Global microRNA expression profile in laryngeal carcinoma unveils new prognostic biomarkers and novel insights into field cancerization

Abstract

Laryngeal carcinoma is still a worldwide burden that has shown no significant improvement during the last few decades regarding definitive treatment strategies. The lack of suitable biomarkers for personalized treatment protocols and delineating field cancerization prevents further progress in clinical outcomes. In the light of this perspective, MicroRNAs could be promising biomarkers both in terms of diagnostic and prognostic value. The aim of this prospective study is to find strong prognostic microRNA biomarkers for advanced laryngeal carcinoma and molecular signatures of field cancerization. Sixty patients were enrolled and four samples were collected from each patient: tumor surface and depth, peritumor normal mucosa, and control distant laryngeal mucosa. Initially, a global microRNA profile was conducted in twelve patients from the whole cohort and subsequently, we validated a selected group of 12 microRNAs with RT-qPCR. The follow-up period was 24 months (SD ± 13 months). Microarray expression profile revealed 59 dysregulated microRNAs. The validated expression levels of miR-93-5p (χ²(2) = 4.68, log-rank p = 0.03), miR-144-3p (χ²(2) = 4.53, log-rank p = 0.03) and miR-210-3p (χ²(2) = 4.53, log-rank p = 0.03) in tumor samples exhibited strong association with recurrence-free survival as higher expression levels of these genes predict worse outcome. Tumor suppressor genes miR-144-3p (mean rank 1.58 vs 2.14 vs 2.29, p = 0.000) and miR-145-5p (mean rank 1.57 vs 2.15 vs 2.28, p = 0.000) were significantly dysregulated in peritumor mucosa with a pattern of expression consistent with paired tumor samples thus revealing a signature of field cancerization in laryngeal carcinoma. Additionally, miR-1260b, miR-21-3p, miR-31-3p and miR-31-5p were strongly associated with tumor grade. Our study reports the first global microRNA profile specifically in advanced laryngeal carcinoma that includes survival analysis and investigates the molecular signature of field cancerization. We report two strong biomarkers of field cancerization and three predictors for recurrence in advance stage laryngeal cancer.

Figure 1

Cluster analysis of significantly dysregulated miRNAs from the global microarray profile of 2549 microRNAs in 12 patients with advanced laryngeal carcinoma. Legend: (1) tumor surface; (2) tumor depth; (3) peritumor mucosa; (4) normal laryngeal mucosa.

Figure 2

Significant difference in expression levels of miR-181b-5p between tumor surface and depth (pairwise comparison, Wilcoxon signed-rank test, z = 2.42, p = 0.015).

Figure 3

Statistically significant differences in the expression levels of miR-144-3p and miR 145-5p between tumor tissue, peritumor mucosa and healthy control laryngeal mucosa. Pairwise comparison using the Friedman test (with Bonferroni correction); miR-144-3p [mean rank]: 1.58 versus 2.14 versus 2.29, p = 0.000; miR-145-5p [mean rank]: 1.57 versus 2.15 versus 2.28, p = 0.000.

Figure 4

(a–c): Kaplan–Meier curves illustrating that patients expressing higher levels of miR-93-5p, miR-210-3p and miR-144-3p had significantly worse survival rates than patients with lower expression levels (χ²(2) = 4.68, log-rank p = .03; χ²(2) = 4.53, log-rank p = .03, χ²(2) = 4.53, log-rank p = .03, respectively). (d–f) Spearman’s rank-order correlation reveals strong associations between miR-93-5p, miR-210-3p and miR-144-3p.

Figure 5

Illustration of some major cascades in cancer with a summary of all validated targets of miR-93-5p, miR-210-3p, miR-144-3p, and miR 145-5p in laryngeal carcinoma published in the literature. Blue color designates tumor suppressor genes, red color—oncogenes and references are given in brackets.