International Journal of Cardiology 170 (2014) 388–393
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Remote ischemic preconditioning with – but not without – metabolic support protects the neonatal porcine heart against ischemia-reperfusion injury Michael R. Schmidt a,⁎, Nicolaj B. Støttrup a, Hussain Contractor b, Janus A. Hyldebrandt c, Mogens Johannsen d, Christian M. Pedersen a, Rune Birkler d, Houman Ashrafian b, Keld E. Sørensen a, Rajesh K. Kharbanda b, Andrew N. Redington e, Hans E. Bøtker a a
Department of Cardiology, Aarhus University Hospital, Skejby, Brendstrupgaardsvej, Aarhus N, DK-8200, Denmark Department of Cardiology, The John Radcliffe, Headley Way, Headington, Oxford, OX3 9DU, United Kingdom c Department of Anesthesiology and Intensive Care, Aarhus University Hospital, Skejby, Brendstrupgaardsvej, Aarhus N, DK-8200, Denmark d Department of Forensic Medicine, Aarhus University Hospital, Skejby, Brendstrupgaardsvej, Aarhus N, DK-8200, Denmark e Division of Cardiology, Hospital for Sick Children, Toronto, M5G 1X8, Canada b
a r t i c l e
i n f o
Article history: Received 28 August 2013 Received in revised form 21 October 2013 Accepted 2 November 2013 Available online 12 November 2013 Keywords: Preconditioning Metabolism Ischemia-reperfusion injury Pediatric Insulin
a b s t r a c t Background: While remote ischemic preconditioning (rIPC) protects the mature heart against ischemiareperfusion (IR) injury, the effect on the neonatal heart is not known. The neonatal heart relies almost solely on carbohydrate metabolism, which is modified by rIPC in the mature heart. We hypothesized that rIPC combined with metabolic support with glucose-insulin (GI) infusion improves cardiac function and reduces infarct size after IR injury in neonatal piglets in-vivo. Methods and results: 32 newborn piglets were randomized into 4 groups: control, GI, GI + rIPC and rIPC. GI and GI + rIPC groups received GI infusion continuously from 40 min prior to ischemia. rIPC and GI + rIPC groups underwent four cycles of 5 min limb ischemia. Myocardial IR injury was induced by 40 min occlusion of the left anterior descending artery followed by 2 h reperfusion. Myocardial lactate concentrations were assessed in microdialysis samples analyzed by mass spectrometry. Infarct size was measured using triphenyltetrazolium chloride staining. Systolic recovery (dP/dtmax as % of baseline) after 2 h reperfusion was 68.5 ± 13.8% in control, 53.7 ± 11.2% in rIPC (p b 0.05), and improved in GI (83.6 ± 18.8%, p b 0.05) and GI + rIPC (87.0 ± 15.7%, p b 0.01). Conclusion: rIPC + GI protects the neonatal porcine heart against IR injury in-vivo. rIPC alone has detrimental metabolic and functional effects that are abrogated by simultaneous GI infusion. © 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Impaired ventricular systolic and diastolic function are important contributors to the mortality and morbidity associated with pediatric cardiac surgery [1,2], particularly after complex procedures in the neonate and young infant. The immature heart is structurally, functionally, and metabolically different from the adult heart and undergoes fundamental changes around the time of birth. For example, it is well known that cardiac energy generation shifts from glycolysis to oxidative lipid metabolism reflecting the increasing mechanical and metabolic demands that occur during the first few postnatal days and weeks, however our understanding of the detailed ontogeny of these changes remains incomplete. Understanding such physiological transitions in the immediately postnatal myocardium is a prerequisite to manipulating
⁎ Corresponding author. Tel.: +45 7845 6020; fax: +45 7845 6022. E-mail address: [email protected] (M.R. Schmidt). 0167-5273/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2013.11.020
the mechanisms of myocardial protection and to devise new strategies for protecting the immature heart during cardiac surgery. Preoperative remote ischemic preconditioning (rIPC), induced by cycles of ischemia and reperfusion of a limb, has recently been shown to prevent myocardial damage during surgery in older children [3] and adults [4]. In contrast, the impact of rIPC on the immature heart remains unknown. This is important, particularly as local ischemic preconditioning, where the cycles of ischemia and reperfusion are applied directly to the tissue subjected to prolonged ischemia, has suggested that the responses may display distinct maturational dependence. Some studies have suggested that the fetal heart is resistant to local ischemic preconditioning, and that its cardioprotective effect only becomes effective after 1–4 weeks (depending on species) after birth [5–7]. In an important recent study by Jones et al., no detectable cardioprotection was achieved by rIPC in hypoxic neonates undergoing cardiopulmonary bypass surgery [8], which was most recently supported by Pepe et al. showing that cardioprotective intracellular signaling mechanisms may already be activated in cyanosed children eligible for
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
surgical correction of tetralogy of Fallot and are not further enhanced by rIPC [9]. However, others have shown effective pharmacologic and ischemic preconditioning in normoxic immature rabbit hearts [10], and isolated neonatal cardiomyocytes [11,12]. Given these inconsistent observations, it is unsurprising that the mechanisms underlying ‘preconditioning resistance’ in the neonate remain unknown, although the immature myocardium appears intrinsically more resistant to hypoxia and possibly ischemia [6,13], raising the possibility that it is already ‘preconditioned’ in some way. Furthermore, the relevance of these studies to remote ischemic preconditioning (rIPC) is questionable, as rIPC does not involve ischemic bursts to the myocardium itself to induce the preconditioned state. However, in a recently published study, we demonstrated that also rIPC prior to ischemia-reperfusion injury may have detrimental effects in the isolated neonatal rabbit heart [14]. Importantly, most experimental studies of ischemic cardiac conditioning have been performed on isolated rodent hearts. While these models are relatively simple to establish and effectively allow for detailed mechanistic studies, they poorly reflect the complex physiology of ischemia-reperfusion injury in the intact animal or human. In this study, we used an in-vivo newborn piglet model to examine the effects of rIPC and metabolic support on function, infarct development and crude indicators of cardiac metabolism during and after regional myocardial ischemia. We used piglets less than 4 days old to study the effects of rIPC in hearts prior to the transition towards mature metabolism. We hypothesized that the adverse effects of IPC attributable to a relative reduction of carbohydrate metabolism could be counteracted by concurrent stimulation of the carbohydrate metabolism by glucose-insulin infusion.
