Post - Acta Physiol 2010 199 43 by Paul Steendijk

Acta Physiol 2010, 199, 43–52

Cardiac function during mild hypothermia in pigs: increased inotropy at the expense of diastolic dysfunction H. Post,1 J. D. Schmitto,2 P. Steendijk,3 J. Christoph,1 R. Holland,2 R. Wachter,1,4 F. W. Scho¨ndube2 and B. Pieske1 1 2 3 4

Department of Cardiology, Medical University of Graz, Graz, Austria Department of Cardiothoracic Surgery, University Hospital of Go¨ttingen, Go¨ttingen, Germany Departments of Cardiology and Cardiothoracic Surgery, Leiden Medical University Center, Leiden, the Netherlands Department of Cardiology and Pneumology, University Hospital of Go¨ttingen, Go¨ttingen, Germany

Received 24 August 2009, revision requested 8 October 2009, revision received 8 December 2009, accepted 17 January 2010 Correspondence: H. Post, Klinische Abt. fu¨r Kardiologie, Medizinische Universita¨t Graz, Auenbruggerplatz 15, 8036 Graz, Austria. E-mail: heiner.post@ meduni-graz.at

Abstract Aim: The induction of mild hypothermia (MH; 33 C) has become the guideline therapy to attenuate hypoxic brain injury after out-of-hospital cardiopulmonary resuscitation. While MH exerts a positive inotropic effect in vitro, MH reduces cardiac output in vivo and is thus discussed critically when severe cardiac dysfunction is present in patients. We thus assessed the effect of MH on the function of the normal heart in an in vivo model closely mimicking the clinical setting. Methods: Ten anaesthetized, female human-sized pigs were acutely catheterized for measurement of pressure–volume loops (conductance catheter), cardiac output (Swan-Ganz catheter) and for vena cava inferior occlusion. Controlled MH (from 37 to 33 C) was induced by a vena cava inferior cooling catheter. Results: With MH, heart rate (HR) and whole body oxygen consumption decreased, while lactate levels remained normal. Cardiac output, left ventricular (LV) volumes, peak systolic and end-diastolic pressure and dP/dtmax did not change significantly. Changes in dP/dtmin and the time constant of isovolumetric relaxation demonstrated impaired active relaxation. In addition, MH prolonged the systolic and shortened the diastolic time interval. Pressure–volume analysis revealed increased end-systolic and end-diastolic stiffness, indicating positive inotropy and reduced end-diastolic distensibility. Positive inotropy was preserved during pacing, while LV end-diastolic pressure increased and diastolic filling was substantially impaired due to delayed LV relaxation. Conclusion: MH negatively affects diastolic function, which, however, is compensated for by decreased spontaneous HR. Positive inotropy and a decrease in whole body oxygen consumption warrant further studies addressing the potential benefit of MH on the acutely failing heart. Keywords diastolic function, hypothermia, systolic function.

The induction of mild hypothermia (MH; 33 C) after out-of-hospital cardiopulmonary resuscitation improves neurological outcome, reduces mortality and has become the guideline therapy since 2005 (International

Liaison Committee on Resuscitation 2005). However, with respect to haemodynamics, the current perception is that ‘mild hypothermia increases systemic vascular resistance, which reduces cardiac output’ (Neumar

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Cardiac function during mild hypothermia

Æ H Post et al.

et al. 2008). Resuscitated patients typically suffer from acute and pre-existing cardiac dysfunction, and to further reduce cardiac output (CO) by MH seems counterintuitive. The clinical implementation of MH therefore is still less than satisfactory (Brooks & Morrison 2008). Temperature has been recognized as a major determinant of myocardial function since more than a hundred years (Langendorff 1897). In isolated papillary muscles and trabeculae from rats (Shattock & Bers 1987, Hiranandani et al. 2006), rabbits (Shattock & Bers 1987), cats (Yeatman et al. 1969, Templeton et al. 1974), swine (Weisser et al. 2001) and human hearts (Weisser et al. 2001), a decrease in temperature increased isometric force development. Also in isolated hearts from rats (Fukunami & Hearse 1989), guineapigs (Nakae et al. 2001), rabbits (Mattheussen et al. 1996) and dogs (Monroe et al. 1964, Templeton et al. 1974, Suga et al. 1988), systolic pressure was higher, and the corresponding end-systolic pressure–volume relationship (ESPVR) was steeper when temperature decreased. In all studies, increased force and pressure development contrasted to delayed contraction, such that at a given heart rate (HR) the duration of systole was prolonged and relaxation was slowed (Mattheussen et al. 1996, Nakae et al. 2001). Studies on the actually ejecting heart in vivo at a level of mild to moderate hypothermia (>30 C) give a less uniform picture. Few studies reported an increased (Weisser et al. 2001) or maintained (London et al. 1988) CO, while the majority observed a decrease in CO (Goldberg 1958, Rittenhouse et al. 1971, London et al. 1988, Oung et al. 1992, Perez-de-Sa et al. 2002, Boddicker et al. 2005, Nishimura et al. 2005) that related to decreased HRs during cooling. At the single beat level, regional myocardial function increased (D’ Ambra et al. 1987) or decreased (Greene et al. 1989, Tveita et al. 1998) with cooling. Inotropy assessed by dP/dtmax increased (Weisser et al. 2001), remained unchanged (Rittenhouse et al. 1971) or decreased (Tveita et al. 1998) during MH. Pressure–volume analyses demonstrated an unchanged (at 35 C) (Fischer et al. 2005) or steeper (at 34 and 32 C) (Nishimura et al. 2005) ESPVR in animals, while in patients undergoing coronary artery bypass surgery, hypothermia flattened the slope of the ESPVR (Lewis et al. 2002). As in vitro, active relaxation was consistently impaired during MH also in vivo (Greene et al. 1989, Tveita et al. 1998, Weisser et al. 2001, Fischer et al. 2005). It is thus unclear from existing data whether MH acts as a positive inotrope also in vivo, and to what extent a putative positive inotropic effect is counteracted by relaxation abnormalities. Potentially confounding factors among the above mentioned studies comprise insufficient control of HR, load, invasive open-chest

