The Effects of Local Anesthetics on Perioperative Coagulation, Inflammation, and Microcirculation

Major surgery is associated with a hypercoagulable and proinflammatory state that persists into the postoperative period (1,2). Perioperative inflammatory responses to trauma can trigger hypercoagulability, especially in patients undergoing vascular surgery, and are associated with vasoocclusive and thromboembolic events—major causes of postoperative morbidity and mortality (3–5). The mechanisms for these effects are poorly understood, but hypercoagulability seems to originate from what is known as the “stress response” to major surgery (4,6). Postoperative changes occur in all aspects of the coagulation system, including increased plasma levels of coagulation factors (1), decreased levels and more rapid inactivation of endogenous coagulation inhibitors (7), enhanced platelet reactivity (8), and impaired fibrinolysis (9). In addition, increasing attention has been given to the close link between hemostasis and inflammation (10), which is minutely influenced by general anesthesia with parenteral opioids (3). However, recent evidence suggests that regional anesthesia has a protective effect against the perioperative stress response. The beneficial effects of the epidural administration of local anesthetic (LA) have been attributed to the changes in physiology induced by neuraxial anesthesia and better pain management (5,11). However, as discussed below, there are hints that the pharmacodynamic effects of LA itself may contribute to these effects. This review is arranged according to the different effects of LAs—first on the hypercoagulable and inflammatory responses to surgery, and second on microcirculation. Stress Response Surgical trauma and stress generate a status that is characterized by vasodilation, increased vascular permeability (inflammation), and sensitization of nociceptors (primary pain) (12). This is a result of a local release of multiple chemical substances, e.g., substance P, bradykinin, serotonin, and prostaglandins (PgE) from a cascade of arachidonic acid metabolites from afferent nerve endings (Fig. 1) (12). In addition to the local response and transmitter release, surgical stress stimulates sympathetic division of the autonomic nervous system. Postganglionic fibers of the sympathetic nervous system secrete norepinephrine as the neurotransmitter. These norepinephrine-secreting neurons are classified as adrenergic fibers. Responses evoked by autonomic sympathetic nervous system stimulation are shown in Table 1. Both responses to surgical stress—the release of neuroendocrinic hormones and the local release of transmitters—are important for the subsequent course of healing and postoperative outcome.Figure 1: Local response to surgical trauma. Surgical trauma and stress result in a local release of multiple chemical substances, e.g., substance P (sP), bradykinin (BK), serotonin (5-HT), and prostaglandins (PgE) from a cascade of arachidonic acid metabolites from afferent nerve endings. HPETE = hydroperoxy-eicosatetraenoic acid. Modified from Ref. 12.Table 1: Responses Evoked by Autonomic Sympathetic and Parasympathetic Nervous System StimulationPerioperative pain control using regional anesthesia techniques may be a powerful tool for reducing perioperative stress (13). Reduced incidences of vascular graft occlusion and thromboembolic complications have been demonstrated, along with decreased perioperative coagulability, when epidural anesthesia and analgesia were compared with general anesthesia (5,14–16). Serum markers that may reflect the humoral stress response (i.e., catecholamines, corticotropin, thromboxane A2 TXA2, and antidiuretic hormone) are decreased by epidural blockade (17–19), but not by general anesthesia (20). Most authors have interpreted the better effectiveness of postoperative pain management with epidural anesthesia as being the underlying reason for the diminished stress response (5,11) compared with intermittent on-demand opioid analgesia. However, stress-mediated increases in serum variables, such as angiotensin (secondary to a more intense renin activity), catecholamines, and TXA2, activate platelets (21,22) and can be prevented by regional anesthesia. A reduction in the stress response with epidural anesthesia and analgesia may therefore indirectly affect platelet aggregation (20). In addition to this indirect effect, LAs directly influence some of the receptors and their signaling pathways i.e., PgE2 (EP1) (23), TXA2(24), and lysophosphatidic acid (25) that are involved in inflammatory processes, platelet activation, nociception, and peripheral pain. A pronounced inhibition of these receptors by LAs in plasma, absorbed from the epidural space, has been demonstrated, and this has been confirmed in other studies and in several experimental models (26,27). The exact mechanisms are unclear, but one effect is at the G protein, for receptors coupled to Gα(q) proteins (28). Other potential mechanisms