Changes in circulating microRNAs after radiochemotherapy in head and neck cancer patients

Abstract

Introduction: Circulating microRNAs (miRNAs) are easily accessible and have already proven to be useful as prognostic markers in cancer patients. However, their origin and function in the circulation is still under discussion. In the present study we analyzed changes in the miRNAs in blood plasma of head and neck squamous cell carcinoma (HNSCC) patients in response to radiochemotherapy and compared them to the changes in a cell culture model of primary HNSCC cells undergoing simulated anti-cancer therapy.

Materials and methods: MiRNA-profiles were analyzed by qRT-PCR arrays in paired blood plasma samples of HNSCC patients before therapy and after two days of treatment. Candidate miRNAs were validated by single qRT-PCR assays. An in vitro radiochemotherapy model using primary HNSCC cell cultures was established to test the possible tumor origin of the circulating miRNAs. Microarray analysis was performed on primary HNSCC cell cultures followed by validation of deregulated miRNAs via qRT-PCR.

Results: Unsupervised clustering of the expression profiles using the six most regulated miRNAs (miR-425-5p, miR-21-5p, miR-106b-5p, miR-590-5p, miR-574-3p, miR-885-3p) significantly (p = 0.012) separated plasma samples collected prior to treatment from plasma samples collected after two days of radiochemotherapy. MiRNA profiling of primary HNSCC cell cultures treated in vitro with radiochemotherapy revealed differentially expressed miRNAs that were also observed to be therapy-responsive in blood plasma of the patients (miR-425-5p, miR-21-5p, miR-106b-5p, miR-93-5p) and are therefore likely to stem from the tumor. Of these candidate marker miRNAs we were able to validate by qRT-PCR a deregulation of eight plasma miRNAs as well as miR-425-5p and miR-93-5p in primary HNSCC cultures after radiochemotherapy.

Conclusion: Changes in the abundance of circulating miRNAs during radiochemotherapy reflect the therapy response of primary HNSCC cells after an in vitro treatment. Therefore, the responsive miRNAs (miR-425-5p, miR-93-5p) may represent novel biomarkers for therapy monitoring. The prognostic value of this exciting observation requires confirmation using an independent patient cohort that includes clinical follow-up data.

Figure 1

Sensitivity of primary HNSCC cell cultures to 5-FU treatment and ionizing radiation. Cell viability (XTT-assay) of HN1957 (a) and HN2092 (b) 48 h after 5-FU treatment relative to DMSO-controls. Each data point represents the mean of three biological replicates. Dashed lines mark the 5-FU concentrations selected for in vitro radiochemotherapy modeling. Survival curves were generated by colony forming assay for HN1957 (c) 6 days after irradiation and HN2092 (d) 12 days after irradiation. Error bars represent the standard deviation of three biological replicates.

Figure 2

Unsupervised hierarchical cluster analysis of samples collected before therapy and after two days of treatment. Samples collected prior to treatment are marked with blue bars, samples collected after two days of treatment (XRT = radiotherapy) are marked with red bars. Cluster 1 and Cluster 2 indicate the two main clusters emerging from unsupervised cluster analysis using blood plasma concentration levels of the top six distinctive miRNAs.

Figure 3

Cell viability (XTT-assay) of primary HNSCC cell cultures. Absorption was measured 24 h after exclusive 5-FU treatment and 5-FU treatment in combination with 2 x 2 Gy in vitro irradiation relative to DMSO-controls. Error bars represent the standard deviation of three biological replicates (*p < 0.05, **p < 0.01).