**9. Support to the central blood volume**

An OLT may be associated with a severe blood loss in about 20% of the patients (Massicotte *et al.*, 2004). Bleeding during OLT is most common during the dissection phase for the cirrhotic liver associated with distended abdominal veins. Accordingly, rapid infusion systems have been developed which that can deliver 1.5 l of heated fluid per minute through 7 F venous catheter is recommended for infusion of fluids and blood products. The rapid infusion machine reports the accumulated amount of fluid (blood) administered and is also suited for infusion of small volumes to children as the fluid can be administrated with an accuracy of millilitres. To adults it is common to administer bolus infusions of 100 ml (occasionally 500 ml) over 1 min since, as mentioned, a lowering of SvO2 by 1 % corresponds typically to a volume deficit of ~100 ml (Ejlersen et al., 1995a). Furthermore, one or two i.v. lines can be placed for supplementary administration of fluid and medication.

Since patients with end-stage liver decease present with serious haemostatic defects, management of coagulation during OLT is a challenge. Coagulopathy can be aggravated by hypothermia and acidemia, associated with increased risk of uncontrolled bleeding and mortality (Lier et al., 2008). Accordingly fluid, including blood products, is preheated and the patients are provided with a warm airflow blanket (e.g. "Bair Hugger") covering the

intended increase in ScO2 (Nissen et al., 2010). Alternatively, hypotension should be considered in relation to a decrease in plasma calcium in response to administration of blood products and calcium should be supplemented to restore the physiological level (1.2 mM). However, the use of phenylephrine may be indicated during reperfusion of the donor liver to reduce peripheral vasodilatation and thereby to centralise of blood accumulated in the splanchnic region. Alternatively, the administration of phenylephrine to increase the blood pressure may be replaced by the use of ephedrine that does not demonstrate the same

PaCO2 has a significant influence on CBF and PaCO2 is regularly monitored by a continuous recording of the end-tidal CO2 tension to maintain a value of, e.g. 4,5 kPa. In that regard OLT is no exception, but with the reduction of the metabolic rate during the anhepatic phase of the OLT, a given setting of ventilation may lower PaCO2 and ventilation then needs to be reduced in order to maintain CBF and ScO2 (Madsen & Secher, 1999). Conversely, with reperfusion of the donor liver, PaCO2 increases again and often to values that exceed the level established in the dissection phase of the operation. Accordingly, ventilation should be increased at, or likely before reperfusion of the liver in order to prevent cerebral hyperperfusion and ventilation is thereafter gradually reduced towards the end of the

With optodes placed over a muscle NIRS monitors muscle oxygen saturation (SmO2) and decreases before central hypovolaemia affects blood pressure. However, haemorrhagic hypotension is likely to be caused by a Bezold-Jarisch-like reflex including loss of sympathetic activity and, therefore, an increase in muscle blood flow and in turn SmO2 (Madsen *et al.*, 1995). SmO2 effectively detects central hypovolaemia (Soller *et al.*, 2008) (Fig. 2) and supplements, or may be used as an alternative non-invasive monitoring modality to SvO2 for detection of a blood loss. Ideally SmO2 provides for an early warning of ongoing haemorrhage and allows for direction of fluid administration before the blood loss affects

An OLT may be associated with a severe blood loss in about 20% of the patients (Massicotte *et al.*, 2004). Bleeding during OLT is most common during the dissection phase for the cirrhotic liver associated with distended abdominal veins. Accordingly, rapid infusion systems have been developed which that can deliver 1.5 l of heated fluid per minute through 7 F venous catheter is recommended for infusion of fluids and blood products. The rapid infusion machine reports the accumulated amount of fluid (blood) administered and is also suited for infusion of small volumes to children as the fluid can be administrated with an accuracy of millilitres. To adults it is common to administer bolus infusions of 100 ml (occasionally 500 ml) over 1 min since, as mentioned, a lowering of SvO2 by 1 % corresponds typically to a volume deficit of ~100 ml (Ejlersen et al., 1995a). Furthermore, one or two i.v. lines can be placed for

Since patients with end-stage liver decease present with serious haemostatic defects, management of coagulation during OLT is a challenge. Coagulopathy can be aggravated by hypothermia and acidemia, associated with increased risk of uncontrolled bleeding and mortality (Lier et al., 2008). Accordingly fluid, including blood products, is preheated and the patients are provided with a warm airflow blanket (e.g. "Bair Hugger") covering the

negative influence on ScO2 (Nissen *et al.*, 2010; Meng *et al.*, 2011).

operation guided by ScO2 as the CO2 load is eliminated by the exhaled air.

CBF and in turn ScO2.

