Article Figures & Data
Figures
- FIG 1.
Flow chart of patients included and excluded in the study. Among the 26 patients included in the study, 6 patients had 2 lesions, 2 patients had 1 lesion with tumor recurrence (TR) and 1 lesion with radiation necrosis (RN), both confirmed with clinical/radiologic follow-up; 1 patient with 2 lesions with RN confirmed with surgery; 1 patient had 1 lesion with TR and 1 lesion with RN confirmed with surgery; 1 patient with 1 lesion with TR confirmed by surgery and 1 lesion with RN confirmed with clinical/radiology follow-up; 1 patient had 1 lesion with TR confirmed with surgery and 1 lesion with TR confirmed with clinical/radiologic follow-up.
- FIG 2.
Sample of a case of recurrent glioblastoma multiforme with a false-negative ASL study due to probable ineffective labeling. Patient had a history of a right frontal grade IV glioma and had undergone resection and chemoradiation 3 years earlier. A, Axial T1-weighted image shows a newly enhancing lesion in the right anterior thalamus. B, ASL-derived CBF reveals a low CBF value in the enhancing lesion (arrow). The CBF in the right hemispheric cortex is lower than on the left side, presumably due to ineffective labeling from dental hardware. DSC MR imaging demonstrates a high, uncorrected rCBV value (C) and leakage-corrected rCBV value (D) at the rim of the enhancing lesion (arrow). DCE MR imaging using MOLLI T1 mapping shows high Vp (E) and Ktrans (F) values in the lesion (arrows). Pathology confirmed glioblastoma recurrence.
- FIG 3.
Example of a case of recurrent glioblastoma multiforme with a false-negative DSC study due to susceptibility artifacts. A, Axial T1-weighted image shows a newly enhancing lesion in the right mesial temporal lobe. B, ASL-derived CBF map demonstrates a marked increase in CBF in the lesion (arrow). C, Corrected CBV map is unreliable in the tumoral region due to susceptibility artifacts (arrow). DCE MR imaging performed without T1 mapping reveals high Vp (D) and Ktrans (E) values in the lesion (arrow), which correlate with the high CBF value.
Tables
Radiation Necrosis Tumor Recurrence P Value Median age (yr) (95% CI) 54 (44–63.2) 59 (50.9–65.3) .34 Female/male 13% (1/12) 21% (4/15) .35 Initial tumor grade: grade 3, grade 4 0, 13 4, 15 .098 Median duration from radiation treatment to study MR imaging (days), (95% CI) 179 (142–462) 275 (191–915) .066 % Lesions with reresection 30% (5/13) 68% (13/19) .27 Median duration from imaging to reresection (days) (95% CI) 20 15 (13.5–26) .96 Median Karfnosky score (range) 80 (75.3–90) 80 (70–90) .43 ↵a Patients with 2 lesions are entered as 2 separate entries for the purpose of this table.
- Table 2:
Comparison of perfusion parameters for lesions diagnosed as tumor recurrence versus radiation necrosis
Radiation Necrosis Tumor Recurrence No. Median 95% CI No. Median 95% CI P Value CBF_ASL (mL/100 g/min) 13 24.1 17.8–38.3 19 36.0 33.9–62.2 .003a rCBV 13 1.81 1.55–3.18 19 3.69 2.02–5.03 .058 Corrected rCBV 13 1.24 0.83–2.89 19 2.50 1.60–3.15 .048 Ktrans_SI (min−1) 13 0.013 0.0062–0.039 19 0.019 0.011–0.048 .28 Vp_SI 13 0.16 0.12–0.29 19 0.27 0.21–0.39 .063 Ktrans_MOLLI (min−1) 11 0.052 0.024–0.094 14 0.055 0.047–0.11 .44 Vp_MOLLI 11 0.17 0.13–0.28 14 0.28 0.20–0.41 .012 Ktrans_SMART (min−1) 11 0.042 0.020–0.075 14 0.046 0.037–0.071 .48 Vp_SMART 11 0.12 0.082–0.23 14 0.21 0.15–0.33 .080 ↵a Significance following Holm Bonferroni correction.
Table 3:Correlogram between different perfusion parameters using the Spearman rank correlation coefficienta
↵a Color coding is red for a very strong correlation (rs >0.8); orange for a strong correlation (0.6 ≤ rs ≤ 0.79), yellow for a moderate correlation (0.4 ≤ rs ≤ 0.59), and green for a weak correlation (rs <0.4).
- Table 4:
Diagnostic accuracy of various parameters in the differentiation between tumor recurrence and radiation necrosis for all lesions (n = 32 lesions) and for subgroup analysis of patients with DCE T1 mapping (n = 25 lesions)a
Parameters No. AUC (95% CI) P Value Optimal Threshold Sensitivity (%) Specificity (%) ASL_CBF 32 0.81 (0.63–0.93) <.001 >30 mL/100 g/min 84 (60–97) 77 (46–95) rCBV 32 0.70 (0.52–0.89) .033 >2.43 69 (43–87) 77 (46–95) Corrected rCBV 32 0.71 (0.51–0.90) .037 >1.54 79 (54–94) 69 (39–91) Ktrans_SI 32 0.61 (0.41–0.82) .27 >0.0093 (min-1) 84 (60–97) 46 (19–75) Vp_SI 32 0.69 (0.51–0.88) .038 >0.18 79 (54–94) 62 (32–86) Ktrans_SI 25 0.66 (0.43–0.88) .17 >0.0092 (min-1) 86 (57–98) 55 (23–83) Vp_SI 25 0.68 (0.46–0.90) .10 >0.18 79 (49–95) 64 (31–89) Ktrans_MOLLI 25 0.59 (0.36-0.82) .44 >0.045 (min-1) 78 (49–95) 45 (17–77) Vp_MOLLI 25 0.80 (0.62–0.98) <.001 >0.19 86 (57–98) 63 (31–89) Ktrans_SMART 25 0.58 (0.34–0.83) .49 >0.024 (min-1) 92 (66–1) 27 (6–61) Vp_SMART 25 0.71 (0.49–0.88) .057 >0.12 93 (66-1) 54 (23–83) ↵a Statistics are listed with their 95% CI.
