Table 4:

Early MRI predictors (0–6 and 0–12 mo) of time to confirmed disability progression status using Cox regression analysis and Kaplan-Meier survival analysis^{^a}

	Cox Regression Analysis		AUC Analysis		Kaplan-Meier Survival Analysis at 80% Specificity
	HR	P Value	AUC	P Value	Cutoff	Sensitivity	P Value
New/enlarging T2 lesions 0–6 mo	1.01	.726	0.56	.293	3.5	0.25	.582
New/enlarging T2 lesions 0–12 mo	1.01	.435	0.59	.072	5.3	0.33	.171
T2-LV absolute change 0–6 mo	1.02	.711	0.54	.501	1.16	0.27	.415
T2-LV absolute change 0–12 mo	1.03	.567	0.57	.190	1.41	0.32	.188
Atrophied T2-LV 0–6 mo	4.23	.04^{^b}	0.61	.022^{^b}	0.09	0.40	.017^{^b}
Atrophied T2-LV 0–12 mo	2.41	.022^{^b}	0.61	.022^{^b}	0.18	0.36	.076
PBVC 0–6 mo	0.89	.431	0.54	.366	−0.83	0.29	.128
PBVC 0–12 mo	0.85	.105	0.55	.379	−1.38	0.21	.704
PVVC 0–6 mo	1.01	.419	0.52	.707	13.46	0.22	.482
PVVC 0–12 mo	1.07	.375	0.53	.451	16.29	0.26	.487
PCVC 0–6 mo	0.92	.123	0.55	.272	−2.28	0.31	.152
PCVC 0–12 mo	0.92	.115	0.62	.011^{^b}	−2.47	0.39	.091

Note:—HR indicates hazard ratio; AUC, area under the curve.
↵a Cox regression and Kaplan-Meier analyses were used to analyze the association of early MRI outcome changes (0–6 and 0–12 mo) and time to development of CDP. The Benjamini-Hochberg correction was used to minimize the false discovery rate, and P values < .05 were considered significant.
↵b Significant P value <.05.