Precision Rechallenge: Liquid Biopsy CGP Refines Patient Selection in Metastatic Colorectal Cancer

Highlights

• The randomized Phase 2 CAVE-2 GOIM trial found that adding the PD-L1 inhibitor avelumab to cetuximab rechallenge did not significantly improve progression-free survival (PFS) or overall survival (OS) compared to cetuximab monotherapy in patients with RAS/BRAF wild-type (WT) metastatic colorectal cancer (mCRC).

• The study confirms the critical role of liquid biopsy comprehensive genomic profiling (CGP) using FoundationOne Liquid CDx to identify patients eligible for anti-EGFR rechallenge.

• A pre-planned ‘negative hyper-selection’ analysis revealed that patients without mutations in a specific 10-gene panel (including KRAS, NRAS, BRAF, EGFR ECD, PIK3CA, and others) achieved significantly better clinical outcomes.

• Negative hyper-selection resulted in a median OS of 15.0 months compared to 11.1 months for those with detectable resistance mutations, supporting the integration of liquid biopsy into the mCRC continuum of care.

Background: The Evolution of Anti-EGFR Rechallenge

Metastatic colorectal cancer (mCRC) remains a significant challenge in clinical oncology, particularly in the refractory setting. For patients with RAS and BRAF wild-type tumors, anti-epidermal growth factor receptor (EGFR) monoclonal antibodies, such as cetuximab and panitumumab, are pillars of frontline and second-line therapy. However, resistance almost inevitably develops, often driven by the emergence of clones harboring mutations in the MAPK pathway, such as KRAS or NRAS mutations, or alterations in the EGFR extracellular domain (ECD).

The concept of ‘rechallenge’ is based on the dynamic nature of clonal evolution. When anti-EGFR therapy is discontinued, the selective pressure on the tumor is removed. Previous research, including the CRICKET and CHRONOS trials, has suggested that sensitive clones may repopulate the tumor over time—a phenomenon known as ‘clonal decay’ of resistant variants. This window of opportunity allows for the re-introduction of anti-EGFR agents. The CAVE-2 GOIM trial was designed to build upon this hypothesis, exploring whether the addition of immune checkpoint inhibition could further enhance the efficacy of this strategy.

Study Design and Methodology

The CAVE-2 GOIM trial (NCT05291156) was a multicenter, randomized Phase 2 study conducted across several Italian oncology centers. The trial enrolled patients with RAS/BRAF WT mCRC who had achieved a clinical benefit (complete response, partial response, or stable disease for at least 4 months) from previous frontline anti-EGFR-based therapy and subsequently progressed. Importantly, patients had to be refractory to fluoropyrimidines, oxaliplatin, and irinotecan.

A key feature of the study was the mandatory use of liquid biopsy for patient screening. Comprehensive genomic profiling (CGP) was performed using the FoundationOne Liquid CDx platform. Patients were randomized in a 2:1 ratio to receive either:

• Arm A: Cetuximab (400 mg/m2 loading dose, then 250 mg/m2 weekly) plus Avelumab (10 mg/kg every 2 weeks).

• Arm B: Cetuximab monotherapy (same dosing schedule).

The primary endpoint was Overall Survival (OS). Secondary endpoints included Progression-Free Survival (PFS), Objective Response Rate (ORR), and safety. A pre-planned exploratory analysis focused on ‘hyper-selection,’ categorizing patients based on the absence (negative) or presence (positive) of specific resistance-associated genomic variants in baseline circulating tumor DNA (ctDNA).

Key Findings: Does Immunotherapy Add Value?

Between August 2022 and December 2024, the trial screened 328 patients, ultimately randomizing 156. The results indicated that the addition of avelumab did not provide the synergistic boost researchers had hoped for. In the intention-to-treat population, median PFS was 5.3 months in Arm A versus 4.3 months in Arm B (HR: 0.78; 95% CI: 0.55-1.10; P=0.158). The primary endpoint, median OS, showed no significant difference between the two arms: 14.8 months for the combination versus 12.9 months for monotherapy (HR: 1.00; 95% CI: 0.65-1.52; P=0.983).

While the combination did not outperform monotherapy, the results for the cetuximab-only arm reinforced the value of rechallenge itself in a molecularly selected population, providing a meaningful survival benefit in a heavily pre-treated cohort.

The Significance of Negative Hyper-Selection

The most impactful data from CAVE-2 emerged from the molecular sub-analysis. The researchers evaluated a ‘negative hyper-selection’ criteria, defined as the absence of pathogenic variants in KRAS, NRAS, BRAF, EGFR ECD, PIK3CA (exon 20), MAP2K1, AKT1, MET, PTEN, and ERBB2. Of the 156 randomized patients, 124 (approximately 79%) met these criteria.

Patients in the negative hyper-selection group fared significantly better across all efficacy metrics, regardless of which treatment arm they were in:

• Median PFS: 5.35 months vs. 3.65 months in the positive group (HR: 0.62; P=0.017).

• Median OS: 15.0 months vs. 11.1 months in the positive group (HR: 0.61; P=0.037).

• Overall Response Rate: Improved significantly in patients without detectable resistance clones.

This data highlights that the presence of even low-frequency resistant clones at the time of rechallenge—undetectable by standard tissue biopsy but visible via ctDNA—is a strong negative predictor of response.

Expert Commentary and Clinical Implications

The failure of avelumab to improve outcomes in this setting reflects a broader challenge in colorectal cancer: most mCRC tumors are ‘immunologically cold’ (microsatellite stable/proficient mismatch repair). While some preclinical models suggested that anti-EGFR therapy might induce immunogenic cell death or increase immune cell infiltration, the CAVE-2 results suggest these effects are insufficient to overcome the immunosuppressive microenvironment of mCRC when paired with a PD-L1 inhibitor alone.

However, the trial is a resounding victory for the ‘liquid biopsy first’ approach. In clinical practice, deciding when to rechallenge a patient has often been a matter of clinical judgment based on the ‘treatment-free interval.’ CAVE-2 provides robust evidence that genomic profiling of ctDNA is a more accurate tool. By filtering out the 20% of patients who harbor latent resistance mutations, clinicians can avoid ineffective treatments and focus on therapies with a higher probability of success.

The study also validates the use of broad-panel CGP over single-gene testing. While KRAS and NRAS are the primary culprits, the inclusion of EGFR ECD mutations and MET amplifications in the hyper-selection panel captures a more complete picture of the tumor’s resistance landscape.

Conclusion: A New Standard for Rechallenge

The CAVE-2 GOIM trial clarifies the role of anti-EGFR rechallenge in the modern era of precision medicine. While the addition of avelumab is not recommended, cetuximab monotherapy remains a viable and effective option for a subset of mCRC patients. The trial underscores that ‘molecular selection’ is the key to unlocking this benefit. Implementing ctDNA-based CGP at the point of progression on second- or third-line therapy should now be considered a standard component of the continuum of care for RAS/BRAF WT mCRC.

Funding and Clinical Trial Registration

The CAVE-2 GOIM trial was supported by the Gruppo Oncologico dell’Italia Meridionale (GOIM). ClinicalTrials.gov Identifier: NCT05291156.

References

Ciardiello D, Martini G, Bielo LB, et al. Cetuximab rechallenge in molecularly selected metastatic colorectal cancer: the randomized CAVE-2 GOIM trial. Ann Oncol. 2025 Dec 22:S0923-7534(25)06334-3. doi: 10.1016/j.annonc.2025.12.014. PMID: 41443411.