In this prospective cohort study, we explored the associations between plasma levels of the soluble form of the complement inhibitors CD46 and CD59 and long-term prognosis after an acute coronary event. After adjusting for age, sex and potential confounders, we found that elevated levels of both proteins were independently associated with increased risk for recurrent coronary events, and sCD59 was strongly associated with the risk for HF hospitalization. The link between sCD59 and incident HF was further supported by positive correlations with LV mass and volumes, indicative of accelerated remodelling, and a negative association with LV systolic function at 1-year after the index event.

During the acute phase of ACS, danger-associated molecular patterns (DAMPs) released by cell necrosis bind to pattern recognition receptors (PRRs) on immune cells, triggering inflammatory responses [13]. PRRs such as mannose-binding lectin (MBL), and collectins trigger the complement system through the lectin pathway [14]. The activated complement fragments trigger an inflammatory response [6] and excessive complement activation was found to exacerbate cardiomyocyte injury [15]. Previous studies have shown that MBL levels are elevated during cardiac ischemia/reperfusion and contribute to cardiomyocyte damage [16, 17]. Accordingly, Weisman et al. and Banz et al. reported that blocking complement activation by soluble CR1 in animal models of MI led to smaller injuries and improved function [18, 19]. These results suggest that the complement system plays an important role in cardiac injury post-MI, and that inhibition of complement pathways might be a useful treatment to prevent post-MI complications.

CD46 and CD59 inhibit excessive complement activation, and higher deposition of these inhibitors in the ischemic myocardial tissue might confer protection against development of post-ACS complications. Early work by Väkevä et al. suggested that the loss of CD59 from the ischemic myocardium allows MAC formation, leading to increased tissue damage [20]. Using biopsies from human infarcted myocardium collected between 1 and 14 days post-mortem, the authors found lower expression or absence of CD59 in the infarcted areas, with concomitant MAC deposition in CD59-negative regions. CD59 was present in small vesicles between the infarcted area and normal tissues, suggesting that it might have been lost through shedding [20]. These findings have been confirmed in a rat model of myocardial ischemia, demonstrating increased presence of complement factors in the myocardium within hours after the acute event. CD59 expression was lost from day 1 onwards and was associated with concomitant MAC deposition [21]. The role of MAC deposition in myocardial loss and dysfunction is further supported by studies showing MAC expression in biopsies from failing human hearts, but not in healthy myocardium [22].

Following its shedding from the myocardium, CD59 can be measured in plasma as sCD59. Plasma levels of sCD59 were previously found to be significantly increased at 4 h and 24 h post-MI compared to healthy controls. Importantly, there was a strong correlation between plasma sCD59 and circulating levels of soluble components of the terminal complement pathway C5b-9, suggesting that sCD59 shedding could allow MAC deposition in the post-ischemic myocardium [9]. There are no previous studies investigating the possible associations between plasma sCD59 and prognosis in ACS patients. However, indications that associations between sCD59 and the extent of tissue damage might exist have been provided by an earlier study on 68 patients with recovery of spontaneous circulation after cardiac arrest [23]. Serum levels of sCD59, sC5b-9, C5a, C3a, C3b, C1q, MBL, Bb, TNF-α, IL-6, neuron-specific enolase (NSE) and S100β have been found to be elevated in these patients compared to controls. sCD59 was positively correlated with sC5b-9, TNF-α, IL-6, S100β and NSE, suggesting that CD59 shedding is proportional with the degree of MAC activation and tissue inflammation. Plasma sCD59 on days 1, 3 and 7 post-cardiac arrest was the strongest predictor of a poor neurological prognosis and mortality at 28 days, supporting its value as a prognostic biomarker [23].

