| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1997;64:163-170
© 1997 The Society of Thoracic Surgeons
Cardiothoracic Unit and Department of Surgery, Hammersmith Hospital, Royal Postgraduate Medical School, London, United Kingdom
Accepted for publication January 25, 1997.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
Methods. Twenty-four patients were prospectively randomized into one of four CPB groups: nonpulsatile hypothermic (NP 28); pulsatile hypothermic (P 28); nonpulsatile normothermic (NP 37); and pulsatile normothermic (P 37). Gastric wall blood flow was assessed using laser Doppler flow measurement and gastric mucosal oxygenation (intramucosal pH), using tonometry.
Results. After 10 minutes of CPB, the NP 28 group had the greatest reduction in gastric wall blood flow (-60.6% ± 3.8%) compared with baseline (p < 0.05). Thirty minutes into CPB, the P 37 group had less gastric mucosal hypoperfusion (-9.7% ± 10.3%) than the NP 28 patients (-53.0% ± 8.6%; p < 0.05). All groups showed a hyperemic response immediately after CPB. No significant differences between the four groups were found for gastric mucosal oxygenation during or after CPB. A progressive decline occurred in this variable during the period 3 to 4 hours after CPB. At this time, total-body oxygen consumption and extraction were at their maximum.
Conclusions. This study found that perfusion protocol can influence mucosal blood flow, but other overriding factors that operate during and after CPB act to cause mucosal hypoxia. These findings, particularly the timing of mucosal hypoxia, may have implications for centers contemplating early extubation or "fast tracking" of patients after CPB.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Gastrointestinal (GI) integrity is now recognized as one of the key factors determining a successful outcome after a cardiac surgical procedure. The lining of the GI tract can be used as a window to monitor the adequacy of tissue oxygenation generally, because at times of hemodynamic stress, it is the first tissue to become compromised [1]. Noninvasive monitoring of gastric and colonic mucosal oxygenation has been shown to be useful as a predictor of outcome in many clinical settings including cardiac surgery [2, 3]. Because the duration of mucosal hypoxia has been found to correlate with outcome after cardiopulmonary bypass (CPB), factors that may prevent mucosal acidosis can be expected to influence postoperative morbidity and mortality [3]. Some variables that influence gastric mucosal oxygenation (gastric intramucosal pH [pHi]) in cardiac patients have already been reported and include volume status and anesthesia [4, 5]. Pharmacologic attempts to improve gastric perfusion during CPB using splanchnic vasodilators have not been successful [6–8]. Dopexamine, a putative splanchnic vasodilator, was found to worsen gastric mucosal hypoxia, a result that may be due to maldistribution of blood flow within the splanchnic bed [6]. The aim of this study was to assess the effects of CPB perfusion protocols on the adequacy of gastric mucosal oxygenation or pHi, its relation to whole-body oxygen utilization, and the correlation of gastric pHi to mucosal blood flow and systemic acid-base indices.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
The study comprised 24 patients. They were prospectively randomized into one of four groups: nonpulsatile hypothermic CPB (NP 28); pulsatile hypothermic CPB (P 28); nonpulsatile normothermic CPB (NP 37); and pulsatile normothermic CPB (P 37).
Anesthesia Protocol
Premedication was standardized to lorazepam, scopolamine hydrabromide, papaverine hydrochloride, and 300 mg of ranitidine hydrochloride the night before operation. Anesthesia was induced with a combination of methohexital sodium, fentanyl, midazolam hydrochloride, and a mixture of oxygen, isoflurane, and air. Muscle relaxation was provided with pancuronium bromide or vecuronium bromide (depending on preoperate ß-blockade). Anesthesia was subsequently maintained using a mixture of air, oxygen, isoflurane, and fentanyl. In the anesthesia room, a Swan-Ganz catheter was sited, together with a tonometer (Tonometrics Division, Instrumentation Corp, Helsinki, Finland) using a technique previously described [9].
