Intraabdominally insufflated carbon dioxide (CO2) during laparoscopy may have a specific effect on splanchnic circulation that may be unrelated to the effects of increased intraabdominal pressure alone. Therefore, the influences of insufflation with CO2 versus air on splanchnic circulation were compared.
Pigs were chronically instrumented for continuous recording of mesenteric artery, portal venous, inferior vena cava, and pulmonary arterial blood flow and portal venous pressure. After induction of anesthesia, CO2 or air was insufflated in 14 and 10 pigs, respectively. With the pigs in the supine position, intraabdominal pressure was increased in steps of 4 mmHg up to 24 mmHg by graded gas insufflation.
During air insufflation, mesenteric artery vascular resistance was unchanged, whereas mesenteric arterial blood flow decreased with increasing intraabdominal pressure. Shortly after CO2 insufflation to an intraabdominal pressure of 4 mmHg, mean arterial pressure, mesenteric arterial blood flow, and mesenteric arterial vascular resistance were increased by 21%, 12% and 9%, respectively. Subsequently, with the onset of CO2 resorption in the third minute, mean arterial pressure declined to baseline values and mesenteric arterial vascular resistance declined to 85% of baseline values, whereas mesenteric arterial blood flow continued to increase to a maximum of 24% higher than baseline values. At steady-state conditions during CO2 insufflation, mesenteric arterial blood flow was increased up to an intraabdominal pressure 16 mmHg but decreased at higher intraabdominal pressures.
In contrast to air insufflation, intraabdominal insufflation of CO2 resulted in a moderate splanchnic hyperemia at an intraabdominal pressure < or = 12 mmHg. At higher intraabdominal pressure values, pressure-induced changes became more important than the type of gas used.
CLINICAL and experimental studies have revealed that pneumoperitoneum can induce hemodynamic changes. In this context, most authors observed decreases in cardiac output, [1–5]whereas others reported no changes [6,7]or increases. Part of this inconsistency is attributable to differences in experimental conditions or design (e.g., patients' position, surgery, anesthesia). These studies did not consider separately the mechanical effects of increased intraabdominal pressure (IAP) and the biochemical properties of the insufflated gases. Although elevation of IAP induces a pressure-related reduction in inferior vena cava blood flow, cardiac preload, and cardiac output, no information is available about the additional hemodynamic effects of carbon dioxide (CO2) used as insufflation gas. Because elevated arterial CO2tension decreases mesenteric vascular resistance or increases splanchnic blood flow, a similar effect of intraabdominally insufflated CO2could be postulated.
Accordingly, the objective of the current study was to clarify the vasodilatatory properties of intraabdominally insufflated CO2on splanchnic arteries and its possible effects on systemic circulation. Therefore, in an experimental study in pigs, the effects of intraabdominal CO (2) insufflation on splanchnic circulation were compared with the effects of air insufflation.
This study was approved by the Governmental Animal Research Committee according to the German Animal Protection Law. Forty pigs (Deutsche Landrasse; 25 female and 15 males; weight, 35 - 51 kg) were chronically instrumented. Animals were anesthetized with propofol and fentanyl. After insertion of a central venous catheter, a left thoracotomy was performed, the pericardium was incised, and a precalibrated ultrasonic transit time flow probe ([empty], 16 - 20 mm; Transonic Systems, Ithaca, NY) was positioned nonconstrictively around the pulmonary artery above the ligamentum arteriosum for measurement of cardiac output. After a median laparotomy, the centrum tendineum of the diaphragm was dissected, and a transit time flow probe was placed around the inferior vena cava ([empty], 12 - 14 mm) cranial to the diaphragm. Additional flow probes were placed around the portal vein ([empty], 10 - 12 mm) and arteria mesenterica cranialis ([empty], 6 - 8 mm). A portal venous catheter was inserted via vena lienalis. All transducer leads and catheters were tunneled subcutaneously into the right axilla below the trapezius muscle, and the thoracic and abdominal cavities were closed.
All animals were allowed to recover during controlled conditions during which clinical examinations, metabolic analysis, and weight gain measurements were performed. After complete recovery (7 - 10 days after surgery), 24 pigs with no signs of impaired gastrointestinal function or intraabdominal lesions were studied.
