The cardioprotective role of the Mediterranean Diet (MD) has long been established mainly due to the outcomes of large epidemiological studies showing that a high adherence to this prudent dietary model is inversely related to the risk of cardiovascular disease (CVD) hard end-points like myocardial infarction (MI), stroke, cardiovascular morbidity and mortality [1, 2]. Although the countries of the Mediterranean basin follow slightly different variations of the MD [3], the consumption of olive oil (OO), as the main dietary fat source, is a central, common trait of all types of the MD. It is therefore obvious that the biological properties of OO, mainly those related to the protection from CVDs, has been studied extensively till now [4]. The cardioprotective properties of OO have been initially attributed to its high content of monounsaturated fatty acids (MUFA), especially oleic acid. However, we now know that a small number of OO’s phytochemicals, especially polyphenols, show a plethora of biological properties which seem to be responsible for the health benefits of OO [5].

Platelets are physiological cellular effectors of haemostasis, innate immunity and tissue regeneration [6]. However, in the last two decades their implication in several mechanisms underlying chronic diseases has been highlighted [7, 8, 9]. A chronically activated or a dysfunctional platelet may initiate and propagate cellular mechanisms in the vascular bed which lead to atherosclerosis, atherothrombosis, vascular inflammation and unregulated haemostasis [10]. The antiplatelet pharmacology, although necessary for the primary and secondary prevention of CVDs, faces many challenges including bleeding complications and resistance of platelets to drug actions [11, 12]. Nutrition is a daily mild modulator of platelet actions and can either directly or indirectly affect platelet functionality. Dietary patterns, micronutrients and phytochemicals able to normalize the biological functions of platelets could be proven beneficial for the prevention of CVDs [13, 14]. The antiplatelet properties of diet is a rather ignored aspect of it, however, taking into account the crucial mediating role of platelets in many pathological mechanisms of CVDs, it can partially explain the cardioprotective properties of prudent dietary patterns such the MD pattern.

This review article aims to highlight the platelet modulating roles of OO polyphenols, trying to critically assess whether those properties could partially explain the cardioprotective role of OO.

Platelets are the smallest blood cells (average diameter 2–5 $\mu$m) and they circulate in numbers ranging from 150 to 350 $\times$ 10${}^9$/L in healthy subjects. They are anucleate, discoid, fragments of megakaryocytes which in turn are derived from pluripotent stem cells in the bone marrow. Platelets are characterized by a rich granule system consisting of $\alpha$-granules, dense granules and lysosomes. Upon activation platelets secrete $\alpha$-granules and dense granules along with granules’ content comprising of a milieu of bioactive components (coagulants, anticoagulants, fibrinolytic proteins, chemokines, adhesion proteins, angiogenic factors, immune mediators, ions, biogenic amines, nucleotides). Platelets are also characterized by an extensive network of surface glycoproteins and other integral proteins by which platelets interact with each other but also with leukocytes, endothelium, subendothelial proteins and cancer cells [15]. Through the secretion of pleiotropic mediators and growth factors, the heterotypic interactions with other cells, their responsiveness to inflammatory stimuli through Toll-like receptors and the induction of neutrophil extracellular traps (NETs), platelets have a central role in haemostasis, thrombosis, inflammation, antimicrobial defense, neurological disorders, tumor growth and metastasis [9, 16].

The dysregulation of the platelet-endothelium crosstalk is the main mechanism underlying the involvement of activated platelets in vascular complications. A healthy, intact, endothelium prevents adhesion and aggregation of platelets by secreting physiological, endogenous antiplatelet agents such as prostacyclin and nitric oxide (NO). However, upon the physical disruption of endothelium, the exposure of subendothelial matrix proteins (von Willebrand Factor, vWF, collagen-COL) triggers the adhesion and activation of platelets to endothelium through their binding to membrane glycoproteins which is the first step of the physiological role of platelets in primary haemostasis [16]. However, a chronic, subclinical, activation of platelets due to metabolic, immune or redox reasons may contribute to atherosclerosis progression in several ways. Activated platelets can adhere to an intact endothelium releasing pro-atherogenic molecules and microparticles. They can promote oxidation of low density lipoprotein (LDL) particles while at the same time they are activated by oxidized LDL (oxLDL) in a positive feedback loop. Taking into account the widely accepted theory of endothelial inflammation as the leading cause of atherosclerosis, it should be mentioned that several antiplatelet drugs (aspirin, clopidogrel) have known anti-inflammatory properties. In addition, heterotypic interactions of platelets with leukocytes promote their recruitment to inflamed endothelium propagating atherogenesis [17].

An atherosclerotic endothelium is prone to arterial thrombotic disease leading to MI and ischemic stroke. The crucial role of platelets in atherothrombosis has been confirmed by epidemiological studies showing the ability of anti-platelet drugs, mainly aspirin, to prevent MI, cardiovascular death, ischemic stroke and total mortality as a secondary prevention [18]. While the role of antiplatelet therapy as a secondary prevention of ischemic diseases is well established, its role in the primary prevention is still debated. This is partly because of the conflicting results of the clinical studies [19, 20] but also because of the increased risk for bleeding.