Acta Physiol 2010, 199, 43–52

preparations, the method and precision of cooling, accumulation of cardiodepressant barbiturates, and the presence or lack of muscular relaxation. We thus aimed to assess the effect of MH on cardiac function in an experimental set-up that would mimic a clinical situation. Human-sized pigs were used in a closed-chest preparation, and the anaesthetic regimen was adapted from patient treatments including complete muscular relaxation to avoid shivering. Hypothermia was induced by a vena cava inferior cooling catheter that allowed for rapid and precise stepwise core temperature reduction. Systolic and diastolic function was evaluated at spontaneous and paced HRs by both pressure–time and pressure–volume analysis. Finally, preload reduction was applied to obtain loadindependent indices of myocardial function.

Materials and methods The study protocol was approved by the local bioethics committee of the district of Braunschweig, Germany, and conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).

Experimental model Ten female domestic pigs (69 5 kg) were fasted overnight, sedated with 0.25 mg kg)1 midazolam and 2.5 mg kg)1 ketamine, and transported to the operating room. Anaesthesia was introduced with 5 mg kg)1 thiopental, a tracheotomy was performed, and the animals were connected to a respirator (Ventilog 2; Draeger, Luebeck, Germany, 60% room air/40% oxygen) supplemented with 0.5% isoflurane. The respirator was set at 20 min)1, 5 mmHg positive end-expiratory pressure, and tidal volume was continuously adjusted to maintain the expiratory carbon dioxide concentration between 4 and 4.5%. Both common carotid arteries and jugular veins were dissected free, and sheaths (Avanti+, 8F, Cordis, Bridgewater, NJ, USA) were introduced into the vessels. Anaesthesia was then maintained with midazolam (0.2 mg kg)1 h)1), fentanyl (20 lg kg)1 h)1) and pancuronium (0.1 mg kg)1 h)1). Under fluoroscopic guidance, a Swan-Ganz catheter was positioned in the pulmonary artery. The catheter was equipped with an electric heating coil (Edwards Lifesciences CCO, Irvine, CA, USA), enabling measurement of cardiac output in minute intervals (Edwards Lifesciences Vigilance I). A pacing wire was placed in the right atrium, and a pressure–volume catheter (CA71103-PL, 7F, 12 electrodes; CD Leycom, Zoetermeer, the Netherlands) connected to a signal processing unit (Sigma-5 DF; CD Leycom) was placed in the left

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Acta Physiol 2010, 199, 43–52

ventricle (LV). The conductance catheter position was optimized to obtain stable pressure–volume loops. Both common femoral veins were dissected free, and sheaths (14F; St. Jude Medical, St Paul, MN, USA) were introduced. Via the left femoral vein, an intravascular vena cava inferior occlusion (VCO) balloon catheter was positioned at the level of the diaphragm (Boston Scientific, Natick, MA, USA). The right femoral vein was used to introduce an intravascular cooling catheter. This catheter consisted of a triple-lobed, helically wound, heparin-coated balloon allowing for the inflow and outflow of temperatured saline from a cooling unit, which was controlled by feed-back from an oesophageal temperature probe (SetPoint System; Radiant Medical, Redwood City, CA, USA). The speed of cooling was approx. 1 C per 30 min. The animals were anticoagulated by a bolus of 10 000 IE heparin, followed by continuous infusion of 1000 IE h)1 heparin through the side arms of each femoral sheath to prevent venous thrombosis. Volume was administered at a fixed rate of 3 mL kg)1 h)1, consisting of gelatine 4% (0.75 mL kg)1 h)1, Gelafundin 4%; Braun, Melsungen, Germany) and balanced crystalloid infusion (2.25 mL kg)1 h)1, Sterofundin; Braun) containing sodium, potassium, magnesium, chloride, lactate, and further enriched with 2.5% glucose. When the instrumentation process was completed, animals were allowed to stabilize for 30 min.

Experimental protocol Measurements were taken at 37, 35, 33 and 32 C. To calibrate the conductance catheter, blood conductivity (q), the slope factor a and parallel conductance (Vc) (10% hypertonic saline infusion) were determined at each temperature before measurements were begun (Steendijk & Baan 2000). During the protocol, a four-lead peripheral ECG and LV pressure recordings were used to monitor the animals. The volume signal was displayed on-line and calibrated offline during later data analysis. After calibration at each temperature, steady-state data were acquired at spontaneous HR, and a pair of blood samples was drawn from the aorta and the pulmonary artery. A maximum of four short-episode inferior vena cava occlusions (£20 beats) were performed until two technically stable runs without arrhythmia had been acquired. The tidal volume of the respirator was temporarily set to zero during vena cava occlusion. At each temperature, HR was increased by right atrial pacing in steps of 25 min)1 as long as pacing was followed by LV contraction, and haemodynamic measurements were repeated at steady state and during vena cava occlusion. When the protocol was finished, the animals were killed by bolus infusion of 80 mmol potassium chloride.

H Post et al.

Æ Cardiac function during mild hypothermia

Blood analysis Blood samples were processed immediately after withdrawal using a blood gas analyser (RapidLab 865; Bayer Healthcare, Leverkusen, Germany) equipped for the measurement of oxygen, carbon dioxide, pH, haemoglobin, sodium, potassium, glucose and lactate. All measurements were corrected for temperature within the analyser.

