Cardiopulmonary Effects of Intrathoracic Insufflation in Dogs

Introduction

The use of minimally invasive surgery as a diagnostic modality and therapeutic tool is increasing in both the human and veterinary fields. Potential benefits of minimally invasive surgery over conventional techniques include decreased pain, healing time, and postoperative complications.^1–4 With the advent of variable angled rigid scopes and flexible endoscopes, the number and variety of surgical procedures being attempted via minimally invasive surgery are rapidly increasing.^1–11

Thoracoscopy can be performed with or without carbon dioxide (CO₂) insufflation of the thoracic cavity. However, laparoscopy requires insufflation to allow visualization of the abdominal viscera. Currently most thoracoscopy procedures in humans are performed without insufflation.¹³¹² One main-stem bronchus is intubated with a specialized endotracheal tube to allow ventilation of the intubated lung and passive collapse of the nonintubated lung. Alternately, one lung is blocked with a specialized endotracheal tube with a balloon-tipped stilet to allow passive or active insufflation-aided collapse of the blocked lung and ventilation of the nonblocked lung.^48–12 Visualization of the thoracic cavity is primarily limited to the site of lung collapse. The available sizes of these specialized endobronchial tubes and the patient’s disease condition may limit the use of single lung-lobe intubation.

The incomplete mediastinum in the dog allows CO₂ insufflation to increase visualization of the entire thoracic cavity compared to single lung intubation. In addition, CO₂ insufflation without single lung-lobe intubation does not require bronchoscopic confirmation of endotracheal tube placement and is technically easier to perform, which can speed lung collapse and thereby reduce operative time.^348–12 There are no size restrictions for patients, and the technique can be performed on patients with diffuse lung pathology that cannot tolerate single lung ventilation. Recommendations for upper limits of intrathoracic pressure with insufflation of the thoracic cavity in humans and dogs are 8 to 10 mm Hg.¹³ However, physiological effects of these intrathoracic pressures on cardiac output (CO) and oxygen (O₂) delivery in dogs have not been thoroughly evaluated.¹⁴⁶⁹¹³ In contrast, laparoscopy and the effects of abdominal insufflation on cardiovascular physiology are well documented, and an upper safe limit has been demonstrated.¹² The significant and severe cardiovascular effects caused by abdominal insufflation demonstrate the need for studies on the effects of intrathoracic insufflation.

The purpose of this study is to quantify the effects of incremental positive-pressure insufflation of the intrathoracic cavity in anesthetized dogs on CO, arterial pressure (AP), central venous pressure (CVP), heart rate (HR), and the percent saturation of hemoglobin with oxygen (SPO₂).

Materials and Methods

This study was approved by the Animal Care and Use Committee at Washington State University. Seven male, mixed-breed dogs weighing 20 to 38 kg were used for this study. All animals were screened for health by physical examination, packed cell volume, total protein, and electrocardiogram (ECG). The dogs were premedicated with acepromazine (0.05 mg/kg body weight, intramuscularly [IM]), atropine (0.04 mg/kg body weight [IM]), and butorphanol (0.2 mg/kg body weight [IM]). Anesthesia was induced with sodium thiopental (10 mg/kg body weight, intravenously [IV]). The dogs were intubated, and a medium plane of anesthesia (i.e., stable respiration and heart rate, abolished laryngeal and palpebral reflexes, and profound muscle relaxation) was established. Anesthesia was maintained with isoflurane in O₂ delivered to a semi-closed circle anesthetic machine^a by a precision out-of-circle vaporizer.^b The dogs were placed on and under circulating heated water blankets to reduce heat loss. Lactated Ringer’s solution was administered IV at 20 mL/kg body weight per hour for the first hour, followed by 10 mL/kg body weight per hour for each subsequent hour under anesthesia. All dogs were part of a terminal surgery teaching laboratory and underwent exploratory laparotomy and visceral biopsies prior to undergoing this experiment. Significant bleeding was not noted in any dog during the laboratory procedures preceding the experimental study.

Once abdominal closure was completed, the dogs were transferred to the experiment group. Anesthesia was maintained with a vaporizer setting of 1.5% isoflurane in O₂ throughout the experiment. An incision was made over the right ventral cervical region, and the right carotid artery was exposed and catheterized^c to allow measurement of AP. A 30-cm catheter^d was placed percutaneously in the left jugular vein, with the tip in the cranial vena cava located near the right atrium to measure CVP. The AP and CVP were each measured directly using a calibrated pressure transducer^e connected to a physiological recorder.^f

An 8-French catheter introducer^g was placed into the right jugular vein, and a thermodilution catheter^h was placed to measure CO. Proper placement of the thermodilution catheter was confirmed by obtaining characteristic pulmonary arterial pressure waveforms, displayed on an oscilloscope, during insertion of the catheter. To measure CO, 5 mL boluses of 5% dextrose solution, chilled to 5° to 8°C, were administered through the proximal port of the thermodilution catheter immediately following the expiratory phase of ventilation. A computerⁱ was used to measure CO.

