Abstract
Genetics is having a major impact on understanding of many complex diseases including stroke. The recent GIGASTROKE consortium identified 89 genetic loci for stroke and its different subtypes, 1 and many more have been identified for magnetic resonance imaging (MRI) markers of cerebrovascular disease such as white matter hyperintensities (WMH).2–4 These findings are providing new information on stroke pathophysiology, and identifying novel treatment targets. 5 However stroke genetic data are also now being in a multitude of other ways, and three papers in this month’s issue demonstrate examples of these different applications.
One common problem in observational studies is to determine whether an association is causal. This can be investigated using genetic data with a technique called Mendelian Randomization. In this approach, genetic variants are used as instrumental variables to approximate the effect of an exposure to seek evidence for a causal association with stroke. The method is based on the random assortment of genetic alleles during meiosis, making the analysis less prone to confounding and reverse causation than observational studies. It is increasingly being used because it can overcome a major limitation of evidence from observational studies namely unmeasured confounding. 6
One factor that we increasingly realize has a major impact on many common diseases including stroke is frailty. Frailty is a distinctive health state in which the ability of older people to cope with acute stressors is compromised by an increased vulnerability brought by age-associated declines in physiological reserve and function across multiple organ systems. 7 Although closely associated with age, multimorbidity, and disability, frailty is a discrete syndrome that is associated with poorer outcomes across a range of medical conditions. 7 It is present in 20−30% of patients presenting with stroke and associated with worse outcomes. 8 However, whether this association with outcome is a direct effect of frailty, or due to unmeasured confounding, is uncertain. To investigate this question Cai and colleagues used Mendelian Randomization, and publish their findings in this months International Journal of Stroke (IJS). 9
They identified genetic instruments for frailty in a genome-wide association study meta-analysis including 175,226 individuals of European descent. Corresponding genetic association estimates for functional outcome after ischemic stroke at 90 days were taken from the Genetic of Ischemic Stroke Functional Outcome (GISCOME) network. They found that genetically predicted higher frailty index (odds ratio (OR) = 5.12; 95% confidence interval (CI) = 1.31–20.09; p = 0.019) was associated with a worse functional outcome, defined as a modified Rankin Scale score of ⩾3, after ischemic stroke. After adjusting for potential confounding traits including body mass index, C-reactive protein, inflammatory bowel disease, and smoking initiation, the association remained. They conclude that their results support a causal role for frailty on poor functional outcome after ischemic stroke, and that treating frailty might represent a potential target for intervention to improve stroke recovery. 9
Similar Mendelian randomization techniques can be used to investigate whether specific drugs might be useful in treatment of stroke. One such drug is sodium valproate. In 2012, one of the first genes identified for stroke was Histone Deacetylase 9 (HDAC9). 10 This was associated specifically with large artery stroke and not other subtypes, and subsequent studies demonstrated it was also associated with atherosclerosis elsewhere in the body, and that the genetic variant appeared to act via promoting atherosclerosis. 11 The commonly used anti-epileptic drug sodium valproate inhibits the HDACs among other actions, and it has been suggested that it might represent a novel treatment for atherosclerosis and stroke. Epidemiological studies report that taking sodium valproate is it is associated with a reduced stroke risk compared to other anti-epileptics supporting this hypothesis. 11 Also in this month’s IJS, Mayerhofer and colleagues use Mendelian Randomization techniques to investigate this question, and ask whether genetic variants that influence seizure response (and therefore efficacy of sodium valproate in the individual) associate with ischemic stroke among valproate users. 12
They derived a genetic score for valproate response using genome-wide association data of seizure response after valproate intake from the Epilepsy Pharmacogenomics Consortium. They then tested this score among valproate users of the UK Biobank for association with incident and recurrent ischemic stroke. A higher valproate response genetic score was associated with higher serum valproate levels, and with lower ischemic stroke risk with a halving of absolute risk in the highest compared to the lowest score tertile. 12 They replicated this finding in the second cohort of valproate users from the Mass General Brigham Biobank. They conclude that their results provide causal support for the effectiveness of valproate in ischemic stroke prevention and that clinical trials are now required in order to confirm this therapeutic effect.