2. Methods 2.1. Design Thirty-two newborn piglets (1–4 days old) were randomized into four groups: Control group (no pretreatment during a 40 min observation period prior to myocardial ischemia), pretreatment with rIPC (4 cycles of 5 min tourniquet occlusion of the lower left limb followed by 5 min reperfusion), pretreatment with glucose-insulin (GI, insulin 100 mU/ kg/h in 20% glucose continuously from 40 minutes prior to ischemia), and pretreatment with combined rIPC + GI (Fig. 1).
2.2. Animals The animals were handled according to the guidelines of the Danish Committee for Animal Research (Dyreforsøgstilsynet, Copenhagen, Denmark). The piglets stayed with their mother until the day of the study.
389
2.3. Experimental preparation and protocol Anesthesia was induced with midazolam (0.5 mg/kg intramuscularly) and azaperone (4 mg/kg intramuscularly) followed by etomidate (0.5 mg/kg intravenously). The piglets were intubated and ventilated. Anesthesia was maintained with an infusion of pentobarbital. Two 4 F sheaths were placed in the left jugular vein and carotid artery, respectively. A 3 F Mikro-Tip high fidelity pressure manometer (Millar Instruments, US) was placed through the arterial sheath into the left ventricle. After randomization, a midline thoracotomy and pericardiotomy were performed and CMA-20 Elite microdialysis probes (CMA/Harvard Apparatus, Sweden) inserted into the free wall of the left (area at risk) and the right (area not at risk) ventricle using an introducer cannula (0.45 mm × 12 mm) (and) fixed with tissue glue (LiquiBand, Devon, UK). Particular attention was paid to place the microdialysis catheters in the area at risk and not at risk, respectively. A suture was placed around the left anterior descending artery 2–3 mm from its origin and the ends inserted into a plastic tube to form a tourniquet. All piglets then underwent 40 min pretreatment according to group allocation, followed by 40 min of regional ischemia (induced by tightening the tourniquet) and 120 min of reperfusion (by releasing the tourniquet). Body temperature was maintained within the normal range (38.5–39.5 °C) using a warming blanket. Hydration was maintained with continuous intravenous infusion of a solution containing isotonic sodium chloride (4 ml/(kg h)) and additional potassium supplied when needed to maintain serum potassium above 4.0 mmol/l. Coronary occlusion was confirmed by epicardial cyanosis and decrease in blood pressure, while reperfusion was verified by epicardial hyperemia. Indices of left ventricular function were acquired continuously, sampled via high-fidelity analogue to digital hardware to dedicated data acquisition software (Notocord®, France) and stored on a computer for later analysis. Continuous indices of systolic (dP/dtmax) and diastolic (dP/dtmin, tau) function were calculated automatically by the software (Fig. 2). At the end of the protocol, the tourniquet around the left anterior descending artery was tightened again and 20 ml of 1% Evans blue solution injected into the left atrium to delineate the area-at-risk. After a few seconds allowing the myocardium to be perfused by the dye, piglets were sacrificed by excision of the heart.
2.4. Microdialysis and assessment of myocardial lactate and inosine levels Following implantation of the microdialysis probes (polyarylethersulphone membrane length 10 mm, molecular cut-off 20 kDa; CMA Elite 20, CMA/Harvard Apparatus, Sweden) a 40 min wash-out period was allowed for lactate and inosine to reach equilibrium in the perturbed tissue before the experimental protocol of 40 min of stabilization, 40 min of regional ischemia, and 120 min (during which microdialysis samples were collected for the first 40 min) of reperfusion was commenced. Perfusion fluid consisted of Krebs–Henseleit solution deoxygenated with 95% N2 and 5% CO2. Pump perfusion velocity (Univentor 801) was 1 μl/min. The sampling time for each fraction was 10 min, and samples were harvested by a customized fraction collector (Univentor 820 Microsampler, Univentor Limited, Zejtun, Malta). All samples were cooled to 8 °C during sampling and stored at −80 °C until time of analysis. Following separation in a Waters ACQUITYTM Ultra-performance liquid chromatography (UPLC) system (Waters Corp., Milford, MA, USA), lactate and inosine were quantitated using a Waters XevoTM triple quadruple tandem mass spectrometer (Waters Corp., Manchester, UK) with a Z-spray electrospray ionization source operating both in the positive and negative ion modes. All analytes were verified using two daughter ions when possible and deuterated internal standards for selected components.[15] The final interstitial concentrations were calculated by correction for relative recovery rate. To ensure sufficient relative recovery, a series of in vitro
Fig. 1. Study design. Four groups all undergoing 40 min of pretreatment, 40 min of ischemia and 2 h of reperfusion.
390
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
LV dP/dt max Control Glucose-insulin
2000
Glucose-insulin + rIPC rIPC
1500
mm Hg/s
Table 1 Group statistics on weigh, sex and age.
1000
500
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190
0
Weight (kg) (mean ± SD) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Male sex (n, %) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Age (days) (median[min;max]) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC
1.21 1.30 1.18 1.26
± ± ± ±
0.41 0.51 0.28 0.40
3 (37.5%) 4 (50%) 5 (62.5%) 3 (37.5%) 2[1,4] 2[1,4] 2[1,4] 2[1,4]
min Fig. 2. Maximal derived pressure (dP/dtmax) average of 10 s recording in 10-min intervals; see Results section for further description and comparison of groups. Error bars: SEM.
experiments were conducted as previously published.[15] Specific recovery rate of lactate was 40% and of inosine 87%. Baseline values and mean concentrations from stabilization (pre-ischemia), regional ischemia, and reperfusion were used for statistical comparison (see below). 2.5. Assessment of myocardial infarction The explanted hearts were frozen immediately at −80 °C for 20 min, the right ventricular free wall removed and the left ventricle subsequently sliced into 2–3 mm transverse sections and incubated in 1.25% 2,3,5-triphenyltetrazoliumchloride dissolved in phosphate buffer (Sigma, St. Louis, Mo, USA) at 37 °C for 5 min at pH 7.4. Subsequently, the hearts were stored overnight in 10 ml of Lillie's solution (4% formaldehyde buffer, VWR International, Denmark). Next day, each heart was weighed and scanned with a flatbed scanner (HP ScanJet 4300C, Hewlett Packard, Palo Alto, CA, USA). The AAR and size of infarction (IS) were assessed by computer planimetry (Photoshop CS3, Adobe systems incorporated, San Jose, California). Infarct size is expressed as a ratio, IS of AAR, weighted with the weight of each individual slice. All analyses were performed by a blinded observer. 2.6. Statistics Hemodynamic measurements and metabolomic results were compared between groups using repeated measures two-way ANOVA with Bonferroni correction comparing intervention groups to controls. As occasional lactate and inosine samples were missing, mean concentrations from pre-ischemia, regional ischemia, and reperfusion were used for the ANOVA analysis. Data satisfied criteria for normal distribution and there were no significant outliers. Infarct size was compared between groups using one-way ANOVA.
