Pharmacology of drugs used in autoimmune dermatopathies in cats and dogs: A narrative review

Mechanism of action

Glucocorticoids are lipophilic and diffuse easily through the cell membranes. Once in the cell cytosol, GCs bind to glucocorticoid receptors (GR) and translocate into the nucleus as GC–GR complexes via binding to proteins known as importins.^{4, 5} In the nucleus, the GR interacts with DNA and proteins to alter gene expression (Figure 1).

The genomic effects of GCs occur via several mechanisms: direct binding of the ligand-bound GR to DNA via positive glucocorticoid response elements (+GREs) or negative GREs (nGREs) and physical interaction with another transcription factor without direct contact with DNA (called ‘tethering’).^{3, 5} Direct binding to +GREs induces transcription of genes that have anti-inflammatory and immunomodulatory properties such as annexin-A1 (ANAX1, also known as lipocortin 1), GC-induced leucine zipper (GLIZ) and mitogen-activated protein kinase phosphatase 1 (MPK1).⁴ By contrast, binding to nGREs inhibits the transcription of genes such as those associated with corticotropin-releasing hormone, melanocyte-stimulating hormone and β-endorphin.⁵

Tethering interferes with NF-κB and activator protein-1 (AP-1)—two transcriptional activators that produce a repertoire of inflammatory cytokines, chemokines and adhesion molecules involved in most, if not all, of the inflammatory and autoimmune dermatoses. It also interferes with key proinflammatory transcription factors from members of the signal transducer and activator (STAT) and nuclear factor of activated T cells (NFAT).³ Finally, GR may also bind to both DNA and protein (composite GRE binding), resulting in the aforementioned anti-inflammatory and immunosuppressive effects.³

Our understanding of the effects of GCs on B cells and antibody production remain incomplete. It has been suggested that GCs have effects on B-cell selection as B cells express the GR throughout development and immature B cells are more sensitive to apoptosis.³ Chronic use of GC is also associated with the inhibition of antibody production in B cells.⁶

Glucocorticoids can also exert their effects via a nongenomic mechanism that occurs more rapidly (within minutes) than the genomic mechanism, which may take hours to days. In the nongenomic pathway, interaction of GCs with membrane-specific GR, cytosolic GR (resulting in the release of a variety of proteins without the need to translocate into the nucleus) and nonspecific interactions with cell membranes leads to alteration of transmembrane currents, signal transduction and intracellular calcium levels^{5, 7}—all of which have anti-inflammatory effects.

Pharmacokinetics and pharmacodynamics

The pharmacokinetics (PK) of GCs in dogs and cats are sparsely reported, possibly because the effects of GCs do not necessarily correlate to the time they are detectable in plasma.⁸ Doses are therefore often derived from human use, or based on the pharmacodynamic (clinical) response in the individual animal. Current literature on prednisone, prednisolone and dexamethasone is summarised below, with an emphasis on their use as immunosuppressive agents. There are no relevant publications on triamcinolone PK currently in the literature.

Prednisone is used clinically in dogs owing to their ability to rapidly absorb the drug and efficiently metabolise it to prednisolone. Following oral administration of prednisone, prednisolone concentrations in the plasma of healthy dogs are approximately sixfold higher than the concentrations of prednisone.⁹ Over a dose range of 0.5–4.0 mg/kg, the kinetics of prednisone and prednisolone were not linear based on a disproportionate increase in drug exposure at the 1 mg/kg dose using the area under the curve (AUC), which is a measure of overall drug exposure. Relatively proportional increases in maximum concentrations were seen, and the half-life remained consistent. This finding is consistent with data in other species and is thought to be due to concentration-dependent protein binding.