are possible, e.g., a reduced variables release or direct LA effects on proteins, or within their signaling cascades. Coagulation Factors and Fibrinolysis Several studies have shown that epidural anesthesia reduces the incidence of thrombotic events (14,15) and is associated with beneficial effects on postoperative outcome (5). These effects are accompanied by a decrease in intraoperative blood loss (14,15), and may be attributed either to physiologic changes induced by neuraxial anesthesia or to pharmacologic effects of LA on the coagulation system. Some putative direct effects of epidurally administered LA on coagulation and fibrinolysis variables have been investigated by determining parts of the coagulatory and fibrinolytic pathway activity. The perioperative increases of factor VIII and von Willebrand factor—the latter mediating attachment of platelets to the vascular wall—could in part contribute to perioperative thrombosis (29). Bredbacka et al. (30) reported less pronounced releases of factor VIII and von Willebrand factor with epidural anesthesia compared with general anesthesia in patients undergoing abdominal hysterectomy, whereas factors II and X remained unchanged in a study by Henny et al. (31). Variables for estimating fibrinolytic activity were measured in the blood of patients undergoing elective lower extremity vascular reconstruction. These studies demonstrated, first, that epidural anesthesia enhances fibrinolytic activity by preventing the postoperative release of plasminogen activator inhibitor-1 protein (4); second, a larger baseline concentration of plasminogen activators; and third, an increased capacity of the venous endothelium to release plasminogen activators (14). In addition, antithrombin III, the principal inhibitor of thrombin activity, which progressively decreases during the early postoperative period (29), returns to preoperative concentrations more rapidly in patients receiving epidural LA (2). All of these studies suggest that the epidural administration of LA is able to reverse, or at least limit, perioperative hypercoagulability by preventing the release of procoagulatory mediators, by inhibiting their signaling pathways, or through increased fibrinolysis. Platelet Aggregation Platelet aggregation is one of the most important steps in hemostasis. Several studies published in the 1980s demonstrated an inhibitory effect of epidurally administered LAs on platelet aggregation (14,31,32). In a study conducted by Hollmann et al. (33), a new device (Clot Signature Analyzer®; Xylum Corp., Scarsdale, NY) was used to evaluate perioperative hypercoagulation. Clot signature analysis is a novel technique for determining interactions between platelets, proteins, and collagen surfaces. It provides a simulated vascular flow environment for assessing both the platelet and coagulation activities of native whole blood. In this study, the use of epidural anesthesia prevented immediate postoperative hypercoagulability, without affecting physiologic aggregation and coagulation processes. However, there is evidence that LA per se affects aggregation. The effect of the epidural administration of bupivacaine on platelet aggregation was studied by Odoom et al. (32) using adenosine diphosphate-induced platelet aggregation in platelet rich plasma in seven patients undergoing endoscopic transurethral prostate resection. The time of incubation of the LA after epidural administration seemed to have a major role, because a time lag of 1 h between the peak plasma concentration and the maximum platelet aggregation was striking. This might explain why, in a study published by Borg and Modig (34), 25-fold larger bupivacaine concentrations in platelet rich plasma were necessary to inhibit platelet aggregation after an incubation period of only 5 min. Odoom et al. (32) considered bupivacaine accumulation within the platelets as an explanation for the time lag. This has been demonstrated by Weksler et al. (35) for propanolol, a β-blocker with LA properties and comparable pharmacologic behavior. In 1977, the authors reported that propanolol inhibited platelet aggregation induced by adenosine diphosphate, epinephrine, collagen, thrombin, and the ionophore A23187. They also showed that platelets accumulated 14C-propanolol in vitro 10- to 30-fold over plasma concentrations. Propanolol has an LA or membrane-stabilizing effect (36). The concentrations of propanolol needed for a membrane effect are larger (10−6–10−4 M) than for β-adrenergic blockade (10−9–10−8 M). Also, the membrane-stabilizing effect is not stereospecific, unlike β-blockade. Thus, d(+) and l(−) propanolol have similar membrane effects, whereas the d(+) isomer has only 1% of the β-blockading capacity of the l(−) isomer. The concentration of propanolol required to inhibit platelet function in vitro (10−7–10−4 M) is large, suggesting a membrane effect. The conclusion inferred by these authors is supported by their finding that the d(+) and l(−) forms are equipotent inhibitors