Drug therapy in feline patients has many potential roadblocks. There are relatively few approved drugs or dose optimization studies in cats, and extrapolating dosages from other species can be difficult because of key differences in feline drug metabolism. Many drugs that are designed for humans or dogs need to be reformulated for cats because of tablet size or palatability. Even with drug reformulation, many cats are difficult to medicate orally. One study of cat owners found that about 25% of drug doses could not be administered as prescribed.1

Cats have important differences in drug metabolism compared to humans and dogs, two species from which feline dosages are sometimes extrapolated. It is well-known that cats are deficient in glucuronidation of some xenobiotics; for example, UDP-glucuronosyltranferase (UGT) activity for acetaminophen is 10-fold lower in cats compared to dogs and humans.2 This is due to a nonfunctional feline pseudogene for UGT1A6,³ the UGT isoform that metabolizes acetaminophen and other phenolic compounds in humans. This same enzyme glucuronidates morphine and serotonin,⁴ and contributes to the metabolism of silybin (in milk thistle).⁵ Evolutionary loss of functional UGT1A6 in felids is thought to be a result of dietary specialization as “hypercarnivores,” with minimal exposure to plant-derived phenolic compounds.⁶ Therefore, glucuronidation of certain drugs is deficient in cats (Table 4.1). However, cats are capable of glucuronidating some drugs, such as the angiotensin II inhibitor telmisartan,²⁰ and endogenous compounds, such as thyroxine²⁴ and bilirubin.²⁵

Cats are also deficient in the enzyme thiopurine methyltransferase, which metabolizes the immunosuppressive drug azathioprine. The activity of this enzyme, which can be measured in red blood cells, is 80%–85% lower in cats compared to dogs.26–28 Because of this, cats are especially sensitive to myelosuppression from azathioprine.²⁹ Individual variability in thiopurine methyltransferase among cats (almost 10-fold) can be attributed to genetic polymorphisms in the feline gene, such that there is overlap between some “high activity” cats and some “low activity” dogs.^27,30 However, azathioprine is not recommended for use in any cat, since other immunosuppressants such as prednisolone or cyclosporine have much better safety profiles.

Compared to humans, cats appear to have low activity for aldehyde oxidase,³¹ a hepatic cytosolic phase I enzyme. Aldehyde oxidase bioactivates the antiviral drug famciclovir to penciclovir, an active metabolite that targets herpesviruses. Since cats are deficient in this pathway, a much higher famciclovir dosage is required in cats (90 mg/kg, orally [PO], every 8 hours) to achieve therapeutic concentrations of active penciclovir³² compared to the label dosage for humans (500 mg, or about 7 mg/kg, PO, every 8 hours). Like cats, dogs are also deficient in aldehyde oxidase activities.³¹ For example, conversion of methotrexate to an active metabolite (catalyzed by aldehyde oxidase) is absent in dogs.³³ It is unclear whether this has an adverse effect on methotrexate efficacy in dogs.

Characterization of cytochrome P450 enzymes (CYP) in cats has lagged that in dogs. Orthologs of many human P450s have been identified in cats,^34–37 but substrate ranges for clinically used drugs are mostly unexplored. Cats have low hepatic activity for at least two key P450 enzymes. CYP2B6, which converts diazepam to nordiazepam,¹⁷ is absent in the feline liver;³⁷ instead, diazepam is primarily metabolized to temazepam in cats, and this metabolite is poorly glucuronidated (Table 4.1).¹⁷ CYP2C, which metabolizes fluoxetine, losartan, and piroxicam in humans,³⁸ has low activity in the feline liver.³⁴ However, the drugs that are metabolized by this CYP in cats are not yet characterized. Additional research on feline cytochrome P450s is needed to better understand the potential for drug-drug interactions and inappropriate dosage extrapolations in cats.

The neonatal period in dogs and cats has been defined as the first 4 weeks of life, with the pediatric period defined as up to 12 weeks of age.39 Although drug therapy is common in human neonates, very few pharmacokinetic studies have been performed in newborns and infants. Given that neonatal pharmacology is even less well-studied in cats, specific and valid recommendations are difficult to make. However, there are certain physiologic differences between neonates and adults (based on studies in humans, dogs, rodents, and occasionally in cats) that can help to guide rational drug therapy in these tiny and rapidly changing patients.

