**2. Normovolaemia**

By definition it seems difficult to defend that any patient should be provided with a fluid overload or be maintained hypovolaemic and yet, even hypovolaemic shock is sometimes treated with sympatomimetic drugs (De Backer D. *et al.*, 2010). For balancing volume administration it is of interest that normovolaemia can not only be defined but also defended during surgery. For healthy supine humans stroke volume, CO, and SvO2 do not respond to expansion of CBV and, therefore for supine humans, a volume administration strategy that secures that the heart operates on the upper ceiling of the Frank-Starling curve maintains the patient normovolaemic (Harms *et al.*, 2003;Jans *et al.*, 2008;Bundgaard-Nielsen *et al.*, 2009c) (Fig. 1).

For individualized goal-directed fluid therapy, volume is administered until a flow related parameter such as stroke volume, CO or SvO2, reaches a maximal value (Jenstrup *et al.*, 1995), i.e. volume is administrated until cardiac function does not depend on preload to the heart. This volume administration strategy is of interest not only because it remains elusive what volume rate a fixed volume administration strategy should aim at (Bundgaard-Nielsen *et al.*, 2009b), but also because an individualized fluid administration strategy, in contrast to a fixed volume strategy, consistently improves outcome for surgical patients (Bundgaard-Nielsen *et al.*, 2007a; Lopes *et al.*, 2007; Donati *et al.*, 2007; Abbas & Hill, 2008; Mayer *et al.*, 2010).

Furthermore during surgery, HR and especially MAP are affected by the anaesthetic agents and by the surgical stress (Ejlersen *et al.*, 1995a). Despite these limitations in the use of HR and MAP to detect deviations in CBV, the capability to balance CBV is of importance for tissue perfusion and oxygenation and notably for oxygenation of the brain (ScO2) (Nissen *et al.*, 2009a), indicating that advanced cardiovascular monitoring is required to secure the well-being of the patient (Yao *et al.*, 2004;Bundgaard-Nielsen *et al.*, 2007a;Murkin *et al.*, 2007) In order to maintain CBV during surgery, it is important that normovolaemia is defined. For supine humans the heart operates on the upper flat part of the Frank-Starling curve (Harms *et al.*, 2003) and to establish and to maintain a maximal resting stroke volume for the heart (or CO) secures that the patient remains normovolaemic during the operation and that fluid administration strategy reduces postoperative complications to an extent that affects the hospital stay (Bundgaard-Nielsen *et al.*, 2007a). Such goal directed fluid therapy was introduced by Shoemaker et al. (Shoemaker, 1972;Shoemaker *et al.*, 1988) in regard to CO but without taking the individual and partly genetically determined differences in CO (Snyder *et al.*, 2006) into account. Accordingly, this chapter focuses on how normovolaemia can be established and maintained during OLT despite the difficulties confronting the definition of normovolaemia by the spontaneous changes in stroke volume, CO and (mixed) venous oxygen saturation (SvO2) during the different phases of the operation. A second goal of this chapter is to introduce devices that can be applied for intraoperative monitoring of CBV and ScO2. Focus is on the importance of maintaining a normal CBV to secure cerebral blood flow (CBF) and ScO2 since these variables are taken to express the integrity of the cardiovascular system and their defence, at least potentially, prevents postoperative complications and cognitive dysfunction (Murkin *et al.*, 2007). In addition, the volume administration strategy

By definition it seems difficult to defend that any patient should be provided with a fluid overload or be maintained hypovolaemic and yet, even hypovolaemic shock is sometimes treated with sympatomimetic drugs (De Backer D. *et al.*, 2010). For balancing volume administration it is of interest that normovolaemia can not only be defined but also defended during surgery. For healthy supine humans stroke volume, CO, and SvO2 do not respond to expansion of CBV and, therefore for supine humans, a volume administration strategy that secures that the heart operates on the upper ceiling of the Frank-Starling curve maintains the patient normovolaemic (Harms *et al.*, 2003;Jans *et al.*, 2008;Bundgaard-Nielsen

For individualized goal-directed fluid therapy, volume is administered until a flow related parameter such as stroke volume, CO or SvO2, reaches a maximal value (Jenstrup *et al.*, 1995), i.e. volume is administrated until cardiac function does not depend on preload to the heart. This volume administration strategy is of interest not only because it remains elusive what volume rate a fixed volume administration strategy should aim at (Bundgaard-Nielsen *et al.*, 2009b), but also because an individualized fluid administration strategy, in contrast to a fixed volume strategy, consistently improves outcome for surgical patients (Bundgaard-Nielsen *et al.*, 2007a; Lopes *et al.*, 2007; Donati *et al.*, 2007; Abbas & Hill, 2008; Mayer *et al.*,

applied during the operation is addressed.

