Cell-free DNA trisomy 21 tests kick ass

NIPT for Down's Syndrome and other chromosomal abnormalities is taking off. A colleague of mine recently had an Ariosa test, paid for privately, and reported real satisfaction with the process. Lin Chitty at UCL Institute of Child Health and Great Ormond Street Hospital recently reported on a model-based analysis of NHS costs and outcomes of NIPT for Down's syndrome. This suggested that NIPT was cost effective if offered at around £50 per test (compare this to the £500-£1000 privately). NIPT is not my area of expertise but I've been watching it as technological developments have often been a little in advance of cell-free DNA work in cancer.

Possibly millions of tests have now been performed. NIPT is being rolled out to patients across the globe at an amazing rate compared to the introduction of other diagnostic tests, and the NHS is getting in on the game. The number of companies offering tests is growing and so are the litigation's. Most recently Illumina filed a new patent infringement suit against Ariosa claiming their Harmony NIPT test infringes a patent for “Multiplex Nucleic Acid Reactions” (one of the patent holders is ex-Illumina, and was an author on the Ariosa paper discussed below). NIPT commonly tests for trisomy 21 (Down’s Syndrome), trisomy 18 (Edwards’ Syndrome) and trisomy 13 (Patau’s Syndrome) and most tests are NGS based, Ariosa's test is array based. You can get an NGS-based NIPT test from ThisIsMy for just £320, tests in North America are as low as £200. Tests are available from: Ariosa Harmony, BGI NIFTY, Genesis Genetics Serenity (Verifi), Illumina Verifi, Natera Panorama, Premaitha IONA, Sequenom MATERNIT21. 


NIPT can also detect cancer: It is still quite early days, and there are inevitable headlines when NIPT or PGD go wrong (HFEA and Daily Mail); but there has been lots of coverage recently on the reporting of cancer diagnoses from NIPT testing, initially from Sequenom and most recently from Illumina (Bianchi et al JAMA 2015). Detecting cancer is low on the list of expected outcomes, abnormal results are usually put down to a foetus positive for one of the syndromes being tested for; if this is not the case then an abnormal placenta, a lost twin and even abnormal sex chromosomes in the mother come higher up the list of possibilities than cancer! The 10 cases reported in the JAMA paper suggest a lower cancer incidence than might be expected in a similar control group. However this is not surprising given that the current NIPT tests are using low-coverage WGS so will miss smaller amplifications and deletions. A “simple” test of this would be to sequence these 125,000 cases to higher coverage and see if incidence goes up (anyone know how many reads are being used in Verifi or similar?). The case of Danielle Bryant, one of patients reported in the JAMA study and interviewed by GenomeWeb and MIT Technology Review, tells of an amazing clinical journey with a positive outcome – hopefully Illumina will continue to monitor her cancer using a ctDNA assay. Other NIPT providers like Ariosa and Natera are likely to have to face up to how this kind of data will be dealt with

Arrays versus NGS for NIPT: The purchase of Ariosa by Roche for $625M is just part of their recent spending spree on biotech. Last year Roche abandoned a bid for Illumina not moving past $55 per share (isn't hindsight wonderful). A paper last year from Ariosa describes their array-based test. This is based on the DANSR assay, developed to target 100s of loci on Chr18 and Chr21 specific regions, ostensibly to reduce the amount of sequencing required in an NIPT test; and reported to require as few as 400,000 reads to generate high-quality T21 calls. The Ariosa paper argues strongly that their array-based test is faster and more accurate – bold claims, but I don’t think they really stand up to scrutiny.

They suggest that WGS of cfDNA generates data indiscriminately from regions of the genome with "little clinical utility for aneuploidy screening". Given the possible interest in screening for cancers one could argue exactly the opposite, and if you want a test that can do more than simply screen for T15/T18/T21 then ‘indiscriminate’ starts to morph into ‘unbiased’.

The more extraordinary claim is the speed of the array-based test versus NGS. They suggest that NIPT can be made simpler and TAT can be brought down by use of the Affymetrix Gene-Titan platform, stating “NGS…involves complex and costly hardware”, but Gene-Titan is anything but simple and cheap, it’s a high-throughput array scanner best run in batches of 96, and the paper reports use of a 384 sample array! A single end 36bp can be run in 7 hours on HiSeq 2500 in rapid mode, add in clustering time and library prep using Nextera XT and a 24hour TAT is achievable; assuming 10M reads per sample then 60 patients could be processed per rapid-run. It seems crazy to suggest that because an array can be scanned in one minute you’ll see lower capital costs (unless your trying to justify this decision to shareholders by pointing to a peer-reviewed article).

The authors also devote a section of the paper to the benefits of not needing to multiplex samples for analysis. Firstly multiplexing on a HiSeq lane is almost identical to running 96 arrays through a Gene-Titan. Secondly they focus on the poor performance of some reported multiplexing experiments quoting a 4-fold variation in sample coverage. Our experience is that carefully processed multiplex pools can get very close to 1:1 ratios and certainly within 20% if not significantly lower.

How good is NIPT: NIPT testing for T21 has been shown to be effective in high-risk pregnancies. A recent paper in NEJM reports on 15,000 first trimester normal pregnancies on the use of Trisomy 21 NIPT compared to standard screening (measurement of nuchal translucency and biochemistry). They reported higher sensitivity, lower false positive rate, and higher positive predictive value for NIPT than for standard screening. NIPT detected Trisomy 21 in 10% of cases (38 of 38) compared with just under 80% with standard screening, and NIPT had  the lowest false positive rate 0.06% compared to 5.4% for standard screening. Others have also reported the success of NIPT for normal patients (e.g. 1914, 2049, 3000 and 11,105 cases in the hyperlinked papers).

I am certain that by the time my kids are having their own kids they'll get NIPT tests and probably have had their genomes screened for carrier status - if not have their WGS data on their iPhone X.