Janux Therapeutics Study update

Good morning, everyone, and thank you for joining us. Earlier today, Janux issued a press release announcing initial Phase I clinical data for JANX011 in healthy volunteers. The press release, today's webcast, and the accompanying presentation slides are available on our investor relations website. Before we begin, I'd like to remind everyone that certain statements made during today's presentation and question-and-answer session may constitute forward-looking statements. These statements are based on current information, assumptions, and expectations, and involve risks and uncertainties that could cause actual results to differ materially from those expressed or implied. Additional information regarding these risks and uncertainties can be found on Janux's periodic filings with the Securities and Exchange Commission. We caution you not to place undue reliance on these forward-looking statements, which speak only as of today. Janux undertakes no obligation to update these statements, except as required by law.

With that, I'd like to turn the presentation over to Dr. David Campbell, President and Chief Executive Officer of Janux.

Good morning, and thanks for joining us. Today, we're presenting the first human data from JANX011, our CD19-targeted ARM product candidate, following completion of the healthy volunteer study. Before diving into the details, I thought it would be helpful to set the stage for today's presentation. At 3,000 micrograms, we dosed five subjects with a single subcutaneous dose and no step-up dosing. Most achieved greater than 90% B-cell depletion, and two of the five reached undetectable peripheral B-cell levels. CRS was grade 1 only. For us, these results do two things. First, they establish JANX011 as a clinical-stage autoimmune asset, and second, they provide the first human validation of the ARM platform. Today, I'll start by putting ARM into the context of the broader Janux portfolio. Then I'll walk through the human data, our development strategy, and the milestones ahead.

Janux now has three precision platforms, and we think they create complementary value drivers across oncology and autoimmune disease. TRACTr remains our oncology foundation, and TRACIr builds on that foundation through localized CD28 co-stimulation, with the potential to deepen and extend responses. ARM extends our precision T-cell engagement strategy into autoimmune disease, where the objective is to deplete disease-driving immune cells with an off-the-shelf biologic. I also want to point out an important distinction with ARM. Unlike TRACTr and TRACIr, ARM does not rely on masking or proteolytic activation. It is a bispecific antibody that achieves selectivity by engaging a different site on the T-cell receptor rather than CD3. Strategically, we think that's important. TRACTr and TRACIr expand what we can do in oncology, while ARM gives us the opportunity to build an autoimmune franchise using many of the same scientific, clinical, and manufacturing capabilities.

We view ARM as more than a single product opportunity. It has the potential to become another long-term growth platform for Janux. With that context, let me turn to JANX011 and the first human data. I want to start with a top-level view of the JANX011 clinical profile, and then we'll go through each of these findings in more detail over the next several slides. As I noted during the introduction, at 3,000 micrograms, we dosed five subjects with a single subcutaneous dose. Most achieved greater than 90% B-cell depletion, and two of the five reached undetectable peripheral B-cell levels. There was no priming dose or step-up dosing, and CRS was grade 1 only. We also saw a clear dose response, depletion persisting through day 42, and T-cell activation concentrated in the intended TRBV6-positive population. However, I would not focus on any one of those readouts in isolation.

It's how the activity, tolerability, and dosing profiles come together. Ultimately, the question is whether that overall profile can support broad use in autoimmune patients. When we put those pieces together, we believe JANX011 has now demonstrated the human pharmacology, tolerability, and intended mechanism needed to provide the clinical foundation for moving into autoimmune development. If you step back and look at the autoimmune treatment landscape, we think there's a pretty clear opportunity. At one end, CAR T has shown the potential for deep and durable responses, but it also comes with complicated manufacturing, lymphodepletion, specialized treatment centers, and management of toxicities like CRS and ICANS. At the other end, antibody-based therapies have attractive safety and convenience profiles, typically with intermittent outpatient dosing.

They've transformed treatment across a number of autoimmune diseases, but generally have not shown the same depth of immune cell depletion and response associated with cellular therapy. Conventional T-cell engagers sit somewhere in between. They've helped validate the importance of deep B-cell depletion, but achieving and maintaining that depletion has often required step-up dosing and frequent repeat dosing, adding a treatment burden and CRS-related safety considerations. That's really the opportunity we designed ARM to address. Combining the depth of T-cell-directed depletion with the practicality of an off-the-shelf biologic. The goal is meaningful depletion with a scalable and convenient treatment approach that can fit into traditional clinical settings. The healthy volunteer data provide an encouraging initial profile with an activity, tolerability, and dosing profile consistent with the strategic position we designed ARM to occupy.

