**2. Blood-membrane interaction: the role of complement, coagulation, kininkallikrein systems and soluble mediators**

### **2.1 Activation of the complement alternative pathway**

Early studies on biocompatibility focused on acute hypersensitivity-like reactions which in some cases were fatal. Various mechanisms were elucidated. Activation of complement was shown by Craddock *et al* in 1977 (Craddock et al, 1977). Hydroxyl radicals, present on the surface of cellulosic membranes, bind with the C3b in the blood and activate the alternative pathway leading to the release of potent anaphylatoxins, C3a and C5a. Both C3a and C5a and their relative desarginated products induce prompt activation and aggregation of polymorphonuclear neutrophils (PMNs) and leukopenia. This is a very rapid process reaching a nadir from 15 to 30 min after initiation of dialysis. Aggregates of PMNs are sequestered particularly in the lung capillaries. Although the extent of the anaphylatoxin generation and of the neutropenia is also patient-dependent, these studies failed to find a relationship with chronic clinical trade-offs despite the hypothesis that recurrent pulmonary sequestration could induce pulmonary fibrosis. Reduction of the hydroxyl groups on the membrane surface or new synthetic polymers reduced the activation of the alternative pathway of the complement cascade. Temperature could also reduce complement activation (*Maggiore Q, personal communication, 1988*). Testing

products (C3a and C5a and their desarginated products) guided the development of less neutropenia-inducing membranes and ultimately to the final development of fully synthetic membranes which have very low if at all capacity to induce complement activation. At that time, coagulation was an important reason for frequent interruptions and delays in the HD sessions. Due to the complex interplay known to occur between the activation of the complement and coagulation systems, it became of great interest to try to reduce the propensity for intravascular coagulation. The development of high-flux membranes and growing awareness of the benefits of convective and convective/diffusive under several contexts (intradialytic cardiovascular stability, better control or the uremic status and fluid control) gave impetus to a large number of enlightening studies on another mechanism of HD bio**IN**compatibility. The contamination by bacterial products, particularly with the widespread use of bicarbonate-based dialysates opened a new era in the field of biocompatibility. The formulation of the "interleukin hypothesis" was *a posteriori* not only the basis for further studies on the monocyte stimulation during HD, but also provided a link between biocompatibility and chronic inflammation. Basically, the evolution of biocompatibility has led us to consider two sides of the same coin: on one side, the biological responses at the blood-membrane interface; on the other hand, the consequences derived from the contact on the membrane performances (e.g. hydraulic permeability and sieving

In this review, we will summarize the most important steps in the evolution from the concept of the blood-dialyzer membrane interaction to that of the whole HD system compatibility. In face of very recent developments of cell-to-cell communication and signal transduction, we will also discuss the new hypothesis for a role of microvesicles (MVs) in cell activation, as well as in tissue and vascular repair. We will not deal with other important aspects of biocompatibility such as the oxidant stress, the relevant role of additives in dialyzer manufacturing, and of leachables and the effects of different sterilization modes.

**2. Blood-membrane interaction: the role of complement, coagulation, kinin-**

Early studies on biocompatibility focused on acute hypersensitivity-like reactions which in some cases were fatal. Various mechanisms were elucidated. Activation of complement was shown by Craddock *et al* in 1977 (Craddock et al, 1977). Hydroxyl radicals, present on the surface of cellulosic membranes, bind with the C3b in the blood and activate the alternative pathway leading to the release of potent anaphylatoxins, C3a and C5a. Both C3a and C5a and their relative desarginated products induce prompt activation and aggregation of polymorphonuclear neutrophils (PMNs) and leukopenia. This is a very rapid process reaching a nadir from 15 to 30 min after initiation of dialysis. Aggregates of PMNs are sequestered particularly in the lung capillaries. Although the extent of the anaphylatoxin generation and of the neutropenia is also patient-dependent, these studies failed to find a relationship with chronic clinical trade-offs despite the hypothesis that recurrent pulmonary sequestration could induce pulmonary fibrosis. Reduction of the hydroxyl groups on the membrane surface or new synthetic polymers reduced the activation of the alternative pathway of the complement cascade. Temperature could also reduce complement activation (*Maggiore Q, personal communication, 1988*). Testing

**kallikrein systems and soluble mediators** 

**2.1 Activation of the complement alternative pathway** 

coefficients).

complement activation (C3a or C5a plasma levels) by highly sensitive ELISA tests has become a standard requirement for the evaluation of biocompatibility ever since along with the precise characterization of the polymer structure (Krieter et al, 2008). It also became clear that synthetic polymers had a very low neutropenia-inducing effect. In some cases such as the polyacrylonitrile membrane, this was also due to the capacity of the membrane to adsorb C3b and the anaphylytoxins thus masking in fact complement activation (Pascual et al 1993) (**Figure 1**).

