Table 2.
Statistical analysis of the relationship between 18F-fluorodeoxyglucose (FDG) uptake metrics and esophagitis grade at the time of the FDG-positron emission tomography (PET) scan, and maximum treatment esophagitis grade (n = 79). Univariate logistic regression p-values reported have been corrected with the Benjamini-Hochberg false discovery rate procedure.
| FDG uptake metric | ≥Grade 2 during PET scan | ≥Grade 2 Treatment Maximum | ||
|---|---|---|---|---|
|
| ||||
| P Value | AUC* | P Value | AUC* | |
| SUVmean | 2.04E-02 | 0.67 | 2.35E-01 | 0.61 |
| SUVmax | 5.23E-03 | 0.73 | 1.88E-03 | 0.75 |
| nSUVmean† | 8.16E-05 | 0.84 | 4.09E-05 | 0.87 |
| nSUVmax | 1.55E-04 | 0.82 | 1.26E-06 | 0.91 |
| nSUVaxMax1‡ | 8.16E-05 | 0.85 | 1.17E-06 | 0.91 |
| nSUVaxMax3 | 5.32E-05 | 0.85 | 9.53E-06 | 0.87 |
| nSUVaxMax5 | 5.75E-05 | 0.84 | 1.52E-05 | 0.86 |
| nSUVaxMax7 | 8.16E-05 | 0.81 | 3.65E-05 | 0.84 |
| nSUVlen20** | 5.32E-05 | 0.80 | 7.37E-06 | 0.89 |
| nSUVlen30 | 5.32E-05 | 0.84 | 1.17E-06 | 0.91 |
| nSUVlen40 | 8.16E-05 | 0.85 | 6.52E-08 | 0.92 |
| nSUVlen50 | 7.15E-04 | 0.81 | 1.26E-06 | 0.87 |
| nSUVlen60 | 1.32E-02 | 0.74 | 4.23E-05 | 0.80 |
| nSUVperc65†† | 8.31E-05 | 0.83 | 1.12E-05 | 0.89 |
| nSUVperc75 | 8.31E-05 | 0.83 | 2.48E-06 | 0.90 |
| nSUVperc85 | 8.16E-05 | 0.84 | 2.48E-06 | 0.90 |
| nSUVperc95 | 4.69E-04 | 0.83 | 1.26E-06 | 0.91 |
Area under the curve.
Voxel mean normalized standard uptake value.
Axial-averaged maximum nSUV for 1, 3, 5, and 7 slices (e.g., nSUVaxMax1 for the 1-slice average).
Axial length of esophageal tissue with at least a given percentage of axial-averaged nSUV response (e.g., nSUVlen20 for a length of esophagus with nSUV ≥ 20% axial-averaged response over the baseline SUV value).
Voxel percentile nSUV values from the sixty-fifth to the ninety-fifth percentiles (e.g., nSUVperc65 for the sixty-fifth percentile).