3. Results A total of 48 newborn piglets were used. Sixteen died due to technical or surgical problems and were excluded from the study, of which ten died prior to randomization and 6 during protocol (1 control, 2 GI, 1 GI + rIPC, 2 rIPC). There was no difference between groups regarding baseline characteristics (weight, sex, age) (Table 1). 3.1. Baseline characteristics Indices of cardiac function did not differ between groups at baseline with regards to left ventricular systolic and diastolic pressure, heart rate (HR), systolic blood pressure, minimal and maximal dP/dt (dP/dtmin and dP/dtmax), end diastolic pressure (EDP) or tau (Table 2). 3.2. Left ventricular function There was no difference between groups regarding left ventricular pressure and HR during the protocol (Table 2). All animals showed a significant drop in dP/dtmax to 41% ± 10% (controls), 20% ± 6% (rIPC), 62% ± 11% (GI) and 60% ± 11% (GI + rIPC) of baseline (p b 0.01 in
all groups) during ischemia and significant post-ischemic recovery to 68.5% ± 13.8% controls, p b 0.05), 53.7% ± 11.2% (rIPC, p b 0.05), 83.6% ± 18.8% (GI, p b 0.01) and 87.0% ± 15.7% (GI + rIPC, p b 0.01), of baseline, compared to late ischemia, (Fig. 3). However, in the rIPC group, dP/dtmax was significantly lower than in controls prior to ischemia (p b 0.05), throughout ischemia (p b 0.01) and during reperfusion (p b 0.01). Similarly, all groups showed significant impairment of diastolic function (i.e. increased dP/dtmin, tau and EDP) (Table 2) during ischemia and significant recovery during reperfusion, but the rIPC group deteriorated more during ischemia (p b 0.05) and recovered less during reperfusion (p b 0.01) compared to controls. 3.3. Lactate and inosine In the area at risk, an expected increase in interstitial lactate occurred during ischemia and reperfusion. However, only the rIPC group was significantly different from the controls (Fig. 3a). In the area not at risk, lactate levels remained stable in control, GI and GI + rIPC groups but increased in the rIPC group (Fig. 3b). In the area at risk, there were no statistically significant differences between groups in inosine concentrations, although a trend towards higher concentrations during reperfusion was noted in the rIPC group. In the area not at risk, inosine concentrations remained stable and there were no differences between groups. 3.4. Infarct size Infarct size relative to area at risk was 16.5% ± 3.6% in the control group and tended to be increased in the rIPC group (19.4% ± 1.7%, ns), was significantly increased in the GI group (24.5% ± 4.2%, p b 0.01) and reduced in the rICP + GI (11.5% ± 2.3%, p b 0.05) (Fig. 4). 4. Discussion This study is the first to examine the in-vivo effects of rIPC in the neonatal heart and also the first study the effect of rIPC on the citric acid cycle. Our results show that rIPC in neonatal piglets modifies neonatal cardiac function and metabolism and adversely affects responses to ischemia-reperfusion injury. Our data show that, even prior to myocardial ischemia, rIPC is associated with hemodynamic compromise as manifest by reduced dP/dtmax. The responses to ischemia-reperfusion were also different amongst the groups. GI treatment alone preserved systolic function but increased infarct size, and rIPC alone was associated with worse functional recovery and a trend towards increased infarct size, compared to controls. However, the combination of rIPC and GI preserved cardiac function and reduced infarct size while maintaining a favorable metabolic profile with low lactate levels. These findings
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
391
Table 2 Indices of cardiac function at selected time points. Values are mean ± standard deviation. P-values reflect comparison to control group calculated by two-way ANOVA. Pre-ischemia 0 min HR (min−1) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Systolic BP (mm Hg) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC dP/dtmax (mm Hg/s) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC dP/dtmin (mm Hg/s) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Tau Control Glucose–Insulin Glucose–Insulin + rIPC rIPC EDP Control Glucose–Insulin Glucose–Insulin + rIPC rIPC
121 141 133 139
± ± ± ±
Ischemia 20 min
37 48 29 26
40 min
Reperfusion 60 min
80 min
P 120 min
160 min
200 min
(vs. Control)
133 135 130 129
± ± ± ±
40 31 28 29
135 138 130 137
± ± ± ±
51 37 38 30
148 150 151 149
± ± ± ±
57 62 66 48
152 152 147 145
± ± ± ±
39 41 55 63
136 138 143 139
± ± ± ±
36 74 46 27
135 136 128 128
± ± ± ±
47 46 37 49
128 140 139 129
± ± ± ±
35 46 45 24
ns ns ns
62.3 ± 21.7 61.9 ± 21.9 64.2 ± 25.6 59.2.3 ± 18.2
61.9 64.3 60.1 58.2
± ± ± ±
23.8 24.5 26.1 20.1
60.1 64.4 57.9 56.5
± ± ± ±
27.6 26.0 23.7 28.1
48.3 50.0 50.3 45.3
± ± ± ±
21.7 33.9 18.2 22.7
47.2 49.9 53.8 46.5
± ± ± ±
22.9 29.9 27.6 25.9
51.1 57.3 60.3 55.8
± ± ± ±
26.4 17.6 20.9 23.1
55.3 60.5 60.4 53.1
± ± ± ±
20.2 24.2 19.1 22.1
58.5 60.1 62.1 55.1
± ± ± ±
24.2 23.8 24.9 22.1
ns ns ns
1365 1283 1389 1381
1361 1023 1561 1267
1289 ± 316 837 ± 128 1567 ± 186 1273 ± 235
889 ± 220 673 ± 311 1102 ± 189 1028 ± 361
559 256 870 837
± ± ± ±
274 234 308 401
956 ± 144 498 ± 178 1019 ± 231 1137 ± 280
938 ± 189 579 ± 121 1167 ± 233 1167 ± 183
936 ± 223 689 ± 206 1162 ± 229 1201 ± 260
p b 0.01 p b 0.01 p b 0.01
−439 −221 −691 −767
± ± ± ±
−818 −404 −899 −931
−777 ± 288 −532 ± 260 −967 ± 265 −1004 ± 178
−890 −582 −951 −958
p b 0.01 p b 0.01 p b 0.01
± ± ± ±
201 221 213 198
± ± ± ±
285 166 158 209
−1051 ± 172 −1169 ± 246 −1019 ± 220 −967 ± 144
−1033 ± 231 −901 ± 203 −1102 ± 305 −832 ± 260
−1006 ± 308 732 ± 164 −1031 ± 331 −800 ± 223
−861 −519 −978 −737
38.2 33.9 40.9 40.1
± ± ± ±
8.4 9.2 9.3 11.3
35.2 48.7 38.3 51.7
± ± ± ±
6.3 18.9 7.9 12.8
42.3 59.0 46.3 58.2
± ± ± ±
31.9 37.2 21.3 19.3
44.3 ± 68.3 ± 42.9 ± 63.3 ±
7.87 6.34 7.83 6.82
± ± ± ±
2.84 3.29 4.28 2.50
7.03 6.91 8.12 7.87
± ± ± ±
3.12 2.71 3.66 2.89
7.16 7.01 8.13 8.09
± ± ± ±
3.91 2.97 4.11 3.87
10.23 ± 4.28 9.78 ± 5.12 10.72 ± 5.59 11.72 ± 6.48
provide a potential explanation as to why achieving cardioprotection in the immature heart is difficult but also raise a series of new questions.
