Advertisement

Long-term Survival Update and Extended RAS Mutational Analysis of the CAIRO2 Trial: Addition of Cetuximab to CAPOX/Bevacizumab in Metastatic Colorectal Cancer

  • Sanne ten Hoorn
    Affiliations
    Amsterdam UMC location University of Amsterdam, Center for Experimental and Molecular Medicine, Laboratory for Experimental Oncology and Radiobiology, Amsterdam, The Netherlands

    Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands

    Oncode Institute, Utrecht, The Netherlands
    Search for articles by this author
  • Linda Mol
    Affiliations
    Clinical Research Department, Netherlands Comprehensive Cancer Center (IKNL), Nijmegen, The Netherlands
    Search for articles by this author
  • Dirkje W. Sommeijer
    Affiliations
    Amsterdam UMC location University of Amsterdam, Center for Experimental and Molecular Medicine, Laboratory for Experimental Oncology and Radiobiology, Amsterdam, The Netherlands

    Amsterdam UMC location University of Amsterdam, Department of Medical Oncology, Amsterdam, The Netherlands

    Flevohospital, Department of Internal Medicine, Almere, The Netherlands
    Search for articles by this author
  • Lisanne Nijman
    Affiliations
    Amsterdam UMC location University of Amsterdam, Center for Experimental and Molecular Medicine, Laboratory for Experimental Oncology and Radiobiology, Amsterdam, The Netherlands

    Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands

    Oncode Institute, Utrecht, The Netherlands
    Search for articles by this author
  • Tom van den Bosch
    Affiliations
    Amsterdam UMC location University of Amsterdam, Center for Experimental and Molecular Medicine, Laboratory for Experimental Oncology and Radiobiology, Amsterdam, The Netherlands

    Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands

    Oncode Institute, Utrecht, The Netherlands

    Amsterdam UMC location University of Amsterdam, Amsterdam Gastroenterology & Metabolism, Amsterdam, The Netherlands
    Search for articles by this author
  • Tim R. de Back
    Affiliations
    Amsterdam UMC location University of Amsterdam, Center for Experimental and Molecular Medicine, Laboratory for Experimental Oncology and Radiobiology, Amsterdam, The Netherlands

    Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands

    Oncode Institute, Utrecht, The Netherlands
    Search for articles by this author
  • Bauke Ylstra
    Affiliations
    Amsterdam UMC location Vrije Universiteit Amsterdam, Department of Pathology, Cancer Center Amsterdam, Amsterdam, The Netherlands
    Search for articles by this author
  • Erik van Dijk
    Affiliations
    Amsterdam UMC location Vrije Universiteit Amsterdam, Department of Pathology, Cancer Center Amsterdam, Amsterdam, The Netherlands
    Search for articles by this author
  • Carel J.M. van Noesel
    Affiliations
    Amsterdam UMC location University of Amsterdam, Department of Pathology, Amsterdam, The Netherlands
    Search for articles by this author
  • Roy J. Reinten
    Affiliations
    Amsterdam UMC location University of Amsterdam, Department of Pathology, Amsterdam, The Netherlands
    Search for articles by this author
  • Iris D. Nagtegaal
    Affiliations
    Radboud Institute for Molecular Life Sciences, Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands
    Search for articles by this author
  • Miriam Koopman
    Affiliations
    Department of Medical Oncology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands
    Search for articles by this author
  • Author Footnotes
    ⁎ Shared senior and corresponding authorship
    Cornelis J.A. Punt
    Footnotes
    ⁎ Shared senior and corresponding authorship
    Affiliations
    Department of Epidemiology, Julius Center for Health Sciences and Primary Care, University Medical Center, Utrecht University, Utrecht, The Netherlands
    Search for articles by this author
  • Author Footnotes
    ⁎ Shared senior and corresponding authorship
    Louis Vermeulen
    Correspondence
    Address for correspondence. Louis Vermeulen MD, PhD, Amsterdam University Medical Centers – University of Amsterdam, Center for Experimental Molecular Medicine (CEMM), Lab. for Experimental Oncology and Radiobiology (LEXOR), Meibergdreef 9, Amsterdam, The Netherlands, 1105
    Footnotes
    ⁎ Shared senior and corresponding authorship
    Affiliations
    Amsterdam UMC location University of Amsterdam, Center for Experimental and Molecular Medicine, Laboratory for Experimental Oncology and Radiobiology, Amsterdam, The Netherlands

    Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands

    Oncode Institute, Utrecht, The Netherlands
    Search for articles by this author
  • Author Footnotes
    ⁎ Shared senior and corresponding authorship
Open AccessPublished:December 02, 2022DOI:https://doi.org/10.1016/j.clcc.2022.11.006

      Abstract

      Background

      Here we present updated survival of the CAIRO2 trial and assessed whether the addition of anti-EGFR to anti-VEGF therapy could still be an effective treatment option for patients with extended RAS/BRAF wildtype and left-sided metastatic colorectal cancer (mCRC).

      Materials and Methods

      Retrospective updated survival and extended RAS and BRAF V600E mutational analysis were performed in the CAIRO2 trial, a multicenter, randomized phase III trial on the effect of adding cetuximab to a combination of capecitabine, oxaliplatin (CAPOX), and bevacizumab in mCRC.

      Results

      Updated survival analysis confirmed that the addition of cetuximab did not provide a benefit on either progression free (PFS) or overall survival (OS) in the intention-to-treat population. With the extended mutational analyses 31 KRAS, 31 NRAS and 12 BRAF V600E additional mutations were found. No benefit of the addition of cetuximab was observed within the extended wildtype group, even when selecting only left-sided tumors (PFS HR 0.96, p = 0.7775). However, compared to the original trial an increase of 6.5 months was seen for patients with both extended wildtype and left-sided tumors (median OS 28.6 months).

      Conclusion

      Adding cetuximab to CAPOX and bevacizumab does not provide clinical benefit in patients with mCRC, even in the extended wildtype group with left-sided tumors. However, in the extended wildtype group we did observe clinically relevant higher survival compared to the initial trial report, indicating that it is important to analyze a broader panel of RAS and BRAF variants using more recent sequencing techniques when assessing survival benefit after anti-EGFR therapy.

