Part 4 · Treatment
The DMT ladder
How MS drugs are sequenced: start on a lower rung and step up after breakthrough disease, or start high from the beginning. See what the 2018 guidelines say, what the observational studies report, and why stopping later in life is still debated.
Published guidance and studies only. This module reports what the ECTRIMS/EAN and AAN guidelines, trials and labels say. The person on the ladder is an illustration, not a case. Which medicine suits a particular person, and when to change it, is a decision for that person and their MS team.
Breakthrough here follows AAN 2018: one or more relapses, two or more new MRI lesions, or more disability on examination, within one year. The two lanes share the same events, so the only difference between them is the rung. Event years are examples, not data. Outcome bars and lines below the lanes are published observational results; lines between published points are schematic, and the studies quoted give no points after year 10. 51,182,190,394
Breakthrough and NEDA-3
What counts as breakthrough disease
Clinicians should discuss switching with people who take their DMT as prescribed, have been on it long enough for it to take full effect, and have one or more relapses, two or more unequivocally new MRI-detected lesions, or increased disability on examination, over a one-year period
. Any one of the three is enough. 394
ECTRIMS/EAN 2018 recommendation 14 (strong) says: Offer a more efficacious drug to patients treated with interferon or glatiramer acetate who show evidence of disease activity.
Recommendation 11 (consensus) sets the yardstick: a reference brain MRI usually within six months of treatment onset
, compared with one typically 12 months after starting treatment
. The 2021 MAGNIMS-CMSC-NAIMS consensus names 3D FLAIR as the core sequence for spotting new lesions. 323,491
NEDA-3: no evident disease activity
- 1 · Relapsesnone
- 2 · MRIno new or enlarging T2 lesions and no gadolinium-enhancing lesions
- 3 · Disabilityno sustained disability progression
The trial version first appeared in a post hoc analysis of AFFIRM: over 2 years, 37% on natalizumab and 7% on placebo had no clinical or radiological activity. In 2015 NEDA was proposed as a treat-to-target goal, with a new baseline MRI once the drug has taken effect and at least yearly MRI after that. In OPTIMUM, 25.0% on ponesimod and 16.4% on teriflunomide met NEDA-3 over 108 weeks. 166,189,236
Escalate, or start high? A live controversy
For starting high
- Harding 2019 (Wales, 592 people): 5-year EDSS change 0.3 with early intensive therapy versus 1.2 with escalation (adjusted β −0.85, 95% CI −1.38 to −0.32). 60% of those who escalated developed sustained disability before escalating. 182
- He 2020 (MSBase and Sweden, 213 versus 253 matched people): starting high-efficacy therapy within 2 years of onset versus at 4–6 years gave mean EDSS 2.3 versus 3.5 at year 10 (difference −0.98, −1.51 to −0.45, over years 6–10). 190
- Brown 2019 (1,555 matched people): initial fingolimod, alemtuzumab or natalizumab carried a lower hazard of conversion to SPMS than initial interferon or glatiramer acetate (HR 0.66, 0.44–0.99; 5-year risk 7% versus 12%). Escalating within 5 years versus later: HR 0.76 (0.66–0.88). 51
- Spelman 2021: 34.5% of Swedish and 7.6% of Danish patients started a highly effective DMT; the Swedish strategy was associated with less 24-week confirmed worsening (HR 0.71, 0.57–0.90). 449
- An Italian Delphi of 65 neurologists (2025) reached strong consensus
to support early HE-DMT initiation to prevent irreversible disability
. 143
Against firm conclusions
- All four studies are observational. People who start high may differ from those who escalate in ways the matching cannot remove (confounding by indication).
