The influence of a modified open lung concept (mOLC) on pulmonary and cardiovascular function during total intravenous anaesthesia (TIVA) in horses was evaluated. Forty-two warmblood horses (American Society of Anesthesiologists class 1 to 2), scheduled for elective surgery (mean [sd] weight 526  kg, age 6.4 [5.4] years) were randomly divided into three groups: ventilation with mOLC, intermittent positive-pressure ventilation (IPPV), and spontaneous breathing. Premedication (0.8 mg/kg xylazine), induction (2.2 mg/kg ketamine and 0.05 mg/kg diazepam) and maintenance of anaesthesia with TIVA (1.4 mg/kg/hour xylazine, 5.6 mg/kg/hour ketamine and 131.1 mg/kg/hour guaifenesin), with inhalation of 35 per cent oxygen in air, were identical in all horses. Heart rate, respiratory rate, mean arterial blood pressure (MAP), pH, and arterial partial pressure of oxygen (paO2) and carbon dioxide (paCO2) were evaluated. Data were collected every 10 minutes from 20 to 90 minutes anaesthesia time. Factorial analysis of variance and Tukey's post hoc test were used for statistical analysis (a=5 per cent). Horses in the mOLC-ventilated group had an overall significantly higher paO2 (16.9 [1.0] v 11.7 [1.34] v 10.5 [0.57] kPa) and lower MAP (93.1 [5.47] v 107.1 [6.99] v 101.2 [5.45] mmHg) than the IPPV and spontaneously breathing groups, respectively.
Statistics from Altmetric.com
ATELECTASIS and pulmonary ventilation/perfusion mismatch leading to hypoxaemia are common problems during general anaesthesia in horses. Many attempts have been made to try to prevent them. Intermittent positive-pressure ventilation (IPPV) is a common method of artificial ventilation to correct hypoventilation or apnoea and decrease the arterial partial pressure of carbon dioxide (paCO2) to normal (Weaver and Walley 1975, Nyman and Hedenstierna 1989, Day and others 1995). The effects of IPPV on the arterial partial pressure of oxygen (paO2) are controversial. Some studies have accomplished a higher paO2 by IPPV (Day and others 1995, Edner and others 2005), while some have shown no significant effect on arterial oxygenation (Weaver and Walley 1975). A possible reason for hypoxaemia during general anaesthesia is the development of atelectasis. There are two major types of atelectasis: compression and gas resorption (Hedenstierna 2003). To reopen collapsed lung tissue in human beings, high inspiratory pressures are followed by an immediate application of positive end-expiratory pressure (PEEP) to prevent the lung from collapsing again (the open lung concept) (Lachmann 1992, Papadakos and Lachmann 2007). There are two strategies to achieve recruitment (reopening) of collapsed alveoli: a stepwise PEEP and positive inspiratory pressure (PIP) titration, or a single hyperinflation in combination with PEEP (Neumann and others 1999). The first strategy has been used in ponies and horses, and led to improved oxygenation (Wettstein and others 2006, Schürmann and others 2008). For more practicability, a modified open lung concept (mOLC) including hyperinflation of the lung during inhalation anaesthesia has been described (Hopster 2007). A previously defined PEEP and three steps of recruitment manoeuvres with PIPs of 60, 80 and 60 cmH2O is used when impaired oxygenation is detected. Horses ventilated by mOLC showed significantly improved pulmonary function compared with horses ventilated by IPPV. In contrast, the application of PEEP without recruitment manoeuvres showed no significant changes compared with horses ventilated by IPPV (Pauritsch 1997).
The influence of gas composition on atelectasis formation and pulmonary shunt plays an important role during general anaesthesia. The recurrence of atelectasis after recruitment manoeuvres depends on the inspired oxygen fraction. In human beings breathing 100 per cent oxygen, re-expanded lung tissue collapsed again after five minutes. When breathing 40 per cent oxygen, recollapse was delayed until 40 minutes (Rothen and others 1995). In horses, Marntell and others (2005) described a significantly higher pulmonary shunt with 100 per cent oxygen compared with breathing air.
Spontaneous breathing is usually maintained during total intravenous anaesthesia (TIVA), but the aim of the present study was to compare different types of ventilation and their effects on pulmonary and cardiovascular function.
Material and methods
The study had a prospective, randomised clinical trial design with horses matched in triplicate.
