Using Large-Scale Multiomic Datasets to Assess the Population Frequency of ADAMTS13 Loss-of-Function Variants in the United States and United Kingdom

Background:

Congenital thrombotic thrombocytopenic purpura (cTTP), also known as Upshaw-Schulman syndrome (USS), is a rare thrombotic microangiopathy caused by homozygous or compound heterozygous germline loss-of-function (LOF) variants in ADAMTS13. The population prevalence of USS and ADAMTS13 LOF variants remains controversial, with estimates varying widely due to study limitations such as ascertainment bias, small sample size, assessment of genetically homogenous populations, and failure to adjust for sex and ancestry. Large-scale whole genome sequencing datasets offer a unique opportunity to address this question with far greater reliability and precision than previously possible. Here, we interrogated two national biorepositories in the United States and United Kingdom to compute the population frequencies of both known USS mutations and novel ADAMTS13 variants associated with reduced plasma ADAMTS13 levels.

Aims:

Determine the prevalence of known and novel USS variants in the United States and United Kingdom using population-scale genomic and proteomic data.

Methods:

We used data from the UK Biobank (UKB, N=452,910) and the NIH All of Us program (AoU, N=242,901) to identify participants with rare (MAF ≤0.1%) germline variants in ADAMTS13. Variants were assigned a functional impact score (FIS) between 0.0 and 1.0, with higher scores indicating variants predicted in silico to have a greater likelihood of causing protein LOF. A score of 1.0 indicates mutations causing stop gained (nonsense), frameshift, or disruption of an essential splice site. ADAMTS13 variant carriers were divided into three groups: FIS ≥0.7, FIS=1.0, and USS-associated variants according to the Universal Protein Resource Consortium (UniProt). Total circulating ADAMTS13 levels in each group were assessed in a subset of UKB participants (N=48,415) using the Olink® Explore 3072 plasma proteomics panel and reported as linearized normalized protein expression (L-NPX) values. National sex and ancestry data used for adjustment of crude prevalence estimates were obtained from the US Census Bureau 2023 population projections and 2021 UK Census.

Results:

We identified 1267 participants in the UKB and 564 in AoU with known USS variants. A further 454 UKB and 386 AoU participants were identified with ADAMTS13 variants at FIS≥0.7. Of detected variants, 99.8% were heterozygous. Mean circulating ADAMTS13 levels were significantly lower in each of the ADAMTS13 variant carrier groups compared to wild-type participants: those with USS variants had a mean plasma level of 61.0% (P < 2.2 x 10^-16), those with FIS=1.0 variants had a mean plasma level of 65.0% (P=1.1 x 10^-8), and the remaining participants with FIS≥0.7 variants had a mean plasma level of 73.1% (P=3.0 x 10^-6). We computed the sex and ancestry-adjusted prevalence of known USS-associated variants to be 0.25% in both the US and UK. In both countries, USS variants were marginally more common among participants of European descent than other ancestries. After adjustment, we did not detect any difference in prevalence between males and females. Including individuals with predicted LOF ADAMTS13 variants (FIS ≥0.7), the population prevalence rises to 0.40% in the US and 0.37% in the UK. Assuming Hardy-Weinberg equilibrium, the expected population frequency of homozygous or compound heterozygous LOF variants in ADAMTS13 is 0.0004%, or approximately 4 per million.

Conclusions:

Our work represents the largest study to date examining the population frequency of LOF variants in ADAMTS13. We found that germline genetic variants in ADAMTS13 resulting in reduced plasma protein levels are likely more common than previously thought. Accordingly, the frequency of USS in the general population may be underestimated.