Teriflunomide fails in Paediatric MS for primary outcome


Teriflunomide (Aubagio) holds a license for adult RRMS with an efficacy of roughly 30% relapse rate reduction versus placebo (dummy tablet).

The long awaited study in paediatric MS (TERIKIDS study) recently published, however has failed to meet its primary endpoint – time to first confirmed clinical relapse in the teriflunomide group vs. placebo. The probability of first confirmed relapse during the trial period was 0.53 in the placebo group vs. 0.39 in the teriflunomide group; the p-value, however, doesn’t reach statistical significance at p=0.29).

Why should this happen, when in principle one anti-inflammatory treatment should work as well in younger patient as in an adult (e.g. fingolimod that currently also holds a license for paediatric MS).

Let’s look at the MRI metrics (aka secondary end points demonstrated in the figure below). Not surprisingly there is a statistically significant treatment effect with a reduction in T2 lesion load by 55% and Gad-enhancing lesions by 75% in the Teriflunomide group. In the TEMSO study (in adult RRMS), the reduction in Gad-enhancing lesions was -0.80; not very different to the TERIKIDS study.

One of the major requirements for a Phase 3 study to meet its primary endpoint(s) is adequate sample size. For instance, in TEMSO there were 721 participants (1:1 randomisation to teriflunomide vs. placebo). For this study to have reached significance for an effect size of 34% relapse rate reduction at a p-value of 0.29, the investigators would have needed to recruit 600 patients for the study to be statistically significant (not n=166 as in this study) – see Editorial accompanying this article.

So disregarding the entire study on the basis of underpowering would be foolish, and this is the view that has been taken in EUROPE whilst not in the US. You can therefore prescribe Teriflunomide in the UK but not across the pond. The ethics of care inequality is not a new one and not unique to MS. But, I and many others alike do feel that regulatory authorities need to move away from out-dated clinical endpoints and review the data that may be very compelling in terms scientific merit.

Figure: Primary efficacy analysis and key secondary imaging analyses(A) Primary efficacy analysis: Kaplan-Meier plot of time to first confirmed clinical relapse during the double-blind period. (B) Key secondary imaging analysis: mean number of new or enlarged T2 lesions during the double-blind period. (C) Key secondary imaging analysis: mean number of gadolinium-enhancing lesions during the double-blind period. HR=hazard ratio


Lancet Neurol. 2021 Dec;20(12):1001-1011. doi: 10.1016/S1474-4422(21)00364-1.

Safety and efficacy of teriflunomide in paediatric multiple sclerosis (TERIKIDS): a multicentre, double-blind, phase 3, randomised, placebo-controlled trial

Tanuja ChitnisBrenda Banwell  Ludwig Kappos  Douglas L ArnoldKivilcim GücüyenerKumaran DeivaNatalia SkripchenkoLi-Ying Cui Stephane Saubadu Wenruo Hu Myriam Benamor Annaig Le-HalperePhilippe TruffinetMarc Tardieu TERIKIDS Investigators

Background: Therapeutic options for children with multiple sclerosis are scarce. Teriflunomide is approved in more than 80 countries for the treatment of adults with relapsing multiple sclerosis. The TERIKIDS study examined the safety and efficacy of teriflunomide in children with relapsing multiple sclerosis.

Methods: The TERIKIDS trial was a multicentre, phase 3, double-blind, parallel-group, randomised, placebo-controlled study conducted at 57 clinical centres in 22 countries in Asia, Europe, the Middle East, North Africa, and North America. The trial enrolled patients aged 10-17 years, diagnosed with relapsing multiple sclerosis and with at least one relapse in the year preceding screening or at least two relapses in the 2 years preceding screening. Patients were randomly assigned (2:1) to oral teriflunomide (dosage equivalent to 14 mg in adults) or matching placebo, using an interactive web and voice response system, for up to 96 weeks. Personnel in all sites and all patients were masked to study treatment in the double-blind period. Early entry into a subsequent 96-week open-label extension phase was possible before the end of the double-blind period for patients with confirmed clinical relapse or high MRI activity (at least five new or enlarged T2 lesions at week 24, followed by at least nine new or enlarged T2 lesions at week 36, or at least five new or enlarged T2 lesions at weeks 36 and 48, or at weeks 48 and 72). The primary endpoint was time to first confirmed clinical relapse by the end of the double-blind period. Key secondary imaging endpoints were number of new or enlarged T2 lesions and number of gadolinium-enhancing lesions per MRI scan. Efficacy endpoints were analysed in the intention-to-treat population, and safety was assessed in all patients randomly assigned to treatment and exposed to the double-blind study medication. This study is registered with ClinicalTrials.gov (trial number NCT02201108) and is closed to recruitment, but an additional optional open-label extension is ongoing.

