NEW YORK (GenomeWeb) – Members of the International Genetics and Translational Research in Transplantation Network (iGeneTrain) are banding together to apply genomics to transplantation studies in the hopes of better understanding and improving transplant patient outcomes.
In a paper in press in the journal Transplantation, the team described the goals and progress made so far for the iGeneTrain effort. When the paper was submitted, that work included 22 genetic studies, though the tally now comes in closer to 30 studies, according to information on the iGeneTrain web site.
Through genome-wide association studies involving more than 32,000 organ donors or recipients of kidney, liver, heart, or lung transplants, the researchers are searching for genetic factors contributing to both short-term transplantation success or failure as well as longer-term complications, including side effects related to the immunosuppressors transplant recipients receive to stave off rejection.
The group eventually hopes to generate multi-omics profiles for at least a subset of those samples, explained iGeneTrain leader Brendan Keating, a transplantation researcher at the University of Pennsylvania.
"This is an area that has been very lacking in genomic studies," he told GenomeWeb. "We're hoping that this helps to get this field moving."
Organ transplant recipients may experience a host of complications — from immune rejection of the newly acquired organ to problems such as diabetes, elevated cholesterol, high blood pressure, or skin cancer that arise as a result of immunosuppression therapy.
"Rejection is more important in certain organs than in others," Keating noted, "but a lot of [other] complications are starting to become a larger source of morbidity than rejection."
"Some of these events don't occur as frequently as they did in the past when we didn't have good immunosuppression," added iGeneTrain investigator Ajay Israni, a nephrology physician and clinical researcher at the University of Minnesota's Hennepin County Medical Center.
Even so, Israni explained, by bringing together large sets of organ recipients and donors, it becomes possible to parse out patterns involving genetic factors that influence patient outcomes, immunosuppression responses, and risk of short- or long-term complications.
At the moment, the iGeneTrain team is focusing much of its attention on transplant features such as graft survival, acute rejection, and post-transplant diabetes onset. In kidney transplant recipients, the group is also looking at delayed graft function.
Israni is optimistic that the studies will make it possible to apply precision medicine in a transplant setting. "If I know that this is a higher risk donor-recipient combination, based on their genetic information I can tailor their immunosuppression accordingly."
The consortium has been in the works for around three years already, Keating noted, during which time team members not only designed a custom chip but also started bringing together existing genotyping data from transplant donors and recipients with documented outcomes.
Keating, Israni, and their colleagues outlined the rationale, features, and performance of the custom Affymetrix array — nicknamed the Tx GWAS array — in another paper published online today in Genome Medicine, where they applied the chip to reference samples from the HapMap and 1000 Genomes projects.
The array includes 780,000 variants peppered across the entire genome, though team members made a special effort to get deep coverage of immune-related sites associated with transplant outcomes in the past — particularly human leukocyte antigen and natural killer cell immunoglobulin-like receptor (KIR) loci — as well as sites implicated in pharmacogenomic associations involving immunosuppression therapies and SNPs tagging rare variants.
The researchers have also imputed genotypes for around 10,000 donor and recipient samples that were not tested using the custom Tx GWAS array, Keating noted.
With the help of data generated for the 1000 Genomes and the Genome of the Netherlands projects, they inferred variant patterns at roughly 88 million sites genome-wide in those samples, including 15 to 16 million high-confidence variants.
That data has been included in an iGeneTrain data freeze. Across all of the organ transplant types, the group plans to look first at variants involved in time to first biopsy-confirmed rejection.
The researchers are in the process of attempting to replicate potential associations between loci outside of the HLA region with features such as time to organ rejection across multiple transplant types, for instance.
Keating noted that such analyses are complicated by non-genetic factors that have been implicated in transplant outcomes — from the quality of the organ the recipient receives to time spent waiting for an appropriate organ.
"There are a lot of very complex variables and that's one of the reasons we needed huge numbers [of recipients and donors]," he said.
In parallel with cross-organ analyses of side effects and complications, members of the team are considering genetic factors that might impact outcomes within specific transplant types.
Among the cross-organ analyses, they are on the hunt for loss-of-function variants that differ between donors and recipients and might contribute to immune reactions to the transplanted organs, for example.
In their paper describing the Tx GWAS array, the researchers presented a pipeline for finding such loss-of-function variants, with the help of data from the Genotype-Tissue Expression (GTEx) project, the Exome Aggregation Consortium, and past studies of loss-of-function mutation patterns in the average human genome.
The initial set of iGeneTrain studies focused on heart, liver, lung, and kidney transplants, which are currently the most common transplanted organs.
The majority of the samples tested so far — around two-thirds — came from kidney transplant donors or recipients, Keating noted, though the group is continuing to add information on thousands of new donor-recipient pairs.
Most, but not all, of the studies have hinged on samples from both transplant recipients and the donor who they received the organ from. Keating noted that there "were some consent issues at some of sites with accessing [transplant] donor DNA."
Though a large proportion of the study's participants so far are Caucasian, the researchers noted that there are still a relatively large number of individuals from other ancestral backgrounds, given the size of the overall donor and recipient pool being tested.
For the most part, the team is analyzing donor-transplant samples using the custom genotyping array or corresponding variants imputed from other chips, though it has collaborated with BGI and Roche to do exome sequencing on around 500 participants, targeting protein-coding regions and, in some cases, untranslated regions.
Investigators involved in iGeneTrain have also used strategies such as metabolomics or RNA sequencing to uncover urine metabolites or transcripts related to acute rejection in kidney transplant recipients.
There are several more transplantation-related phenotypes of interest such as delayed graft function, cancer recurrence, and so on that the team hopes to investigate in the future.
Members of the group are also keen to delve into pharmacogenomics patterns associated with immunosuppression therapy response as well as secondary complications caused by such treatment, including skin cancer.
The researchers emphasized that the iGeneTrain group is open to new members, provided potential collaborators are willing to follow the consortium's imputation pipeline and/or analyses with the custom transplant genotyping.
Generally speaking, members of the consortium would like to have at least 12 months of clinical follow-up data for transplant recipients included in their organ rejection-related studies, though longer follow-up times are preferable, particularly for kidney transplant studies where chronic rejection may occur.
The team is also hoping to collaborate with organ transplant centers and/or organ procurement centers in the US and beyond to tap into available DNA-containing donor and recipient samples as well as the long-term transplant outcome information accrued for the recipients.
"A number of outcomes from those recipients are on federal databases," Keating said. "So there's deep phenotypic data — outcomes data — on these people."
Researchers at the University of Pennsylvania secured institutional review board approval to link donor and recipient data from a local organ procurement center, he noted. "For us to be able to do that at a number of different sites across the US or worldwide — we think that's a very powerful model to quickly get at long term outcomes, especially when the data already exists in a CLIA-type environment."
Funding for the iGeneTrain effort comes from a variety of sources, ranging from U01 and R01 grants that team members have received from the National Institutes of Health to discretionary funding from transplant groups.
To date, the researchers estimate that more than $30 million has been invested in the collection, phenotyping, and genotyping of samples in the datasets.