In man and laboratory animals, prednisone is converted to prednisolone in the liver via the hepatic enzyme 11-β-hydroxysteroid dehydrogenase type 1.¹⁰ Studies in cats show that this enzyme, while present to a certain degree in the liver, is not capable of dehydrogenase activity when evaluated in the context of the ability to convert cortisol to cortisone.¹¹ This may explain the low conversion of prednisone to prednisolone in cats. Following administration of prednisone, only approximately 20% of drug is converted to prednisolone. The prednisolone AUC following administration of prednisolone is reported as 3230.6 ng/mL/h versus 672.63 ng/mL/h following administration of prednisone.¹² These differences may represent an inability to convert the drug to prednisolone in the liver, yet could also be a result of lower oral absorption of prednisone. Because of this, prednisolone is recommended for use in cats over prednisone.

Prednisolone has excellent oral absorption in both dogs and cats and can be used for long-term administration in either species.^{12, 13} As discussed, it is preferred over prednisone in cats, and should be used in dogs with altered hepatic function, where metabolism of prednisone to prednisolone may be diminished. In over-conditioned cats [body condition score (BCS) of 4–5 of 5], plasma concentrations of prednisolone were, on average, twofold higher than the concentrations in normal cats (BCS 3–3.5 of 5), suggesting that the drug may be better dosed on lean body mass.¹⁴

Data on dexamethasone PK in dogs and cats are even scarcer. Dogs administered 1 mg/kg of dexamethasone in alcohol (intravenously) or as isonicotinate (intramuscularly) had drug detectable for 10 h post-administration.¹⁵ Using a very sensitive assay, dexamethasone was detected in the plasma of Greyhound dogs for ≤96 h post-administration of approximately 0.05 mg/kg dexamethasone sodium.¹⁶ Despite the long detection time, dexamethasone suppression of endogenous cortisol production was not significant after 24 h. The PK of oral dexamethasone in dogs is not published. There are studies available that report the PK of oral dexamethasone in cats following a single dose of injectable dexamethasone sodium phosphate administered at 0.05 or 0.2 mg/kg.^{17, 18} Oral absorption is adequate, although variable in cats. Willis-Goulet et al. (2003) also examined transdermal absorption of dexamethasone in cats. Following transdermal administration, plasma concentrations were low to undetectable; this route should not be used.¹⁸

In general, immunosuppressive doses of GCs are higher and require more frequent administration than anti-inflammatory doses. For example, in dogs, prednisolone plasma concentrations necessary to suppress cortisol, neutrophil and lymphocyte responses are 0.04, 10 and 22.5 ng/mL, respectively.¹⁹ At 1 mg/kg prednisolone i.v., plasma neutrophil counts increased and lymphocyte counts decreased, and had returned to baseline within the suggested 24-hour dosing interval.

Species differences in responsiveness to GCs also need to be taken into account. Cats are less responsive to the anti-inflammatory, immunosuppressive and adverse effects of GCs compared to dogs.²⁰ This is due to a reduced number of GR in the skin and liver of cats compared to dogs.²¹ Those receptors present are also lower-affinity receptors in cats. Therefore, doses of GCs used in cats are higher than those used in dogs.

Glucocorticoid resistance (i.e. absence of clinical response to optimal dosage of GCs) has been reported in humans with conditions such as asthma,²² nephrotic syndrome²³ and ulcerative colitis.²⁴ In animals, GC resistance has been recognised in several diseases such as canine mast cell tumour,²⁵ feline lymphoma²⁶ and canine protein-losing enteropathy.²⁷ The exact pathomechanism of GC resistance in animals is not well-understood, yet low expression of GR and alterations in the metabolic pathways have been suggested.^{28, 29} Although GC resistance in feline and canine autoimmune dermatoses has not been reported, the possibility exists, especially if there is poor response to GC therapy during the induction phase.

