Table 2.
Comparison of dosimetric parameters according to vertebral compression fracture.
| Parameters | Mean value* |
Comparison of mean value | AUC† | Logistic regression |
|||
|---|---|---|---|---|---|---|---|
| VCF+ (95% CI) | VCF− (95% CI) | p-value* | p-value‡ | OR | 95% CI | ||
| PTV_Dmin (cGy) | 1159 (985, 1333) | 1049 (1006, 1093) | 0.44 | 0.574 | – | – | – |
| PTV_Dmean (cGy) | 1896 (1617, 2174) | 1771 (1722, 1820) | < 0.01 | 0.587 | – | – | – |
| PTV_Dmax (cGy) | 2264 (1854, 2674) | 2085 (2033, 2137) | 0.02 | 0.641 | – | – | – |
| PTV_D10% (cGy) | 2247 (1926, 2569) | 2002 (1960, 2044) | < 0.01 | 0.620 | – | – | – |
| PTV_D50% (cGy) | 1974 (1791, 2158) | 1803 (1755, 1851) | 0.01 | 0.595 | – | – | – |
| PTV_D60% (cGy) | 1915 (1746, 2084) | 1746 (1696, 1796) | 0.02 | 0.593 | – | – | – |
| PTV_D70% (cGy) | 1859 (1692, 2026) | 1684 (1632, 1737) | 0.02 | 0.596 | – | – | – |
| PTV_D80% (cGy) | 1793 (1622, 1964) | 1617 (1563, 1671) | 0.02 | 0.600 | – | – | – |
| PTV_D90% (cGy) | 1706 (1527, 1884) | 1526 (1472, 1581) | 0.02 | 0.609 | – | – | – |
| PTV_D95% (cGy) | 1663 (1455, 1816) | 1443 (1389, 1496) | 0.01 | 0.633 | – | – | – |
| PTV_V10 Gy (cm3) | 83 (59, 107) | 56 (48, 63) | < 0.01 | 0.677 | – | – | – |
| PTV_V15 Gy (cm3) | 78 (54, 103) | 46 (39, 53) | < 0.01 | 0.710 | < 0.01 | 1.11 | 1.05–1.13 |
| PTV_V20 Gy (cm3) | 31 (8, 53) | 14 (11, 18) | 0.01 | 0.623 | – | – | – |
| PTV_V25 Gy (cm3) | 11 (0, 33) | 1 (0, 2) | < 0.01 | 0.554 | – | – | – |
| PTV volume (cm3) | 83 (59, 107) | 57 (49, 64) | < 0.01 | 0.668 | – | – | – |
Abbreviations: VCF, vertebral compression fracture; VCF+, patients with VCF; VCF-, patients without VCF; CI, confidence interval; AUC, area under the curve; OR, odds ratio; PTV, planning target volume; Dmin, minimal radiation dose to PTV; Dmean, mean radiation dose to PTV; Dmax, maximal radiation dose to PTV; Dx%, radiation dose receiving × % volume of PTV; Vx Gy, PTV volume receiving more than × cGy.
*Analyzed by the Student’s t-test.
†Area under the receiver operating characteristics curve.
‡Analyzed by a multiple logistic regression analysis and stepwise selection.