(Sunday, 7/14/2019)
Room: ePoster Forums
Purpose: Current immunotherapies are ineffective in patient treatments and response rates remain at about 20%. Predictive models, distinguishing responders from non-responders, could help guide clinical practice. We hypothesize that MR-textures hold predictive potential for the outcome of immunotherapy treatments.
Methods: In fifteen 6-8 week old female BALB/c mice, 4T1 tumors were established subcutaneously on the right flank by injection of 1x106 cultured tumor cells in 100 µL of matrigel. After tumor volumes reached ~200 mm3 the cohort was randomized in three equal-sized groups. First group was used as control, second group received PDL1 inhibitor (2 x 100pmole) as a monotherapy, and the third group received radiotherapy (3 x 8 Gy) followed by intraperitoneal administration of the PDL1 inhibitor (2 x 100pmole). The treated animals were MRI-imaged before therapy, tumors were segmented, and 92 geometric-, first-, second-, and third-order texture features were extracted. The primary end point of the study was to establish correlation between tumor histopathology and the extracted texture features. The tumors were harvested at the same time and measured for weight, and the histopathology slides were stained with Ki-67 as a marker of cell proliferation for an assessment of response to therapy.
Results: The average weight of the tumors in the control, immunotherapy, and immunotherapy+RT groups were 2.56, 2.44, and 1.41 grams respectively, with average Ki-67 values of 0.4644, 0.4223, and 0.042 (larger Ki-67 values correspond to higher proliferative indexes and indicate a more resistant responsive). In both treated cohorts (immunotherapy alone, and immunotherapy+RT) MRI textures correlated with Ki-67 results. In the immunotherapy cohort 14 features correlated with the histopathology, while in the immunotherapy+RT group 32 features followed the Ki-67 trend.
Conclusion: The results suggest that pre-treatment MR textures may predict treatment response for immunotherapy alone as well as for combined immunotherapy and RT regimens.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