Oral drug absorption may differ in newborn kittens compared to adult cats. Immaturity of gastric parietal cells leads to a relatively high gastric pH in neonates; for example, gastric pH is greater than 3.0 though 5 weeks of age in puppies.⁴⁰ High gastric pH may decrease the bioavailability of drugs that require an acid environment for absorption, such as iron supplements, ketoconazole, and itraconazole.^41–43 Fluconazole may be better absorbed in neonates, since its absorption is not affected by gastric pH, at least in humans.⁴⁴ Oral drug absorption can also be impaired during nursing, due to drug binding by cationic milk components such as calcium and magnesium (Fig. 4.1).⁴⁵ For example, the oral bioavailability of enrofloxacin is less than 35% in nursing kittens, compared to 85% by the subcutaneous route.⁴⁶

Hepatic CYP450 activities are low in newborns, but approach and even exceed adult levels by 7 weeks of age, as shown in puppies (there are no equivalent studies in cats).47 This may represent an evolutionary response to a wider variety of dietary chemicals encountered at weaning. Immature CYP450 content in neonates is associated with delayed hepatic clearance of some drugs. For example, lidocaine and theophylline show delayed elimination in puppies less than 1 to 2 weeks old.^48,49 However, by the time most of our patients present for first vaccination, hepatic function has matured.

Newborn kittens have decreased glomerular filtration rate (GFR) until 9 weeks of age, when GFR reaches rates found in adult cats.⁵⁰ Prior to 9 weeks, kittens may be at greater risk for fluid overload due to impaired solute and water excretion, and for toxicity from renally eliminated drugs such as aminoglycosides. Classic early warning signs of nephrotoxicity, such as granular casts, are not consistently observed in neonatal pups given gentamicin, despite the development of renal tubular lesions and impairment of GFR.⁵¹ Aminoglycosides should therefore be avoided whenever possible in these very young patients. In contrast, enrofloxacin, despite its renal excretion, is cleared efficiently in kittens as young as 2 weeks of age⁴⁶ and does not appear to require dose reductions in this age group. Notably, kittens are less susceptible to cartilage damage from fluoroquinolones compared to puppies, with no cartilage lesions reported at three times the label dose for enrofloxacin or orbifloxacin (Freedom of Information summaries, http://www.fda.gov). However, marbofloxacin has a narrower safety margin for cartilage lesions, and is not approved for cats less than 1 year of age.

A geriatric veterinary patient has been defined as having reached 75% of its expected lifespan.52 Aging leads to global effects on kidney function, liver blood flow, body composition, and cardiovascular function that can increase the risk of drug-induced morbidity. Adverse drug reactions are reported to be two to three times higher in elderly human patients compared to younger adults.^53,54 Some of this risk can be attributed to patient confusion and errors in self-dosing; however, pharmacokinetic and pharmacodynamic factors are also involved.

Age-related decreases in renal function may be the most important factor affecting drug dosing in geriatric human patients.⁵³ Even patients without overt azotemia are likely to have decreased GFR associated with aging. This may be compounded by subclinical dehydration, since older patients tend to have decreased total body and interstitial water.⁵³ In geriatric cats, the prevalence of chronic kidney disease is quite high, with approximately 65% of cats more than 15 years old diagnosed with International Renal Interest Society (IRIS) stage 2 or higher chronic kidney disease on random screening.⁵⁵ This may lead to decreased elimination and increased toxicity of renally cleared drugs in older cats. Enrofloxacin has been associated with retinotoxicity in elderly cats at the label dose of 5 mg/kg/day.⁵⁶ Since this ocular toxicity is dose-dependent,⁵⁷ toxicity seen in older cats is likely due to decreased renal drug clearance. While orbifloxacin, marbofloxacin, and pradofloxacin are also cleared by the kidneys, they are less retinotoxic at higher dosages,⁵⁷ and may be safer for geriatric cats.

Aging is associated with modest decreases in liver function, including decreased liver mass and variable reductions in CYP450 function in elderly human patients.⁵⁸ Decreased liver blood flow also occurs with aging and can lead to decreased clearance of certain drugs.⁵⁹ For example, propofol is a “blood flow-limited” drug, and its clearance is diminished in older humans.⁶⁰ Propofol clearance is also decreased in geriatric dogs, with higher plasma drug concentrations and apnea seen in some older dogs given standard dosages.⁶¹ Other drugs that show impaired clearance in elderly human patients, due to liver or renal impairment or other factors, are listed in Box 4.1. Although we do not have comparable studies in cats, these drugs should probably be dosed conservatively in older cats, with education of owners for careful monitoring for adverse effects.