± ± ± ±
147 169 178 144
37.1 41.9 20.2 30.6
40.5 70.3 40.9 60.2
± ± ± ±
220 94 257 155
38.4 38.3 24.4 29.3
10.16 ± 3.87 9.76 ± 4.27 9.91 ± 3.92 10.72 ± 4.29
± ± ± ±
197 161 229 217
± ± ± ±
205 161 280 251
40.1 ± 66.3 ± 38.3 ± 52.3 ±
38.8 36.3 23.8 31.4
36.3 62.3 37.9 47.6
± ± ± ±
33.9 39.1 23.8 47.1
36.8 ± 63.4 ± 33.4 ± 41.3 ±
21.5 46.1 26.7 36.4
p b 0.01 ns p b 0.05
8.28 8.16 8.12 9.67
3.82 3.09 2.99 4.92
7.04 7.38 8.00 8.82
± ± ± ±
2.37 3.92 3.41 3.28
7.67 7.80 8.28 9.26
2.97 3.82 4.82 3.46
ns ns p b 0.05
± ± ± ±
± ± ± ±
GI + rIPC groups, the inosine increase is attenuated, again most likely due to increased substrate availability and preserved ATP production.
4.1. Metabolism 6.1. Cardiac function Given the complete lack of prior data, it would be naive to imagine that the present study would provide a complete mechanistic description of the detrimental metabolic effects of rIPC on the immature heart, however the abnormalities observed do provide clues as to their origins. 5. Hypoxia, glucose metabolism and glycogen turnover The neonatal myocardial tolerance towards hypoxia may cause or reflect a different effect of cardioprotective interventions. It has been proposed that ischemic conditioning in effect induces adaptation to hypoxia [16], which may not be relevant or possible in the neonatal heart. It is not known, however, whether the hypoxia tolerance in neonatal organs is a consequence of immature cell and organelle structure or of the carbohydrate dependent metabolism. Glycogen represents an important energy source in the immature myocardium. Glycogen levels are very high and modifiable by stress (such as functional ischemia/tachycardia) and substrate supplementation (glucose–insulin infusion)[17]. 6. Lactate and inosine Inosine is a terminal metabolite in the breakdown of ATP, and rising inosine concentrations are believed to reflect extreme ATP depletion occurring during hypoxia/ischemia. In the control group such a rise in inosine concentration occurs during ischemia – as expected – followed by a drop during reperfusion when ATP production is reestablished. However, in the rIPC group, the inosine concentrations increase further during reperfusion indicating a longer lasting impaired ATP-production, which is supported by increased lactate levels in the rIPC group. In the GI and
Generally, cardiac function recovered much better than seen in adult porcine hearts exposed to a similar IR insult [18]. Nonetheless, both systolic and diastolic functions were significantly impaired in the rIPC group and significantly better preserved in the GI and GI + rIPC groups compared to controls. While the adverse effect of rIPC alone on cardiac function may purely reflect metabolic impairment, improved function in the groups receiving insulin may be partially explained by the direct inotropic effect of insulin itself [19,20]. This is supported by the increase in dP/dtmax occurring as soon as GI infusion was initiated, prior to ischemia at a time when substrate supplementation would presumably be redundant. Conversely, the early decline in cardiac function in the rIPC group began after the second cycle of rIPC and also prior to ischemia. As discussed above, this hemodynamic response was unique to the rIPC group and indicates an adverse effect of rIPC in the “healthy” neonatal heart that is amplified by IR injury.
6.2. Infarct size It has previously been shown that the neonatal heart is more resistant to IR injury than the adult heart [11]. Our study supports this observation. The infarct size in the control animals (less than 20% of area at risk) was less than half that observed in our previous study of adult pigs subjected to an identical IR injury [21]. Nonetheless, while the absolute decrease in infarct size observed with the combination of rIPC + GI may seem minor, it still represents a relative reduction of around one quarter (from approximately 16% in controls to approximately 12%), and would almost certainly be clinically relevant if translated to human studies.
392
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
Fig. 3. Concentrations (mean +/− SEM) of inosine and lactate in the area at risk (left ventricle, ischemic territory) and area not at risk (right ventricle) during 40 min of pretreatment, 40 min of ischemia and 40 min of reperfusion; see Results section for further description and comparison of groups. Error bars: SEM. * indicates significant difference from control group by repeated measures two-way ANOVA.
6.3. Clinical perspectives Generally, translating cardioprotective strategies to clinical use has proven difficult [22,23], although rIPC has successfully been introduced in the treatment of acute [24] and chronic [25,26] heart disease in adults. The present study supports our recent finding of detrimental effects of rIPC in isolated neonatal rabbit hearts [14]. Indeed, our data have several important implications for future studies of rIPC in neonates and young infants. Firstly, our observation of impaired metabolic function
and myocardial performance mandates that any clinical study should include careful evaluation of the effects of the rIPC stimulus itself, prior to inclusion as a potential therapy in clinical trials. Secondly, careful consideration should be made for maintenance of substrate delivery, potentially with a GI infusion, during the rIPC stimulus and subsequent IR insult, to minimize any potential adverse effects, and harness potential benefit. Given that GI infusion alone significantly increased infarct size in our study, careful preclinical studies of the optimal balance between rIPC stimulus and degree of concomitant metabolic support will be required before the use of therapeutic rIPC could be supported in this patient population.