Several pro-clinical and clinical studies have also shown that platelets play a central role in malignancy and metastatic potential of cancer cells. The crosstalk between platelets and malignant cells in a reciprocal way leads to the progression of malignant lesions but also to cancer-induced thrombosis. Chronic activation of platelets drives immune cells to the site of tumor sustaining the development of malignant lesions. Platelets contribute to an invasive phenotype of cancer cells and protects them during their presence in the circulation conferring by this way to their metastatic potential. On the other hand, cancer cells can induce platelet aggregation, activation and release of their granule content among which growth and angiogenic factors support tumor survival and promotion [9]. Randomized clinical trials with aspirin have shown that doses of 300 mg/day are effective in the primary prevention of colon-rectal cancer [21] and less efficient for other types of cancer [22, 23]. However, recent randomized controlled trials failed to show a protection of a low dose of aspirin for overall cancer incidence and cancer-related mortality [24]. Similar conflicting results were obtained when dual antiplatelet therapy with aspirin plus purinergic receptor, P2Y12 inhibitors was studied [25, 26].

Chronic activation of platelets is also a risk factor for neurological disorders such as Alzheimer’s disease (AD). Higher platelet activation is associated with faster cognitive decline in AD [27] and CAD patients [28]. Activated platelets can metabolize amyloid-beta (A$\beta$) precursor protein into its metabolites contributing by this way to the deposition of A$\beta$ while the involvement of platelet to chronic vascular inflammation is another mechanism that may link activated platelets with AD [29].

Type 2 diabetes mellitus (T2DM) is probably the most characteristic non-communicable disease which is characterized by hyperactive and dysfunctional platelets while at the same time is a risk factor for all the aforementioned chronic diseases. The diabetic platelet is characterized by an increased platelet turnover, increased intracellular Ca${}^{2+}$, upregulation of P2Y12 signaling and oxidative stress, increased P-selectin and glycoprotein expression, impaired response to NO and prostacyclin, resistance to aspirin, decreased membrane fluidity. The main drivers of these cellular abnormalities are the metabolic disturbances characterizing T2DM such as insulin resistance, hyperglycemia, diabetic dyslipidemia, obesity and subclinical inflammation [30, 31].

It is therefore obvious that chronically activated, dysfunctional platelets mediate the progression of the majority of non-communicable diseases in a pleiotropic fashion. Therefore, antiplatelet therapy remains an attractive therapeutic means which however hasn’t reached the expected targets according to the preclinical studies. A significant proportion of the patients receiving antiplatelet therapy acquire some sort of resistance to it while at the same time bleeding remains an important side effect. It is therefore obvious that the consumption of foods demonstrating antiplatelet activity may be a less drastic but on the other hand a more sustainable way of achieving daily antiplatelet therapy, either alone or in combination with antiplatelet drugs.

Olive oil is the oil obtained from the fruit of olive trees (Olea europaea L.). According to the Annex VII (Part VIII) of Regulation (EC) No 1308/2013 of the European Union, OOs can be classified into 8 quality categories, namely extra virgin olive oil (EVOO), virgin olive oil (VOO), OO, olive pomace oil, lampante, refined olive oil, refined-olive pomace oil and crude pomace OO. Among them the first four types of OOs are the edible ones. EVOO has the highest quality in terms of organoleptic characteristics, low acidity (0.8%) and it is obtained directly from the fruit by mechanical means only. VOO is similarly obtained by mechanical means only but it is inferior to EVOO mainly to its higher acidity (2%) and lower physicochemical and sensory properties. OO is a blend of refined olive oil with small amounts of either EVOO or VOO. It is inferior to EVOO and VOO and its acidity should not exceed 1%. Finally, olive-pomace oil is a mixture of refined olive pomace oil with small amounts of EVOO or VOO. It has the lowest quality of all edible oils, its acidity should not exceed 1% [32].

All OOs are composed of a saponifiable and an unsaponifiable fraction. The saponifiable fraction accounts for almost 98% of OO composition and it consists of triglycerides, diglycerides, monoglycerides and free fatty acids (FA). It is characterized by a high percentage of MUFA, mainly oleic acid (55–83%), and lower amounts of polyunsaturated fatty acids (PUFA), mainly linoleic acid (3–21%) and saturated fatty acids (SFA) (palmitic acid: 8–20% and stearic acid: 0.5–5%). However, the unsaponifiable fraction of OOs, although low in quantity, confers the beneficial biological properties of OO due to the presence of a milieu of over 250 biologically active phytochemicals. These include squalene (800–8000 mg/Kg), $\beta$-carotene and lutein (4–10 mg/Kg), sterols (1000–3000 mg/Kg), triterpenic compounds (200–300 mg/Kg), tocopherols/tocotrienols (250–350 mg/Kg) and polyphenols (200–1500 mg/Kg) [33].

Olive oil phenolics are secondary plant metabolites, produced by olive cells to combat pests and bacteria. They are responsible for the aroma and flavor of OOs but they also confer oxidative stability to the oil. Tocopherols/tocotrienols are the main lipid soluble phenolic compounds while at least 40 hydrophilic phenolic compounds have been identified in OO. According to their chemical structure can be classified into phenolic acids (photocatechouic, vanillic acid), phenolic alcohols (hydroxytyrosol-HT, tyrosol-TYR), secoiridoids (oleuropein OE, oleuropein aglycone OEA, ligstroside aglycone), flavonoids (luteolin-LU, apigenin-AP), lignans and hydroxy-isochromanes. EVOO contains the highest amount of total phenolics (200–800 mg/Kg), followed by VOO (60–350 mg/Kg) and refined OO (60–200 mg/Kg). In terms of biological properties and quantity, TYR, HT, their secoiridoid derivatives oleacein and oleocanthal, OEA and ligstroside are the most important phenolic compounds in OO [32, 33, 34].

