Cancer Imaging (Oct 2023)

Concordance between head and neck MRI and histopathology in detecting laryngeal subsite invasion among patients with laryngeal cancer

  • Issa Mohamad,
  • Taher Abu Hejleh,
  • Monther Qandeel,
  • Maysa Al-Hussaini,
  • Sami Koro,
  • Ayat Taqash,
  • Abdelatif Almousa,
  • Fawzi Abuhijla,
  • Ramiz Abuhijlih,
  • Fatenah Ajlouni,
  • Akram Al-Ibraheem,
  • Dima Abu Laban,
  • Tariq Hussein,
  • Ebrahim Mayta,
  • Wisam Al-Gargaz,
  • Ali Hosni

DOI
https://doi.org/10.1186/s40644-023-00618-y
Journal volume & issue
Vol. 23, no. 1
pp. 1 – 9

Abstract

Read online

Abstract Background Accuracy of head and neck MRI (HN-MRI) in predicting tumor invasion of laryngeal site/subsites in patients with laryngeal cancer prior to laryngectomy is poorly evaluated in the literature. Therefore, we aim to evaluate the diagnostic value of HN-MRI in accurate pre-operative estimation of tumor invasion to laryngeal subsites in patients with laryngeal cancer. Methods Patients with laryngeal cancer who underwent HN-MRI for cancer staging and underwent total laryngectomy between 2008 and 2021 were included. Sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy of HN-MRI in predicting tumor invasion of laryngeal subsites were calculated based on concordance between the HN-MRI and histopathological results. Results One hundred and thirty-seven patients underwent total laryngectomy [primary: 82/137(60%), salvage 55/137(40%)]. The utilization of HN-MRI resulted in the downstaging of 16/137 (11.6%) patients and the upstaging of 8/137 (5.8%) patients. For the whole cohort, there was a significant discordance between HN-MRI and histopathology for T-category; out of 116 cT4a disease, 102(87.9%) were confirmed to have pT4a disease, and out of 17 cT3 disease, 9(52.9%) were confirmed to have pT3 disease, p < 0.001. The MRI overall diagnostic accuracy of predicting tumor invasion was 91%, 92%, 82%, 87%, 72%, 76%, 65% and 68% for base of tongue, arytenoid, vocal cord, posterior commissure, pre-epiglottic space, cricoid cartilage, inner thyroid cortex, and subglottis, respectively. Conclusions In patients with laryngeal cancer undergoing total laryngectomy, HN-MRI demonstrates promising accuracy in predicting tumor invasion of specific laryngeal subsites (e.g., base of tongue). Our findings showed the potential of HN-MRI as a valuable tool for pre-operative planning and treatment decision-making in this patient population.

Keywords