ns p
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Remote ischemic preconditioning with – but not without – metabolic support protects the neonatal porcine heart against ischemia-reperfusion injury Michael R. Schmidt a,⁎, Nicolaj B. Støttrup a, Hussain Contractor b, Janus A. Hyldebrandt c, Mogens Johannsen d, Christian M. Pedersen a, Rune Birkler d, Houman Ashrafian b, Keld E. Sørensen a, Rajesh K. Kharbanda b, Andrew N. Redington e, Hans E. Bøtker a a
Department of Cardiology, Aarhus University Hospital, Skejby, Brendstrupgaardsvej, Aarhus N, DK-8200, Denmark Department of Cardiology, The John Radcliffe, Headley Way, Headington, Oxford, OX3 9DU, United Kingdom c Department of Anesthesiology and Intensive Care, Aarhus University Hospital, Skejby, Brendstrupgaardsvej, Aarhus N, DK-8200, Denmark d Department of Forensic Medicine, Aarhus University Hospital, Skejby, Brendstrupgaardsvej, Aarhus N, DK-8200, Denmark e Division of Cardiology, Hospital for Sick Children, Toronto, M5G 1X8, Canada b
a r t i c l e
i n f o
Article history: Received 28 August 2013 Received in revised form 21 October 2013 Accepted 2 November 2013 Available online 12 November 2013 Keywords: Preconditioning Metabolism Ischemia-reperfusion injury Pediatric Insulin
a b s t r a c t Background: While remote ischemic preconditioning (rIPC) protects the mature heart against ischemiareperfusion (IR) injury, the effect on the neonatal heart is not known. The neonatal heart relies almost solely on carbohydrate metabolism, which is modified by rIPC in the mature heart. We hypothesized that rIPC combined with metabolic support with glucose-insulin (GI) infusion improves cardiac function and reduces infarct size after IR injury in neonatal piglets in-vivo. Methods and results: 32 newborn piglets were randomized into 4 groups: control, GI, GI + rIPC and rIPC. GI and GI + rIPC groups received GI infusion continuously from 40 min prior to ischemia. rIPC and GI + rIPC groups underwent four cycles of 5 min limb ischemia. Myocardial IR injury was induced by 40 min occlusion of the left anterior descending artery followed by 2 h reperfusion. Myocardial lactate concentrations were assessed in microdialysis samples analyzed by mass spectrometry. Infarct size was measured using triphenyltetrazolium chloride staining. Systolic recovery (dP/dtmax as % of baseline) after 2 h reperfusion was 68.5 ± 13.8% in control, 53.7 ± 11.2% in rIPC (p b 0.05), and improved in GI (83.6 ± 18.8%, p b 0.05) and GI + rIPC (87.0 ± 15.7%, p b 0.01). Conclusion: rIPC + GI protects the neonatal porcine heart against IR injury in-vivo. rIPC alone has detrimental metabolic and functional effects that are abrogated by simultaneous GI infusion. © 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Impaired ventricular systolic and diastolic function are important contributors to the mortality and morbidity associated with pediatric cardiac surgery [1,2], particularly after complex procedures in the neonate and young infant. The immature heart is structurally, functionally, and metabolically different from the adult heart and undergoes fundamental changes around the time of birth. For example, it is well known that cardiac energy generation shifts from glycolysis to oxidative lipid metabolism reflecting the increasing mechanical and metabolic demands that occur during the first few postnatal days and weeks, however our understanding of the detailed ontogeny of these changes remains incomplete. Understanding such physiological transitions in the immediately postnatal myocardium is a prerequisite to manipulating
⁎ Corresponding author. Tel.: +45 7845 6020; fax: +45 7845 6022. E-mail address: [email protected] (M.R. Schmidt). 0167-5273/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2013.11.020
the mechanisms of myocardial protection and to devise new strategies for protecting the immature heart during cardiac surgery. Preoperative remote ischemic preconditioning (rIPC), induced by cycles of ischemia and reperfusion of a limb, has recently been shown to prevent myocardial damage during surgery in older children [3] and adults [4]. In contrast, the impact of rIPC on the immature heart remains unknown. This is important, particularly as local ischemic preconditioning, where the cycles of ischemia and reperfusion are applied directly to the tissue subjected to prolonged ischemia, has suggested that the responses may display distinct maturational dependence. Some studies have suggested that the fetal heart is resistant to local ischemic preconditioning, and that its cardioprotective effect only becomes effective after 1–4 weeks (depending on species) after birth [5–7]. In an important recent study by Jones et al., no detectable cardioprotection was achieved by rIPC in hypoxic neonates undergoing cardiopulmonary bypass surgery [8], which was most recently supported by Pepe et al. showing that cardioprotective intracellular signaling mechanisms may already be activated in cyanosed children eligible for
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
surgical correction of tetralogy of Fallot and are not further enhanced by rIPC [9]. However, others have shown effective pharmacologic and ischemic preconditioning in normoxic immature rabbit hearts [10], and isolated neonatal cardiomyocytes [11,12]. Given these inconsistent observations, it is unsurprising that the mechanisms underlying ‘preconditioning resistance’ in the neonate remain unknown, although the immature myocardium appears intrinsically more resistant to hypoxia and possibly ischemia [6,13], raising the possibility that it is already ‘preconditioned’ in some way. Furthermore, the relevance of these studies to remote ischemic preconditioning (rIPC) is questionable, as rIPC does not involve ischemic bursts to the myocardium itself to induce the preconditioned state. However, in a recently published study, we demonstrated that also rIPC prior to ischemia-reperfusion injury may have detrimental effects in the isolated neonatal rabbit heart [14]. Importantly, most experimental studies of ischemic cardiac conditioning have been performed on isolated rodent hearts. While these models are relatively simple to establish and effectively allow for detailed mechanistic studies, they poorly reflect the complex physiology of ischemia-reperfusion injury in the intact animal or human. In this study, we used an in-vivo newborn piglet model to examine the effects of rIPC and metabolic support on function, infarct development and crude indicators of cardiac metabolism during and after regional myocardial ischemia. We used piglets less than 4 days old to study the effects of rIPC in hearts prior to the transition towards mature metabolism. We hypothesized that the adverse effects of IPC attributable to a relative reduction of carbohydrate metabolism could be counteracted by concurrent stimulation of the carbohydrate metabolism by glucose-insulin infusion.
2. Methods 2.1. Design Thirty-two newborn piglets (1–4 days old) were randomized into four groups: Control group (no pretreatment during a 40 min observation period prior to myocardial ischemia), pretreatment with rIPC (4 cycles of 5 min tourniquet occlusion of the lower left limb followed by 5 min reperfusion), pretreatment with glucose-insulin (GI, insulin 100 mU/ kg/h in 20% glucose continuously from 40 minutes prior to ischemia), and pretreatment with combined rIPC + GI (Fig. 1).