After induction of anesthesia, two 7-French Swan Ganz catheters (Abboth, Wiesbaden, Germany) were placed into the abdominal aorta and the pulmonary artery. The leads of the flow probes were exposed and connected to the transient time ultrasonic flowmeter (T201; Transonic Systems). An intraabdominal insufflation catheter was inserted into the right lower abdomen via a minilaparotomy. After complete instrumentation, the anesthetized animals were allowed to stabilize for 3 h.
Anesthesia was induced with 2 mg/kg propofol and 2 [micro sign]g/kg fentanyl and maintained with 10 - 12 mg [middle dot] kg-1[middle dot] h-1propofol and 2 - 4 [micro sign]g [middle dot] kg-1[middle dot] h-1fentanyl according to clinical signs of inadequate anesthesia during the placement of the catheters. To eliminate the influences of fluctuating anesthetic depth, the doses of propofol and fentanyl necessary to maintain adequate anesthesia, as initially defined, were kept constant over time. The trachea was intubated, and mechanical ventilation (fractional inspired oxygen tension = 0.4) was set to maintain an end-tidal partial pressure of CO2of 34 +/- 2 mmHg by adjusting the respiratory rate. Body temperature was maintained at baseline values (range +/- 0.5 [degree sign]C) by heating blankets. Initially, all animals received a bolus dose of Ringer's lactate (10 - 20 ml/kg) until a central venous pressure of 6 mmHg was established, followed by infusion of 2 ml [middle dot] kg-1[middle dot] h-1Ringer's lactate.
Intraabdominal pressure and gas insufflation flow were recorded and regulated by adjustment of the Abdominal-CO2-Insufflator[trade mark sign](Waltz Elektronik GmbH, Rohrdorf, Germany) connected to the intraperitoneal trochar. Mean arterial pressure (MAP), central venous pressure (CVP), mean portal venous pressure (PVP), heart rate, mesenteric arterial blood flow (MABF), portal venous blood flow (PVBF), inferior vena cava blood flow (IVCF), and cardiac output (CO) were continuously recorded (100 values/s). Systemic vascular resistance (SVR) and mesenteric artery vascular resistance (MAVR) were calculated using the following formulas:
SVR =(MAP - CVP)[middle dot] 79.9 [middle dot] CO-1[dyne [middle dot] s [middle dot] cm-5
MAVR =(MAP - PVP)[middle dot] 79.9 [middle dot] MABF-1[dyne [middle dot] s [middle dot] cm-5
Blood samples in heparinized syringes were drawn according to the study protocol from all instrumented vessels and marked with the corresponding indices: arterial (a), pulmonary artery (v), and portal vein (pv). Whole blood CO2partial pressure was measured using a blood gas analyzer (IL1306 [registered sign]; Instrumentation Laboratory, Munich, Germany). Central venous concentrations of catecholamines in plasma were measured using high-performance liquid chromatography: epinephrine (normal values, 50 - 100 pg/ml; sensitivity, 2 pg/ml) and norepinephrine (normal values, 70 - 390 pg/ml; sensitivity, 2 pg/ml). Concentrations of vasopressin were assayed by direct radioimmunoassay (Hermann Bierman GmbH, Bad Nauheim, Germany; normal values for humans, 0.0 - 6.7 pg/ml). Because reference values for pigs have not yet been published, relative changes during the course of the study were analyzed.
Elimination of CO2(VCO2), oxygen consumption (VO2), and respiratory quotient (RQ) were determined using the Engstrom Elvira [trade mark sign] ventilator system (Gambro Engstrom AB, Broma, Sweden) with integrated metabolic monitor. VCO2, VO2and RQ were calculated every 60 s. Assuming constant metabolism, the endogenously produced CO2(VendoCO2) can be calculated at any time t using the RQ0measured before insufflation :
VendoCO2(t)= RQ0[middle dot] VO2(t)[ml/min]
Provided that arterial CO2tension is kept constant, the reabsorbed CO2is eliminated and not stored in the body. Thus, expired CO2volume consists of the endogenously produced and the reabsorbed CO (2), which permits calculation of CO2resorption (Delta VCO2)12:
Delta VCO2(t)= VCO2(t)- VendoCO2(t)= VCO2(t)- RQ0[middle dot] VO2(t)[ml/min]
For evaluation of the effects of CO2on the target variables, the peritoneal cavities of 14 and 10 animals were insufflated with CO2and air, respectively. All other procedures of the study protocol were identical in both groups. With animals in the supine position, IAP was applied in steps of 4 mmHg, from 0 up to 24 mmHg. Gases were insufflated with a constant flow of 1 1/min. The target IAP level was maintained constant using a flow of 16 l/min over 20 min. Between the ninth and eleventh minute at each stable IAP, the blood samples were drawn. After IAP reached 24 mmHg, pneumoperitoneum was deflated. Thirty minutes after desufflation, all parameters were registered again.
