BALTIMORE – Using nanopore sequencing, researchers at Columbia University Irving Medical Center and their collaborators have developed a rapid and low-cost assay to detect fetal aneuploidy in a variety of reproductive samples.
Described in a study published in the New England Journal of Medicine earlier this month, the test, named short-read transpore rapid karyotyping (STORK), can identify aneuploidy across the genome within hours while only costing $200 or less per sample. Additionally, the relatively low setup cost for the test makes it suitable to be deployed in point-of-care settings.
"Across all of reproductive medicine, there is this disaster situation of aneuploidy, where you have extra or missing chromosomes," said Zev Williams, division chief of reproductive endocrinology and infertility at Columbia University and the lead investigator of the study. Because aneuploidy is by far the most common cause of miscarriage, being able to quickly and inexpensively detect and diagnose the defect has important implications across reproductive care, he added.
According to Williams, the traditional aneuploidy testing method is G-banded karyotyping, where cells taken during prenatal screening procedures are cultured and then processed for observation under the microscope. That method, however, can take weeks to derive a result and requires living cells.
Meanwhile, fluorescence in situ hybridization (FISH), a molecular cytogenetic technique, offers a more rapid way to detect certain chromosomal abnormalities, but the downside of this method is that it is not genomewide and can only target a limited number of chromosomes.
Besides cytogenetics, Williams said there are also DNA-based methods, such as chromosomal microarrays and next-generation sequencing, for aneuploidy detection. However, these methods, as with karyotyping and FISH, typically need to be carried out in high-complexity testing labs that have made significant upfront investments in equipment.
"All of these methods require you to take a sample and send it to a centralized reference lab," said Williams. "We realized that if you can get something that was quick and low cost, that would really be a game changer for patients."
Toward that goal, the team sought to develop an assay using nanopore sequencing, which offers, with the Oxford Nanopore Technologies MinIon, a low upfront instrument cost. To boost the test's speed, the team looked to perform short-read sequencing.
"For our application, we actually don't need long reads; long reads actually waste time," Williams explained. "The way our method works is, you can think of it almost as a counting approach, rather than a sequencing approach. We just need to know the identity of that little fragment of DNA."
That said, Williams noted that the biggest technical bottleneck for the team to overcome during method development was to adapt the nanopore sequencing platform, which is optimized for long-read sequencing, for short DNA fragments.
"Nanopore-based sequencing was developed for long-read sequencing," he said. "Interestingly, the hard part is not making small DNA, it's what to do with that DNA to allow the nanopore sequencer to be able to read it." To address the bottleneck, Williams said the team had to design ways to make nanopore sequencing read the short DNA fragments "much more efficiently." One of the strategies, for instance, was to load the flow cell with around 0.4 pmol short-read library, which generated 30 percent more sequencing data than the manufacturer’s recommended loading condition.
For their study, Williams and his team applied STORK to 218 blinded reproductive samples that included products of conception after spontaneous pregnancy loss, tissue from chorionic villus sampling, amniotic fluid from amniocentesis, and trophectoderm biopsy specimens from embryos undergoing preimplantation genetic testing. After genomic DNA extraction and short-read library preparation, the samples were sequenced using the Oxford Nanopore MinIon flow cell, and the results were analyzed in real time using designated algorithms.
In general, the data showed that STORK results were highly concordant with those from standard clinical testing across the sample types. In terms of speed and cost, the study showed that, when testing a single sample, the sequencing time can be condensed to 10 minutes at a cost of $200. When 10 samples are multiplexed and sequenced simultaneously, the sequencing time was two hours, and the cost was less than $50 per sample.
Regarding potential concerns about the accuracy of nanopore sequencing, Williams said that because STORK uses a counting-like approach, the sequences only need to be accurate enough to show where that read aligns in the genome. "We tried to take advantage of the strength of the technology, which is its speed, its low cost, its portability, and its ease of use, but not have to use its limitations in terms of accuracy," he said.
The method is "very impressive," said Aleksandar Rajkovic, an Ob/Gyn professor at the University of California, San Francisco and the chief genomics officer of UCSF Health. "Whenever you shorten the time from sample collection to diagnosis, you are making an improvement."
According to Rajkovic, karyotyping and chromosomal microarray analysis, which are currently widely used to detect aneuploidy, usually take between seven and 14 days to accomplish. Karyotyping, he said, tends not to provide high resolution for prenatal samples, and it may miss certain cell types due to the culturing step. "Whenever you grow cells, you may lose certain populations of cells," Rajkovic said. "You could have a mixture of normal and abnormal cells, and the abnormal cells may not grow as well."