Olive oil is probably the main cardioprotective component of the MD according to the results of observational and dietary intervention studies [35]. The emblematic PREDIMED study showed that the adoption of a MD type of diet with EVOO as the main fat source resulted in a 30% reduction of CVDs (acute MI, stroke, death from CVD causes) [36]. The cardioprotective properties of OO were initially attributed to its high MUFA content. However, recent meta-analyses of cohort studies have shown that the dietary intake of either total fat, SFA, MUFAs or PUFA were not associated with the risk of cardiovascular disease while the dietary source (animal or vegetable) of MUFA may play a role on the association of MUFA and CVDs [37, 38]. Therefore, nutritionists seeked alternative explanations for the beneficial properties of EVOO. Among all phytochemicals of OO, its unique phenolics seems to be responsible for the majority of its cardioprotective properties, although the research is ongoing. Numerous in vitro and preclinical studies have demonstrated the pleiotropic actions of phenol extracts or individual EVOO phenolics on inflammation, endothelial dysfunction, coagulation, dyslipidemia and oxidative stress [32, 39]. These pathological mechanisms underlie the majority of chronic disease and their regulation by OO phenolics can partly explain the health promoting properties of OO. Although certain molecular targets are linked to certain phenolics it seems that they act in a more general way as nutritional hormetins triggering lifelong adaptive processes preventing by this way the biological aging [40].

Taking into account the central role of platelets in the pathogenesis of chronic diseases, especially CVDs, and the cardioprotective properties of OO as part of the MD, in the next chapters of this review we investigate the relationship of OO phenolics with platelet functions as another mechanistic explanation of the OOs’ health benefits.

We searched the literature in PubMed and Scopus databases until March 2022 using a combination of the following keywords: olive oil, virgin olive oil, extra virgin olive oil, phenolics, polyphenols, phenols, tyrosol, hydroxytyrosol, oleocanthal, oleuropein, oleacein, secoiridoids, platelets, platelet aggregation, platelet adhesion, platelet rich plasma (PRP). We included in vitro/ex vivo studies, animal studies and dietary intervention studies in humans. At least one assay of platelet functionality should be included in these studies. In cell and animal studies, the active ingredients should be either EVOO or VOO with a known amount of total phenolics, isolated phenolic extracts with some sort of phenolic characterization, isolated or synthetic phenolic molecules and synthetic derivatives of them. For the human studies, we selected dietary intervention trials where at least one intervention group included the administration of EVOO, VOO with a known amount of total phenolics. We have to mention that in many articles OO was administered as a placebo to clinical trials investigating the health properties of PUFAs. Those studies were excluded from this review since the OOs were not characterized for their phenolic composition.

Cellular studies in isolated platelets is a convenient way for screening the antiplatelet properties of bioactive compounds and investigating the possible mechanisms by which these compounds exert their antiplatelet effects. The characteristics and the main outcomes of the in vitro studies exploring the modulatory role of OO phenolic extracts on platelet functions are presented in Table 1 (Ref. [41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53]).

The majority of the studies have used PRP or whole blood (WB) from healthy human donors as the source of platelets. The most common platelet assays that are utilized in those studies are agonist-induced platelet aggregation and thromboxane B2 (TxB${}_2$) production. The endogenous agonists more frequently used for the ex vivo activation of platelets were COL, adenosine diphosphate (ADP), thrombin (THR) and arachidonic acid (AA).

The studies so far have shown that several phenolic compounds and extracts of OO are able to inhibit platelet activation. The strongest evidence exists for HT since many cell studies have demonstrated its anti-aggregating potential. HT was able to inhibit COL-, ADP- and Thrombin Receptor Activating Peptide (TRAP)-induced platelet aggregation in human PRP [46, 47, 48], THR-induced aggregation in washed human platelets [49], COL-, ADP- and AA-induced aggregation in WB of healthy donors [50, 51]. Only one study could not demonstrate the inhibitory potential of HT at 100 $\mu$M against COL-induced platelet aggregation in the WB of healthy donors. Among the other phenolic compounds, OE [46, 48, 49, 52], OEA [49], TYR [49], deacetoxy oleuropein aglycone (DHPG) [47, 53] and hydroxytyrosol acetate (HTA) [50, 53] have shown anti-aggregating properties in at least one study. The majority of studies pre-incubated platelets for 10 min before the addition of the aggregating agents since this time seems to maximize inhibition of platelet aggregation [46].

Phenol-rich extracts from OOs [49], waste water (WW) [46], alperujo (a solid by-product of OO extraction) [47] and steam-treated alperujo [53] also demonstrate potent antiplatelet properties. Phenolic extracts seems to have a better antiplatelet action compared to pure compounds probably due to the synergistic effect of the phenolics. A phenolic extract from WW containing a mixture of phenolic compounds with a concentration comparable to pure HT (100 $\mu$M) has a better anti-aggregating activity compared to isolated compounds [46]. In addition, an alperujo extract containing HT and DHPG has a small but significant antiaggregating activity against COL and TRAP-induced aggregation of human platelets while the isolated compounds have negligible effect [47]. Future studies should investigate whether OO by-products could serve as a rich, cheap, source of polyphenols able to formulate nutraceuticals and supplements with cardioprotective properties [54].

