Table 3.
Bexarotene | Placebo | Bexarotene‐placebo change | ||||
---|---|---|---|---|---|---|
Voxel number | Unadjusted mean (SD) change in lesional voxel MTR (pu) | Voxel number | Unadjusted mean (SD) change in lesional voxel MTR (pu) | Adjusted bexarotene‐placebo difference (95% CI) | p‐value | |
All voxels | 10,921 | −0.04 (2.68) | 8,384 | −0.1 (2.32) | 0.32 (−0.23, 0.87) | 0.227 |
All voxels from pure WM lesions | 9,650 | −0.08 (2.48) | 7,125 | 0.06 (1.93) | 0.21 (−0.33, 0.75) | 0.445 |
All voxels from pure GM lesions | 1,271 | 0.26 (3.85) | 1,259 | −0.98 (3.72) | 1.45 (0.16, 2.74) | 0.036 |
All voxels from pure CGM lesions | 1,040 | 0.48 (4.06) | 929 | −0.85 (4.12) | 1.25 (−0.20, 2.70) | 0.105 |
All voxels from pure DGM lesions | 231 | −0.74 (2.48) | 330 | −1.33 (2.2) | 2.24 (0.44, 4.04) | 0.041 |
Effect of baseline voxel MTR: voxels from pure WM lesions | ||||||
Quartile 1 [lowest] | 2,225 | 0.41 (2.71) | 1,968 | 0.53 (2.01) | 0.69 (0.13, 1.25) | 0.020 |
Quartile 2 | 2,199 | 0.15 (2.53) | 1,995 | 0.02 (1.89) | 0.39 (−0.16, 0.94) | 0.175 |
Quartile 3 | 2,553 | −0.15 (2.30) | 1,640 | −0.01 (1.82) | 0.15 (−0.39, 0.70) | 0.592 |
Quartile 4 [highest] | 2,673 | −0.61 (2.28) | 1,521 | −0.44 (1.82) | 0.04 (−0.50, 0.59) | 0.871 |
Interaction test comparing treatment group differences between the quartiles | <0.0001 | |||||
Effect of baseline voxel MTR: voxels from pure CGM lesions | ||||||
Quartile 1 [lowest] | 283 | 1.18 (4.61) | 209 | 0.48 (3.66) | 0.81 (−0.71, 2.34) | 0.306 |
Quartile 2 | 266 | 1.16 (3.93) | 226 | −0.40 (3.86) | 1.06 (−0.44, 2.56) | 0.181 |
Quartile 3 | 251 | 0.50 (3.4) | 241 | −0.77 (3.54) | 1.26 (−0.23, 2.74) | 0.112 |
Quartile 4 [highest] | 239 | −1.09 (3.68) | 253 | −2.43 (4.69) | 1.62 (0.11, 3.13) | 0.047 |
Interaction test comparing treatment group differences between the quartiles | 0.047 | |||||
Effect of baseline voxel MTR: voxels from pure DGM lesions | ||||||
Quartile 1 [lowest] | 50 | 0.60 (2.85) | 90 | −0.87 (2.46) | 2.19 (0.36, 4.02) | 0.046 |
Quartile 2 | 47 | −0.54 (2.03) | 93 | −0.76 (2.19) | 2.07 (0.25, 3.89) | 0.057 |
Quartile 3 | 65 | −1.16 (1.92) | 75 | −1.27 (1.61) | 1.75 (−0.05, 3.56) | 0.094 |
Quartile 4 [highest] | 68 | −1.61 (2.19) | 72 | −2.68 (1.83) | 2.03 (0.17, 3.88) | 0.064 |
Interaction test comparing treatment group differences between the quartiles | 0.517 | |||||
Effect of baseline voxel T1: voxels from pure WM lesions | ||||||
Quartile 1 [lowest] | 2,264 | −0.79 (2.58) | 1,929 | 0.3 (1.7) | 0.16 (−0.45, 0.78) | 0.611 |
Quartile 2 | 1,783 | 0.21 (2.18) | 2,411 | 0.06 (1.96) | −0.02 (−0.60, 0.56) | 0.952 |
