Table 2.

Summary of the main studies supporting the beneficial effects of early use of HE-DMTs in MS patients

Study	Study design	Comparisons of treatment strategy(ies)	MS patients	Main findings
Harding et al. [9]	Retrospective single-center data from the southeast Wales	EIT vs ESC ESC: IFN-β, GA, DMF, FTY, TERI EIT: NAT, ALEM	EIT: n = 104 ESC = n = 488	5-year change in EDSS score was lower in the EIT vs ESC (mean [SD] = 0.3 [1.5] vs 1.2 [1.5], p < 0.001), which remained significant after adjustment for relevant covariates (β = − 0.85; 95% CI − 1.38; − 0.32, p = 0.002) Median time to sustained accumulation of disability was 6.0 (95% CI 3.17; 9.16) years for EIT and 3.14 (2.77; 4.00) years for ESC (p = 0.05)
Brown et al. [10]	Retrospective data from the MSBase registry	HE-DMT vs ME-DMT Early vs late escalation from ME-DMT to HE-DMT (≤ 5 vs > 5 years) ME-DMT: IFN-β, GA HE-DMT: FTY, NAT, ALEM	ME-DMT: n = 407 HE-DMT: n = 211	Initial treatment with HE-DMT vs ME-DMT associated with a significant lower risk of SPMS conversion ( HR 0.66, 95% CI 0.44; 0.99, p = 0.046); 5-year absolute risk: 7% vs 12% Escalation from ME-DMT to HE-DMT within 5 years vs later associated with a significant lower risk of SPMS conversion ( HR 0.76, 95% CI 0.66; 0.88, p < 0.001; 5-year absolute risk: 8% vs 14%
He et al. [11]	Retrospective data from the MSBase registry and the Swedish MS registry with ≥ 6 years of follow-up	Early vs late HE-DMT start after clinical onset (0–2 vs 4–6 years) HE-DMT: RTX, OCRE, MTX, ALEM, NAT	Early HE-DMT: n = 213 Late HE-DMT: n = 253	Early vs late HE-DMT start was associated with a significantly lower EDSS score after 6 years (mean EDSS [standard deviation] = 2.2 [1.6] vs 2.9 [1.8], p < 0.001), which persisted throughout each year of follow-up until the 10th year after disease onset with a difference between groups of − 0.98 (95% CI − 1.51; − 0.45; p < 0.0001) across the 6–10 year follow-up period
Buron et al. [12]	Retrospective data from the Danish registry	First treatment: HE-DMT vs ME-DMT ME-DMT: IFN-β, GA, TERI, DMF HE-DMT: FTY, NAT, CLAD, DAC, ALEM, OCRE	ME-DMT: n = 194 HE-DMT: n = 194	At 4-year follow-up, HE-DMT vs ME-DMT associated with a significantly lower probability of a 6-month confirmed EDSS score worsening (16.7% [95% CI 10.4%; 23.0%] vs 30.1% [95% CI 23.1%; 37.1%]; HR 0.53 [95% CI 0.33; 0.83], p = 0.006)
Spelman et al. [13]	Retrospective data from the Danish and Swedish national MS registries	Swedish vs Danish cohorts ME-DMT: IFN-β, GA, TERI, DMF HE-DMT: FTY, NAT, RTX, ALEM, OCRE	Danish cohort: n = 1994 (92.3%) ME-DMT, n = 165 (7.7%) HE-DMT Swedish cohort: n = 1769 (65.5%) ME-DMT, n = 931 (34.5%) HE-DMT	The Swedish vs Danish treatment strategy associated with  29% in the rate of 24-week confirmed disability worsening ( HR 0.71 [95% CI 0.57; 0.90], p = 0.004)  24% in the rate of reaching an EDSS score ≥ 3.0 ( HR 0.76 [95% CI 0.60; 0.90], p = 0.03)  25% in the rate of reaching an EDSS score ≥ 4.0 ( HR 0.75 [95% CI 0.61; 0.96], p = 0.01)
Uher et al. [14]	Retrospective analyses from Avonex-Steroids-Azathioprine (n = 166), Study of Early IFN-β1a Treatment (n = 180), and in a cohort from the quantitative MRI project (n = 1557)	Escalation from ME-DMT to HE-DMT ME-DMT: IFN-β, GA, TERI, DMF HE-DMT: FTY, NAT, MTX	n = 94 (609 MRI scan)	BVL rates substantially decreased following treatment escalation (before: mean = 0.45, 95% CI − 0.54; − 0.37 vs after: mean = − 0.10, 95% CI − 0.13; − 0.07). Such differences were confirmed in adjusted mixed models, where treatment escalation resulted in a mean reduction of the BVL rate by 0.29% (β = − 0.29, 95% CI − 0.40; − 0.19, p < 0.001) Effects were only measurable two years after escalation to a HE-DMT
Iaffaldano et al. [60]	Retrospective analyses from the Italian MS Registry	EIT vs ESC ESC: IFN-β, GA, DMF, FTY, TERI, AZA EIT: FTY, NAT, MTX, ALEM, OCRE, CLAD	EIT: n = 364 ESC = n = 364	EIT vs ESC showed significantly lower mean annual EDSS changes (p < 0.02), with the differences in mean EDSS changes increasing from 0.10 (95% CI 0.01; 0.19, p = 0.03) at 1 year to 0.30 (95% CI 0.07; 0.53, p = 0.009) at 5 years and to 0.67 (95% CI 0.31; 1.03, p = 0.0003) at 10 years
Hanninen et al. [15]	Retrospective data from the Finnish registry	First treatment: HE-DMT vs ME-DMT ME-DMT: IFN-β, GA, TERI, DMF HE-DMT: NAT, ALEM, OCRE, RTX	ME-DMT: n = 1771 HE-DMT: n = 154	HE-DMT vs ME-DMT associated with a significantly lower probability of a 6-month confirmed EDSS score worsening (28.4% [95% CI 15.7; 39.3] vs 47.0% [95% CI 33.1; 58.1], p = 0.013)

ALEM alemtuzumab, AZA azathioprine, BVL brain volume loss, CI confidence interval, CLAD cladribine, DAC daclizumab, DMF dimethyl fumarate, EDSS Expanded Disability Status Scale, EIT early intensive treatment, ESC escalation, FTY fingolimod, GA glatiramer acetate, HE-DMT high-efficacy disease-modifying therapy, HR hazard ratio, IFN-β interferon beta, ME-DMT moderate-efficacy disease-modifying therapy, MRI magnetic resonance imaging, MTX mitoxantrone, NAT natalizumab, OCRE ocrelizumab, RTX rituximab, TERI teriflunomide