Improved Detection of Transosseous Meningiomas Using 68Ga-DOTATATE PET/CT Compared with Contrast-Enhanced MRI

Wolfgang G. Kunz; Lisa M. Jungblut; Philipp M. Kazmierczak; Franziska J. Vettermann; Andreas Bollenbacher; Jörg C. Tonn; Christian Schichor; Axel Rominger; Nathalie L. Albert; Peter Bartenstein; Maximilian F. Reiser; Clemens C. Cyran

doi:10.2967/jnumed.117.191932

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
FIGURE 1.
Example of meningioma with transosseous extension. A 56-y-old woman presented with worsening headache. CE-MRI was performed and revealed large, homogeneously enhancing mass in left temporopolar region, consistent with meningioma (lower row). Before surgical resection, patient was examined using ⁶⁸Ga-DOTATATE PET/CT, which additionally demonstrated transosseous extension as assessed by strong tracer uptake extending into sphenoid wing. This was not evident on nonenhanced T2- or T1-weighted images or on T1-weighted CE-MR images. Extension of meningioma into sphenoid wing was found during surgical resection and later confirmed by pathologic evaluation of transosseous secretory meningioma (World Health Organization grade I). w = weighted.
Download figure
Open in new tab
Download powerpoint
FIGURE 2.
Comparison of estimated volumes in ⁶⁸Ga-DOTATATE PET/CT and CE-MRI. Correlation scatterplots depict estimated volumes using ⁶⁸Ga-DOTATATE PET/CT (x-axis) against estimated volumes using CE-MRI (y-axis). Spearman correlation coefficients (r) were calculated. Gray line represents perfect positive correlation. Red line is calculated using least-squares fit of data. Data are based on 82 meningioma patients with pathologic exclusion or inclusion of osseous involvement. In cases of imaging-based false-negative osseous involvement, intraosseous part was attributed no volume (0 mL). Wilcoxon signed rank test demonstrated significant differences between PET and CE-MRI for volume estimation of total and intraosseous part of transosseous meningiomas (each with P < 0.001).
Download figure
Open in new tab
Download powerpoint
FIGURE 3.
Comparison of estimated volumes stratified for meningioma location. Plots depict estimated volumes for total meningioma, intraosseous part, and extraosseous part using ⁶⁸Ga-DOTATATE PET/CT (open circles) and CE-MRI (solid circles). Data are based on 105 meningioma patients, among whom 90 had imaging-based diagnosis of transosseous extension. Statistical comparison was performed using Wilcoxon signed rank test. Bold P values indicate statistical significance. CPA = cerebellopontine angle.
Download figure
Open in new tab
Download powerpoint
FIGURE 4.
Example of estimated intraosseous volume discrepancy between ⁶⁸Ga-DOTATATE PET/CT and CE-MRI in patient with skull base meningioma. A 75-y-old woman was followed up with ⁶⁸Ga-DOTATATE PET/CT and CE-MRI after partial resection of transosseous transitional meningioma (World Health Organization grade I) in left cerebellopontine angle. Because of worsening and radiating pain in occipital region and hypoglossal nerve palsy, she was evaluated for stereotactic radiosurgery. Estimated volumes of transosseous extension were significantly larger on ⁶⁸Ga-DOTATATE PET/CT than on CE-MRI (28.0 vs. 7.5 mL), whereas volumes of extraosseous part agreed well (25.3 vs. 22.6 mL). w = weighted.

View popup

TABLE 1

Characteristics of Extraosseous and Transosseous Meningiomas

Parameter	Variable	Extraosseous (n = 15)	Transosseous (n = 67)	P
Age (y)		55 (51–66)	49 (47–61)	0.073
Male sex		6 (40%)	13 (19%)	0.087
Pathology
WHO grade	I	9 (60%)	56 (84%)	NA
	II	6 (40%)	9 (13%)
	III	0 (0%)	2 (3%)
Histologic subtype	Transitional	5 (38%)	22 (38%)	NA
	Meningothelial	2 (15%)	15 (26%)
	Microcystic	0 (0%)	4 (7%)
	Fibroblastic	0 (0%)	1 (2%)
	Secretory	0 (0%)	5 (9%)
	Atypical	6 (46%)	9 (16%)
	Anaplastic	0 (0%)	2 (3%)
Qualitative imaging
Meningiomatosis		9 (60%)	26 (39%)	0.134
Location	Parasagittal/parafalcine	5 (33%)	10 (15%)	NA
	Convexity	6 (40%)	12 (18%)
	Sphenoid wing	3 (20%)	42 (63%)
	Cerebellopontine	0 (0%)	2 (3%)
	Suprasellar/parasellar	1 (7%)	1 (1%)
PET/CT for osseous involvement	No signs of	13 (87%)	0 (0%)	NA
	Suggestive of	0 (0%)	1 (1%)
	Consistent with	2 (13%)	66 (99%)
CE-MRI for osseous involvement	No signs of	9 (60%)	12 (18%)	NA
	Suggestive of	5 (33%)	19 (28%)
	Consistent with	1 (7%)	36 (54%)
Peritumoral edema	None	8 (53%)	33 (49%)	0.858
	Mild	3 (20%)	18 (27%)
	Extensive	4 (27%)	16 (24%)
Quantitative imaging
Meningioma volume in PET/CT (mL)		3.3 (1.6–4.7)	12.8 (6.8–32.9)	<0.001^*
Meningioma volume in CE-MRI (mL)		2.5 (1.4–4.7)	10.6 (4.6–22.0)	0.001^*
Meningioma SUV_max		7.6 (4.3–13.9)	14.2 (10.0–22.4)	0.011^*
Meningioma SUV_mean		2.7 (1.9–3.0)	4.3 (3.1–5.9)	0.001^*
Pituitary SUV_max		15.8 (12.9–20.0)	15.7 (11.6–19.3)	0.852
Pituitary SUV_mean		3.7 (3.5–4.2)	4.0 (3.3–4.7)	0.587

