Chapter 3 Anesthesia for ophthalmic surgery
The advances in veterinary anesthesia have paralleled those in veterinary ophthalmic surgery, and have resulted in less anesthetic risk, and improved patient care and management. Studies in different animal species, including humans, suggest frequent similar ophthalmic responses to tranquilizers, narcotic analgesics, dissociative anesthetics, inhalational general anesthetics, and neuromuscular relaxants; however, species differences can occur. Combinations of critical peri-, intra-, and postoperative topical and systemic drugs are common in veterinary ophthalmic patients, and should be accommodated by the choice of general anesthetic agents and protocols. At the same time, concurrent general anesthesia and ophthalmic needs must avoid drug selections that are contraindicated or incompatible when used simultaneously. A significant number of ophthalmic patients are old and may have other diseases that may influence the choice of general anesthetics. In this chapter the impact on the eye and associated structures of drugs administered as part of general anesthesia will be summarized.
Intraocular pressure (IOP) results from a relative equilibrium between aqueous formation, aqueous humor outflow, and the resistance of the fibrous tunics, e.g., the cornea and sclera, to pressure. The different drugs administered to tranquilize, sedate, and/or anesthetize animals may affect IOP directly by influencing the aqueous humor dynamics, or indirectly by causing hypercapnia, changes in extraocular muscle tone, hypoxemia, and hypothermia. Most general anesthetics lower IOP through actions on the central nervous, respiratory, and circulatory systems. The reduction in IOP is also related directly to the depth of general anesthesia. Most general anesthetics seem to lower IOP by an increase in the rate of aqueous humor outflow. Drugs that directly cause ocular hypotension can also produce ocular hypertension secondary to respiratory depression and acidosis that sometimes occurs with prolonged general anesthesia.
As a general observation, drugs that produce an abrupt increase in arterial blood pressure will result in a moderate increase in IOP. This elevation in IOP is usually transient as the aqueous humor dynamics rapidly readjust and return to normal levels of IOP. The major percentage of the resistance of aqueous humor outflow is determined by the episcleral venous pressure. Drugs that produce marked increases in central venous pressure and episcleral venous pressure can also temporarily elevate IOP. The increased central venous pressure may also expand the anterior and posterior uveal vascular beds, indirectly increasing IOP. Expansion of the uveal vascular channels may produce pressure on the vitreous when the globe has been opened, and force the vitreous, its patellar fossa, and even the lens toward the anterior chamber.
Drug-induced changes in extraocular muscle tone can influence IOP. As most animal species have lower scleral rigidity than humans, relaxation of the extraocular and retrobulbar muscles markedly decreases the pressure of these orbital tissues upon the globe. The ketamine-associated elevation in IOP that occurs in humans is thought to be directly related to the increase in extraocular muscle tone. In lightly anesthetized dogs, intravenous administration of succinylcholine can result in a short-term elevation in IOP. This 5–10 min elevation in IOP is thought to be related to the unusual sensitivity of the extraocular muscles to succinylcholine, and the initial muscle fasciculations that occur during the onset of the drug’s action. When the insertions of the extraocular muscles are severed in the cat, succinylcholine administration does not change IOP.
Animals, in general, possess lower ocular rigidity than humans. As a result, when either the cornea or the sclera is incised and IOP released, the entire globe tends to collapse. The sclera in both the dog and cat possesses elastic fibers, in addition to the major complement of collagen, and as a result the sclera lacks rigidity. When the globe is collapsed, corneal and intraocular surgical procedures are more difficult to perform. The level of IOP in animals with low ocular rigidity also enhances the effects of retrobulbar muscle tone. Once the globe has collapsed, extraocular muscle tone may become of greater concern, tending to distort and displace forward the vitreous, lens, and anterior uvea.
Corneal abrasions, drying of the cornea, conjunctival irritation, and reduced tear formation have been associated with general anesthetics in humans and animals. Ketamine in cats has been associated with corneal drying, although the individual roles of reduced tear formation rates and loss of the protective blink reflex have not been differentiated. Hence, in cats undergoing ketamine anesthesia the corneas should be protected by copious amounts of ophthalmic petroleum-based bland ointment and/or the eyelids closed temporarily by tape.
The same applies in dogs, and petroleum-based bland ointment should be applied to both eyes, depending on the type of ophthalmic surgery. The rate of aqueous tear formation in dogs, as determined by Schirmer’s tear test, after combinations of subcutaneous atropine, intravenous thiamylal sodium, and halothane or methoxyflurane was reduced by about 70% within 10 min and by 97% after 60 min. Another study has indicated that subcutaneous atropine reduces Schirmer’s tear test levels in normal dogs by about 55% at 60 min after drug administration. In dogs and cats with reduced levels of tear formation, the administration of parenteral and/or topical atropine can abruptly lower Schirmer’s tear test levels to zero, and initiate the clinical signs of keratoconjunctivitis sicca.