A couple of studies have shown that the ability of the phenolic compounds to inhibit platelet aggregation depends on the aggregating agent. Two isochromans found in EVOO inhibited the platelet response to AA and COL, but not to ADP. The authors of this study support the hypothesis that these isochromanans, through their radical scavenging activity can inhibit only the initial, redox sensitive phases of platelet activation, induced by COL and AA [41]. HT, HTA and aspirin showed a lower inhibition (higher IC${}_{50}$ values) of ADP-induced platelet aggregation compared to COL-induced platelet inhibition probably because these phenolics interact with the COL-induced eicosanoid production rather than the lowering of Ca${}^{2+}$ which mainly mediates ADP-induced platelet aggregation [50].

The ability of the phenolic compounds to inhibit platelet activation depends on the biological source of platelets indicating that phenolics can interact with platelet either directly (PRP) or indirectly through the production of platelet modulating agonists/antagonists from other blood cells. For example, HT and OE could not inhibit COL-induced platelet aggregation in WB [42] while the same molecules inhibited COL-induced aggregation in human PRP [46]. This was demonstrated more emphatically in the study of Correa et al. [50], which showed that HTA is a better inhibitor of HT in WB while they showed similar inhibitory properties in PRP. Both phenolics could inhibit platelet thromboxane A${}_2$ (TXA${}_2$) production but HTA could also indirectly inhibit platelet aggregation in WB by stimulating NO production by leukocytes [50].

The structure of the phenolic compounds, seems to determine their antiplatelet properties. In a study comparing different phenolic compounds HT was a better inhibitors of COL-induced platelet aggregation compared to OE, LU, AP [46]. Similarly, in the study of Togna et al. [41], only the most lipophilic isochroman derivatives showed a significant inhibitory potential against COL-induced aggregation. Whether the clinical condition of the donor could affect the antiplatelet properties of the phenolic compounds has not been investigated. Just one study showed that OE has a slightly better antiaggregating activity in normal platelets compared to diabetic ones at low concentrations (10 $\mu$M) [52].

The ability of OO phenolics to inhibit platelet aggregation can be attributed to several mechanisms but the inhibition of TxA${}_2$ production from eicosanoids seems to be the most relevant one. HT and HTA can inhibit COL- and THR-induced TxB${}_2$ production in PRP and WB of healthy volunteers and the production of TxB${}_2$ and 12-hydroxy-5,8,10,14-eicosatetraenoic acid (12-HETE) during clotting [46, 50]. Taking into account that HT and HTA can also inhibit prostacyclin production in WB it seems than the inhibition of TxA${}_2$ production is due to inhibition of cyclooxygenase rather than TxB synthetase [50]. OO isochromans can also inhibit arachidonic acid mobilization from platelet membrane phospholipids induced by THR and COL indicating a direct inhibition of PLA2 [41]. OE and HT can also interfere with THR-induced calcium mobilization. This effect was greater in platelets from T2DM subjects. The antioxidant properties of OE and HT could explain the reactive oxygen species (ROS)-mediated entry of Ca${}^{2+}$ into activated platelets [52]. OE (10 $\mu$M) and HT (100 $\mu$M) could also inhibit THR-induced protein tyrosine phosphorylation both in healthy donors and diabetic subjects [52]. Phenolic extracts from OO with a high and low phenolic content were able to inhibit the degrading enzyme of cyclic 3’ 5’ adenosine monophosphate (cAMP), cAMP phosphodiesterase, preventing by this way the lowering of intraplatelet cAMP levels, which is a crucial mechanism of platelet activation [49]. Whether the antioxidant properties of phenolics could explain their antiplatelet actions is again a controversial and not well-studied issue. The radical scavenging activities of a series of OO phenolic compounds are not correlated with their ability to inhibit platelet aggregation in PRP of healthy volunteers [46]. On the other hand, Italian EVOOs could downregulate the activation of platelet nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (NOX) activation induced by AA. NOX can induce the production of ROS which in turn contribute to the production of the platelet agonist 8-iso-PGF2 alpha. The polyphenol and tocopherol content of the EVOOs correlated with the inhibitory activity probably through the scavenging of H${}_2$O${}_2$ [53]. HT (100 $\mu$M) and OE (500 $\mu$M) could also reduce the secretion of pro-atherogenic molecules from platelet granules such as platelet-derived growth factor (PDGF)-AB and soluble CD40L (sCD40L) [48] and suppress the COL-stimulated phosphorylation of heat shock protein 27 (HSP27) in human platelets, leading to the reduction of phosphorylated-HSP27 release which is able to activate the release of pro-inflammatory cytokines from macrophages [48]. HT is also inhibitor of the platelets’ monoamine oxidase isoform (MAO)-B. Taking into account that the primary structure of MAO-B in platelets and brain is almost identical this indicates a possible role of HT in alleviating Parkinson disease symptoms [43]. Finally, proteomic analysis revealed that phenolic-rich extracts, specifically an alperujo extract containing HT and DHPG, have acute effects on platelet proteome regulating multiple pathways involved in platelet aggregation, coagulation, signalling and apoptosis [47].

Several studies have shown that the derivatization of OO phenolics may lead to molecules with better antiplatelet activity. The acetylated derivatives of TYR, which are also found in olives and OOs are two orders of magnitude more potent inhibitors of platelet-activating factor (PAF)-induced aggregation in washed rabbit platelets than tyrosol itself [44]. Acetylated HT has a better inhibitory activity against COL, ADP, AA-induced aggregation in WB and PRP compared to HT [50]. Hydrophobic derivatives of HT (alkyl HT ethers) are better inhibitors of COL, AA, ADP-induced platelet aggregation in WB of healthy volunteers although they similarly inhibited TxB${}_2$ production [51].