Data analysis and statistics Haemodynamic and conductance data were analysed offline by CircLab software (custom-made by P. Steendijk). Steady-state data were taken from two complete respiratory cycles. End-diastole was defined as the time point of zero crossing of dP/dt before its rapid upstroke. End-systole was defined as the time point of maximum pressure–volume ratio. Haemodynamic parameters analysed included HR, absolute systolic and diastolic time intervals (tsys and tdia, ms), systolic and diastolic intervals as a fraction of the cardiac cycle (%tsys, %tdia), end-diastolic (Ped) and end-systolic (Pes) LV pressure, the maximum and minimum of the first derivative of LV pressure (dP/dtmax and dP/dtmin), the monoexponential time constant of isovolumic relaxation (s), and enddiastolic and end-systolic volume (Ved and Ves). Stroke volume (SV) was calculated as Ved ) Ves. Whole body oxygen consumption (VO2) was calculated as the difference of aortic and pulmonary artery oxygen content multiplied by CO. Data acquired during VCO were used to derive pressure–volume relations (Burkhoff et al. 2005). Linear regressions were performed between end-diastolic pressure and volume (end-diastolic pressure–volume relationship, EDPVR) and ESPVR. As vena cava occlusion generated the rather linear part of the EDPVR, fitting by exponential equations did not improve the correlation coefficients and was therefore not applied in the present study. The slope and intercept of each relationship was calculated and averaged among the animals for a given temperature and HR. Only runs with an r-value >0.9 were included in the analysis. To provide physiologically relevant measures of diastolic and systolic function, enddiastolic and end-systolic volume intercepts were calculated at levels of pressures representative of the overall mean of the acquired data (Burkhoff et al. 2005). Systolic function was further assessed by calculating the slope of the relationship between stroke work and preload, and by calculating stroke work at a representative end-diastolic volume. LV end-diastolic pressure data obtained during left atrial pacing were plotted as a function of the ratio of the diastolic time interval and s, and the relationship was fitted by an exponential equation assuming an exponential decay.

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Cardiac function during mild hypothermia

Æ H Post et al.

Acta Physiol 2010, 199, 43–52

No adverse events in terms of arrhythmias or haemodynamic instability occurred during the protocol. Parallel conductance (Vc) (83 7, 86 7, 82 7 and 86 8 mL), the slope factor a (0.45 0.03, 0.43 0.03, 0.43 0.03 and 0.42 0.03) and blood conductivity (q) (132 4, 128 4, 126 4 and 129 6 ms) were not significantly different at 37, 35, 33 and, respectively, 32 C.

while Ped increased significantly at 32 C. The absolute and relative duration of systole was increased during hypothermia, whereas the absolute diastolic time interval remained unchanged (Table 1). dP/dtmax remained constant during hypothermia (P = 0.09), but dP/dtmin decreased markedly at lower temperatures in parallel with an increase in s (Fig. 3). Ved, Ves and SV did not change significantly. Arterial oxygen saturation remained at 99–100%, while central venous oxygen saturation increased and whole body oxygen consumption decreased progressively during hypothermia. Blood pH and levels of glucose, lactate, sodium, potassium and carbon dioxide remained in the normal range in both aortic and pulmonary artery blood samples during hypothermia. With the fixed infusion regime outlined above, haemoglobin decreased from 10.9 0.3 g dL)1 at 37 C to 10.2 0.3 g dL)1 at 35 C, 9.9 0.4 g dL)1 at 33 C (P < 0.05 vs. 37 C) and 9.6 0.4 g dL)1 at 32 C (P < 0.05 vs. 37 C).

LV function, systemic haemodynamics and metabolism at spontaneous HR

LV function and systemic haemodynamics during right atrial pacing

Heart rate decreased in relation to body temperature (Figs 1 and 2). Pes and CO did not change significantly,

Pacing frequency was increased until a mechanical alternans would occur. The highest common HR

All data are given as mean SEM. To compare parameters determined at 37, 35, 33 and 32 C, we performed one-way anova for repeated measurements. A two-way anova for repeated measurements was used to compare parameters determined at 37 and 33 C at different HRs. Post hoc analysis was performed by Tukey’s test. P < 0.05 indicated significant differences.

Results

150 125 100 75 50 25 0

LVP (mmHg)

ED ES

dPdt–1 (mmHg/s)

3000 1500

150 125 100 75 50 25 0

3000

1500

LVP (mmHg)

dPdt–1 (mmHg/s)

–1500 –1500

–3000

–3000 –3000 0.8 1.2 1.6 2.0 33 °C

Spontaneous

175 150 125 100 75 50 25 0 3000 1500 0

–1500 0.8 1.2 1.6 2.0 37 °C

ED ES

–1500 0.8 1.2 1.6 2.0 37 °C

–3000 0.8 1.2 1.6 2.0 33 °C

100 bpm

Figure 1 Original tracings of left ventricular (LV) pressure and LV dP/dt during normothermia and hypothermia at spontaneous heart rate (left) and during pacing at 100 min)1 (right). LV pressure fall is slowed and dP/dtmin is decreased at 33 C. The diastolic interval is shortened during hypothermia, which is potentiated during pacing with a subsequent increase in end-diastolic LV pressure. X-axis is time in seconds. ED, end-diastole; ES, end-systole.

LVPes (mmHg)

Heart rate (bpm) 90

140

130

80 *

120

6500

220

6000

200

5500

180

5000

160

4500

140

110

70 65

Cardiac output (ml/min) 7000 240

37 35 33 Temperature (°C)

100

37 35 33 Temperature (°C)

4000

37 35 33 Temperature (°C)

120

VO2 (mL min–1)

* *

37 35 33 Temperature (°C)

Figure 2 Heart rate and whole body oxygen consumption are deceased with hypothermia, whereas left ventricular end-systolic pressure and cardiac output do not change significantly. *P < 0.05 vs. 37 C.

