KRAS G13D Mutation and Sensitivity to Cetuximab or Panitumumab in a Colorectal Cancer Cell Line Model

ABSTRACT

BACKGROUND:

The treatment of metastatic colorectal cancer (mCRC) includes drugs targeting the epidermal growth factor receptor (EGFR). Mutation in codon 12 or 13 in the Kirsten rat sarcoma viral oncogene homolog (KRAS) gene, downstream of the EGFR, evokes constitutive activation of the RAS/RAF/MAPK signaling pathway and correlates with resistance to anti-EGFR monoclonal antibody (mAb) therapies. However, a retrospective study reported that a proportion of patients with the KRAS G13D mutation may respond to cetuximab. A similar analysis for panitumumab was not as conclusive. We sought to determine the sensitivity of CRC cell lines to cetuximab or panitumumab treatment and to investigate the correlation of the KRAS mutational status of the CRC cell lines to the responsiveness to cetuximab or panitumumab.

METHODS:

To determine the responsiveness of CRC cell lines to cetuximab or panitumumab, cell lines were treated with an optimized concentration of each mAb, and proliferation assays were conducted.

RESULTS:

After treatment with cetuximab or panitumumab, at the optimum concentration of 8 μg/well, the KRAS G13D mutant cell lines HCT-116, LoVo, and T84 showed intermediate sensitivity to both treatments, between the resistant KRAS G12V mutant cell line SW480 and the sensitive KRAS wild-type cell line LIM1215. One of the G13D cell lines was significantly more sensitive to panitumumab than to cetuximab (P = .02).

CONCLUSION:

The specific KRAS mutation determines the responsiveness to anti-EGFR monoclonal antibody treatment, corresponding to reported clinical observations.

The treatment of metastatic colorectal cancer (mCRC) has improved over recent years, with targeted therapies providing additional benefit to standard chemotherapy. The two most important targets for mCRC treatment are epidermal growth factor receptor (EGFR) and vascular endothelial growth factor (VEGF). Mutation of the KRAS gene is now known to be predictive of nonresponse to EGFR-targeted mAb therapy, either as monotherapy or in combination with irinotecan- or oxaliplatin-based chemotherapy.^1,2

Cetuximab (a chimeric human-murine IgG1 mAb) and panitumumab (a fully humanized IgG2 mAb) target the EGFR and act by binding to the EGFR on tumor cells, blocking the downstream intracellular signaling pathways. A member of this downstream cascade is KRAS, and evidence has suggested that patients with KRAS mutations do not benefit from the addition of cetuximab or panitumumab, either alone or in addition to standard chemotherapy.^2,3 Mutation of KRAS results in constitutive downstream activation of the EGFR pathway, propagating further signaling events and making the EGFR inhibitors ineffective. A retrospective analysis of early trials of cetuximab therapy suggested that KRAS exon 2 mutation occurs in 27–43% of patients with mCRC tumors, and the reported objective response rate (ORR) was 0 in this group.⁴ These reports also established superior ORRs with EGFR inhibitors in wild-type (WT) KRAS tumors.^5,6 Subsequent analysis of large randomized trials involving both cetuximab and panitumumab have confirmed the predictive nature of the KRAS mutation.⁴ As a consequence, KRAS testing has been made mandatory for patients with mCRC before treatment with cetuximab or panitumumab.⁷ However, there is growing evidence of the existence of an array of mutations that in turn influence the responsiveness to an anti-EGFR treatment, and their roles are not fully understood.⁴ Overall KRAS mutations, if they include exons 3 and 4 in addition to 2, are likely to be found in approximately 45–55% of all colorectal cancer specimens.⁸

In a retrospective study by De Roock et al,⁹ it was evident that a proportion of patients with KRAS G13D mutation do respond to cetuximab. The largest retrospective analysis, conducted by Peeters et al,¹⁰ to evaluate three phase III trial studies involving the alternate anti-EGFR drug panitumumab (first line, second line, and monotherapy) revealed that KRAS G13D was unfavorably associated with panitumumab treatment effects on overall survival (OS) but not on progression-free survival (PFS) or response rate. These discrepant results may reflect subtle differences between the two antibodies to EGFR—for example, chimeric vs. fully humanized. There is also a report of activity of panitumumab after cetuximab failure, adding further evidence to potential differences in activity.¹¹

Based on these retrospective studies and the conflicting results, we sought to explore, in a preclinical CRC cell line model, the sensitivity and/or resistance to both cetuximab and panitumumab treatment and to investigate the correlation of the KRAS mutational status of the CRC lines to the responsiveness to these agents.

