Do circulating long non-coding RNAs (lncRNAs) (LincRNA-p21, GAS 5, HOTAIR) predict the treatment response in patients with head and neck cancer treated with chemoradiotherapy?

Long non-coding RNAs (lncRNAs) have been shown to be aberrantly expressed in head and neck cancer (HNC). The aim of the present study was to evaluate plasma levels of three lncRNA molecules (lincRNA-p21, GAS5, and HOTAIR) in the treatment response in HNC patients treated with radical chemoradiotherapy (CRT). Forty-one patients with HNC were enrolled in the study. Most of the patients had nasopharyngeal carcinoma (n = 27, 65.9 %) and locally advanced disease. Blood was drawn at baseline and treatment evaluation 4.5 months after therapy. lncRNAs in plasma were measured by semiquantitative PCR. Treatment response was evaluated according to clinical examination, RECIST and PERCIST criteria based on magnetic resonance imaging (MRI), and positron emission tomography with computed tomography (PET/CT) findings. Complete response (CR) rates were 73.2, 36.6, and 50 % for clinical investigation, PET/CT-, or MRI-based response evaluation, respectively. Predictive value of lncRNAs was investigated in patients with CR vs. those with partial response (PR)/progressive disease (PD). We found that post-treatment GAS5 levels in patients with PR/PD were significantly higher compared with patients with CR based on clinical investigation (p = 0.01). Receiver operator characteristic (ROC) analysis showed that at a cutoff value of 0.3 of GAS5, sensitivity and specificity for clinical tumor response were 82 and 77 %, respectively. Interestingly, pretreatment GAS5 levels were significantly increased in patients with PR/PD compared to those with CR upon MRI-based response evaluation (p = 0.042). In contrast to GAS5, pretreatment or post-treatment lincRNA-p21 and HOTAIR levels were not informative for treatment response. Our results suggest that circulating GAS5 could be a biomarker in predicting treatment response in HNC patients.