A Novel Next-Generation Sequencing Based Assay for Non-Invasive Prenatal Testing of Sickle Cell Disease without Paternal DNA

Introduction: Over 300,000 infants are born with sickle cell disease (SCD) every year worldwide, including at least 1,000 in the US. Prenatal diagnosis by amniocentesis or chorionic villus sampling is available; but high cost, invasiveness, and risk of miscarriage limit their use. Recently, non-invasive prenatal testing (NIPT) has become commonplace for aneuploidies, including Trisomy 21. These non-invasive tests operate by genetic analysis of the cell-free fetal DNA (cffDNA) present in maternal blood. The safety and accuracy of NIPT have been key drivers for its rapid and widespread adoption. Yet, no NIPT for SCD or other hemoglobinopathies have been commercialized to date, despite the large numbers of patients affected in the US and worldwide. While de novo mutations can only be of fetal origin and can be identified by available next-generation sequencing (NGS) methods, NIPT for recessively inherited disorders is more challenging. This is because a mother who is a carrier for a recessive disorder contributes a high level of background pathogenic DNA molecules. Therefore, a key technical challenge of NIPT for recessive disorders is developing an assay sensitive enough to detect <5% deviation from 50% allele fraction. To overcome this challenge, we have developed and optimized an NIPT for SCD by assessing the relative mutation dosage of fetal SCD and beta-thalassemia DNA through a novel molecular counting strategy using NGS.

Objectives: The primary objective of this study is to evaluate the performance of a novel NIPT for sickle cell disease.

Methods: The SCD NIPT assay and associated custom bioinformatics analysis were performed on cfDNA obtained from a training cohort of non-pregnant compound heterozygotes for SCD. The SCD NIPT assay was then performed on a validation cohort of pregnant women with either SCD or sickle cell trait (SCT). The accuracy of the SCD NIPT was evaluated by comparison with newborn screening results.

Results: Non-pregnant individuals with genotype HbSE, HbSC, or HbS/beta-thalassemia were included as a training cohort to establish the precision and accuracy of the assay for measuring HbS allele fraction from cfDNA. As expected, the HbS allele fraction in these individuals was 0.500 (standard deviation = 0.011, n = 26), and there was no detectable fetal fraction in these samples. Both training and validation cohort results matched the theoretical limit of detection set by the number of cell-free HBB DNA molecules in plasma. The precision and accuracy of the HBB assay on cfDNA were then used in conjunction with >1000 pre-clinical samples (mixtures of sheared SCT and SCD genomic DNA) to determine analytical sensitivity >98% and specificity >99%, even in the absence of paternal DNA.

Conclusion: We have developed an assay for non-invasive prenatal testing of sickle cell disease. The results obtained to date indicate that the assay reliably detects fetal SCD when the fetal fraction is as low as 5%, the same limit as aneuploidy NIPT. A fetus with SCD has already been identified, and follow-up is ongoing with >20 pregnancies. Since the HBB NIPT is highly targeted, sequencing cost is <$30 per sample. The ability to ascertain fetal SCD status based only on maternal blood will be valuable in clinical settings where the father is unavailable or sample collection would be inconvenient or time-consuming. Several Phase I/II and Phase III trials for curing SCD or beta-thalassemia using autologous gene-editing of stem cells are currently in progress. SCD NIPT could be particularly useful for deciding to bank umbilical cord blood as a source of stem cells for future gene-editing cures.