A number of monogenic forms of stroke can cause small vessel disease, of which the most common is Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy (CADASIL). This is conventionally thought of as a rare disease with a prevalence of about 2 to 5 in 100,000, 13 but recent data have surprised those in the field by demonstrating that similar typical CADASIL mutations in the NOTCH3 gene occur in about one in every 500 normal individuals in populations such as UK biobank.14,15 Interestingly these mutations in apparently “unaffected individuals” are nevertheless associated with an increased risk of future stroke and dementia. 16 These findings are blurring the distinction between monogenic stroke and common polygenic stroke, and making us re-evaluate how we look at the distinction. We learn more about this area from a study from Taiwan by Lin and colleagues in this month’s IJS. 17 They used genome sequencing data from 114,000 Han Chinese participants from the Taiwan Biobank. Of these, 0.95% harbored the pathogenic NOTCH3 R544C variant, by far the most common CADASIL variant in Taiwan. The R544 C carriers had an increased history of both stroke (OR: 2.52, 95% confidence interval (CI) (1.45, 4.37)), and dementia (OR: 30.1, 95% CI (3.13, 289.43)). In individuals with the variant, both hypertension and diabetes mellitus were independently associated with stroke risk. These data confirm how common these typical CADASIL mutations are in the general population, and how they are particularly common in Far Eastern populations including Taiwan. They also confirm that they predict stroke risk, and in those with these variants conventional cardiovascular risk factors such as hypertension are major predictors of whether stroke occurs. 16 This emphasizes that even in monogenic stroke conventional risk factors are important in increasing risk, and should be actively treated.
Cerebral small vessel disease remains one of our major challenges in stroke medicine. 18 Increasingly common with age, and prevalent throughout the globe particularly in low and middle income countries (LMICs), 19 it is a major cause of both ischemic and hemorrhagic stroke as well as vascular cognitive impairment and dementia. Characteristic MRI features include lacunes, WMH, and cerebral microbleeds, 20 all of which have been associated with the degree of cognitive impairment and dementia.21–23 Another MRI feature of small vessel disease are cerebral microinfarcts (CMI). Initially seen on 7T MRI, CMI have now been described on clinical 3T scans. In an excellent review in this month’s IJS, Huang and colleagues provide an update on their detection, prevalence, and association with symptoms. 24 They describe how CMIs are small ischemic lesions invisible to the naked eye at brain autopsy, while the larger ones (0.5–4 mm in diameter) can be visualized on MRI. They can be detected both on diffusion-weighted imaging (DWI) as incidental small DWI-positive lesions, and on structural MRI for those confined to the cortex and in the chronic phase. Acute DWI-positive CMI may progress in a number of ways; either into chronic cortical CMIs, WMH, or may disappear radiologically (but not necessarily histologically). Huang and colleagues describe how CMIs have heterogeneous causes, not only being caused by cerebral small disease but also by other pathologies, including and large vessel disease and cardioembolism. CMIs appear to have clinical relevance being associated with outcomes, including cognitive impairment, and neurobehavioral disturbances. It has been suggested that they may contribute to the cortical atrophy which is a characteristic feature of cerebral small vessel disease. This is an excellent review providing a useful update on this rapidly evolving field.
One presentation I occasionally come across is acute uniocular loss of vision due to non-arteritic central retinal artery occlusion (CRAO). It is unclear whether this should be treated in the same way as an acute ischemic stroke, and prioritized for thrombolysis. To help guide us, Shahjouei and colleagues performed a systemic review of both intravenous and intra-arterial thrombolysis for non-arteritic CRAO, which is also published in this issue. 25 They retrieved data from 771 patients described in 72 publications. Visual improvement was reported in 74.3% of patients who received intravenous tPA within 4.5 h, and in 60.0% of those who received intra-arterial tPA within 24 h. They conclude that early thrombolytic therapy with tPA is associated with enhanced visual recovery in CRAO, and that it should be given as soon as possible, although further studies are required to refine the optimum time window.
Finally carotid stenting is used to carotid stenosis. One complication is perioperative hypotension, which as I have personally experienced can be quite severe on occasion. In this issue, Yang and colleagues perform a systematic review which reports its prevalence to be high at 28.6%. 26 They identified a number of risk factor for it which include older age, and bulb stenosis. It is most often related to stretching of the vessel by the balloon and/or stent and stimulation of carotid baroreceptors and consequent fall in vascular tone. It usually only lasts a brief time, although occasionally be prolonged. It is certainly something to be aware of.