Table 2.

Efficacy of median recurrent cutoff and previously published newly diagnosed cutoffs (Neal et al.²⁴) for discriminating OS and PFS across all treatment groups

		Treatment group	Median cutoffs	P value	Neal cutoffs	P value
DG T1Gd	OS	All treatments	106.7	.00114	78	9.25E−05
		BevAlone	134.4	.18	78	.21
		BevCyto	95.9	.00036	78	9.25E−05
	PFS	All treatments	106.7	.00063	93	.00011
		BevAlone	134.4	.059	93	.078
		BevCyto	95.9	.00085	93	.00085
DG FLAIR	OS	All treatments	74.4	.80	78	.54
		BevAlone	121.3	.74	78	.80
		BevCyto	57.8	.55	78	.33
	PFS	All treatments	74.4	.51	93	.25
		BevAlone	121.3	.91	93	.62
		BevCyto	57.8	.10	93	.08

Analysis was performed for both T1Gd- and FLAIR-based DG scores. Significant log-rank test P values underlined.