Stereotactic body radiation therapy for liver oligometastases: Predictive factors of local response by 18F-FDG-PET/CT

Objective: To investigate metabolic parameters as predictive of local response after stereotactic body radiation therapy (SBRT) for liver-oligometastases. Methods: Inclusion criteria of the present retrospective study were: (a) liver oligometastases with controlled primary tumor; (b) absence of progressive disease ≥6 months; (c) metastases ≤ 3; (d) evaluation of SBRT-response by means of 18-fludeoxyglucose-PET/CT for at least two subsequent evaluations; (e) Karnofsky performance status >80; (f) life-expectancy >6 months. The following metabolic parameters were defined semi-quantitatively for each metastases: (1) standardized uptake value (SUVmax; (2) SUV-mean; (3) metabolic tumor volume (MTV), tumor volume with a SUV ≥3, threshold 40%; (4) total lesion glycolysis (TLG), i.e. the product of SUV-mean and MTV. Local control was defined as absence of recurrence in the field of irradiation. Results: 41 liver metastases were analyzed. Pre-SBRT, median SUV-max was 8.7 (range, 4.5-23.59), median SUV-mean was 4.6 (range, 3-7.5), median MTV was 5.7 cc (range, 0.9-80.6) and median total lesion glycolysis was 24.1 (range, 3.6-601.5). At statistical analysis, metastases with SUV-mean >5 (p 0.04; odds ratio 4.75, sensitivity = 50%, specificity = 82.6%, area under the curve 0.66) and SUV-max >12 (p 0.02; odds ratio 5.03, sensitivity = 69%, specificity = 70%, area under the curve = 0.69) showed higher rates of infield-failure compared to the remaining lesions. Conclusion: According to current findings, pre-SBRT SUV-max and SUV-mean could be predictable of local response in liver oligometastases. Advances in knowledge: Present findings could support the hypothesis that fludeoxyglucose-PET/CT may be a powerful tool to predict tumor control. Specifically, current results might be helpful for clinicians in the decision- making process regarding liver oligometastatic patient selection as well as the individual therapy stratification distinguishing between slowly local progressing patients and rapidly progressing patients.