Rajkovic also thinks that in the prenatal setting, the improved testing speed can help women or couples alleviate their worry about aneuploidy in their pregnancy. "If you can provide them with reassurance that everything is fine at least on the chromosomal level," he said, "it can be very reassuring, and it can decrease that anxiety in the couples." In terms of providing actionable information, faster test results can allow patients to decide more quickly how they want to manage their pregnancy, he added.
One limitation of STORK, Rajkovic noted, is that because it is based on counting, it does not provide information about single nucleotide polymorphisms (SNP), which is important for diagnosing uniparental disomy or regions of homozygosity. Similarly, he would like to see how good STORK can detect microdeletions and microduplications, which he said is not clear from the paper.
Commenting on the new method's cost, Rajkovic thinks it remains to be seen how it will be translated to the expenses relayed to patients. "Whenever you put something in the clinical setting, you have to pay for bioinformatics, you have to pay for technicians, and that, of course, increases the costs," Rajkovic said. "The question is not how much it really costs, but how much you are going to charge the patient."
Beyond that, he said, the biggest question is whether insurance will cover this test. "The problem in prenatal testing is that insurance companies do not always reimburse prenatal testing very well," he explained. "Chromosomal microarrays are still not very well reimbursed; whole-exome and whole-genome sequencing are not reimbursed in the prenatal setting because insurance companies don't think that the information is going to change the course of the pregnancy."
To that, Williams said the goal for the team is to bring the cost of the test further down. "I think we should be able to offer [the test] for under $200," he said, adding that hopefully, the reduced cost can allow patients and clinicians to get answers without having to be beholden to insurance companies.
"By running a rapid test that confirms aneuploidy, the majority of the cases will have a diagnosis," said Svetlana Madjunkova, director of reproductive genetics at the Canadian Reproductive Assisted Technology (CReATe) Fertility Centre. "Then you can take your time to do proper comprehensive testing for the patients that need further evaluation."
In 2020, Madjunkova and her colleagues also published a study in the New England Journal of Medicine, where her team was able to detect structural rearrangements and aneuploidy concurrently in embryos using nanopore sequencing with the MinIon device.
According to Madjunkova, whose lab has been using nanopore sequencing since 2016, the difference between the assay developed by her group and STORK is that the former uses nanopore long-read sequencing, which can also determine balanced translocations and rearrangements in embryos. In terms of turnaround time, Madjunkova said her assay can be completed in several hours, depending on the number of samples, while STORK can finish sequencing in as little as 10 minutes.
While the CReATe Fertility Centre currently offers a nanopore sequencing-based clinical test to select patients to help diagnose structural rearrangements, Madjunkova said that for aneuploidy testing, the facility primary replies on low-pass whole-genome sequencing on the Illumina platform to meet the clinic's high throughput of typically more than 1,000 embryos a week.
In the case of Illumina sequencing, she said the turnaround time from library preparation to sequencing result is around 24 hours, plus a few more hours for bioinformatic analysis. The cost for the Illumina-based assay, Madjunkova said, is around $200 per sample.
Mirroring the Columbia researchers' point, Madjunkova thinks the main advantages of the nanopore-based assays are that they require lower upfront capital investment, which makes the technology more accessible for smaller labs, and that it is relatively easy to use.
Therefore, she thinks nanopore sequencing-based aneuploidy tests can "fill in the gaps" for labs that do not have easy access to centralized reference labs for more comprehensive tests. Madjunkova also said the high speed of these assays also makes them more suitable to be deployed in point-of-care settings. In addition, she thinks that the fast test turnaround can also help patients who experienced a miscarriage receive closure, alleviating some of the self-blame or hopelessness.
Although Madjunkova considers the fast-evolving nature of nanopore sequencing exciting, she also noted it poses somewhat of a hurdle for clinical adoption, which often requires a locked-in workflow. "I can attest that it is challenging, [especially] if you want to implement all the new things that become available," she said.
However, Madjunkova said that as with any clinical test, evolving or changing the protocols is something that can cause additional work for a clinical lab, but it's necessary. "My philosophy is that we will only move or change clinical workflows at the moment we see way better returns for the change," she added.
Meanwhile, at Columbia, Williams said his team is gearing up to offer STORK as a diagnostic test to patients. His team has filed a patent for the test and has submitted the assay to New York state for clinical approval.
"As soon as it gets approved, we're excited to be able to offer that to patients," he said.