Skip to main content
Premium Trial:

Request an Annual Quote

GWAS in Children IDs Novel Gene Locus Linked to Post-Operative Pain, Morphine Dose


NEW YORK (GenomeWeb) – Researchers have conducted a genome-wide association study in children to retrospectively try to identify genetic markers that might predispose them to having greater post-operative pain, necessitating a higher morphine dose.

The study, the first of its kind in kids, analyzed samples from more than 600 children between ages four and 18 using Illumina arrays. The children hadn't been previously treated with opioids, and had been admitted to the hospital for day surgery for tonsillectomy and adenoidectomy.

The primary outcomes of the study, led by Scott Cook-Sather, a pediatric anesthesiologist at The Children's Hospital of Philadelphia, was to identify SNPs associated with morphine dose requirements in these children. The researchers also assessed markers to gauge if they were linked to high or low post-operative pain scores.

"While not reaching significance … our collective evidence implicates the TAOK3 locus," Cook-Sather and colleagues wrote in a paper published in Pain. According to the publication, researchers concluded that SNPs in the TAOK3 locus were associated with increased morphine requirement in European Caucasian children, but not in African American children. However, the TAOK3 SNPs were associated with greater post-operative pain in both pediatric groups.

The TAOK3 site hasn't been previously associated in the literature with morphine sensitivity. However, genes in the locus encode a protein that plays a role in signal transduction of neurons that transmit pain sensations.

"The important thing about studying characteristics and phenotypes in kids is that the heritability of the genetic traits is stronger in children than in adults," Cook-Sather told PGx Reporter. "Adults accumulate various environmental influences." For example, over his life an adult patient may have used other opioids for various pain symptoms, consumed copious amounts of alcohol, or become smokers, all of which could influence his ability to respond to analgesics.

"At the DNA level, they may have methylated DNA from epigenetic phenomena [associated] with living in the world," Cook-Sather said. "Looking at children, it's perhaps easier to tease out some baseline genetic differences … [and] arguably, one can say you're coming closer to answering pure questions about pain and analgesic response without these other mitigating factors."

As part of a discovery set, researchers genotyped samples from nearly 300 European Caucasian children, analyzing 136 samples on the Illumina Human-Hap550 SNP array and 157 samples on the Illumina Human610-Quad version 1 SNP array. These tests included more than 500,000 common SNPs, which Cook-Sather and colleagues sifted through for genotype-phenotype associations.

Using approximately 75 samples from European Caucasian children, researchers replicated their findings from the discovery set with regard to two top TAOK3 SNPs — rs795484 and rs1277441 – using a TaqMan assay from Life Technologies. Meanwhile, they genotyped 250 samples from African American children.

These two TAOK3 SNPs account for 8 percent of the ten-fold variance in morphine requirement, the researchers estimated. This is comparable to what clinical factors, such as age, body mass, and overall health status, contribute to morphine dose variance.

Ultimately, the association with the TAOK3 SNPs fell just short of statistical significance. GWAS studies require several hundred to a thousand patients in order to be able to conclude with confidence that the results are not due to chance. However, Cook-Sather has confidence in the TAOK3 finding based on the strength of the signal seen in the study and SNPs in linkage disequilibrium with the TAOK3 locus.

"Imputation at the TAOK3 region identified 29 additional variants associated with morphine requirement," the study authors noted in the paper. "Significantly, minor allele variants showed clear gene dose effects and, with regard to the interrelated [high] post-operative pain outcome, minor allele frequencies were enriched in both African American and European Caucasian cohorts."

Furthermore, Cook-Sather pointed out that his team's identification of SNPs in the TAOK3 locus was supported by the replication cohort with regard to its association with morphine dose requirements in European Caucasian children, and in both European Caucasian and African American cohorts with regard to the markers' association to high pain. Lastly, because the involvement of TAOK3 in these phenotypes is "mechanistically plausible," Cook-Sather said his group "felt very confident that we could argue that this is a real signal and not one that was spurious from a small sample size."

In the study, the researchers didn't look for markers of response within any specific gene. Instead, they conducted a GWAS in order to identify genes in an unbiased way that might be significantly associated with morphine response and post-operative pain.

"TAOK3 soared above all the other genes in the 500,000 SNPs we looked at," Cook-Sather said. "And the closest gene that we know of that does regulate morphine response, mainly [OPRM1], was several orders of magnitude less significant just by statistics, than was TAOK3. And the CYP450 system did not come into play."

Cook-Sather and colleagues didn't find any CYP450 markers implicated in morphine response likely because morphine is an active drug. CYP450 2D6 polymorphisms can speed up or slow down people's ability to metabolize codeine, which is a prodrug that depends on CYP2D6 for conversion to morphine.

"You could argue that the phenotype that we looked at, namely acute post-operative pain … or pain that is experienced within one to two hours of morphine exposure may less likely be subject to metabolism genes" such as the CYP450 or glucuronosyltransferase genes, Cook-Sather said. He explained that due to the fact that morphine is an active drug, the metabolism genes could be associated with pain later in the post-operative setting, when the patient is at home, for example. However, immediately after the surgery, these types of metabolism genes aren't associated with pain, according to the latest study.

Cook-Sather noted that he was particularly proud of the primary endpoint in this study, which measured pain in terms of whether patients were comfortable enough to be released from the hospital. This, he felt, reduced the subjectivity of pain scores. "There are so many studies that look at pain at 10 minutes [after surgery], pain at 30 minutes, and pain at 24 hours," he said. "But one person's pain [that they score as] 5 out of 10 could be another person's pain [threshold] of 7 over 10."

Instead, children in this study were not sent home until they felt comfortable to do so. "I just think that is a robust clinical endpoint that others have not necessarily used," he said. Studying a pediatric population also adds to the complexity. Although teenagers in the study may have been able to score their pain accurately, that type of survey is harder to conduct on a four-year old trial participant.

"If you've never been run over by a truck, you have no idea what 10 out of 10 pain is," Cook-Sather said. "If you've never had an operation and your throat is barbequed from electrocautery, you have no idea what severe pain is." Given these challenges with using pain scores, Cook-Sather said he was confident that his team "looked at a composite phenotype that is possibly more valuable."

Cook-Sather's group will follow this study with a larger trial, looking at gene associations in a thousand patients. In the study, researchers will also investigate the pharmacokinetic parameters implicated in morphine response.

The Children's Hospital of Philadelphia's Department of Anesthesiology, Critical Care Medicine, and Center for Applied Genomics provided funds for this study.

Filed under