By Monica Heger
Next-generation sequencing of dried blood spots could be a more cost-effective method for monitoring HIV drug resistance than PCR-based genotyping from plasma samples and could also enable broader access to HIV drug resistance testing, according to researchers with the Public Health Agency of Canada.
In a pilot study, published recently in Antiviral Therapy, researchers from the National HIV and Retrovirology Laboratories in Canada used next-gen sequencing on Roche's 454 GS FLX to screen for drug resistance on previously collected dried blood spot samples from 48 patients with HIV-1.
The samples were part of a prior study, published in Antiviral Therapy in 2007, that compared dried blood spots to plasma for HIV genotyping in patients who were newly diagnosed with HIV-1 and had not yet received treatment.
In the current study, the researchers wanted to test the ability of next-generation sequencing to identify drug-resistance variants from the dried blood spot samples as a lower-cost method of monitoring drug resistance. They used the GS FLX for the sequencing, and compared the results to the genotyping results and to Sanger sequencing.
The researchers used a tagged, pooled, amplicon sequencing strategy on two lanes of the GS FLX, creating three overlapping amplicons from each patient encompassing the entire HIV-1 PR gene and the first 237 codons of the RT gene.
The team found that the results of the 454 sequencing were concordant with Sanger sequencing, and even picked up a variant that Sanger sequencing failed to detect.
"Pooled 454 pyrsosequencing is an efficient and cost-effective [next-gen sequencing] method for determining HIV drug resistance prevalence using [dried blood spot] specimens," the authors concluded. "This technique can lower barriers of cost and throughput and might improve accessibility to HIV [drug resistance] surveillance" in resource-limited settings.
Dried blood spots have a number of advantages over plasma, including ease of collection and storage, said James Brooks, chief of the National Laboratory for HIV Genetics at the Public Health Agency of Canada and senior author of the study. Collecting plasma requires tourniquets, needles, and sophisticated laboratory infrastructure to then separate out the plasma from the cellular components. Dried blood spots, on the other hand, require just a method for drawing a small amount of blood and the paper on which the blood is dried and stored.
Additionally, he added, plasma must be stored at negative 80 degrees C, which can be challenging in areas with unstable power supplies. Dried blood spots are much more durable.
The genotyping study published in 2007 established the "equivalence of dried blood spots and plasma collected in field conditions in drug-naïve people," Brooks said. The next step was to find a lower-cost method of screening those samples for drug resistance.
Current costs for genotyping HIV specimens from dried blood spots run between $125 per sample for in-house methods to more than $400 per sample for commercial kits, according to the researchers, who estimated that their method could reduce those costs by 40 percent compared to in-house methods and by nearly 90 percent compared to commercial testing.
[ pagebreak ]
While sequencing with the 454 machine is more expensive than other next-gen systems, Brooks said that the read lengths from the 454 are suited for the "particular region of the HIV genome that we're interested in looking at. He added that the Ion Torrent is an "interesting technology," but its read lengths are not yet long enough. "For our application, [454 is] the winner today."
The investigators found that aside from being cheaper than conventional genotyping methods, next-gen sequencing is also better at picking up rare variants.
In the pilot study, the researchers only looked at variants present in 20 percent of more of the reads, and even at that level Sanger sequencing still missed one variant present in 20.4 percent of reads.
Depending on the depth of coverage, however, next-gen sequencing can detect very rare variants, below 1 percent frequency.
Deciding how deep to sequence and thus what level of variants to screen for is one of the next steps the team must figure out. Brooks said that the team would likely reanalyze some existing studies to evaluate the clinical significance of variants at 10 percent frequency. "We can reliably pick those up in a cost-effective way," he said.
Drawing that cutoff point is tricky, said Brooks, because while there is some evidence that rare variants predict therapeutic failure, not everyone with rare variants fails.
"The field hasn't yet generated enough data for us to have a good understanding for why people fail [drug treatments] with a minor variant present at a certain frequency and other people do not," he said.
Another requirement before sequencing of dried blood spots could be implemented clinically is to automate as much of the sample prep and bioinformatics pipeline as possible, he said.
For this step, he said that the team is working with researchers from Canada's National Microbiology Laboratory in Winnipeg, which has a bioinformatics core that will help develop software for data analysis.
Additionally, he said, the team will test the technique on patients who have already received drug treatment. There has been some concern that collecting dried blood spots might not be a good sampling method for patients who have already received therapy because the virus could change after receiving treatment.
"The current circulating virus may be different from the virus that gets archived," Brooks said. "So we don't know whether it's going to be an appropriate way of collecting specimens in the treatment-experienced population."
Ultimately, Brooks said the goal is to create a pipeline that would "take the processed output from the pyrosequencing and generate a report that will have the drug resistance pattern."
The technique could then be implemented in Canada as well as other countries. For instance, the laboratory is part of the World Health Organization's HIV drug resistance laboratory network that is accredited for testing HIV samples for drug resistance.
"Ultimately, once we've established this pipeline, we'd like to use this technique as a way of conducting both domestic HIV drug resistance surveillance and to also use it for our work that we're involved in with [WHO]," said Brooks.
Brooks said that as part of that network, the lab receives requests from other countries — often low- or middle-income nations — to test samples. Typically the lab sponsors the costs of these genotyping tests, "but if we can do this in a more cost-effective manner, then we can deliver this service to a greater number of clients," he said.
The same technique could be applied to other diseases as well, such as hepatitis. Brooks said the he is currently conducting a surveillance of hepatitis C patients with the Public Health Agency of Canada, doing genotyping of dried blood spots. "There's no reason why we couldn't use the same [sequencing] technology to lower the costs of genotyping," he said.
Have topics you'd like to see covered by Clinical Sequencing New? Contact the editor at mheger [at] genomeweb [.] com.