NEW YORK – Mission Bio is working with researchers to conduct proof-of-concept studies in acute myeloid leukemia with its single-cell multiomics measurable residual disease (scMRD) assay to generate data that it hopes will bridge the gap between research and clinical use of its Tapestri technology.
The San Francisco-based firm recently made its single-cell multiomics analysis platform available through an early-access program with the goal to move the technology from the research space closer to the clinic. The firm has developed a clinical workflow for its MRD assay, and according to Mission Bio Chief Medical Officer Todd Druley, the assay has been thoroughly tested and is ready for the next step.
"Our sensitivity and specificity are comparable with what's being used out in the field for clinical diagnostics for people that have blood cancers," he said. "But what we don't have is any actual patient data that's clinically annotated."
The company identified AML as a promising area to develop a new MRD assay due to significant unmet clinical need. MRD testing is not as advanced in AML as it is in other blood cancers. For example, the ClonoSeq next-generation sequencing-based in vitro diagnostic assay is cleared by the US Food and Drug Administration for multiple myeloma, chronic lymphocytic leukemia, and acute B lymphoblastic leukemia. Adaptive Technologies, the test's maker, touts its ability to detect one cancer cell in a million and accurately predict relapse and overall survival.
In contrast, using flow cytometry and NGS combined — the most accurate currently available methods — about 30 percent of AML patients who test negative for MRD eventually relapse. The challenge is that AML is very difficult to distinguish from clonal hematopoiesis, a normal aging process, and generally lacks individual mutations that are definitive signs of MRD.
Another challenge with AML MRD testing is that it can only be done using a bone marrow biopsy, whereas less invasive MRD tests using peripheral blood are available for other blood cancers. Mission Bio said it hopes to adapt its AML MRD test for peripheral blood in the future, but currently the scMRD test relies on bone marrow biopsy samples.
AML or clonal hematopoiesis?
The European LeukemiaNet MRD Working Party, a group dedicated to standardizing MRD testing, recommends MRD testing in AML to assess risk of relapse and guide treatment decisions and as a potential surrogate endpoint for overall survival in drug trials.
Multiparametric flow cytometry is the most commonly used method for MRD testing in AML. Currently, flow cytometry assays have a limit of detection between 0.05 and 0.1 percent. That translates to detection of one cell out of 1,000 to 2,000. For certain subtypes of AML, such as NPM1-mutated AML and core binding factor AML, which are characterized by inversions on chromosome 16 or translocations on chromosomes 8 and 21, RT-PCR-based assays are available with sensitivity of about one cell in a million.
As another option, NGS panels are typically offered within institutions as laboratory-developed tests but are often not designed for MRD testing, said Nicholas Short, an acute leukemia specialist at MD Anderson Cancer Center. For NGS tests, Short said sensitivity is around 1 percent, which is not ideal for an MRD test, and results can be ambiguous if the prognostic significance of certain mutations is unclear.
"One challenge we have is determining, are these really leukemia or are they a pre-leukemic state, clonal hematopoiesis of indeterminate potential?" said Short. "If you have a residual TET2 mutation, is that actually the disease, or is that just a pre-leukemic state that may or may not actually predispose you to relapse?" For those reasons, most labs are now combining flow cytometry with NGS or individual tests for select driver mutations to maximize MRD detection.
"This has been the hardest malignancy to pin down testing for that is reproducible and interpretable in a way that can be widely used in practice," said John Byrd, a physician-scientist at the University of Cincinnati and chief medical officer of the Leukemia and Lymphoma Society's Beat AML master clinical trial.
Byrd pointed out that interpretation of MRD results in AML is often subjective, despite many studies attempting to refine the testing. Byrd compared distinguishing clonal hematopoiesis and AML to distinguishing light brown and medium brown hair. "With the evolution of next-generation sequencing, initially we thought this was going to be the nirvana," he said. "It was going to take us to understanding and enable [us] to see small clones of AML where this would be predictive."
One distinguishing feature, according to Byrd, is that clonal hematopoiesis typically has just one or two AML-related mutations, whereas AML will tend to have more disease-associated variants. That creates an opportunity for technologies like Mission Bio's Tapestri platform that can detect individual clones bearing multiple mutations.
A Mission Bio spokesperson said that while Tapestri's sensitivity is comparable to flow cytometry-based methods, Tapestri's ability to detect co-localized single nucleotide variants, copy number variants, chromosomal anomalies, and surface immunophenotypes "offers unparalleled granularity to monitor clonal selection before, during, and after treatment has been administered."
Predicting relapse, guiding treatment
At MD Anderson, an AML patient in remission typically receives MRD testing by flow cytometry followed by NGS. Some patients can have MRD assessed using a PCR assay.
Short said MRD results are mainly used to decide whether a patient in remission should proceed to an allogeneic stem cell transplant. "In general, regardless of what someone's baseline risk appears to be based on their cytogenetic and molecular features, if they are persistently MRD-positive, then we would recommend transplant for them," said Short.
For example, someone who is persistently MRD positive with a very favorable risk profile would be sent to transplant. Conversely, Short said an MRD-negative patient with an adverse risk profile, would also be recommended for transplant "because we know that those patients are still very likely to relapse."
