during treatment. Despite these constraints, our study contributes to the growing evidence supporting anti-EGFR re-challenge in selected mCRC patients and presents a need for more extensive prospective studies.