Forty-two healthy warmblood horses (classified under the American Society of Anesthesiologists physical states classification system as normal and healthy  or having mild systemic disease ) weighing more than 400 kg were randomly divided into three equal groups: ventilation with mOLC, IPPV and spontaneous breathing. Each group contained six horses placed in dorsal recumbency and eight horses placed in lateral recumbency depending on the surgical intervention. The type of surgical interventions were similar in number in each group (Table 1). After one horse was assigned to a group, a horse matched for similar surgical intervention and weight was assigned to one of the remaining two groups, and the next horse of similar surgical intervention and weight was placed in the remaining group. All horses were referrals to the Clinic for Horses at the University of Veterinary Medicine Hannover for elective surgery. Horses were determined to be healthy by clinical examination and preoperative arterial blood gas analysis. In the event of an intraoperative paO2 lower than 8 kPa horses were excluded from measurement and the inspired oxygen fraction was increased.
Horses were premedicated with 0.8 mg/kg xylazine (Xylazin 2 per cent; CP-Pharma), and anaesthesia was induced by use of 2.2 mg/kg ketamine (Narketan 10; Vétoquinol) and 0.05 mg/kg diazepam (Diazepam AbZ 10 mg Ampullen; AbZ Pharma). All drugs were administered intravenously. After tracheal intubation, horses were connected to a circle system operating in a semi-closed mode and breathed a mixture of 35 per cent oxygen in air. Maintenance of anaesthesia was achieved by continuous intravenous infusion of 1.4 mg/kg/hour xylazine, 5.6 mg/kg/hour ketamine and 131.1 mg/kg/hour guaifenesin (Myolaxin; Vétoquinol) using a volumetric pump (Infusomat fmS; B. Braun Melsungen). In the event of anaesthesia lasting longer than 90 minutes, maintenance of anaesthesia was changed over to inhalational anaesthesia using isoflurane.
IPPV was performed by use of a modified Bird Mark 7 Servo-Respirator with a PIP of 25 to 30 cmH2O, respiratory frequency of 5 to 8 breaths/minute and inspiration:expiration (I:E) ratio of 1:2 (Table 2). In the mOLC-ventilated group, a PEEP of 10 to 12 cmH2O was added and recruitment was performed when the paO2 was below the limit of 18.7 kPa. Recruitment manoeuvres consisted of a PIP of 60, 80 and 60 cmH2O, respectively, over three consecutive breaths with an I:E ratio of 1:1. In between the PIP, normal PEEP was administered at the end of each breath. Respiratory pressures were measured using a modified Pitot tube (Horse-lite flowmeter) located between the y-piece and endotracheal tube. The Pitot tube was calibrated according to Moens and others (2009) with a 3 litre syringe (Hans Rudolph). The limit of 18.7 kPa ≈140 mmHg paO2 for recruitment manoeuvres was calculated using the alveolar gas equation:
from which 30 per cent (±3 per cent) was subtracted because of the assumable limitations of gas exchange during anaesthesia. Where pAO2 is the alveolar partial pressure of oxygen, pB is the barometric pressure (767 mmHg), pH2O is the partial pressure of water vapour (47 mmHg), FiO2 is the fraction of inspired oxygen, paCO2 is the arterial pressure of carbon dioxide, and R is the respiratory quotient (a ratio of carbon dioxide production to oxygen consumption, assumed to be 0.8) (Robinson 2009).
Cardiovascular monitoring consisted of measurement of the heart rate derived from the ECG (lead II) and arterial blood pressure. A catheter was placed either in one of the branches of the facial artery or in the third metatarsal artery and was connected to a DTX-Plus pressure transducer (Becton Dickinson). The level of the transducer was identical in lateral and dorsal recumbency on the level of the heart base. Mean arterial blood pressure (MAP) and heart rate were shown on an anaesthesia monitor (Kardiokap 5; Datex-Ohmeda) and were recorded every 10 minutes.
Respiratory frequency and end-tidal carbon dioxide (EtCO2) concentration were determined using a side-stream capnograph, calibrated before each measurement, and recorded every 10 minutes.
Blood gas samples were collected every 10 minutes from the arterial catheter used for blood pressure measurement, and pH, paO2, arterial oxygen saturation (SaO2) and paCO2 were measured immediately after collection of samples by a blood gas analyser with correction for body temperature.
The statistical analysis was performed with SAS v 9.1. For calculation of the linear model, the MIXED procedure was used. To test the influence of time and ventilation parameters, a two-way analysis of variance and Tukey's post hoc test were used. Data are presented as the arithmetic mean (sd). Values of P<0.05 were considered to be statistically significant.
There was no significant difference between the three groups (P>0.05) in terms of the mean (sd) age or bodyweight of the horses, or the number of horses in each group, the number undergoing different types of surgery or the number of dorsal or lateral recumbency (Table 1).
Anaesthesia was induced and maintained in all horses without complications, with minor variations in heart rate, respiratory rate and MAP. During 90 minutes of TIVA, horses showed a surgical depth of anaesthesia with minor characteristics of palpebral reflex, muscle tone, tear production and eyeball position.
