Materials and Methods

Case selection—Medical records of cats anesthetized for placement of a feeding tube at the Hospital for Animals at Cornell University from January 1995 through December 2004 were assessed. Inclusion criteria included cytologic confirmation by a pathologist of hepatic lipidosis with no other major systemic disease processes (eg, neoplasia or heart failure), anesthetized for placement of a feeding tube without other surgical procedures, complete ICU records, and recorded outcome (eg, discharge from ICU).

Medical records review—Data extracted included age, body weight, sex, administration of propofol (yes or no), dosage of propofol administered, type of inhalant anesthetic used, type of feeding tube (esophagostomy tube, PEG tube, or gastrotomy tube), duration of anesthesia, duration of surgery, number of hours in the ICU after placement of feeding tube, administration of blood products (whole blood, plasma, or packed RBCs), and survival (discharge from ICU). Cats were separated into 2 groups on the basis of those that received propofol (propofol group) and those that did not receive propofol (nonpropofol group).

Statistical analysis—Because of skewed data on box-and-whiskers plots, nonparametric analysis was used. Data were reported as median and range. Associations with propofol use were tested by use of the Fisher exact test for dichotomous variables and by use of the Wilcoxon rank sum test for continuous variables. Spearman rank correlation was used to test whether the dosage of propofol was associated with amount of time in the ICU. A Kaplan-Meier survival curve and the logrank test were used to test whether propofol use affected the interval until discharge from the ICU; data were censored on the basis of cats that died or were euthanatized. All tests were interpreted as 2-sided, with significance defined as values of P ≤ 0.05. Because of concerns about potential iatrogenically induced morbidity and fatalities (eg, improper placement of a feeding tube), adjustments were not made for multiple comparisons.

Results

Data from 44 cats were included in the analysis. There were 21 females and 23 males. Age ranged from 3 to 15 years (median, 8 years), and body weight ranged from 1.8 to 9.0 kg (4.0 to 19.8 lb), with a median of 4.8 kg (10.6 lb). Twenty-seven cats received propofol for anesthetic induction. Induction agents administered to the other 17 cats included ketamine (n = 2 cats), ketamine plus midazolam (3), ketamine plus diazepam (4), etomidate (1), isoflurane via face mask (5), and sevoflurane via face mask (2).

Anesthesia was maintained by administration of isoflurane in oxygen in 42 cats, whereas anesthesia was maintained by administration of sevoflurane in oxygen in the other 2 cats. Placement of a feeding tube involved esophagostomy tubes in 12 cats (8 in the propofol group and 4 in the nonpropofol group), PEG tubes in 23 cats (16 in the propofol group and 7 in the nonpropofol group), and gastrotomy tubes in 9 cats (3 in the propofol group and 6 in the nonpropofol group).

Groups did not differ with regard to age or body weight (Table 1). Similarly, groups did not differ with regard to survival (Table 2). Dosage of propofol in cats that received blood products ranged from 4.0 to 7.0 mg/kg (1.8 to 3.2 mg/lb), with a median of 4.5 mg/kg (2.0 mg/lb), which did not differ significantly (P = 0.96) from the dosage for cats that did not receive blood products (range, 2.2 to 28.8 mg/kg [1.0 to 13.1 mg/lb]; median, 5.0 mg/kg [2.3 mg/lb]). Number of hours in the ICU also was not significantly (P = 0.24) correlated with the dosage of propofol (Spearman rank correlation, r = −0.24; n = 26 cats). Cats that survived received dosages of propofol that were significantly (P = 0.012) higher, compared with dosages for nonsurvivors. Dosages for the 24 survivors in the propofol group ranged from 2.2 to 28.8 mg/kg (median, 5.0 mg/kg; first quartile, 4.0 mg/kg), whereas the dosage for each of the 2 nonsurvivors in the propofol group was 2.5 and 3.1 mg/kg (1.1 and 1.4 mg/lb), respectively.

Although there was no significant association between use of propofol and survival, there was a significant (P = 0.034) association between number of hours in the ICU and survival (median, 117 hours for survivors and 48 hours for nonsurvivors). Number of hours until discharge from the ICU did not differ significantly (P = 0.94; log-rank test in the survival analysis) between the propofol and nonpropofol groups.

Seven of 44 (16%) cats did not survive until discharge from the ICU. Five cats (1 in the propofol group and 4 in the nonpropofol group) died of natural causes. Two cats (1 in each group) were euthanatized.