Cardiopulmonary Bypass Protocol
Cardiopulmonary bypass was instituted using two-stage venous and aortic cannulation. Bypass was standardized as follows:
Regulation of Mean Arterial Blood Pressure
Mean arterial blood pressure was adjusted if it was greater than 70 mm Hg for 5 minutes or greater than 80 mm Hg at any time using the following protocols: during CPB, MAP was reduced using either glyceryl trinitrate infusion or boluses of midazolam; after CPB during anesthesia, it was reduced with either fentanyl or isoflurane; and immediately after CPB and in the intensive care unit, mean arterial pressure was reduced using glyceryl trinitrate or propofol infusions. The mean arterial pressure was increased if it fell to less than 50 mm Hg for more than 5 minutes or was less than 35 mm Hg at any time using the following protocols: during the anesthesia period before and after CPB, it was adjusted with (1) volume loading after assessment of pulmonary artery wedge pressure with a Swan-Ganz catheter (first unit, Haemacel, and thereafter human albumin solution 4.5%) and (2) dopamine hydrochloride; and during CPB, it was augmented using the 
agonist methoxamine hydrochloride.
Tonometry and Gastric Laser Doppler Flow Measurements
Gastric wall blood flow was measured by laser Doppler flow (LDF) measurements using an MBF3 meter (Moor Instruments Ltd, Axminster, Devon, UK) in combination with a probe design described previously [10]. The principle of measuring LDF is that an electromagnetic wave reflected from a moving object will undergo a shift in frequency relative to a stationary observer. The shift in frequency is proportional to the velocity of the moving object. The MBF3 meter uses a solid-state laser diode to generate laser light at a wavelength of 780 nm with maximum accessible power of 2 mW.
A 2-m interconnecting cable with separate fiberoptic pathways for incident and reflected light is attached proximally to the meter. At its distal end, it is attached to a second cable (the probe cable), which is 2 m in length and has only one optical pathway terminating distally in a mucosal probe. A viscous optical coupling solution (refractive index matched) is applied at the interface to ensure efficient optical transmission. The probe cable is rotated axially to achieve maximum signal intensity before the screw-lock assembly is tightened. The reflected signal is photo-detected, amplified and processed by an analog processor. This allows the determination of mean red blood cell flux, measured as perfusion units. When the flux signal is displayed against time, it broadly resembles an electrocardiographic trace of ventricular fibrillation. The positive offset corresponds to mean perfusion, whereas the oscillations are believed to be caused by local blood flow changes resulting from factors such as vasomotion.
Prior to clinical use, the system is calibrated by immersing the distal tip of the probe in a reference solution consisting of a 0.5% wt/vol suspension of polystyrene microspheres with a mean diameter of 0.48 µm (Seradyn Inc, Indianapolis, IN). The Brownian motion–dependent movement of particles provides a calibration reference. However, the units of flux are arbitrary and cannot be related to tissue perfusion in milliliters per minute. After calibration of the laser Doppler meter, the laser Doppler probe was attached to the tip of a primed Trip gastric tonometer (Tonometrics Inc, Hopkinton, MA) using Micropore surgical tape (3M Medical Division, St. Paul, MN).
After induction of general anesthesia, the tonometer/laser Doppler assembly was passed orally and sited in the stomach. Position was confirmed by instilling air into one of the channels of the tonometer and auscultating over the epigastrium and by aspirating stomach contents. Gastric air and fluid was aspirated to ensure good mucosa/probe coaptation. Once the patient was positioned on the operating table, the probe cable was connected to the interconnecting cable through the screw-lock mechanism. This system was found to provide greatest signal stability for sequential blood-flow measurements and facilitated the determination of simultaneous blood flow and mucosal oxygenation from adjacent regions of the stomach. Steady-state baseline measurements 10 minutes before the onset of CPB were recorded, and subsequent measurements were expressed as a percentage of this value. Laser Doppler flow readings were recorded both on the hard disc of a computer and as a hard copy printout. Laser Doppler flow values were then averaged over 2-minute intervals for the duration of the study.
The gastric tonometer was primed using the technique previously described [10]. A 20- to 30-minute period of equilibrium was employed, and pHi values were calculated [3]. All arterial blood gas measurements were determined using the same blood gas analyzer (ABL4; Radiometer, Copenhagen, Denmark).
