Now, there is accumulating evidence that suggests that radiopharmaceutical therapies can induce genotoxic stress within the marrow. This can sometimes cause prolonged cytopenias and rarely therapy-related myeloid neoplasm. Now, why is the marrow exposed? PSMA is not thought to be expressed within the hematopoietic niche. Rather, perhaps the marrow's exposed through circulating activity or skeletal crossfire. In talking about the spectrum of hematologic disorders, it's important that we use common language. Clonal hematopoiesis has already been referenced a few times during this conference. When I mention clonal hematopoiesis, I'm referring to mutations that can occur within hematopoietic stem and progenitor cells. Those, in turn, confer a competitive advantage to select cells in the marrow. We know that CH is very common, even in patients who don't have cancer, even in patients who aren't exposed to genotoxic therapies.
It happens naturally with aging, but it can be further enriched by some of these genotoxic exposures. The two states that I'm going to focus on are CHIP and CCUS. When I refer to CHIP, I'm speaking about the presence of myeloid compartment mutations in the absence of cytopenias, whereas CCUS has those same mutations in the presence of cytopenias. It's important to frame CH as a precursor state. I view this as a biologic substrate, upon which additional instability accumulates over time and can rarely progress to therapy-related myeloid neoplasms. Now, what are the mechanisms by which this occurs? We all have somatic mosaicism happening in tissues all over our body, including the marrow. As we age, this increases, and then we give a radioligand therapy or other anti-cancer therapies that exert selective pressures within the marrow. You get preferential expansion of clones that harbor pertinent mutations. We think the DDR alterations are most pertinent here.
Then, the RLT resistant clones can then outcompete the wild-type cells, resulting in cytopenias at later time points. What is the evidence that we have from hematologic toxicity from our phase-three trials? Here, you'll see VISION, PSMAfore, and PSMAddition. They did a beautiful job outlining the rates of anemia, thrombocytopenia, leukopenia. We see other terms of interest too, like pancytopenia and bone marrow failure. PSMA-4 reported that, on average, cytopenias after exposures to lutetium will improve within two months, but what you're not seeing, at least in the data that was available a few months ago, are any reported cases of MDS or leukemia. Now, in the absence of that data, what can we gather from other sources? There are two large case series that I want to highlight, one from Australia and one from my institution. At Peter Mac, they looked at 381 patients who had exposure to lutetium. Subsequently, 5 patients or 1% were diagnosed with a therapy-related myeloid neoplasm.
Now, that includes MDS, APL, and AML. At our own institution, we looked at our first 405 patients that were treated. 18 of those patients, subsequently, underwent a bone marrow biopsy for persistent hematologic dysfunction. We found nine cases of MDS, seven cases of CCUS. Now, more recently, we've updated this cohort, and now we have 14 cases of MDS and 12 cases of CCUS. What also stands out to me from this data is the latency from the first cycle of lutetium until the bone marrow biopsy. It was less than a year here. Now I want to acknowledge that this was a heavily-pretreated group of patients. They were VISION-like patients who had had exposure to radiotherapy to the prostate, metastasis-directed therapy, taxanes, platinums in some situations, but that is a shorter latency that's been described classically with alkylators and Topoisomerase inhibitors. Is this the true incidence of therapy-related myeloid neoplasm? I don't think it is, and I'll tell you why.
When we looked at our own experience, we treated very few patients who had meaningful baseline cytopenias, but when we evaluated those patients at three, six, and 12 months post-treatment, there was an enrichment in grade one, grade two, and even grade three cytopenias among patients who didn't have further anti-cancer therapies. Now, it is a very small percentage of those patients that subsequently had a bone marrow biopsy to give us definitive diagnoses. Now, we talk about CH. This is a big umbrella term. I think we all recognize that and that these alterations don't carry uniform biologic risk. Now it's our view that select alterations that involve epigenetic regulators, like DNMT3A, type two, ASXL1, these are expanding slowly and may remain silent for years. We worry more about the DDR alterations. In our own series, 72% of the patients who had bone marrow biopsy had alterations within DDR genes, and PPM1D was particularly enriched. Now, the same pattern has been described in patients who are post PARP inhibitors.
Now, yesterday, Dr. Tolmeijer did a beautiful job outlining the results of therapy. She described rates of clonal hematopoiesis in 77% of patients, using the low VAF, that were treated with either lutetium or cabazitaxel in that trial. Now, what she also noted was that treatment-emergent CH occurred more commonly after lutetium, as compared to cabazitaxel, and then the preexisting clones also expanded more frequently and to a greater magnitude following treatment with lutetium. I also found it interesting that they report the same observation that we had last year. PPM1D was particularly enriched in these patients post-treatment with lutetium. In summary, it's my view that our clinical trials data is unable to provide good estimates of the rates of therapy-related myeloid neoplasms post-Pluvicto. Persistent hematologic dysfunction and therapy-related myeloid neoplasms, though, have been described. They've been described in these large and very rigorous real-world data sets, but I worry it's an underestimate.
We have emerging data that suggests that maybe these are emerging a little bit more quickly after radiopharmaceutical therapy, noting though that our patients are heavily-pretreated with other genotoxic therapies. Most importantly, I think we must frame CH as a risk modifier and emphasize the uncertainties, including how the trajectory of CH will ultimately correlate with late hematologic outcomes. So, what's needed? What's needed now is what we're doing, right? We're transitioning from these post-hoc toxicity studies to prospective cohorts. We're collecting blood before, during, and after treatment. Now, it's my view that the risk of therapy-related myeloid neoplasms can be mitigated through better patient selection and better treatment sequencing, but we've got to collect the data to do this. Our team is doing what a lot of folks in academics do, right? We take a very simple complex, and then we make it look much more complex on the screen.
Essentially, all I'm saying here is that we're considering the inputs, right? What are the prior exposures that a patient has? How much radiation have they received? What's the marrow reserve? And then you try to put patients into high-risk or low-risk, and that informs surveillance strategies down the road. We recognize that there are patients that need to be monitored more closely, including the clonal evolution over time, and I do think we need to be better about getting bone marrow biopsies to define the process that lead to persistent cytopenias. So with that, I want to say thank you, in particular to Dr. Yael Kusne. She's a leukemia doctor at Mayo Clinic, Arizona, who has led our institutional efforts. She's mentored by Mrinal Patnaik, Geoff Johnson, Matt Thorpe, wonderful colleagues in nuclear medicine. Dr. Miguel Muniz is here today, presenting a poster and Ali Tarhini. They help us with our databasing efforts at the institution, so thank you all very much.