Article Figures & Data
Figures
- FIGURE 1.
NTCP computation scheme. (Top) NTCP computation using simulated lobule dose distribution obtained from synthetic arterial tree. (Bottom) Potential individualization of lobule dose distribution derived from difference between simulated and patient voxel 90Y biodistribution. 90Y biodistribution was computed from synthetic arterial tree convolved with PET resolution to allow its comparison with typical 90Y time-of-flight PET imaging of patient after glass microsphere radioembolization not crossing tumors (both for delivered dose of 120 Gy to liver).
- FIGURE 2.
Lobule radiosensitivity hierarchy model. (Left) Schematic representation of lobule. (Courtesy of Educational Resource Center, College of Veterinary Medicine at University of Georgia; art by Will McAbee.) Because hexagon side-to-side distance is about 1,200 μm, almost all lobule structures are closer than 500 μm from small arteriole or venule (highlighted in red) still transporting blood from only surviving triad (right one). (Right) Lobule is assumed to be serial tissue regarding portal triads, hepatic cells, and centrilobular vein subsets, whereas each subset is assumed to be parallel tissue regarding their respective subunits.
- FIGURE 3.
Microscale dose distributions. (A) Total lobule fraction as function of BED delivered to hepatocytes and centrilobular vein subsets. (B) Total lobule fraction as function of BED of less irradiated portal triad. Lobule dose distributions were derived from resin and glass microsphere transport dynamics in synthetically grown hepatic arterial tree (8) for constant number (2.9) of portal arteries in 81% of triads (C) and for several portal arteries per triad following Poisson law of 2.4 mean (P). Distributions were computed for predicted TD50 in two-thirds liver radioembolization (Table 1). Similar lobule dose distributions are obtained for the 2 radioembolization devices regarding less irradiated portal triad (B).
- FIGURE 4.
Comparison of observed and predicted liver NTCP. Shown are Lyman–Burman Kutcher fit (red line) of NTCP observed in SPECT/CT dosimetry clinical study (asterisks) mixing whole-liver (35 patients), right-lobe (35 patients), and left-lobe (3 patients) resin radioembolization; SD (vertical bars); 95% confidence interval (dashed line); liver absorbed dose (D); predicted NTCP in whole-liver and two-thirds liver resin radioembolization (RE) computed using new radiobiologic model (black lines); and predicted NTCP accounting for proportion of whole-liver, right-lobe, and left-lobe RE present in each observed point (circles). (Data are from Strigari et al. (3) and from written communication with Strigari, October 2013). CI = confidence interval.
- FIGURE 5.
TD50 as function of targeted liver volume fraction Vf for different microsphere activities computed using radiobiologic model. Circles are Monte Carlo simulations. Curves are fit using Equation 8. Red and blue colors correspond to glass and resin microspheres, respectively, at day of calibration. Purple color corresponds to glass microspheres used 8 d after calibration. Diamond is TD50 observed in right-lobar radioembolization using glass microspheres (4). Square is TD50 observed in mixing of right-lobar and whole-liver radioembolization using resin microspheres (3).
Tables
- TABLE 1
Observed and Predicted Doses (Gy) Giving NTCP of 0.5 for Different Irradiation Modalities
Clinical studies EUDKB50 (11) EUDJH50 (12) Predicted TD50 (this work) Modality and liver fraction Observed TD50 Observed BED50 EBRT 1/1 43 (5) 77 77 77 43 EBRT 4/5 53 (5) 95 77 70 52 EBRT 2/3 64 (5) 115 78 66 64 Resin 1/1 + 2/3 52* (3) 93 239 78 55 Glass 2/3 100 (4) 252 433 77 99 ↵* Mixed whole-liver and right-lobe radioembolization.
Supplemental Data
Files in this Data Supplement:
Jump to section
Related Articles
Cited By...
- No citing articles found.