High-Dose Nusinersen and the Future of SMA Treatment

The FDA's March 2026 approval of a higher-dose nusinersen (Spinraza; Biogen) regimen for spinal muscular atrophy (SMA) marks the latest chapter in a therapeutic success story. When nusinersen was first approved in December 2016 at a dose of 12 mg, it fundamentally altered the natural history of the disease. The subsequent approvals of onasemnogene abeparvovec-xioi (Zolgensma; Novartis) in 2019 and risdiplam (Evrysdi; Genentech) in 2020 brought about a 3-therapy landscape that has since reshaped clinical practice, newborn screening protocols, and long-term expectations for patients across the SMA severity spectrum.

Against this backdrop, the clinical rationale for a higher nusinersen dose was grounded not in inadequacy of the original regimen, but in the recognition that the drug's favorable safety profile left room to explore whether additional efficacy remained achievable. The DEVOTE clinical program, a 3-part study designed to evaluate the 28-mg maintenance dose and a 50-mg loading dose, demonstrated meaningful incremental benefit—a finding that was methodologically demanding to establish precisely because the 12-mg baseline was already effective.

The FDA's approval of the new regimen adds a clinically meaningful option, but also raises practical questions about insurance authorization and administration logistics that treating physicians will need to navigate. Thomas Crawford, MD, is a pediatric neurologist at Johns Hopkins who has been involved in SMA care and research for more than 30 years, spanning the era before any disease-modifying therapy (DMT) existed through the current multi-agent landscape. NeurologyLive® spoke with Crawford following the new approval to get his take on the regulatory decision and what it means for day-to-day clinical practice. In the conversation, he addressed the practical implications of the new dosing protocol, offered his view on how clinicians should think about the 3 approved DMTs relative to one another, and discussed what he sees coming next—including an antimyostatin therapy anticipated for approval later this year, an implantable intrathecal catheter program under study, and EVOLVE-SMA, a new functional classification system he is actively promoting to track the changing nature of SMA in the post-DMT era.

NeurologyLive: Can you give some background context about the FDA’s recent approval of the high dose regimen for nusinersen?

Thomas Crawford, MD: Nusinersen was initially approved in late 2016 at a dose of 12 milligrams for everybody. Since then, there was some thought—we chose that dose—we thought it was a large dose—but indeed, there was no reason to consider not giving a higher dose, because it had now been found to be incredibly safe. So the idea that there might have been some efficacy left on the table with that choice led Biogen to try to study a higher dose.

Now, that's really hard to do, because the initial 12-milligram dose was just so extraordinary. Babies were living that used to die. How do you study an improvement where the marginal benefit is going to be relatively small, as opposed to the incredible “they're living where they didn't before” kind of study that required a small sample size? So for that Biogen put together a series of studies—it's complicated to try to tease out if there was a meaningful difference between the new high dose at 28 milligrams compared to the old 12 milligrams—but indeed, they were able to show benefit.

What was your immediate reaction/feelings about the FDA’s decision?

I was very pleased to see that the approval went through. The remarkable thing is, there was absolutely no difference in safety. This drug is safe. The major difficulty is that it has to be administered by an intrathecal approach, which is not great—but there has been no concern about the safety differences between the 12 and the 28 milligrams. With that, I was very pleased to see that the FDA approved it. They took their time about it, but we got it.

What are the main things you think doctors should know about the new approval?

The main thing is that there's really no reason not to go forward with the 28 milligrams. It is the same price, with one exception: the new protocol starts with a single dose at 50 milligrams, as opposed to the continuous dose at 28. The 28-milligram dose is the same price to the insurance companies, but we have to obtain a single authorization for the 50-milligrams because it's on a per-milligram basis required by the FDA. So that first loading dose is more expensive.

There's no reason not to do it.

Can you discuss your view on the future of SMA therapeutics in general?

SMA has been really exciting since the original antisense oligonucleotide approval. We've now got 3 different therapies, all classed as DMTs—we know SMA is a slowly degenerative disease, and these therapies—all in their individual ways—appear to largely either cease or very strongly slow the rate of further degeneration.

As near as I can tell, there's really no difference in efficacy between the 3 therapies. We've never had a head-to-head comparison, and won't, because it would be too expensive and too difficult to show a difference because the difference is going to be very small.