Between the minutes 12 and 15 at each IAP, mean values of all continuously recorded values, their related parameters, blood gas values, and concentrations of the vasoactive hormones in plasma were used for analysis of the effects of insufflation gases and IAP. For the period between minutes 9 and 15, steady-state conditions were assumed to have been achieved (<10% variation in the subsequent period with stable IAP). The respective values were subjected to a two-way repeated-measures analysis of variance with the within-groups factor, IAP, the between-groups factor, insufflation gas, and their interaction (gas x IAP). Once IAP proved to be significant (P < 0.05), a repeated-measures analysis of variance was performed in each group separately. Once interaction effects (gas x IAP) proved to be significant (P < 0.05), a further two-way repeated-measures analysis of variance was performed. The within-groups factor was restricted to an IAP of 0 mmHg and the clinically often recommended IAP of 12 mmHg. The effects of insufflation with CO2or air to an IAP of 12 mmHg were judged by the interaction term: gas x 12 mmHg.
Time analysis of the effects of intraabdominal insufflation on MAP, MABF, MAVR, and Delta VCO2was performed using two-way repeated-measures analysis of variance with the within-groups factor, time (three measuring intervals: 80 - 20 s before insufflation, and 150 - 210 and 540 - 600 s after starting insufflation), the between-groups factor, insufflation gas, and their interaction (time x gas). Once time or interaction effect proved to be significant (P < 0.05), within-group CO2multiple comparisons (time) were performed using Dunnett's test.
Baseline values ([almost equal to] 10 min before starting the study protocol at an IAP of 0 mmHg) were compared with values obtained 30 min after desufflation of the pneumoperitoneum to verify the time stability of the model using paired t tests (P < 0.10). In addition, the 90% confidence intervals of this comparison were related to mean baseline values. All statistical analysis was performed using StatView for Windows 4.55 (Abacus Concepts, Inc., Berkeley, CA).
Insufflation of Air
During insufflation of air, heart rate did not change significantly. Systemic vascular resistance and MAP were increased, and CO and IVCF were decreased with increasing IAP (Table 1). Central venous pressure (Table 1) and PVP (Table 2) were increased in parallel with changes in IAP. Mesenteric artery vascular resistance was constant over the entire IAP range, whereas MABF and PVBF decreased in an IAP-dependent manner (Table 2). In air-insufflated animals, no CO2resorption from the peritoneum was observed, and portal venous and mixed venous CO2partial pressures were unchanged (Table 3). Vasoactive hormones did not change because of the increasing IAP (Table 4).
Insufflation of CO2
In CO2-insufflatedanimals, heart rate and MAP were increased in relation to increases in IAP. Cardiac output and IVCF were increased at IAP values <or= to 12 mmHg; further increases in IAP resulted in a continuous decrease. Systemic vascular resistance was unchanged homologous to CO. Central venous pressure was increased in parallel with IAP (Table 1). Mesenteric artery blood flow and PVBF were increased up to IAP values <or= to 16 mmHg. Further increases in IAP resulted in a decrease in MABF and PVBF. Mesenteric artery vascular resistance was changed in a direction opposite that of MABF (Table 2). CO2resorption was increased in an IAP-dependent manner up to IAP values of 12 mmHg. To achieve stable arterial CO2partial pressures, the respiratory rate was increased from 10 +/- 1 to 16 +/- 3 min-1. Despite further increases in IAP, CO2resorption and respiratory rate remained constant. Portal venous and mixed venous CO2partial pressures were increased continuously (Table 3). The levels of vasopressin, epinephrine, and norepinephrine were altered in an IAP-dependent manner (Table 4).