Findings: Between July 24, 2014, and the date of last patient visit on Oct 25, 2019, 185 patients were screened for eligibility, 166 (90%) were enrolled, and 109 were randomly assigned teriflunomide and 57 were randomly assigned placebo. 102 (94%) of 109 and 53 (93%) of 57 completed the double-blind period. Switch to the ongoing open-label extension because of high MRI activity was more frequent than anticipated in the placebo group (14 [13%] of 109 patients in the teriflunomide group vs 15 [26%] of 57 in the placebo group), decreasing the power of the study. After 96 weeks, there was no difference in time to first confirmed clinical relapse with teriflunomide compared with placebo (hazard ratio 0·66, 95% CI 0·39-1·11; p=0·29). Teriflunomide reduced the number of new or enlarged T2 lesions versus placebo by 55% (relative risk 0·45, 95% CI 0·29-0·71; p=0·00061), and the number of gadolinium-enhancing lesions by 75% (relative risk 0·25, 0·13-0·51; p<0·0001). Adverse events occurred in 96 (88%) patients in the teriflunomide group and 47 (82%) patients in the placebo group; serious adverse events occurred in 12 (11%) patients in the teriflunomide group and 6 (11%) patients in the placebo group. Nasopharyngitis, upper-respiratory-tract infection, alopecia, paraesthesia, abdominal pain, and increased blood creatine phosphokinase were more frequent with teriflunomide than with placebo. During the double-blind phase, four patients in the teriflunomide group had pancreatic adverse events (two with acute pancreatitis and two with pancreatic enzyme elevation), of which three events led to treatment discontinuation.

Interpretation: No significant difference in time to first confirmed clinical relapse was found, possibly because more patients than expected switched from the double-blind to the open-label treatment period because of high MRI activity. Key secondary imaging analyses and a prespecified sensitivity analysis of probability of relapse or high MRI activity suggest that teriflunomide might have beneficial effects in children with relapsing multiple sclerosis by reducing the risk of focal inflammatory activity.

About the author

Neuro Doc Gnanapavan


  • The question is who cares? Why would anyone ever prescribe a low-efficacy drug like teriflunomide to a child whose brain and spinal cord are being destroyed?

    • Probably because we would expect the paediatric population to be on a DMT for a lot lot longer than an adult, which is a big unknown in terms of their safety profile. Therefore an efficacious drug even if it’s 34% in terms of relapse rate reduction but safe is a good option.

  • It breaks my heart to think that children are having to deal with this awful disease. Researching children with MS was supposed to provide insights into what kicks off this disease – I’m assuming children who get MS have been infected with ebv (which goes against the view that getting ebv early in childhood is a good thing). An ebv vaccine can’t come quickly enough – particularly for children at higher risk of getting MS.

    • Not all have encountered EBV, the rates are low in the Eastern countries. But, studying this is helpful in understanding the MS prodrome.

  • Why would they conduct such study when it can be easily predicted the study is going to be underpowered?

    Gambling on regulatory bodies will be sensible? Unexpected low enrollments? Hoping for a miracle result of the drug in a different population for some magical reason?

    I understand more options is good but these kids on placebo!! Aren’t kids the best population to test IRT approach of anti-CD20’s and there has been usable of rituximab on kids and they had similar safety profile + very very effective?

    • I think you will always struggle to recruit the numbers needed in paediatric MS. There is a lot of pressure to complete studies over the 2 year period based on financial pressures. A 5y+ will not be on the cards.

      You’re correct about RTX’s expected safety profile in children, however, we know that this is not the case with COVID-19 with the greatest odds of ending up on ITU with RTX. It is good to have a range of options. In my opinion, what should change is that the regulatory authorities accept the secondary outcomes as the proof of principle that a drug works in the same way in children as it does in adults if the data is available.



Recent Posts

Recent Comments