Guts & Glory: Both of Cynata’s iPSC Trials Flop, GVH Phase 2 and KOA Phase 3

Sad moment for the industry, and especially for Australia’s iPSC developer, Cynata Therapeutics. 

We obtained a copy of their investor webinar announcing the results from its two iPSC-derived mesenchymal stem cell (MSC) trials. The Phase 3 trial for knee osteoarthritis (KOA) and the Phase 2 trial for graft-versus-host disease both failed to meet their primary or secondary endpoints (though both appeared safe and well tolerated).

The webinar opens with executive chairman, Jeff Brooks, “Like everybody else, we’re quite more than disappointed and quite shocked by the results we’ve seen last week. I just can’t tell you how disappointed we are. Everybody’s put their heart and soul into this for a number of years…”

Send me the newsletter, I agree to the privacy policy*

Yes

Subscribe

Send me the newsletter, I agree to the privacy policy*

Yes

Subscribe

Dr Kilian Kelly, CEO of Cynata, continued with readouts and details of the trials. Since then, the stock has lost over 96% of its value:

Phase 3 KOA

Unlike GVH, Cynata has not attempted a previous trial for KOA. However, there is a wealth of data for this indication; it seems to be the hottest target and the largest market for stem cell therapy. Everybody’s got a bum knee.

The existing data can be mixed, but there are promising therapies such as Cartistem by Medipost (Korea), which is umbilical cord blood-derived MSCs + scaffold. They’ve had positive Phase 3s in Japan/Korea, and will begin in the US this year. 

Building on that, Cynata appears to be the first to attempt a Phase 3 iPSC-derived MSC trial for this indication (though I could be wrong). The active group had a decrease in pain “in line with expectations”, a 50%, but…. The same was true for the placebo group, with about 48.1% of patients achieving a similar reduction. 

“Needless to say, this was not statistically significant,” Dr. Kelly said. The secondary endpoint, cartilage thickness assessed via MRI, showed greater cartilage loss in the active group than in the control group. 

Dr. Kelly made a few interesting remarks on why this may have occurred:

• The issue wasn’t that the stem cell group didn’t report results; it reported great results, in fact. The problem was that placebo performed just as well. Pain is obviously a subjective endpoint, and OA trials are well known for large placebo effects, especially when the administration involves an injection as opposed to a pill. 
• Probably one of the most interesting takes was how the amplified enthusiasm around stem cell therapies may have affected outcomes. It’s all over social media, often seen as a panacea; there’s a lot of hype. If a patient hears Joe Rogan say stem cells regrew his rotator cuff, and believes they’re getting the same therapy, they’re very likely to report progress on subjective outcomes.
  • Mayo Clinic did an interesting study on this. Patients with bilateral KOA received bone marrow-derived MSCs in one knee and saline in the other; both knees improved at about the same rate.
  • Duke conducted a study that randomized patients to receive steroids (control), stromal vascular fraction (adipose-derived MSCs), bone marrow-derived MSCs, or umbilical cord tissue, all of which showed similar improvements.

Phase 2 GVH

This was perhaps more surprising, because the only proven MSC indication is actually GVH. There is one single FDA-approved MSC therapy, Ryoncil, which is allogeneic, bone marrow-derived, and effective when administered IV. 

Given that, Cynata ran a Phase 1 using its iPSC-derived MSCs for this indication, with good results. A total of 15 patients were treated with CYP-001:

• Overall response rate by Day 100 was 86.7% – 13 out of 15 patients showed an improvement in GvHD severity by at least one grade compared to baseline
• Complete response rate by Day 100 was 53% – GvHD signs and symptoms completely resolved in 8 out of 15 patients

However, in Phase 2, the results were inconsistent with those in Phase 1. Dr. Kelly noted the primary endpoint (overall response rate) saw a “very slightly higher… rate in the active group of about 57.7%, but that was compared to 54.8% in the control group. We hoped it would have been much higher than that”

On the secondary endpoints (day 28 complete response rate, day 100 overall response rate/overall survival rate, etc.), “it’s largely a similar pattern with no significant differences between groups”. The overall survival rate at day 100 was “nominally higher” in the active vs. control group, but it was “hard to kind of draw any conclusions from”. 

He addressed the question, how is this possible after a great Phase 1? 

• The Phase 1 had a “slightly different” patient population. It was comprised of patients who had already failed to respond to steroids, and at the time, there were no other approved therapies available.
• But, by the time Phase 2 began, the treatment landscape changed. Ruxolitinib (a JAK inhibitor) had been approved and became the standard of care for GVH. To avoid withholding an approved therapy from very sick patients, Cynata instead enrolled newly diagnosed patients before ruxolitinib would typically be given. 
• However, many patients in the trial ultimately received ruxolitinib anyway. According to the trial protocol, once they switched therapies, they were counted as non-responders. Dr. Kelly questioned whether clinicians “waited long enough to allow the MSCs to work,” though he acknowledged physicians appropriately prioritized patient care over the study.
• Cynata also considered combining its iPSC-derived MSCs with ruxolitinib, but felt that approach had drawbacks. Their cells have demonstrated a “very clean safety profile,” while ruxolitinib is associated with significant side effects. Combining the two could eliminate that safety advantage, and proving an additional benefit over a drug with roughly a 60% response rate would likely require a much larger trial.
• Another option was enrolling only patients who had already failed both steroids and ruxolitinib. However, Dr. Kelly said there are very few such patients, and those who do exist are typically “very sick,” making recruitment and trial execution extremely challenging.

What’s next? 

Despite the disappointing outcomes, Dr. Kelly maintained that the results do not invalidate the underlying iPSC platform. “I still believe in the platform and its ability to make MSCs that are very consistent and functional and potent.” He added that the company’s challenge is now determining how future development can be funded.

For now, the company is cutting costs, expects to receive an Australian R&D tax incentive later this year, and is evaluating all strategic options while continuing its kidney transplant study, which is funded by an external partner. 

Given these results, it begs an uncomfortable question. If an iPSC-derived MSC product failed to separate from placebo in a Phase 3 trial, what would happen if many of the cell therapies currently marketed in the U.S., including BMAC and birth tissue products, were held to the same standard? How much of the benefit is true biological effect, and how much is placebo?