Use for autoimmune dermatopathies in dogs and cats

Immunosuppressive doses of systemic GCs, either as monotherapy or in combination with other immunosuppressants/immunomodulators, are the mainstay and often first-line therapy for many autoimmune dermatoses in cats and dogs. GC is considered as the standard first-line treatment for the induction phase of therapy in all untreated patients with active disease, except for those patients for whom GCs are contraindicated. Pemphigus foliaceus, autoimmune subepidermal blistering dermatoses (AISBD) [e.g. mucous membrane pemphigoid (MMP) and epidermolysis bullosa acquisita (EBA)], cutaneous lupus erythematosus (CLE) and pemphigus vulgaris (PV) are examples of autoimmune dermatoses that are treated with systemic GCs. Dosing regimens vary and are likely need to be tailored to individual patients. In one comparative study, dogs with PF treated with oral GC pulse therapy (10 mg/kg once daily for three days, consecutively, followed by a dose of <2 mg/kg/day between pulses) had higher proportions of clinical remission (CR) during the first three months and a lower average maximal oral GC dosage given between pulses.³⁰ The benefit of pulse therapy over conventional daily GC dosing was not apparent in a small case series of 12 cats with PF.³¹ In a critically appraised topic on the dosage of prednisone or prednisolone for the treatment of canine PF, there was no compelling evidence that oral prednisone or prednisolone at the dose of 2 mg/kg/day was effective in induction and/or maintenance of CR, and the authors suggested that a higher dose is more likely to be associated with CR.³² Likewise, high doses of oral GCs (≥3 mg/kg/day for dogs and ≤3 mg/kg twice daily for cats)—often in combination with another immunosuppressant—have been recommended for the treatment of canine and feline PV,³³ and canine uveodermatological syndrome (UDS).³⁴ In cats with PF, GC monotherapy was the most common treatment administered at the time of CR.³¹

Adverse effects

Owing to the wide distribution of GR in all nucleated cells, the chronic use of GC results in many systemic adverse effects. These include iatrogenic hyperadrenocorticism, gastrointestinal ulceration, cutaneous atrophy, diabetes mellitus resulting from insulin resistance, opportunistic infections and delayed wound healing. Chronic dosing of GCs in cats can result in different dermatological adverse effects to those in dogs; cats tend to develop extreme thinning of the skin with possible tearing and curling of the pinna caused by attenuation of the cartilage, whereas dogs uniquely develop calcinosis cutis. Long-acting methylprednisolone acetate has been associated with the induction of congestive heart failure in cats.³⁵ This is believed to be the consequence of a shift in fluids resulting in an increased plasma volume secondary to glucocorticoid-induced hyperglycaemia; it is noteworthy that this conclusion was from a study that was conducted over 24 days and, therefore, the aforementioned pathomechanism could be a short-term adverse effect.³⁶ Although the anti-inflammatory dose of the intermediate-acting steroid prednisolone (1–2 mg/kg/day, p.o.) given to healthy cats with allergic dermatitis for 14 days did not result in significant haemodynamic and echocardiographic changes,³⁷ these results should not be extrapolated to cats requiring immunosuppressive doses of GC for the long-term treatment of autoimmune dermatoses.

Adverse effects of GCs are related to both the dose used and the duration of therapy. In humans, the cumulative dose administered is often calculated to determine the risk for adverse effects developing. One study of patients being treated with steroids showed that the risk for complications developing increased by 3%–8% for each 1 g of cumulative steroid dosed (most frequently prednisone or prednisolone).³⁸ To the best of the author's knowledge, similar data are not available in veterinary species, yet chronicity of dosing should be considered when determining risk for adverse effects. One strategy to potentially decrease risk for adverse effects is to dose on a mg/m² basis, rather than a mg/kg basis. Large-breed dogs had higher C_max and AUC when dosed with prednisolone at 2 mg/kg compared to 40 mg/m², and to small dogs dosed at 2 mg/kg.³⁹ Optimal dosing regimens still require further study.