"There's really no such thing as being MRD-negative," explained Byrd. "We should say MRD not detectable, because our technology can only get down to a certain threshold below which there could still be minimal residual disease, but we're just not detecting it."
The default for intermediate-risk patients is also a transplant, depending on other factors such as age, comorbidities, and donor availability. Short noted, though, that if an intermediate-risk patient rapidly achieves MRD-negative status, this might reduce the strength of transplant recommendation.
At MD Anderson, AML patients receive bone marrow biopsies every three to four months for the first one to two years following remission, and are tested for MRD every time. In community practices, patients may be tested for MRD upon completion of treatment, but typically don't have repeat bone marrow biopsies in remission until they have signs of relapse.
MRD testing can also influence therapy decisions for patients who are not transplant candidates, or who are not yet ready to do a transplant. Those patients might be directed to further therapy or a clinical trial.
Byrd recalled the case of a 53-year-old woman with AML bearing mutations in DNMT3A and NPM1. Her flow cytometry-based MRD at the end of induction chemotherapy was negative for evidence of leukemia, and NGS testing couldn't detect the two mutations. Due to risk factors including weight and comorbid illnesses, Byrd said, "we elected to give her consolidation [chemotherapy] with four courses of cytarabine rather than going to an allogeneic transplant."
While some AML patients are served well by the currently available MRD tests, many could benefit from improved assays. Short said that includes patients at intermediate risk, for whom the decision about whether to proceed to transplant is not clear, particularly those who are MRD-negative by current assays. "If you're MRD-positive, I already know you need to go to transplant," said Short. "But if you're MRD-negative [and at intermediate risk], we know many of those patients can do very well without transplant."
With its early-access program, Mission Bio now hopes to establish the clinical utility of its scMRD protocol as a method to provide more granular data for these types of clinical decisions. First, Druley said, the firm wants to determine the benefit of identifying the mutations and surface protein expression on a single-cell level. The other question the company is interested in is, "Will oncologists take action with these data?" Druley said. "Our goal is ultimately to show that patients benefit when doctors do take action from these data sooner in a patient's [treatment] course."
First steps toward proof of concept
Wenbin Xiao, an assistant attending hematopathologist at Memorial Sloan Kettering, has had an interest in improving the sensitivity and specificity of MRD assays in AML for many years. His initial idea was to sort single cells into individual wells on a plate and do a colony formation assay followed by genomic sequencing. But that would have been a "massive" undertaking, Xiao acknowledged, "considering how many cells you are going to do and how many patients you have."
Not long after, Xiao joined Ross Levine's laboratory at MSK. Levine, a blood cancer researcher at MSK, had done some studies with the Tapestri platform. "Immediately, I thought it would be an ideal platform for my research question," said Xiao, who is participating in Mission Bio's early-access program for the scMRD assay.
He and his group analyzed 30 samples from AML patients as a proof of principle for the platform. Xiao used Tapestri to combine single-cell DNA sequencing data with immunophenotype of the cells based on 42 surface markers and is preparing a manuscript for publication. In Xiao's study, the scMRD assay had a sensitivity of 0.01 percent, up to tenfold better than a comparable flow cytometry-based assay. "In the future, we would like to focus on the markers that are relevant to MRD detection and potential therapeutic targets," Xiao said.
Because that study was small, Xiao is planning a larger study that he hopes will provide more information to distinguish pre-leukemic and leukemic status and identify high-risk clones that correlate with relapse. For this, he said his team will need to longitudinally collect many patients' samples. "We plan to do a study targeting 200 or 300 patients and correlate the single-cell results at diagnosis, [remission] stage, and relapse stage," Xiao said, adding that ideally those samples would come from a clinical trial and be used for prospective analysis. If that doesn't work out, his group will collaborate with a tissue bank to obtain samples for retrospective analysis.
Although cost is a major challenge in a study like this, Xiao's group is able to reduce spending by multiplexing at least five patients into a single run. Then, the sequence data from the five patients can be separated again in the computational analysis stage. A second challenge, Xiao said, is the cell yield from the platform. "Currently, the cell yield is 10 to 20 percent," he said. "That means we are not getting information from about 80 percent of the cells. If we can increase that, it would be a huge help."
Sanam Loghavi, an MD Anderson hematopathologist and molecular pathologist, is also participating in Mission Bio's early-access program. Loghavi is analyzing samples from nine patients treated at MD Anderson and banked for future research using the Tapestri platform. She'll measure the sensitivity and specificity of the assay for detecting MRD and attempting to differentiate residual AML from clonal hematopoiesis in patients who were treated for AML.
Loghavi and colleagues aim to analyze at least a million cells from each sample. They selected samples with mutations specific to clonal hematopoiesis, including some samples that were negative by conventional MRD testing methods, some that were borderline positive for MRD, and some that were unequivocally MRD positive.
The team will compare their results against data on traditional MRD testing platforms including flow cytometry and bulk NGS. "A unique aspect of the Tapestri platform is that it is able to pair DNA mutation analysis on a single-cell level with phenotypic data and protein expression," said Loghavi, noting that this information is typically obtained separately from flow cytometry and NGS analysis.