of platelet aggregation, serotonin release, and platelet adhesion to collagen. Furthermore, they showed that practolol, a potent β-adrenergic blocking drug that lacks membrane stabilizing activity, has no effect on platelet function. Thus, the authors concluded that the membrane effect, and not β-adrenergic blockade, seems to be crucial for the action of propanolol on platelets (35). A significant correlation was observed among bupivacaine plasma levels, bupivacaine incubation time, and the inhibition of all platelet aggregation variables, which suggests that this effect is caused by the LA itself (32). A 1977 report by Cooke et al. (37) supports this view; they found that an IV application of a lidocaine bolus followed by continuous infusion for 6 postoperative days (leading to plasma concentrations of 4 × 10−6 to 2 × 10−5 M) reduced the risk for deep vein thrombosis (DVT) without increasing bleeding risks in patients undergoing elective hip surgery. Notably, after discontinuation of lidocaine (between postoperative days 7 and 14), 41% of these patients without thrombosis during therapy developed deep vein thrombosis. No differences were found between the groups with regard to blood loss or transfusion requirements. Thus, lidocaine seems to reduce the risk for thrombosis without increasing bleeding, and this effect shown in the study by Cooke et al. (37) must be attributable to a direct effect of the LA rather than to the effects of neuraxial block produced by epidural anesthesia. Cytosolic mobilization of calcium from intracellular and extracellular sites is one of the earliest events in the activation process of platelets (38). There is evidence that LA blocks the Ca2+ uptake in platelets (39), as well as the influx of extracellular calcium, through selective inhibition of the Ca2+-dependent adenosine triphosphatase (40). An alternative hypothesis highlights the role of coagulation-associated pathways such as TXA2 signaling. TXA2 is a potent stimulator of platelet aggregation (41,42), and is one of the major mediators released during surgery (43,44). Kohrs et al. (27) studied the effects of bupivacaine (1–10 μM) on whole blood coagulation measured by using coagulation analysis and activated clotting time. Bupivacaine, in clinically relevant concentrations of 1–2 μM, influenced whole blood clotting characteristics. Thromboxane receptor antagonism increased activated clotting time in this study, confirming a role for thromboxane in coagulation. Bupivacaine inhibited TXA2 signaling, but seemed to block additional factors as well. Depolarization (45), changes in membrane microviscosity (46), and an increase in sodium uptake (47), are additional stimulatory factors during early platelet aggregation. Whether LA might influence these stimulatory factors of platelet aggregation has not been studied. Inflammatory Response Surgical trauma is associated with the activation of inflammatory pathways. Depending on the procedure and the occurrence of any complications, this may lead to a generalized systemic inflammatory response syndrome. Links between inflammation and coagulation are emerging in several areas such as the cellular and protein levels. At the cellular level, leukocytes bind tissue factor and other coagulation factors/complexes to their membranes, and through this mechanism activate the plasmatic coagulation cascade (48). Leukocytes are actively and specifically recruited to platelet clots (49). Platelets can cover leukocytes and thereby facilitate the adhesion of leukocytes to endothelium, diapedesis into the tissue, and release of inflammatory mediators. These are key events in the inflammatory response (50). At the protein level, coagulation factors, such as tissue factor, its endogenous inhibitor protein tissue factor pathway inhibitor, and factor Xa, are increasingly accepted as inflammatory proteins (51). There is no doubt that the plasminogen activator urokinase has a major role in tissue remodeling after inflammatory or ischemic insults (52). Although the clinical benefits solely related to the antiinflammatory effects of LAs have not been demonstrated, there is growing evidence from in vitro and animal studies that supports the view that LAs can modulate these events. Lidocaine reduces leukocyte adherence and delivery to inflamed tissues (53). Ropivacaine very potently reduces rolling and adhesion of leukocytes to the inflamed vessel wall, as assessed by intravital microscopy (54). Lidocaine and tetracaine reduce superoxide release from polymorphonuclear neutrophils in vitro(55). Polymorphonuclear neutrophils from cardiac patients who had received lidocaine perioperatively as an antiarrhythmic therapy, reduced superoxide production to 20%(56). LAs inhibit the phagocytic activity of leukocytes (57). These antiinflammatory effects are in part explained by an inhibition of leukocyte priming (55), an activation step that determines the amplitude of the response to activation (58). Some mechanisms of these antiinflammatory