Purpose: To investigate dose-responses of regional liver functions from Gadoxetic Acid uptake rates (k1) mapped by DGAE-MRI and to relate them to baseline global and regional liver functions.
Methods: 25 patients with HCC were enrolled in an IRB-approved adaptive SBRT trial. The patients had DGAE-MRI of liver preRT, and one-month after 3 or 5 fx of SBRT on a 3T scanner to quantify k1 maps. The accumulated doses up to the post scan were converted into EQD2 (=2.5) and co-registered to pre/post-RT k1 maps. Percentage changes in k1 from pre- to post-RT were analyzed in dose bins of every 5-Gy and 5 pre-RT k1 bins. Dose-response patterns of the whole group and individuals were analyzed by regression analysis. Reponses patterns were further clustered into two groups using k-means and tested the group differences with ANOVA. Correlations between percentage changes in k1 and pre-, post-RT or changes in Child-Pugh scores were analyzed using correlations.
Results: The averaged percentage change in k1 of the whole group showed a linear relationship with 0.44%/Gy with an adjusted R^2=0.99 and a significant greater response at the highest pre-RT k1 value (P<0.0001). Two groups were clustered based on the dose-response patterns at low doses, with one group showing a linear dose-response to all doses and another having a non-linear but over-response to low doses and a linear response for the doses > 80 Gy. Responses of the two groups were significantly different (P<0.000002). The response patterns were not significantly correlated with pre, post-RT or changes in Child-Pugh scores.
Conclusion: Different dose-response subgroups (over-sensitive to low doses vs not) were identified based on changes in GA uptake rates from pre- to post-SBRT in the patients with HCC. This could have an impact on how to spare liver functional regions in personalized SBRT for HCC.
Funding Support, Disclosures, and Conflict of Interest: NIH/NCI P01 CA059827
Not Applicable / None Entered.
Not Applicable / None Entered.