LAUSR

Purpose To check the impact on toxicity and on overall survival (OS) of a personalized two compartment (tumor vs non tumor) dosimetric treatment planning method in the treatment of hepatocarcinoma with 90Y glass microspheres. Methods 43 Child A strictly selected patients (series 1) had been previously treated with 90Y glass microspheres according to the standard mono-compartment planning of 120 Gy to the injected liver lobe. A retrospective analysis of 99mTc-MAA pre-treatment SPECT/CT images had allowed to determine NTCP and TCP [1] . A decompensation risk of 15% was adopted as planning limit, corresponding to NTCP15 = 75 Gy. Liver absorbed dose is averaged on the organ volume excluding tumors. Such planning method was then applied to 116 less selected Child A patients (series 3, comparable to series 1). Results With respect to series 1, administered activity in series 3 was doubled in half of patients, reduced of 50% in a quarter of patients. Toxicity rate was maintained below 15%. Patient were stratified as advanced (with Portal Vein Thrombosis (PVT)) vs intermediate (NO PVT). In the PVT subgroups, lesions had the same size distribution. Treatment planning improved median OS from 8 to 12 months (with Mantel Cox test (p = 0.067) i.e. close to significance). In the NO PVT subgroups the median OS was 17 vs 15 months (n.s.) i.e. maintained, despite the larger tumor sizes. Conclusions methodological limits were absence of scatter correction and mismatch between simulation and 90Y PET dosimetry. Results are really encouraging, since are the first indications of improved outcome thanks to dosimetric treatment planning.