MATERIALS AND METHODS

Cell Lines and Reagents

CRC lines, HCT-116, T84, LoVo (all KRAS G13D mutant), and SW480 (KRAS G12V mutant) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). LIM1215 CRC line (KRAS WT) was a kind gift from the Ludwig Institute (Melbourne, Australia). The cell lines were cultured in 75-mL tissue culture flasks (Greiner Bio-One, Frickenhausen, Germany) in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin, 100 μg/mL streptomycin, and 200 μg/mL glutamine (all from Gibco-Life Technologies, Grand Island, NY, USA) at 37°C and 5% CO₂, according to the protocol provided by ATCC. The cell lines were tested with the MycoAlert mycoplasma detection kit (Lonza, Sydney, NSW, Australia) and were found to be free of mycoplasma contamination throughout the experimental procedure.

Monoclonal Antibodies

Cetuximab (Erbitux; Merck Serono, Frenchs Forest, NSW, Australia) at 5 mg/mL and panitumumab (Vectibix; Amgen, Thousand Oaks, CA, USA) at 20 mg/mL were generously provided by the Medical Oncology Unit Pharmacy of The Queen Elizabeth Hospital, Adelaide, Australia.

Proliferation Assay and Optimization of Antibody Concentrations

The CellTiter 96® AQueous nonradioactive cell proliferation assay kit (Promega, Madison, WI, USA) was used to assess the resistance and sensitivity of these CRC lines to the monoclonal antibodies. The cells were seeded into 96-well plates (Greiner Bio-one) at 5 × 10³ cells/mL in a total volume of 100 μL medium and incubated for 24 hours in a 5% CO₂ humidified atmosphere at 37°C. The cells were treated in triplicate with cetuximab or panitumumab at concentrations of 0.5–10 μg per well, to determine the optimal concentrations for the proliferation assay. Another set of cells on the plate was treated in triplicate with 10 μg of monoclonal mouse IgG1 or IgG_2A isotype control antibody (R&D Systems, Minneapolis, MN, USA) as a negative control. The plates were incubated for 72 hours at 37°C in a 5% CO₂ incubator, and all wells were then treated with 20 μL of MTS/PMS solution and incubated for a further 1.5 hours. The absorbance was read with the Fluostar Optima instrument (BMG Labtech, Durham, NC, USA) at a wavelength of 490 nm. Proliferation was recorded as a percentage of that obtained for the isotype control antibody-treated cells (100%).

Statistical Analysis

Statistical analyses were performed with Students t-test in the Prism 6 program (GraphPad Software, La Jolla, CA, USA) with P < .05 deemed significant.

RESULTS

Optimization of Antibody Concentrations

The optimal treatment concentration to provide the maximum inhibition of proliferation for all the cell lines for cetuximab and panitumumab was determined to be 8 μg/well (Supplementary Figures 1–10).

Correlation Between KRAS Status and Responsiveness to Cetuximab or Panitumumab Treatment

After cetuximab treatment (Figure 1A), the G13D cell lines, with a mean proliferation rate of 53.5% relative to the isotype control, were significantly more sensitive than SW480 (KRAS G12V mutation) with a proliferation rate of 72.3% (P = .04), but were less sensitive than LIM1215 (KRAS WT) (proliferation 29.5%; P = .012). After panitumumab treatment (Figure 1B), the G13D cell lines (mean proliferation, 44.7%) again showed intermediate sensitivity between SW480 (proliferation, 71.4%; P = .0003) and LIM1215 (proliferation, 29%; P = .004). Comparing the efficacy of cetuximab vs. panitumumab within each of the 5 cell-lines showed that only LoVo (G13D) had a significant difference in response, being more sensitive to panitumumab (P = .02) (Figure 2).

Figure 1.

Response of KRAS G12V, G13D and WT cell lines to (A) 8 μg cetuximab treatment and (B) 8 μg panitumumab treatment. Dashed line: mean proliferation, as a percentage of isotype control antibody treatment for the G13D cell lines.

Figure 2.

Percentage proliferation relative to isotype control antibody, comparing cetuximab to panitumumab for each of the cell lines. *Significance at P = .02.

DISCUSSION

Anti-EGFR mAbs are effective at prolonging disease-free survival and OS and in improving response rates in patients with KRAS WT mCRC.^1,12 Mutation in KRAS exon 2, which leads to EGFR-independent, constitutive activation of RAS, was shown to predict for resistance to both anti-EGFR agents (cetuximab and panitumumab), and subsequently this was confirmed in the retrospective KRAS analysis of the two pivotal studies of anti-EGFR therapy.^2,3

In our in vitro study, the SW480 CRC line with the KRAS G12V mutation had the highest proliferation rate, showing resistance to both the anti-EGFR mAbs cetuximab and panitumumab, when compared with other cell lines with the KRAS G13D mutation and KRAS WT cells. The 3 cell lines with KRAS G13D (HCT-116, LoVo, and T84), showed an intermediate level of responsiveness to cetuximab and panitumumab, with the proliferation rate decreasing as the treatment concentration increased, indicating a dose–response activity. However, the responsiveness was not as great as LIM1215, a KRAS WT cell line, which had the lowest proliferation rate when treated with cetuximab and panitumumab. The KRAS WT cell line showed the highest response rate to both cetuximab and panitumumab, as expected.