Swan-Ganz Measurements
Standard thermodilution methodology was used to determine the cardiac output (CI) before and after CPB. At any given time point, three readings were made, and the mean value was calculated. Blood was simultaneously sampled from the arterial line and the Swan-Ganz pulmonary artery port (mixed venous blood) to determine total-body oxygen consumption, delivery, and extraction using the following formulas:
 | (1) |
 | (2) |
 | (3) |
 | (4) |
Statistical Analyses
Differences between groups were first analyzed by analysis of variance using the Kruskal-Wallis test; subsequent differences between any two groups were then evaluated using the Mann-Whitney U test. For comparison within the same group, the Wilcoxon signed-rank test was employed. A two-tailed p value of less than 0.05 was accepted as denoting a significant difference. Spearman's rank-order correlation analysis was used to evaluate the relationship between any two variables. Values are shown as the mean ± the standard error.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
for the four groups. (For clarity of illustration, the standard errors have been omitted in this and subsequent figures).
|
).
Total-Body Oxygen Utilization
TOTAL-BODY OXYGEN CONSUMPTION.
The changes in 
O2 are plotted in Figure 2A. There were no significant differences between the groups before or after CPB. Patients who received hypothermic CPB (28°C), NP 28 and P 28 groups, had a reduction in O2 after 20 minutes of CPB to 38.6 ± 5.1 mL•min-1•m-2 and 35.8 ± 7.1 mL•min-1•m-2, respectively; O2 values were significantly lower than corresponding normothermic group values (p < 0.05). After rewarming, a significant differ ence was found between the NP 28 and NP 37 groups for a period of 2 hours after CPB; thereafter, a significant difference was detected between the P 28 and P 37 groups 5 hours after the end of CPB.
|
O2) are plotted in Figure 2B. Unlike the differences observed for O2, there were no differences between the groups for O2. The onset of CPB was associated with a reduction in O2 in all four groups (p < 0.05), which subsequently increased in the period after CPB. Thereafter, the values remained constant but lower than baseline.
TOTAL-BODY OXYGEN EXTRACTION.
The modulations in total-body oxygen extraction are plotted in Figure 2C. The changes in O2 are reflected in the percent extraction fraction. Hypothermic CPB resulted in a reduction in the extraction of oxygen by the body in both of these groups, whereas normothermic patients had a progressive increase in oxygen extraction fraction. As an example, 20 minutes after the onset of CPB, the extraction fraction was 13.1% ± 1.9% and 31.3% ± 1.7% for the P 28 and P 37 groups, respectively (p < 0.05). In the period immediately after CPB (up to 60 minutes after CPB), there was an initial plateau in the oxygen extraction followed by further increases. A zenith in oxygen extraction occurred in all four groups between 3 and 5 hours after the end of CPB. At this stage, the oxygen extraction fraction for all four groups had approximately doubled compared with baseline (p < 0.05).
MIXED VENOUS OXGEN SATURATION.
The changes in mixed venous oxygen saturation together with the mean total bypass time and the number of grafts in each of the four groups are provided in Table 1. There were no significant differences for the total CPB time or the number of grafts. Differences in oxygen saturation mirrored the changes in O2 and oxygen extraction. Under conditions of nonpulsatile flow, mixed venous oxygen saturation values were significantly higher in the hypothermic group than in the normothermic group 20 minutes into CPB and 50 minutes, after its cessation (p < 0.05). Under pulsatile flow conditions, saturation values were higher during CPB and 20 minutes and 2 hours after its cessation in the hypothermic group (p < 0.05). Thereafter, no significant differences were found between the four groups, but desaturation to levels significantly lower than baseline persisted to the end of the study period.
|
. During CPB, a significant intergroup difference was observed at only one time point (20 minutes after the onset of CPB) between the P 37 and the P 28 groups, which had pHi values of 7.37 ± 0.03 and 7.49 ± 0.03, respectively (p < 0.05). All four groups, regardless of CPB protocol, had a progressive reduction in gastric pHi in the period after CPB. Two hours after the end of CPB, all groups had a pHi significantly lower than baseline (p < 0.05). A nadir for the gastric pHi was reached between 3 and 5 hours after CPB. Four hours after CPB, all four groups had a mean pHi less than or equal to 7.35, indicative of mucosal hypoxia.
|
; for clarity of presentation, only data of the NP 28 group are plotted. This group was chosen because it represents an established regimen. The graph demonstrates that reciprocal changes in gastric pHi and blood flow occur during the rewarming phase of CPB and in the immediate period after CPB.
|
O2 with the development of gastric mucosal acidosis at increasingly higher levels of O2 (Fig 5A).
|
).
