loading page

CORRELATION BETWEEN PET-CT UPTAKE VALUES AND PATHOLOGICALY FEATURES IN HEAD AND NECK CANCER
  • Atakan Sarıgül,
  • Vahit Mutlu
Atakan Sarıgül
Ataturk Universitesi Arastirma Hastanesi

Corresponding Author:[email protected]

Author Profile
Vahit Mutlu
Ataturk Universitesi Arastirma Hastanesi
Author Profile

Abstract

Objectives: To evaluate the correlation between SUV-Max values and pathological outcomes in head and neck squamous cell carcinoma (HNSCC) and determine the predictive power of SUV-Max for disease prognosis. Data Sources: Retrospective analysis of medical records and PET-CT imaging results from patients diagnosed with HNSCC at our institution between 2014 and 2023. Review Methods: Examination of SUV-Max values from F18-FDG PET-CT scans and their association with pathological findings such as perineural invasion, lymphovascular invasion, and neck lymph node metastasis. Statistical analysis was conducted to establish cutoff values and assess the significance of correlations. Results: Our study identified significant cutoff values for PET-CT SUV-Max that correlate with the pathological features of head and neck cancer. For primary tumors, a SUV-Max cutoff of 14.71 predicted neck metastasis with a sensitivity of 67.6% and specificity of 64.2%, demonstrating moderate diagnostic accuracy with an AUC of 0.648. Perineural invasion was optimally predicted at a cutoff of 13.28, with a sensitivity of 74%, specificity of 67.3%, and an AUC of 0.728. Similarly, a cutoff of 13.28 for lymphovascular invasion yielded a sensitivity and specificity of 63%, with an AUC of 0.628. Additionally, neck lymph node metastasis was effectively assessed with a SUV-Max cutoff of 2.74, achieving a sensitivity of 62.2%, specificity of 67%, and an AUC of 0.694. Conclusion: SUV-Max values from PET-CT scans are significant predictors of pathological outcomes in HNSCC, aiding in the stratification of patient prognosis and guiding clinical decision-making.
Submitted to Clinical Otolaryngology
Submission Checks Completed
Assigned to Editor
Reviewer(s) Assigned