Circulating and salivary microRNA expression analysis in nasopharyngeal carcinoma : potential for diagnostic biomarker

Abstract

Nasopharyngeal carcinoma (NPC) is among the most frequently reported cancer in Malaysia where it is usually diagnosed at late stages. The inconvenient and painful contemporary diagnostic methods for NPC discourage the population at risk from being screened at early stages. Thus, the discovery on non-invasive biomarkers for early detection of NPC is warranted. MicroRNAs (miRNAs), a class of RNAs that have been found to circulate stably in body fluids, are widely studied as potential biomarkers for NPC. While many countries with high NPC prevalence are actively studied extracellular miRNAs as diagnostic biomarker for NPC, such studies in Malaysia were scarce. Thus, this study aimed to identify the circulating and salivary miRNAs as early detection biomarker of the disease by analysing their differential expression. Blood and saliva samples were collected from 37 NPC and 37 control subjects in the states of Pahang and Kelantan, and were subjected to miRNA extraction. miRNA extracts from NPC (n=10) and control (n=11) plasma samples were used to screen the differentially expressed miRNAs using Taqman® Low Density Array card A and B. The significant differentially expressed (p<0.05) miRNAs in plasma of NPC subjects were selected for validation using 36 plasma and 37 saliva samples of NPC and control subjects. The consistent differential expression of miRNAs in plasma, as well as their corresponding miRNAs in saliva, were further analysed to evaluate their diagnostic performance. The selected miRNAs were cross-validated to select the best combination of miRNA model in predicting NPC. ROC curve analysis was used to evaluate the diagnostic performance of miRNA models. The result on circulating miRNA screening showed that eleven miRNAs were significantly differentially expressed (p<0.05) in NPC as compared to control subjects. The validation on eight selected miRNAs revealed that four miRNAs, namely hsa-miR-150, hsa-miR-205, hsa-miR-639 and hsa-miR-889, were consistently differentially expressed in plasma of NPC as compared to control subjects with significant result. The cross-validation showed three similar circulating and salivary miRNAs were selected as the best model in predicting NPC. Additionally, two circulating miRNAs, namely hsa-miR-150 and hsa-miR-205, and one salivary miRNA, namely hsa-miR-144#, were also proposed as miRNA models for diagnostic performance analysis due to their significant expression. The ROC curve analysis demonstrated that the model with two circulating miRNAs, namely hsa-miR-150 and hsa-miR-205, with adjustment for known risk factors of NPC was the best model in predicting NPC (AUC = 0.865) and early stage NPC (AUC = 0.860). Therefore, the present study proposed the two miRNAs, namely hsa-miR-150 and hsa-miR-205, as minimal invasive diagnostic biomarkers for early detection of NPC. The findings from this study is expected to provide more evidences for the gaps in discovery phase of biomarker development pipelines for NPC.