Comparing outcomes from CT and PET imaging after tumor liver radioembolization with 90Y-microspheres might suggest more appropriate choice of response evaluation

Objectives Radioembolization(RE) with 90Ymicrospheres is promising treatment of unresectable liver tumors. RECIST criteria based on CT imaging are predominantly applied for patients’ follow-up. Conversely, many authors consider FDG-PET functional most useful for response evaluation. Aim of this study was to compare the objective responses derived by quantitative parameters of CT vs.PET staging images.

Methods 20 patients underwent radioembolization (resin 90Y-microsferes, 0.8-2.8GBq) after basal FDG PET/CT, 99mTcMAA SPECT, dosimetry. Patients were monitored with PET & CT, 6weeks(16pts), 12weeeks (12pts), and >6months (7pts) after RE. Lesions were monitored as for maximal StandardizedUptakeValue(SUV) and mass extension(mCT). The %variations of mCT & SUV vs. basal examinations were evaluated to establish Complete-Response(CR), Partial-Response(PR), Stable-Disease(SD), Progressive-Disease(PD)based on published criteria. Dose-response relationship was also investigated.

Results 22 lesions were studied. Median(range) absorbed doses was 82(30-140)Gy. No correlation emerged between mCT and SUV %variations. At 12weeks lesion responses were: 9%(CR), 55%(PR), 27%(SD), 9%(PD) from CT(mCT); 21%(CR), 21%(PR), 16%(SD), 42%(PD) from PET(SUV). PET was often able to anticipate the response as compared with CT, mainly in case of progression. Considering responses from PET, tumor doses(Gy) ranged: (50-140)CR+PR; (95-110)SD; (30-120)PD, showing no significant correlation but definitively lower PD rate for doses>80-100Gy.

Conclusions PET parameters seems to predict response more properly. Often CT seemed to underestimate response, being not able to account for cell death rapidity and time delay for liver remodeling. Although more data are needed to confirm results, useful indications emerged to identify most appropriate parameters predictive of RE responses