Mechanism of action

Ciclosporin A is classified as a calcineurin inhibitor; its primary immunosuppressive effect is the inhibition of T-lymphocyte function. Calcineurin is an intracellular protein that activates gene transcription factors by dephosphorylation. When an antigen (e.g. self-antigen) binds to the T cell receptor (TCR), it triggers calcium ion release from the endoplasmic reticulum. The increase of intracellular calcium ions causes the activation of several downstream signalling effectors such as mitogen-activated protein (MAP) kinases, protein kinase C (PKC) and calcineurin.⁴⁰ Activated calcineurin dephosphorylates nuclear factor of activated T cells (NFAT), thereby allowing its translocation into the nucleus, which subsequently upregulates transcription of genes important for innate and adaptive immunity such as interleukin (IL)-2, IL-4, tumour necrosis factor (TNF)-α and TNF-γ.⁴¹ Interleukin-2 in particular, is a potent T-lymphocyte stimulator: it induces proliferation and differentiation of naïve CD8+ T cells into effector T cells by promoting secretion of granzyme B and perforins; it stimulates the proinflammatory activity of T helper (Th)-dependent B cells, antigen-presenting cells, mast cells, basophils and eosinophils; and it influences the differentiation of CD4⁺ T cells into Th1 or Th2, and subsequently affects the proliferation and differentiation of natural killer (NK) cells and B cells.^42-45

Ciclosporin A acts by binding to intracellular cyclophilin, which creates a complex that has a high affinity for calcineurin.⁴¹ The binding of the ciclosporin–cyclophilin complex with calcineurin inhibits its function of dephosphorylating NFAT and thus prevents the translocation of NFAT into the nucleus. Without NFAT in the nucleus, the aforementioned inflammatory cytokines are not transcribed and full T-lymphocyte activation is impaired (Figure 1).

Pharmacokinetics and pharmacodynamics

The pharmacokinetics and pharmacodynamics of CsA have been explored extensively in dogs,⁴¹ and to a lesser degree, cats. Formulation is important and the FDA-approved formulations for dogs and cats should be used. The use of generic human formulations in dogs has been shown to result in a threefold variability in drug absorption between formulations⁴¹; likewise compounded formulations are not recommended. With a microemulsified formulation (Atopica), the bioavailability of CsA in cats and dogs when administered per os was reported to be 25%–29%⁴⁶ and 35%,⁴⁷ respectively. An oral solution (Cyclavance) is also available and has been shown to be bioequivalent to Atopica in dogs with a relative bioequivalence of approximately 101%.⁴⁸ Administration of food along with CsA reduces the mean bioavailability by 22% in dogs,⁴⁹ although this did not affect the clinical outcome in 15 dogs treated for AD.⁵⁰ However, this result cannot be extrapolated when treating cats or dogs with autoimmune dermatoses as no data are currently available. Other routes of administration of CsA include transdermal and subcutaneous. A study involving six healthy cats reported that the absorption of transdermal CsA in the majority of the cats is poor.⁵¹ Median concentrations in that study following seven days of oral administration were 2208 ng/mL at two hours post-dosing, compared to 37 ng/mL at two hours after transdermal applications of equivalent doses.

Ciclosporin is primarily metabolised in the liver by the cytochrome P450 3A (CPY3A) family of metabolising enzymes. Several drugs that also are metabolised by this enzyme will affect the blood concentration of CsA if given concurrently. Of these drugs, azole antifungals (e.g. ketoconazole and fluconazole) are often used along with CsA (especially in large dogs) because they decrease the metabolism of CsA, and therefore increase blood and skin CsA concentrations.⁵² This enables a CsA dose reduction of 50%–70%,⁴⁵ which reduces the overall cost of therapy. The interaction of CsA with other drugs is reviewed elsewhere.^{41, 45} Ciclosporin is also a substrate for p-glycoprotein efflux pumps, which may result in other significant drug–drug interactions. This also implies that caution must be used when dosing CsA in dogs with the MDR1 (ABCB1-1Δ) mutation. Animals heterozygous for this mutation may develop excessive immunosuppression at lower-than-expected doses, potentially resulting in severe infections.⁵³ Because a similar mutation has been identified in cats,⁵⁴ this caution may apply to affected feline patients as well.

