Data-driven identification of predictive risk biomarkers for subgroups of osteoarthritis using interpretable machine learning.

Osteoarthritis (OA) is increasing in prevalence and has a severe impact on patients' lives. However, our understanding of biomarkers driving OA risk remains limited. We developed a model predicting the five-year risk of OA diagnosis, integrating retrospective clinical, lifestyle and biomarker data from the UK Biobank (19,120 patients with OA, ROC-AUC: 0.72, 95%CI (0.71-0.73)). Higher age, BMI and prescription of non-steroidal anti-inflammatory drugs contributed most to increased OA risk prediction ahead of diagnosis. We identified 14 subgroups of OA risk profiles. These subgroups were validated in an independent set of patients evaluating the 11-year OA risk, with 88% of patients being uniquely assigned to one of the 14 subgroups. Individual OA risk profiles were characterised by personalised biomarkers. Omics integration demonstrated the predictive importance of key OA genes and pathways (e.g., GDF5 and TGF-ß signalling) and OA-specific biomarkers (e.g., CRTAC1 and COL9A1). In summary, this work identifies opportunities for personalised OA prevention and insights into its underlying pathogenesis.

CancerDetector: ultrasensitive and non-invasive cancer detection at the resolution of individual reads using cell-free DNA methylation sequencing data.

The detection of tumor-derived cell-free DNA in plasma is one of the most promising directions in cancer diagnosis. The major challenge in such an approach is how to identify the tiny amount of tumor DNAs out of total cell-free DNAs in blood. Here we propose an ultrasensitive cancer detection method, termed 'CancerDetector', using the DNA methylation profiles of cell-free DNAs. The key of our method is to probabilistically model the joint methylation states of multiple adjacent CpG sites on an individual sequencing read, in order to exploit the pervasive nature of DNA methylation for signal amplification. Therefore, CancerDetector can sensitively identify a trace amount of tumor cfDNAs in plasma, at the level of individual reads. We evaluated CancerDetector on the simulated data, and showed a high concordance of the predicted and true tumor fraction. Testing CancerDetector on real plasma data demonstrated its high sensitivity and specificity in detecting tumor cfDNAs. In addition, the predicted tumor fraction showed great consistency with tumor size and survival outcome. Note that all of those testing were performed on sequencing data at low to medium coverage (1× to 10×). Therefore, CancerDetector holds the great potential to detect cancer early and cost-effectively.

Enhanced platelet adhesion and aggregation by endothelial cell-derived unusually large multimers of von Willebrand factor.

Endothelial cells synthesize and secrete von Willebrand factor (VWF) multimers, including unusually large forms (ULVWF), which are usually cleaved into smaller multimers found in normal plasma (P-VWF). Thrombotic thrombocytopenic purpura (TTP) is a microangiopathic disorder characterized by systemic attachment of platelets to inadequately cleaved ULVWF multimers. We have compared ULVWF and P-VWF in their capacity to become immobilized onto surfaces in vitro and their ability to mediate platelet adhesion. We have also used functional assays to directly compare ULVWF forms with purified P-VWF in mediating platelet aggregation in solution. At comparable concentrations, ULVWF is more effectively adsorbed onto glass surfaces than P-VWF and supports increased platelet adhesion. ULVWF is also significantly more potent than P-VWF in mediating both shear-induced platelet aggregation and ristocetin-mediated platelet agglutination.

Aggregometry detects platelet hyperreactivity in healthy individuals.

Aggregometry is widely used to assess platelet function, but its use in identifying platelet hyperreactivity is poorly defined. We studied platelet aggregation in 359 healthy individuals using the agonists adenosine diphosphate (ADP), epinephrine, collagen, collagen-related peptide, and ristocetin. We also assessed the reproducibility of these assays in 27 subjects by studying them repeatedly on at least 4 separate occasions. Healthy subjects exhibited considerable interindividual variability in aggregation response to agonists, especially at concentrations lower than those typically used in clinical laboratories. For each agonist tested at these submaximal concentrations, a small proportion of individuals demonstrated an unusually robust aggregation response. Subjects who exhibited such in vitro hyperreactivity to one agonist tended to demonstrate a similar response to others, suggesting that hyperreactivity is a global characteristic of platelets. Epinephrine and collagen-related peptide were especially reliable and efficient in detecting hyperreactivity. For epinephrine, excellent reproducibility persisted for up to 3 years, and hyperreactivity was associated with female sex and higher fibrinogen levels (P < .02). We recommend these assays as appropriate candidates for future studies requiring accurate assessment of increased platelet reactivity. These include clinical studies to improve risk assessment for arterial thrombosis, as well as genetic studies to establish determinants of the hyperreactive platelet phenotype.

Duration of exposure to high fluid shear stress is critical in shear-induced platelet activation-aggregation.

Platelet functions are increasingly measured under flow conditions to account for blood hydrodynamic effects. Typically, these studies involve exposing platelets to high shear stress for periods significantly longer than would occur in vivo. In the current study, we demonstrate that the platelet response to high shear depends on the duration of shear exposure. In response to a 100 dyn/cm2 shear stress for periods less than 10-20 sec, platelets in PRP or washed platelets were aggregated, but minimally activated as demonstrated by P-selectin expression and binding of the activation-dependent alphaIIbbeta3 antibody PAC-1 to sheared platelets. Furthermore, platelet aggregation under such short pulses of high shear was subjected to rapid disaggregation. The disaggregated platelets could be re-aggregated by ADP in a pattern similar to unsheared platelets. In comparison, platelets that are exposed to high shear for longer than 20 sec are activated and aggregated irreversibly. In contrast, platelet activation and aggregation were significantly greater in whole blood with significantly less disaggregation. The enhancement is likely via increased collision frequency of platelet-platelet interaction and duration of platelet-platelet association due to high cell density. It may also be attributed to the ADP release from other cells such as red blood cells because increased platelet aggregation in whole blood was partially inhibited by ADP blockage. These studies demonstrate that platelets have a higher threshold for shear stress than previously believed. In a pathologically relevant timeframe, high shear alone is likely to be insufficient in inducing platelet activation and aggregation, but acts synergistically with other stimuli.

Platelet aggregation and activation under complex patterns of shear stress.

Arterial stenosis results in a complex pattern of blood flow containing an extremely fast flow in the throat of stenosis and a post-stenosis low flow. The fast flow generates high shear stress that has been demonstrated in vitro to activate and aggregate platelets. One potential problem of these in vitro studies is that platelets are invariably exposed to a high shear stress for a period that is significantly longer than they would have experienced in vivo. More importantly, the role of the post-stenosis low flow in platelet activation and aggregation has not been determined. By exposing platelets to a shear profile that contains both high and low shear segments, we found that platelets aggregate when they are exposed to a high shear stress of 100 dyn/cm(2) for as short as 2.5 s, a period that is significantly shorter than those previously reported (30-120 s). Platelet aggregation under this condition requires a low shear exposure immediately after a high shear pulse, suggesting that post-stenosis low flow enhances platelet aggregation. Furthermore, platelet aggregation under this condition is not activation-dependent because the CD62P expression of sheared platelets is significantly less than that of platelets treated with ADP. Based on these findings, we propose that shear-induced platelet aggregation may be a process of mechanical crosslinking of platelets, requiring minimal platelet activation. This process may function as a protective mechanism to prevent in vivo irreversible platelet activation and aggregation under temporary high shear.