Fig. 1. Pathways involved in blood-membrane interactions. LTB4 denotes leukotriene B4, PAF, platelet-activating factor, IL-1, interleukin-1, TNF-, tumor necrosis factor.

### **2.2 Activation of the coagulation system**

Numerous acquired hemostatic abnormalities have been identified in chronic renal failure. HD adds to these disturbances as it repetitively implies turbulent blood flow, high shear stress, and contact of blood to artificial surfaces. Anticoagulation in HD is targeted to prevent activation of coagulation during the procedure. Most anticoagulant agents inhibit the plasmatic coagulation cascade. Still commonly used is unfractionated heparin, followed by low-molecular-weight heparin preparations with distinct advantages. Immune-mediated heparin-induced thrombocytopenia constitutes a potentially life-threatening complication of

The Evolution of Biocompatibility: From Microinflammation to Microvesiscles 97

apheresis (Owen et al., 1994) or in Hemodiafiltration (HDF) with regeneration of the ultrafiltrate (Tetta C, Wratten ML, unpublished observation, 2001). The appearance of signs and symptoms of a hypersensitivity-related event can be dramatic in the practice of HD. The complexity of the causal factors and the underlying mechanisms are often difficult to unveil (Arenas et al., 2006). The majority of reported cases have been due to ethylene oxide (ETO) (Poothullil et al., 1975), triggered by both immunoglobulin E (IgE) and non-IgE factors (Johansson et al., 2001). However, a considerable number of publications have focused on other HD substances and materials such as heparins, different dialyser membranes, iron, erythropoietin, polyacrylonitrile AN69® high flux membranes, latex, antiseptic or formaldehyde (Ebo et al, 2006). Many different underlying mechanisms have been postulated. Hypersensitivity reactions have been estimated to occur in ~4/100000 dialysis treatments. A postal survey of all HD centres in the UK suggested that 1/20 to 1/50 patients may be susceptible to anaphylactoid reaction to a new hemodialyser at some time in between, while the risk of reaction occurring with any single HD session is~1/1000 to 1/5000. Although it is likely that many reactions are unrecognized or unreported, the scale of the problem is larger than many nephrologists have suspected (Nicholls et al., 1987).

Fig. 2. Activation of the kinin-kallikrein system and generation of bradykinin (BK)

heparin therapy requiring immediate switch to nonheparin alternative anticoagulants. Danaparoid, lepirudin, and argatroban are currently being used for alternative anticoagulation, all of which possess both advantages and limitations. Recently citrate has been proposed as anticoagulant in maintenance HD (Wright et al, 2010). In the past, empirical strategies reducing or avoiding heparin were applied for patients at bleeding risk, whereas nowadays regional citrate anticoagulation is increasingly used to prevent bleeding by allowing procedures without any systemic anticoagulation. Avoidance of clotting within the whole hemodialyzer circuit is not granted. Specific knowledge of the mechanisms of coagulation, the targets of the anticoagulants in use, and their respective characteristics constitutes the basis for individualized anticoagulation aimed at achieving full patency of the circuit throughout the procedure. Patency of the circuit is an important prerequisite for optimal HD quality. Intrinsic coagulation Hageman factor XII as well as other coagulation factors are also activated (Fischer, 2007). However, the activation of the coagulation is a very complex phenomenon that may be enhanced by different independent factors other than the membrane surface *per se* such as: the dynamics at the dialyzer heads, defects in the hollow fibre cutting of the polyurethane, any condition that predisposes for blood to be stagnant. The activation of coagulation by a membrane in a dialyzer is difficult to assess given the above-mentioned factors and the host's response to the anticoagulation regime put in place (**Figure 1**).

### **2.3 Activation of the kinin-kallikrein system**

Surface activation of Factor XII induces the kinin-kallikrein that ensues in the generation of bradykinin (**Figure 2**).