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Acta Physiol 2010, 199, 43–52

H Post et al.

Table 1 Parameters at spontaneous heart rate 37 C tsys (ms) tdia (ms) Ped (mmHg) Ved (mL) Ves (mL) SV (mL) ven O2-Sat (%)

338 396 9 138 69 69 73.3

35 C 7 35 1 13 7 9 1.8

395 394 11 138 64 73 76.7

Systolic time (% cardiac cycle)

33 C 7* 35 1 12 9 7 2.3

437 392 12 129 59 70 79.5

32 C 14* 34 1 11 8 5 1.6*

464 383 13 134 65 70 80.2

64 *, # *, # 62 60 * 58 # 56 * 54 52 : 37 °C * : 33 °C 50 48 46 44 60 80 100 120 140 160 Heart rate (bpm)

21* 35 1* 12 10 5 1.3*

tsys, absolute duration of systole; tdia, absolute duration of diastole; Ped, left ventricular (LV) end-diastolic pressure; Ved, end-diastolic LV volume; Ves, end-systolic LV volume; SV, stroke volume; ven. O2-Sat, pulmonary artery oxygen saturation. *P < 0.05 vs. 37 C.

56 54

* *

52 48 46 37 35 33 Temperature (°C)

dP/dtmin (mmHg/s)

–1000 2200 –1400

100/min

: 37 °C : 33 °C

sp.

15 150/min

125/min

sp.

100/min

5 2

tdia/τ

τ (ms) 120

100 *

* 80

1600

–1600

1400

–1800

1200

–1200

1800

dP/dtmax (mmHg s–1)

2000

125/min 20

The predominant effect of hypothermia on LV pressure–volume loops was an up- and leftward shift, such that a given end-diastolic and end-systolic pressure was reached at lower volumes (Fig. 5). With linear regression analysis, there was no significant effect on the slopes of EDPVR and ESPVR (Table 3), but both relations showed a leftward shift evidenced by a decrease in the respective x-axis intercepts. Accordingly, calculated volumes for an end-diastolic pressure of 10 mmHg were decreased during hypothermia (Fig. 6). Similarly, calculated volumes for an end-systolic pressure of 120 mmHg were decreased during hypothermia, both at spontaneous HR and at 100 min)1. Preload recruitable stroke work tended to increase with hypothermia, and calculated stroke work for a given end-diastolic volume increased significantly with cooling. At HRs higher than 100 min)1 during MH, VCO frequently caused arrhythmia, so that no reliable loops could be determined.

A constant (C) was included as Ped could not reach zero values in the presence of positive end-expiratory pressures. Curve fit parameters were as follows: C = 9.4 0.7 mmHg; A = 701 599 mmHg; B = 1.4 0.3; r2 = 0.95; P = 0.001.

Pressure–volume relationships

Ped ¼ C þ A expð Bðtdia =sÞÞ

Ped (mmHg)

Figure 4 Pacing prolongs the relative duration of systole and decreases end-diastolic volume, which is potentiated during hypothermia. During pacing, left ventricular end-diastolic pressure increases to a far greater extent during hypothermia than during normothermia. This increase can be predicted from the ratio of diastolic time (tdia) and s. *P < 0.05 vs. spontaneous heart rate; #P < 0.05 vs. 37 C.

reached by all animals was 150 min)1 at 37 C, 125 min)1 at 35 and 33 C, and 100 min)1 at 32 C. Pacing caused a relative prolongation of systole that was potentiated at 33 C (Fig. 4), such that the decrease in absolute diastolic time during pacing was substantially more pronounced during hypothermia than at 37 C (Table 2). Pacing at 37 C slightly increased Ped and reduced Ved; these changes were again potentiated at 33 C. Pes was reduced at the highest pacing rates applied. dP/dtmax remained constant during pacing at 37 C and decreased during hypothermia. dP/dtmin remained decreased at 33 C vs. 37 C. SV decreased progressively during pacing in both groups. The increase in Ped could be predicted accurately from the ratio of diastolic time and s (tdia/s) by the following exponential equation:

Systolic time (% cardiac cycle)

Æ Cardiac function during mild hypothermia

37 35 33 Temperature (°C)

–2000

37 35 33 Temperature (°C)

Figure 3 The systolic time interval is increased progressively during hypothermia. dP/dtmax does not increase significantly (P = 0.09), while the changes in dP/dtmin and s indicate a profoundly slowed relaxation. *P < 0.05 vs. 37 C. 2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Cardiac function during mild hypothermia

Æ H Post et al.

Table 2 Parameters at paced heart rate sp

100 min

Acta Physiol 2010, 199, 43–52 33 °C

LVP (mmHg) )1

125 min

140

37 °C

150 min

120

tsys (ms) 37 C 338 7 311 33 C 437 14 366 tdia (ms) 37 C 396 35 275 33 C 392 34 226 Pes (mmHg) 37 C 111 4 112 33 C 125 8 123 dP/dtmax (mmHg s)1) 37 C 1519 138 1485 33 C 1908 206 1850 dP/dtmin (mmHg s)1) 37 C 1770 92 1868 33 C 1252 94 1486 s (ms) 37 C 50 2 47 33 C 100 15 71 Ved (mL) 37 C 138 13 130 33 C 129 11 111 SV (mL) 37 C 69 9 58 33 C 70 5 56

4* 5*

274 5* 297 5*

244 5*

5* 6*

201 5* 181 6*

151 5*

100 80 60 40

5 8

109 4 112 9*

101 4*

20 0

121 1499 98 1491 100 185 1735 191* 86 1861 69 1770 60 83* 1387 121 1* 3*

45 1* 66 2*

42 1*

13 10

115 11 86 9*

97 10*

7* 4*

45 5* 40 4*

34 4*

Pes, left ventricular (LV) end-systolic pressure; dP/dtmax, maximum first derivative of LV pressure; dP/dtmin, minimum first derivative of LV pressure; s, isovolumic relaxation constant; for other abbreviations, see Table 1. *P < 0.05 vs. spontaneous heart rate. P < 0.05 vs. 37 C.