Discussion

In the study reported here, we did not obtain any evidence that the use of propofol increased morbidity or fatalities in cats with hepatic lipidosis. Propofol is a commonly used anesthetic because of its favorable pharmacodynamic and pharmacokinetic characteristics. Although propofol typically is metabolized via glucuronide conjugation in the liver, there is also important extrahepatic metabolism because clearance exceeds hepatic blood flow.¹ For example, in humans with cirrhosis and in pigs whose livers were removed during transplantation surgeries, the pharmacokinetics of propofol are virtually unchanged.^3,6 In cats and sheep, pulmonary uptake contributes substantially to propofol clearance,^2,7 and renal clearance is evident in humans.⁷

Hepatic lipidosis is a common syndrome associated with poor nutrition. It is a condition in which large vacuoles of lipid accumulate in hepatocytes; however, the condition is reversible. In cats, hepatic lipidosis typically is associated with inappetence and can be primary or secondary to other disease processes that lead to inappetence.⁵ Treatment of animals with hepatic lipidosis centers on reinstitution of nutritional components, which is usually accomplished via a feeding tube. Considering the pharmacokinetic characteristics of propofol (particularly in patients with liver disease or dysfunction), propofol is commonly chosen as the induction agent for cats with hepatic liposis that are anesthetized for placement of a feeding tube.

Concern has been raised about the administration of propofol to cats with hepatic lipidosis. This concern is based on the fact that cats have low concentrations of glucuronyl transferase and thus the clearance of propofol could be slowed. Additionally, propofol increases oxidative stress for feline RBCs and therefore can cause formation of Heinz bodies when administered repeatedly via constant rate infusion for several consecutive days.⁶ Although repeated short-duration anesthesia is not associated with adverse hematologic changes in cats that do not have hepatic disease,⁸ some clinicians have argued that cats with hepatic disease are at greater risk of hemolytic consequences.^a Because the study reported here was a retrospective case series, it was not possible to evaluate hemolysis on the basis of changes in the PCV. In many of the cats, PCV was not routinely determined and fluid administration was not standardized. An estimation of clinical hemolysis was indirectly assessed by evaluating the need for blood products and the number of hours spent in the ICU. Our assumption was that cats that had clinically relevant hemolysis would require more blood products and spend more time in the ICU. However, the propofol group did not require additional blood products and did not spend additional time in the ICU, compared with results for the nonpropofol group. However, it is possible that some cats that would have benefited from administration of a blood product did not receive it because of clinician choice or owner reluctance.

Similarly, records analysis did not allow us to assess generalized morbidity. Generalized morbidity was indirectly assessed by analysis of the number of hours spent in the ICU. A limitation of this assessment is that the number of hours spent in the ICU could have been affected by variables other than morbidity. Those variables could have included the time of day that a cat was admitted or the particular clinician who had primary responsibility for the care of each cat. However, there was no difference in the number of hours spent in the ICU for cats in the propofol or nonpropofol groups, and it is difficult to imagine that the time of day a cat was admitted to the ICU or the clinician with primary responsibility for care would have been associated with the decision to use propofol.

Survival was a more direct assessment. Of the 44 cats included in the study, only 7 died (2/27 in the propofol group and 5/17 in the nonpropofol group). Although there was not a significant difference between groups and relatively few cats died, the risk of death was negatively associated with the use of propofol (ie, the risk of death was lower for cats that were administered propofol). Three types of feeding tubes were used (esophagostomy tubes, gastrotomy tubes via surgery, and PEG tubes). We believed it was possible that the increased invasiveness of a surgical approach could have contributed to increases in morbidity and fatalities; however, there was no significant difference in duration of anesthesia between cats in which a gastrotomy tube was placed and the other cats, and type of tube was not associated with propofol administration.

Selection of cats for inclusion required that they did not have concurrent disease. This would have biased the selection for cats that had primary rather than secondary hepatic lipidosis. It is possible that these cats were not as systemically ill as other cats with secondary hepatic lipidosis; morbidity and survival may differ in cats with systemic illness. Similarly, the impression that cats with hepatic lipidosis do worse after receiving propofol^a may have more to do with the severity of illness in the cats with hepatic lipidosis referred to veterinary medical teaching hospitals. It is likely that cats with mild or moderate hepatic lipidosis are being treated in general or secondary practices and that only the cats with the most severe hepatic lipidosis are referred to tertiary care facilities, such as veterinary teaching hospitals.

Data were collected from medical records of cats treated during a 9-year period. It is possible that the standard of care changed during that time period; however, it was beyond the scope of this study to assess such changes. We conclude that the use of propofol to induce anesthesia in cats with primary hepatic lipidosis for the purpose of placing a feeding tube does not increase morbidity or fatalities.