**2. Normovolaemia** 

*et al.*, 2009c) (Fig. 1).

2010).

Fig. 1. Relationship between systemic haemodynamic variables, mixed venous oxygen saturation and tilt angle. BP, systolic, mean and diastolic blood pressure; HR, heart rate; CO, cardiac output; SvO2 mixed venous oxygen saturation. Variables are mean ± S.E.M. # P < 0.05 vs. 0 deg. Dashed line represents the supine position; 70 a, 70 deg head-up tilt for 10 min; 70 b, last minute of 70 deg head-up tilt for nine subjects (From Harms et al., 2003 with permission)

Based on administration of a crystalloid or a colloid, an inherent difficulty for individualized goal directed fluid therapy is, however, that a reduction in haematocrit is associated with an increase in CO, i.e. normovolaemic haemodilution increases CO (Krantz *et al.*, 2005). In other words, it is not CO but SvO2 that is the regulated variable since the red cells create their own flow regulation through the release of ATP and NO when oxygen is released from oxyhaemoglobin (Gonzalez-Alonso *et al.*, 2006). To direct fluid administration on the basis of establishing maximal values for stroke volume or CO requires that there is added a rule to limit the fluid administered. A common algorithm implies that a 10%, or larger increase in stroke volume justifies further administration of 200–250 ml of colloid, thereby minimizing the risk of creating a fluid overload (Bundgaard-Nielsen *et al.*, 2007a). In contrast, during isovolaemic haemodilution, SvO2 remains stable until the haemoglobin level is reduced by approximately 50% (Krantz *et al.*, 2005) and volume administration based on the recording of ScO2 is therefore widely independent of the type of fluid used.

When CBV is normalized by fluid to establish a maximal SvO2, the administration of 100 ml fluid results a ~1% increase in SvO2 for the adult patient (~ 70 kg) (Ejlersen *et al.*, 1995a) and that relationship applies also to children when the volume is adjusted according to body weight. For supine humans, SvO2 is on an average 75% (Harms *et al.*, 2003) but for patients undergoing OLT, SvO2 is typically ~ 85% (Ejlersen *et al.*, 1995b) reflecting that for these patients CO is larger (7-9.5 l/min) (Table 1) than for a reference population (6.5 l/min).


HR, hear rate; CI, cardiac index; SvO2, mixed venous saturation; TA, thoracic electric admittance; MAP, mean arterial blood pressure; CVP, central venous pressure; PAMP, pulmonal arterial mean pressure; SVRI, systemic vascular resistance index; PVRI, pulmonal vascular resistance index (Modified from Skak et al., 1997).

Table 1. Cardiovacular variables during OLT

Reperfusion of the grafted liver is associated with peripheral vasodilatation and although CO is likely to increase (Table 1), situations associated with peripheral vasodilatation, as during heating (Wilson *et al.*, 2009), are likely to reduce CBV. The importance of establishing normovolaemia from the induction of anaesthesia is illustrated by an increase in ScO2 as determined by near infrared spectroscopy (NIRS) and similarly determined muscle oxygenation (SmO2) since both these indices of tissue oxygenation increase in parallel with CO as the heart becomes filled with blood and decrease with the filling of the heart during a bleeding episode (Fig. 2).

