Although papillary thyroid cancer can remain indolent, associated lymph node metastases and recurrence rates are approximately 50% and 20%, respectively. Omics-based medicine has led to the discovery of predictive biomarkers that can be used to predict tumor progression and clinical outcomes. We aimed to develop a noninvasive omics-driven blood test to allow accurate risk stratification and help tailor individual patient treatment plans. RNA sequencing (seq) and microRNA analysis of The Cancer Genome Atlas and Gene Expression Omnibus datasets were employed to identify an epigenetic prognostic panel. Integrated bulk assay for transposase-accessible chromatin-seq and RNA-seq experiments confirmed the results. Sixty-two paired tumor and adjacent control thyroid tissues and 67 blood samples (62 papillary thyroid cancer and 5 controls) were analyzed for validation using sequencing and real-time polymerase chain reaction and correlated to clinical outcomes. A liposome-exosome fusion clustered regularly interspaced short palindromic repeats (CRISPR)-fluorescent detection system miRNA assay was developed. A predictive risk nomogram was generated and tested for performance. Our miRNA panel (miR-146b-5p and miR-221-3p) from tissue and blood was associated with aggressive features and was located within accessible chromatin regions. The miRNA risk score and prognostic nomogram showed higher accuracy in predicting lymph node metastases (miR-146b: area under the curve [AUC] 0.816, sensitivity 76.9%; miR-221: AUC 0.740, sensitivity 79.5%) and recurrence (miR-146b: AUC 0.921, sensitivity 75.0%; miR-221: AUC 0.756, sensitivity 70.0%; p < 0.001) than staging and American Thyroid Association risk stratification. CRISPR-based miRNA assays showed upregulation in the blood of cancer cohorts. CRISPR-based detection of miR-146b and miR-221 in the blood of thyroid cancer patients is a reliable and noninvasive tool for real-time assessment and prognostication that has great potential to provide a direct impact on the care of these patients.
Read full abstract