Horses in the mOLC group showed a significantly higher paO2 at all individual time points than horses in the other groups (Fig 1). There was no significant difference between the IPPV and spontaneous breathing groups. The combined measurement of all time points showed a significantly higher paO2 in the mOLC group compared with the IPPV and spontaneous breathing groups. To achieve a minimum of 18.7 kPa paO2, a mean of 3.4 recruitment manoeuvres were necessary during anaesthesia.
There was no significant difference in MAP between the mOLC and spontaneous breathing groups, but there was a difference in mean values over time. There was a significant (P<0.001) decrease in MAP in all three groups during anaesthesia. Horses in the IPPV group showed a significantly higher MAP than horses in the mOLC group (Fig 2). Administration of inotropes was not necessary.
Horses in the spontaneous breathing group had a significantly higher respiratory frequency than those in the other groups (spontaneous breathing 8 [0.82], mOLC 5.1 [0.32], IPPV 5 [0.23] breaths/minute; P<0.01) and additionally a lower pH (spontaneous breathing 7.33 [0.01], mOLC 7.42 [0.01], IPPV 7.42 [0.01] mmHg; P<0.001). Furthermore, a significantly higher EtCO2 (spontaneous breathing 51.3 [1.5], mOLC 37.6 [0.9], IPPV 37.1 [0.92] mmHg; P<0.001) and a significantly higher paCO2 (spontaneous breathing 55.4 [1.28], mOLC 43.3 [1.25], IPPV 44.7 [0.54] mmHg; P<0.001) was observed in horses in the spontaneous breathing group compared with those in other groups.
Eleven horses were excluded from analysis due to technical problems, such as alteration of recumbency (two horses), or hypoxaemia (paO2 <8 kPa), occurring during anaesthesia (nine horses). The hypoxaemic horses had to be treated by raising the FiO2 to improve oxygenation. Six of these nine horses were in the spontaneous breathing group and were placed in dorsal recumbency; one was assigned to the IPPV group and placed in dorsal recumbency, one was assigned to the PPV group and placed in lateral recumbency, and one was assigned to the mOLC group and placed in dorsal recumbency.
The use of open lung concept ventilation in horses during inhalation anaesthesia has been reported (Hopster 2007, Schürmann and others 2008). In those studies, horses had a significantly higher paO2 compared with horses ventilated by IPPV but also a considerable decrease in MAP. Depression of cardiovascular function when using the open lung concept is a common secondary effect and can be explained by the high intrathoracic pressures that occur during recruitment manoeuvres (Levionnois and others 2006, Wettstein and others 2006). High PIPs compress the large vessels, such as the aorta and cranial and caudal vena cava, and thus reduce preload and afterload of the heart. Furthermore, the increase in positive pressure in the thorax correlates positively with resistance of blood flow through the lung vessels and results in less filling of the left atrium and a subsequent reduction in blood pressure. The reduced cardiac output results in reduced oxygen delivery to the tissues. For adequate oxygen supply to the tissues, sufficient paO2 and blood pressure are both necessary. In the present study, a similar phenomenon could be observed in the mOLC group compared with the spontaneous breathing group. On one hand, there was an improved paO2 in the mOLC group, but on the other, there was impaired cardiovascular function.
Most horses in the mOLC group showed better oxygenation compared with horses in the other groups, independent of recumbency and surgical intervention. The frequently described atelectasis (Nyman and others 1990, Rothen and others 1993, Hedenstierna 2003) could be successfully treated by recruitment manoeuvres in combination with PEEP (Nyman and others 1987, Wettstein and others 2006, Levionnois and others 2006, Hopster 2007). IPPV combined with PEEP was sufficient to reach the set limit of 18.7 kPa in only two horses, whereas in all the other horses in the mOLC group manoeuvres had to be applied.
Mechanical ventilation for horses in the mOLC group was performed with a predefined PEEP of 10 to 12 cmH2O and was not individually titrated for each horse, as recommended in the literature (Agrò and others 2004, Wettstein and others 2006, Papadakos and Lachmann 2007). This simplified method proved to be effective in previous studies (Hopster 2007, Schulte-Bahrenberg 2008). As an indication of adequate gas exchange in the lung, a paO2 of 18.7 kPa was considered to be the lower limit based on the presumption that the paO2 increases linearly to the FiO2. Due to the assumed limitation in gas exchange during general anaesthesia, 30 per cent was subtracted.
The application of only 10 cmH2O PEEP is controversial. Wilson and McFeely (1991) and Heinrichs (1992) detected a significant increase in paO2 due to enlargement of alveoli and consequently an enlargement in the surface for gas exchange, whereas the majority of authors described no or only a minimal improvement (Nyman and Hedenstierna 1988, Wilson and Soma 1990, Pauritsch 1997).
