Evaluation of tumor response to antiangiogenic therapy in patients with recurrent gliomas using contrast-enhanced perfusion-weighted magnetic resonance imaging techniques: A meta-analysis

Table 1 Quality assessment of the included studies

Study	Risk of bias				Applicability concerns
	Patient selection	Index test	Reference standard	Flow and timing	Patient selection		Index test	Reference standard
Iwamoto et al[29], 2010	L	U	H	L	L	L		L
Piludu et al[30], 2015	L	U	U	L	L	L		L
Kalpathy-Cramer et al[31], 2017	L	U	U	L	L	L		L
O’Neill et al[32], 2016	L	U	H	L	L	L		L
Schmainda et al[4], 2014	L	U	U	L	L	L		L
Hilario et al[33], 2016	L	U	U	L	L	L		L

L: Low risk; U: Unclear risk; H: High risk.

Table 2 The basic characteristics of the included studies

Studies	Country	Patients, n	Median age	Study design	Imaging modality	Imaging biomarkers	MR scanner tesla	Anti-angiogenic treatment	Clinical question	Standard reference	TP	FP	FN	TN
Iwamoto et al[29], 2010	United States	11	53 (29-73)	Pro-spective	DSC, DCE	Ktrans, rCBV	NR	Pazopanib	DSC and DCE vs response	Macdonald criteria RANO criteria	1	0	6	4
Piludu et al[30], 2015	Germany	27	54 (33-77)	Pro-spective	DCE	nIAUGC, Ktrans	3.0 T	Bevacizumab	DCE vs response	RANO criteria	6	0	5	16
Kalpathy-Cramer et al[31], 2017	United States	10	62 (51-74)	Pro-spective	DSC, DCE	Ktrans, rCBV, rCBF	3.0 T	Tivozanib	DSC and DCE vs response	RANO criteria	1	1	4	4
O’Neill et al[32], 2016	United States	12	NR	Pro-spective	DCE	Ktrans, Ve	1.5 T	VEGF Trap	DCE vs response	Macdonald criteria	1	0	0	11
Schmainda et al[4], 2014	United States	36	34 (30-68)	Retro-spective	DSC	rCBV, stdrCBV	1.5 T or 3.0 T	Bevacizumab	DSC vs response	Macdonald criteria RANO criteria	6	0	0	30
Hilario et al[33], 2016	Spain	24	52.5 (31-74)	Retro-spective	DSC	Leakage volume (CBV-LCCBV)	1.5 T	Bevacizumab	DSC vs response	RANO criteria	14	0	0	10

TP: True positive; FP: False positive; TN: True negative; FN: False negative; DCE: dynamic contrast-enhanced; DSC: Dynamic susceptibility.

Table 3 Clinical responses of gliomas to antiangiogenic treatments

Study	Patient, n	RANO criteria				Macdonald criteria
		Responders		Non-responders		Responders		Non-responders
		CR	PR	SD	PD	CR	PR	SD	PD
O’Neill et al[32], 2016	12	-	-	-	-	0	1	0	11
Piludu et al[30], 2015	27	0	6	5	16	-	-	-	-
Schmainda et al[4], 2014	36	0	6	0	30	-	-	-	-
Iwamoto et al[29], 2010	11	-	-	-	-	0	1	6	4
Hilario et al[33], 2016	24	0	14	0	10	-	-	-	-
Kalpathy-Cramer et al[31], 2017	10	1	1	4	4	-	-	-	-

CR: Complete response; PR: Partial response; SD: Stable disease; PD: Progressive disease. *Response as defined using the Macdonald criteria and Response Assessment in Neuro-Oncology (RANO) working group guidelines.

Table 4 True positive, false positive, true negative, and false negative as well as the sensitivity, specificity, likelihood ratio, and diagnostic odds ratio for the meta-analysis of antiangiogenic treatment evaluation

Study	TP	FP	FN	TN	Sensitivity (95%CI)	Specificity (95%CI)	LR+ (95%CI)	LR- (95%CI)	DOR (95%CI)
Iwamoto et al[29], 2010	1	0	4	6	0.2 (0.01-0.72)	1 (0.54-1)	3.5 (0.17-70.94)	0.81 (0.49-1.340)	4.33 (0.14-132.32)
Piludu et al[30], 2015	6	0	5	16	0.55 (0.23-0.83)	1 (0.79-1)	18.42 (1.14-296.83)	0.47 (0.25-0.880	39 (1.88-810.44)
Kalpathy-Cramer et al[31], 2017	1	1	4	4	0.2 (0.01-0.72)	0.8 (0.28-0.99)	1 (0.08-11.93)	1 (0.54-1.86)	1 (0.05-22.18)
O’Neill et al[32], 2016	1	0	0	11	1 (0.03-1)	1 (0.72-1)	18 (1.06-304.71)	0.26 (0.02-2.89)	69 (0.96-4951.23)
Schmainda et al[4], 2014	6	0	0	30	1 (0.54-1)	1 (0.88-1)	57.57 (3.65-906.93)	0.07 (0.01-1.05)	793 (14.37-43,746.61)
Hilario et al[33], 2016	14	0	0	10	1 (0.77-1)	1 (0.69-1)	21.27 (1.42-319.53)	0.03 (0-0.53)	609 (911.16-33,236.62)
Pooled results	29	1	13	77	0.69 (0.53-0.82)	0.99 (0.93-1)	12.84 (4.54-36.28)	0.35 (0.22-0.53)	24.44 (7.19-83.06)
Heterogeneity test					I2 = 81.30%, P = 0.0001	I2 = 12.2%, P = 0.3369	I2 = 19.1%, P = 0.2892	I2 = 81.5%, P = 0.0001	I2 = 53.5%, P = 0.0563

TP: True positive; FP: False positive; TN: True negative; FN: False negative; LR+: Positive likelihood ratio; LR-: Negative likelihood ratio; DOR: Diagnostic odds ratio; CI: Confidence interval.