Conversely, to maintain the commonly accepted 70% value for SvO2 (Rivers, 2006) is likely to represent a 1.5 l volume deficit for the patient undergoing OLT, considering the 1% reduction in SvO2 to 100 ml blood volume relationship during hypovolaemia (Ejlersen *et al.*, 1995a). A 1.5 l volume deficit is so large that it compromises MAP and ScO2 (Secher *et al.*, 1992) since a ~30 % reduction of the blood volume and, hence CBV elicits a Bezold-Jarischlike reflex including a critical reduction in CBF (van Lieshout *et al.*, 2003;Secher & van

is reduced by approximately 50% (Krantz *et al.*, 2005) and volume administration based on

When CBV is normalized by fluid to establish a maximal SvO2, the administration of 100 ml fluid results a ~1% increase in SvO2 for the adult patient (~ 70 kg) (Ejlersen *et al.*, 1995a) and that relationship applies also to children when the volume is adjusted according to body weight. For supine humans, SvO2 is on an average 75% (Harms *et al.*, 2003) but for patients undergoing OLT, SvO2 is typically ~ 85% (Ejlersen *et al.*, 1995b) reflecting that for these patients CO is larger (7-9.5 l/min) (Table 1) than for a reference population (6.5

> **Anhepatic phase**

**Reperfusion End of** 

**operation** 

the recording of ScO2 is therefore widely independent of the type of fluid used.

**Dissection phase** 

HR (bpm) 95 90 87 97 CI (l m-1 min-1) 4.4 3.3 5.0 4.5 SvO2 (%) 85 81 86 82 TA (Ohm) 29 28 29 26 MAP (mmHg) 88 89 85 87 CVP (mmHg) 10 10 14 11 PAMP (mmHg) 18 17 26 21 SVRI (mmHg m2 min l-1) 142 195 105 120 PVRI (mmHg m2 min l-1) 11 14 9 12 Temperature 36 35 35 35

HR, hear rate; CI, cardiac index; SvO2, mixed venous saturation; TA, thoracic electric admittance; MAP, mean arterial blood pressure; CVP, central venous pressure; PAMP, pulmonal arterial mean pressure; SVRI, systemic vascular resistance index; PVRI, pulmonal vascular resistance index (Modified from

Reperfusion of the grafted liver is associated with peripheral vasodilatation and although CO is likely to increase (Table 1), situations associated with peripheral vasodilatation, as during heating (Wilson *et al.*, 2009), are likely to reduce CBV. The importance of establishing normovolaemia from the induction of anaesthesia is illustrated by an increase in ScO2 as determined by near infrared spectroscopy (NIRS) and similarly determined muscle oxygenation (SmO2) since both these indices of tissue oxygenation increase in parallel with CO as the heart becomes filled with blood and decrease with the filling of the heart during a

Conversely, to maintain the commonly accepted 70% value for SvO2 (Rivers, 2006) is likely to represent a 1.5 l volume deficit for the patient undergoing OLT, considering the 1% reduction in SvO2 to 100 ml blood volume relationship during hypovolaemia (Ejlersen *et al.*, 1995a). A 1.5 l volume deficit is so large that it compromises MAP and ScO2 (Secher *et al.*, 1992) since a ~30 % reduction of the blood volume and, hence CBV elicits a Bezold-Jarischlike reflex including a critical reduction in CBF (van Lieshout *et al.*, 2003;Secher & van

l/min).

Skak et al., 1997).

bleeding episode (Fig. 2).

Table 1. Cardiovacular variables during OLT

Fig. 2. Venous (SvO2) and muscle oxygen saturation (SmO2) plotted together with filling of the left ventricle (LVAd), and volume balance in a patient exposed to an episode of intraoperative bleeding and following volume expansion (C. Tollund unpublished)

Lieshout, 2009;Madsen & Secher, 1999). The Bezold-Jarisch-like reflex also provokes a marked (30-fold) increase in plasma vasopressin (Sander-Jensen *et al.*, 1986) with longlasting effect on urine production and could explain a potential difficulty in maintaining a reasonable urine production after surgery. Typically, the patients are in need of 0.5 l of volume before OLT (Ejlersen *et al.*, 1995b), a value that corresponds to that found also for other groups of patients before surgery (Jenstrup *et al.*, 1995;Bundgaard-Nielsen *et al.*, 2007b;Bundgaard-Nielsen *et al.*, 2009a). If it is felt desirable to maintain urine production, it is important that such an initial volume deficit is corrected and, eventually, a larger fluid load may be required than that which establishes that the patient has been provided with a normal blood volume. For additional volume threatment of patients, the administration of lactated Ringer solution is preferable to the administration of saline (Waters et al., 2001).