      Keywords

      Introduction

      For patients with mismatch repair proficient (pMMR) metastatic colorectal cancer (mCRC), standard first-line treatment is cytotoxic chemotherapy combined with targeted agents directed against either the epidermal growth factor receptor (anti-EGFR) or against the vascular endothelial growth factor (anti-VEGF). Even though improvement in survival has been shown with these regimens in clinical trials, this effect is less clear in daily practice and 5-year survival rates remain poor.
      • Howlader N.N.A.
      • Krapcho M
      • Miller D
      • et al.
      SEER Cancer Statistics Review, 1975-2016.

      SEER Cancer Stat Facts: Colorectal Cancer. National Cancer Institute. Bethesda, M., https://seer.cancer.gov/statfacts/html/colorect.html. Accessed January 4, 2022.

      • Hamers P.A.H.
      • Elferink M.A.G.
      • Stellato R.K.
      • et al.
      Informing metastatic colorectal cancer patients by quantifying multiple scenarios for survival time based on real-life data.
      CRC is a heterogeneous disease and inhibition of a single signal-transduction pathway is unlikely to provide the most beneficial treatment results.
      • Punt C.J.
      • Koopman M.
      • Vermeulen L.
      From tumour heterogeneity to advances in precision treatment of colorectal cancer.
      Therefore, a combination of targeted agents might be a effective strategy.
      Although preclinical and early clinical studies have suggested that the dual EGFR/VEGF inhibition will increase antitumor activity,
      • Ciardiello F.
      • Bianco R.
      • Damiano V.
      • et al.
      Antiangiogenic and antitumor activity of anti-epidermal growth factor receptor C225 monoclonal antibody in combination with vascular endothelial growth factor antisense oligonucleotide in human GEO colon cancer cells.
      • Saltz L.B.
      • Lenz H.J.
      • Kindler H.L.
      • et al.
      Randomized phase II trial of cetuximab, bevacizumab, and irinotecan compared with cetuximab and bevacizumab alone in irinotecan-refractory colorectal cancer: the BOND-2 study.
      • Shaheen R.M.
      • Ahmad S.A.
      • Liu W.
      • et al.
      Inhibited growth of colon cancer carcinomatosis by antibodies to vascular endothelial and epidermal growth factor receptors.
      • Tonra J.R.
      • Deevi D.S.
      • Corcoran E.
      • et al.
      Synergistic antitumor effects of combined epidermal growth factor receptor and vascular endothelial growth factor receptor-2 targeted therapy.
      this was not confirmed in the CAIRO2 study. This randomized phase III study of capecitabine, oxaliplatin, and anti-VEGF with or without anti-EGFR therapy (cetuximab) in mCRC resulted in an unexpected significantly shorter progression-free survival (PFS) and inferior quality of life in the intention-to-treat population (ITT).
      • Tol J.
      • Koopman M.
      • Cats A.
      • et al.
      Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer.
      Similar results were obtained in two other trials with comparable design.
      • Hecht J.R.
      • Mitchell E.
      • Chidiac T.
      • et al.
      A randomized phase IIIB trial of chemotherapy, bevacizumab, and panitumumab compared with chemotherapy and bevacizumab alone for metastatic colorectal cancer.
      ,
      • Saltz L.
      • Badarinath S.
      • Dakhil S.
      • et al.
      Phase III trial of cetuximab, bevacizumab, and 5-fluorouracil/leucovorin vs. FOLFOX-bevacizumab in colorectal cancer.
      Together these results have ruled out the combined use of anti-VEGF and anti-EGFR monoclonal antibodies with chemotherapy as treatment option for mCRC. However, it has not been explored wether this combination may be benificial for a specific subgroup of patients with extended mutational characterization.
      It is known that anti-EGFR is only effective in a subgroup of mCRC patients. Activating KRAS, NRAS and BRAF V600E mutations have been associated with primary resistance against anti-EGFR and even an inverse outcome following anti-EGFR treatment.
      • De Roock W.
      • Claes B
      • Bernasconi D.
      • et al.
      Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis.
      ,
      • Vale C.L.
      • Tierney J.F.
      • Fisher D.
      • et al.
      Does anti-EGFR therapy improve outcome in advanced colorectal cancer? A systematic review and meta-analysis.
      In the CAIRO2 trial, the worst PFS was indeed seen in cetuximab treated patients with a KRAS (exon 2, codon 12 or 13) mutation. However, we currently know that mutational variants other than KRAS exon 2 (codon 12 or 13) mutations and sideness are also predictive for resistance to anti-EGFR treatment.
      • Dienstmann R.
      • Salazar R.
      • Tabernero J.
      Overcoming Resistance to Anti-EGFR Therapy in Colorectal Cancer.
      We hypothesized that we could identify a subgroup for whom the combined use of anti-VEGF and anti-EGFR will be effective. We therefore assessed the effect of the addition of cetuximab in the CAIRO2 trial in the extended RAS and BRAF V600E wildtype group.

      Methods

      Study Design and Participants

      The CAIRO2 study was an open label, multicenter, randomized phase III trial conducted in 79 centers in the Netherlands (ClinicalTrials.gov identifier: NCT00208546).
      • Tol J.
      • Koopman M.
      • Cats A.
      • et al.
      Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer.
      Patients with previously untreated mCRC were randomly assigned to receive treatment with capecitabine, oxaliplatin and bevacizumab with (CBC group) or without the addition of cetuximab (CB group). Extensive information on inclusion criteria, randomization process and treatment schedules has been described previously.
      • Tol J.
      • Koopman M.
      • Cats A.
      • et al.
      Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer.
      Between June 2005 and December 2006 755 patients were randomized and 736 were eligible for the ITT population (368 in each treatment group).
      The primary endpoint of the CAIRO2 trial was PFS. Secondary endpoints were overall survival (OS), tumor response (RECIST), duration of response, quality of life and safety. From patients of whom resected tumor tissue was available DNA was extracted for KRAS (exon 2) mutational analysis. Patients were not selected for KRAS wildtype status, as data on the predictive value of KRAS mutation status for the outcome of anti-EGFR therapy were not available at the start of the CAIRO2 study.
      The survival data were updated using the NKR (Dutch cancer registry). Updated survival data was obtained from all 736 patients of the ITT population in June 2020.
      The CAIRO2 trial was approved by the national ethics committee on research involving human subjects Arnhem–Nijmegen and conducted in agreement with the declaration of Helsinki. Written informed consent was obtained from all patients.