2.2. Animals The animals were handled according to the guidelines of the Danish Committee for Animal Research (Dyreforsøgstilsynet, Copenhagen, Denmark). The piglets stayed with their mother until the day of the study.
389
2.3. Experimental preparation and protocol Anesthesia was induced with midazolam (0.5 mg/kg intramuscularly) and azaperone (4 mg/kg intramuscularly) followed by etomidate (0.5 mg/kg intravenously). The piglets were intubated and ventilated. Anesthesia was maintained with an infusion of pentobarbital. Two 4 F sheaths were placed in the left jugular vein and carotid artery, respectively. A 3 F Mikro-Tip high fidelity pressure manometer (Millar Instruments, US) was placed through the arterial sheath into the left ventricle. After randomization, a midline thoracotomy and pericardiotomy were performed and CMA-20 Elite microdialysis probes (CMA/Harvard Apparatus, Sweden) inserted into the free wall of the left (area at risk) and the right (area not at risk) ventricle using an introducer cannula (0.45 mm × 12 mm) (and) fixed with tissue glue (LiquiBand, Devon, UK). Particular attention was paid to place the microdialysis catheters in the area at risk and not at risk, respectively. A suture was placed around the left anterior descending artery 2–3 mm from its origin and the ends inserted into a plastic tube to form a tourniquet. All piglets then underwent 40 min pretreatment according to group allocation, followed by 40 min of regional ischemia (induced by tightening the tourniquet) and 120 min of reperfusion (by releasing the tourniquet). Body temperature was maintained within the normal range (38.5–39.5 °C) using a warming blanket. Hydration was maintained with continuous intravenous infusion of a solution containing isotonic sodium chloride (4 ml/(kg h)) and additional potassium supplied when needed to maintain serum potassium above 4.0 mmol/l. Coronary occlusion was confirmed by epicardial cyanosis and decrease in blood pressure, while reperfusion was verified by epicardial hyperemia. Indices of left ventricular function were acquired continuously, sampled via high-fidelity analogue to digital hardware to dedicated data acquisition software (Notocord®, France) and stored on a computer for later analysis. Continuous indices of systolic (dP/dtmax) and diastolic (dP/dtmin, tau) function were calculated automatically by the software (Fig. 2). At the end of the protocol, the tourniquet around the left anterior descending artery was tightened again and 20 ml of 1% Evans blue solution injected into the left atrium to delineate the area-at-risk. After a few seconds allowing the myocardium to be perfused by the dye, piglets were sacrificed by excision of the heart.
2.4. Microdialysis and assessment of myocardial lactate and inosine levels Following implantation of the microdialysis probes (polyarylethersulphone membrane length 10 mm, molecular cut-off 20 kDa; CMA Elite 20, CMA/Harvard Apparatus, Sweden) a 40 min wash-out period was allowed for lactate and inosine to reach equilibrium in the perturbed tissue before the experimental protocol of 40 min of stabilization, 40 min of regional ischemia, and 120 min (during which microdialysis samples were collected for the first 40 min) of reperfusion was commenced. Perfusion fluid consisted of Krebs–Henseleit solution deoxygenated with 95% N2 and 5% CO2. Pump perfusion velocity (Univentor 801) was 1 μl/min. The sampling time for each fraction was 10 min, and samples were harvested by a customized fraction collector (Univentor 820 Microsampler, Univentor Limited, Zejtun, Malta). All samples were cooled to 8 °C during sampling and stored at −80 °C until time of analysis. Following separation in a Waters ACQUITYTM Ultra-performance liquid chromatography (UPLC) system (Waters Corp., Milford, MA, USA), lactate and inosine were quantitated using a Waters XevoTM triple quadruple tandem mass spectrometer (Waters Corp., Manchester, UK) with a Z-spray electrospray ionization source operating both in the positive and negative ion modes. All analytes were verified using two daughter ions when possible and deuterated internal standards for selected components.[15] The final interstitial concentrations were calculated by correction for relative recovery rate. To ensure sufficient relative recovery, a series of in vitro
Fig. 1. Study design. Four groups all undergoing 40 min of pretreatment, 40 min of ischemia and 2 h of reperfusion.
390
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
LV dP/dt max Control Glucose-insulin
2000
Glucose-insulin + rIPC rIPC
1500
mm Hg/s
Table 1 Group statistics on weigh, sex and age.
1000
500
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190
0
Weight (kg) (mean ± SD) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Male sex (n, %) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Age (days) (median[min;max]) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC
1.21 1.30 1.18 1.26
± ± ± ±
0.41 0.51 0.28 0.40
3 (37.5%) 4 (50%) 5 (62.5%) 3 (37.5%) 2[1,4] 2[1,4] 2[1,4] 2[1,4]
min Fig. 2. Maximal derived pressure (dP/dtmax) average of 10 s recording in 10-min intervals; see Results section for further description and comparison of groups. Error bars: SEM.
experiments were conducted as previously published.[15] Specific recovery rate of lactate was 40% and of inosine 87%. Baseline values and mean concentrations from stabilization (pre-ischemia), regional ischemia, and reperfusion were used for statistical comparison (see below). 2.5. Assessment of myocardial infarction The explanted hearts were frozen immediately at −80 °C for 20 min, the right ventricular free wall removed and the left ventricle subsequently sliced into 2–3 mm transverse sections and incubated in 1.25% 2,3,5-triphenyltetrazoliumchloride dissolved in phosphate buffer (Sigma, St. Louis, Mo, USA) at 37 °C for 5 min at pH 7.4. Subsequently, the hearts were stored overnight in 10 ml of Lillie's solution (4% formaldehyde buffer, VWR International, Denmark). Next day, each heart was weighed and scanned with a flatbed scanner (HP ScanJet 4300C, Hewlett Packard, Palo Alto, CA, USA). The AAR and size of infarction (IS) were assessed by computer planimetry (Photoshop CS3, Adobe systems incorporated, San Jose, California). Infarct size is expressed as a ratio, IS of AAR, weighted with the weight of each individual slice. All analyses were performed by a blinded observer. 2.6. Statistics Hemodynamic measurements and metabolomic results were compared between groups using repeated measures two-way ANOVA with Bonferroni correction comparing intervention groups to controls. As occasional lactate and inosine samples were missing, mean concentrations from pre-ischemia, regional ischemia, and reperfusion were used for the ANOVA analysis. Data satisfied criteria for normal distribution and there were no significant outliers. Infarct size was compared between groups using one-way ANOVA.
