Pediatric stroke and ADAMTS genes

In a study of 270 children with ischemic stroke and their parents, Arning and colleagues report associations with genetic variants in several genes of the ADAMTS family.¹ 

Pediatric stroke is a rare but devastating disease, with substantial short-term mortality that in survivors often leads to lifelong debilitating intellectual and motoric defects.²  The etiology of pediatric stroke is complex.³  There is a moderate genetic component, and several genetic abnormalities, particularly in the coagulation system, have been associated with its occurrence. These include common variants such as prothrombin 20210A and factor V Leiden, well-established risk factors for venous thrombosis. It is self-evident that these variants can only contribute to a minor extent to the occurrence of pediatric stroke, given the wide discrepancy between the incidence of pediatric stroke (∼ 1 in 20 000) and the prevalence of these variants (∼ 1 in 20). Moreover, even for venous thrombosis, where prothrombotic abnormalities usually stand out more than for arterial disease, most carriers of these variants never develop thrombosis. The association with arterial disease for these and other prothrombotic variants is weak, in adults as well as in children, with a nearly negligible risk-enhancing effect.⁴  So, there must be more. Because pediatric stroke has also been associated with arterial malformations, sickle cell disease, and infectious diseases, a complex multicausal etiology is likely.

Searching for the cause of a rare disorders with a complex etiology is a daunting enterprise due to the large numbers of patients required to disentangle the various causes. In this issue of Blood, Arning and colleagues from several centers in Germany report on a search for genetic causes of pediatric stroke. In a nationwide survey, 270 children with stroke (mean age at the event < 6 years) were included in the study, and analyzed together with their parents. The authors applied the technique of a large number of single nucleotide polymorphisms (SNPs) scattered over the genome, but rather than a genome-wide association study (GWAS) in which patients are contrasted to unrelated healthy reference individuals, they used the family structure in a so-called transmission disequilibrium test (TDT).

The TDT is a clever use of information of relatedness between subjects, in the absence of extended pedigrees allowing true linkage analysis. Because it requires DNA from both parents, it is particularly useful for the genetic analysis of pediatric disorders. First proposed in 1993,⁵  the test directly compares the transmission of the candidate allele from a heterozygous parent to the affected child with the expected tranmission probability of each allele (of 50%), and by definition takes household effects into account. Because TDT belongs to the family of linkage analysis and not association analysis, it is not affected by population stratification. Therefore, the results of this analysis are more persuasive than those of previous studies in unrelated individuals.

They succesfully genotyped over 300 000 single nucleotide variants, as well as a large number of variants prevously identified as potential risk factors in adult ischemic stroke. Of the latter, none provided a clear signal, but several variants stood out from the larger group. Four variants were suggestive based on the statistical threshold used, of which 2 were intragenic, in ADAMTS2 and ADAMTS12, with odds ratios close to 2, and minor allele frequencies of 20% to 50%. Moderate associations, below the preset threshold for the discovery, were found for ADAMTS13 and ADAMTS17.

These proteins belong to a class of enzymes called a disintegrin and metalloproteinase with thrombospondin motif (ADAMTS), that contain 18 proteins, ADAMTS1 to ADAMTS19. These proteins share a similar structure, with a metalloproteinase domain, a disintegrin-like domain, and a thrombospondin-type motif. However, they do not have similar function, nor are they located on the same chromosome. Deficiencies of the 2 most well-known proteins in the family even lead to diametrically opposed symptoms: a deficiency of ADAMTS2, which is involved in collagen assembly, leads to Ehlers-Danlos syndrome with hyperelastic skin, hyperflexible joints, and fragile vessel walls causing easy bruisability and aneurysmatic bleeding. ADAMTS13 is the enzyme that cleaves procoagulant von Willebrand factor (see figure), and a deficiency leads to microvascular thrombosis known as thrombotic thrombocytopenic purpura (TTP). Other ADAMTS enzymes are involved in cartilage metabolism, angiogenesis, and cell adhesion, while for several the function is not completely known. Hence, the findings of Arning et al on pediatric stroke are intriguing and beg for replication and laboratory follow-up.

It is tempting, and Arning et al are duly tempted, even though the association is not the strongest, to hypothesize further on a possible involvement of ADAMTS13, because this is the only one in the ADAMTS family of proteins known to be involved in thrombosis. However, this association is only clear for severe deficiencies of ADAMTS13, which is an extremely rare disorder, and the variants reported here do not cause such a deficiency. For moderately reduced levels of ADAMTS13, which are not related to TTP, the association with arterial disease is controversial, with studies from our centers showing both a protective and a deleterious effect of reduced levels.^6-8  Therefore, the results of the study by Arning and colleagues need to be replicated first, and if the association with ADAMTS13 variants stands, subsequent analyses should focus on associations between the variants and changes in level or function of the proteins in carriers of the implicated variants. The suggestion of the authors that ADAMTS13 concentrate might have a role in pediatric stroke is premature, despite some positive results in experimental stroke in animal models, and years of fundamental and clinical research are needed before such a conclusion could be reached, if ever.

The study by Arning et al at the moment has no clinical implications. Its importance lies in an elegant genetic analysis of a relatively large group of patients with a devastating and enigmatic disease. The findings are intriguing, and may lead to new insights in the pathogenesis of the disorder, but it remains to be seen which these insights will be.