      Mutation Analysis of KRAS, NRAS, and BRAF V600E

      In the primary analysis of the CAIRO2 study KRAS (exon 2) and BRAF V600E mutation analysis was performed on primary or metastatic samples by a pyrosequencing approach for 528 (KRAS) and 520 (BRAF V600E) patients from which formalin-fixed paraffin-embedded (FFPE) tissue was available.
      • Tol J.
      • Koopman M.
      • Cats A.
      • et al.
      Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer.
      ,
      • Tol J.
      • Nagtegaal I.D.
      • Punt C.J.
      BRAF mutation in metastatic colorectal cancer.
      For the extended mutational analysis of the current study all patients with available primary tumor tissue and previously reported KRAS exon 2 wildtype and BRAF V600E wildtype tumors were included. KRAS (exon 2,3,4), NRAS (exon 2,3,4) and BRAF (exon 15) were analyzed (Supplementary Table 1). Two different methods were used for mutational analysis. For 58 samples whole exome sequencing (WES) data was available,
      • Smeets D.
      • Miller I.S.
      • O’Connor D.
      • et al.
      Copy number load predicts outcome of metastatic colorectal cancer patients receiving bevacizumab combination therapy.
      the remaining 215 wildtype samples with available DNA were analyzed using a next generation sequencing (NGS) custom panel.
      The methods for the WES are described in Smeets et al. 2018.
      • Smeets D.
      • Miller I.S.
      • O’Connor D.
      • et al.
      Copy number load predicts outcome of metastatic colorectal cancer patients receiving bevacizumab combination therapy.
      Briefly, DNA libraries were prepared using the KAPA library preparation kit (KAPA Biosystems), according to the manufacturer's instructions. Libraries were quantified using the Quant-iT™ PicoGreen™ dsDNA Assay Kit. After confirmation of successful library construction, whole exome enrichment was performed using the SeqCapV3 exome enrichment kit (Roche) following the manufacturer's instructions. The enriched libraries were sequenced on a HiSeq2500, paired end 100bp. Target coverage was 60x. For those samples where median coverage of the published data was below 60x, additional sequencing was performed on a HiSeq4000, paired end 150 bp. All sequencing data was reanalyzed to specifically call KRAS, NRAS and BRAF mutations in the same regions as the custom NGS panel (Supplementary Table 1). Sequencing reads were aligned to the human reference genome hg19 using bwa mem version 0.7.10. Picard tools version 1.111 MarkDuplicates was used for duplicate marking. Variant calling was performed using Mutect2 in tumor-only mode and SnpEff version 4.3t was used for gene annotation. Variants were then manually inspected for presence of KRAS (exon 2, 3 and 4), NRAS (exon 2, 3 and 4) or BRAF V600E mutations in the same regions as the custom targeted NGS panel.
      NGS was performed using a custom targeted NGS amplicon panel from the pathology department of KRAS (exon 2, 3 and 4), NRAS (exon 2, 3 and 4) and BRAF V600E (Supplementary Table 1). The DNA libraries were prepared using the Ion AmpliSeq Library Kit 2.0 according to the manufacturer's instructions. Libraries were quantified using the Qubit 3.0 Fluorometer. Tumor DNA libraries were sequenced on an Ion 530 chip in the Ion GeneStudio S5 System (ThermoFisher). The target sequencing depth was 1,500× per amplicon. Sequences were analyzed using SeqNext software v4.1.2 (JSI Medical Systems GmbH, Ettenheim, Germany). For mutation calling a variant allele fraction (VAF) cutoff value of 5% was used.

      CMS Classification

      Samples were classified into the main molecular subtypes, CMS2/3 and CMS4, using the immunohistochemistry classifier.
      • Ten Hoorn S.
      • Trinh A.
      • de Jong J.
      • Koens L.
      • Vermeulen L.
      • et al.
      Classification of Colorectal Cancer in Molecular Subtypes by Immunohistochemistry.
      ,
      • Trinh A.
      • Trumpi K.
      • De Sousa E.M.F.
      • et al.
      Practical and Robust Identification of Molecular Subtypes in Colorectal Cancer by Immunohistochemistry.
      CMS1 is determined based on MMR status (dMMR) and excluded from the analyses reported here due to low numbers (n = 7).

      Statistical Analysis

      The updated survival analyses were performed on the ITT population (n = 736). For the extended mutational analysis all patients with known mutation status and new extended mutation status were analyzed (n = 522) (Figure 1). Baseline characteristics were compared using the Pearson Chi-squared test for categorical variables and unpaired t-test for continuous variables. Unknowns were excluded for testing variables. The PFS and OS were calculated using the Kaplan-Meier method. Comparisons between the different treatment groups (CB and CBC) were calculated using the log-rank test. The effect of treatment on survival was estimated with the Cox proportional hazards model. All statistical analyses were performed in R (version 4.0.5).

