Anaesthetic management of caesarean sections in dogs

A caesarean section is the surgical removal of fetuses from the pregnant bitch. Caesarean section is mainly indicated by dystocia, but elective caesarean sections are increasingly common for certain breeds of dog. Elective caesarean section may be performed for:

• Brachycephalic breeds
• Dogs that have a history of having required caesarean section or of difficult parturition
• Perceived lower risk or cost than emergency management
• High-value litters
• Litters of known large or small size
• Dams carrying the litter of a sire that was much larger than the dam, such as after accidental mating
• Convenience (occasionally).

The VetCompass research programme has reported breed-associations and outcomes of caesarean section across the UK (O’Neill et al, 2017; 2019). In summary:

• Breeds such as the Staffordshire Bull Terrier, Golden Retriever and Jack Russell Terrier were more commonly presented for dystocia overall
• The risk of dystocia was greater in brachycephalic and toy breeds
• Brachycephallic breeds were subsequently 1.5 times more likely to proceed to caesarean section
• Age and weight were a risk factor
• Bitches aged 3-5 years were three times more likely to encounter dystocia than younger bitches
• Dams weighing less than 10 kg or 40-45 kg are at increased risk of dystocia.

Surgical urgency has been linked to an increased mortality rate (Brodbelt et al, 2008), which is often the case with unplanned caesarean sections. There have been recent studies investigating protocols for caesarean sections elected before parturition (De Cramer et al, 2017; Groppetti et al, 2019; De Cramer and Nbthling, 2020). One study put the mortality rate associated with caesarean section at between 2.6% for conservative caesarean section and 4.2% for caesarean section with ovariohysterectomy (Conze et al, 2020). However, other studies have reported much lower mortality rates (Moon et al, 1998; 2000; Metcalfe et al, 2014).

Anaesthetic

The anaesthetic plan must aim for maternal and neonatal survival, presenting a unique challenge to the anaesthetist. Exposure of the fetuses to anaesthetic drugs, since many cross the placenta (Griffiths and Campbell, 2015; Groppetti et al, 2010; 2019), can depress mentation at birth. This creates a delicate balance between depth of anaesthesia, analgesia and fetal vitality.

Fetal survival, particularly following caesarean section, is heavily dependent on the initial transition from inside to outside the womb (Hillman et al, 2012), and so this period must be supported as much as possible.

As with any anaesthetic, there is no one standard method. However, some authors advocate a certain approach, summarised here (Mason, 2006; Ryan and Wagner, 2006a; Duke-Novakovski et al, 2016):

• Preparation of the patient conscious and with minimal handling
• Intravenous cannulation, clipping of the surgical site and an initial skin clean
• The patient is transferred to theatre either before or immediately after induction and intubation
• Injectable induction with either propofol or alfaxalone intravenously (or mask induction with inhalant anaesthetic agent) without premedication
• Anaesthesia is maintained with inhalant anaesthesia
• Opioid analgesia can be administered intravenously after removal of the final fetus.

Many bitches requiring emergency caesarean section are mentally depressed enough at the time of surgery to allow the omission of a sedative protocol. However, it is a largely unbalanced anaesthetic, since propofol and isofluorane lack the analgesia often provided by a multimodal approach balanced by premedicants or intraoperative medications. Fractious patients may require a propofol dose towards the upper end of the dose range for induction, risking neonatal hypoxia (since propofol is associated with hypotension and apnoea). The suggested dose of propofol for dogs not premedicated is 6-7 mg/kg (Allerton, 2020), although dosing to effect is advised. Other authors suggest 3-8 mg/kg for un-premedicated caesarean section (Mason, 2006). Intramuscular premedication may be justified to place an intravenous cannula without excessive stress, but the side effects need to consider altered fetal and maternal physiology. Some evidence suggests that sex steroids around parturition have anti-nociceptive actions (Cogan and Spinnato, 1986; Gintzler and Liu, 2012). However, this is unlikely to be sufficient for abdominal surgery.

