Zolgensma ‘Transformative’ for SMA and Gene Therapies to Come, Jerry Mendell Says in Interview

From its very first clinical trial — using  a dose widely frowned upon — Zolgensma has had a “transformative” impact on spinal muscular atrophy (SMA) and on gene therapy as a whole, said Jerry Mendell, MD, a principal investigator across Zolgensma’s clinical program.

“We did our best, I think, to make SMA treatment available and in a groundbreaking study,” Mendell, a practicing neurologist and researcher at Nationwide Children’s Hospital in Columbus, Ohio, said in an interview with SMA News Today.

Continuing results from its pivotal Phase 1 trial (NCT02122952) in 15 type 1 infants and its long-term extension study (NCT03421977), which were influential in Zolgensma’s recent U.S. approval, are changing how gene therapy’s potential is viewed.

Jerry Mendell, principal investigator for the gene therapy Phase 1 trial. (Courtesy of Nationwide Children’s Hospital)

“Gene therapy is the hottest item on the block now. But there was a time when we first got into this trial, where there wasn’t a person in the world who believed that gene therapy would work. We have to remember that,” Mendell said.

“There were a lot of naysayers. Just go to the internet and you’ll find all over the place people who were saying: ‘You shouldn’t do that, you’re just going to create a vegetable.’”

Chief among reasons for the success of what Novartis defines as its one-time therapy was the high dose given — 1.1×10¹⁴ vector genome per kilogram of body weight (vg/kg) — to 12 of the 15 babies, now toddlers, in the pivotal trial.

“This is the highest dose of virus that had ever been given in a clinical trial, and most investigators — and even the FDA — were very concerned about giving this much,” Mendell said. But views he has heard from experts since, including scientists “who’ve been in gene therapy world for quite awhile,” point to that dose greatly aiding gene therapy research.

“Basically, the response was … what you did was change the field, so that we can move forward with clinical efficacy and safety to a degree that we’ve never seen before,” he said of reactions at a recent, invitation-only meeting of gene therapy researchers in Miami. “And your outcome measures demonstrate that.”

Executives and researchers at AveXis — which led Zolgensma’s development and is now part of Novartis — and Novartis have also called the therapy “transformative” for SMA. Mendell agreed, but specified: “It’s transformative because this much virus has never been given, this much efficacy never achieved.”

What was learned in testing and refining what became Zolgensma is now aiding work in other chronic and rare diseases. “The safety and efficacy of the doses we used in this clinical trial now are being applied, as we speak, to other clinical trials that are not SMA,” he said.

One example is Duchenne muscular dystrophy, “one of my passions for the last 50 years,” Mendell said. “We’re doing a big gene therapy trial in Duchenne right now, and guess what dose we’re using: the same as for SMA. Not a surprise.” This Phase 1/2 study (NCT03375164) is testing the safety and efficacy of a microdystrophin gene therapy, delivered using an AAV viral vector, in children with Duchenne.

A pivotal trial in the making

SMA results from mutations in the SMN1 gene that significantly lower levels of a working SMN protein, leading to a swift loss of motor neurons, the specialized cells that control muscle contraction. Zolgensma uses a viral vector called AAV9, which crosses the blood-brain barrier to deliver a functional copy of SMN1 and restore production of a working and full-length protein in motor nerve cells.

But the road to opening a first low- and high-dose trial of a gene therapy in 15 infants with SMA type 1, which Mendell led at Nationwide starting in 2014, was not always smooth.

Work in the delta7 mouse model of SMA — developed by Arthur Burghes at Ohio State University College of Medicine, which Mendell’s hospital is part of — had shown that only the higher dose led to “100 percent” motor neuron rescue. This was key, as the FDA considers success in changing a disease’s natural history in an animal model a fair judge of benefit to people. (Natural history refers to how a disease would progress without treatment.)

But the agency’s first priority is safety, Mendell said, and regulators favored testing Zolgensma, then known as AVXS-101, in patients with the milder SMA type 2 disease before using it on babies with type 1, its most common and severe form.

“So we went to the FDA and showed them we had safety data to cover this dose, and we also had efficacy data to defend it,” Mendell said. “We pretty much stayed firm on the fact that there was scientific logic” in the trial’s design, including “a very rapid loss of neurons that takes place in the first six to nine months of the SMA child’s life.”

If a physician is “to have any chance of rescuing a baby,” he added, “you have to get under that, get the treatment before you have that severe neuronal loss so you can rescue those nerve cells.”

In this dose-escalating study, a low dose went to the first three patients — 6.7×10¹³ vg/kg — via intravenous (IV), or systemic, infusion. Early trials with patients grouped to receive increasing doses of an investigative treatment is not unusual, Mendell said, although gene therapy investigators now “try to discourage” them in FDA applications “largely as a result of the SMA study.”