A pulse oximeter was placed on the animal’s tongue to monitor SPO₂.^j A thermometer probe^j was inserted 10 cm into the rectum to measure core body temperature. Heart rate was monitored using the pulse oximeter and continuous ECG.

The dogs were then placed on mechanical ventilation.^k Respiratory rate was initially set at 12 breaths per minute with a tidal volume of 20 to 24 mL/kg body weight, which was sufficient to cause a maximal inspiratory pressure of 25 cm H₂O. Once the catheters were secured, the dogs were positioned in right lateral recumbency. After 5 minutes of stabilization, experimental variables CO, systolic and diastolic AP, SPO₂, CVP, HR, ventilation rate, and temperature were recorded immediately following expiration in the ventilation cycle. Mean arterial blood pressure was inadvertently not recorded due to an error in the design of the study variables record sheet used during the experiment.

A blunt metal urinary catheter^l was inserted through a small skin incision at the left fifth intercostal space and introduced into the intrathoracic cavity. The pressure of the intrathoracic cavity was allowed to equilibrate with atmospheric pressure (0 mm Hg), and after 5 minutes of stabilization, the baseline experimental variables were recorded.

A CO₂ insufflator^m was attached to the blunt metal urinary catheter via a three-way stopcock to regulate intrathoracic pressure (IP). Intrathoracic pressure was increased to 3 mm Hg (the lowest setting allowed by the insufflator), and experimental variables were recorded after 5 minutes. Intrathoracic pressure was increased by increments of 1 mm Hg, and after 5 minutes at each new pressure, the experimental variables were recorded. As IP increased, tidal volume and respiratory rate were adjusted to maintain SPO₂ >85%. However, maximum inspiratory pressure was maintained <25 cm H₂O [18 mm Hg] above IP to protect the lungs from hyperinflation trauma.¹⁴,15 Once SPO₂ remained <85% despite increases in minute volume, insufflation was discontinued and the CO₂ was released from the pleural cavity. After IP returned to atmospheric pressure for 5 minutes, the experimental variables were again recorded.

The dogs were euthanized at the end of the study with pentobarbital (395 mg/2 kg body weight, IV).

Data was analyzed using a one-way repeated measures analysis of variance (ANOVA). Pairwise comparisons were performed using a Dunnet’s test, with values obtained at baseline before introduction of the intrathoracic catheter as the controls. Significance was set at P≤0.05.

Results

Mean core body temperature before insufflation was 35.0°C, and no change in core temperature occurred in any of the dogs during the experiment. Four dogs were under anesthesia for 6 hours prior to measurement of baseline experimental variables. Three dogs were under anesthesia for 8.5 hours prior to measurement of baseline experimental variables. The median time to complete the experiment once the animals were instrumented was 110 minutes. No significant differences in experimental variables were found between these two groups. Consequently, data from all seven dogs was pooled for the subsequent analyses.

Results of increasing IP on CO, AP, SPO₂, CVP, and HR are recorded in the Table. Cardiac output and systolic and diastolic APs decreased significantly at 3 mm Hg IP. Significant decreases in SPO₂ were noted at 10 mm Hg IP. Significant increase in CVP was noted at 6 mm Hg IP. Heart rate was significantly decreased at 5 to 6 mm Hg IP.

Discussion

Cardiac output decreased significantly with minimal insufflation of the intrathoracic cavity and dropped by almost 50% from baseline at 10 mm Hg, which is the recommended maximum safe insufflation pressure.¹³ Similar results were reported for pigs in which cardiac index (CI) decreased by 70% when IP was increased to 10 mm Hg during thoracic insufflation.¹⁶ One study, in which single lung intubation and intrathoracic insufflation were used, experienced marked cardiovascular compromise when the IP was inadvertently allowed to rise above 8 mm Hg.⁴ While the exact mechanism for a decrease in CO is not known, it has been hypothesized that an increase in IP above atmospheric pressure collapses central veins and increases pulmonary resistance.^16–19 This combination of events decreases the volume of blood returning to the heart and increases the pressure required to eject blood from the right heart. The end result is a decrease in stroke volume and CO. In the pig study, CI returned to within 5% of baseline levels after insufflation was discontinued, and IP decreased to atmospheric pressure.¹⁶ Similar results were obtained in this study and in the study in which IPs in dogs were inadvertently increased above 8 mm Hg.⁴ This suggests that the decrease in CO observed during intrathoracic insufflation is only a transient event and completely reversible following normalization of IP.

The effect of thoracic insufflation on HR is difficult to understand because of the multitude of factors that control HR.²⁰ In this study, the HR remained essentially unchanged throughout almost the entire range of IPs, and it was slightly decreased at IPs of 5 and 6 mm Hg. This effect on HR occurred despite the fact that both systolic and diastolic APs were significantly decreased at an IP of only 3 mm Hg, and both continued to drop further throughout the range of IPs attained in this study. The authors had anticipated that decreased arterial blood pressure would result in an increase in HR.²⁰ However, the net effect of the various reflexes responsible for controlling HR resulted in little change, or even a small decrease in HR, despite the significant drop in arterial blood pressure observed during thoracic insufflation.