Physiological changes during pregnancy

Cardiovascular and blood pressure

Uterine blood flow and fetal supply is entirely dependent on mean arterial pressure as there is no uterine mechanism for autoregulation. Any change to the dam, such as stress, pain, excessive anaesthetic depth or hypovolaemia, will affect the fetuses too.

There are many cardiovascular changes during human pregnancy (Sanghavi and Rutherford, 2014). Haemodynamic changes have not been as well investigated in bitches as they have been in humans, but are thought to be similar (Almeida et al, 2017). To maintain mean arterial pressure, cardiac output/work increases by 30-50% (Tan and Tan, 2013) as a result of:

• Increased heart rate, especially at parturition
• Preload increases by about 45% as a result of increased plasma volume (Sanghavi and Rutherford, 2014) - this appears to be exacerbated by a greater number of fetuses (Tan and Tan, 2013)
• Decreased afterload as a result of systemic and renal vasodilation (Sanghavi and Rutherford, 2014).

At parturition, these changes are exaggerated by as much as 80% as a result of myometrial contractions forcing blood into systemic circulation, pain increasing heart rate and catecholamine release (Sanghavi and Rutherford, 2014). This creates acute demand on the heart, and reduces its reserve capacity to compensate for increased demand. Bitches with pre-existing heart disease, in particular, can struggle.

Because of the dam’s higher blood volume, signs of hypotension/shock can be delayed in some cases. There is evidence that the baroreceptor reflex is less functional during pregnancy (Brooks and Keil, 1994; Brooks et al, 2012), reducing the dam’s ability to compensate for volume changes. On the other hand, preoperative and intraoperative stressors could overstretch the heart and lead to decompensation.

Management

Gentle handling and calm should be prioritised until the bitch’s stability has been established. Non-essential procedures, such as blood sampling, can be delayed until the animal is anaesthetised or can be performed at placement of the intravenous line. Transfer between induction and theatre should be as quick as possible.

Hypotension must be closely monitored and addressed early. Specific guidelines have been produced for managing hypotension in humans following spinal anaesthesia for caesarean section (Kinsella et al, 2018). These advocate maintaining systolic arterial pressure above 90% of the preinduction systolic arterial pressure. No such guidelines are available for managing hypotension during canine caesarean section but, extrapolating from experience in human medicine, the need to closely monitor and intervene early can be pre-empted.

Pregnant animals may be less responsive to vasopressors than expected (Paller, 1984) and consequences for the fetus must be considered when dosing approaches the top end of recommended ranges.

Other recommendations include:

• Hydration status and blood pressure monitoring should be available before induction
• Keep vaporiser settings appropriately low to maintain acceptable depth of anaesthesia, while reducing the risk of hypotension
• Ephedrine can be used to induce vasoconstriction. Heart rate may increase and acidosis may occur in neonates (Reynolds and Seed, 2005). Phenylephrine has become preferred in humans to avoid fetal acidosis (Ngan Kee et al, 2009), although evidence is not strong in high-risk caesarean section (Heesen et al, 2019)
• If hypotension results from bradycardia, glycopyrrolate is recommended over atropine, as it is less able to cross the placenta
• The discussed cardiovascular changes reduce cardiac reserve. Liberal fluid therapy/boluses could lead to decompensation earlier than would be expected in another animal without pre-existing cardiac disease. Hypovolaemia should be closely attended to, and hypotension must be addressed cautiously:
• Some authors advocate a pre-emptive bolus of crystalloids at 20 ml/kg for epidural-induced hypotension in bitches (Jones, 2001). A single dose of ephedrine may be preferable (Steagall, 2017), but may cause fetal acidosis (Duke-Novakovski et al, 2016)
• Use of intravenous colloids should be considered for treatment of maternal hypotension (Duke-Novakovski et al, 2016).