Disease states also may affect CsA PK. In an experimental model of diabetic dogs, overall drug exposure (as measured by AUC) was decreased by 52%, clearance was significantly increased and, subsequently, half-life was significantly decreased (9.32 h vs. 22.56 h) compared to healthy dogs.⁵⁵ The mechanism for this is not fully understood, although it was speculated to be caused by increased clearance secondary to hyperglycaemia or alterations in the lipid profile of these dogs. Similar changes in PK were reported in human diabetic transplant patients placed on CsA,⁵⁶ and it would be assumed that a similar phenomenon be seen in diabetic cats.

In order to determine the effectiveness of CsA in dogs, there are two types of tests; therapeutic drug monitoring (TDM) and quantitative reverse transcription polymerase chain reaction (qRT-PCR). However, there is a poor correlation between CsA blood concentration and the clinical response in dogs with AD,⁴⁹ and therefore TDM should only be interpreted in conjunction with clinical assessment of skin lesions and signs. Although data on this correlation in dogs or cats with autoimmune dermatoses are lacking, it is reasonable to suggest that clinical response plus TDM is superior to TDM alone when using oral CsA as a treatment. An alternative way to determine the effect of CsA dosing in a particular animal is through a pharmacodynamic assay that measures cytokine (IL-2) gene expression using qRT-PCR technology to detect the percentage suppression in the animal.⁵⁷

Use for autoimmune dermatopathies in cats and dogs

Ciclosporin A is used to treat many autoimmune dermatoses in cats and dogs; these include, and are not limited to, PF, CLE, PV, UDS, perianal fistula, sebaceous adenitis and ischaemic dermatopathies.^{31, 33, 34, 58-60} The efficacy of oral CsA for the treatment of generalised discoid and vesicular variants of CLE has been reported in a comprehensive review.⁵⁸ In a recent retrospective study, CR of PF was achieved in nine of 11 dogs treated with oral CsA along with oral GC; of these, in five dogs, the disease was maintained in CR with tapered oral CsA monotherapy. The authors of that study proposed that the efficacy of CsA in these cases may be associated with inhibition of B-cell activation (via inhibition of Th cell function), reduction in metalloproteinase-9 expression and blocking of the c-Jun N-terminal kinase (JNK) and p38 signalling pathways that are involved in the pathogenesis of pemphigus.⁶¹ Ciclosporin A may be effective as monotherapy for cell-mediated autoimmune dermatoses such as chronic cutaneous lupus erythematosus and sebaceous adenitis as studies showed that they are associated with cell-mediated autoimmunity involving high expression of T lymphocytes in the epidermis and dermis,⁵⁸ and sebaceous glands,⁶² respectively. However, for canine and feline PF, the usage of CsA as the sole immunosuppressant for maintenance therapy may not be sufficient for all patients, yet it is valuable as a steroid-sparing agent. Tables 1 and 2 summarise the response to CsA in selected autoimmune dermatoses in cats and dogs. Clinicians need to be aware that treatment of autoimmune dermatoses with CsA is considered an extra-label use in cats and dogs.

Owing to the difference in the MoAs, the combination of CsA and azathioprine (AZA) as immunosuppressive agents has been used successfully to treat dogs diagnosed with PF,^{64, 65} PF with concurrent immune-mediated thrombocytopaenia⁷⁶ and idiopathic aplastic pancytopenia.⁷⁷ Because CsA and AZA target different pathways, the risk of myelosuppression from the combination can be expected to be no higher than that from AZA monotherapy, yet there are still no data available to make this conclusion and severe immunosuppression may still occur. Therefore, administration of these combinations should be carefully considered.

Adverse effects

The most common adverse effect reported with CsA is gastrointestinal (GI) upset, with vomiting (28%) and diarrhoea (14%) the most frequently reported clinical signs in dogs.⁷⁸ Gastrointestinal upset is usually transient and resolves with dose reduction. Anecdotally, storing CsA capsules (Atopica) in the freezer for 30–60 min before oral administration reduces the incidence of vomiting,⁷⁹ and the author has found this beneficial in preventing vomiting in some dogs. When stored in a –20°C freezer for one month, the stability and absorption of CsA in dogs is not impacted.⁸⁰ Other adverse effects include gingival hyperplasia, cutaneous papillomatosis and opportunistic infections. Opportunistic bacterial (Nocardia spp., Burkholderia cepacian complex)^81-83 and fungal (Alternaria spp., Curvularia spp. and Aspergillus spp.)^84-86 infections have been reported in dogs treated with CsA, particularly when used in combination with GC.

