Sunday, May 5, 2019

Zolgensma versus nusinersen for spinal muscular atrophy

This is a big moment for people diagnosed with spinal muscular atrophy, or SMA, a rare and potentially lethal genetic disorder that destroys muscles. For decades, there was no way to change the trajectory of their disease. They now have one marketed medicine, and this month, chances are they’ll have another: a gene therapy that promises a long-lasting treatment, if not an outright cure, through a one-time dose.

This weekend at the annual American Academy of Neurology meeting, patients, their families, and doctors will gain more insight about the gene therapy, Zolgensma, which is owned by drug giant Novartis (NYSE: NVS), and how it might stack up against the approved medicine nusinersen (Spinraza), owned by Biogen (NASDAQ: BIIB). They can also look forward to the latest clinical data from an SMA drug, risdiplam, from Roche, that, if successful, would be the first that a patient could take orally—a big deal, because the most severe cases of SMA are in newborns and infants, and Spinraza requires chronic spinal infusions.

“It’s ridiculously exciting,” says Jahannaz Dastgir, a pediatric neurologist at Goryeb Children’s Hospital in Morristown, NJ. “It’s a great time to be a doctor.”

All the new information comes amid anticipation that the FDA this month will approve Zolgensma. With the agency’s green light, it would be the second approved gene therapy in the US, and one of just a handful around the world.

Zolgensma will also face something other gene therapies haven’t: competition. Approved in late 2016, Spinraza has already proven effective and, after early hiccups, has become a big seller for the beleaguered Biogen, with $1.7 billion in sales in 2018 and $518 million in the first quarter of 2019.

Novartis has high hopes for Zolgensma, too, having paid $8.7 billion to buy its developer AveXis in 2018. Its success or failure will be a bellwether for the economics of gene therapy. (Nationwide Children’s Hospital in Columbus, OH, where the therapy was developed, will be watching closely too.)

If both Spinraza and Zolgensma are available, doctors, payers, patients, and their families will face tough medical, logistical, and economic decisions. So far, Spinraza has far more data to support it. But it has a $750,000 first-year price tag and requires a few spinal infusions a year at a $375,000 annual cost thereafter, for life. Zolgensma could cost $1 million or more (Novartis has hinted much more) for a single dose, theoretically a bargain if it saves lives and negates downstream medical and social costs that SMA patients and their families would otherwise face.

A recent survey of 30 physicians in the US and Europe by the investment bank Jefferies suggested that a majority of newly diagnosed SMA patients, as well as those currently on Spinraza, will get Zolgensma. Jefferies predicts $2.6 billion in peak sales for Zolgensma.

It’s possible that the best results could come from combination therapy, but that hasn’t been tested and the costs would be exorbitant.

Alex Fay, a pediatric neurologist at UCSF Benioff Children’s Hospital in San Francisco, CA, says he would be hesitant to switch patients if Spinraza is well tolerated and working. Adding more complication, says Fay, is the fast progress of the disease. “Those decisions are going to have to be made pretty quickly,” says Fay.

Information revealed soon could make those decisions easier. Babies diagnosed with Type 1 SMA, the most common and deadly form of the disease, often die before the age of two. Type 2 patients may never be able to walk, while Type 3 patients can walk initially before losing strength later in life. In all types, it seems that the earlier the treatment, the more benefit.

Thus far, all public Zolgensma data have been in babies with Type 1. There will be more of that at AAN. Studies presented at the meeting this weekend will also, for the first time, reveal Zolgensma’s effects on more moderate forms of SMA, and in patients who haven’t shown symptoms yet. Those data could help determine Zolgensma’s eventual reach.

More than 7,500 patients across several SMA types have now received Spinraza, some as long as six years. Biogen recently used that experience to turn up the heat on Novartis.

