Predictive Model of Paraaortic Lymph Node Involvement in cN0 Locally Advanced Cervical Cancers: PET/CT Technology...

Area of Science:

• Oncology
• Nuclear Medicine
• Radiology

Background:

• Locally advanced cervical cancer (LACC) poses challenges in detecting occult paraaortic lymph node (PALN) involvement.
• Accurate prediction of PALN metastasis is crucial for treatment planning and patient outcomes.

Purpose of the Study:

• To develop and validate a predictive model for occult PALN involvement in LACC patients using 18F-FDG PET/CT parameters.
• To evaluate the impact of time-of-flight (TOF) reconstruction technology on PET/CT performance for PALN detection.
• To compare the performance of the proposed model against existing clinical practice scores.

Main Methods:

• Retrospective observational cohort study of 71 LACC patients undergoing 18F-FDG PET/CT before PALN surgical staging.
• Collected PET/CT parameters included tumor SUVmax, tumor metabolic tumor volume (MTV), and pelvic lymph node involvement.
• Multivariate regression analysis and ROC curves were used to develop and assess the predictive model.

Main Results:

• Occult PALN involvement was identified in 12.7% of patients.
• The derived PET model incorporated pelvic lymph node counts, tumor-to-liver SUV ratio, and the absence of TOF technology (OR 21.4).
• The model achieved 77.8% sensitivity and 88.7% specificity, with performance significantly influenced by TOF usage. No significant difference was found compared to clinical practice scores (p=0.55).

Conclusions:

• PET/CT technology, specifically the use of time-of-flight, significantly affects the detection of occult PALN involvement in LACC.
• The proposed PET-based model demonstrates potential for improved prediction of PALN metastasis.
• Consideration of TOF technology is recommended for future revisions of PET-based scoring systems in LACC management.