Horses anaesthetised by TIVA using a combination of an a2-agonist, ketamine and guaifenesin commonly maintain a higher mean arterial blood pressure than horses maintained by inhalational drugs (Taylor and others 1998, McMurphy and others 2002). Despite the application of recruitment manoeuvres in the present study, no horse needed to be supplemented by administration of inotropes for improvement of cardiovascular function and for achieving a MAP of more than 70 mmHg. This suggests that recruitment manoeuvres during TIVA are possible in the majority of cases.
In the present study, a mixture of xylazine, ketamine and guaifenesin was administered by a volumetric pump at a constant rate for 90 minutes to provide standard conditions for each animal. Therefore, an individual dose for each horse and its depth of anaesthesia could not be provided. The doses of xylazine and guaifenesin were comparable, but the ketamine dose of 5.6 mg/kg/hour used in the present study was higher than that used in other studies (Greene and others 1986, Young and others 1993, Spadavecchia and others 1999, Muir and others 2000); however, this dosage was used in preliminary tests at the Equine Clinic, School of Veterinary Medicine Hannover, and as a consequence was chosen for use in the present study.
The respiratory frequency of horses in the spontaneous breathing group was higher over the whole 90-minute anaesthetic period compared with that of the mechanically ventilated horses. A reason for this was the predetermined frequency of approximately five breaths/minute that was set on the ventilator for the mOLC and IPPV groups. Another, although not surprising, result is the slight respiratory acidosis observed, which arises from the lower pH and the increased paCO2. To avoid stimulation of the respiration regulatory centre and a non-physiological change in acid-base balance during mechanical ventilation, the respiratory parameters were set to accomplish a maximum of 40 to 45 mmHg paCO2. So, despite the higher respiratory rate observed in the spontaneous breathing group, a smaller minute ventilation could be detected by an increase in paCO2. A reason for this could be increased dead-space ventilation and increased development of atelectasis due to the pressure of the gastrointestinal tract on the diaphragm and, thus, the resulting limited lung function. Another possible reason for increased paCO2 are smaller tidal volumes occurring during spontaneous breathing.
In anaesthetised horses, mild hypercapnia leads indirectly to improved cardiovascular function through stimulation of the sympathetic nervous system with an increased systemic blood pressure and higher myocardial contractility (Wagner and others 1990, Khanna and others 1995, Taylor 1998). Therefore, it is not clear whether the higher MAP over time in the spontaneous breathing group compared with the mOLC group was a result of the absence of mechanical ventilation or the mild hypercapnia. However, when horses were mechanically ventilated with mild hypercapnia, they showed a similar cardiovascular function to that observed in horses in the spontaneous breathing group (Mizuno and others 1994, Khanna and others 1995). Additionally, Wagner and others (1990) detected that horses ventilated with mild hypercapnia showed a better cardiovascular function than horses ventilated with normocapnia. Nevertheless, horses in the IPPV group showed higher, but non-significant, mean MAP values over time compared with horses in the spontaneous breathing group; therefore, no influence of carbon dioxide could be detected in the present study.
Maintenance of general anaesthesia in horses using TIVA and breathing 35 per cent oxygen generally showed adequate arterial oxygen saturation above 90 per cent. Six of nine horses excluded from the present study due to hypoxaemia were breathing spontaneously and were positioned in dorsal recumbency. This suggests that the combination of dorsal recumbency and spontaneous breathing may increase the likelihood of hypoxaemia. A possible reason for this could be a changed ventilation-perfusion mismatch with an increased intrapulmonary shunt volume and an increased incidence of the development of atelectasis (Nyman and Hedenstierna 1989, Nyman and others 1990). Most general anaesthesia performed under field conditions involves horses that are spontaneously breathing air. If it is assumed that a mixture of 35 per cent oxygen in air in combination with dorsal recumbency is not suitable to provide a paO2 of more than 8 kPa over 90 minutes, then general anaesthesia under field conditions would probably cause hypoxaemia.
To ensure an uncomplicated recovery period, surgery was completed within 90 minutes of anaesthesia or, in 30 of 42 horses, maintenance of anaesthesia was changed to inhalational anaesthesia. Given this limitation, the influence of different types of ventilation on the recovery period could not be evaluated. The need to change from TIVA to isoflurane after 90 minutes is another important limitation of TIVA and implies that surgeries with an assumed time of more than 90 minutes should be performed only in isoflurane-equipped facilities.
The results of the present study provide new information about mechanical ventilation during TIVA. Thus, ventilation of horses using a mOLC with inhalation of 35 per cent oxygen in air can be recommended during TIVA. Insufflation of oxygen can be applied even under field conditions and is likely to improve paO2.
Provenance not commissioned; externally peer reviewed
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.