**9. Support to the central blood volume** 

supplementary administration of fluid and medication.

upper part of thorax, the neck, and arms and that arrangement is usually able to maintain the patient's temperature above 36º. Furthermore, frequent adjustment of CBV according to the individualized goal directed fluid administration strategy prevents marked changes in pH and arterial lactate and frequent monitoring of the plasma calcium level, with appropriate substitutions, supports coagulation competence.

Yet, in order to preserve coagulation competence, there remains some disagreement. While one review finds that correction of coagulation defects with fresh frozen plasma (FFP) and platelets does not reduce the blood loss and has a negative influence on outcome (Dalmau et al., 2009), the local experience is that timely administration of FFP and platelets improves outcome (Johansson *et al.*, 2010), a strategy supported by other studies demonstrating a decrease in blood loss and improvement in 24 hour and 30-day survival following early use of plasma and platelets (Holcomb, 2010; Holcomb et al., 2008). Discrepancy between views likely relate to which patients are addressed with the need for FFP and platelets indicated only for patients presenting with a deficit in consequence of their liver disease or provoked by major haemorrhage.

It is important to monitor the OLT patient's coagulation competence and in that regard a determination of the protrombin time (PT) and activated partial protrombin time (APPT) seems inadequate since these variables do not correlate well with clinically coagulopathy or bleeding condition (Murray *et al.*, 1999; Segal & Dzik, 2005). For on-line evaluation of haemostasis, thrombelastographhy (TEG) is used as pioneered during OLT by Kang et al (Kang *et al.*, 1985). TEG records the viscoelastic changes during coagulation by analysing whole blood placed in a rotating cup (Fig. 10) while a pin suspended in the blood from a torsion wire records the resistance to motion.

Fig. 10. Thrombelastograph technology and measured variables: The clotting time (R), the angle (α) representing the progressive increase in clot strength, the maximal clot strength (MA) and the fibrinolysis (LY) (From Johannson et al,. 2010 with permission)

Four TEG parameters are regularly reported: the clotting time (R), the angle (α) representing the progressive increase in clot strength, the maximal clot strength (MA), and fibrinolysis (LY) (Fig. 10) (Johansson, 2009). The TEG reported values correlate well with the clinical bleeding conditions and are recommended to direct blood product treatment together with a platelet count and check of the haemoglobin concentration (Plotkin et al., 2008). The TEG analysis can be performed in the laboratory and is locally displayed in real-time in the operating theatre to enable for early intervention, e.g. by administration of platelets and plasma in case of significantly attenuated coagulation competence. In this endeavour, it is considered that transfusion of platelets to 100 X 109 L-1 (rather than to the commonly used guideline of 50 X 109 L-1) secures coagulation competence and additional administration of platelets is routinely performed before the reperfusion of the donor liver associated with significant use of platelets (Johansson *et al.*, 2010). Thus, the coagulation competence of the patient's blood is accentuated by the administration of FFP and saline-adenine-glucosemannitol (SAGM) erythrocyte suspension to a haemoglobin concentration of 6 mM (haematocrit 30%), making sure that plasma calcium does not decrease and calcuimcloride is administered frequently to maintain a reference value of 1.2 mM.

Yet it has to be accepted that some patients have intraoperative increased consumption of fibrinogen and if a diffuse bleeding in combination with a reduced α and MA manifest, we suggest monitoring functional fibrinogen to decide whether the reduction in MA relates to platelet or to fibrinogen function. Attention has also been directed to the endothelial barrier function in response to haemorrhagic shock and the role of glycocalyx appears important to endothelial permeability, intracellular dysfunction, and oedema (Holcomb, 2011).

Until recently, infusion of aprotinin was an option for OLT and aprotinin reduces haemorrhage by hindering fibrinolyses and thereby stabilizes the formed blood clots. Aprotinin has, however, been withdrawn from the marked (Dietrich, 2009) and tranexamic acid is the (cheaper) alternative to be administrated before surgery and again before reperfusion of the donor liver (Takagi et al., 2009).

Further refinement of treatment includes control of plasma concentration of magnesium (Skak et al., 1996) and corrected if low (reference value 0.8 mM) and maintenance of the blood glucose or, conversely, administration of insulin in case the blood glucose level increases beyond 10 mM. Surprisingly, the blood glucose level does not decrease during the anhepatic phase of OLT, presumably because the kidneys supplement glucose production (Lauritsen *et al.*, 2002). Also it should be considered that during massive administration of blood products, HR might be affected by the potassium concentration of Sag-M blood of approximately 50 mM. It is advised that massive administration of blood is paralleled by infusion of adrenaline (4-6 mg kg-1 min-1) to eliminate potassium from blood.