      Results

      Patients

      For all patients in the ITT population (n = 736) updated survival was obtained in June 2020. In the CB group 356 patients (97%) and CBC group 354 patients (96%) had died.
      For the extended mutational analysis 522 patients were included in the analysis, of whom 249 with previously known KRAS or BRAF V600E mutated tumors. 273 patients with wildtype tumors underwent the extended RAS and BRAF V600E mutational analyses (Figure 1). Baseline characteristics of the extended RAS and BRAF V600E subgroup were well balanced between the treatment groups and representative for the ITT population (Table 1 and Supplementary Table 2).
      Table 1Baseline Characteristics Extended RAS and BRAF Subgroup
      CharacteristicsCB Group (n = 266)CBC Group (n = 256)P Value
      Age (years)0.716
       Median (range)63.8 (57.6-69.5)63.0 (56.9-69.7)
      Sex (%)0.255
       Male148 (55.6%)156 (60.9%)
       Female118 (44.4%)100 (39.1%)
      WHO performance status0.124
       0171 (64.3%)181 (70.7%)
       195 (35.7%)74 (28.9%)
       No data0 (0%)1 (0.4%)
      Serum lactate dehydrogenase level
      Normal (0) or abnormal (1), according to the cutoff values of each individual center.
      0.732
       0169 (63.5%)166 (64.8%)
       197 (36.5%)88 (34.4%)
       No data0 (0%)2 (0.8%)
      Previous adjuvant therapy0.920
       No222 (83.5%)211 (82.4%)
       Yes44 (16.5%)44 (17.2%)
       No data0 (0%)1 (0.4%)
      Primary tumor site0.073
       Left-sided105 (39.5%)116 (45.3%)
       Right-sided68 (25.6%)71 (27.7%)
       Rectum87 (32.7%)60 (23.4%)
       No data6 (2.3%)9 (3.5%)
      Time of metastasis0.373
       Synchronous150 (56.4%)154 (60.2%)
       Metachronous116 (43.6%)100 (39.1%)
       No data0 (0%)2 (0.8%)
      Resection of primary tumour
       No257 (96.6%)246 (96.1%)1.000
       Yes8 (3.0%)7 (2.7%)
       No data1 (0.4%)3 (1.2%)
      Pearson Chi-squared test used for categorical variables and unpaired t-test used for continuous variables. Unknowns were excluded for testing variables.
      Abbreviations: CB = capecitabine and oxaliplatin plus bevacizumab; CBC = capecitabine and oxaliplatin plus bevacizumab and cetuximab; WHO = World health organisation.
      a Normal (0) or abnormal (1), according to the cutoff values of each individual center.

      Updated Survival

      For the updated median OS of the ITT population there was no statistically significant difference between the treatment arms, with 20.3 months (95% CI 18.0-24.1) in the CB group compared with 20.0 months (95% CI 18.3-21.4) in the CBC group (HR 1.14, 95% CI 0.98-1.32), p = 0.0831; Figure 2A and Table 2A). The updated median PFS was 10.6 months (95% CI 9.6-12.2) in the CB group compared with 9.5 months (95% CI 8.5-10.5) in the CBC group (HR 1.12, 95% CI 0.97-1.30), p = 0.123) (Table 2A).
      Table 2Comparing Survival Data
      nMedian PFSMedian OS
      A. Total Group
      Original paper
       CB Group36810.720.3
       CBC Group3689.419.4
      P value
      Comparisons between the different treatment groups (CB and CBC), calculated using the log-rank test.
      0.010.16
      Updated survival
       CB Group36810.620.3
       CBC Group3689.520.0
       P value
      Comparisons between the different treatment groups (CB and CBC), calculated using the log-rank test.
      0.120.08
      B. Updated survival in subgroup
      Extended KRAS/BRAF Wildtype
       CB Group10512.225.4
       CBC Group9012.425.8
      P value
      Comparisons between the different treatment groups (CB and CBC), calculated using the log-rank test.
      0.970.93
      Extended KRAS/BRAF Wildtype and left-sided
       CB Group9212.826.8
       CBC Group7212.629.2
      P value
      Comparisons between the different treatment groups (CB and CBC), calculated using the log-rank test.
      0.780.84
      C. Subgroup (treatment arms combined)
      Original paper
      KRAS Wildtype31610.622.1
      KRAS and BRAF Wildtype26611.324.5
      Updated survival
       New KRAS/BRAF Mutant748.517.7
       Extended KRAS/BRAF Wildtype19512.325.6
      P value
      Comparisons between the new KRAS/BRAFmutant and extended KRAS/BRAF wildtype subgroups, calculated using the log-rank test.
      0.0070.02
       Extended KRAS/BRAF Wildtype and left-sided16412.628.6
      Abbreviations: CB = capecitabine and oxaliplatin plus bevacizumab; CBC = capecitabine and oxaliplatin plus bevacizumab and cetuximab; ITT = intention-to-treat population; OS = overall survival; PFS = progression-free survival.
      a Comparisons between the different treatment groups (CB and CBC), calculated using the log-rank test.
      b Comparisons between the new KRAS/BRAFmutant and extended KRAS/BRAF wildtype subgroups, calculated using the log-rank test.
      Of the ITT population 52 (7%) of the patients were long-term survivors (>72 months). Baseline characteristics of these long survivors were a slightly younger age (59 versus 62 years), lower serum LDH, left-sided primary tumors and a single metastatic site. These patients tolerated more cycles of systemic therapy and had better objective response rates. There was no apparent effect of treatment arm (Supplementary Table 3).