Comparison between Insufflation Gases
With respect to all investigated IAP levels from 0 - 24 mmHg, insufflation of CO2resulted in a significantly different pattern (gas x IAP, P < 0.05) of all measured vasoactive hormones, CO2resorption, portal and mixed venous CO2partial pressures, and in all hemodynamic parameters except SVR and PVP (next to last column in Table 1, Table 2, Table 3, Table 4). At the clinically often recommended IAP of 12 mmHg, insufflation of CO2resulted in a significantly different pattern (gas x 12 mmHg, P < 0.05) of CO2resorption, portal and mixed venous CO2partial pressures, and in all hemodynamic parameters except SVR, PVP, and CVP (last column in Table 1, Table 2, Table 3, Table 4).
Time analysis at the IAP increasing from 0 to 4 mmHg revealed marked transient alterations in the CO2group. Shortly after CO2insufflation was started, MAP (121% of baseline), MABF (112% of baseline), and MAVR (109% of baseline) were increased up to the third minute. With the onset of CO2resorption in the third minute, MAP declined to baseline values and MAVR to 85% of baseline values, whereas MABF continued to increase (120% of baseline). At the ninth minute, a plateau was achieved for all parameters (Figure 1). Similar changes were observed during the first minutes of increasing IAP from 4 to 8 mmHg. Up to the third minute after insufflation to 8 mmHg, MAP was increased to 117% of steady-state values at 4 mmHg, MABF to 108%, and MAVR to 110%, respectively. With the onset of a further increase in CO2resorption, MAP and MAVR decreased again. At the ninth minute, a plateau was achieved for all parameters (values at steady state are shown in Table 1and Table 2). There was again a significant time effect regarding the gas used in the investigated parameters MAP, MABF, MAVR, and CO2resorption (time x gas, P < 0.05). At IAP values >or= to 12 mmHg, changes in MAP and MAVR did not show this biphasic character (time and time x gas, not significant).
Time Stability of the Model
For all measured variables except CO2resorption, baseline values did not differ significantly (P < 0.10) from values 30 min after desufflation in either group. The 90% confidence interval of the differences between baseline and postdesufflation values was <5% of baseline values for IVCF, MAP, portal venous CO2partial pressure, vasopressin, SVR, CO, epinephrine, and PVBF and <10% of baseline values for arterial CO2partial pressure, MABF, mixed venous CO2partial pressure, heart rate, norepinephrine, MAVR, PVP, and CVP. In the CO2group, CO2resorption remained elevated 30 min after desufflation (P < 0.001); however, the 90% confidence interval of the differences between baseline and postdesufflation values for CO2resorption was 8 ml/min, which is <10% of the maximum change.
There is controversy regarding the effects of pneumoperitoneum on splanchnic perfusion. [13–15]In this study, at low IAP values (<12 mmHg) after CO2insufflation increases in MABT and PVBF ([almost equal to] 25%) were observed during steady-state conditions. These increases in splanchnic perfusion and in CO can be explained either by decreases in afterload or by increases in preload. The increases in IVCF of [almost equal to] 20% may suggest the latter mechanism. Compression of the intraabdominal capacitance vessels may shift blood into the central veins. In air-insufflated animals with identical IAP values, however, IVCF and CO were unchanged. This is in agreement with the results of Barnes, who reported no increase in preload, CO, or splanchnic perfusion during hydrostatic elevation of IAP to 10 mmHg. Consequently, it may be concluded that the augmentation of blood flows up to an IAP of 12 mmHg is related to effects of CO2on the vasculature rather than to effects of increased IAP.
At IAP values >12 mmHg, blood flows decreases in both groups. Air insufflation resulted in a decrease of blood flows starting from baseline flow values, whereas with CO2insufflation, blood flows were decreased starting from the increased flow levels. These findings are consistent with previous results, which show that decreases in IVCF were induced at IAP values >or= to 20 mmHg. [1,2,16]This effect, clinically known as “supine vena cava syndrome,” was of less importance in our study, possibly because, in pigs, relative blood volume in lower extremities is lower than in humans. Another explanation may be that fluid substitution before gas insufflation prevented considerable cardiovascular changes induced by volume redistribution.