In conclusion, OO phenolic compounds, mainly HT and OE show antiplatelet properties affecting multiple pathways of platelet activation. However, the concentrations used in the in vitro experiments can not be achieved in blood even under postprandial conditions. Therefore, in vitro experiments may facilitate a rapid screening and may give some hints on the mechanism involved in the antiplatelet properties of OO phenolics but from a nutritional point of view the outcomes of these experiments should be translated to real clinical settings with caution.

The studies utilizing animal models to investigate the antiplatelet properties of EVOO, VOO, phenolic extracts or pure phenolic compounds are shown in Table 2 (Ref. [55, 56, 57, 58, 59, 60, 61]). The majority of the studies used rats as the preferred animal species while only one study used New Zealand rabbits. The pathological phenotypes induced by the researchers were either atherosclerosis (induced by an atherogenic diet) [55], diabetes (induced by streptozotocin) [56, 57] while the rest of the intervention studies were conducted in healthy animals. The investigators administered either VOO [55, 56, 58], EVOOs with different concentrations of phenolics [59], HT [57, 60], HTA [60], DHPG [57] and alkyl ether derivatives of HT [61]. The interventions lasted from 0 days (acute administration) to 3 months while the experimental oils/molecules were administered by oral gavage in the majority of the studies.

The most important outcome of these studies is that the platelet inhibition by OO phenolics is accompanied by improvements of the pathological phenotypes. OO could slow atherosclerosis progression concomitantly with inhibition of platelet hyperreactivity. Specifically, when an atherogenic diet was accompanied with the supplementation of VOO (15%) to New Zealand rabbits the atherosclerotic lesions of the arteries were improved concomitantly with the lowering of platelet aggregation, TxB${}_2$ and lipid peroxide production [55]. VOO could also attenuate ex vivo brain damage concomitantly with inhibition of platelet hyperreactivity. Ex vivo brain damage in fresh brain slices of rats after hypoxia-reoxygenation was attenuated with the administration of two doses of VOO and this improvement was accompanied with the lowering of ex vivo platelet aggregation and TxB${}_2$ production in rats. It seems that platelet inhibition could play a key role in the improvement of throbogenesis and brain ischemia and may play a role in the prevention of ischemic stroke [58]. Finally, VOO can act synergistically with acetylsalicylic acid (ASA) to inhibit vascular complications in the retina of diabetic rats. Administration of VOO (0.5 mL/Kg/day per os) attenuated the activation of platelet activity induced by diabetes (platelet aggregation, TxB${}_2$ production, reduction of aortic prostacyclin and NO production). This effect was accompanied by improvements of retinal ischemia measured as the retinal surface area occupied by peroxidase-permeable vessels [56].

The dosage of the phenolic compounds is a major determinant of their antiplatelet properties. HT and HTA inhibited COL-induced platelet aggregation in WB of rats in a dosage range of 1–100 mg/Kg/day. Their ability to inhibit platelet aggregation was positively correlated with the dosage of the phenolic compounds [60]. Priora et al. [59] compared the antiplatelet effects of 3 OOs with a similar fatty acid (FA) content but different amounts of minor polar compounds (MPC). They conducted both acute (one dose) and subacute (12 d) experiments demonstrating the superiority of the EVOO-enriched with MPC to inhibit platelet aggregation. The MPC contained HT, TYR, elenolic acid, OEA but mainly deacetoxy oleuropein aglycone (DACOLAG) and secoiridoid derivatives [59]. The synergism between phenolic compounds could explain the superiority of whole OOs or phenolic extracts against pure phenolic compounds to inhibit platelet activation. A recent study of De La Cruz Cortés et al. [57], clearly showed the synergistic effect of HT and DHPG in similar proportions found in EVOO to attenuate COL-induced platelet aggregation of diabetic rats and to spare prostacyclin synthesis by the aorta.

The investigation of the mechanisms underlying the antiplatelet effects of OO phenolics demonstrated once more the ability of VOO and its phenolics to inhibit TxB${}_2$ synthesis. The administration of VOO inhibited the ex vivo TxB${}_2$ production in an experimental model of atherosclerosis in rabbits [55], in healthy rats in a dose dependent manner [58] and in streptozotocin-induced diabetic rats [56]. HT and HTA, at a dosage scheme of 100 mg/Kg/day inhibited Ca${}^{2+}$-induced TxB${}_2$ synthesis in WB [60]. The ability of EVOO enriched with MPCs to inhibit only reversible platelet aggregation could be explained by the radical scavenging properties of phenolics since ROS mainly affect the activation of fibrinogen receptors in platelets but not the release of granules by platelets [59]. Another indirect mechanism of platelet inhibition is the stimulation of NO and prostacyclin synthesis by the endothelium. VOO stimulated prostacyclin synthesis in the aorta of rabbits fed an atherogenic diet [55]. VOO containing 250 mg/Kg total phenols partially reversed inhibition of Ca${}^{2+}$-induced prostacyclin production by the aortas of streptozotocin-induced diabetic rats. This effect could not be observed by aspirin which inhibits aortic prostacyclin production due to endothelial cyclooxygenase (COX) inhibition [56]. HT and HTA were able to increase ex vivo NO synthesis in a manner proportionally similar to aspirin [60]. An alternative explanation for the indirect antiplatelet properties of phenolics is offered by Priora et al. [59] who showed that administration of EVOO rich in MPCs to rats is able to inhibit ex vivo ADP-induced platelet aggregation in PRP which is accompanied by a reduction of reduced homocysteine (rHcy) levels. Previous studies have shown that rHcy has pro-aggregating properties increasing TxB${}_2$ production in human platelets [62].