Both systolic and diastolic APs decreased significantly at 3 mm Hg IP. Surprisingly, the systolic AP, while decreased significantly, did not fall below clinically acceptable levels of 80 mm Hg or to the degree that might be expected considering the marked decrease in CO.²¹ Arterial pressure is the result of blood volume in the vessel as well as the tone of the smooth muscle in the vessel wall. These, in turn, are determined by CO, baroreceptor reflexes, and centrally mediated nervous responses to arterial blood pressure.²² The authors suggest that the decrease in systolic and diastolic APs is predominantly the result of the decrease in CO. The reason that the systolic AP did not drop as low as may have been expected, compared to the marked drop in CO, is unclear. The other factors that affect AP may have increased blood pressure either to overcome the higher CVP created by the increased IP, or in response to centrally mediated nervous detection of decreased AP. A similar comparison of diastolic AP decreases being “clinically” relevant was not possible because the clinical significance of diastolic AP as a cardiovascular monitoring parameter has not been clearly defined. Diastolic AP was the only experimental parameter that failed to return to within 5% of baseline after the positive IP was removed. This may be the result of not allowing the diastolic AP enough time to return to preexperimental levels, or it may indicate that there are some nontransient effects of positive-pressure intrathoracic insufflation. The authors suspect that this was due to not allowing the diastolic AP enough time to return to preexperimental levels, but the possibility of some nontransient effects of positive-pressure intrathoracic insufflation on diastolic AP cannot be excluded.

Central venous pressure increased in direct relationship to IP. Under normal circumstances, CVP is a function of venous return and ability of the heart to pump blood from the right atrium. Further, an increase in CVP would usually be expected to increase CO via the Frank-Starling mechanism. However, with increased IP due to thoracic insufflation, CVP increases due to extramural pressure on central veins. This, in turn, increases resistance to blood returning to the heart, thereby decreasing CO despite an increase in CVP.

The SPO₂ did not decline in this study until IP exceeded 9 mm Hg. Gross estimates of partial pressure of oxygen (PaO₂) can be determined by extrapolation from SPO₂ using the O₂-hemoglobin saturation curve.21,23 At 8 mm Hg IP, PaO₂ was estimated in the authors’ study to be >50 mm Hg, the minimum acceptable level for maintaining normal O₂ delivery to the body.21 However, SPO₂ is not a direct measurement of O₂ delivery to the tissues.24 Flow rate and volume also contribute to O₂ delivery; therefore, local tissue hypoxia may occur in patients with low CO and a normal pulmonary system despite normal SPO₂ values.21 The end result is that a patient being monitored by SPO₂ as an indicator of tissue oxygenation may suffer significant, life-threatening tissue hypoxia despite a normal SPO₂.

The animals used for this experiment were chosen because the authors felt that they would be adequate models, provided that there was not excessive hemorrhage or other complications during the teaching laboratory. Prolonged isoflurane anesthesia up to 4.5 hours has not been shown to have significant effects on CO and HR.²⁵ While temporal effects of isoflurane anesthesia have been seen in other species (i.e., horses and humans), there are currently no studies demonstrating the effects of isoflurane anesthesia and recumbency in dogs for >4.5 hours.²⁶²⁷ In addition, comparisons with other species and anesthetic gases are difficult to support, as the type of gas and species can dramatically affect cardiopulmonary variables. In addition, the use of animals already slated for euthanasia decreased the number of animals required for terminal experiments.

A number of recent articles in the veterinary literature have shown that different degrees of intrathoracic visualization are required, depending on the procedure being performed.^{47–10132829} For the most part, these studies evaluate models where insufflation was not performed (e.g., passive insufflation alone) or was performed using single lung-lobe intubation and short-duration insufflation to speed collapse of the nonintubated lung. In these studies, the authors have concluded or postulated that insufflation may increase visualization and can be helpful or required in order to complete the surgery.^47–10

Conclusion

Because of associated significant and severe decreases in CO at even low IPs, sustained positive-pressure thoracoscopy should be used with caution in dogs. There are no reported critical lower limits of CO; therefore, a specific or safe IP to remain at or below cannot be ascertained from the data of this study. However, the authors conclude that increased IP and duration of insufflation should be kept to a minimum to decrease the negative effects on CO. The authors also conclude that the variables routinely used to monitor patients under anesthesia, such as AP and HR, may not be reliable indicators of cardiovascular status in dogs undergoing insufflation-aided thoracoscopy. The authors suggest that SPO₂ and serial blood gases in addition to the routinely measured variables may be better indicators of the cardiovascular status in dogs undergoing insufflation-aided thoracoscopy.

Alternatives to sustained positive-pressure intrathoracic insufflation include short-term positive-pressure intrathoracic insufflation, single lung ventilation, or a combination of these techniques. Single lung intubation is associated with fewer cardiopulmonary changes when compared to the results of this study.^4101130 Further studies are required to evaluate the extent of intrathoracic visualization provided and the cardiopulmonary effects produced by all of the above techniques in dogs.