Pleasure to be here at the Bank of America conference in Las Vegas. I'm Bassil Dahiyat, CEO of Xencor. We're a monoclonal antibody engineering company. This is our safe harbor slide. You can, of course, always check our filings with the SEC, the 10-Qs and 10-Ks. We are a protein engineering company, and what we've is engineering monoclonal antibodies to make them better therapeutics. Essentially engineering their immunological properties and their stability and their structures to make better drugs. We have built really a large intellectual property base around specifically that bottom half of antibodies. It's shown in blue on that diagram on your right. That's called the Fc domain. That's a part of the antibody that controls many of these features.
We've built a large toolkit that lets us create different kinds of antibody drugs with different advantages, and use that to build both a large internal pipeline, as well as a number of partnerships. Our internal pipeline is highlighted by a growing suite of what are called bispecific antibodies in oncology. We have five internal programs in clinical testing now. A partner has a sixth, and we're going to add a seventh in the clinic in the coming weeks. We've also, as I said, had a number of partnerships shown on the bottom here. Most recently, we announced one with Genentech to exploit these Fc domain tools to build better cytokine drug candidates.
In all cases, this is using the same suite of Fc domains that are plug and play and let us build a broad portfolio rapidly on the discovery side and focus our efforts on developing better drugs. Last, most recently, one of our partners, Alexion, launched commercially a drug called Ultomiris that uses one of our proprietary Fc domains for creating longer half-life. The idea of engineering the bottom half of the antibody is to really add another degree of freedom to what nature has been doing and what the biotech industry had done for the first sort of 20 years of antibody work, which was engineering at the tips. You make an antibody that sticks to a given antigen and hope that the natural antibody's function does something great.
Sometimes like for Rituxan or Soliris, it does, and you have a wonderful drug that helps a lot of people. In other cases, you really need extra oomph. At the top right, you see bispecific antibodies in the sort of first generation iteration that don't really look like antibodies. They try to stitch together two antigens using artificial structures. You see cytokine drugs, sort of the first biotech drugs. A lot of challenges in manufacturing of them, in their modularity and their persistence. By adding Fc domains, you can create entirely new structures that look much more antibody-like. The edges you get here are antibodies naturally persist a long time. They're easy to make. They're very stable. Nature evolved them to be that way, and we in biotech have exploited that.
By Fc engineering, we can access these properties now in a wide array of ways. We've built this suite of Fc domains to truly be plug and play by engineering different features in antibodies, based on a specific function of Fc domains and heightening or optimizing that in each given case. On the left in green, you have our Xtend Fc domain. That's an Fc domain that takes the natural property of antibodies, which is to persist in the circulation much longer than most proteins, and heighten that. We can triple the circulating half-life of an antibody. That's the Fc domain used by our partner, Alexion, in ULTOMIRIS. As you go across that slide, you can see each different feature of antibodies that you accentuate. You have your higher cytotoxicity. The natural cytotoxicity of antibodies kills tumor cells. Ours enhances at 30 to 40 fold.
That's now in phase III testing with our partner MorphoSys in the drug tafasitamab. We have our Immune Inhibitor Fc domain. Lastly, we have our bispecific Fc domain, where we've taken that natural dimeric structure and broken that symmetry and put it back together in a way that greatly expands the suite of molecules we can make. In all cases, these are very small numbers of changes to the Fc structure, just a couple, in most cases. You're better than 99% identical to natural antibodies, in which case the goal here is not to break what's already good. We use this plug-and-play suite of tools to build a broad internal pipeline. These are the molecules we built ourselves and are advancing with our own assets and financing. Lot of purple here. This is the growth of what our bispecific Fc toolkit.
We see an enormous amount of potential here, really untapped still, as the field starts using the new tools like Xencor and other companies have built for bispecific antibodies that act like regular antibodies. We're going to be exploring a lot of different biologies here. This is the pipeline of licensed assets. In most cases, we've never touched these molecules. It's almost like a software licensing model. We take an Fc we've already created, allow a partner to use it in a limited way. They pay us up-fronts, milestones, and royalties. One marketed product at the top, soon to be filed, a BLA for we hope to be an approval in about one year's time by our partner MorphoSys in oncology. On down the list, to our most recent deal with Alexion, sorry, Astellas, that we announced just about one month ago.