In humans, body composition can also change with age, including decreased total body water and lean muscle mass.54 Cachexia and sarcopenia are important emerging syndromes in small animal medicine.⁶² Older cats may be cachectic, and highly lipophilic drugs, including all anesthetics, should be dosed conservatively in these patients. Conversely, some older cats may be obese. In obese cats, relatively polar drugs with poor fat distribution should be dosed based on estimated ideal (lean) body weight.⁶³ For example, dose reductions of 15% to 20% were calculated for obese cats treated with the polar drug gentamicin, based on differences in pharmacokinetics compared to lean cats.⁶⁴ Lipophilic drugs with high fat distribution, such as propofol and benzodiazepines, can be dosed based on measured body weight in obese patients.⁶⁵

Finally, compensatory physiologic responses are impaired with age. These changes include decreased sensitivity to circulating catecholamines, diminished renal autoregulatory responses, a loss of respiratory muscle mass with decreased ventilatory capacity, and increased susceptibility to hypotension.⁵⁴ These changes lead to a higher risk of complications during treatment with anesthetics or other drugs that affect blood pressure or heart rate. In fact, advancing age is an independent risk factor for postanesthetic complications and mortality in cats.^66,67 For both pharmacokinetic and pharmacodynamics reasons, initial dosages of midazolam, propofol, and fentanyl are reduced by 50% to 75% in geriatric human patients, ⁵⁴ and similar reductions may be wise in geriatric cats.

Renal disease leads to decreased filtration of many drugs and their active metabolites, as well as impaired tubular secretion of drugs such as famotidine and trimethoprim.68 Renal disease is also associated with less obvious effects on drug disposition, including decreased renal drug metabolism and impaired binding of acidic drugs to albumin.⁶⁹ These alterations can each increase the risk of drug toxicity in patients with kidney disease.

In renal disease, dosage reductions are indicated for any drug with a relatively narrow margin of safety that is either primarily eliminated by the kidneys or that has an active metabolite that is eliminated by the kidneys There is little information to guide dosage adjustments in cats with chronic kidney disease. In humans, dose adjustments are typically made when GFR, as measured by creatinine clearance, drops to about 0.7–1.2 mL/kg/min, depending on the drug’s therapeutic index.⁷⁰ Based on the relationship between GFR and serum creatinine in cats,⁷¹ this is equivalent to serum creatinine concentrations of approximately 2.5 to 3.5 mg/dL (221–309 μmol/L). Therefore, in the absence of specific data in cats, it is reasonable to consider dosage adjustments for renally cleared drugs when the serum creatinine reaches this range (Table 4.2). In humans with kidney disease, doses for renally cleared drugs are typically 25% to 75% of the standard daily dosage.⁷⁰

Ampicillin and amoxicillin are renally excreted, but have wide safety margins, so dose adjustments are probably not clinically necessary. Cephalothin can cause lipid peroxidation and nephrotoxicity in animal models and can be nephrotoxic in combination with aminoglycosides in older human patients.72 Therefore, dosage reductions of this cephalosporin may be indicated in veterinary patients with renal disease. For more expensive beta lactam derivatives, such as meropenem, dose adjustments are recommended in humans when creatinine clearance dips below 0.7 mg/mL/kg; prolongation of the dosing interval is recommended.⁷⁸

Aminoglycosides are cleared by the kidneys and are also dose-dependent nephrotoxins; these antibiotics should be avoided in pre-existing renal disease. For patients with renal disease that develop resistant gram-negative infections, other antimicrobials (such fluoroquinolones, cefoxitin, ceftazidime, or piperacillin) should be substituted whenever possible. When aminoglycosides are necessary, rehydration and concurrent fluid therapy (intravenous [IV] or subcutaneous [SC]) are recommended, since hypovolemia is a risk factor for aminoglycoside nephrotoxicity in humans.⁷⁹ In addition, amikacin should be considered over gentamicin, since it is less nephrotoxic in human patients⁸⁰ and may be less nephrotoxic in cats as well.⁸¹ Aminoglycosides are contraindicated in combination with furosemide or nonsteroidal anti-inflammatory drugs (NSAIDs), each of which can exacerbate nephrotoxicity.^79,82

If aminoglycosides are used in human patients with reduced renal function, the dosage is adjusted by extending the dosing interval (e.g., dosing every 48 hours instead of every 24 hours).⁸³ Aminoglycosides are concentration-dependent antimicrobials (i.e., bacterial kill correlates with high peak drug concentrations), but nephrotoxicity correlates with high trough drug concentrations.⁸⁴ Aminoglycoside drug dosages should be adjusted to keep trough serum drug concentrations (right before the next dose) below 2 µg/mL.⁸⁵ Actual measurement of trough drug concentrations is ideal, but results must be available rapidly to be useful in clinical decision making. A practical monitoring approach is to examine fresh urine sediments daily for granular casts, which indicate renal proximal tubular damage and can be seen days before azotemia develops.⁸⁶ The development of granular casts suggests that the drug should be discontinued, unless continued treatment is necessary for a life-threatening infection. Aminoglycoside toxicity is lessened in cats if therapy can be limited to 5 days or less, whenever possible.⁸⁷