Discussion It is unclear whether the robust positive inotropic effect of hypothermia seen in vitro actually occurs in the intact heart in vivo, and to what extent it may be limited by

100

120

140

160

LV volume (mL)

Figure 5 Original tracings of left ventricular pressure–volume loops during transient vena cava inferior occlusion at 37 and at 33 C. Heart rate is paced at 100 min)1. There is an up- and leftward shift in the loops at hypothermia, such that a given end-systolic pressure is reached at lower volumes, indicating increased inotropy. Lines indicate the end-systolic pressure– volume relationship.

relaxation abnormalities. The present study thus set out to test the effects of mild hypothermia on cardiac function in an experimental model close to the current therapeutic application of hypothermia in patients.

Metabolic balance during MH Heart rate decreased in the present study, while Pes, LV volumes and CO showed only minor changes not reaching statistical significance. In contrast, whole body oxygen consumption decreased substantially by 25% from 37 to 33 C. This is noteworthy, as the animals were in deep narcosis under full muscular relaxation, and thus the fall in oxygen consumption can be fully ascribed to the induction of hypothermia. Considering the resulting increase in central venous oxygen

Table 3 Slopes of linear regression analysis of the respective relationships 37 C EDPVRsp (mmHg mL)1) ESPVRsp (mmHg mL)1) ESPVR100/min (mmHg mL)1) PRSW (mmHg) PRSW100/min (mmHg) SW at Ved = 130 mL (mmHg mL) SW100/min at Ved = 130 mL (mmHg mL)

0.083 0.89 0.87 62 57 6353 6550

35 C

0.011 0.12 0.13 5 4 608 539

0.095 0.90 0.95 69 59 8560 7151

33 C

0.008 0.10 0.13 6 7 839* 950

0.093 0.86 0.99 65 63 8541 7961

32 C

0.006 0.10 0.13 7 4 673* 575*

0.104 0.92 1.03 67 68 8624 8024

0.018 0.16 0.18 8 6* 663* 629*

EDPVRsp, end-diastolic pressure–volume relationship at spontaneous heart rate; ESPVRsp, end-systolic pressure–volume relationship at spontaneous heart rate; ESPVR100/min, same at pacing with 100 min)1; PRSW, preload recruitable stroke work at spontaneous heart rate; PRSW100/min, same at pacing with 100 min)1; SW at Ved = 130 mL, calculated stroke work at an end-diastolic volume of 130 mL at spontaneous heart rate; SW100/min at Ved = 130 mL, same at pacing with 100 min)1. *P < 0.05 vs. 37 C.

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Acta Physiol 2010, 199, 43–52

200

Calculated volume (mL) at Ped = 10 mmHg

H Post et al. Calculated volume (mL) at Pes = 120 mmHg

100

-sp.HR-

-100 bpm-

180 160 140

* 60

120 40 100 20

80 60 37°C 35°C 33°C 32°C

0 37°C 35°C 33°C 32°C

37°C 35°C 33°C 32°C

Figure 6 With decreasing temperatures, the calculated volume at an end-diastolic pressure of 10 mmHg is decreased, indicating impaired end-diastolic distensibility (spontaneous heart rate). The calculated volume at an end-systolic pressure of 120 mmHg is decreased during hypothermia both at spontaneous heart rate (left) and at 100 min)1 (right), indicating increased inotropy. *P < 0.05 vs. 37 C.

saturation, while other metabolic parameters such as pH and arterial lactate concentration remained constant, we demonstrated that the metabolic balance was fully preserved during MH in the present study. It is thus conceivable that the moderate drop in CO seen in previous studies during MH was not indicative of an impaired oxygen supply, but rather reflected a reduced metabolic rate that could be met with less CO.

Cardiac inotropy during MH In isolated pig myocardial muscle strips, the increase in twitch force during hypothermia is characterized by a prolongation of time to peak tension and time to 50% relaxation. With this slowing of contraction, the maximum rate of force development remained constant in spite of increased maximum force (Weisser et al. 2001). This pattern is in contrast to, for example, the effect of beta-adrenoreceptor agonists, which typically increase both maximum force and speed of contraction. A dichotomy between speed and force of contraction during hypothermia was also described in isolated hearts, where end-systolic elastance, but not dP/dtmax increased during cooling (Fukunami & Hearse 1989). Also in the present study in vivo, dP/dtmax increased to only a minor degree below the level of significance (P = 0.09), which is not unexpected as the systolic time interval increased during MH. On the other hand, pressure–volume loops were shifted leftwards, implying increased inotropy. Evaluation of the ESPVR has to consider both the x-axis intercept (V0) and the slope of the ESPVR (Ees), such that the placement of the ESPVR in total at the level of physiological pressure values is assessed (Burkhoff et al. 2005). In the present study, the slope of the ESPVR remained constant, while calculated end-systolic volumes at 120 mmHg decreased progres-

Æ Cardiac function during mild hypothermia

sively during hypothermia. This parallel leftward shift in the ESPVR thus indicates increased inotropy in spite of an unchanged Ees. This is further supported by a trendwise increase in the slope of the relationship between stroke work and preload (PRSW), and a significant increase in calculated stroke work at a given preload (Table 3). In this context, data from the study by Lewis et al. (2002) in patients undergoing bypass surgery can be interpreted differently. The authors report that the slope of the ESPVR was lower during MH, and conclude that MH exerts a negative inotropic effect. However, at a given end-systolic pressure of, for example, 90 mmHg, end-systolic volumes were lower rather than higher vs. control; this indicates an increase rather than a loss of inotropy. Taken together, our data show that positive inotropy during hypothermia in vivo becomes apparent not in terms of accelerated pressure development (minor increase in dP/dtmax), but in terms of increased end-systolic stiffness (leftward shift in ESPVR). Of note, it is unlikely that this shift is secondary to increased afterload, as end-diastolic volumes tended to decrease with cooling.