We have the ability to do these deals and not dilute the value of the platform by carving very narrowly what a partner can do and not having to deploy our own resources to support the partnership. I'll touch on the technology in a little more detail, going into our bispecific Fc domain. This is what generates the most questions amongst our investors, I'll spend some time on this. The idea here is you have on the far left, sort of the first-generation bispecifics, which were just stitching together the antigen binding domains from two antibodies and jamming them together. What we wanted to do was bring natural antibody properties to these molecules that were short acting, rapidly clear from the body, hard to make, very unstable, and very difficult to engineer or tune for particular potencies and affinities.
Unlike natural antibodies, which make great drug scaffolds. We engineered the Fc domain to create these two different halves of a structure, but inherently built into the Fc domain, so your final structure looks like a regular antibody. We wanted to make that in a way that's easily purified and that still maintains the very robust stability of regular antibodies. After a number of years of effort, we pulled together such a sort of feature set. From there, it's off to the races. We've, in a plug-and-play way, built our first three candidates with our bispecific toolkit. In parallel, by the way, we were supporting or really allowing our partners, Amgen and Novartis, to build their own molecules. These three are all in clinical testing right now. We've reported data on the first one on the left, XmAb14045, targeting AML.
In the middle is XmAb13676. On the right is XmAb18087. These target different tumors. In the middle, B-cell malignancies, on the right, neuroendocrine tumors, and gastrointestinal stromal tumors. By targeting different tumor antigens, but always on the other side, targeting a molecule called CD3, which is a T-cell target. If you can bind CD3 and bring that close to a tumor, you can really light up that T-cell and have it kill that target cell. That's the thesis there. In all cases, our preclinical data showed active molecules that persisted and had their activity persist a long time, a challenge that had been difficult for first-generation bispecifics. The first bit of clinical data we have is on XmAb14045 in AML, a very difficult tumor type. In this molecule, the tumor target is CD123, which exists on AML cells.
It also exists on various cells in the myeloid lineage. The drug was engineered, again, in this plug-and-play way with our standard bispecific Fc domain. We had initial data at ASH in December of last year. We had about a 28% rate of complete remissions or complete remissions with incomplete platelet recovery, or incomplete hematologic recovery, or so-called CR/CRI rate, which is the metric typically used. This was in a very heavily pretreated patient population. On the top, you look at pretty promising early efficacy. That was a rate of efficacy when we reached active doses after dose escalating very cautiously and carefully. With these new agents, in particular, we have to be cautious. On the bottom, you look at the tolerability profile. The typical toxicity of this kind of bispecific is what's called cytokine release syndrome. As you turn on T-cells, they release cytokines.
These are signaling factors that can create a sort of immune response that can often overwhelm patients if it's not managed well, and they can have severe sequelae, even death from that overactivation of the immune system. In this case, you can see as you track to the right on that plot, different days of dosing on day one, dosing after a week, the third week of dosing, fourth week, and each week you can see that rate declines. That's a general phenomenon. Once you can get through that earlier initial toxicity, you have a more tolerable drug. Now, this is a once-a-week dose agent, which is again, a big step above first generation bispecifics that were typically infused continuously by a pump that was worn by the patient for weeks at a time.
That was an initial proof of concept sort of for the platform, an initial look at promising data here. We are currently with our partner, Novartis, making plans for the next sets of clinical trials we want to do with this agent. The next suite of bispecifics we created, let us use this generic toolkit to create new molecules. This is our approach at looking at general T cell activation against tumors. The same approach that are done by the checkpoint class of drugs like nivolumab or pembrolizumab, for turning T cells on generally against tumors. In our case, we want to use something special about bispecific's ability to bind two targets at once. In particular, the ability to tune these to bind preferentially T cells that only have both targets. That favors T cells in the tumor microenvironment, a pretty well-established result in the literature.