Fluoroquinolones are also renally cleared. While they are not overtly nephrotoxic, they can cause dose-dependent retinotoxicity in cats.⁵⁷ Dosage adjustments in renal disease may be particularly important for enrofloxacin, which is more retinotoxic (retinal lesions at four times the label dosage) compared to other veterinary fluoroquinolones in cats.⁵⁷ Although the optimal method for fluoroquinolone dosage adjustment is not established in cats, extending the dosing interval may be most appropriate,⁸⁸ since fluoroquinolones are also concentration-dependent antimicrobials.

Potentiated sulfonamides should be used with caution in azotemic patients, due to decreased renal clearance. Dosage reductions are recommended for the generic drug trimethoprim-sulfamethoxazole in human patients.⁸⁹ Dose reductions are even more important for trimethoprim–sulfadiazine. Sulfadiazine is relatively insoluble and can precipitate as drug crystals in the renal tubules, especially at high concentrations or in acid urine. This has led to hematuria, urolithiasis, and even acute kidney injury in humans.⁷⁷ Because of this, sulfadiazine should be avoided in cats that are dehydrated, on urinary acidifying diets, or have acute kidney injury or chronic kidney disease.

Furosemide is renally cleared and can cause significant dehydration and hypokalemia, which can cause further renal decompensation. Furosemide should be avoided in cats with underlying renal disease, unless there is good rationale (i.e., fulminant congestive heart failure). Cats treated with furosemide should be monitored closely for dehydration, hypokalemia, and worsened azotemia, with routine evaluation of skin turgor, body weight, body condition score, packed cell volume and total protein, serum potassium, and renal indices at each recheck.

Angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers (e.g., telmisartan) are recommended to reduce proteinuria in cats with renal disease (International Renal Interest Society Guidelines; http://iris-kidney.com). Benazepril may be preferable to enalapril in cats with substantial azotemia. Benazepril does not depend solely on renal elimination and does not require dose adjustment in moderately azotemic cats.⁷⁴ It is uncommon for ACE inhibitors to cause systemic hypotension at therapeutic dosages in cats, but these drugs can decrease GFR at high dosages, particularly in combination with dehydration or concurrent furosemide administration. It is therefore important to frequently monitor clinical hydration, blood urea nitrogen (BUN), creatinine, and electrolytes in cats treated with ACE inhibitors; for example, initially after 1 week, after 1 month, and then every 3 months, depending on clinical status.

Nonsteroidal anti-inflammatory drugs can adversely affect GFR in patients with hypovolemia or underlying renal disease, by blocking the elaboration of renal prostaglandins that otherwise auto-regulate renal blood flow.⁹⁰ For analgesia in cats with kidney disease, buprenorphine provides an alternative to NSAIDs, with comparable analgesic efficacy in cats.⁹¹ If an anti-inflammatory effect is needed, NSAIDs should be dosed conservatively, and cats should be monitored carefully. For example, meloxicam has been implicated in episodes of acute kidney injury in cats, leading to a label warning in some countries against chronic use in this species (Metacam; Boehringer Ingelheim). However, at a low chronic dosage of 0.01–0.03 mg/kg/day, meloxicam was well tolerated by older cats with osteoarthritis in two studies, including cats with stable IRIS stage 2 chronic kidney disease.^92,93 This is an off-label indication, however, and requires informed owner consent and periodic monitoring for renal decompensation and gastrointestinal (GI) bleeding. It is important to note that coxibs (cyclooxygenase [COX]-2 selective NSAIDs) have the same potential for adverse renal events as older NSAIDs in humans.⁹⁴ Both COX-1 and COX-2 are expressed constitutively in the kidney (this has also been shown in cats),⁹⁵ and are important for regulating renal blood flow.⁹⁶

In human patients with inflammatory liver diseases without liver failure, drug clearance is not substantially affected.97 With progressive loss of liver function, however, both liver blood flow and hepatic biotransformation pathways become impaired, and drugs that rely on biotransformation can accumulate (Box 4.2). For these drugs, reductions to 25%–50% of regular dosages are recommended for human patients with impaired liver function (Box 4.2).⁹⁸ These same recommendations may apply to cats with hepatic lipidosis, portosystemic shunts, acute hepatotoxicosis, severe cholestasis, and advanced infiltrative liver diseases, although we do not have comparable studies in cats.