Last week it published results in Neurology, the AAN’s medical journal, from a long-term study in “later-onset” patients, aged 5 to 19, who were likely to develop Type 2 or Type 3 SMA. Each group showed improvements on tests of motor function; historical data suggest they should get weaker. A couple patients with Type 3 SMA even regained the ability to walk during the trial, Biogen said.

Citing the study and other data supporting Spinraza, Biogen CEO Michel Vounatsos was adamant on an April 24 conference call that the drug “will remain the standard of care for SMA for years to come.”

The presentations this weekend will shed more light on the potential benefits and risks of the new world of SMA treatments, but there will plenty of questions left unanswered. Here we break down four key SMA topics that will be under intense discussion.

Fast Access: SMA is a battle against time. Neurons die and don’t come back. Muscles waste away and are replaced by scar tissue and fat. The muscle-wasting is particularly fast for babies with Type 1. Time to treatment is of the essence. They may never sit up, roll over or walk, and muscles for breathing don’t develop properly, forcing the use of ventilators or tubes inserted into their windpipe (what’s known as a tracheostomy), and the risk of deadly respiratory infections.

The stakes are also high in other forms of SMA. More time untreated could mean abilities lost—like being able to lift your arm over your head, or type on a computer—that are never regained.

Delays in getting Spinraza after its 2016 launch have been ironed out, says Dastgir. She has treated about 30 SMA patients total, and a couple of them have received their first Spinraza dose two weeks after diagnosis. (Dastgir is on the AveXis scientific advisory board and has accepted speaking fees from Biogen.)

One little boy Fay treated was already on a ventilator when he got Spinraza. He still needed a tracheostomy. Novartis has to be ready with Zolgensma. “If there’s any chance of delays early on I’m just going to put kids on Spinraza,” Fay says.

The FDA Label: The FDA approved Spinraza for all forms of SMA, before Biogen had evidence for many of those patients.

How will clinicians and insurers react if the FDA approves Zolgensma for all SMA patients? The plan is to infuse it into the spine of Type 2 or 3 patients and into the bloodstream of babies with Type 1. Fay wants to see if one method will be as effective as the other.

Positive data at AAN from the STRONG study (in Type 2 patients) and the SPR1NT study (in pre-symptomatic patients) would help Zolgensma’s case. Fay will wait for more follow-up. “I don’t want to use Zolgensma prematurely,” he says.

Durability: The main selling point for gene therapy is staying power, which Novartis is counting on with Zolgensma. The longer it works, the more it should stave off downstream healthcare costs. But perhaps not as much as one might expect. In older patients, their SMA might not advance, which is “huge and possibly life-saving of itself,” says Dastgir. “But the damage has been done.” They’ll still need monitoring and treatment.

For newborns who receive treatment, the goal is to have them “walk and run,” says Dastgir, “but honestly, I have no idea what to expect.”

It’s also not clear if Zolgensma’s savings would include avoiding the cost of Spinraza. It would need to prove as effective as Spinraza, and last long enough to be worth its cost. That evidence will require years of patient follow-up.

Getting the right dose of Zolgensma will be important. A second shot might not work because the patient would develop an immune defense against the virus used as a delivery system. Some SMA patients will have that defense before any treatment. Nine of 177 (5 percent) of children under five screened for Zolgensma thus far have had these traits and weren’t eligible for treatment. There’s a test for pre-existing immunity, but how will practitioners deal with results that come back in a gray area? Will insurers refuse to pay for treatment?

“If costs were not an issue then we would want to do combination therapy,” Fay says. But they will be. “I honestly don’t know how insurance is going to handle it,” he says.

Safety: Thus far, Spinraza has proven largely safe. But there are important caveats. Patients need to get multiple spinal taps per year, for life. It’s an uncomfortable procedure, at best. Some patients might require sedation, which is worrisome because of potential long-term effects on developing brains. “We try to avoid repeated anesthesia in kids, as there are some risks,” Fay says.