Although the topical effects of general anesthetics have not been reported in large and special species animals, petroleum-based bland ointment is applied liberally to protect the corneoconjunctival surfaces during prolonged general anesthesia.
The oculocardiac or oculorespiratory cardiac reflex was first described by Aschner and Dagnini in 1908 in two simultaneous but independent reports. The afferent aspect of this reflex is carried in the long and short ciliary nerves, the ciliary ganglion, and the ophthalmic branch of the trigeminal nerve via the gasserian ganglion to the trigeminal sensory nucleus. Short internuncial fibers within the reticular formation connect the trigeminal sensory nucleus to the visceral motor nucleus of the vagus nerve and its descending nerve to complete the efferent limb to the heart. The afferent limb of the ophthalmic division of the trigeminal nerve does not appear to be unique. Intraorbital stimulation of the third, fourth, and sixth cranial nerves will also produce consistent respiratory prolongation, but more variable cardiac responses.
There appears to be significant species differences for the oculorespiratory cardiac reflex (documented in dogs, cats, horses, and birds), and whether the cardiac, the respiratory, or a combination of both components occurs. In addition to the induced bradycardia, in some species such as the dog, the concurrent respiratory depression can be more profound. The reflex may be initiated by a number of ophthalmic manipulations, including ocular pressure massage for glaucoma, intraorbital injections of local anesthetics (which are also used to block this reflex), surgical traction of the extraocular muscles, and manipulations of the eyelid muscles. In dogs under general anesthesia, neuromuscular blocking agents, and controlled ventilation, only the cardiac portion of this reflex can be appreciated.
There may be some individual animal variations relative to the oculorespiratory cardiac reflex in the dog and cat, with only some animals demonstrating this reflex consistently. The primary effect in the cat seems to be respiratory; in the dog, respiratory depression is the dominant response, but bradycardia can develop. To manage the oculorespiratory cardiac reflex, a number of strategies have been developed. To diminish or completely block the vagal effect on the heart, intravenous atropine is the standard treatment. Unfortunately, atropine administration yields inconsistent results. Consequently, the rationale to administer parenteral atropine preoperatively in dogs differs among veterinary anesthesiologists. One school recommends against the routine administration of parenteral atropine preoperatively. If bradycardia develops during the surgical procedure, surgery is temporarily halted and atropine is administered intravenously. Other veterinary anesthesiologists continue to recommend routine use of preoperative parenteral atropine to prevent the potential oculorespiratory cardiac reflex from developing.
Unfortunately, intravenous atropine may not only increase the heart rate in the dog, but also increase the possibility of ventricular dysrhythmias. The intravenous dose of atropine to treat and/or prevent the oculorespiratory cardiac reflex in the dog seems critical. In children, although the prophylactic use of parenteral atropine seems to lower the incidence of the oculocardiac reflex, it has also been associated with severe and prolonged ventricular dysrhythmias. Low levels (0.015 mg/kg IV) of atropine in dogs with an experimentally induced oculorespiratory cardiac reflex may actually enhance respiratory depression. Higher doses of atropine (0.023–0.04 mg/kg) may eliminate the bradycardia but prolong the apnea. Of these two complications, clinical management of apnea with controlled ventilation is the most feasible solution. An alternative to atropine in the dog is glycopyrrolate (0.01 mg/kg IV, usually given in two divided doses; often the second dose is not necessary) which appears as effective in preventing the oculorespiratory cardiac reflex but produces tachycardia.
Under most circumstances, if the oculorespiratory cardiac reflex develops during ophthalmic surgery, surgery is suspended for several minutes and the depth of general anesthesia is increased. Less aggressive surgery is then slowly resumed while the respiratory and cardiac rates are carefully monitored. Fortunately, the onset of the oculorespiratory cardiac reflex is usually in the early aspects of surgery, and in intraocular surgical procedures before critical manipulations have begun.
During the induction of general anesthesia most injectable and inhalational anesthetics produce a downward and inward rotation of the eye that limits access to the cornea, anterior chamber, and anterior segment. As the globe is rotated ventromedially, the nictitating membrane simultaneously protracts to nearly cover the cornea. Some degree of enophthalmia also develops, decreasing further the exposure of the cornea and globe for surgery. In the large and giant breeds of dogs, access to the eye is already limited, and these drug effects can severely compromise surgical exposure of the eye. This poor positioning of the globe can handicap the surgeon by impairing observation of the entire cornea and anterior segment, increasing the difficulties of surgical manipulations, and unnecessarily prolonging the surgery.
