Table 4.
Summary of potential predictive molecular biomarkers for response to the epidermal growth factor receptor (EGFR)–targeted monoclonal antibodies cetuximab and panitumumab in metastatic colorectal cancer*
| Relationship to response | Biomarker | First author (reference) |
| Predicts lack of response and now incorporated into clinical practice† | KRAS mutation | Amado (27); Bokemeyer (28); Van Cutsem (29); Table 2, this article; (20,130,131) |
| Very likely to predict lack of response | Mutation or lack of expression of PTEN; mutation of BRAF or PIK3CA | Frattini (58); Perrone (38); Benvenuti (33); Di Nicolantonio (40); Sartore-Bianchi (72); Di Nicolantonio (132) |
| May predict lack of response‡ | Increased HER2 gene copy number | Finocchiaro (100) |
| May predict increased likelihood of response | Increased EGFR gene copy number§ | Table 3, this article |
| Increased EGFR phosphorylation‡ | Personeni (81) | |
| Overexpression of alternative EGFR ligands (amphiregulin and/or epiregulin)‡ | Khambata-Ford (59) | |
| pAKt overexpression‡ | Razis (133) | |
| Other potential markers | Markers of angiogenesis and cell cycle regulation; transcription factors (VEGF, IL-8, COX-2, cyclin D, NFκB)‡ | Vallböhmer (108); Nagashima (109); Zhang (92) |
Data are based on analysis of tumor tissue from patients participating in clinical trials. COX-2 = cyclooxygenase-2; HER2 = human epidermal growth factor-2; IL-8 = interleukin-8; NFκB = Nuclear factor kappa B; pAkt = phosphorylated Akt; VEGF = vascular endothelial growth factor.
Based on data from a study that prospectively defined biomarker analysis and included a large number of patients (13).
Limited preliminary data.
Data need to be confirmed in large patient datasets, preferably with prospective study design.