• Kiran B Jadhav, Vandana Shah, Amit Mahajan, Prachur Malhotra, Ghanshyam Parmar, and Jinesh Shah.
• Department of Oral Pathology and Microbiology, Vasant Dada Patil Dental College and Hospital, Maharashtra University of Health Sciences, Sangli, India. Electronic address: dr.kiranjadhav@yahoo.com.
• Lancet Oncol. 2022 Jul 1; 23 Suppl 1: S26.

BackgroundOral squamous cell carcinoma is a substantial health burden and one of the most common cancers worldwide. 40% of patients with oral squamous cell carcinoma have metastasis to cervical lymph nodes. Modern diagnostic aids for the assessment of lymph node metastasis have some limitations and drawbacks. miR-21 targets genes associated with the metastatic process in oral squamous cell carcinoma. The aim of the study was to evaluate the sensitivity and specificity of miR-21 for the assessment of cervical lymph node metastasis in patients with oral squamous cell carcinoma.MethodsThis work was conducted at Sumandeep Vidyapeeth, Vadodara, India. Unstimulated whole saliva and tumour tissue was obtained from patients with a clinically suspicious oral squamous cell carcinoma. The assessment of cervical lymph node metastasis was done before surgery by imaging techniques (CT or MRI) and post-surgically confirmed by histopathological examination of excised lymph nodes. miR-21 expression was evaluated using real-time PCR. Data were analysed for correlation analysis, cutoff values, sensitivity, and specificity. Kappa statistics were applied to assess the degree of agreement between the lymph node metastasis and miR-21 expression.Findings130 patients diagnosed with oral squamous cell carcinoma were included. miR-21 expression showed a significant correlation with cervical lymph node metastasis, with a diagnostic accuracy of 65-72% in saliva and 69-82% in tumour tissue. The mean cutoff value, defined as the value of fold (ie, to the power of) change indicating maximum sensitivity and specificity of miR-21 expression, was 2·32 cycle threshold (ct) for miR-21-5p (sensitivity 42·6%, specificity 90·3%) and 2·16 ct for miR-21-3p (sensitivity 60·3%, specificity 83·9%) in saliva, and 1·80 ct for miR-21-5p (sensitivity 76·5%, specificity 61·3%) and 0·89 ct for miR-21-3p in tumour tissue (sensitivity 82·4%, specificity 80·6%). We observed that when miR-21 expression is above the cutoff score, the probability of lymph node metastasis was higher. The independent t test showed a significant correlation (p<0·001) between cervical lymph node metastasis and miR-21 expression in saliva and tumour tissue, but not for miR-21-3p expression in tumour tissue (p=0·11). Very good agreement (Cohen's kappa=0·63) was observed between tumour tissue miR-21-3p and cervical lymph node metastasis, with a specificity of 80·60% and a sensitivity of 82·40%. The statistical analysis for correlation between saliva and tumour tissue miR-21 expression and age, sex, site of tumour (eg, buccal mucosa, tongue), and tobacco consumption habits did not show any significant correlation, but a significant correlation was observed with one-way ANOVA testing for the comparison between TNM stage and miR-21-5p and miR-21-3p expression in saliva and tumour tissue (p<0·0001).InterpretationmiR-21 expression in saliva and tumour tissue samples from patients with oral squamous cell carcinoma showed high diagnostic accuracy for assessment of cervical lymph node metastasis, and it could be used as an alternative for the assessment of cervical lymph node metastasis before surgery.FundingNone.Copyright © 2022 Elsevier Ltd. All rights reserved.

22 keywords...

hide…