      Extended RAS and BRAF V600E Mutational Analysis

      Of the 273 patients with KRAS and BRAF V600E wildtype tumors and of whom primary tumor FFPE tissue was available for the extended molecular analyses, 269 patients (98.5%) were successfully analyzed. Additional mutations were found in 27.5% (n = 74) of the analyzed tumors. KRAS was mutated in 31 (11.4%), NRAS in 31 (11.4%) and BRAF V600E in 12 (4.4%). RAS and BRAF V600E mutations were mutually exclusive. Although KRAS exon 2 was sequenced during the original trial, we found nine additional patients with a mutation in KRAS exon 2 in the previously reported wildtype patients. For BRAF V600E we established additional mutations in tumors of 12 patients (Table 3).
      Table 3RAS and BRAF Mutation Status
      ExtendedTotalCB GroupCBC Group
      (n = 273)(n = 522)(n = 266)(n = 256)
      Genotypen (%)n (%)n (%)n (%)
      Wildtype (RAS and BRAF)195(71.4)195(37.4)105(39.5)90(35.2)
      KRAS exon 2 mutant
      KRAS G12A14(2.7)5(1.9)9(3.5)
      KRAS G12C1(0.4)21(4)11(4.1)10(3.9)
      KRAS G12D1(0.4)67(12.8)36(13.5)31(12.1)
      KRAS G12E1(0.2)1(0.4)0(0)
      KRAS G12R5(1)1(0.4)4(1.6)
      KRAS G12S2(0.7)10(1.9)5(1.9)5(2)
      KRAS G12V1(0.4)56(10.7)33(12.4)23(9)
      KRAS G13C1(0.2)0(0)1(0.4)
      KRAS G13D2(0.7)35(6.7)18(6.8)17(6.6)
      KRAS G13E1(0.2)1(0.4)0(0)
      KRAS V14L1(0.4)1(0.2)1(0.4)0(0)
      KRAS Q22K1(0.4)1(0.2)1(0.4)0(0)
      KRAS exon 3 mutant
      KRAS A59E1(0.4)1(0.2)0(0)1(0.4)
      KRAS Q61H5(1.8)5(1)1(0.4)4(1.6)
      KRAS Q61R1(0.4)1(0.2)0(0)1(0.4)
      KRAS exon 4 mutant
      KRAS A146T10(3.7)10(1.9)4(1.5)6(2.3)
      KRAS A146V2(0.7)2(0.4)1(0.4)1(0.4)
      KRAS K117N3(1.1)3(0.6)1(0.4)2(0.8)
      KRAS Total31(11.4)235(45)120(45.1)115(44.9)
      NRAS exon 2 mutant
      NRAS G12A1(0.4)1(0.2)0(0)1(0.4)
      NRAS G12C1(0.4)1(0.2)1(0.4)0(0)
      NRAS G12D8(2.9)8(1.5)3(1.1)5(2)
      NRAS G12S1(0.4)1(0.2)1(0.4)0(0)
      NRAS G12V1(0.4)1(0.2)0(0)1(0.4)
      NRAS G13R1(0.4)1(0.2)0(0)1(0.4)
      NRAS G13V1(0.4)1(0.2)0(0)1(0.4)
      NRAS exon 3 mutant
      NRAS Q61A1(0.4)1(0.2)0(0)1(0.4)
      NRAS Q61H3(1.1)3(0.6)2(0.8)1(0.4)
      NRAS Q61K9(3.3)9(1.7)4(1.5)5(2)
      NRAS Q61L3(1.1)3(0.6)1(0.4)2(0.8)
      NRAS Q61R1(0.4)1(0.2)0(0)1(0.4)
      NRAS Total31(11.4)31(5.9)12(4.5)19(7.4)
      BRAF mutant (V600E)12(4.4)57(10.9)27(10.2)30(11.7)
       Unknown4(1.5)4(0.8)2(0.8)2(0.8)
      Abbreviations: CB = capecitabine and oxaliplatin plus bevacizumab; CBC = capecitabine and oxaliplatin plus bevacizumab and cetuximab.

      Survival of the Extended RAS and BRAF V600E Wildtype Subgroup

      The extended wildtype cohort was defined as those patients with wildtype RAS and BRAF V600E tumors in both the original as well as the extended mutational analyses (n = 195). In this extended wildtype cohort PFS was not significantly different between the treatment groups (HR 1.01, 95% CI 0.75-1.34, p = 0.9681, with 12.2 months in the CB (95% CI 10.5-15.0) and 12.4 months in the CBC group (95% CI 10.6-13.2). For median OS also no difference was seen for the extended wildtype cohort between the treatment groups (HR 1.01, 95% CI 0.76-1.35, p = 0.9342, with 25.4 months in the CB (95% CI 20.7-30.7) and 25.8 months in the CBC group (95% CI 22.9-32.2) (Table 2B).
      As it is known that right-sided primary tumors most probably do not benefit from anti-EGFR treatment,
      • Tejpar S.
      • Stintzing S.
      • Ciardiello F.
      • et al.
      Prognostic and predictive relevance of primary tumor location in patients with ras wild-type metastatic colorectal cancer: retrospective analyses of the crystal and fire-3 trials.
      we also calculated the survival of patients with left-sided and extended wildtype tumors (n = 164). Again, no significant difference was seen with both PFS and OS between the treatment groups (PFS: CB 12.8 months versus CBC 12.6 months, HR 0.96, 95% CI 0.70-1.31, p = 0.7775; OS CB 26.8 months versus CBC 29.2 months, HR 0.97, 95% CI 0.71-1.33, p = 0.841) (Figure 2B and Table 2B).
      Figure 2
      Figure 2Overall survival curves. Six-year overall survival update for (A) the different treatment arms: CB (capecitabine and oxaliplatin plus bevacizumab) and CBC (CB and cetuximab); (B) the extended RAS and BRAF V600E wildtype and left-sided subgroup compared to patients with a RAS or BRAF V600E mutation or right-sided tumour; (C) according to RAS and BRAF V600E mutational status; (D) according to RAS and BRAF V600E mutational status and treatment arm.
      However, when combining both treatment groups (as no significant difference between treatment group was observed) median PFS and OS of the extended wildtype subgroup was 3.5 months higher compared to the KRAS wildtype subgroup of the original report (25.6 months versus 22.1 months) (Table 2C). Of the 273 patients which were wildtype and underwent extended mutational analyses a significant difference in survival of the extended wildtype subgroup compared to the newly mutated group was observed when the patients with additional mutations were removed (extended wildtype versus new RAS/BRAF V600E mutant PFS: 12.3 months versus 8.5 months, p = 0.007; OS: 25.6 versus 17.7 months, p = 0.02 (Table 2C and Supplementary Figure 1). Median survival for patients with previous known KRAS or BRAF
      V600E mutations versus newly found mutations remained unchanged. When restricting to left-sided primary and extended wildtype tumors the OS was even higher, adding another 3.0 months (from 25.6 to 28.6 months) (Table 2C). In total the survival of the extended wildtype and left-sided primary tumor subgroup was 6.5 months higher compared to the original KRAS wildtype subgroup.

      Prognostic and Predictive Effect of RAS and BRAF V600E Mutation Status in the Total Group of Patients

      We explored the prognostic effect of the mutation status in the total group, combining both treatment groups. Patients with RAS and BRAF V600E wildtype tumors had significantly better OS compared to patients with either RAS or BRAF V600E mutations (RAS/BRAF wildtype 25.6 months (95% CI 23.0-29.5) versus RAS mutation 19.8 months (95% CI 17.7-22.1), HR 1.32, 95% CI 1.09-1.60, p = 0.004; and versus BRAF V600E mutation 13.6 months (95% CI 9.61-16.6), p < 0.001). BRAF V600E mutations showed the worst survival compared to both RAS mutations (HR 1.70, 95% CI 1.27-2.26, p < 0.001) and patients with wildtype tumors (HR 2.27, 95% CI 1.68-3.08, p < 0.001) (Figure 2C).
      With regards to the predictive effect, patients with a RAS mutation had a significantly worse survival when treated with the addition of cetuximab as compared to the CB group (HR 1.36, 95% CI 1.07-1.75, p = 0.01387) with a median OS of 23.2 months for the CB group (95% CI 19.8-27.7) compared with 18.3 months in the CBC group (95% CI 14.7-20.6). No treatment interaction was seen for BRAF V600E mutations (HR 1.03, 95% CI 0.61-1.75, p = 0.9088) (Figure 2D).