Diastolic function during MH In the present study, the induction of MH impaired diastolic function by three factors. First, LV relaxation was delayed by small decreases in temperature, as indicated by significant changes in dP/dtmin and s already at the small step from 37 to 35 C. Second, with the prolongation of systole, the diastolic time interval remained constant during hypothermia-induced bradycardia, which is clearly different from resting bradycardia during normothermia. Finally, we demonstrate a leftward shift in the EDPVR, representing a third aspect of diastolic dysfunction during MH. We addressed the relevance of slowed relaxation by right atrial pacing. At 37 C, pacing led to an increase in the systolic time fraction interval and of Ped; these changes were potentiated during MH. According to the monoexponential fit of s, the diastolic time period necessary for complete active relaxation, when estimated as a 97% decay of LV pressure, can be calculated as 3.5 · s (Weisfeldt et al. 1978, Leite-Moreira & Gillebert 2000). We plotted Ped data obtained at spontaneous and paced HRs as a function of the ratio of diastolic time and s, and indeed observed a close relationship with a steep increase at ratio levels of less than 3.5. These data indicate that during pacing, incomplete LV relaxation and reduced LV suction in early diastole significantly contribute to elevated enddiastolic pressures during MH. This finding further provides experimental evidence for Hay’s hypothesis (Hay et al. 2005) who predicted – in a computer simulation – that at a given HR prolongation of systole

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Cardiac function during mild hypothermia

Ă&#x2020; H Post et al.

and slowed relaxation could potentially increase enddiastolic pressures, in particular at higher HRs. However, in contrast to the model assumption of Hay et al. (2005), increased Ped during MH in the present study occurred at significantly lower end-diastolic volumes, and the leftward shift in EDPVR at spontaneous HR during MH corresponded to ratio values >4. This indicates that passive LV stiffness is also increased during MH.

Potential subcellular mechanisms of altered cardiac function during MH On the myocyte level, cardiac contraction is coupled to electrical excitation by phasic release of calcium from the sarcoplasmic reticulum. While several studies have addressed the effect of moderate and deep hypothermia on excitationâ&#x20AC;&#x201C;contraction coupling, only few studies have been performed at mild hypothermia (>30 C). Hypothermia prolongs the cardiomyocyte action potential (Bjornstad et al. 1993), which could theoretically increase cardiomyocyte calcium influx and content. However, in cardiac muscle strips from rats and pigs, sarcoplasmic reticulum calcium content was unchanged (Weisser et al. 2001), and the calcium transient was delayed, but not increased in its amplitude at levels of MH (Weisser et al. 2001, Hiranandani et al. 2006). It is thus unlikely that MH exerts a positive inotropic effect by higher intracardiomyocyte calcium levels. This is in contrast to, for example, beta-adrenoceptor activation, which increases sarcoplasmic reticulum calcium content, and typically accelerates and increases the calcium transient, thereby increasing the number of crossbridges recruited for force development. The increase in inotropy by hypothermia therefore is more energy efficient than, for instance, by adrenaline, as only little additional energy is required to handle intracellular calcium (Suga et al. 1983, 1988). At unchanged calcium transients, an increased calcium responsiveness of the myofilaments has to explain increased force development during MH. Such increase can in principal be caused by increased calcium sensitivity or an increase in calcium-activated force (Lee & Allen 1997). While skinned cardiomyocyte calcium sensitivity was increased at 25 C (Sprung et al. 1994), no change was observed at 29 C (Harrison & Bers 1989) or at the level of MH (Nakae et al. 2001). Also alkalosis, which would increase myofilament calcium sensitivity, was not observed between 37 and 32 C in isolated hearts (Kusuoka et al. 1991). The latter study, however, demonstrated an increase in calcium-activated force (Kusuoka et al. 1991). Of note, this finding related to temperatures >30 C, while at lower temperatures, an opposite effect was observed in both mammalian skeletal and myocardial preparations

Acta Physiol 2010, 199, 43â&#x20AC;&#x201C;52

(reviewed in Kusuoka et al. 1991). Mechanisms of increased calcium-activated force during MH are unclear, but may comprise a slower reaction rate of calcium and troponin C, and/or slowed crossbridge cycling (Henderson & Cattell 1976). Both mechanisms could well explain prolonged contraction and relaxation together with increased force development. Lower relaxation rates are further likely to result from a temperature-dependent slowing of sarcoplasmic calcium ATPase activity (Inesi & Watanabe 1967), such that resequestration of calcium during diastole is delayed. We further demonstrated a leftward shift in EDPVR during hypothermia. This was observed before in isolated hearts (Remensnyder & Austen 1965, Templeton et al. 1974) and attributed to a higher viscosity of the heart muscle during MH (Templeton et al. 1974). Also, phosphorylation changes in cardiomyocyte structural proteins, as, for instance, titin, have been linked to an altered EDPVR (Kass et al. 2004); however, this has not been tested during MH, yet.