How do we do that? We make each side of the antibody only stick sort of okay in terms of how tight to the target. When you have both together, you get this cooperativity effect, like a zipping up of a zipper or Velcro. We have put two of these in the clinic, and we have a third one coming soon, all with the same hypothesis, each testing a different biology because the biology is really the unknown here. We have wonderful molecular tools to test biologies. We are doing that. The biology here is tricky. On the top, we go after a very well-validated pair of targets, PD-1 and CTLA-4.
In the middle, we wanted to have both PD-1 blockade to let the T cells get turned back on, but also prime the pump with a molecule that actively, if you bind it, actively turns on T cells called an agonist. In this case, we selected ICOS. How did we select it? We literally tried a whole bunch of different agonist partners in the context of a bispecific antibody for the one that preclinically was most active. An empirical exercise. Ditto on the bottom. We wanted an agent, a bispecific that could be used on top of a PD-1 antibody. The assumption being many patients are going to be treated with those. What is the best way to add to that with triple checkpoint blockade? Again, we empirically tried a number of pairs. CTLA-4 and LAG-3, two well-known targets, worked the best at activating T cells preclinically.
All of these hypotheses are being tested. This year, we should have data from the first, the top one, XmAb20717, late in the year, showing initial bioactivity and pharmacokinetic data and tolerability data. Last, the most recent addition to this suite of molecules we are building with our bispecifics is the subject of our Genentech deal. It is an engineered cytokine. We use the Fc domain analysis, robust, stable scaffold to array a cytokine, IL-15, which is highly selective for T cells and natural killer cells to amplify them. We wanted to tune that molecule to be better tolerated than the natural molecules and last longer than the sort of the rapid spike of activity you typically see with cytokines, followed by a long tail off. We did something counterintuitive.
We turned down the potency of these naturally very potent molecules, put them on our heterodimeric Fc for long persistence, that combination gives what we hope to be a best-in-class profile, and our partner, Genentech, seemed to agree. We partnered this program specifically because Genentech has the scale and the capital to rapidly deploy this molecule if it succeeds in its initial dose escalation phase I study to be proved tolerable. To be expanded into many different combination trials very rapidly. That's the way to really properly exploit this kind of asset. We still share 45% of the worldwide P&L if it ultimately gets to commercialization.
This slide just shows the different types of bispecific antibodies or cytokines in the case of the last one that we're developing right now, sort of coded by the little symbol on the left, the red CD3, the blue tumor microenvironment, now the yellow cytokine, as well as a couple of the partnered that are not disclosed in class to show that we've built a large pipeline. We're trying a lot of biologies using our toolkit because we don't know what's going to be the right biology, but we have tools to do it. We're very early to the game in bispecifics, seeing issues just starting down this path we think will be a fruitful one. We want to be at the lead in a number of ways. I'll touch now on our autoimmune program of obexelimab. This is a B-cell inhibiting antibody. It is not bispecific.
It uses a different engineered Fc domain for quieting down the immune system. It's basically a highly active B-cell inhibitor that doesn't kill B cells. The typical way to inhibit B cells is using rituximab. That's often used in autoimmune disease, and you just try to wipe out as many B cells as you can to quiet down whatever immune activity they have got. That's used successfully in RA. A drug similar to rituximab, ocrelizumab, is used in MS very successfully. We wanted to have an agent that could be very simply given subcutaneously, that could be well-tolerated, and that had that reversibility that you might want if you want to remove the immune suppression. In short, obexelimab has successfully shown that from biomarker studies, very encouraging data in phase II studies in IgG4-related disease, a newly defined autoimmune disease, as well as in lupus and RA.
We're currently, as we focus on our bispecific antibodies in oncology, we're seeking partnerships for this program right now. I'll close with a snapshot at milestones for the year. In addition to trial initiations, both ourselves and by supporting Genentech's IND of our IL-15 program. We expect to have later in the year initial phase I readouts from three of our bispecifics. All of this activity, this broad activity, is supported by a balance sheet that we've built from both upfront payments from our partnerships as well as financings. We ended Q1 with $650 million in cash or cash equivalents. That included about $135 million in receivables as of March 31st that we have already received in April.
We have a balance sheet, and with revenues coming in from expected royalties, we think we can support this campaign of broad development, hopefully deep into clinic development for a number of years. Thank you very much.