Blogged by Gian Marco De Marchis, MD MSc FESO
In her talk on “Mendelian Randomization in Stroke”, Jemma Hopewell outlined that more than half of drugs thought promising will fail. Mendelian Randomization (MR), which is based on the natural occurrence of genotype (e.g. genotype aa vs. AA), can provide insight into life-long effects of risk factors, while traditional clinical trials provide shorter term insights. MR is not a replacement for RCT, but offers opportunities to explore causal effects and inform trial design.
Mira Katan talked on the clinical applications of blood-based biomarkers. She explained that – basically – we have three types of clinical biomarkers – etiological, diagnostic, prognostic markers. It takes time for biomarkers to be translated into clinical practice. The troponin assay was developed only in the 80’s – years before it became routinely available in the clinic. Why does it take such a long time? There are several phases in the way to clinical application – discovery, validity, derivation, validation, randomized clinical trials. She gave one example. Blood MR-proANP is associated with cardioembolic stroke – this was shown in a derivation study and in a geographically and timely independent cohort sturdy. This kind of validation supports the link between MR-proANP and cardioembolic stroke. MR-proANP is also associated with atrial fibrillation – it can thus be used to select stroke patients which need a more in-depth work-up for atrial fibrillation, and – possibly – select patients for oral anticoagulation in the context of randomized clinical trials.
Guido Falcone talked on the genetics of intracerebral hemorrhage. Why use genetics? The first reason is the help with risk predictions – for instance, in lobar ICH, the proportion of the variance in the risk explained by genetic variation is 48%. The second reason is that genetics can help identify novel targets in human disease. Several landmark studies linked APOE (E2/E4) to ICH recurrence, risk, volume, spot sign in. Genome-wide association studies (GWAS) of ICH identified 1q22 as a risk locus for deep ICH. The third reason is that genetics can help in causal inference. CETP is associated with inverse correlation between lipid levels and risk of ICH. The fourth reason is that genetics can help in risk prediction, as shown in a UK Biobank with 500,000 individuals. APOE-e has been associated with the risk of warfarin-related lobar ICH. In future, genetic analysis of neuroimaging phenotypes will be combined for a more comprehensive assessment.
Anne Joutel spoke about new insights in the mechanisms of ICH in the collagen type IV disease. In a mouse model of lobar ICH, microbleeds originate from capillaries – never from arteries – and arise from a transient leakage of the blood brain barrier (BBB). Oppositely, in a mice model of collagen type IV disease (Col4a1), deep ICH originates from arteries with focal degeneration of smooth muscle cells. A “second hit” seems necessary. Flat-mounted retina preparation of a Col4a1-mutant mouse showed smooth muscle cell loss in arteries, and – in the more distal transitional segment – smooth muscle cell amplification (“hypermuscularization”) These findings were replicated in the brain, i.e. not only in the retina, of Col4a1 mutant mice. Hypermuscularization of transitional segments could be the second hit that leads to deep ICH. The combination of arterial wall weakening with a more distal stiffening due to hypermuscularizazion can ultimately lead to ICH.