Dastgir has avoided sedation in all her patients so far, using unconventional methods such as letting parents, gowned and masked, stay with their kids. But she acknowledges treatments that don’t require sedation—at least not repeated sedation—will be attractive: “Psychologically, Zolgensma will be more appealing to parents, as a one-time dose.”

Biogen last year published a retrospective study of eight patients in their Spinraza clinical trials and concluded that the anesthesia, over 61 procedures, was “safe and effective.” (Biogen did not respond to requests for comment by press time.)

Zolgensma has also appeared safe. The most consistent issue has been a spike in liver enzymes, alleviated with a short course of steroids. It’s a common occurrence in gene therapies that are delivered, like Zolgensma, with the help of adeno-associated viruses (AAVs).

However, Novartis reported in April that one Type 1 patient died in a trial in Europe, and the investigator deemed it “possibly” related to treatment. Novartis executives later said the death could have been due to the steroids and that regulators weren’t concerned. But as SVB Leerink analyst Mani Foroohar mentioned recently in a research note, “steroid treatment is part and parcel of AAV administration,” leaving the question whether the steroid course could limit Zolgensma’s reach, he wrote.

Noting how sick Type 1 patients are, Fay of UCSF said he’ll be more concerned if it’s determined the death was caused by an immune attack against the gene therapy. The autopsy results have yet to be released.

1 comment:

  1. In order to provide policymakers with a broad view of cost-effectiveness, we also examined costs per LY gained. This approach values any life extension, even at a very low quality of life, as equal to life extension at full health. Cost per LY (life year) gained does not capture improvements in quality of life as intended by ICER’s stated goal of highlighting an “equal value for life-year gained” (evLYG) measure, but in this case it was not possible to construct this measure, and viewing results of both the cost per LYG and the cost per QALY (quality-adjusted life yeat) gained will ensure that policymakers can feel confident that they are considering information that poses no risk of discrimination against this patient group. For Spinraza in presymptomatic SMA, we estimated the cost per LYG as $652,000 from the health care sector perspective. For Zolgensma in patients with symptomatic Type I SMA, at a placeholder price of $2 million the corresponding finding in the health care sector analysis was $182,000 per LYG. In this analysis, Zolgensma’s price would need to be approximately $1.5 million to meet a $150,000 per LYG threshold. We performed multiple additional sensitivity and scenario analyses to address multiple avenues of uncertainty. We conducted numerous scenario analyses to explore questions about the best way to model the connection between motor skill improvements and quality of life, the impact of different time horizons and of a societal perspective on modeling results, and the relevance of substantial non-drug health care costs that continue to accrue when a treatment extends life. Except for one scenario analysis, which took a 10-year time horizon, we assumed in all other analyses that the short-term benefits of both treatments persist for a lifetime. Although there remains substantial uncertainty about whether this will prove true, input from clinical experts and judgments based on the mechanism of action of the two treatments leads us to believe that our base-case assumption of lifetime durability of benefit, while it may be viewed as optimistic by some, is the best starting point for a judgment of the value of these treatments at this time.

    For Spinraza, when accounting for model input uncertainty through scenario and one-way sensitivity analyses, the incremental cost effectiveness ratios did not fall below $670,000 per QALY gained. The results were most sensitive to the length of survival, the costs associated with treating people with SMA, and the utilities in both the “sitting” and “not sitting” health states. Results from the probabilistic sensitivity analyses found that Spinraza had a zero likelihood of achieving costeffective thresholds of less than $500,000 per QALY gained.

    For Zolgensma, when accounting for model input uncertainty through scenario and one-way sensitivity analyses the range in the incremental cost-effectiveness ratios was $199,000 to $406,000 per QALY gained. The results were most sensitive to the length of survival, health care costs, and utility in both the “sitting” and “walking” health states. Results from the base-case probabilistic sensitivity analysis found that Zolgensma had a 0.1% chance of being cost effective at thresholds of $150,000 per QALY but 100% chance of being cost-effective at thresholds above $300,000 per QALY gained.