Bradykinin is physiologically under the tight control of very potent kinases that are able to promptly lyse the molecule and inactivate its potent vasodilator activity. In certain conditions, however, the lytic effect of this kinase is deficient. This occurs in patients under therapy with angiotensin converting enzyme enzyme (ACE) inhibitors. However, there are patients who experience hypersensitivity-like phenomena, that can be reconducted to bradykinin generation, even in the absence of concomitant therapy with ACE inhibitors. The explanation of this phenomenon came from pioneering studies on angio-edema, a rare but potentially fatal condition (Adam et al., 2002). These reactions are mainly associated to defects in the enzymatic activity of the aminopeptidase P (**Figure 2**). Bradykinin acts through two types of tissue receptors: R1 are mostly located in the skin and respiratory tissues (lungs and bronchi), while R2 are mostly found in the gastrointestinal tract. The overproduction of bradykinin may lead to two different clinical presentations: the first is mainly characterized by a rapid developing skin flushing, hypotension, and dyspnoea. These reactions may be mild but very severe, fatal episodes of shock have been described. In the second instance, these reactions, which were for some time unexplained, occur after 1 h-2 hr of HD treatment, may but may be not associated with the use of ACE inhibitors. The patient has severe diarrhoea which requires immediate interruption of the extracorporeal treatment. This manifestation may unpredictably recur and disappears upon disconnection. Bradykinin-induced reaction, may in principle occur following the contact with any foreign surface. Their potential, unpredictable severity should call for immediate action even in patients with mild forms. The commonest causes have been the use of strongly negative surfaces such as AN-69 membranes (Tielemans et al., 1990), or adsorbents used in LDL

heparin therapy requiring immediate switch to nonheparin alternative anticoagulants. Danaparoid, lepirudin, and argatroban are currently being used for alternative anticoagulation, all of which possess both advantages and limitations. Recently citrate has been proposed as anticoagulant in maintenance HD (Wright et al, 2010). In the past, empirical strategies reducing or avoiding heparin were applied for patients at bleeding risk, whereas nowadays regional citrate anticoagulation is increasingly used to prevent bleeding by allowing procedures without any systemic anticoagulation. Avoidance of clotting within the whole hemodialyzer circuit is not granted. Specific knowledge of the mechanisms of coagulation, the targets of the anticoagulants in use, and their respective characteristics constitutes the basis for individualized anticoagulation aimed at achieving full patency of the circuit throughout the procedure. Patency of the circuit is an important prerequisite for optimal HD quality. Intrinsic coagulation Hageman factor XII as well as other coagulation factors are also activated (Fischer, 2007). However, the activation of the coagulation is a very complex phenomenon that may be enhanced by different independent factors other than the membrane surface *per se* such as: the dynamics at the dialyzer heads, defects in the hollow fibre cutting of the polyurethane, any condition that predisposes for blood to be stagnant. The activation of coagulation by a membrane in a dialyzer is difficult to assess given the above-mentioned factors and the host's response to the anticoagulation regime put in place

Surface activation of Factor XII induces the kinin-kallikrein that ensues in the generation of

Bradykinin is physiologically under the tight control of very potent kinases that are able to promptly lyse the molecule and inactivate its potent vasodilator activity. In certain conditions, however, the lytic effect of this kinase is deficient. This occurs in patients under therapy with angiotensin converting enzyme enzyme (ACE) inhibitors. However, there are patients who experience hypersensitivity-like phenomena, that can be reconducted to bradykinin generation, even in the absence of concomitant therapy with ACE inhibitors. The explanation of this phenomenon came from pioneering studies on angio-edema, a rare but potentially fatal condition (Adam et al., 2002). These reactions are mainly associated to defects in the enzymatic activity of the aminopeptidase P (**Figure 2**). Bradykinin acts through two types of tissue receptors: R1 are mostly located in the skin and respiratory tissues (lungs and bronchi), while R2 are mostly found in the gastrointestinal tract. The overproduction of bradykinin may lead to two different clinical presentations: the first is mainly characterized by a rapid developing skin flushing, hypotension, and dyspnoea. These reactions may be mild but very severe, fatal episodes of shock have been described. In the second instance, these reactions, which were for some time unexplained, occur after 1 h-2 hr of HD treatment, may but may be not associated with the use of ACE inhibitors. The patient has severe diarrhoea which requires immediate interruption of the extracorporeal treatment. This manifestation may unpredictably recur and disappears upon disconnection. Bradykinin-induced reaction, may in principle occur following the contact with any foreign surface. Their potential, unpredictable severity should call for immediate action even in patients with mild forms. The commonest causes have been the use of strongly negative surfaces such as AN-69 membranes (Tielemans et al., 1990), or adsorbents used in LDL

(**Figure 1**).

bradykinin (**Figure 2**).