The study of De La Cruz et al. [56] has also demonstrated the synergistic effect of VOO containing 250 mg/Kg total phenols with aspirin in inhibiting platelet aggregation in streptozotocin-induced diabetic rats. Treatment of diabetic rats with both aspirin and VOO produces greater inhibition of ADP-induced platelet aggregation compared to the administration of aspirin and VOO separately. Both VOO and aspirin could inhibit TxB production by platelets due to COX inhibition while at the same time VOO phenolics act on aortic endothelium preventing the inhibition of vascular NO synthesis. This effect increases the prostacyclin to TxB ratio which is indicative of the thrombogenic potential of the vascular bed [56].

The animal studies have also proved the observation from cell studies that the increased lipophilicity of phenolic compounds and their derivatives is related with better antiplatelet properties. The administration of the acetylated derivative of HT to rats produces greater inhibition of COL-induced aggregation in WB (ID${}_{50}$: 16 mg/Kg/day) compared to HT (ID${}_{50}$: 48 mg/Kg/day) [60]. Muñoz-Marín J et al. [61] synthesized a series of alkyl ether derivatives of HT which are more lipophilic and stable. Compared to HT, the administration of its alkyl ether derivatives to rats reduced maximum intensity of COL-induced aggregation in a dose and structure dependent manner. The highest inhibition was achieved by hexyl derivative (59%). The alkyl ether derivatives were also better inhibitors of ex vivo Ca${}^{2+}$-induced TxB${}_2$ production in a dose and structure dependent manner. The highest inhibition was achieved by hexyl ether derivatives at 20 and 50 mg/Kg/day [61].

In conclusion, pre-clinical studies in animal models proved the ability of VOO and its phenolic compounds to inhibit ex vivo platelet activation and this inhibition was correlated with improvements of pathological phenotypes related to ischemic diseases. Although the dosage schemes for the VOO (0.25–1 mL/Kg/day) can be realistically achieved in the Mediterranean countries, this is not the case for the administration of pure phenolic compounds which are given in supraphysiological, pharmacological doses. Therefore, the outcomes of the latter studies have a value from a pharmacological point of view rather than from a nutritional perspective.

Numerous studies have already used OO as a control/placebo to dietary interventions investigating the effect of other types of fats and oils (mainly PUFAs) on platelet activation. Olive oil was utilized as a biologically inert oil and as a source of MUFAs. In most cases, no characterization of the phenolic content was provided although, for the majority of the studies the usage of refined OO was implied. These studies are not included in this review since their design does not provide evidence for the ability of OO phenolics to inhibit platelet activation in humans.

A recent observational study demonstrated that obese patients who were frequent consumers of OO ($>$4 times/week) demonstrated lower THR-induced P-selectin expression on the membanes of platelets compared to less frequent consumers (1 time/week). This study clearly showed that increased consumption of OO correlates with lower platelet activation however the types of OO consumed by the volunteers of the study were not identified [63].

The interventional studies with EVOOs/phenolic extracts in humans are presented in Table 3 (Ref. [45, 64, 65, 66, 67, 68, 69, 70, 71, 72]). One study coming from the PREDIMED trial investigated the effect of a MD regime enriched with either EVOO (MD-EVOO) or nuts (MD-Nuts) on platelet count. Both MD interventions were able to keep platelet counts within the normal range taking into account the trend for an increase of platelet count over time. In addition, both MD interventions decreased the risk of developing thrombocytopenia while blunted the association of thrombocytopenia with all-cause mortality. However, since the beneficial alterations of platelet counts were observed with both MD interventions, they can’t be attributed to EVOO phenolics [64]. Another study from a small subcohort of the PREDIMED [65] compared the effect of interventions on the concentration of plasma microvesicles (MVs) from different cellular origins. Such MVs have both pro-coagulant [73] and pro-inflammatory actions [74] and they are found in higher concentrations in patients suffering from a major adverse cardiovascular event [75]. The study demonstrated that only the MD-Nut arm was able to reduce platelet derived MVs (PAC-1+/AV+ and CD62P+/AV+) and after one year follow up platelet-derived microparticles were lower in the nuts group after compared with the other two groups. Therefore, it seems that EVOO does not have a role on this mechanism although further studies are required.