Experimental limitations We aimed to provide a clinically relevant set-up and used anaesthetics and analgesics that are as well used in humans. Of note, pancuronium, fentanyl and midazolam do not exert overt direct cardio-depressant effects, and isoflurane is metabolized only to a minor extent. We thus minimized the likelihood that drug accumulation during hypothermia-induced slowing of metabolism or a stress response to cooling would bias cardiac function data. However, with this regime, baseline LV ejection fraction was approx. 50% and lower than in conscious pigs and humans. Our findings have thus to be viewed under the reservation of a depressed baseline cardiac function.

Conclusion We show that MH profoundly reduces whole body oxygen demand that outweighs a decrease in cardiac output. MH has a small but consistent positive inotropic effect in terms of end-systolic stiffness, while the duration of contraction is prolonged and dP/dtmax remains unchanged. Diastolic function is impaired by shortening of the diastolic time interval, delayed active relaxation and decreased end-diastolic distensibility; however, this is largely compensated for by spontaneous bradycardia The increase in inotropy and the decrease in whole body oxygen consumption warrant further experimental and clinical studies addressing the potential benefit of hypothermia when applied to the acutely failing heart and during cardiogenic shock.

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Acta Physiol 2010, 199, 43–52

Conflict of interest No conflict of interest is declared.

References Bjornstad, H., Tande, P.M., Lathrop, D.A. & Refsum, H. 1993. Effects of temperature on cycle length dependent changes and restitution of action potential duration in guinea pig ventricular muscle. Cardiovasc Res 27, 946–950. Boddicker, K.A., Zhang, Y., Zimmerman, M.B., Davies, L.R. & Kerber, R.E. 2005. Hypothermia improves defibrillation success and resuscitation outcomes from ventricular fibrillation. Circulation 111, 3195–3201. Brooks, S.C. & Morrison, L.J. 2008. Implementation of therapeutic hypothermia guidelines for post-cardiac arrest syndrome at a glacial pace: seeking guidance from the knowledge translation literature. Resuscitation 77, 286– 292. Burkhoff, D., Mirsky, I. & Suga, H. 2005. Assessment of systolic and diastolic ventricular properties via pressure– volume analysis: a guide for clinical, translational, and basic researchers. Am J Physiol Heart Circ Physiol 289, H501– H512. D’ Ambra, M.N., Magrassi, P., Lowenstein, E., Kyo, S., Austen, W.G., Buckley, M.J. & LaRaia, P.J. 1987. Myocardial temperature variation: effect on regional function and coronary flow in dogs. Am J Physiol 252, H448–H455. Fischer, U.M., Cox, C.S., Jr, Laine, G.A., Mehlhorn, U. & Allen, S.J. 2005. Mild hypothermia impairs left ventricular diastolic but not systolic function. J Invest Surg 18, 291–296. Fukunami, M. & Hearse, D.J. 1989. The inotropic consequences of cooling: studies in the isolated rat heart. Heart Vessels 5, 1–9. Goldberg, L.I. 1958. Effects of hypothermia on contractility of the intact dog heart. Am J Physiol 194, 92–98. Greene, P.S., Cameron, D.E., Mohlala, M.L., Dinatale, J.M. & Gardner, T.J. 1989. Systolic and diastolic left ventricular dysfunction due to mild hypothermia. Circulation 80, III44– III48. Harrison, S.M. & Bers, D.M. 1989. Influence of temperature on the calcium sensitivity of the myofilaments of skinned ventricular muscle from the rabbit. J Gen Physiol 93, 411– 428. Hay, I., Rich, J., Ferber, P., Burkhoff, D. & Maurer, M.S. 2005. Role of impaired myocardial relaxation in the production of elevated left ventricular filling pressure. Am J Physiol Heart Circ Physiol 288, H1203–H1208. Henderson, A.H. & Cattell, M.R. 1976. Length-induced changes in activation during contraction. A study of mechanical oscillations in strontium-mediated contractions of cat and frog heart muscle. Circ Res 38, 289–296. Hiranandani, N., Varian, K.D., Monasky, M.M. & Janssen, P.M. 2006. Frequency-dependent contractile response of isolated cardiac trabeculae under hypo-, normo-, and hyperthermic conditions. J Appl Physiol 100, 1727–1732. Inesi, G. & Watanabe, S. 1967. Temperature dependence of ATP hydrolysis and calcium uptake by fragmented sarcoplasmic membranes. Arch Biochem Biophys 121, 665–671.

H Post et al.

