While many in the cytogenetics world are concerned with array findings of unclear clinical significance, and how to communicate that information back to patients, a recent review conducted within Cleveland Clinic Laboratories found that it was the lab's sample preparation method that had actually led to skewed or unclear results.
Diane O'Neill, a cytogeneticist at CCL, told attendees of the Association of Genetic Technologists meeting held earlier this month in Las Vegas that the lab implemented comparative genomic hybridization-based testing for genetic abnormalities in patients last year, but, over time, experienced a decrease in the quality of the results.
That motivated the group to look into every aspect of its chromosomal microarray analysis, including DNA extraction. According to O'Neill, the lab decided to try 5-Prime's extraction method for obtaining DNA to run on arrays, and was pleased with the results, ultimately leading CCL to abandon its previous DNA extraction kit, which is manufactured by Qiagen.
Now, CCL is developing a training program for its DNA extraction lab, and is eager to share its experience with other labs adopting oligonucleotide CGH for constitutional cytogenetics.
"We are excited because nobody really talks about DNA extraction when they are troubleshooting arrays," O'Neill told BioArray News at the conference. "Maybe it's better to start from the beginning and fine tune that process before one even runs an array," she said. "Arrays are expensive," she added, "that was a problem for us before — we were having to repeat samples."
Another issue was interpreting unusual results. O'Neill said that poor DNA quality "significantly shifted probe heads," but that the alteration-calling algorithm would still call the changes, and that they sometimes made it past internal filters.
"We keep track of every notation that is made on every patient, and we started to notice this wasn't only happening with benign copy number changes," O'Neill said. "We could see there were definitely mixed artifacts with mixed oligos." She stressed that in these instances, the lab would rerun its analysis from beginning to end, and that it did not affect results reported to patients.
To overcome that, the lab not only switched extraction kits to 5-Prime from Qiagen, but also began development of a quality management program that ensures certain quality control metrics are met throughout the CMA process.
"They won't be able to deliver the DNA until they hit certain metrics," said O'Neill. "That way, by the final outcome, we are going to have nice, clean data that is interpretable," she said.
'Not a Product Issue'
Though CCL's experience would suggest that 5-Prime's extraction kit performed better than Qiagen's, O'Neill said that it's "not a product issue," and that the kit just happened to work in CCL's lab environment.
"I don't think that one kit is better than the other but for our applications, we are getting better genomic DNA with the 5-Prime," she said.
She said that CCL's DNA extraction lab initially selected Qiagen kits because it had used them to run other tests. The group also selected manual DNA extraction, as opposed to automatic, because of performance issues.
A number of issues that impacted the Qiagen kits' performance may have been operator-related variability, as well as ozone levels, which can damage DNA quality. "We moved into a brand new building with windows," said O'Neill. "No one ever would have thought that Cleveland has an ozone issue, but we do."
The switch to 5-Prime isn't the only change CCL is experiencing at the moment. It initially implemented constitutional CGH testing on the Roche NimbleGen platform, but Roche discontinued making CGH arrays last year. Now, CCL is validating constitutional CGH on Agilent Technologies' microarray platform.
O'Neill anticipated that it would begin offering constitutional CGH testing on the Agilent platform next month. She also said that the lab would like to implement array-based prenatal and oncology testing in the future.