Enrolled at between 6 and 7 months — beyond the typical six-month threshold for symptom onset in SMA type 1 — these three infants were weaker than the 12 later brought into the trial. Dramatic improvements were not seen, Mendell said, adding that in all 15 treated children, there was no “eureka moment” but “a slow process and one that evolved over time.”

Mendell is a clinical translational scientist — meaning he works to “translate” lab research into treatments — and as such said he  “generally take[s] a conservative view about efficacy.” As children gained in “motor milestones rather than regressing,” starting with breathing and swallowing better, eating and showing “limb control,” his and his team’s confidence slowly grew.

“Early on, I can specifically tell you that we remained skeptical and really had to be convinced,” and view “repeated confirmation” of gains, Mendell said. “Gradually … we saw patients who were then beginning to sit … a huge milestone. But it didn’t come about as a ‘eureka’ moment, it came about as an evolving improvement in motor function.”

All these milestones, and the simple fact that treated infants were living longer than SMA type 1 natural history typically allowed, were “game-changers.”

“But you take any one of infants and you’d see a slow evolution; you could map it out,” Mendell said. “Until you do that repeatedly, you always consider that there could be an outlier.”

Latest available results of this Phase 1 trial show survival and motor gains at 24 months post-treatment. Other significant improvements include 11 toddlers able to control their heads, sit without support for 30 seconds or more, and eat normally. As of March 8, all in the long-term START study (NCT03421977) show no loss of therapeutic benefit for as long as 4.3 years post-treatment.

“That supports the whole idea of long-term efficacy,” Mendell said.

And — with Spinraza (nusinersen, by Biogen), first approved for all disease types in December 2016 — the availability of Zolgensma strongly justifies the need for newborn screening for SMA, he added. Five states — Missouri, Minnesota, New York, Pennsylvania, and Utah — now routinely screen newborn babies for this disease; 13 others, including Ohio, are beginning pilot screening or considering doing so.

Zolgensma’s approval, Mendell said firmly, “means that there will be unequivocal newborn screening.” A pre-symptomatic infant was among those treated in ongoing SPR1NT (NCT03505099) at Nationwide: “We had the opportunity to treat an infant and never have them develop symptoms.”

Treating all with SMA

Mendell hopes that Zolgensma, like Spinraza, will eventually treat all with SMA, regardless of type or age.

“Absolutely,” he said, when asked. “There may be a group that is beyond the threshold of response” — older type 2 and 3 patients with only a “potential” for “slower” neuron rescue — but “that remains to be seen on the clinical level without speculation, because it’s been such a powerful treatment.”  

As approved by the FDA, Zolgensma may treat all SMA types but only in pediatric patients. It can be given as a one-time intravenous (IV) infusion to pre-symptomatic infants through toddlers up to 2 years old diagnosed with, or likely to develop, types 0, 1, 2, or 3.

Older children and adults with SMA types 2 and 3 will be treated with Zolgensma as an intrathecal (IT) injection directly into the spinal canal, a form of administration not yet approved.

A Phase 1 trial, called STRONG (NCT03381729) and recruiting type 2 patients between 6 and 60 months (5 years) of age, is delivering Zolgensma via intrathecal (spinal canal) injection. This change “was a scientific and clinical decision” to target the nervous system directly, Mendell said, to enable use of a “far less” amount of the virus than bloodstream delivery.

It’s safer, he added, both in avoiding an impact on other organs — possible with the systemic IV route — and in lowering the amount of antibodies against the AAV9 viral vector used as a delivery tool, which could produce an immune reaction. In fact, early findings presented at the 2019 American Academy of Neurology (AAN) Annual Meeting showed low rates of AAV9 antibodies in STRONG trial patients and not high enough to exclude them from participating.

A lower dose might also mean a less costly treatment, Mendell added.

A higher dose may also be possible — early studies in intrathecal delivery showed no significant toxicity at viral amounts at least 50 percent higher than the proposed treatment dose, he said, adding “I think you’ll see that some of these higher doses are tested in the clinic.”

Mendell has been at the forefront of SMA and muscular dystrophy research for decades; his first muscular dystrophy gene therapy trial dates to 1999. He was appointed director of Nationwide’s Center for Gene Therapy in 2004, and put neuromuscular disease at its forefront. He also recruited Brian Kaspar, now chief scientific officer at AveXis, and credits Kaspar with seeing the “potential” of developing a gene therapy using Burghes’ SMA mouse model.

In the end, he stressed, Zolgensma’s success comes down to the willingness of industry and academics to work together.

“We developed the product here at Nationwide Children’s, but AveXis was smart enough to adopt it,” Mendell said. That company, largely on the success of its pivotal Phase 1 trial, was acquired by Novartis in April 2018 for $8.7 billion.