However, valid conclusions on the direct effect of OO phenolics to platelet functionality can only be drawn by studies consisting of intervention groups differing in the amount of the administered phenolics and keeping FA profile similar. A small cross-over trial from Oubiña et al. [66] supplemented post-menopausal women with either EVOO (total polyphenols: 108 mg/Kg) or high-oleic acid sunflower oil (HOSO) (total polyphenols: 25 mg/Kg). The two oils had similar oleic acid content but HOSO contained higher concentrations of $\alpha$-tcocpherol, total sterols and much lower concentration of squalene. Despite the modest concentration of total polyphenols in the EVOO, its administration lowered TxB${}_2$ production in PRP induced by ADP compared to HOSO. However, no differences were observed in the concentration of TxB${}_2$ and 6-keto-prostaglandin-F1$\alpha$ (a stable analogue of prostacyclin in 24 h urine samples). The authors attributed this difference to the higher amount of linoleic acid in the HOSO diet, however according to the aforementioned cell and animal studies, the increased content of polyphenols in EVOO may play a role for the observed antiplatelet effect of EVOO [66]. Stronger evidence for the ability of EVOO to attenuate chronic platelet activation was obtained by the cross-over VOLOS study. This study demonstrated the ability of EVOO, at doses (40 mL/day), comparable to those consumed by the Mediterranean populations to reduce serum TxB${}_2$ production by 20%. EVOO also improved plasma antioxidant capacity raising the possibility that the redox modulation of EVOO polyphenols may play a role on EVOOs antiplatelet properties [67]. The outcomes of VOLOS were confirmed by a very small, single-armed study in Type 1 diabetic patients who are administered with a HT-rich phenolic extract from olive mill wastewaters for 4 consecutive days corresponding to the amount of HT typically found in 25–50 g of olive fruit. A significant reduction (47%) of serum TxB${}_2$ after 30-min clotting was observed at the fourth day indicating the ability of dietary HT to attenuate platelet aggregability in vivo [68]. Another study in female patients with fibromyalgia compared an EVOO (248 mg/Kg total polyphenols) with an ROO of similar fatty acid profile but poorer in polyphenols (152 mg/Kg total polyphenols) to affect cardiovascular risk factors [69]. The EVOO did not have an effect of platelet count, mean platelet volume (MPV) and platelet distribution width (PDW) while ROO lowered PDW compared to EVOO. These results are in contrast with the aforementioned PREDIMED study where MD-EVOO could modulate platelet counts [64], however we have to mention that this study was much smaller, the intervention lasted only 3weeks and the difference in the total polyphenol content between EVOO and ROO was rather small.

The postprandial state is a daily, metabolically challenging period for the human metabolism and several homeostatic mechanisms are activated for the proper handling of macronutrients. Energy dense meals can stress those mechanisms leading to a transient, postprandial hyperglycemia, hyperlipidemia often accompanied by proinflammatory processes and oxidative stress [76]. Taking into account that OO is usually consumed as part of a meal and that postprandial dysmetabolism is a risk factor for atherogenesis, the ability of EVOO to handle postprandial oxidative stress may give nutritional relevant explanations for its cardioprotective properties. Under this perspective, Carnevale et al. [70] conducted two postprandial studies comparing the effect of a typical Mediterranean lunch with or without corn oil (CO) and EVOO on platelet ROS production, platelet lipid peroxidation and activation of NOX. The studies showed that EVOO almost completely blunted the postprandial increases of these oxidative stress indices. In vitro studies have shown that the postprandial antioxidant effects of EVOO can be attributed to both its vitamin E and polyphenol content [70]. A similar follow up study was conducted a few years later from the same group. Just before a typical lunch healthy volunteers received either placebo or 20 mg OE, a precursor of HT in OO. The outcomes of the study clearly showed that OE could minimize postprandial lipid peroxidation in platelets and p47phox phosphorylation. The latter is the cytosolic subunit of platelets’ Nox2 and its phosphorylation leads to Nox2 activation. The in vivo results were confirmed by in vitro data showing that HT reduced p47phox phosphorylation and isoprostane formation [45]. The results of the aforementioned studies indicate that platelets are an important source of ROS postprandially and a target for the antioxidant phenols of EVOO. Moreover, the downregulation of Nox2-derived oxidative stress can improve hyperglycemia by increasing the bioavailabilty of incretins as the authors showed in the same study [45]. Finally, a study of Agrawal K et al. [71] comparing the acute effect of 40 mL EVOO rich either in oleocanthal/oleacein or TYR has shown that oleocanthal/oleacein rich EVOOs can acutely (in 2 h) reduce COL-induced platelet activation without affecting oxylipin production by platelets. The inhibition of platelet aggregation correlated with the oleocanthal content while the inhibition of eicosanoid production with the EVOO total phenolic content. Oleocanthal is strong inhibitor of COX acting similarly to aspirin in inhibiting platelet activation [71].

A relatively recent research trend in the field of nutrition is the enrichment of foods with bioactive extracts or pure compounds in an attempt to formulate food products with pleiotropic bioactions. Under this perspective the group of Widmer et al. [72] produced an EVOO enriched with green tea catechins (EGCG) and compared it with the non-enriched EVOO in early-atherosclerotic patients. Among other parameters, they measured platelet count and they found that in the combined EVOO groups OO supplementation could lower platelet count at the end of the supplementation period with no differences between groups. Subclinical inflammation and cytokines such as interleukin 6 (IL-6) are potent thrombopoietic factors [77] so the observed reduction in platelet counts could be attributed to an improvement of subclinical inflammation. Although white blood cells (WBC) counts and plasma soluble intercellular adhesion molecule-1 (sICAM) were reduced by both EVOOs other inflammatory markers were not affected significantly implying that other mechanisms could be involved in the modulation of platelet counts after the consumption of EVOO [72].

Conclusively, the data from human studies are promising but again not persuasive mainly because the majority of studies are under-powered and heterogenous concerning the type of the supplemented EVOOs/extracts, the dosage schemes, the subjects and the assays of platelet functionality. We clearly need more large cohort studies, like the ones coming from the PREDIMED trial, where the modulation of platelet indices by OO phenolics should be linked to hard end-points. If so, then the notion that the antiplatelet properties of OO phenolics may be partly responsible for their cardioprotective properties will be strengthened.

