Compass Therapeutics 3rd Annual Guggenheim Healthcare Innovation Conference

Thanks so much, Michael. Thanks for inviting us. Thanks for having us here.

Sure. Of course, I am not going to argue with you. It has been somewhat of a difficult year. Luckily, we are extremely well-funded. We just announced our earnings this morning. We are funded into 2028 for all of our development programs. I think, like all difficult times, right, really focus on the mission. We are really focused on developing a set of new drugs. We have got four drugs in the clinic. We can talk about each of those individual drugs, and we are really focused on advancing these programs to so-called inflection points, which will be driven by the clinical data. So, we are just really, really focused on not the difficulty, but the real opportunity with developing these drugs in these indications.

Sure. In addition, of course, to the ESMO presentation, we simultaneously published all of the data in the Journal of Clinical Oncology, on October 24, online. I think, what have we learned about tovecimig? I think, let's start sort of very, very simply, if you don't mind. The drug clearly works, right? There's just no question about it. We more than tripled the overall response rate. Progression-free survival had a hazard ratio of 0.44, a hazard ratio on PFS that you don't see every day. There's clearly activity with this drug. And I was very encouraged. We had a whole series of meetings with KOLs at ESMO, and every meeting we had, the KOLs started the meeting by saying, "Well, this drug clearly works." So, we have an unequivocal efficacy signal, and I think that's a very, very good place to start the conversation.

Sure. Let's talk about the endpoint first. On the overall survival endpoint, the study allowed crossover, of course. More than half the patients who were randomized to the control arm crossed over to receive combination therapy with tovecimig. And even though they crossed over, of course, in an intent-to-treat analysis, they are, quote-unquote, "counted" with the chemotherapy control arm. That's just the way the statistics work. So 85% of the patients in the study received combination therapy. The study was effectively a randomized trial of tovecimig versus paclitaxel versus tovecimig plus paclitaxel. The overall survival in the intent-to-treat analysis was not different. Fascinatingly, during the review of the draft of the JCO manuscript, one of the reviewers of that manuscript specifically asked for the Two-Stage Estimation method, and we did that, and that showed a hazard ratio of 0.85 when you correct for crossover.

So, this question about the hazard ratio being above one, hopefully we have put that question at least to rest. The second half of your question, of course, is very complicated, right? How does this all connect on the regulatory front? We submitted a meeting package in the summer. As I have said publicly, we were not afforded a meeting, and all of the responses were written responses only. We never really had a discussion about the survival data. As I sit here today, I don't really know what the answer to your question is. What we have said publicly, and I'll say again today, we're working to re-engage with the FDA. We're right in the middle of that right now.

We expect to have a meeting with the FDA, and hopefully, like many other companies, which you know these companies full well, over the past six to nine months, have had rather dramatic reversals of their FDA trajectory based on a different approach. So, we're trying that different approach. We're right in the middle of that. We'll see. Hopefully, that will put us in a position where we can say something more definitive about the regulatory path.

We received written feedback, which was very short feedback. We attempted to go back and get that clarified, and the clarification was not super helpful to us. So we were in this very difficult position where we felt like we needed to make a disclosure. FDA, quote-unquote, "recommended" that we do a study to demonstrate a survival advantage. So that's what we disclosed.

The RPSFT analysis could not be done, because the statistical assumptions for that method were not met. So, how persuasive will it be? I do not know the answer to that. I will say, we only very recently disclosed that, and that entire analysis was not included in our original submission to the FDA in the summer.