↵* Statistically significant.
WHO = World Health Organization; NA = not applicable.
Data are count followed by percentage for categoric variables and median followed by interquartile range for continuous variables. Proportion analysis was tested using χ² test. Nonparametric testing for continuous variables was performed using Mann–Whitney U test.

View popup

TABLE 2

Diagnostic Performance of ⁶⁸Ga-DOTATATE PET/CT and CE-MRI for Detection of Osseous Involvement in Intracranial Meningiomas

Method	Sensitivity	Specificity	Positive LR	Negative LR	Prevalence
Overall (n = 82)
PET/CT	98.5 (92.0–99.9)	86.7 (60.0–98.3)	7.39 (2.03–26.9)	0.02 (0.00–0.12)	81.7 (71.6–89.4)
CE-MRI	53.7 (41.1–66.0)	93.3 (68.1–99.8)	8.06 (1.20–54.2)	0.50 (0.37–0.66)	81.7 (71.6–89.4)
Preoperative (n = 39)
PET/CT	100 (88.4–100)	77.8 (40.0–97.2)	4.50 (1.33–15.3)	0.00 (NA)	76.9 (60.7–88.9)
CE-MRI	53.3 (34.3–71.7)	100 (66.4–100)	∞ (NA)	0.47 (0.32–0.68)	76.9 (60.7–88.9)
Postoperative (n = 43)
PET/CT	97.3 (85.8–99.9)	100 (54.1–100)	∞ (NA)	0.03 (0.00–0.19)	86.0 (72.1–94.7)
CE-MRI	54.1 (36.9–70.5)	83.3 (35.9–99.6)	3.24 (0.53–19.9)	0.55 (0.33–0.91)	86.0 (72.1–94.7)

LR = likelihood ratio; NA = not applicable.
Data are percentages (or ratios) followed by 95% confidence interval. Imaging diagnosis of “consistent with” was set as test-positive. McNemar test showed significant differences in diagnostic performance of PET/CT and CE-MRI (P < 0.001).

View popup

TABLE 3

Receiver-Operating-Characteristic Analysis of Imaging Parameters for Osseous Involvement

Parameter	Area under curve	P
PET/CT qualitative assessment	0.932 (0.830–1.000)	<0.001^*
PET/CT meningioma volume	0.803 (0.681–0.925)	<0.001^*
PET/CT meningioma SUV_mean	0.778 (0.640–0.916)	0.001^*
CE-MRI qualitative assessment	0.773 (0.637–0.909)	0.001^*
CE-MRI meningioma volume	0.768 (0.641–0.895)	0.001^*
PET/CT meningioma SUV_max	0.710 (0.557–0.864)	0.011^*
PET/CT peritumoral edema	0.508 (0.342–0.675)	0.919
CE-MRI peritumoral edema	0.477 (0.307–0.646)	0.778

↵* Statistically significant.
95% confidence intervals are in parentheses.

View popup

TABLE 4

PET/CT Measurements in Transosseous Meningiomas

Parameter	Overall (n = 67)	Parafalcine/parasagittal (n = 10)	Convexity (n = 12)	Sphenoid wing (n = 42)	Cerebellopontine (n = 2)	Suprasellar/parasellar (n = 1)
SUV_max total	14.2 (10–22)	10.8 (7–12)	15.6 (10–31)	14.8 (11–26)	7.3 (6–9)	12.6 (13–13)
SUV_max intraosseous	13.4 (10–22)	10.0 (7–11)	15.6 (10–31)	14.2 (11–26)	7.3 (6–9)	12.6 (13–13)
SUV_mean total	4.3 (3–6)	3.7 (3–4)	4.7 (3–7)	4.5 (4–6)	2.6 (3–3)	4.5 (4–4)
SUV_mean intraosseous	4.6 (3–7)	3.4 (3–4)	5.7 (3–8)	5.0 (4–8)	2.7 (3–3)	4.8 (5–5)
Volume total (mL)	12.8 (7–33)	6.4 (4–15)	13.3 (5–18)	17.7 (10–47)	5.6 (3–9)	9.9 (10–10)
Volume intraosseous (mL)	6.8 (3–15)	2.7 (1–5)	5.1 (2–11)	9.3 (6–21)	4.1 (2–7)	6.8 (7–7)

Data are median followed by interquartile range. SUV_mean in total meningioma compared with intraosseous part of meningioma showed statistically significant differences on Wilcoxon signed rank testing (P < 0.001).