So the differences between them are in the perception of deficit. In the case of nusinersen, the perception of deficit is the spinal taps. There's clearly a lot of problems with that, because it's a difficult procedure. In the case of Zolgensma, the gene transfer, it initially was only available to babies younger than 23 months by FDA approval, but actually less than 6 months by the initiating studies. More recently, an intrathecal approach has been approved, but it's based on a fairly small number of patients, and I still have questions about larger issues about safety that I think need to be established. Risdiplam, which is the oral drug, is also effective and has been shown to have widespread effect—but there are a certain number of patients who find it difficult to tolerate because of more mild side effects. And I think that for the very youngest babies, there's still some reason for concern that safety hasn't been validated, although they have their approval from the FDA.

So it's the perception of deficit, rather than the actuality of a demonstrated perception, that determines which therapy you want to go forward with.

What are the next steps in advancing treatment personalization in SMA?

The first and most important thing is that we need to get these DMTs into patients as soon as the diagnosis is made. That is greatly advantaged by the now-universal access to newborn screening programs. When the newborn screening call is made, there's a need to move with all speed to try to get one of the therapies in.

Now, there's been a practice advancement in that many babies are now receiving bridge dosing with risdiplam before going on to either continue with risdiplam, or proceed to gene transfer or nusinersen dosing afterwards. I think that's a good thing. It hasn't been studied formally, but there's hardly any plausible reason why not to do that. It does require some skill and interfacing with the various insurance companies to accomplish, so that takes some extra work—but really, we want to move quickly.

There have also been some other advances. Biogen has an ongoing trial to enable an implantable catheter so that the possiblity of avoiding the need for the spinal taps would be very much advantaged, if and when that study is completed. It also has the advantage that there’s hopes—not yet demonstrated, just hopes—that we could implant the catheter with the end of the port relatively high in the spinal cord, so that should improve delivery, as well.

Then there's the opportunity for nonDMT therapies. In the wings right now is an antimyostatin therapy that is presently given intravenously, and possibly later could be given subcutaneously—the generic name is apitegromab, made by Scholar Rock. We expect that approval to be sometime this year, so that would be useful. It makes people stronger, which is pretty remarkable. We'll see what the price is and what the label is from the FDA—obviously we’re waiting with anticipation to see what that's going to be.

There are also some other therapies that are involved in some more complicated ways that we think could either improve SMN enhancement, or other backdoors to try to make motor neurons stronger—from NMD Pharmaceuticals and some other ideas people have around. So we're not done. That's the cool thing. After making some really extraordinary—some of the most dramatic—changes in the course of a fatal disease possible, I want to emphasize that we're not done. There's still more stuff to do.

Is there anything else you want to add?

There's a complexity here. With newborn screening, we are no longer seeing large numbers of new babies with profound weakness entering the system. I joke that we're not seeing kids in motorized wheelchairs running around and nailing me in the kneecaps when I go to conventions. That's really, really good.

On the other hand, the two-copy kids—the kids with the more severe genotypes who are treated at birth by a DMT—are not normal. And in the past, they would have died. So I think the 10- and 20-year prospect is that we're going to have a much larger percentage of kids who are modestly weak: able to walk, or barely not walk. They're going to survive—and they would have succumbed before. So their numbers are going to start increasing, so I expect there's going to be a change in the population of SMA from severe disease with profound weakness in all the people who were rescued after a lot of damage was done, to this more indolent form as these kids grow older. It will be interesting to see if the total debility of SMA increases or decreases as a consequence of this.

I’m promoting a new metric for SMA called EVOLVE-SMA that is a very simple way of assigning how severely affected people are by what their functional disabilities are. It's divided into tiers, with the top tier being a person who can climb up a curb by themselves, and just below that, people who can get out of a car, go into a store, pick something up from the store, and walk out on their own. Each tier is defined by some function that is important, and critically, the tiers don't overlap—so you're not going to have people passing a higher tier but not passing a lower tier. I would like people to start using the tier-system to categorize their populations. Remarkably, this can be done by patients. They can assign their own tiers as accurately as a doctor or a therapist can.

With that, we should be able to do a population analysis of how much debility is associated with SMA over the next decade—and by that means, see what the new nature of SMA is in the postDMT era. So I'm in promotion mode for that, I'd like people to take on that.

This transcript has been edited for clarity.