Both antibodies are likely to have a similar activity in vitro, although there are differences in their action in vivo, probably because they are different immunoglobulin isotypes. Panitumumab, like other IgG2 antibodies, was thought not to mediate antibody-dependent cellular cytotoxicity (ADCC), a recognized action of IgG1 antibodies, including cetuximab,¹³ although a recent report has demonstrated that panitumumab may in fact induce ADCC via myeloid effector cells in vitro, a mechanism that may contribute to its efficacy in vivo.¹⁴ Further, a report of panitumumab activity after cetuximab failure again suggested some differences in the mechanism of action.¹⁵

The possibility of a difference between these drugs has been further highlighted by the differential response in the G13D mutant subgroup in retrospective analyses: De Roock et al⁹ reported that patients (n = 571) receiving any cetuximab-based treatment regardless of concomitant chemotherapy, showed a statistically significant longer OS and PFS in those with KRAS G13D mutations than did patients with any other KRAS mutation (OS, 7.6 months, 95% CI, 5.7–20.5, and PFS, 4.0 months, 95% CI, 1.9–6.2 vs. OS, 5.7 months, 95% CI, 4.9–6.8, and PFS, 1.9 months, 95% CI, 1.8–2.8).⁹ These findings were supported in a recent systematic review and meta-analysis of 10 studies of cetuximab treatment in patients with mCRC. Those with KRAS G13D tumors had a significantly higher ORR and longer PFS and OS than did patients with KRAS codon 12-mutated tumors, and lower ORR with shorter PFS and OS than patients with KRAS WT tumors.¹⁶ These data thus support a better prognosis for patients with G13D mutations, in keeping with the results in our cell line analysis. However, clinical outcomes for panitumumab therapy appear to be less well defined: Peeters et al¹⁰ concluded from their pooled analysis of randomized trials for 1053 KRAS mutant patients treated with panitumumab that although no mutant KRAS allele was consistently shown to be a significant predictive factor for PFS or OS, patients with mutant KRAS codon 12 or 13 mCRC tumors were unlikely to benefit from panitumumab therapy.¹⁰ In contrast, our in vitro study suggests that G13D mutant cells show sensitivity to panitumumab that falls between the sensitive WT cells (P = .004) and the resistant G12V cells (P = .0003).

A look at the potential differences in codon 12 vs. 13 mutations shows that our overall results concur with those in an in vitro study done by Guerrero and colleagues in 2000.¹⁷ They found that the malignant potential of tumor cells may be influenced, not only by the presence or absence of RAS mutations, but also by its molecular nature. KRAS codon 13 mutation was reported to have a reduced transforming capacity, as compared with KRAS codon 12 mutation in experimental systems. Their study suggested that tumors carrying KRAS codon 13 mutations are less aggressive than those with codon 12 mutations. Also, colorectal tumors carrying KRAS codon 12 mutations showed a lower apoptotic rate than did tumors lacking this mutation and conferred a more aggressive tumor phenotype by altering the threshold of apoptotic induction; in contrast KRAS codon 13 mutations reduced this threshold. Their results indicate that cells carrying KRAS codon 13 mutations or overexpressing the WT allele may have reduced survival and could be selected against in the adenoma-to-carcinoma transition.¹⁷ This possibility would explain the higher proliferative capacity of the codon 12 G12V cell line in our study, which exhibited definite resistance to cetuximab and panitumumab when compared with the G13D cell lines, which showed a degree of sensitivity to both anti-EGFR drugs. It has recently been established that negative feedback involving EGFR mediates cross-talk between the RAS-MEK-ERK pathway and the parallel AKT pathway through phosphorylation of key signaling molecules and that this activation of AKT varies depending on specific mutations including in KRAS and BRAF. Furthermore, this study highlighted the need for combined inhibition of MEK and EGFR to inactivate both pathway end points.¹⁸ This variability in the activation of AKT, depending on the particular KRAS mutation, explains the discrepant growth inhibition responses to EGFR inhibition in tumor cells with differing KRAS mutations in our study.

CONCLUSIONS

In our preclinical CRC cell line model, there were differences in response to anti-EGFR treatments, depending on the particular KRAS mutation. SW480 with KRAS G12V mutation was resistant to both cetuximab and panitumumab, whereas all of the KRAS G13D-mutated cell lines showed intermediate sensitivity to both treatments. Of note, one of the G13D cell lines showed greater sensitivity to panitumumab than to cetuximab. Further prospective data are needed to clarify the role of the KRAS G13D mutation for anti-EGFR-based treatments in mCRC; an Australian-led international trial, the AGITG ICECREAM study (ACTRN12612000901808) is currently under way. This is a randomized phase II study of cetuximab alone or in combination with irinotecan in mCRC patients with KRAS WT and in patients harboring a G13D mutation. Results from such trials in the future should inform the selection of patients for anti-EGFR therapy and the need for additional targeted therapies, based on genotype, so that a more personalized approach to treatment can be realized.