3. Results A total of 48 newborn piglets were used. Sixteen died due to technical or surgical problems and were excluded from the study, of which ten died prior to randomization and 6 during protocol (1 control, 2 GI, 1 GI + rIPC, 2 rIPC). There was no difference between groups regarding baseline characteristics (weight, sex, age) (Table 1). 3.1. Baseline characteristics Indices of cardiac function did not differ between groups at baseline with regards to left ventricular systolic and diastolic pressure, heart rate (HR), systolic blood pressure, minimal and maximal dP/dt (dP/dtmin and dP/dtmax), end diastolic pressure (EDP) or tau (Table 2). 3.2. Left ventricular function There was no difference between groups regarding left ventricular pressure and HR during the protocol (Table 2). All animals showed a significant drop in dP/dtmax to 41% ± 10% (controls), 20% ± 6% (rIPC), 62% ± 11% (GI) and 60% ± 11% (GI + rIPC) of baseline (p b 0.01 in
all groups) during ischemia and significant post-ischemic recovery to 68.5% ± 13.8% controls, p b 0.05), 53.7% ± 11.2% (rIPC, p b 0.05), 83.6% ± 18.8% (GI, p b 0.01) and 87.0% ± 15.7% (GI + rIPC, p b 0.01), of baseline, compared to late ischemia, (Fig. 3). However, in the rIPC group, dP/dtmax was significantly lower than in controls prior to ischemia (p b 0.05), throughout ischemia (p b 0.01) and during reperfusion (p b 0.01). Similarly, all groups showed significant impairment of diastolic function (i.e. increased dP/dtmin, tau and EDP) (Table 2) during ischemia and significant recovery during reperfusion, but the rIPC group deteriorated more during ischemia (p b 0.05) and recovered less during reperfusion (p b 0.01) compared to controls. 3.3. Lactate and inosine In the area at risk, an expected increase in interstitial lactate occurred during ischemia and reperfusion. However, only the rIPC group was significantly different from the controls (Fig. 3a). In the area not at risk, lactate levels remained stable in control, GI and GI + rIPC groups but increased in the rIPC group (Fig. 3b). In the area at risk, there were no statistically significant differences between groups in inosine concentrations, although a trend towards higher concentrations during reperfusion was noted in the rIPC group. In the area not at risk, inosine concentrations remained stable and there were no differences between groups. 3.4. Infarct size Infarct size relative to area at risk was 16.5% ± 3.6% in the control group and tended to be increased in the rIPC group (19.4% ± 1.7%, ns), was significantly increased in the GI group (24.5% ± 4.2%, p b 0.01) and reduced in the rICP + GI (11.5% ± 2.3%, p b 0.05) (Fig. 4). 4. Discussion This study is the first to examine the in-vivo effects of rIPC in the neonatal heart and also the first study the effect of rIPC on the citric acid cycle. Our results show that rIPC in neonatal piglets modifies neonatal cardiac function and metabolism and adversely affects responses to ischemia-reperfusion injury. Our data show that, even prior to myocardial ischemia, rIPC is associated with hemodynamic compromise as manifest by reduced dP/dtmax. The responses to ischemia-reperfusion were also different amongst the groups. GI treatment alone preserved systolic function but increased infarct size, and rIPC alone was associated with worse functional recovery and a trend towards increased infarct size, compared to controls. However, the combination of rIPC and GI preserved cardiac function and reduced infarct size while maintaining a favorable metabolic profile with low lactate levels. These findings
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
391
Table 2 Indices of cardiac function at selected time points. Values are mean ± standard deviation. P-values reflect comparison to control group calculated by two-way ANOVA. Pre-ischemia 0 min HR (min−1) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Systolic BP (mm Hg) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC dP/dtmax (mm Hg/s) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC dP/dtmin (mm Hg/s) Control Glucose–Insulin Glucose–Insulin + rIPC rIPC Tau Control Glucose–Insulin Glucose–Insulin + rIPC rIPC EDP Control Glucose–Insulin Glucose–Insulin + rIPC rIPC
121 141 133 139
± ± ± ±
Ischemia 20 min
37 48 29 26
40 min
Reperfusion 60 min
80 min
P 120 min
160 min
200 min
(vs. Control)
133 135 130 129
± ± ± ±
40 31 28 29
135 138 130 137
± ± ± ±
51 37 38 30
148 150 151 149
± ± ± ±
57 62 66 48
152 152 147 145
± ± ± ±
39 41 55 63
136 138 143 139
± ± ± ±
36 74 46 27
135 136 128 128
± ± ± ±
47 46 37 49
128 140 139 129
± ± ± ±
35 46 45 24
ns ns ns
62.3 ± 21.7 61.9 ± 21.9 64.2 ± 25.6 59.2.3 ± 18.2
61.9 64.3 60.1 58.2
± ± ± ±
23.8 24.5 26.1 20.1
60.1 64.4 57.9 56.5
± ± ± ±
27.6 26.0 23.7 28.1
48.3 50.0 50.3 45.3
± ± ± ±
21.7 33.9 18.2 22.7
47.2 49.9 53.8 46.5
± ± ± ±
22.9 29.9 27.6 25.9
51.1 57.3 60.3 55.8
± ± ± ±
26.4 17.6 20.9 23.1
55.3 60.5 60.4 53.1
± ± ± ±
20.2 24.2 19.1 22.1
58.5 60.1 62.1 55.1
± ± ± ±
24.2 23.8 24.9 22.1
ns ns ns
1365 1283 1389 1381
1361 1023 1561 1267
1289 ± 316 837 ± 128 1567 ± 186 1273 ± 235
889 ± 220 673 ± 311 1102 ± 189 1028 ± 361
559 256 870 837
± ± ± ±
274 234 308 401
956 ± 144 498 ± 178 1019 ± 231 1137 ± 280
938 ± 189 579 ± 121 1167 ± 233 1167 ± 183
936 ± 223 689 ± 206 1162 ± 229 1201 ± 260
p b 0.01 p b 0.01 p b 0.01
−439 −221 −691 −767
± ± ± ±
−818 −404 −899 −931
−777 ± 288 −532 ± 260 −967 ± 265 −1004 ± 178
−890 −582 −951 −958
p b 0.01 p b 0.01 p b 0.01
± ± ± ±
201 221 213 198
± ± ± ±
285 166 158 209
−1051 ± 172 −1169 ± 246 −1019 ± 220 −967 ± 144
−1033 ± 231 −901 ± 203 −1102 ± 305 −832 ± 260
−1006 ± 308 732 ± 164 −1031 ± 331 −800 ± 223
−861 −519 −978 −737
38.2 33.9 40.9 40.1
± ± ± ±
8.4 9.2 9.3 11.3
35.2 48.7 38.3 51.7
± ± ± ±
6.3 18.9 7.9 12.8
42.3 59.0 46.3 58.2
± ± ± ±
31.9 37.2 21.3 19.3
44.3 ± 68.3 ± 42.9 ± 63.3 ±
7.87 6.34 7.83 6.82
± ± ± ±
2.84 3.29 4.28 2.50
7.03 6.91 8.12 7.87
± ± ± ±
3.12 2.71 3.66 2.89
7.16 7.01 8.13 8.09
± ± ± ±
3.91 2.97 4.11 3.87
10.23 ± 4.28 9.78 ± 5.12 10.72 ± 5.59 11.72 ± 6.48
provide a potential explanation as to why achieving cardioprotection in the immature heart is difficult but also raise a series of new questions.