Several strategies have been developed to correct the rotation of the globe and exposure difficulties associated with general anesthetics. Unfortunately, most of these remedies to improve exposure may also result in some additional operative risks. Sutures may be placed in the anterior sclera or the rectus muscle insertions, and anchored to the eyelid specula or drapes. Scleral clips may be used similarly. Retrobulbar injections with saline positioned directly into the extraocular muscle cone to push the eye forward, or external to the extraocular muscle cone to turn the globe, may also produce noticeable inward compression of the posterior segment and additional pressure on the vitreous body. The animal cornea and sclera unfortunately lack rigidity, unlike humans and primates in general, and with traction or compression these tunics may become distorted. For certain types of conjunctival and corneal surgery, the distortion of the globe associated with these procedures may be inconsequential. However, when intraocular surgical procedures are planned, any preventable pressure on the globe, in whole or in part, should be avoided. Administration of the different neuromuscular blocking agents has replaced the need for extrabulbar injections to manipulate the position of the globe for surgery.
Pupil size has been used historically to monitor the depth of general anesthesia. Without local control by topical mydriatics or miotics, pupil size may vary from marked dilatation to pinpoint in the lighter levels of general anesthesia, to progressive mydriasis with deep general anesthesia. For conjunctival and corneal surgical procedures, pupil size is often adjusted preoperatively depending upon the concurrent ophthalmic disease. Often the pupil is dilated. Drug-induced iridocycloplegia helps reduce the pain associated with preoperative anterior uveitis, and pupil dilatation reduces the likelihood of posterior synechiae formation.
In the event of corneal and intraocular surgery, control of pupil size may become critical. Maximum mydriasis is essential for cataract extraction; preoperative pupillary dilatation is usually achieved with 0.3% scopolamine combined with 10% phenylephrine, 1% atropine, or a combination of 1% atropine and 10% phenylephrine. Topical non-steroidal anti-inflammatory agents, such as 0.03% sodium flurbiprofen, can also assist in the maintenance of pupillary dilatation. Prostaglandins appear to be released when the anterior chamber is entered surgically and initiate strong miotic activity. Endocapsular phacoemulsification of canine cataracts requires maximal mydriasis. Without the combination of topical mydriatics, topical and parenteral corticosteroids, and non-steroidal anti-inflammatory agents, the microsurgical refinements and higher success rates for canine cataract surgery would not have been possible.
The extraocular muscles are well developed in the dog and, in addition to the four rectus and two oblique extraocular muscles, include the retractor oculi muscle that inserts onto the sclera under the rectus muscle insertions and behind the globe’s equator. This bulk of extraocular muscles may produce pressure and indent the posterior segment of the globe, even with optimal general anesthesia. The extraocular muscle pressure, combined with the low scleral rigidity of the dog, seems to be more important once the anterior chamber has been entered, as during cataract and lens removal. If general anesthetics also increase central venous pressure, additional orbital pressure on the globe may develop from the extensive venous plexuses within the orbit.
In the cat, the effects of the extraocular muscles during general anesthesia seem less important than in the dog. This may be caused by the poorly developed cat extraocular muscles and the limited orbital space. As a result, increased pressure on the posterior segment is less important and does not appear to be a problem clinically.
Several strategies have been developed to address the potential extraocular muscle pressure and its adverse effects when the anterior chamber has been entered surgically. Neuromuscular blocking agents have now become routine for canine and equine intraocular surgery; in addition to greatly reducing extraocular muscle tone, these agents result in optimal eye position for microsurgery.
Preanesthetic medications are designed to facilitate a smooth induction of general anesthesia, and help prevent possible drug-related complications. The controversial routine use of parenteral atropine as an anticholinergic agent has already been discussed. Parenteral glycopyrrolate (0.01 mg/kg IM) is preferred because of fewer cardiac effects.
Sedatives and tranquilizers are often employed preoperatively, especially before intraocular surgery. Both sedatives and tranquilizers lower IOP, probably by increasing the outflow of aqueous humor. Among the phenothiazine tranquilizers, acepromazine maleate is the most frequently recommended. Acepromazine maleate (0.03–0.1 mg/kg IM) is frequently utilized, not only because of the resultant tranquilization, but also for its anti-arrhythmogenic effect associated with the stabilization of the myocardium against catecholamine stimulation and arrhythmogenic agents. The phenothiazine tranquilizers also possess an anti-emetic action perioperatively. Both postoperative vomiting and retching in humans can elevate IOP indirectly by abrupt venous pressure increases. The same effect probably occurs in dogs.
