Table 2. Multivariable Cox Proportional Hazards Regression Models to Test Associations Between sNfL Levels and Risk of 6-Month CDW and EDSS of 3 Using the Cutoff Values of 10 pg/mL and a z Score of 1.5^a.

Variables	HR (95% CI)
	6-mo CDW			EDSS of 3
	Development cohort (n = 327)	Validation cohort (n = 251)	Total (N = 578)	Development cohort (n = 327)	Validation cohort (n = 251)	Total (N = 578)
sNfL cutoff of 10 pg/mL
sNfL>10 pg/mL	2.39 (1.39-4.12)b	1.61 (1.07-2.42)c	1.88 (1.37-2.60)d	4.12 (2.18-7.77)d	2.03 (1.23-3.33)b	2.48 (1.69-3.64)d
Age at first relapse	1.01 (0.98-1.03)	1.01 (0.99-1.03)	1.01 (0.99-1.02)	1.01 (0.98-1.04)	1.01 (0.98-1.03)	1.01 (0.99-1.03)
Male sex	2.02 (1.22-3.32)b	1.21 (0.78-1.86)	1.50 (1.10-2.07)c	1.58 (0.88-2.84)	1.47 (0.87-2.47)	1.38 (0.95-2.02)
Baseline EDSS score	1.44 (1.18-1.75)d	1.18 (0.91-1.53)	1.20 (1.03-1.39)c	1.66 (1.37-2.05)d	2.08 (1.55-2.79)d	1.55 (1.35-1.78)d
T2 lesion load	1.42 (1.07-1.88)c	1.09 (0.82-1.44)	1.24 (1.03-1.49)c	1.35 (1.01-1.16)c	1.05 (0.74-1.49)	1.27 (1.04-1.57)c
Time from first relapse to sample obtainment, mo	1.06 (0.99-1.13)	1.00 (0.95-1.06)	1.01 (0.98-1.05)	1.08 (1.01-1.16)c	1.02 (0.96-1.09)	1.04 (0.99-1.09)
Proportion of time of DMT usee	0.38 (0.17-0.78)b	0.63 (0.37-1.08)	0.58 (0.38-0.87)b	0.25 (0.11-0.56)b	0.31 (0.16-0.59)d	0.35 (0.22-0.58)d
Proportion of time of HE-DMT usef	0.26 (0.09-0.77)c	0.19 (0.04-0.89)c	0.26 (0.11-0.61)b	0.05 (0.01-0.21)d	0.26 (0.06-1.09)	0.11 (0.04-0.30)d
AIC	671	1020	1938	543	682	1415
C statistic	0.72	0.63	0.65	0.82	0.72	0.75
z Score cutoff of 1.5
z Score >1.5	2.06 (1.21-3.52)b	1.56 (1.03-2.38)c	1.80 (1.30-2.50)d	2.99 (1.65-5.40)d	1.94 (1.16-3.23)c	2.13 (1.46-3.11)d
Age at first relapse	1.01 (0.99-1.04)	1.01 (0.99-1.03)	1.01 (0.99-1.03)	1.02 (0.99-1.05)	1.01 (0.99-1.04)	1.02 (1.00-1.04)c
Male sex	1.90 (1.15-3.13)c	1.25 (0.81-1.93)	1.49 (1.08-2.05)c	1.43 (0.80-2.58)	1.59 (0.94-2.69)	1.36 (0.93-1.98)
Baseline EDSS score	1.45 (1.19-1.77)d	1.20 (0.93-1.55)	1.20 (1.03-1.39)c	1.67 (1.36-2.04)d	2.11 (1.57-2.84)d	1.55 (1.34-1.78)d
T2 lesion load	1.46 (1.11-1.94)b	1.09 (0.82-1.45)	1.26 (1.05-1.52)c	1.39 (1.04-1.87)c	1.07 (0.75-1.51)	1.31 (1.07-1.61)c
Time from first relapse to sample obtainment, mo	1.06 (0.99-1.13)	1.00 (0.95-1.06)	1.02 (0.98-1.06)	1.08 (1.01-1.16)c	1.02 (0.96-1.09)	1.04 (0.99-1.08)
Proportion of time of DMT usee	0.39 (0.19-0.79)b	0.61 (0.35-1.04)	0.56 (0.37-0.85)b	0.26 (0.12-0.59)b	0.28 (0.14-0.54)d	0.34 (0.21-0.56)d
Proportion of time of HE-DMT usef	0.27 (0.09-0.79)c	0.19 (0.04-0.91)c	0.26 (0.11-0.61)b	0.06 (0.01-0.25)d	0.27 (0.07-1.12)	0.12 (0.04-0.32)d
AIC	674	1021	1941	552	683	1422
C statistic	0.72	0.63	0.65	0.82	0.72	0.75

Abbreviations: AIC, Akaike information criterion; CDW, confirmed disability worsening; DMT, disease-modifying treatment; EDSS, Expanded Disability Status Scale (scores range from 0 to 10 in 0.5-unit increments, with higher scores representing higher levels of disability); HE-DMT, high-efficacy DMT; HR, hazard ratio; sNfL, serum neurofilament light chain.

^a Multivariable models were adjusted by age at first relapse, sex, T2 lesion load, time from first relapse to blood sample obtainment, proportion of time that the patient received DMT, excluding HE-DMT (obtained by dividing the time of DMT use by the time of disease evolution until the outcome of interest), and proportion of time of HE-DMT use.

^b P < .01.

^c P < .05.

^d P < .001.

^e DMTs, excluding HE-DMTs, included subcutaneous or intramuscular interferon-beta, glatiramer acetate, teriflunomide, dimethyl fumarate, fingolimod, oral cladribine, daclizumab, azathioprine, and tacrolimus.

^f HE-DMTs included natalizumab, alemtuzumab, ocrelizumab, rituximab, ofatumumab, and mitoxantrone.