Nkarta Study update

Please go ahead. Thank you, moderator, and welcome, everyone.

We are grateful you could join us this morning. I would like to remind everyone that today's call will include forward-looking statements, and actual results may differ materially. Please refer to the risk factors in our most recent quarterly report on Form 10-Q filed with the SEC for a discussion of the factors that could cause those differences. I want to open today by thanking our patients and their care partners for their trust and their participation in our study. Without them, the search for a better treatment wouldn't be possible. We are here for them and not the other way around. We are excited to be here at ACR Convergence in Orlando, where two world-renowned rheumatologists are presenting initial data for our investigational CAR NK cell therapy, NKX019.

Dr. Guillermo Valenzuela, founding medical director of Integral Rheumatology and Immunology Specialists, shared our initial myositis data yesterday in an oral presentation. Our systemic sclerosis data will be presented in about two hours during a poster session by Dr. Dinesh Khanna, director of the Scleroderma Program at the University of Michigan. On Wednesday morning, our overall safety data across all autoimmune indications will be the subject of another oral presentation by Dr. Valenzuela. Here is our agenda for today's call. Shawn Rose, our chief medical officer and head of R&D, will walk you through the data. President Nadir Mahmood will review market opportunities and provide an update on our cash runway. I'll come back for a quick recap, and then we'll open it up for questions. This morning, we are presenting the first data from our autoimmune program in myositis and systemic sclerosis.

These two indications will now be our sole focus as we work towards pivotal trials next year. We will carefully follow the rheumatoid arthritis patients we've treated so far and have been encouraged by responses to date for safety and preliminary efficacy, along with patients enrolled with other indications. However, the focus of our time and resources now is on myositis and systemic sclerosis. We will revisit the other indications as time and resources permit. The bulk of our data we'll highlight this morning are from myositis and scleroderma patients with at least 90 days of follow-up, treated at the highest dose we've reached so far. That's three doses of 4 billion cells given on days 0, 3, and seven for a total of 12 billion cells.

Now, at Nkarta, we've always believed the safety profile of NK cells would enable patients to receive cell therapy in a more accessible and convenient way. Our engineered CAR NK cells are administered on demand without apheresis in an outpatient setting, without an overnight hospital stay, just a two-hour observation period after each dose, without the necessity for patients to travel hundreds of miles to a major academic center, and without a hematology oncology cell therapist needed to be on call in the event of a life-threatening immune reaction. Also, our allogeneic CAR NK cells can be redosed, a key advantage over current autologous CAR-T approaches. Today, we will present proof-of-concept data for NKX019 in autoimmune disease for patients given a three-dose cycle of cells following preconditioning.

Now, the three things that stand out about the data are, first, we're seeing a dose response in line with what you would see with a traditional small molecule therapeutic or biologic. Second, we're seeing impressive early efficacy in a range we believe to be comparable to what has been reported for CAR-T. And third, the safety profile of NKX019 remains highly promising and, importantly, free of immune effector cell toxicities. So to date, no CRS, no neurotoxicity, no IEC-HS, and no GVHD. So before Shawn goes through the data, I want to talk about our ongoing dose optimization work and some key translational findings that are not reported in our data analysis today. The data we're presenting today reflects three infusions of NKX019 administered on days 0, 3, and seven following preconditioning lymphodepletion. The graph on the left shows NKX019 exposure and the dose response we've observed.

As you can see, each step up in dose from 1 billion to 2 billion to 4 billion cells given on day 0, 3, and seven produces a higher peak and a larger area under the curve. The graph on the right shows host T-cell levels recovering from lymphodepletion between day four and day seven. Looking back at the graph on the left, you'll notice that both peak levels and overall exposure from the final dose are reduced due to the return of those T cells. In other words, today's data most likely reflect a reduced contribution from that third dose. To address this, we recently amended the Ntrust protocol to compress the dose schedule. We are now dosing patients on days 0, 2, and 4. This way, all three infusions are administered before those T cells recover.

