Clinicians have long known that about 30% of acute myeloid leukemia (AML) patients in what appears to be complete remission will die within three years of treatment. Observational research combining CIBMTR® clinical data with matching biospecimens from multiple repositories has worked out potential mechanisms and suggested at least one novel therapeutic approach for improved outcomes.

“We’ve known the data for 50 years,” said  Christopher Hourigan, DM, DPhil, FRCP, professor and director of the Fralin Biomedical Research Institute Cancer Research Center, Virginia Tech. “There just hasn’t been a way to operationalize it.”

Hourigan was one of the featured speakers in the Feb. 12 session Concurrent: Driving Translational Research Supported by Linkage of CIBMTR Clinical Data and Biospecimens [CIBMTR].

Observational research using patient registries has identified three genomic drivers of AML relapse: FLT3 internal tandem duplication (FLT3-ITD), mutated NPM1, or co-mutations. The Molecular Evaluation of AML Patients After Stem Cell Transplant to Understand Relapse Events (MEASURE) study is designed to transform associations linking those mutations to poor survival following hematopoietic cell transplantation (HCT) into actionable prospective evidence.

Hourigan noted that early registry studies confirmed that patients with AML can be risk stratified using minimal residual disease (MRD) testing. The CIBMTR biobank has long collected blood samples before allogeneic HCT (alloHCT) with data collected on FLT3, NPM1, IDH1/2, and KIT mutation status with relapse and survival data.

Early studies found that FLT3-ITD and/or NPMI mutations are strongly associated with relapse and death with no safe level of persistence. Current evidence supports the use of single-gene testing for FLT3-ITD and mutated NPMI at a single time point before alloHCT.

MEASURE will explore a novel multi-target, patient-personalized, repeat testing strategy at multiple centers across the United States. Initial data is expected during the summer of 2025.

“Moving our field forward requires cooperation, carefully annotated biospecimens, and robust audited clinical datasets,” Hourigan said. “CIBMTR infrastructure is an on-ramp to clinically impactful translational research.”

Biosamples contain multiple molecular predictors of HCT outcomes. Markers of cellular aging, germline variation, and somatic alterations may all play roles in refining HCT patient risk stratification and donor selection to improve outcomes.

Age is a clear factor in HCT outcomes, said Stephen Spellman, MBS, vice president of research at NMDP℠ and senior scientific director for CIBMTR. The HCT patient population is aging with more than 25% of recipients age 65 or older today. Older patients exhibit a number of aging phenotypes that can adversely affect outcomes.

“Donor age also makes a difference,” he said.

HCT outcomes tend to be better when donors are under 30 years versus 30 years and older. Older donor age adds an additional element of risk to HCT.

Telomere length may be a key element. Telomere shorten with each cell division, and cells with shorter telomeres can be less successful. A study in aplastic anemia found that transplantation using cells with longer telomere length improved outcomes by about 20% versus shorter telomere length. In malignant diseases, post-HCT telomere shortening reduced two-year survival by half.

Germline genetics and somatic alterations can affect outcomes. Severe aplastic anemia (SAA) comes in two forms, inherited and acquired. Patients with inherited causes do not respond to immunosuppression and need special HCT regimens as a first line of therapy. In acquired SAA, only young patients with a matched sibling donor receive first-line HCT.

Germline genetic testing is appropriate for all patients with SAA, Spellman said. Unrecognized carriers have double the risk of mortality compared to recognized carriers.

Biobanks have also played a role in characterizing mechanisms that underly the severity of SARS-CoV-2 infection. Researchers began looking at human leukocyte antigen (HLA) variations as key determinants of immune response soon after the COVID-19 pandemic appeared.

“The problem is that most (COVID-19) infections are mild or asymptomatic,” said Jill Hollenbach, PhD, MPH, professor of neurology, University of California, San Francisco. “Most studies focus on extreme phenotypes and worse disease with worse outcomes. We wanted to look at asymptomatic disease.”

Multiple studies identified protective and risk alleles, but it was not clear which alleles played leading roles. The prospective HLA/COVID-19 Citizen Science Project enrolled about 30,000 individuals with existing HLA data at NMDP who monitored and reported their COVID-19 symptoms using a phone app. Virtually all (99.9%) were never hospitalized. Based on their documented HLA status pre-COVID, HLA-B*15:01 emerged as a key protective allele.

“We had a strong dose effect,” Hollenbach said. “Those with two copies were more than eight times more likely to have no symptoms, and those with one copy 2.4 times more likely to be asymptomatic.”

The B*15:01 epitope results were replicated in independent datasets, Hollenbach said. B*15:01-positive biosamples taken years, sometime decades, before the emergence of SARS-CoV-2 showed the same response. Memory CD8+ T cells recognized SARS-CoV-2 thanks to a similar domain on the spike protein found on seasonal coronaviruses that cause the common cold. A single B*15:01 protein, NQKLIANQP, allows similar binding to both the SARS-Cov-2 and seasonal CoV peptides.

This and other sessions at the 2025 Tandem Meetings | Transplantation & Cellular Therapy Meetings of ASTCT® and CIBMTR® is available for on-demand viewing for registered attendees.