In this study, we evaluated the association between soluble P-selectin levels at dialysis initiation and ASCVD. We found that baseline soluble P-selectin levels were associated with composite ASCVD, cardiovascular mortality and sudden cardiac death among males. The association with cardiovascular mortality and sudden cardiac death persisted despite adjustment for other inflammatory markers and platelet count.
P-selectin is involved in the early stages of atherogenesis by recruiting leukocytes into developing atherosclerotic plaques. In mouse models of atherosclerosis, P-selectin gene knockout slowed development of atherosclerotic lesions [13
], and overexpression led to increased plaque burden [15
]. In humans, higher soluble P-selectin levels have been associated with greater atherosclerotic plaque burden [16
] and myocardial infarction [18
]. In this study, we detected a marginal association between soluble P-selectin levels and aggregate ASCVD events among males after adjustment for demographics and cardiovascular risk factors, but the association did not persist after the adjustment for other inflammatory markers.
We detected a stronger, more robust association between soluble P-selectin and cardiovascular mortality and, particularly, sudden cardiac death among males. Autopsy studies in humans have demonstrated that intracoronary thrombosis and microemboli play a major role in the pathogenesis of sudden cardiac death [19
]. Previous work in dogs has shown that thrombotic occlusion of coronary arteries has a greater propensity to induce malignant ventricular arrhythmias than mechanical occlusion of coronary arteries by balloon inflation [20
]. This suggests that the process of thrombosis and platelet aggregation results in electrical instability in the heart independent of the amount of ischemic myocardium.
Previous human studies also support the hypothesis that greater expression of P-selectin is associated with unstable arrhythmias. Blood from humans with a prior history of ventricular fibrillation complicating myocardial infarction demonstrate greater expression of platelet P-selectin in response to challenge with lipopolysaccharide than their counterparts with myocardial infarction and without ventricular fibrillation [21
Recently, genetic polymorphisms in the P-selectin gene (SELP) have been found to differ between patients with a history of ventricular fibrillation after myocardial infarction and those with uncomplicated myocardial infarction, again suggesting a pathogenic role of P-selectin in electrical destabilization of the myocardium after an ischemic injury [22
Our findings of increased risk of cardiovascular mortality and, particularly, sudden cardiac death in patients with higher levels of soluble P-selectin are consistent with these prior reports and could indicate those at higher risk for sudden death. We also extend these findings by demonstrating that soluble P-selectin levels measured in a group at high risk for sudden cardiac death are associated with fatal disease over long-term follow-up.
Notably, in this study there was no association between soluble P-selectin and CVD in females. The distribution and correlates of soluble P-selectin by sex were similar and the number of cardiovascular events in males and females was also similar. Therefore, it is unlikely that differences in clinical characteristics by sex account for the observed heterogeneity. Although the exact mechanisms underlying this sex difference are not clear, a large volume of in vitro and in vivo work suggests that there are differences in endothelial function by sex. In particular, estrogen has been shown to increase factors which protect against endothelial activation, such as nitric oxide and endothelial-derived hyperpolarizing factor [23
]. It is possible that increases in endothelial protective factors in females mitigate the effects of activated platelets making this pathway less important in the pathogenesis of fatal CVD. In our sensitivity analysis, we did not see differences when restricting our analysis to only older women; however, our sample size was limited making it difficult to explore this relationship and how it may be affected by age and other factors in women. The differences in risk associated with soluble P-selectin levels by gender should be confirmed in future studies.
Our study has several limitations. Membrane-bound P-selectin expression can be modified quickly in vivo and therefore a single measure of soluble P-selectin may not be a reliable indicator of overall exposure. Additionally, while some of the hypothesized effects of P-selectin in promoting early atherosclerotic plaque development may occur over time, other effects may be more acute in nature and could not be addressed with our study design.
Our study does, however, have many strengths. The CHOICE cohort is comprised only of incident dialysis patients. We have a prospective design with detailed ascertainment of baseline comorbidities and complete ascertainment of events for all participants over a clinically significant interval. We collected extensive covariate information on participants, including other inflammatory markers, allowing us to isolate the association of soluble P-selectin independent of these markers.
It has been previously shown that out-of-hospital sudden cardiac death makes up a significant proportion of the observed deaths in the CHOICE cohort, with a cumulative incidence of 20.4% after 8 years of follow-up [2
]. The inflammatory markers interleukin-6 and C-reactive protein were both strongly associated with risk of sudden cardiac death in this cohort [2
]. In this study, we demonstrate that soluble P-selectin is strongly associated with cardiovascular mortality and, in particular, sudden cardiac death independent of these established inflammatory markers. In aggregate, these results suggest the potential role of multiple inflammatory markers representing different components of the inflammatory cascade and thrombotic response, including cellular adhesion molecules, cytokines and generalized inflammatory markers as risk factors for sudden cardiac death and cardiovascular mortality.