A recent study by researchers at Cincinnati Children's Hospital Medical Center has demonstrated that SNP microarrays can identify regions of homozygosity that aid in clinical diagnosis as well as identify instances of close parental relatedness.
The study, published Aug. 2 in Genetics in Medicine, documents the Cincinnati group's experience using several array platforms over the course of four years to calculate the likelihood of parental consanguinity based on the measurement of multiple large homozygous regions spread across multiple chromosomes.
In a paper published in Genetics in Medicine earlier this year, the same CCHMC team reported results of a survey of laboratories that offer SNP array-based constitutional testing, indicating that labs differ widely in how they calculate and report ROH results, especially if associated with a probability of consanguineous parentage (BAN 7/17/2012).
Kristen Sund, a genetic counselor at CCHMC and the recent study's first author, told BioArray News this week that the new study offers more explicit evidence of both the clinical diagnostic potential of measuring and reporting ROH as well as the ethical issues associated with this information.
"It's not going to be a surprise to people that the technology can do this. There have been several review papers, but there haven't been publications that go into detail on what groups have seen [in their] patients and what they do with the information," Sund said.
Importantly, she said, the Cincinnati group's newest paper clearly showed that ROH data can aid diagnosis. "Some people might say this is just an incidental finding, why do we look at this? … But we're not just looking at this because its fun. We are looking at this because it can help us make diagnoses," she explained.
Sund explained that single homozygous regions, or multiple large ROH on a single chromosome, can aid in diagnosis by indicating uniparental disomy. "But then even in these cases where we're finding these regions not due to UPD, but because the parents are related by blood, there is also clinical utility," she said.
In the group's recent study, Sund and her coauthors describe two cases where ROH analysis yielded candidate genes, and eventual clinical diagnoses in collaboration with the patient's physicians. She said the team has since seen a number of additional cases where looking at ROH yielded candidate genes for further clinical testing.
According to Sund, this diagnostic utility serves to highlight even more the ethical and legal implications for reporting ROH results in cases where there is evidence of parental blood relationships, she said.
"In some cases it might not matter so much," she said. "The range [of percent homozygosity] we saw was from 0.9 percent homozygosity up to 30.1 percent. Basically that tells us we see a lot of very distant parental relationships, but we also see parents who are first-degree relatives, which could be an incestuous relationship."
The group's previous paper demonstrated a wide range of approaches for reporting ROH results that could indicate close parental relationships. Some groups reported feeling a duty to report results that indicate instances of incest. Others cited a duty to report evidence of child abuse while a third set of respondents said they always communicate instances of likely consanguinity to clinicians.
Sund said the American College of Medical Genetics is currently drafting a set of recommendations on how to handle such cases, but to her knowledge, these guidelines will only deal with the legal and ethical questions of when results need to be communicated and how they should be shared with physicians, patients, or legal authorities.
The guidelines are not expected to resolve other issues, such as how laboratories should decide on ROH cutoff points and other analytical approaches to identify likely consanguineous relationships.
In the report, Sund and her team outlined ROH results culled from samples sent to their lab over three years and analyzed with three Illumina SNP microarray platforms. The researchers picked out samples with ROH greater than 10 megabases in length, occurring on more than one chromosome, as those meeting the cutoff point for consanguinity.
Of more than 3,000 SNP microarrays analyzed during the three-year period, 59, or 1.8 percent, met those criteria, with homozygosity percentages ranging from 0.9 to 30.1 percent.
According to the study, homozygosity in offspring of blood-related relationships is expected to be 25 percent, 12.5 percent, and 6.25 percent, for first-, second-, and third-degree relatives, respectively. By calculating a confidence interval for predicted inbreeding, the researchers were able to categorize the suspected degree of relationship for the parents of each subject.
For example, individuals with homozygosity exceeding 21.3 percent fell above the group's cutoff for suspected offspring of first-degree relatives in the study. Eleven of the 59 subjects met this threshold, the authors reported, while the remaining 48 represented the offspring of parents who were more distantly related.
The researchers found that in two cases of known consanguineous families, homozygosity mapping actually led to an eventual molecular diagnosis via the identification of candidate genes in homozygous regions.
In one example, the team found an ROH common to both the study proband and an affected sibling, containing a mutated gene, HSD17B4, responsible for the disease phenotype.
In a second family, the researchers identified eight regions of interest from ROH in common between two female siblings, which they used to hone in on a diagnosis that eventually led to effective treatment for the sisters' severe epileptic symptoms.
To highlight the ethical and legal implications of such findings, Sund and her colleagues compared their microarray results to clinicians' reports and patient medical charts to categorize each case as either consistent or not consistent with families' self-reported family history.
Sund said the results of this analysis highlight the fact that interpreting an ROH finding can be complicated, especially in geographical regions where consanguineous relationships are more common.
"It's going to be hard to distinguish between parents with close relationships and multiple generations of consanguinity," she said.
"I think we really need to focus on the fact that [measuring homozygosity] can be used as a clinical tool," Sund said.
While "we're still not 100 percent sure of what our role is and what the clinician's role is in communicating this to authorities when necessary or how to go about it … I think the really important thing that comes out of this paper is just that people start talking about this."
Especially important, she said, is a continuing discussion of how to best design cutoffs and analyses for the most accurate prediction of parental blood relationships.
In both the recent study and the group's previous paper, Sund and colleagues highlight several issues, including ROH size, and the use of either the whole genome or only the autozome, in calculating homozygosity percentages.
In the latest study, the group compared measures including the X chromosome against those without, writing that the inclusion of the X chromosome may increase calculations of percent homozygosity in some cases. The authors wrote that this suggests it may be "prudent" to exclude the X chromosome from calculations.
Overall, the group wrote, "we do not suggest including an estimate of parental degree of relatedness in a clinical laboratory report. Such calculations are only an estimate and cannot account for multiple generations of consanguinity."
Although ROH measures were successful in classifying some cases, not all patients fit cleanly, and many described a family history in conflict with the microarray results. Only a small number of the patients analyzed in the study had clinically reported parental consanguinity, so it was difficult to confirm the accuracy of the array-based predictions, the group concluded.
However, the authors wrote, the study clearly demonstrated that SNP microarrays have the potential to identify a "high degree of parental relatedness including potentially illegal incestuous relationships."
"What we've come to understand," Sund said, "is that our biggest responsibility occurs when one of the parents is underage or mentally incapacitated."
In the lab, she said, "we don't always have all the information we need. So my role is to speak with the clinician, and say these are some things you might want to think about and if the parent is underage or incapacitated, that’s something that might need to be reported."