Various clinical, molecular and genetic risk factors for VTE have been identified, but prediction of VTE risk remains suboptimal. Widely used clinical risk scores, such as the Wells and Geneva risk scores, have been shown to be marginally less effective than the clinical Gestalt of reasonably experienced physicians, and the large majority of individuals who receive a positive D-dimer result turn out not to have VTE upon further examination (96% of such individuals in a real-life example). The cost/benefit ratio and selection criteria for genetic screening for congenital thrombophilia are so unclear that guidelines advise to forego it altogether. Predicting recurrence risk is even more difficult; for this task, no clinical signs or symptoms, and no molecular markers have yet been found to be reliably predictive. A rough initial categorization, of provoked versus unprovoked VTE based on the presence or absence of transient or chronic risk factors in the acute phase, has helped somewhat in the risk stratification of patients at risk of recurrent VTE, but it is highly likely that a large proportion of patients are still exposed to anticoagulation for unnecessarily long periods of time.
Evidently, the risk factors that are known to predispose to VTE in both acute and long-term settings only marginally correspond with a patient’s actual risk of VTE. Two patients of seemingly similar clinical risk profiles (e.g., positive D-dimer, high Wells pre-test probability, unprovoked etiology of the VTE) may be affected differently by the presence of risk factors (e.g., specific comorbidity, advanced age, impaired renal function), and may respond entirely differently to the same treatment. It is highly plausible that our current, macroscopic view of such patients obscures important differences between them, which may only be revealed by taking a finer-grained perspective. The proteome is likely to reflect much more precisely which patients have a high probability of being at risk of VTE, and which patients only appear to be at risk. Contrasting the protein expression profiles of clinically similar patients with different outcomes will help elucidate which molecular mechanisms are of primary importance in eliciting and sustaining VTE risk. Besides etiological mechanisms, the proteomes of patients with acute VTE may provide additional hints into the long-term risk of VTE, such as those relating to thrombus composition and fibrinolytic potential. In this proposal, we request funding to facilitate a series of studies in which the protein expression profiles of individuals conventionally thought to be at elevated risk of VTE, but who exhibit differential outcomes of interest (VTE vs no VTE), are directly compared using innovative proteomic analysis technology and advanced machine learning methods. Through these studies, we hope to elucidate the different molecular pathways through which VTE arises, and potentially through which VTE risk persists for longer durations. The results of this proposed project may be useful for the development of new models for risk prediction and stratification for use in clinical practice.
Vincent ten Cate, MSc.