NEW YORK (GenomeWeb) – More than 4.5 million people need blood every year in the US and Canada, according to Blood Centers of the Pacific. And while most people know just a handful of blood types — A, B, O, Rh positive and Rh negative — there are actually a couple hundred blood group antigens on everyone's red blood cells, making accurate blood group tying prior to transfusion critical.
Currently, standard methods rely on SNP genotyping or serological typing, which involves identifying the specific antigens present. But now, researchers at the University of Washington and Bloodworks Northwest are looking to use next-generation sequencing methods to create higher resolution test.
During a presentation at the Advances in Genome Biology and Technology meeting in Marco Island, Fla., last month, Jason Underwood, who leads technology development at the University of Washington's Northwest Genomics Center, said that he and his collaborators at Bloodworks NW have designed a 42-gene panel that they are testing on around 1,100 patients, comparing the NGS approach to a SNP chip that evaluates 23 known variations.
The goal of the panel is to create a better method for blood typing individuals that undergo blood transfusions. Individuals with diseases like sickle cell, thalassemias, and some cancer have to undergo multiple transfusions in their life and are at an increased risk of developing allosensitization, particularly if the match is not perfect.
Although a SNP genotyping chip — the Precise Type HEA assay, developed by Immucor — received US Food and Drug Administration clearance last year, such approaches may not pick up variants that are specific to non-European populations.
And, although such chips offer advantages over serological typing, they are biased toward the population they are built on, Underwood said. Most chips are based predominantly on European genomes, which can cause problems when evaluating individuals from other ancestries, since blood group mismatches are more likely to occur between individuals of different ethnicities.
For instance, African Americans are disproportionately impacted by sickle cell disease, and thus require many blood transfusions. The standard chip approach is not always good enough, since it cannot take into account population-specific variants.
"When you transfuse someone, you expose them to red blood cells and the patient's immune system can have a response that can cause negative reactions," Meghan Delaney, the medical director of Red Cell Genomics at Bloodworks NW, told GenomeWeb. Such reactions can cause serious problems, and in rare cases be fatal, and also make it harder to find a blood donor for the patient in the future.
In a study of over 9,000 samples that looked at SNP genotyping compared to serology typing, the UW and Bloodworks NW researchers found that there was a substantial discrepancy, particularly in individuals that were not of European descent. For individuals from Asian or African descent, the SNP chip disagreed with the serology-based typing approximately 4.7 percent of the time.
Delaney added that using chips designed largely on European donors will not perform as well for rare blood types on individuals of different ancestry.
The team's 42-gene panel includes all exons, 2,000 bases upstream and 1,000 bases downstream of the exons, and selected introns. The group is running all samples on the Illumina HiSeq using 100-base paired end reads, and running 96 samples on the MiSeq with 250-base paired end reads. The team is also testing two different library prep protocols with different insert sizes.
From the first 469 samples of the 1,100-patient cohort, Underwood said his team has already identified cases where sequencing was able to explain discrepancies seen when comparing the SNP genotyping technique to serology typing.
For instance, he said, looking at what is known as the Kidd Blood Group, the gene SLC14A1, there are two known common antigens — Jka and Jkb. There were some cases where the SNP chip typed the individual as positive, but typing by serology did not find the expected antigens.
"This is important because if a person does not have a particular antigen, you don't want to give them blood with that antigen," Underwood said. However, the NGS approach was able to explain the discrepancy.
In many cases, Underwood said, the NGS approach revealed that the person had a loss-of-function mutation, so that the gene did not produce the protein, an anomaly the SNP chip wouldn't have picked up. In addition, he said, some of those loss-of-function variants seemed to be enriched in individuals of Asian descent.
Another area where the researchers found discrepancies between the SNP genotyping test and serology typing was in the RHCE gene. The RH region is highly polymorphic and there are over 500 alleles, some of them structural variants, which are undetectable by the SNP chip. Underwood said that the sequencing approach has been able to call the deletions that contribute to loss of function and slight differences between RHD and RHCE, which are around 94 percent homologous.
Jill Johnsen, an assistant member of Bloodworks NW, said that sequencing-based approaches would likely not immediately replace but could "initially be a powerful molecular tool to complement" existing methods. "Initially, the real value of this will be to solve the difficult cases," she said, for instance, when SNP chip typing and serological typing are discrepant.
In the future though, as sequencing costs come down, it would be a "wonderful way to do high-resolution blood typing" for patients that receive multiple transfusions and are at a high risk of having adverse events. Already, for such "challenging cases, their workups can be quite expensive, so the NGS panel may be justified," Johnsen said.
"There is a lot of diversity of blood group antigens in the world," Underwood said, "so making a more universal panel would be a benefit" and improve the success rate of transfusions.
Underwood added that a cost-benefit analysis of the panel has not yet been done. However, one reason for keeping the panel small is to keep cost down. He said the team does not yet have a commercialization strategy or plans to bring the test through FDA clearance.