      Predictive Effect of the CMS

      As the extended RAS and BRAF V600E wildtype subgroup did not show a benefit from the combination of anti-EGFR and anti-VEGF, we explored the predictive effect of the consensus molecular subtypes (CMSs). In the extended wildtype group there were 72 CMS2/3 and 51 CMS4 patients. For CMS2/3 patients there was a numerically but statistically non-significant increased median OS in the CBC group of 31.9 months versus 23.0 months in the CB group (HR 0.75, 95% CI 0.46-1.21, p = 0.246). No treatment effect was seen for CMS4 (HR 0.78, 95% CI 0.56-1.07, p = 0.127) (Figure 3).
      Figure 3
      Figure 3Overall survival curves for the consensus molecular subtypes. Overall survival stratified by the consensus molecular subtypes (CMS) in the extended RAS and BRAF Wildtype cohort for (A) CMS2/3 and (B) CMS4. Abbreviations: CB, capecitabine and oxaliplatin plus bevacizumab; CBC, CB and cetuximab.

      Discussion

      In this updated survival analysis and extended analysis of RAS/BRAF V600E mutation status of the CAIRO2 study we show that the addition of anti-EGFR to first-line treatment with anti-VEGF and CAPOX does not improve outcome in mCRC patients. However, survival was substantially higher compared to the original trial when only patients with extended RAS/BRAF V600E wildtype and left-sided tumors were considered.
      For the population with an anti-EGFR resistant RAS mutant genotype, the detrimental effect on survival by the addition of anti-EGFR treatment was confirmed.
      • Bokemeyer C.
      • Bondarenko I.
      • Hartmann J.T.
      • et al.
      Efficacy according to biomarker status of cetuximab plus FOLFOX-4 as first-line treatment for metastatic colorectal cancer: the OPUS study.
      ,
      • Douillard J.Y.
      • Oliner K.S.
      • Siena S.
      • et al.
      Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer.
      Patients with RAS mutated tumors in the control arm with chemotherapy plus bevacizumab showed a similar OS compared to patients with wildtype tumors, implying that the significant lower median OS for the RAS mutant tumors could be attributed to the detrimental effect of cetuximab in RAS mutated tumors. Hence, not only a lack of response but worse survival is seen when treating RAS mutated tumors with anti-EGFR therapy. In the PACCE trial addition of panitumumab to chemotherapy plus bevacizumab also resulted in a decrease in PFS and excess toxicity.
      • Hecht J.R.
      • Mitchell E.
      • Chidiac T.
      • et al.
      A randomized phase IIIB trial of chemotherapy, bevacizumab, and panitumumab compared with chemotherapy and bevacizumab alone for metastatic colorectal cancer.
      The same holds true for the study of Saltz et al. where patients with KRAS mutant tumors showed inferior PFS when cetuximab was added to chemotherapy and bevacizumab.
      • Saltz L.
      • Badarinath S.
      • Dakhil S.
      • et al.
      Phase III trial of cetuximab, bevacizumab, and 5-fluorouracil/leucovorin vs. FOLFOX-bevacizumab in colorectal cancer.
      Patients harboring a BRAF V600E mutation are known to be characterized by a dismal prognosis.
      • Caputo F.
      • Santini C.
      • Bardasi C.
      • et al.
      BRAF-mutated colorectal cancer: clinical and molecular insights.
      In our study we confirmed the poor prognosis of these patients, compared to patients with RAS mutated and RAS/BRAF wildtype tumors. However, no predictive effect was observed in this trial for the addition of cetuximab. Even though both RAS and BRAF V600E mutations activate the pathway downstream of EGFR thereby causing resistance to anti-EGFR therapy, there appears to be a difference in the clinical effect dependent on the specific mutation. Patients with BRAF V600E mutated tumors show no difference in survival between the two treatment combinations, but patients with RAS mutated tumors had worse survival with the addition of cetuximab compared to the control arm.
      Interestingly we discovered additional RAS exon 2 and BRAF V600E mutations in tumors which were classified as wildtype in the primary mutational analyses. This can be explained by the improved quality and efficiency of DNA sequencing and increased accuracy of software for calling of mutations between the primary analyses in 2008 and the current analyses. When removing the patients with the newly found mutations from the wildtype subgroup we showed an important increase in survival of the extended wildtype subgroup. These findings imply that older mutational analyses might underestimate the number of mutants and if retrospective analyses are performed using older mutation data it should be considered to re-analyze the wildtype cohort for additional mutations.
      Two possible explanations for the lack of benefit in our extended wildtype subgroup might involve the negative interaction of anti-EGFR when combined with capecitabine.
      • Chan D.L.
      • Pavlakis N.
      • Shapiro J.
      • et al.
      Does the chemotherapy backbone impact on the efficacy of targeted agents in metastatic colorectal cancer? A systematic review and meta-analysis of the literature.
      Firstly, the often-decreased total dose intensity and hence efficacy due to increased toxicity from capecitabine-containing regimens. A second, speculative, hypothesis is decreased cytotoxic activity of capecitabine as the required metabolic activation is reduced due to cetuximab-induced G1 arrest.
      • Chan D.L.
      • Pavlakis N.
      • Shapiro J.
      • et al.
      Does the chemotherapy backbone impact on the efficacy of targeted agents in metastatic colorectal cancer? A systematic review and meta-analysis of the literature.
      Therefore, it might be interesting to repeat a similar study in an infusional-5-FU or irinotecan-based chemotherapy regimen combined with anti-EGFR. Another interesting step could be to expand the mutational panel with other mutations associated with anti-EGFR resistance, including HRAS, PIK3CA exon 20, PTEN, MAP2K1 and amplifications involving MET, ERBB2, KRAS and also including amplifications which increase the sensitivity to anti-EGFR therapy e.g. EGFR and IRS2.
      • De Roock W.
      • Claes B
      • Bernasconi D.
      • et al.
      Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis.
      ,
      • Dienstmann R.
      • Salazar R.
      • Tabernero J.
      Overcoming Resistance to Anti-EGFR Therapy in Colorectal Cancer.
      ,
      • Bardelli A.
      • Corso S
      • Bertotti A.
      • et al.
      Amplification of the MET receptor drives resistance to anti-EGFR therapies in colorectal cancer.
      • Bertotti A.
      • Papp E.
      • Jones S.
      • et al.
      The genomic landscape of response to EGFR blockade in colorectal cancer.
      • Perrone F.
      • Lampis A.
      • Orsenigo M.
      • et al.
      PI3KCA/PTEN deregulation contributes to impaired responses to cetuximab in metastatic colorectal cancer patients.
      We also did not account for possible secondary resistance, which can occur during treatment with anti-EGFR.
      • Dienstmann R.
      • Salazar R.
      • Tabernero J.
      Overcoming Resistance to Anti-EGFR Therapy in Colorectal Cancer.
      ,
      • Bertotti A.
      • Papp E.
      • Jones S.
      • et al.
      The genomic landscape of response to EGFR blockade in colorectal cancer.
      ,
      • Arena S.
      • Bellosillo B.
      • Siravegna G.
      • et al.
      Emergence of multiple EGFR extracellular mutations during cetuximab treatment in colorectal cancer.
      Therefore, we cannot fully exclude that dual EGFR/VEGF inhibition may still be beneficial in a small subset.
      Interestingly, in a small subgroup analysis we showed the predictive potential of stratifying the extended wildtype group according to the CMS, with a numerically difference in median OS for CMS2/3 tumors treated with the combination of anti-EGFR and anti-VEGF as compared to no addition of anti-EGFR. However, the sample size of this small subgroup was too small to show any significant effects, which we have shown before in this CAIRO2 wildtype cohort for the original KRAS mutations.
      • Trinh A.
      • Trumpi K.
      • De Sousa E.M.F.
      • et al.
      Practical and Robust Identification of Molecular Subtypes in Colorectal Cancer by Immunohistochemistry.
      There could also have been a switch from CMS2 to CMS4 subtype, contributing to acquired cetuximab resistance which was not accounted for in this analysis as we only classified primary tumor tissue prior to treatment.
      • Woolston A.
      • Khan K.
      • Spain G.
      • et al.
      Genomic and transcriptomic determinants of therapy resistance and immune landscape evolution during anti-EGFR treatment in colorectal cancer.
      More recently we have shown that the efficacy of anti-EGFR in the different subtypes is dependent on the chemotherapy backbone. Efficacy using an oxaliplatin backbone, as used in the CAIRO2 study, was restricted to left-sided CMS2/3 tumors, which is in line with our results.
      • Ten Hoorn S.
      • Sommeijer D.W.
      • Eliott F.
      • et al.
      Molecular subtype-specific efficacy of anti-EGFR therapy in colorectal cancer is dependent on the chemotherapy backbone.
      In conclusion, based on the here reported retrospective analysis of the CAIRO2 trial, adding cetuximab to CAPOX and bevacizumab does not provide any clinical benefit in the updated survival analyses and extended wildtype group. However, in the extended wildtype group we did observe a clinically relevant higher survival compared to the initial trial report, indicating that it is important to analyze a broader panel of RAS and BRAF variants using more recent sequencing techniques when assessing survival benefit after anti-EGFR therapy.