Similar to dogs, the most common adverse effects of oral CsA in cats are diarrhoea and vomiting.⁸⁷ In cats that were previously infected with feline herpesvirus-1 (FHV-1), administration of oral CsA at the dose of 7 mg/kg/day for 42 days could result in reactivation of FHV-1 yet the clinical signs were mild and self-limiting.⁸⁸ Acute fatal toxoplasmosis secondary to CsA therapy has been reported in cats^{89, 90} yet is still considered rare. In a study by Lappin et al.,⁹¹ cats receiving oral CsA 42 days after experimental infection with Toxoplasma gondii did not show reactivation of toxoplasmosis and oocyst shedding. It is unlikely that toxoplasma-naïve cats would develop clinical toxoplasmosis induced by oral CsA at 7 mg/kg/day, yet measures to reduce the risk of exposure (e.g. prevent hunting and consuming raw meat) should be emphasised to cat owners. Finally, subcutaneous administration of CsA (50 mg/mL injection USP; Sandimmune, Novartis) was administered to 11 cats with nonseasonal allergic dermatitis; although all cats showed a significant decrease in Feline Dermatitis Extent and Severity Index (FeDESI) and pruritus Visual Analog Scale (PVAS) scores, five cats were withdrawn from the study; reasons for withdrawal include behavioural changes and owner-perceived lack of efficacy (n = 1), injection site adverse reaction (n = 2) and owner discomfort in the administration of the subcutaneous injections (n = 2).⁹² Cats receiving chronic, high-dose CsA following renal transplants had a higher incidence of neoplasia compared to control populations of cats that did not receive transplants.^93-95

Azathioprine

Azathioprine is a drug that was first used to reduce the risk of organ transplant rejection.⁹⁶ It is a pro-drug of 6-mercaptopurine (6-MP) that exerts its immunosuppressive effects by interfering with nucleotide synthesis. In addition to its use as a drug to prevent transplant rejection, AZA is used to treat Crohn's disease, ulcerative colitis and AD in humans; the latter as an extra-label use for moderate to severe AD that is nonresponsive to CsA.⁹⁷ In dogs, a pilot study on the usage of AZA for canine AD concluded that this drug has insufficient efficacy and carries too high a risk for adverse effects (compared to CsA) to be recommended as a treatment.⁹⁸

Mechanism of action

After oral ingestion, AZA is absorbed in the intestinal tract. In the intestinal wall, liver and red blood cells (RBCs), AZA is converted to 6-MP.⁹⁹ Thereafter, 6-MP undergoes further conversion via three different metabolic pathways associated with three different enzymes: xanthine oxidase (XO), thiopurine-S-methyltransferase (TPMT) and hypoxanthine-guanine phosphoribosyltransferase (HPRT).^{96, 99} The metabolic pathways associated with XO and TPMT result in the formation of 6-thiouric acid and 6-merthymecaptopurine (6-MMP), respectively. These two metabolites have minimal to no immunosuppressive effects.^{97, 99} However, the metabolic pathway involving HPRT converts 6-MP to 6-thioguanine nucleotide (6-TGN) which has cytotoxic effects (Figure 2).