Quartile 3 | 2,810 | 0.18 (2.62) | 1,383 | 0.18 (2.07) | 0.02 (−0.54 0.58) | 0.951 |
Quartile 4 [highest] | 2,792 | 0.05 (2.31) | 1,402 | −0.42 (1.95) | 0.30 (−0.26, 0.85) | 0.298 |
Interaction test comparing treatment group differences between the quartiles | 0.643 | |||||
Effect of baseline voxel T1: voxels from pure CGM lesions | ||||||
Quartile 1 [lowest] | 231 | 1.81 (4.58) | 261 | −0.86 (4.69) | 1.08 (−0.74, 2.90) | 0.258 |
Quartile 2 | 348 | 0.34 (4.05) | 144 | −0.63 (3.74) | 1.44 (−0.26, 3.15) | 0.112 |
Quartile 3 | 259 | −0.08 (3.79) | 233 | −1.21 (3.77) | 1.49 (−0.04, 3.03) | 0.070 |
Quartile 4 [highest] | 201 | −0.11 (3.35) | 291 | −0.66 (4.01) | 1.50 (−0.06, 3.07) | 0.074 |
Interaction test comparing treatment group differences between the quartiles | 0.104 | |||||
Effect of baseline voxel T1: voxels from pure DGM lesions | ||||||
Quartile 1 [lowest] | 45 | −2.36 (2.88) | 95 | −2.07 (1.79) | 2.35 (0.33, 4.38) | 0.052 |
Quartile 2 | 40 | 1.49 (2.78) | 100 | −0.72 (2.38) | 2.04 (0.11, 3.97) | 0.072 |
Quartile 3 | 71 | −0.2 (1.41) | 69 | −0.25 (2.26) | 1.77 (−0.23, 3.78) | 0.121 |
Quartile 4 [highest] | 74 | −1.45 (1.8) | 66 | −2.29 (1.51) | 2.15 (0.17, 4.13) | 0.066 |
Interaction test comparing treatment group differences between the quartiles | 0.711 | |||||
Effect of voxel band number: voxels from pure WM lesions | ||||||
Band 1 [nearest ventricles] | 184 | −0.15 (1.69) | 242 | −0.03 (1.72) | −0.15 (−0.73, 0.43) | 0.612 |
Band 2 | 451 | 0.13 (2.25) | 596 | 0.24 (1.82) | −0.10 (−0.67, 0.47) | 0.727 |
Band 3 | 732 | −0.32 (2.35) | 760 | −0.07 (1.84) | −0.05 (−0.61, 0.51) | 0.858 |
Band 4 | 1,067 | −0.44 (2.33) | 930 | −0.29 (1.73) | 0.05 (−0.50, 0.60) | 0.857 |
Band 5 | 1,729 | −0.30 (2.33) | 1,051 | −0.18 (1.76) | 0.24 (−0.31, 0.79) | 0.389 |
Band 6 | 1,796 | −0.05 (2.45) | 1,101 | 0.04 (1.90) | 0.17 (−0.38, 0.72) | 0.552 |
Band 7 | 1,649 | 0.03 (2.44) | 1,087 | 0.09 (1.98) | 0.18 (−0.36, 0.73) | 0.515 |
Band 8 | 1,393 | 0.17 (2.86) | 838 | 0.25 (2.16) | 0.09 (−0.46, 0.64) | 0.743 |
Interaction test comparing treatment group differences between the bands | 0.004 | |||||
Effect of voxel band number: voxels from pure CGM lesions | ||||||
Band 1 (inner) | 512 | 1.02 (4.04) | 434 | −0.87 (4.05) | 2.97 (1.25, 4.70) | 0.003 |
Band 2 (outer) | 528 | −0.04 (4.00) | 495 | −0.83 (4.18) | 1.84 (0.36, 3.31) | 0.024 |
Interaction test comparing treatment group differences between the bands | 0.0003 |
p values and 95% CIs are for the differences between bexarotene and placebo after adjustment (for the baseline value of that measure and the four binary minimisation factors). MTR, magnetisation transfer ratio; NA, not applicable.