± ± ± ±
147 169 178 144
37.1 41.9 20.2 30.6
40.5 70.3 40.9 60.2
± ± ± ±
220 94 257 155
38.4 38.3 24.4 29.3
10.16 ± 3.87 9.76 ± 4.27 9.91 ± 3.92 10.72 ± 4.29
± ± ± ±
197 161 229 217
± ± ± ±
205 161 280 251
40.1 ± 66.3 ± 38.3 ± 52.3 ±
38.8 36.3 23.8 31.4
36.3 62.3 37.9 47.6
± ± ± ±
33.9 39.1 23.8 47.1
36.8 ± 63.4 ± 33.4 ± 41.3 ±
21.5 46.1 26.7 36.4
p b 0.01 ns p b 0.05
8.28 8.16 8.12 9.67
3.82 3.09 2.99 4.92
7.04 7.38 8.00 8.82
± ± ± ±
2.37 3.92 3.41 3.28
7.67 7.80 8.28 9.26
2.97 3.82 4.82 3.46
ns ns p b 0.05
± ± ± ±
± ± ± ±
GI + rIPC groups, the inosine increase is attenuated, again most likely due to increased substrate availability and preserved ATP production.
4.1. Metabolism 6.1. Cardiac function Given the complete lack of prior data, it would be naive to imagine that the present study would provide a complete mechanistic description of the detrimental metabolic effects of rIPC on the immature heart, however the abnormalities observed do provide clues as to their origins. 5. Hypoxia, glucose metabolism and glycogen turnover The neonatal myocardial tolerance towards hypoxia may cause or reflect a different effect of cardioprotective interventions. It has been proposed that ischemic conditioning in effect induces adaptation to hypoxia [16], which may not be relevant or possible in the neonatal heart. It is not known, however, whether the hypoxia tolerance in neonatal organs is a consequence of immature cell and organelle structure or of the carbohydrate dependent metabolism. Glycogen represents an important energy source in the immature myocardium. Glycogen levels are very high and modifiable by stress (such as functional ischemia/tachycardia) and substrate supplementation (glucose–insulin infusion)[17]. 6. Lactate and inosine Inosine is a terminal metabolite in the breakdown of ATP, and rising inosine concentrations are believed to reflect extreme ATP depletion occurring during hypoxia/ischemia. In the control group such a rise in inosine concentration occurs during ischemia – as expected – followed by a drop during reperfusion when ATP production is reestablished. However, in the rIPC group, the inosine concentrations increase further during reperfusion indicating a longer lasting impaired ATP-production, which is supported by increased lactate levels in the rIPC group. In the GI and
Generally, cardiac function recovered much better than seen in adult porcine hearts exposed to a similar IR insult [18]. Nonetheless, both systolic and diastolic functions were significantly impaired in the rIPC group and significantly better preserved in the GI and GI + rIPC groups compared to controls. While the adverse effect of rIPC alone on cardiac function may purely reflect metabolic impairment, improved function in the groups receiving insulin may be partially explained by the direct inotropic effect of insulin itself [19,20]. This is supported by the increase in dP/dtmax occurring as soon as GI infusion was initiated, prior to ischemia at a time when substrate supplementation would presumably be redundant. Conversely, the early decline in cardiac function in the rIPC group began after the second cycle of rIPC and also prior to ischemia. As discussed above, this hemodynamic response was unique to the rIPC group and indicates an adverse effect of rIPC in the “healthy” neonatal heart that is amplified by IR injury.
6.2. Infarct size It has previously been shown that the neonatal heart is more resistant to IR injury than the adult heart [11]. Our study supports this observation. The infarct size in the control animals (less than 20% of area at risk) was less than half that observed in our previous study of adult pigs subjected to an identical IR injury [21]. Nonetheless, while the absolute decrease in infarct size observed with the combination of rIPC + GI may seem minor, it still represents a relative reduction of around one quarter (from approximately 16% in controls to approximately 12%), and would almost certainly be clinically relevant if translated to human studies.
392
M.R. Schmidt et al. / International Journal of Cardiology 170 (2014) 388–393
Fig. 3. Concentrations (mean +/− SEM) of inosine and lactate in the area at risk (left ventricle, ischemic territory) and area not at risk (right ventricle) during 40 min of pretreatment, 40 min of ischemia and 40 min of reperfusion; see Results section for further description and comparison of groups. Error bars: SEM. * indicates significant difference from control group by repeated measures two-way ANOVA.
6.3. Clinical perspectives Generally, translating cardioprotective strategies to clinical use has proven difficult [22,23], although rIPC has successfully been introduced in the treatment of acute [24] and chronic [25,26] heart disease in adults. The present study supports our recent finding of detrimental effects of rIPC in isolated neonatal rabbit hearts [14]. Indeed, our data have several important implications for future studies of rIPC in neonates and young infants. Firstly, our observation of impaired metabolic function
and myocardial performance mandates that any clinical study should include careful evaluation of the effects of the rIPC stimulus itself, prior to inclusion as a potential therapy in clinical trials. Secondly, careful consideration should be made for maintenance of substrate delivery, potentially with a GI infusion, during the rIPC stimulus and subsequent IR insult, to minimize any potential adverse effects, and harness potential benefit. Given that GI infusion alone significantly increased infarct size in our study, careful preclinical studies of the optimal balance between rIPC stimulus and degree of concomitant metabolic support will be required before the use of therapeutic rIPC could be supported in this patient population.
ns p
