Abstract
Introduction: Squamous cell(SqK) & adenocarcinoma(AdK) are most common NSCLC types;histology & staging are crucial.Conventional imaging are sometimes inaccurate in assessment of grading & lymph node(LN)metastases.
Aims/Objectives: To investigate predictive value of PET Metabolic tumor volume (MTV) with histology & differentiation. To evaluate a predictor of N2 metastasis: SUV LN/T [ratio between maximum standardized uptake values (SUVmax) of tumor and regional LN].
Methods: 310 consecutive patients (2006-2014) with NSCLC included. All underwent preoperative PET. MTV was delineated on PET;non-avid area excluded as MTV. SUVmax within MTV was calculated.Decision to perform first-line surgery was taken in multidisciplinary meeting. All were treated by pulmonary resection & systematic LN dissection. Comparison was performed with t–test. Relationships were evaluated using Pearson's coefficients.Values plotted nearest upper left corner from Receiver Operating Characteristics(ROC)analyses indicated best diagnostic accuracy. Using appropriate cut-offs,sensitivity,specificity,accuracy were calculated. Diagnostic accuracy was compared using Areas under ROC Curves.
Results: MTV values were significantly higher for SqK vs AdK. Significant differences were observed among MTV and grading. There was positive correlation between MTV & SUVmax of primary tumor(r=0.75). Incidence of N2 LN involvement was 10.6%. SUV LN/T highly correlate with LN pathology(r=0.79). ROC curves for SUV LN/T cut-off≥0.65 was predictive of N2 malignancy.
Conclusions: MTV could help to predict biological information like proliferative activity,histological subtypes,tumor grading. SUV LN/T may be a predictor of N2 metastasis.
- Copyright ©ERS 2015
