Predicting Tumor Response to Neoadjuvant Chemotherapy based on Pretreatment 18F-FDG PET/CT and Clinical Parameters in Locally Advanced Oral Squamous Cell Carcinoma

Introduction: The surgical treatment of locally advanced oral squamous cell carcinoma (OSCC) poses significant challenges due to its complexity and high risk of recurrence, representing a current clinical dilemma. Neoadjuvant chemotherapy (NC) has emerged as a promising strategy to reduce tumor size and recurrence risk, becoming a focal point in OSCC treatment research. However, due to tumor heterogeneity, not all patients are sensitive to NC, leading to variable treatment outcomes. Therefore, we aimed to analyze pretreatment ¹⁸F-FDG PET imaging parameters and clinical features in OSCC patients to predict the efficacy of NC and identify patients likely to benefit from it, thereby aiding in precise clinical stratification before treatment.

Methods: A retrospective analysis was conducted on 89 pathologically confirmed locally advanced OSCC patients in stage III-IV, including 62 primary and 27 recurrent cases. The study focused on pre-NC ¹⁸F-FDG PET/CT imaging parameters and clinical characteristics. Based on the RECIST 1.1 criteria, the efficacy of NC was evaluated using MRI after two chemotherapy cycles, while calculating the depth of response (DpR) for tumor. Univariate and multivariate regression analyses were utilized to identify significant predictive markers for overall NC efficacy and post-treatment DpR.

Results: In the NC non-responsive group, primary OSCC patients exhibited significantly larger tumor sizes, metabolic tumor volume (MTV), and total lesion glycolysis (TLG) compared to the responsive group (P=0.036，P=0.004，P=0.001). MTV and TLG were identified as independent predictors of treatment efficacy (P=0.019，P=0.012). With cut-off values set at MTV=11.7cm³ and TLG=84.2, the areas under curve (AUC) of the receiver operating characteristic (ROC) were 0.739 and 0.782, respectively (P=0.004, P<0.001). Pathological grading, standardized uptake value (SUV), MTV, and TLG emerged as independent predictors of post-NC DpR (P=0.011，P<0.05，P<0.001，P<0.001). Pathological grading, maximum SUV normalized to lean body (SULmax) and MTV account for 52.9% (R²) of the variance in DpR without significant collinearity (VIF<5.0). Among recurrent patients, those in the NC responsive group showed significantly higher SUVs compared to the non-responsive group (P<0.01).

Conclusions: ¹⁸F-FDG PET/CT imaging is a valuable tool for predicting the therapeutic response of OSCC patients to NC. MTV and TLG serve as independent predictive factors for the efficacy of NC in primary OSCC patients. Furthermore, tumor pathological grading, SUV, MTV, and TLG are independent predictors of post-treatment DpR. SUV holds potential significant predictive value for the NC response in recurrent OSCC patients.

• Download figure
• Open in new tab
• Download powerpoint