Æ Cardiac function during mild hypothermia

International Liaison Committee on Resuscitation. 2005. International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science with Treatment Recommendations. Part 4: advanced life support. Resuscitation 67, 213–247. Kass, D.A., Bronzwaer, J.G. & Paulus, W.J. 2004. What mechanisms underlie diastolic dysfunction in heart failure? Circ Res 94, 1533–1542. Kusuoka, H., Ikoma, Y., Futaki, S., Suga, H., Kitabatake, A., Kamada, T. & Inoue, M. 1991. Positive inotropism in hypothermia partially depends on an increase in maximal Ca(2+)-activated force. Am J Physiol 261, H1005–H1010. Langendorff, O. 1897. Untersuchungen am u¨berlebenden Sa¨ugetierherzen. U¨ber den Einfluss von Wa¨rme und Ka¨lte auf das Herz der warmblu¨tigen Thiere. Pflu¨gers Arch 61, 355– 400. Lee, J.A. & Allen, D.G. 1997. Calcium sensitisers: mechanisms of action and potential usefulness as inotropes. Cardiovasc Res 36, 10–20. Leite-Moreira, A.F. & Gillebert, T.C. 2000. The physiology of left ventricular pressure fall. Rev Port Cardiol 19, 1015– 1021. Lewis, M.E., Al Khalidi, A.H., Townend, J.N., Coote, J. & Bonser, R.S. 2002. The effects of hypothermia on human left ventricular contractile function during cardiac surgery. J Am Coll Cardiol 39, 102–108. London, M.J., Sybert, P.E., Mangano, D.T., Fisher, D.M., Bainton, C.R. & Hickey, R.F. 1988. Surface-induced hypothermia: effects on coronary blood flow autoregulation and vascular reserve. J Surg Res 45, 481–495. Mattheussen, M., Mubagwa, K., Van Aken, H., Wusten, R., Boutros, A. & Flameng, W. 1996. Interaction of heart rate and hypothermia on global myocardial contraction of the isolated rabbit heart. Anesth Analg 82, 975–981. Monroe, R.G., Strang, R.H., Lafrage, C.G. & Levy, J. 1964. Ventricular performance, pressure–volume relationships, and O2 consumption during hypothermia. Am J Physiol 206, 67–73. Nakae, Y., Fujita, S. & Namiki, A. 2001. Isoproterenol enhances myofilament Ca(2 + ) sensitivity during hypothermia in isolated guinea pig beating hearts. Anesth Analg 93, 846–852. Neumar, R.W., Nolan, J.P., Adrie, C., Aibiki, M., Berg, R.A., Bottiger, B.W., Callaway, C., Clark, R.S., Geocadin, R.G., Jauch, E.C. et al. 2008. Post-cardiac arrest syndrome: epidemiology, pathophysiology, treatment, and prognostication. A consensus statement from the International Liaison Committee on Resuscitation (American Heart Association, Australian and New Zealand Council on Resuscitation, European Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Asia, and the Resuscitation Council of Southern Africa); the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiovascular Surgery and Anesthesia; the Council on Cardiopulmonary, Perioperative, and Critical Care; the Council on Clinical Cardiology; and the Stroke Council. Circulation 118, 2452–2483. Nishimura, Y., Naito, Y., Nishioka, T. & Okamura, Y. 2005. The effects of cardiac cooling under surface-induced

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x

Cardiac function during mild hypothermia

Æ H Post et al.

hypothermia on the cardiac function in the in situ heart. Interact Cardiovasc Thorac Surg 4, 101–105. Oung, C.M., English, M., Chiu, R.C. & Hinchey, E.J. 1992. Effects of hypothermia on hemodynamic responses to dopamine and dobutamine. J Trauma 33, 671–678. Perez-de-Sa, V., Roscher, R., Cunha-Goncalves, D., Larsson, A. & Werner, O. 2002. Mild hypothermia has minimal effects on the tolerance to severe progressive normovolemic anemia in swine. Anesthesiology 97, 1189–1197. Remensnyder, J.P. & Austen, W.G. 1965. Diastolic pressure– volume relationships of the left ventricle during hypothermia. J Thorac Cardiovasc Surg 49, 339–351. Rittenhouse, E.A., Ito, C.S., Mohri, H. & Merendino, K.A. 1971. Circulatory dynamics during surface-induced deep hypothermia and after cardiac arrest for one hour. J Thorac Cardiovasc Surg 61, 359–369. Shattock, M.J. & Bers, D.M. 1987. Inotropic response to hypothermia and the temperature-dependence of ryanodine action in isolated rabbit and rat ventricular muscle: implications for excitation–contraction coupling. Circ Res 61, 761–771. Sprung, J., Stowe, D.F., Kampine, J.P. & Bosnjak, Z.J. 1994. Hypothermia modifies anesthetic effect on contractile force and Ca2+ transients in cardiac Purkinje fibers. Am J Physiol 267, H725–H733. Steendijk, P. & Baan, J. 2000. Comparison of intravenous and pulmonary artery injections of hypertonic saline for the assessment of conductance catheter parallel conductance. Cardiovasc Res 46, 82–89.

Acta Physiol 2010, 199, 43–52 Suga, H., Hisano, R., Goto, Y., Yamada, O. & Igarashi, Y. 1983. Effect of positive inotropic agents on the relation between oxygen consumption and systolic pressure volume area in canine left ventricle. Circ Res 53, 306–318. Suga, H., Goto, Y., Igarashi, Y., Yasumura, Y., Nozawa, T., Futaki, S. & Tanaka, N. 1988. Cardiac cooling increases Emax without affecting relation between O2 consumption and systolic pressure–volume area in dog left ventricle. Circ Res 63, 61–71. Templeton, G.H., Wildenthal, K., Willerson, J.T. & Reardon, W.C. 1974. Influence of temperature on the mechanical properties of cardiac muscle. Circ Res 34, 624–634. Tveita, T., Ytrehus, K., Myhre, E.S. & Hevroy, O. 1998. Left ventricular dysfunction following rewarming from experimental hypothermia. J Appl Physiol 85, 2135–2139. Weisfeldt, M.L., Frederiksen, J.W., Yin, F.C. & Weiss, J.L. 1978. Evidence of incomplete left ventricular relaxation in the dog: prediction from the time constant for isovolumic pressure fall. J Clin Invest 62, 1296–1302. Weisser, J., Martin, J., Bisping, E., Maier, L.S., Beyersdorf, F., Hasenfuss, G. & Pieske, B. 2001. Influence of mild hypothermia on myocardial contractility and circulatory function. Basic Res Cardiol 96, 198–205. Yeatman, L.A., Jr, Parmley, W.W. & Sonnenblick, E.H. 1969. Effects of temperature on series elasticity and contractile element motion in heart muscle. Am J Physiol 217, 1030– 1034.

2010 The Authors Journal compilation 2010 Scandinavian Physiological Society, doi: 10.1111/j.1748-1716.2010.02083.x