Tilting of the patient into slight lateral recumbency has long been advocated for caesarean section, because the gravid uterus can compress the aorta and vena cava in humans. This phenomenon appears not to be problematic in bitches (Probst and Webb, 1983; Probst et al, 1987; Tan and Tan, 2013), likely as a result of implantation in the uterine horns, off of midline, as opposed to human foetal implantation in the uterine body, which lies midline over the aorta. However, tilting should be considered if a single fetus is confirmed.

Hypertension may be a result of pain, which is discussed further on in the article.

Haematological effects

Production of red blood cells increases alongside plasma volume, but disproportionately, resulting in a dilutional anaemia. Packed cell volume (PCV) generally falls into the low-normal category around parturition (Robertson, 2016; Kimura and Kotani, 2018).

Except for planned caesarean section, many patients will present to the clinic after several hours of labour. The effort of labour, in addition to the usual history of hyporexia, can result in signs of dehydration.

Other abnormalities of consideration include glucose and calcium, which may both be low with prolonged dystocia.

Management

A blood sample should be collected before induction if possible. A mid-normal PCV, or high blood urea nitrogen/creatinine, suggests fluid deficits may need correcting before administration of any pre-medicants or inhalant anaesthetics, as these may precipitate hypotension. Colloids may be more appropriate than crystalloids to prevent overload.

Pulmonary compromise

The gravid uterus reduces the functional residual capacity of the lungs by about 10-25% in humans (Tan and Tan, 2013). This may cause areas of compression atelectasis through increased abdominal pressure on the diaphragm. Functional residual capacity is the volume of gas left in the lungs at the end of expiration - the gas left for oxygen transfer during exhalation that prevents atelectasis. A similar change to functional residual capacity is suspected in dogs, though not confirmed (Tompoulidou et al, 2018). The degree of this compromise may not directly translate to animal species. Additionally, oxygen demand is increased by the fetus and stress. These changes increase the risk for desaturation of haemoglobin.

Delivery of oxygen to tissues depends on the Bohr effect - the effect of blood CO₂ (and consequently pH) on the oxygen dissociation curve. During pregnancy, progesterone lowers the maternal ‘normal’ partial pressure of carbon dioxide in arterial blood (PaCO₂) to account for fetal acid production (including fetal CO₂), resulting in respiratory alkalosis. This supports efficient gas transfer at the placenta.

Blood buffering systems largely maintain pH. When hypercapnia develops, these buffering systems can become overwhelmed, resulting in rapid acidosis, which prevents efficient gas transfer at the placenta (Omo-Aghoja, 2014).

The minimum alveolar concentration of inhalational gases are about 25-40% lower during pregnancy, possibly as a result of progesterone (Erden et al, 2005). Owing to the aforementioned changes in tidal volume and drug sensitivity, changes in anaesthetic depth can be more rapid in caesarean section patients, and require frequent assessment.

Pregnant animals are at higher risk of vomiting (Tan and Tan, 2013) and, therefore, aspiration pneumonia. Gastric emptying is delayed, oesophageal sphincter tone is reduced, and the uterus places pressure on the stomach. Insufficient analgesia can also cause regurgitation.

Management

Pre-oxygenation with 100% oxygen before propofol induction and apnoea increases desaturation time from about 70s to 300s in healthy animals (McNally et al, 2009) and facemasks were superior to flow-by in this respect (Ambros et al, 2018). Consequently, fetal oxygenation should be better maintained.

If possible, the oesophagus should be suctioned at the beginning and end of anaesthesia, as silent regurgitation is not uncommon. Rapid intubation should be facilitated, ideally with a cuffed endotracheal tube, with the bitch in sternal recumbency and head elevated. Extubation may be delayed slightly during the recovery period. Mask induction is possible if familiar, but provides little protection against regurgitation.

The vaporiser setting should be kept as low as possible to reduce the risk of hypotension.

Pulse oximetry, capnography and blood pressure monitoring should be available. End-tidal carbon dioxide should be kept between 35 and 45 mmHg and mean arterial pressure between 65 and 75 mmHg. Intermittent ventilation may be required to maintain these parameters. Excessive ventilation may cause hypocapnia, which will also affect oxygen dissociation at the placenta.