Compressing the dose regimen is also more convenient for patients. They can now complete a three-dose regimen in five days. This is effectively dose escalation since patients to date may not have been getting the full benefits of that final dose. As I said, we believe our early data in both myositis and scleroderma is comparable to CAR-T. We believe this data could improve further when patients get full exposure from all three of those doses. We also have the option under the Ntrust protocol to go to 6 billion cells per dose for a total of 18 billion cells, and we plan to exercise this option and enroll a handful of patients at the 6 billion cell level on the compressed schedule of day 0, 2, and 4. Why? Because our safety profile to date gives us that flexibility, and we don't want to leave any efficacy or durability on the table.

That's the advantage of having a cell therapy with a clean safety profile and a dose response. We owe it to ourselves and to our patients to enter pivotal development at the optimal dose. One final point that's important to address, it concerns the recent safety news from a competitor's trial involving a different investigational NK cell therapy administered with fludarabine as part of the lymphodepletion regimen. Since fludarabine was introduced into Nkarta's lymphodepletion regimen, our program has had a consistent safety standard across all studies for patients that have kidney impairment. Outlined in our protocol, patients with compromised kidney function are required to receive a reduced dose of fludarabine as part of their preconditioning.

Furthermore, beyond a defined level of renal impairment, patients are not eligible to enroll in the Ntrust trials at all. As I mentioned, our protocol also gives us the option to retreat patients, and we've retreated three patients in the study who did not respond to their first cycle of therapy. Each received a second cycle of lymphodepletion, followed by three infusions of 4 billion cells. That's a total of 12 billion cells per protocol. All three of the patients have been safely retreated to date with LD and cells with three months, two months, and one month of follow-up respectively. We'll provide additional safety and preliminary efficacy data on these three patients in the near future. You'll hear more on this from Shawn, who will now take you through the initial data for NKX019.

Thank you, Paul. Good morning, everyone. It's a privilege to present these results. Let me begin by orienting you to the Ntrust study design. It reflects what Paul just described. NKX019 is dosed like a traditional biologic in the outpatient setting. Patients receive three days of lymphodepleting preconditioning, followed by three doses of NKX019 via infusion. Dosing began at 1 billion cells per dose, then we dose escalated to 2 billion cells per dose, and then to 4 billion. Paul already covered a recent refinement of the schedule based on our translational findings and why we're exploring 6 billion cells per dose in preparation for pivotal trials. Under our study protocol, we assess the response for up to two years and follow safety for up to 15 years. On the right are the endpoints. First and foremost, we evaluated safety and tolerability.

In anticipation that fludarabine levels increase in patients with compromised kidney function, our protocol from the start has included a safety standard for patients with compromised kidney function. Patients with compromised kidney function are required to have a reduced dose of fludarabine as part of their lymphodepleting preconditioning. Beyond a defined level of impairment, patients are not eligible to enroll. We also assessed NKX019 exposure and B-cell depletion and recovery, which allows us to determine whether we're achieving immune reset. For efficacy, we used established measures in each disease. In myositis, that's the total improvement score. TIS40 is a moderate response. TIS60 and above are considered major responses. TIS40 is our target because it is the ACR and EULAR-defined threshold for a moderate response. TIS combines muscle strength, patient and physician assessments, and other disease measures, so a moderate response reflects broad improvement in the disease.

The criteria are now widely used as an endpoint in myositis trials, including trials that supported approval in dermatomyositis. We also measure MMT8 for muscle strength and CDASI activity score for skin disease. In systemic sclerosis, we looked at changes in skin thickening with the modified Rodnan skin score, or mRSS, and evaluated overall improvement using the rCRISS composite score. Because the disease affects multiple organ systems, rCRISS measures skin, lung function, physical function, and the patient's and clinician's global assessments. Here, we're aiming for rCRISS-25 or higher, which means at least a 25% improvement across multiple of those five domains. A multi-domain response reflects a more holistic improvement in disease activity. That could meaningfully advance care for these patients, who are typically treated with therapies that target a single organ system or domain.