      Data Availability

      The WES data are deposited at the EMBL-EBI under accession code EGAS00001002617 and EGAS00001006319 (additional sequencing for samples with low coverage). The data are available under restricted access.

      Clinical Practice Points

      • There was no subgroup who benefitted from the combined addition of anti-EGFR and anti-VEGF to a chemotherapy backbone in metastatic colorectal cancer in the CAIRO2 trial. Using more recent sequencing techniques, multiple additional RAS and BRAF mutations were detected and the extended wildtype left-sided subgroup had superior overall survival (28.6 months) compared to the original trial report. This indicates that it is important to analyze a broader panel of RAS and BRAF variants using state of the art sequencing techniques when assessing survival benefit after anti-EGFR therapy.

      Acknowledgements

      We thank Dirk H.F.B. van Essen for performing the Whole Exome Sequencing experiments. Louis Vermeulen received consultancy fees/research funding from Bayer, MSD, Genentech, Servier and Pierre Fabre, but these had no relation to the presented work. The rest of the authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. This work was supported by the Dutch Cancer Society (KWF) and the New York Stem Cell Foundation (NYSCF).

      Appendix. Supplementary materials

      References

        • Howlader N.N.A.
        • Krapcho M
        • Miller D
        • et al.
        SEER Cancer Statistics Review, 1975-2016.
        National Cancer Institute, Bethesda, MD2019 (based on November 2018 SEER data submission, posted to the SEER web site, April 2019)
      1. SEER Cancer Stat Facts: Colorectal Cancer. National Cancer Institute. Bethesda, M., https://seer.cancer.gov/statfacts/html/colorect.html. Accessed January 4, 2022.