ARM came out of a relatively simple question: How many T-cells do you actually need to engage in order to get the biology that you want? CD3 T-cell engagers have clearly shown that redirecting T-cells can generate very meaningful biology. The question we asked was whether you really needed to broadly recruit the T-cell compartment to get that biology. Once we thought we might not need to recruit every T-cell, we investigated a number of different T-cell receptor binding regions. We were looking for a population that could give us the activity we wanted with limited cytokine release. At the highest level, the T-cell receptor is the antigen recognition machinery on a T-cell. You have alpha and beta chains that make up the T-cell receptor itself, which is associated with the CD3 signaling complex.

The important distinction is that CD3 is broadly shared across T-cells, whereas different T-cells express different alpha and beta chains. This work led us to the beta chain binders, in which there are approximately 30 different beta chains in our T-cell repertoire. We then looked across a number of naturally occurring V beta populations. We focused on three things: the proportion of circulating T-cells they represent, limited expression on barrier T-cells that are important for antimicrobial protection, and limited representation in autoimmune disease. Based on that overall profile, we selected TRBV6 for JANX011. TRBV6 represents approximately 10% of circulating T-cells, so by design, the majority of the T-cell compartment remains outside the directly targeted population. We believe preserving that broader T-cell repertoire has the potential to be important for maintaining host immunity over time.

On the target side, JANX011 binds CD19, providing broad B-cell coverage to support the objective of deep depletion across relevant B-cell populations. Importantly, we are not limited by that 10% starting point. The biology is intended to be adaptive. We have shown that as target cells are encountered, the selected T-cell population can expand and differentiate into the effector cells needed for depletion. As the target is eliminated, the stimulus falls away, and the population can contract. This is similar to our own normal adaptive immune response. It is important to note that expansion is not the objective. The objective is engaging enough of the right T-cells to do the job. That is really the ARM thesis: generate the pharmacology we want while limiting the breadth of immune activation, and Janux is our first clinical test of that thesis.

Later in the presentation, we will show first human data demonstrating that T-cell activation was indeed concentrated within the intended TRBV6-positive subset, providing evidence that the molecule is behaving as designed. The healthy volunteer study gave us the first opportunity to test that thesis in humans, and I would emphasize one thing upfront. This study was designed to establish a starting profile, not the final clinical regimen. We evaluated single subcutaneous doses ranging from 30 to 3,000 micrograms in order to understand the human pharmacology and activity-tolerability relationship well enough to design the patient studies. Importantly, the study incorporated a predefined healthy volunteer B-cell depletion ceiling of approximately 95% in order to minimize infection risk, reflecting the risk-benefit considerations appropriate for this population. So we were not trying to establish the maximum achievable biological effect.

We were trying to determine whether JANX011 behaved the way we designed it and whether the data gave us a clear path into patients, and that's what I'll show you next. You've already seen the overall profile. Here, I want to focus specifically on the depletion pharmacology. On the left, you can see a clear dose-dependent response following a single subcutaneous administration and interesting durability profiles even at the less active doses. On the right, you can see the five subjects dosed at 3,000 micrograms. Most achieved greater than 90% peripheral B-cell depletion, with two reaching undetectable levels. I do want to point out the one patient who didn't have as deep a response. We've now determined that was due to suboptimal PK levels, exposure levels in that patient. So it was well-defined why that patient had a lower response than the others.

Importantly, this wasn't simply an early effect. We continued to see depletion through day 42. Again, I wouldn't look at the depletion results in isolation. What matters is the context. One subcutaneous dose, no step-up dosing, and only grade 1 CRS, including in both of those subjects who reached undetectable B-cell levels. For us, that's an important answer. Selective T-cell recruitment was able to generate deep B-cell depletion in humans without broadly recruiting the T-cell compartment through CD3. On the previous slide, we showed that JANX011 can drive deep B-cell depletion following a single dose. The next question is the other side of the design thesis. What was the inflammatory response associated with achieving that level of depletion? Looking first at CRS, there were no events among the 16 subjects treated from 30 to 300 micrograms.