Harnessing the power of genetics to improve transplant safety

Science Spotlight

Harnessing the power of genetics to improve transplant safety

From the Petersdorf Group, Clinical Research Division

July 16, 2018

Transplantation of organs, tissues or cells has revolutionized treatment of many diseases. For example, bone marrow transplantation was developed at Fred Hutch (then a division of the Pacific Northwest Research Foundation) in the 1960s and is now performed approximately 20,000 times per year in the United States (1) to treat blood cancers and other blood disorders. However, transplant-associated mortality remains significant because donor tissue often mounts an immune response against the recipient, or vice versa. Physicians make every effort to pair each transplant recipient with a compatible donor, but this is not always possible and the determinants of compatibility are only partially understood.

The human leukocyte antigen (HLA) genes, which encode proteins that display potentially foreign molecules to the immune system, are the primary metric used to assess compatibility; a donor-recipient pair is said to be HLA-matched if they have similar alleles at certain HLA genes on chromosome 6. However, this practice does not capture the full breadth of genetic variability that might influence transplant success. In addition, “clinical experience shows that many patients who receive transplants from HLA mismatched donors do have successful transplants, indicating that in certain cases mismatches are well-tolerated,” explains Dr. Effie Petersdorf, a member of the Clinical Research Division. Thus, a deeper understanding of which HLA mismatches are associated with increased risk to patients will help to improve transplant outcomes through avoidance of those mismatches, while understanding which mismatches do not increase risk will help physicians broaden the use of selected mismatched donors and thereby improve availability of blood and marrow transplantation to patients in need.

With the goal of improving outcomes for transplant patients, Dr. Petersdorf and her colleagues in Clinical Biostatistics study how genetic factors influence transplant success. In previous work, the researchers hypothesized that undetected genetic variation within a highly polymorphic region of the genome known as the major histocompatibility complex (MHC), which contains the HLA loci and many other immune system-related genes, could contribute to risks after transplantation. To test this idea, Dr. Petersdorf’s team collected data from over 2,500 transplants in which donors and recipients had at least one HLA mismatch. The researchers screened patients and their donors for over 1,000 single nucleotide polymorphisms (SNPs) within the MHC and asked whether any SNP genotypes or SNP mismatches were associated with a higher or lower risk of transplant failure. This analysis, which was published in the journal Blood in 2013, yielded twelve SNPs that are candidate determinants of transplant success.

In follow-up work published last month in the Journal of Clinical Oncology, the researchers sought to validate their findings in an independent cohort of 1,500 HLA-mismatched transplant patients. This analysis confirmed that a recipient’s genotype at a SNP known as rs429916 is significantly associated with transplant-associated mortality; patients carrying an A nucleotide at rs429916 had significantly worse outcomes compared to those with a C. However, SNP analysis on its own does not identify causal gene(s); it only helps researchers narrow their search to a particular region of the genome. Thus, Dr. Petersdorf and her colleagues set out to determine which gene near the rs429916 SNP is responsible for influencing transplant outcomes. By looking for linkage disequilibrium, or a non-random association between particular alleles, the authors found that rs429916A patients tend to also have an allele called DOA*01:01:05 at the nearby HLA-DOA locus. However, the DOA*01:01:05 allele did not alter the sequence or expression of the HLA-DOA protein and is thus unlikely to influence transplant-associated mortality.

Partial amino acid sequence of HLA-DRB

Patients whose genomes encode the FEY motif at residues 26-28-30 of the HLA-DRβ protein have a significantly elevated risk of mortality following an HLA-mismatched transplant.

Image provided by Dr. Effie Petersdorf

Further linkage disequilibrium analysis showed that DOA*01:01:05-rs429916A-positive patients tended to have alleles of HLA-DRB1 that carry the amino acids FEY at residues 26-28-30 of the HLA-DRβ protein. Importantly, the authors found that patients with the FEY sequence were at a significantly higher risk for transplant-associated mortality. Because amino acids 26-30 are located in the peptide-binding groove of HLA-DRβ, differences at these positions may affect the peptide repertoire that can be presented by this protein and thus influence immune responses. Determining how this variation influences transplant outcomes will be an important avenue of future research.

The finding that certain motifs at residues 26-28-30 of HLA-DRβ increase the risk of mortality after an HLA-mismatched transplant may be useful in the clinic for identifying high-risk patients and offering them personalized transplant options. HLA-DRB1 alleles with FEY are much more common in African American and Hispanic American populations compared to Caucasians and Asian Americans, but the deleterious effect of this allele was consistent across all ancestries. Says Dr. Petersdorf: “this information will help clinicians assess the risks to their patients and better understand possible reasons why the success of transplantation may depend on the patient’s ancestral background.”

Petersdorf EW, Stevenson P, Malkki M, Strong RK, Spellman SR, Haagenson MD, Horowitz MM, Gooley T and Wang T. (2018) Patient HLA Germline Variation and Transplant Survivorship. Journal of Clinical Oncology.

This work was supported by the National Institutes of Health and the Office of Naval Research.

(1) Center for International Blood and Marrow Transplant Research (CIBMTR)