Once the fetuses are removed, a ventilatory sigh or small amount of positive end-expiratory pressure may help to reinflate areas of compression atelectasis, improving oxygenation and recovery. A slight tilt to raise the head throughout surgery (reverse Trendelenburg position) may reduce pressure on the diaphragm and thorax, relieving respiratory work and improving functional residual capacity.

Analgesia

Many products are not licensed, or contraindicated, for use in pregnant animals. The owner should be informed of the risks and rationale for using such products, and consent for off-licence use should be recorded if required. Appropriate analgesia will allow reduction of anaesthetic agents and smooth recovery, milk letdown and interaction with neonates.

Local analgesia

Local analgesics generally prevent negative systemic effects of analgesic drugs, and can be used preoperatively, intraoperatively or postoperatively.

Infiltration of local anaesthetic agents into the skin can be very effective at reducing surgical stimulation and improving postoperative analgesia (Cicirelli et al, 2022a). These blocks can be performed by a registered veterinary nurse under direction of a veterinary surgeon in the UK.

Line blocks are injected subcutaneously along the proposed incision before or after surgery, which would be immediately above the linea alba for caesarean section (avoiding mammary tissue as much as possible), as demonstrated in Figure 1.

Splash blocks apply local anaesthetic topically to a tissue during surgery, for example over the linea alba once repaired. For splash blocks, local anaesthetic should be applied to a relatively dry site with little active vascular ooze, to prevent removal or washing away.

Manipulating the ovary and ovarian pedicle is often acutely painful. Lidocaine can be directly applied to the ovary as a splash block or injected into the ovarian pedicle (Cicirelli et al, 2022a; 2022b).

The transversus abdominis plane block is another option, but requires experience and ultrasonography for placing it (Schroeder et al, 2011).

Epidural analgesia

An epidural injection places analgesic and/or anaesthetic drugs in the space immediately above the dura mater of the spinal cord, blocking sensation caudally. Systemic uptake is low, making it an attractive option for caesarean section. One study reported the time from induction to surgery was only 5 minutes longer when an epidural was used, though this was likely in a referral population (Martin-Flores et al, 2021).

Local anaesthetics paralyse all nerve types, and sensory and sympathetic fibres are more sensitive than motor fibres (Jones, 2001). Consequently, lumbar sympathetic outflow is often reduced, causing transient hypotension. Dogs undergoing caesarean section may be more sensitive to this. Pregnant bitches also require a lower drug volume because the epidural space is reduced (Jones, 2001; Dias et al, 2018). If morphine is included, buprenorphine should not be used postoperatively because it will compete at the mu-opioid receptors.

One study reported lower isoflurane requirements and better Apgar scores at 5 and 20 minutes after extraction, with lower isoflurane requirements and better Apgar scores when lidocaine alone was used (Antonczyk and Ochota, 2022). However, a study on epidural methadone showed no differences in maternal or fetal cardiovascular parameters compared with intramuscular methadone. Apgar scores were similar, despite lower umbilical methadone levels in the epidural group (Romagnoli et al, 2019).

Apgar scoring

Neonates can be triaged using Apgar scoring, invented by Virginia Apgar for human beings. Conveniently, Apgar is also a mnemonic for:

• Appearance
• Pulse
• Grimace
• Activity
• Respiration.

An Apgar score is a subjective snapshot, so serial assessments by the same observer are recommended (Vassalo et al, 2015). A modified Apgar has been established for animals (Veronesi et al, 2009), which takes minutes and requires only a stethoscope. Exact details of how to score have been described (Veronesi, 2016). The modified Apgar assesses (Table 1):

• Heart rate
• Respiratory rate
• Reflex irritability
• Mobility
• Mucus membrane colour.

