As EGFR is a self-antigen, the frequency of EGFR-specific CTL is expected to be low in the peripheral blood of HNSCC patients, and the ability of these cells to recognize EGFR
+ tumor cells to be weak. Using a sensitive and specific method available for the detection of rare peptide-specific T cells, we have been successful in establishing that EGFR-specific CD8
+ T cells are present in the circulation of HNSCC patients with high EGFR scores. Additionally, the frequency of EGFR-specific CTL in the peripheral blood of HNSCC patients correlated strongly with the EGFR expression on tumor samples. This correlation suggests that EGFR over-expression on the tumor cells clearly induces T cell responses in the periphery. Interestingly, only the combination of advanced tumor size (T2-4) and high EGFR score (> 7) was followed by a significant increase of EGFR-specific CTL. In case of small tumors (T1), the total amount of EGFR antigen expressed on the cell surface is probably insufficient to induce immune T cell response. In conclusion, the host immune response is too slow to inhibit tumor growth in its early stage. Moreover, it has been reported that suboptimal antigen doses presented by DC induce the development of Treg, while high antigen doses favor development of effector T cells [
16].
Additionally to confirming the existence of EGFR-specific CTL, we have succeeded in expanding autologous EGFR-specific T cells of HNSCC patients
in vitro. Expanded EGFR-specific CTL recognized EGFR on the surface of target cells, irrespective of whether these targets actively expressed the peptide or if they were exogenously loaded with EGFR peptides. This finding introduces the option for expanding EGFR-specific CTL
ex vivo for adoptive immunotherapy of HNSCC in addition to conventional surgery and chemo-radiotherapy [
17].
The current results complement our earlier studies of cellular immune responses to other tumor-associated antigens, such as wild-type p53 peptides and HPV-16. Using the tetramer-based technique in previous studies, elevated frequencies of HPV-specific CTL were detected in HNSCC patients with HPV/p16
+ tumors [
18]. In another tetramer-based study, the frequency of CTL specific for the self-antigen p53 showed an inverse correlation to p53 expression in the tumor. The frequency of p53-specific CTL was increased in HNSCC patients whose tumors had a normal p53 expression, whereas it was decreased in tumors with high p53 expression, which was explained by epitope loss under immune pressure [
11]. Further tetramer studies indicated that p53-specific CTL decreased in the peripheral blood after surgery of HPV
+ HNSCC but not in HPV-negative HNSCC [
19]. The detection of EGFR-specific CTL in the circulation of HNSCC is in line with other studies which used different EGFR-specific peptides. Andrade et al. found, that treatment of tumor cells with cetuximab increased their recognition by EGFR-specific CTL
in vitro[
20].
Despite the elevated frequency of EGFR-specific CTL in the circulation of HNSCC patients with high EGFR score, tumor growth was not inhibited. These results are counterintuitive but may be explained by one or more of the following events:
(a) the EGFR-peptide is presented in association with HLA-A2.1 on the tumor surface in a confirmation unrecognizable by T cells [
21], or DC in tumor-bearing individuals might have impaired antigen presenting capability [
22]. Consequently, adaptive immune responses to the tumor peptides are inefficient, and frequencies of EGFR-specific CTL remain low.
(b) Alternatively, apoptosis of tumor-specific T cells might be responsible for their low frequencies. As shown by Albers et al., annexin expression, which indicates apoptosis, is increased in wild-type p53-specific CTL compared to non-tumor specific T cells in HNSCC [
12].
(c) The presence of tumor-induced suppression in HNSCC patients, as evidenced by increased proportions of myeloid derived suppressor cells, tumor-derived microvesicles, and regulatory T cells at the tumor site and in the peripheral circulation may account for lack of immune responses to EGFR peptides [
23].
(d) Not only antigen presentation on the cell surface, but also the intracellular turnover of the protein might determine and modulate antigen recognition by the immune system, as observed for p53 [
24]. As p53 and EGFR both are self-antigens, this might also be true for EGFR recognition. Nevertheless, despite these various difficulties, EGFR-specific CTL were detectable in the peripheral blood of HNSCC patients and could be expanded
in vitro. Importantly, we found a strong correlation of specific T cell frequency and EGFR expression on tumor cells. Thus, the impairment most likely accounting for the insufficient EGFR-specific immune response in HNSCC patients might be related to the dose of antigen and tumor-derived immune suppression. Considering the presented results, the number of EGFR-specific CTL before and after tumor therapy in correlation to the frequency of regulatory T cells would be of high interest and will be addressed in future longitudinal studies. Further, our results suggest that subsequent studies of tumor therapy should not be limited to the monitoring of tumor regression. They should also focus on the effect which therapy has on various cell populations of the immune system, including regulatory T cells, MDSC, Th-17-cells, and antigen-specific CTL as well as their cytokine expression profile.