GASTRIC MUCOSAL OXYGENATION AND ARTERIAL OR VENOUS BLOOD PH.
No significant correlation was present between arterial or venous blood pH and gastric pHi (Fig 5C).
GASTRIC MUCOSAL OXYGENATION AND ARTERIAL BICARBONATE CONCENTRATION.
A significant positive correlation was found between gastric pHi and arterial bicarbonate concentration (r = 0.50) (Fig 5D).
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
However, the evaluation of perfusion alone in the context of CPB is misleading without considering tissue oxygenation. Total-body oxygen utilization was very similar for all four groups, ie, there was a progressive increase in O2 regardless of the CPB protocol. This rise in O2 was interrupted in the hypothermic patients by core cooling during the early phase of CPB. No notable difference was observed between the hypothermic or normothermic groups in the period after CPB for the oxygen extraction fraction. The O2 fell with the onset of CPB because of reductions in hematocrit, not alterations in the CI, which for many patients was increased by the institution of CPB. In the period after CPB, this reduction in O2 may limit adequate oxygenation of tissues in vital organ beds in the face of increased oxygen demands.
Without establishing the effect on oxygen utilization in different organ beds, there has nevertheless been a drift toward warm heart surgery on the basis that patients can be extubated faster after normothermic CPB. The results of total-body oxygen utilization from this study suggest that both hypothermic and normothermic patients behave in a very similar fashion. Demands by the body on oxygen delivery reach a peak between 3 and 5 hours after CPB regardless of the CPB protocol. Although the oxygen utilization after CPB is very similar, because of the immediate and progressive increase in oxygen consumption in normothermic patients during CPB, the margin for development of ischemic injury is less than it is for hypothermic patients. The effect of enhancing O2 by increasing the CI for normothermic patients from 2.4 to the physiologic level of 3.2 L•min-1•m-2 warrants further investigation.
This study has confirmed our previous finding that significant mucosal hypoperfusion occurs during CPB [10], but the development of mucosal hypoxia as demonstrated by many studies is not a simple consequence of mucosal hypoperfusion during CPB [3, 9, 15], especially as for at least 1 hour after CPB, hyperemia was found in all patient groups. The development of mucosal hypoxia appears to coincide with excessive demands by the body on oxygen supply. As mucosal hypoxia was accompanied by an increase in local blood flow in the period after CPB, it is tempting to speculate that this may be due to inappropriate blood flow distribution to the submucosa/ muscularis at the expense of the mucosa. The zeniths of O2 and total-body oxygen extraction and the development of gastric mucosal hypoxia all coincide between 3 and 5 hours after the end of CPB. This appears to be a critical time for the patient in terms of oxygen utilization. The mucosa is particularly susceptible to mucosal hypoxia because of the unique arrangement of the villous microvasculature, which results in the countercurrent exchange of oxygen [16].
The correlations undertaken for gastric pHi suggest that as O2 increases, the gastric mucosa becomes progressively more susceptible to hypoxia. The mixed venous oxygen saturation also appears to be a good indicator of the development of mucosal hypoxia, again because it reflects the balance between O2 and O2. This may be misleading, however, as mixed venous sampling cannot indicate villus tip oxygenation because of variable degrees of arteriovenous shunting [17]. This association may be unique to the patient who has just undergone CPB and consequently has mixed venous desaturation because of the increased total-body demand, which also coincides with the increased splanchnic oxygen demand.