- AAN 2018 called for research comparing early higher-potency treatment with stepped care. 394
- Baseline DELIVER-MS data show milder disease among people willing to be randomised, and 67% of the observational arm chose early high-efficacy therapy. 457
- Over half of the trials in the 2024 Cochrane network meta-analysis were industry-sponsored, and its authors found the evidence beyond 2 years insufficient. 171
| Randomised trial | Design | Primary outcome | Status on 20 October 2026 |
|---|---|---|---|
| DELIVER-MS (NCT03535298) | Pragmatic randomised trial with a parallel observational cohort, 816 enrolled; early high-efficacy monoclonal antibodies versus escalation 371,457 | Brain volume loss from baseline to month 36 | Active, not recruiting; primary completion estimated July 2027; no results 344 |
| TREAT-MS (NCT03500328) | 900 treatment-naive people, mean age 36, mean EDSS 2.5; higher-efficacy versus traditional DMT strategies 330 | Sustained EDSS-plus worsening | Completed July 2026; no results posted and no outcome paper found 343 |
The guidelines in their own words
Both main guidelines date from 2018. No 2024–2026 update of either was found; the newest consensus is the 2025 Italian Delphi above. ECTRIMS/EAN grades are strong, weak or consensus; AAN levels are A (must), B (should) and C (may). 143,323,394
| Situation | ECTRIMS/EAN 2018 (21 recommendations) 323 | AAN 2018 (30 recommendations) 394 |
|---|---|---|
| CIS | R2 (strong): Offer interferon or glatiramer acetate to patients with CIS and an abnormal MRI with lesions suggestive of MS who do not fulfil criteria for MS | Starting 7 (B): prescribe DMT after a single demyelinating event with two or more characteristic brain lesions to people who decide they want this therapy. Starting 8 (C): clinicians may instead follow untreated, inactive CIS or relapsing MS with imaging at least yearly for five years |
| Active relapsing MS | R3 (strong): Offer early treatment with DMDs to patients with active relapsing-remitting MS as defined by clinical relapses and/or MRI activity | Starting 9 (B): Clinicians should offer DMTs to people with relapsing forms of MS with recent clinical relapses or MRI activity |
| Highly active MS | No separate list in the wording quoted here | Starting 14 (B): Clinicians should prescribe alemtuzumab, fingolimod, or natalizumab for people with MS with highly active MS |
| Breakthrough | R14 (strong): Offer a more efficacious drug to patients treated with interferon or glatiramer acetate who show evidence of disease activity | Switching 1 (B): discuss switching for one or more relapses, two or more unequivocally new MRI-detected lesions, or increased disability on examination, over a one-year period |
| Progressive MS | All weak: R5 interferon for active SPMS, noting the dubious efficacy; R6 mitoxantrone; R7 ocrelizumab or cladribine for active SPMS; R8 consider ocrelizumab for PPMS | Starting 17 (B): Clinicians should offer ocrelizumab to people with PPMS who are likely to benefit from this therapy unless there are risks of treatment that outweigh the benefits |
| Stopping a strong drug | R16 (consensus): consider starting another highly efficacious drug, weighing half-life and the potential for resumed disease activity or even rebound (particularly with natalizumab) | Switching 9 (A): counsel that relapse or MRI activity risk rises within six monthsof stopping natalizumab |
| Stable for years | R18 (weak): Consider continuing a DMD if a patient is stable (clinically and on MRI) and shows no safety or tolerability issues | Stopping (B): advocate that stable patients continue their current DMT unless the patient and physician decide a trial off therapy is warranted. (C): may advise stopping in SPMS without relapses or enhancing lesions in people not ambulatory (EDSS 7 or more) for at least 2 years |
| Mitoxantrone | R6 (weak), as above | (B): clinicians should not prescribe mitoxantrone to people with MS unless the potential therapeutic benefits greatly outweigh the risks |
The two guidelines differ on ocrelizumab in PPMS: AAN gives a Level B should offer
, ECTRIMS/EAN a weak consider
, for a trial effect of HR 0.76 with an upper confidence limit of 0.98 (ORATORIO). 322,323,394
Later life: de-escalation and stopping
DISCOMS asked whether stable older people could stop their DMT safely. It randomised 259 people aged 55 or older, with no relapse for 5 years and no new MRI lesion for 3 years, to stop or continue. Press Play to watch new disease activity accumulate in each arm over 2 years. 96
Only the 2-year values are published; the month-by-month filling is schematic. Each grid is 100 people. 96
What the guidelines say about stopping
AAN 2018 (Level B) advocates that stable patients continue their current DMT unless the patient and physician decide a trial off therapy is warranted
, and (Level C) allows advising stopping in SPMS without relapses or enhancing lesions in people who have not been ambulatory (EDSS 7 or more) for at least 2 years. ECTRIMS/EAN 2018 R18 (weak) favours continuing a stable, well-tolerated DMD, and R17 (weak) warns of resumed disease activity or even rebound when stopping treatment, particularly with natalizumab
. 323,394
Rebound: why some drugs are not simply stopped
After stopping natalizumab, relapses in RESTORE appeared from 4–8 weeks after the last dose. After stopping fingolimod, the label reports severe increases in disability, usually within 12 weeks. The timing and the numbers are animated in Safety & monitoring. 154,163
Beyond the ladder: aHSCT
Autologous haematopoietic stem cell transplantation is not a rung of either 2018 guideline in this module. It is a specialist treatment whose place is still debated, and its evidence is covered in Toward a cure.