        • Hamers P.A.H.
        • Elferink M.A.G.
        • Stellato R.K.
        • et al.
        Informing metastatic colorectal cancer patients by quantifying multiple scenarios for survival time based on real-life data.
        Int J Cancer. 2021; 148: 296-306
        • Punt C.J.
        • Koopman M.
        • Vermeulen L.
        From tumour heterogeneity to advances in precision treatment of colorectal cancer.
        Nat Rev Clin Oncol. 2017; 14: 235-246
        • Ciardiello F.
        • Bianco R.
        • Damiano V.
        • et al.
        Antiangiogenic and antitumor activity of anti-epidermal growth factor receptor C225 monoclonal antibody in combination with vascular endothelial growth factor antisense oligonucleotide in human GEO colon cancer cells.
        Clin Cancer Res. 2000; 6: 3739-3747
        • Saltz L.B.
        • Lenz H.J.
        • Kindler H.L.
        • et al.
        Randomized phase II trial of cetuximab, bevacizumab, and irinotecan compared with cetuximab and bevacizumab alone in irinotecan-refractory colorectal cancer: the BOND-2 study.
        J Clin Oncol. 2007; 25: 4557-4561
        • Shaheen R.M.
        • Ahmad S.A.
        • Liu W.
        • et al.
        Inhibited growth of colon cancer carcinomatosis by antibodies to vascular endothelial and epidermal growth factor receptors.
        Br J Cancer. 2001; 85: 584-589
        • Tonra J.R.
        • Deevi D.S.
        • Corcoran E.
        • et al.
        Synergistic antitumor effects of combined epidermal growth factor receptor and vascular endothelial growth factor receptor-2 targeted therapy.
        Clin Cancer Res. 2006; 12: 2197-2207
        • Tol J.
        • Koopman M.
        • Cats A.
        • et al.
        Chemotherapy, bevacizumab, and cetuximab in metastatic colorectal cancer.
        N Engl J Med. 2009; 360: 563-572
        • Hecht J.R.
        • Mitchell E.
        • Chidiac T.
        • et al.
        A randomized phase IIIB trial of chemotherapy, bevacizumab, and panitumumab compared with chemotherapy and bevacizumab alone for metastatic colorectal cancer.
        J Clin Oncol. 2009; 27: 672-680
        • Saltz L.
        • Badarinath S.
        • Dakhil S.
        • et al.
        Phase III trial of cetuximab, bevacizumab, and 5-fluorouracil/leucovorin vs. FOLFOX-bevacizumab in colorectal cancer.
        Clin Colorectal Cancer. 2012; 11: 101-111
        • De Roock W.
        • Claes B
        • Bernasconi D.
        • et al.
        Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis.
        Lancet Oncol. 2010; 11: 753-762
        • Vale C.L.
        • Tierney J.F.
        • Fisher D.
        • et al.
        Does anti-EGFR therapy improve outcome in advanced colorectal cancer? A systematic review and meta-analysis.
        Cancer Treat Rev. 2012; 38: 618-625
        • Dienstmann R.
        • Salazar R.
        • Tabernero J.
        Overcoming Resistance to Anti-EGFR Therapy in Colorectal Cancer.
        Am Soc Clin Oncol Educ Book. 2015; 35: e149-e156
        • Tol J.
        • Nagtegaal I.D.
        • Punt C.J.
        BRAF mutation in metastatic colorectal cancer.
        N Engl J Med. 2009; 361: 98-99
        • Smeets D.
        • Miller I.S.
        • O’Connor D.
        • et al.
        Copy number load predicts outcome of metastatic colorectal cancer patients receiving bevacizumab combination therapy.
        Nat Commun. 2018; 9: 4112
        • Ten Hoorn S.
        • Trinh A.
        • de Jong J.
        • Koens L.
        • Vermeulen L.
        • et al.
        Classification of Colorectal Cancer in Molecular Subtypes by Immunohistochemistry.
        Methods Mol Biol. 2018; 1765: 179-191
        • Trinh A.
        • Trumpi K.
        • De Sousa E.M.F.
        • et al.
        Practical and Robust Identification of Molecular Subtypes in Colorectal Cancer by Immunohistochemistry.
        Clinical Cancer Research. 2017; 23: 387-398
        • Tejpar S.
        • Stintzing S.
        • Ciardiello F.
        • et al.
        Prognostic and predictive relevance of primary tumor location in patients with ras wild-type metastatic colorectal cancer: retrospective analyses of the crystal and fire-3 trials.
        JAMA Oncol. 2017; 3: 194-201
        • Bokemeyer C.
        • Bondarenko I.
        • Hartmann J.T.
        • et al.
        Efficacy according to biomarker status of cetuximab plus FOLFOX-4 as first-line treatment for metastatic colorectal cancer: the OPUS study.
        Ann Oncol. 2011; 22: 1535-1546
        • Douillard J.Y.
        • Oliner K.S.
        • Siena S.
        • et al.
        Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer.
        N Engl J Med. 2013; 369: 1023-1034
        • Caputo F.
        • Santini C.
        • Bardasi C.
        • et al.
        BRAF-mutated colorectal cancer: clinical and molecular insights.
        Int J Mol Sci. 2019; : 20
        • Chan D.L.
        • Pavlakis N.
        • Shapiro J.
        • et al.
        Does the chemotherapy backbone impact on the efficacy of targeted agents in metastatic colorectal cancer? A systematic review and meta-analysis of the literature.
        PLoS One. 2015; 10e0135599
        • Bardelli A.
        • Corso S
        • Bertotti A.
        • et al.
        Amplification of the MET receptor drives resistance to anti-EGFR therapies in colorectal cancer.
        Cancer Discov. 2013; 3: 658-673
        • Bertotti A.
        • Papp E.
        • Jones S.
        • et al.
        The genomic landscape of response to EGFR blockade in colorectal cancer.
        Nature. 2015; 526: 263-267
        • Perrone F.
        • Lampis A.
        • Orsenigo M.
        • et al.
        PI3KCA/PTEN deregulation contributes to impaired responses to cetuximab in metastatic colorectal cancer patients.
        Ann Oncol. 2009; 20: 84-90
        • Arena S.
        • Bellosillo B.
        • Siravegna G.
        • et al.
        Emergence of multiple EGFR extracellular mutations during cetuximab treatment in colorectal cancer.
        Clin Cancer Res. 2015; 21: 2157-2166
        • Woolston A.
        • Khan K.
        • Spain G.
        • et al.
        Genomic and transcriptomic determinants of therapy resistance and immune landscape evolution during anti-EGFR treatment in colorectal cancer.
        Cancer Cell. 2019; 36: 35-50.e9
        • Ten Hoorn S.
        • Sommeijer D.W.
        • Eliott F.
        • et al.
        Molecular subtype-specific efficacy of anti-EGFR therapy in colorectal cancer is dependent on the chemotherapy backbone.
        Br J Cancer. 2021; 125: 1080-1088