For assessment of transient changes, mesenteric perfusion parameters were analyzed in a time-dependent manner. A biphasic response was observed after CO2insufflation from an IAP of 0 to 4 mmHg. In the first phase (2 - 3 min) after the start of CO2insufflation, increases in MAP, MABF, and MAVR were observed. Previous studies suggest that such initial increases in MAP after intraabdominal insufflation of CO2with or without increased arterial CO2partial pressure are related to sympathetic activation, vasopressin release, [2,16,17]or CO2-relatedreflex mechanisms of mesenteric capacitance vessels. Because we observed unchanged concentrations of vasopressin and catecholamines in plasma 9 min after initiation of each IAP level, CO2insufflation without surgical trauma induced neither a release of vasopressin nor a systematic activation of catecholamines.
The second phase (4 - 9 min) is characterized by decreases in MAP and MAVR and by a further increase in MABF, coincident with the start of CO (2) resorption. In anesthetized dogs, alterations of splanchnic perfusion by arterial CO2tension have been investigated previously. Arterial hypercapnia (arterial CO2tension 65 mmHg) has been shown to induce vasodilation and, consecutively, to increase mesenteric perfusion by 40%. Although in the current study, arterial CO2partial pressure was kept constant, portal venous CO2content was increased, probably because of splanchnic CO2resorption. This indicates contact of capillaries with insufflated CO2. CO2, however, is known to dilate arterial capillaries or precapillary arterioles regulating MAVR. The coincidence of measurable CO2resorption and the decrease in mesenteric artery resistance support this conclusion. Similar effects have been reported by Brandt, who found increased inferior MABF values when CO2was insufflated into the colons of dogs.
The biphasic response of mesenteric hemodynamic and an increase in CO2resorption could be observed only at the first (0 - 4 mmHg) and the second (4 - 8 mmHg) steps of IAP increase. This may be explained by the finding that, up to an IAP of <or= to 12 mmHg, additional vessels on the peritoneal surface were influenced by CO2. Thus, the biphasic response of mesenteric vessels may be characteristics for the initial exposure of peritoneal tissues to the insufflated CO2.
The smooth muscle tone of mesenteric vessels influences SVR in a relevant ways. Accordingly, systemic hemodynamics were influenced by mesenteric CO2effects. In clinical studies, however, these local CO2effects may be masked by volume depletion, patient position, anesthesia, or surgery-induced sympathetic stimulation, explaining the divergent reports of changes in CO.
Time stability of the model can be influenced by the preparation, the anesthetic regimen, and the study protocol. We used a chronic model with only small acute interventions within a period of 3 h before measurements. Therefore, it is unlikely that the preparation may have influenced the measurements. Other interactions could be caused by side effects of the drugs. Propofol dilates mesenteric vessels, and fentanyl decreases splanchnic vascular resistance in a dose-dependent manner. In our study, however, alterations of splanchnic perfusion were reversible 30 min after desufflation during unchanged anesthetic infusion rates. Conversely, there was no evidence for inadequate anesthetic depth, such as trembling, sweating, or movements. The study protocol is characterized by a progressive sequence of increases in IAP. It cannot be excluded that acute alterations of the investigated parameters at each level depend on the proceeding IAP. Our results, however, derive from steady-state conditions, which are probably independent from the type of insufflation.
The results of this study suggest three clinical consequences regarding the insufflation of CO2. First, the increase in MAP with the beginning of CO2insufflation implies the risk of hypertension. Second, sustained abdominal insufflation with CO2results in increasingly greater resorption requiring increases in ventilation to maintain arterial normocapnia. Three, despite arterial normocapnia, splanchnic CO2tension is increased, which causes splanchnic hyperperfusion and these changes are associated with increases in CO. With IAP values >12 mmHg, pressure-induced changes become more important and reduce PVBF and MABF. We conclude that splanchnic circulation is not impaired during insufflation with CO2or air up to an IAP value of <or= to 16 mmHg, as used clinically.
The authors thank Prof. Dr. W. Erhardt (Institut fur Experimentelle Chirurgie) for help with surgery and anesthesia in pigs; Dr. P. Luppa and Dr. R. Probst (Institut fur Klinische Chemie) for measurement of concentrations of catecholamines in plasma; Prof. Dr. R. Senekowitsch (Nuklearmedizinische Klinik) for measurement of concentrations of vasopressin in plasma; and Prof. Dr. K. Ulm (Institut fur Medizinische Statistik und Epidemiologie) for instruction in statistical analysis.