OBJECTIVES: In patients with metastatic pancreatic neuroendocrine tumors (PanNETs), the Ki-67 index is objectively assessed by pathologists
however, liver tumor burden (LTB) depends on the subjective judgment of physicians. This study aimed to elucidate the usefulness of the semi-automated 3D volumetric assessment of LTB in patients with PanNET. MATERIALS AND METHODS: We retrospectively reviewed 29 patients (40 computed tomographies [CTs]) with metastatic PanNETs. LTB was measured using a semiautomated 3D volumetric software program (volumetric assessment) or evaluated independently by 6 clinicians using CT imaging (visual assessment). The treatment map was classified into 3 groups based on LTB and Ki-67 index. RESULTS: Visual and volumetric assessments of the LTB were well correlated. The LTB was significantly higher on visual assessment than volumetric assessment (P <
0.01). Categorization on the map was consistent between the visual and volumetric evaluations in 23 patients (equal group). The remaining 6 patients were overestimated by visual assessment (overestimated group). Progression-free survival was significantly longer in patients in the 'equal group' than the 'overestimated group' (981 vs 366 days, P <
0.01). CONCLUSIONS: This pilot study revealed a good correlation between visual and volumetric assessments, and visual assessment overestimated LTB, compared to volumetric assessment.