Scores can be supplemented with a number of biochemical parameters (Vassalo et al, 2015; Castagnetti et al, 2017; Groppetti et al, 2019), of which blood glucose and lactate are usually accessible using a drop of blood taken from the paw pad or ear tip on a hand-held glucometer/lactatometer. Most neonates will have a degree of acidosis at birth, which in normal neonates often corrects after 60 minutes.

Neonatal survival for the first 24 hours is positively correlated with Apgar score. Veronesi et al (2009) grouped neonates into critical (0-3), low (4-6) and normal (7-10) groups, identifying pups that need greater attention. In the original study, roughly half of the critical group did not survive 24 hours, whereas extremely few of the low group did not survive if appropriate intervention was made.

Investigation into the suitability of the modified Apgar score in two brachycephalic breeds showed these pups can appear less viable but still improve significantly between 5 and 60 minutes after birth (Batista et al, 2014). More support may be required for these litters. A revised Apgar also suggested that small breeds can appear more distressed but have similar viability to larger breeds with higher scores (Veronesi et al, 2022).

Management of neonates

For low/critical neonates, supplemental oxygen, spot glucose measurements and warming should be started. Reversible anaesthetic drugs should be antagonised as soon as possible. Stimulating breathing is a priority, to reverse hypoxaemia and remove gaseous anaesthetics. Pale/cyanotic mucous membranes and bradycardia indicate hypoxaemia. Solving hypoxaemia will resolve bradycardia in most cases.

Inspiration can be supported by:

• Bulb syringes to remove airway obstructions. Swinging should be avoided, as it can cause cerebral haemorrhage and neck injuries
• Tactile stimulation of the chest, genitals, lumbar area, and umbilicus
• Sharp stimulation of the Jen Chung/Renzhong acupuncture point (Lee et al, 1977; Traas, 2008)
• Atipamezole can be given subcutaneously at 50 μg/puppy (De Cramer et al, 2017) and naloxone intranasally/sublingually/subcutaneously/intramuscularly at 0.04 mg/kg
• Doxapram has fallen out of favour unless neonates are apnoeic or refractory to other interventions, because it can make hypoxia worse through increased metabolism (Ryan and Wagner, 2006b).

If blood glucose is low, transmucosal glucose can be supplied. Suckling should be encouraged or 50% dextrose can be supplied orally if neonates are alert (Traas, 2008).

Warming reduces the neonate’s energy expenditure. Neonates should be thoroughly dried with hand towels to reduce thermal losses. Bair huggers, hair dryers, ‘hot hands’ or hot water bottles covered by towels can all be effective if used carefully. If radiant heat sources are used, neonates should be spontaneously moving and have space to move away. Incubators are also effective and should be set to 32°C and 60% humidity (Traas, 2008).

Sufficient and experienced staff members are required for resuscitation of critical neonates. Excellent direction on interventions for these patients can be found in Traas (2008) (Figure 2).

Conclusions

Anaesthesia of pregnant dogs requires knowledge of specific changes to the physiology of the at-term dam in order to ensure survival of mother and neonates. These changes can make the anaesthetic less stable, and the surgical team should be well prepared to deal with rapid changes in parameters, such as blood pressure and oxygenation, or preempt these changes to support the bitch. Analgesic premedicants may be omitted from the anaesthetic protocol as a result of their effect on fetuses, but local anaesthetic agents can be considered because of their lack of systemic effects while contributing to a balanced anaesthetic. Being prepared to rapidly assess neonates through Apgar scoring allows efficient interventions and care to be given to neonates that may benefit most.

Key Points

• The physiology of the pregnant bitch at the time of c-section is significantly different to the non-pregnant bitch, and these changes must be understood to administer appropriate interventions or medications.
• Analgesia may sometimes be foregone in favour of neonatal stability. Local and epidural analgesia appears safe and should be considered in order to provide a balanced anaesthetic, making the anaesthetic more stable and improving the bitch’s comfort postoperatively.
• Apgar scoring should be used for all neonates to direct timely support for survival. The team should be familiar with specific interventions described here and elsewhere.