Epidermal growth factor receptor status and ligand expression influence the treatment sensitivity of HNSCC cells and may be useful as predictive markers.

Our data suggest that the amount of EGF is a determinant of the tumor cell proliferation rate and the response to cetuximab treatment in tongue cancer. Thus, EGF is a potential predictive biomarker of poor cetuximab response and a possible treatment target.

High sensitivity to cetuximab was strongly associated with a treatment-induced reduction in pEGFR both in vivo and in vitro in a panel of HNSCC cell lines, suggesting that EGFR and pEGFR dynamics could be used as a predictive biomarker for cetuximab treatment response.

Abstract. The present study was undertaken to evaluate the possibility of using a panel of proteins and single nucleotide polymorphisms (SNPs) involved in apoptosis, growth control, and DNA repair as predictive markers for cisplatin sensitivity. For this purpose the intrinsic cisplatin sensitivity (ICS) was determined in 39 cell lines derived from squamous cell carcinomas of the head and neck using a colony-forming assay. In these cell lines and in normal oral keratinocytes (NOK), the expression of epidermal growth factor receptor (EGFR), Hsp70, Bax, Bcl-2, Bcl-XL, survivin, and COX-2 was determined. Moreover, the p53, MDM2, FGFR4, XPC, XPD, XRCC1, and XRCC3 genes were analyzed for the presence of specific single nucleotide polymorphisms (SNPs). Pearson's correlation test showed that EGFR was the only protein that was significantly correlated to the ICS (r=0.388, p=0.015). The combination of EGFR, Hsp70, Bax, and Bcl-2 gave the strongest correlation (r=0.566, p≤0.001), whereas Bax alone had the second highest influence on the ICS. Furthermore, all four SNPs within genes involved in DNA repair, i.e. XPC, XPD, XRCC1, and XRCC3, tended to influence the ICS. In order to find the combination of factors, on both protein and gene levels, with the highest correlation to ICS, a multivariate statistical calculation was performed. Our results indicate that SNPs in DNA repair genes (XRCC3 241 and XPD 751 ) influence the ICS and together with the expression of EGFR, Hsp70, Bax, and Bcl-2, they could predict the cisplatin sensitivity of head and neck cancer cell lines (r=0.614, p≤0.001).

Head and neck squamous cell carcinoma (HNSCC) originates from the epithelial lining of the upper aerodigestive tract. It accounts for over 90 % of the malignancies found in the head and neck region. 600,000 new cases of HNSCC occur each year worldwide. Apart from causing painful lesions, HNSCC directly impacts the patient’s fundamental functions such as breathing and eating and also can disrupt the patient’s senses such as smell, taste, speech and even vision. Most cases of HNSCC require a combination of different treatments such as surgery, chemotherapy (primarily cisplatin based), and radiotherapy. Treatment decisions are largely based on the size of the tumor, the involvement of local lymph nodes, and distant spread. Treatment resistance and local recurrence are significant problems and to date no form of clinical treatment sensitivity  prediction is available. A majority of HNSCC tumors overexpresses the epidermal growth factor receptor (EGFR). This receptor is involved in proliferation and DNA repair and is the target of a monoclonal antibody named cetuximab that selectively binds and inhibits EGFR. It is the only targeted therapy available to HNSCC patients and reserved for late stage patients in  Sweden. Numerous investigators have searched for predictive markers and we hypothesized that since HNSCC is a very heterogeneous disease a single factor would not be able to predict the treatment outcome. In paper I we explore a panel of predictive factors using a point system, called the number of negative points (NNP), in which we could combine both proteins and genetic variations in an attempt to find a set of markers that could predict the intrinsic cisplatin sensitivity (ICS). The expression level of EGFR, Hsp70, Bax, Bcl-XL, survivin, and COX-2 was determined in 39 HNSCC cell lines. Moreover, the p53, MDM2, FGFR4, XPC, XPD, XRCC1, and XRCC3 genes were analyzed for the presence of specific single nucleotide polymorphisms (SNPs). Pearson’s correlation tests showed that EGFR was the only protein that alone correlated to ICS (r=0.388, P=0.015). The strongest correlation to ICS was found when combining SNPs in XRCC3 and XPD with the expression of EGFR, Hsp70, Bax, and Bcl-2 using the NNP system (r=0.614 P≤0.001). In paper II we assess the intrinsic radiosensitivity (IR), the ICS, and the intrinsic cetuximab sensitivity (ICmabS) as well as their combinations in 25 HNSCC cell lines established from HNSCC biopsies taken at the Department of Otorhinolaryngology and Head and Neck Surgery at Linköping University Hospital. Furthermore we investigate the EGFR status (consisting of EGFR gene copy number, EGFR mRNA, EGFR protein, pEGFR), pAkt and mRNA levels of the seven known EGFR ligands. No correlation was found between the different treatment sensitivities. Cetuximab treatment response was significantly correlated to epiregulin (EREG) mRNA expression (r=-0.408, P=0.043). Cetuximab resistant cell lines tended to have low levels of pEGFR (P=0.13) while resistant cell lines had a significantly lower expression of EGFR pro...