**2.3 Activation of the kinin-kallikrein system** 

apheresis (Owen et al., 1994) or in Hemodiafiltration (HDF) with regeneration of the ultrafiltrate (Tetta C, Wratten ML, unpublished observation, 2001). The appearance of signs and symptoms of a hypersensitivity-related event can be dramatic in the practice of HD. The complexity of the causal factors and the underlying mechanisms are often difficult to unveil (Arenas et al., 2006). The majority of reported cases have been due to ethylene oxide (ETO) (Poothullil et al., 1975), triggered by both immunoglobulin E (IgE) and non-IgE factors (Johansson et al., 2001). However, a considerable number of publications have focused on other HD substances and materials such as heparins, different dialyser membranes, iron, erythropoietin, polyacrylonitrile AN69® high flux membranes, latex, antiseptic or formaldehyde (Ebo et al, 2006). Many different underlying mechanisms have been postulated. Hypersensitivity reactions have been estimated to occur in ~4/100000 dialysis treatments. A postal survey of all HD centres in the UK suggested that 1/20 to 1/50 patients may be susceptible to anaphylactoid reaction to a new hemodialyser at some time in between, while the risk of reaction occurring with any single HD session is~1/1000 to 1/5000. Although it is likely that many reactions are unrecognized or unreported, the scale of the problem is larger than many nephrologists have suspected (Nicholls et al., 1987).

Fig. 2. Activation of the kinin-kallikrein system and generation of bradykinin (BK)

The Evolution of Biocompatibility: From Microinflammation to Microvesiscles 99

Fig. 3. Microhemolysis is the release of small quantities of hemoglobin (micro- or nanomolar) from erythrocytes. The tyrosine of a hemoglobin molecule can undergo a transition to a reactive free radical. This can react with other protein tyrosine residues to form a dityrosine molecule. Microhemolysis occurs during the HD procedure in which the erythrocytes are slightly damaged and tend to "leak" very small quantities of hemoglobin. This is a very common phenomena in HD and should not be confused with gross hemolysis.

**4.1 From system biocompatibility to systemic chronic inflammation** 

cardiovascular disease observed in the uremic population (Foley et al 1998).

The concept that inflammation underlines many diseases once considered to be linked to degenerative processes has revolutionized the approach to the research into the pathogenesis and new therapeutics alike. In the field of cardiovascular disease, the process of endothelial dysfunction, vascular damage and atherosclerosis is now seen as a continuum (Libby et al., 2002). Cardiovascular disease is among the leading cause of morbidity and mortality in CKD patients on maintenance HD (US Renal Data System, 1997; Parfey & Foley, 1999). Even before reaching the state of chronic kidney disease Stage 5, patients with chronic renal failure present signs of chronic inflammation. Once patients are on HD, the risk of cardiovascular death is approximately 30 times higher than in the general population, and still remains 10 to 20 times higher after stratification for age, gender, and presence of diabetes. Traditional risk factors seem inadequate to explain the remarkable prevalence of

Inflammatory mechanisms play a relevant role in the development and progression of atherosclerosis (Ross, 1999) and heart failure (Vasan et al., 2003). Epidemiological studies in the general population have shown that even minor elevations of C-reactive protein (CRP), an acute phase reactant that markedly increases during an inflammatory response (Ridker PM, et al., 1997) predict the development of coronary heart disease and cardiac failure

**4. The evolution of treatment biocompatibility** 

**4.1.1.1 Systemic Chronic Inflammation** 

### **2.3.1 Soluble mediators**

Many soluble mediators are produced and released following the blood-membrane interaction. Products of the phospholipase A2 such as platelet-activating factor (PAF) and leukotrienes are released by the direct interaction of PMNs and platelets with complementactivating membranes. Although in the presence of blood, the mechanisms of production of PAF and leukotrienes can not be readily differentiated from the activation, as they follow the same kinetics, we could show that for PAF for example, its production and release could be observed in complement-independent conditions such as in the absence of plasma by purified cells incubated with flat HD membranes (Tetta et al., 1996). A large number of studies have also suggested the occurrence in the plasma of lytic enzymes normally present in the vacuoles of inflammatory cells such as elastases, and metalloenzymes. The release of these lytic enzymes is caused by a phenomenon named by cell physiologists as "frustrated phagocytosis".