CD46 acts as a cofactor for FI to induce the degradation of C3b and C4b [5]. So far, there are very few studies focusing on the expression and role of CD46 during and after AMI. Yan et al. have found that genes encoding complement proteins, receptors and inhibitors, including CD46 and CD59, were overexpressed in circulating peripheral blood mononuclear cells from AMI patients compared to healthy controls and patients with stable angina pectoris. These findings suggest activation of the complement system and an increase in its inhibitors on circulating immune cells during AMI [24]. A necropsy study on myocardial specimens collected from 50 deceased patients within 2 days post-MI has shown elevated immunoreactivity for both CD46 and CD59 compared to myocardium of controls who had died due to accidents. The expression of complement proteins, MAC and the complement inhibitors was predominantly located in the coronary endothelium, suggesting deposition of complement factors from the blood stream, but was also present on single cells infiltrating the necrotic myocardium, possibly immune cells [25].

We also found that plasma levels of sCD46 and sCD59 were significantly correlated with the levels of inflammatory cytokines, chemokines and metalloproteinases. To our knowledge, this is the first study to demonstrate positive associations between plasma levels of soluble complement regulatory proteins and a broad spectrum of pro-inflammatory immune mediators in the acute post-ACS period. We hypothesize that both CD46 and CD59 shedding and the potent cytokine and chemokine release into the circulation are simultaneously involved in the acute phase of the post-ischemic immune response and myocardial inflammation. However, our study design cannot directly address this hypothesis, as we cannot establish causal or temporal links between these events. Elevated plasma IL-6 has previously been shown to be associated with poor prognosis post-MI [26]. High IL-8 has also been associated with larger infarction, worse LV recovery and adverse prognosis in STEMI patients [27]. CCL2, CCL3, CCL4 and CX3CL1 are important chemokines involved in inflammatory and reparatory immune processes [28]. Increased levels of CCL3 and CX3CL1 in plasma have previously been linked with increased risk for future MACE in ACS patients, suggesting that the immune responses driven by these chemokines in the acute phase are predominantly detrimental [29, 30]. Matrix metalloproteinases (MMPs) are enzymes involved in tissue damage, repair and remodelling through digestion of extracellular matrix proteins. High plasma MMP3 in the acute phase has been linked to LV dysfunction and increased mortality post-MI. Evidence for a potential direct detrimental role of MMP7 in MI has been provided by experimental work showing that elevated levels of the metalloproteinase post-MI affect electrical conductivity in the myocardium, leading to increased animal mortality. Deletion of MMP7 reversed these effects [31].

Our study has some limitations that have to be acknowledged. As this is an association study, it cannot provide causal links between sCD59, sCD46 and adverse events in this population. As discussed above, we speculate that the shedding of CD46 and CD59 from the myocardium and their subsequent increase in the circulation may lead to impaired tissue protection, leading to MAC-mediated cellular damage. Secondly, follow-up blood samples and echocardiographic data were only collected from a small sub-group of patients over 75 years of age. Consequently, we cannot exclude that this sub-group might have been underpowered to detect associations between sCD46 and sCD59 at 6-weeks and the outcomes, and the described correlations between sCD59 and echocardiographical parameters of cardiac dysfunction and remodelling cannot be directly extrapolated to younger subjects. However, the associations between sCD59 and incident HF during follow-up were valid in the entire study population. Lastly, as the data is expressed as arbitrary units, we cannot establish cut-off values for the use of sCD46 and sCD59 as a prognostic biomarker for MACE and HF in clinical practice. Our study should be considered as hypothesis-generating.Further studies based on measurements of absolute values by other methods are necessary to confirm our findings and define the potential use of sCD46 and sCD59 as biomarkers in clinical practice.

In conclusion, we show that high sCD59 levels in plasma during the acute phase of an ACS are associated with long-term risk for recurrent cardiovascular events, cardiac remodelling and dysfunction, and incident HF. sCD46 also presented associations with incident MACE, but not with HF. We speculate that sCD46 and sCD59 release from the infarcted myocardium might allow MAC deposition and complement-induced cardiac injury, reflected by elevated levels of immune mediators and metalloproteinases. Our data suggest that therapeutic restoration of these complement inhibitory mechanisms or disruption of MAC deposition in the ischemic myocardium might reduce damage and prevent adverse events during follow-up. The value of sCD46 and sCD59 as independent prognostic biomarkers or as a surrogate biomarkers to monitor the efficiency of therapeutic interventions targeting immune and inflammatory mechanisms has to be confirmed in future studies.