Radioresistance of head and neck cancer
Radiotherapy is the most important treatment modality in head and neck cancer, with two thirds of patients treated with (chemo-)radiotherapy (1). With altered fractionated radiotherapy, the locoregional control rates for earlier stages are encouraging, but for stage III and IV tumors, locoregional control remains around 50% (2), leaving considerable need for improvement. Factors that contribute to control of the tumor are tumor site, stage, treatment schedule and dose, tumor volume, and HPV status (3–5). However, even after correcting for these factors, there are still differences in control rates. Such differences may result from differences in tumor microenvironment, tumor cell properties like hypoxia, rapid repopulation between fractions, the fraction of cancer stem cells or intrinsic radiosensitivity (6).
Intrinsic or cellular radiosensitivity is a term used to describe the process of one tumor cell being more resistant than another on the basis of different intracellular mechanisms, independent of microenvironmental factors.
An appropriate way to study intrinsic radiosensitivity is therefore in tissue culture in which potential confounding factors can be reduced or eliminated. It has indeed been shown that intrinsic cellular radiosensitivity significantly determines the outcome of radiotherapy in head and neck cancer (7). However, these data were attained using functional (cell survival) studies, giving limited or no information on genes or pathways involved and thus providing little help to the treating physician on how to improve treatment for patients with radioresistant tumors. We therefore searched for genetic and thus potentially assessable and targetable factors that affect intrinsic radioresistance in head and neck cancer.
mRNA to study radioresistance
mRNA profiling has been used to study radioresistance in cell lines. To date, however, such experiments have been mostly performed on either one or two cell lines only, or on the NCI-60 cell line panel, which contains no head and neck squamous cell carcinoma (HNSCC) lines (8, 9). Because it is known that radiosensitivity is partly dependent on the tissue of origin (e.g., lymphomas are more sensitive than solid tumors), use of such a cell line panel to predict HNSCC radiosensitivity is of questionable value. Therefore, Hall and colleagues attempted to identify a robust gene signature associated with intrinsic radiosensitivity on a series containing 16 cervical and 11 HNSCC cell lines. Unfortunately, they failed to identify such a set (10). Possibly this could be attributed to the fact that mRNA levels alone give an incomplete picture of active processes in the cell, as other factors can influence translation to protein. Among these are microRNAs (miR).
miRs are genomically encoded small pieces of single-stranded RNA of around 22 nucleotides each of which can silence hundreds of genes (11). More than 1,000 miRs have been identified so far, estimated to regulate expression of at least 60% of all genes (12). miRs regulate gene expression by binding to their (partly) complementary sequence on mRNA molecules, resulting in reduced protein production (13, 14). miRs can reduce protein production by causing degradation of mRNAs or by inhibiting translation. Multiple modes of silencing thus seem to exist that can be active concurrently (15, 16).
Ionizing radiation has been shown to induce significant changes in miR expression in 6 cancer cell lines (17). miRs playing a role in radioresistance have been described, although experiments were done in cell line pairs and not in a larger panel of cell lines (18–20).
The goal of this study was therefore to get a better insight into the genetic causes of intrinsic radioresistance in head and neck cancer cells focusing on miR expression. Using a large panel of HNSCC cell lines, we aimed to answer the following questions: (i) Do miR/mRNA expression changes induced by irradiation correlate with radioresistance?; (ii) Can we identify mRNAs that correlate with radioresistance?; (iii) Can we identify driving miRs that correlate with radioresistance?; (iv) If so, are these miRs and their targets related to certain pathways or processes?; and (v) Finally, do these miRs correlate with radiotherapy response in patients with laryngeal cancer? The answers to these questions should lead to a better understanding of radioresistance in this disease and therefore provide guidance toward more individualized treatment.