effects have been identified (59), going beyond changes in the ionic configuration of the plasma membrane, and some are still awaiting more detailed elucidation. One study showed an inhibition of PGE2 EP1 membrane receptors by bupivacaine (23), a receptor that is particularly relevant in inflammation and peripheral pain. Other receptors involved in inflammation and pain are affected by LAs. The lysophosphatidic acid receptor, which is involved in platelet aggregation, inflammation, and wound healing (25,60), and the thromboxane receptor (24,61), are inhibited. All of this evidence suggests direct effects of LAs on G protein-coupled receptors. One site of action of LAs within the G protein-coupled receptor pathway seems to be at the Gα(q) protein (28), but there is increasing evidence indicating more than one mechanism of action of LAs. Butterworth et al. (62) demonstrated LA inhibition of β2-adrenergic receptor binding at micromole concentration. LAs affect ion channels, e.g., sodium channels (63) and G protein-activated K+ channels (64). Microcirculation Although controversial, the incidence of thrombosis of vascular grafts in patients undergoing lower extremity revascularization seems to be significantly reduced after epidural anesthesia. Amputation and the incidence of revision for graft occlusion as clinical markers for graft failure are reduced (3,5,13). The effects on coagulation described above probably reflect only one possible mechanism by which epidurally administered LA may modulate clinical outcome. Tuman et al. (5) described a sustained increase in blood flow to the legs in patients with occlusive atherosclerotic disease because of extended sympathetic block with the postoperative administration of epidural LA. Sympathetic block caused by epidural anesthesia increased blood flow in the legs of patients with occlusive atherosclerotic disease in a study by Haljamae et al. (65). These findings are supported by a retrospective study by Scott et al. (66), who reported a smaller rate of microvascular complications in free skin flaps to the lower extremity. They suggested that pain relief and vasodilation (sympathicolysis) with epidural LA improve conduit blood flow to the free flap, as well as the microvascular flow distribution within the flap. Another group (67) has discussed possible prevention of postoperative vasospasm by LA. One approach to elucidating the direct effects of LA in the microvascular bed is intravital microscopy. After topical application of lidocaine, thrombus formation was reduced and microcirculatory blood flow was restored after laser-induced microvascular injury in a hamster cheek pouch model. A reduction of adhesion was observed both between all blood cells and between these cell clusters and the vessel wall (68). The mechanisms are not yet elucidated; however, it seems that the route of application has a role. Thoracic epidural anesthesia increases gut mucosal blood flow in rats by increasing blood cell velocity, and reduces adverse intermittent flow in the villous microcirculation, despite a reduction in perfusion pressure caused by sympathicolysis (69). Other investigations suggest a systemic effect of the LA on PGE2(23), TXA2(24), lysophosphatidic acid receptor (25), and other G protein-coupled receptors (28), which are of major importance for regulating regional blood flow. This might be a contributory mechanism to neuraxial blockade and sympathicolysis caused by thoracic epidural anesthesia. Conclusion Evidence shows that the prolonged use of epidurally administered LAs affect perioperative coagulation and microcirculation, and have the potential to improve the perioperative outcome (13,70). However, there is evidence that the systemic effects of LAs resorbed from the epidural space contribute to some of these actions. The mechanisms of the direct effects of LA remain unclear, but they affect the modulation of coagulation, fibrinolysis, inflammation, and platelet aggregation as well as the microcirculation without increased blood loss or more perioperative infections. Unfortunately, there are few studies dealing with systemically administered LA in general, and in respect to clinical outcome. Therefore, a discrimination between the direct systemic effects of LA and the effects of the neuraxial blockade, for instance, in epidural anesthesia and analgesia cannot be made. In the future, additional control groups in clinical studies, consisting of patients receiving systemically administered LAs, should be included if ethically appropriate. There is a potential for an improved outcome after surgical procedures that is associated with significant perioperative stress reduction when regional anesthesia is used. The question arises, therefore, whether patients with contraindications to regional anesthesia should receive LAs IV for the surgical procedure. Although studies in this field have been neglected, there is growing evidence that some patients would benefit from systemically administered LAs.