A plethora of studies so far clearly demonstrate that the hyperactivity of platelets, especially under insulin resistance conditions (obesity, metabolic syndrome, diabetes), can augment the pathogenesis of several non-communicable disease including CVDs, cancer and neurological disorders. At the same time recent, large, cohort studies and meta-analyses have shown that the MD and its main component, EVOO, are able to beneficially modulate the progression of those disease. It is also more than clear, that the phenolic content of OO (both its quantity and its composition) is responsible, either alone or in combination with the other micrconstituents of OO for its biological properties. Therefore, the notion that OO phenolics have antiplatelet actions and this is an important mechanism by which EVOO exert its health protective actions is justified. The cellular and animal studies clearly show that EVOO phenolics, mainly HT, are able to inhibit the activation of platelets induced by several endogenous agonists and pathologies. They also showed that OO by-products are a promising source of phenolic extracts which can be utilized in the development of supplements and nutraceuticals with promising antiplatelet actions. In addition, they demonstrated that small chemical modifications of OO phenolics towards increased lipophilicity may lead to the development of derivatives with augmented bioavailability and antiplatelet actions. However, the outcomes of the pre-clinical studies are difficult to be translated to humans mainly because the dosages and the chemical forms of the phenolics used to these studies are much higher and different to that found in human circulation. Despite the heterogeneity of the few human trials on the field so far, the results are promising showing that EVOO can exert antiplatelet actions in real, acute or long-term, trials and at least part of this antiplatelet action can be attributed to the phenolic content of EVOOs. However, we clearly need better, well-powered studies to give answers on the ideal composition of phenolics in EVOO, the daily doses that they should be administered according to the pathology of the population and mechanistic issues concerning the bioavailability and the mode of action on platelets. Last but not least the antiplatelet properties of OO phenolics should be correlated with hard end-points. In any case, the antiplatelet properties of OO phenolics is a promising, emerging mechanism which may explain some of the health properties of EVOO and the MD.

A$\beta$, Amyloid beta peptide; AA, Arachidonic Acid; AD, Alzheimer’s disease; ADP, Adenosine diphosphate; AP, Apigenin; ASA, Acetylsalicylic acid; BW, Body Weight; cAMP, cyclic 3’ 5’ adenosine monophosphate; cGMP, Cyclic guanosine monophosphate; CO, corn oil; COL, Collagen; COX, cyclooxygenase; CVD, Cardiovascular Disease; DACOLAG, Deacetoxy oleuropein aglycone; DHPG, 3,4-dihydroxyphenylglycol; DR, diabetic rats; EGCG, enriched with green tea catechins; EVOO, Extra virgin olive oil; FA, Fatty acids; GAE, Gallic acid equivalents; HC, high MPC concentration; 12-HETE, 12-hydroxy-5,8,10,14-eicosatetraenoic acid; HHT, 12-hydroxy-5,8,10-heptadecatrienoic acid; HOSO, High-oleic acid sunflower oil; HSP, Heat Shock Protein; HT, Hydroxytyrosol; HTA, Hydroxytyrosol acetate; HVA, Homovanillic alcohol; ID${}_{50}$, Inhibitory dose for 50% inhibiton; IL-6, interleukin 6; LC, low MPC concentration; LDL, low density lipoprotein; LFD, low fat diet; LU, Luteolin; MAO, Monoamine oxidase; MD, Mediterranean Diet; MI, myocardial infarction; MPC, Minor Polar Compounds; MPV, Mean Platelet Volume; MUFA, Monounsaturated fatty acids; MVs, Microvesicles; NADPH, Nicotinamide adenine dinucleotide phosphate; NDR, Nondiabetic rats; NETs, Neutrophil Extracellular Traps; NLD, normolipemic diet; NO, nitric oxide; NOX, NADPH oxidase; OE, Oleuropein; OEA, Oleuropein aglycon; OO, Olive Oil; oxLDL, oxidized LDL; PAF, Platelet-activating factor; PDW, Platelet distribution width; PDE, phosphodiesterase; PDGF, Platelet-Derived Growth Factor; PE, Phenolic extracts; PGF, prostaglandin; PLT, platelet; PRP, Platelet rich plasma; PPF, Polymeric phenolic fractions; PUFA, Polyunsaturated fatty acids; rHcy, reduced homocysteine; ROS, reactive oxygen species; ROO, Refined Olive Oil; QU, flavonoids quercetin; SA, Sodium Arachidonate; sICAM, Plasma soluble intercellular adhesion molecule; sCD40L, Soluble CD40L; sNOX2-dp, soluble NOX2–derived peptide; SFA, saturated fatty acids; SFAED, saturated fatty acid-enriched diet; TPL, Total Phosholipids; T2DM, Type 2 Diabetes Mellitus; THR, Thrombin; TRAP, Thrombin Receptor Activating Peptide; TxA${}_2$, Thromboxane A2; TxB${}_2$, Thromboxane B2; TYR, Tyrosol; WB, Whole Blood; WBC, white blood cells; WW, Waste Water.

TN contributed to the conception, design and acquisition of the review data, was involved in revising critically the study’s content and also gave his final approval for the version to be published. MEK contributed to the acquisition of the review data and she drafted the manuscript. All authors read and approved the final manuscript.

Not applicable.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.