Xylazine is not recommended perioperatively in ophthalmic patients because it can cause vomiting and severe bradycardia. Acepromazine may slightly prolong the recovery from general anesthesia, hypothermia, and arterial hypotension, but usually provides a smoother, less traumatic recovery.
Most narcotics seem to slightly lower IOP in those animal species studied. The two major advantages of narcotics are that: 1) these drugs are potent analgesics; and 2) they can be chemically antagonized if drug reversal is necessary. Unfortunately, most of these agents except for meperidine are also potent vagotonic and respiratory depressants. Use of narcotic derivatives prior to and following ophthalmic surgery has become more frequent. Occasionally, if the postoperative recovery becomes traumatic, parenteral narcotics are quite effective, probably because of the analgesic effects.
Most barbiturates lower IOP in animals. This ocular hypotension seems to result from depression of the diencephalon, an increased facility of aqueous humor outflow, and relaxation of the extraocular muscles. Ultrashort-acting barbiturates, such as thiopental and thiamylal (8–12 mg/kg IV) are effective induction agents. The reduction in IOP after administration of these drugs seems to be related to relaxation of the extraocular muscles and an increase in aqueous humor outflow rather than from arterial blood pressure changes. As these agents are potent respiratory depressants, endotracheal intubation should follow immediately after barbiturate administration. Intubation should be standard protocol in all ophthalmic surgical patients. Intermittent and often copious lavage of the corneal and conjunctival surfaces during surgery may exit the nasolacrimal system and accumulate in the mouth and pharynx.
Ketamine may be an exception to the rule for injectable anesthetics. Elevated IOP has been associated with ketamine, with increased tone of the extraocular muscles in humans. Ketamine, recommended for the cat, has been reported to either not change or increase IOP in the cat by 10%. Ketamine, used alone, is not recommended for the dog because of its tendency to produce seizures. Ketamine, a dissociative anesthetic, may be injected intramuscularly for the induction of general anesthesia in cats. Ketamine is often combined with diazepam in dogs to reduce the possibility of seizures and produce muscular relaxation. An anticholinergic, such as atropine, is also administered to minimize salivation. After general anesthesia is sufficiently deep to permit intubation, general inhalational anesthesia may be initiated for longer duration surgeries.
There are other injectable agents that are now used as induction agents for small animals, and experiences with some of these agents, such as propofol, midazolam, and Telazol®, have been excellent. Propofol, as an induction agent, has a recommended intravenous dose in small animals of 6 mg/kg, and then to effect, and has largely replaced the barbiturates. After rapid, smooth and excitement-free onset of general anesthesia, the duration is also relatively short (range 2.5–9 min). Usually administered as a slow bolus injection (to avoid apnea) for the induction of inhalational anesthesia, propofol can be injected repeatedly; however, its short duration of effect requires several injections for relatively short periods of time. Recovery after propofol is usually very rapid, and excitement free. Propofol is thought to lower IOP. Propofol has become the preferred induction agent world-wide in small animals, and the sole general anesthetic for many short-term ophthalmic procedures.
Propofol (2,6-diisopropylphenol) is only soluble in water, and is mixed immediately before use. It comes in sterile glass ampoules and without preservatives. It fits a two-compartment open model, with rapid distribution from the plasma into the tissues and rapid metabolic clearance from plasma. It is metabolized by conjugation primarily by the liver and kidney. It is administered as an intravenous bolus at doses that range from 2.5 mg/kg (sedated dog) to 8 mg/kg (unsedated dog) to allow tracheal intubation and the initiation of inhalational anesthesia. Propofol’s anesthesia is quite brief; in unsedated dogs recovery is only 15 min.
Propofol can also be used for the maintenance of anesthesia administered by continuous infusion or intermittent bolus. Propofol has minimal analgesic effects, and drugs with analgesic effects, such as opiates, should be administered concurrently. Propofol lowers IOP in humans, and this effect has also been reported in dogs (a decline of 26%).
Midazolam is a water-soluble benzodiazepine. It does not induce anesthesia when used alone; hence midazolam is often combined with ketamine, or one of the ultrashort-acting thiobarbiturates (thiamylal or thiopental) to induce general anesthesia. Telazol® consists of equal parts of a dissociative agent, tiletamine, and a benzodiazepine, zolazepam. Once in solution, Telazol® has a limited shelf-life of 4 days at room temperature and 14 days at 4°C. The recommended dose for dogs is 6.6–13.2 mg/kg IM or SC, and for cats 9.7–15.8 mg/kg IM or SC. After deep intramuscular injection of Telazol®, onset of anesthesia is within 2–5 min and the recovery to walking requires 3–5 h. Induction of anesthesia with Telazol® is usually smooth, but recovery can be traumatic.