As has been previously documented, blood gas analysis (either arterial or venous) cannot be used to predict mucosal hypoxia (see Fig 5C), as the blood pHi is a reflection of systemic acid-base status and does not indicate regional changes in tissue oxygenation [17]. The arterial bicarbonate was more closely related to mucosal hypoxia. Blood bicarbonate concentration falls as more is used to buffer acid production by anaerobic metabolism. Tissue anaerobic metabolism ensues when oxygen demand exceeds supply required for aerobic adenosine triphosphate production. Previous studies [18] have documented that the pHi indicates mucosal hypoxia well before alterations in systemic indices such as arterial blood gases are found. Tonometry remains a useful noninvasive clinical method for evaluating qualitative changes in GI mucosal oxygenation.
This study has demonstrated the importance of measuring local perfusion and oxygenation simultaneously. Although the two variables are interrelated, they can clearly change in opposite directions (see Fig 4) depending on the metabolic demands of the cell. Regional blood flow is determined by the CI, the vasomotor tone of local resistance vessels, and the patency of the microvasculature. The adequacy of cellular oxygenation is determined by the availability of oxygen, a function of tissue perfusion, and also by the ability of the cells to take up oxygen and by the viability of the normal mitochondrial oxygen handling. After CPB, metabolic demands are likely to be high as a result of the oxygen debt that may have been incurred with hypoperfusion during CPB; this will increase oxygen demands by the membrane ion pumps to restore intracellular equilibrium. The profound activation of blood components and release of cytokines and stress hormones can also be expected to increase the metabolic rate [19].
At times of increased oxygen demands by the body as a whole, the GI tract appears to be particularly susceptible to the development of cellular hypoxia. This mucosal hypoxia is well tolerated in the majority of patients, as the postoperative course of all our patients was normal with no mortality or major morbidity. However, all the patients selected for this study lacked major comorbidities, which may increase the risk of GI injury during CPB. There is clearly a need to undertake controlled studies to evaluate the benefits of normothermia as regards the suitability of patients for early extubation after CPB. The criteria used for early extubation may also need to include an assessment of splanchnic oxygenation as well as routine scrutiny of respiratory status, body temperature, and arterial blood gases.
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Hultstrom, L. Jansson, B. Bodin, and O. Kallskog Moderate hypothermia induces a preferential increase in pancreatic islet blood flow in anesthetized rats Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2007; 293(3): R1438 - R1443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-P. Braun, S. M Jakob, T. Volk, U. R Doepfmer, M. Moshirzadeh, S. Stegmann, P. M Dohmen, and C. Spies Arterio-venous gradients of free energy change for assessment of systemic and splanchnic perfusion in cardiac surgery patients Perfusion, November 1, 2006; 21(6): 353 - 360. [Abstract] [PDF]
|
|
|
|
|
|
S. K Ohri and T. Velissaris Gastrointestinal Dysfunction Following Cardiac Surgery Perfusion, July 1, 2006; 21(4): 215 - 223. [PDF]
|
|
|
|
 |
|
 A. Nygren, A. Thoren, E. Houltz, and S.-E. Ricksten Autoregulation of human jejunal mucosal perfusion during cardiopulmonary bypass. Anesth. Analg., June 1, 2006; 102(6): 1617 - 1622. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Velissaris, A. Tang, M. Murray, A. El-Minshawy, D. Hett, and S. Ohri A prospective randomized study to evaluate splanchnic hypoxia during beating-heart and conventional coronary revascularization Eur. J. Cardiothorac. Surg., June 1, 2003; 23(6): 917 - 924. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Jakob Splanchnic Blood Flow in Low-Flow States Anesth. Analg., April 1, 2003; 96(4): 1129 - 1138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Thoren, M. Elam, and S.-E. Ricksten Jejunal Mucosal Perfusion Is Well Maintained During Mild Hypothermic Cardiopulmonary Bypass in Humans Anesth. Analg., January 1, 2001; 92(1): 5 - 11. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Lebuffe, C. Decoene, A. Pol, A. Prat, and B. Vallet Regional Capnometry with Air-Automated Tonometry Detects Circulatory Failure Earlier Than Conventional Hemodynamics After Cardiac Surgery Anesth. Analg., November 1, 1999; 89(5): 1084 - 1084. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