6-TGNs can be considered ‘false’ nucleotides (e.g. nonfunctioning purines). During the normal cell cycle, in the S phase, DNA is synthesised from the pairing of nucleotides. The generation of 6-TGNs ‘provides’ a pool of nonfunctioning purine, which when incorporated into DNA, results in mutation and subsequent cessation of the cell cycle. This effect is most profound in cells that are actively dividing such as lymphocytes (B and T cells) and thrombocytes. Amidotransferase enzymes and purine ribonucleotide interconversion are also inhibited by 6-TGNs, therefore reducing the formation of purine nucleotides.⁹⁹ Finally, AZA and its toxic (active) metabolites affect T cell migration and adhesion, and reduce survival and proliferation of T cells through inhibition of RAS-related C3 botulinum toxin substrate-1 (RAC1) and/or B-cell lymphoma-extra large (BCL-XL).⁹⁹ The anti-inflammatory effects of the metabolites from AZA also expand to nonimmune cells such as endothelial cells. 6-MP is also shown to inhibit the function of RAC1, which is important for the formation of ICAM-1 and VCAM-1.¹⁰⁰ As such, leucocyte adhesion to endothelial cells is impaired.

Because AZA is a cytotoxic drug, it should be used with caution with other alkylating agents that interfere with DNA synthesis (e.g. cyclophosphamide) as it may lead to profound myelosuppression. There is some evidence that concurrent use of GC and AZA may increase the risk of acute pancreatitis.^{101, 102} Concurrent administration of allopurinol with AZA is not recommended because the antagonism of xanthine oxidase may interfere with the metabolism of AZA.¹⁰³

Pharmacokinetics and pharmacodynamics

Despite a relatively long history of use of AZA in veterinary medicine, there are no published PK studies in either dogs or cats. In humans, the beneficial effects of azathioprine may take weeks to months to occur.

Use for autoimmune dermatopathies in cats and dogs

In dogs, AZA is used to treat several autoimmune skin diseases such as PF, PV, pemphigus vegetans, vesicular cutaneous lupus erythematosus (VCLE) and several variants of autoimmune subepidermal dermatoses (Table 3) owing to its effect on both B and T cells. It is often used as a steroid-sparing agent. Interestingly, AZA along with oral GC is the most common therapy at the time of optimal disease control in canine PF,¹⁰⁹ UDS³⁴ and canine PV.³³ The use of AZA in cats is not recommended owing to the high risk of myelosuppression (see Adverse effects section below).

Adverse effects

In humans, TPMT activity is variable and correlates with clinical outcomes and therefore, the measurement of TPMT activity is used to assess therapeutic efficacy and toxicity.¹²⁴ Low activity of TMPT increases the risk of myelosuppression in people. Likewise, TPMT activity is detected in RBCs of dogs, cats and horses,¹²⁵ and is highly variable in dogs.¹²⁶ One study reported that Giant Schnauzers had much lower TPMT activity, whereas Alaskan Malamutes had higher TPMT activity when compared to the other dog breeds.¹²⁶ Although lower TPMT activity may influence the development of myelosuppression in some dogs, there could be other pathways involved in AZA-induced myelosuppression because some dogs that experience marked myelosuppression did not have a deficiency of TPMT activity.¹²⁷ More importantly, the measurement of TPMT activity in dogs thus far were from RBCs and therefore it is necessary to investigate further correlation (if any) between the activity of canine RBC TPMT and other organs that are more involved in the metabolism of AZA, such as the liver.

Compared to dogs, blood TPMT activity in cats is much lower.¹²⁵ Therefore, it is reasonable to assume that the risk of myelosuppression in cats is significantly higher if treated with AZA and this drug should be avoided in cats.

Hepatotoxicity [defined as a twofold increase in alanine aminotransferase (ALT) above the upper limit of the reference interval] is another adverse effect of AZA in dogs; this may be idiosyncratic or dose-dependent. In one study that included 34 dogs, the prevalence of hepatotoxicity was 15%, with the median time to onset of 14 days (range 13–22 days).¹²⁸ However, in a more recent study, the prevalence of AZA-induced hepatotoxicity was 5% when AZA was administered every other day with tapering doses of GC.¹²⁹ Therefore, it is recommended that liver enzymes are routinely monitored within 2–3 weeks of initiation of AZA therapy.

Mechanism of action

Chlorambucil is classified as an alkylating agent.¹³² In the liver, chlorambucil is converted into its active metabolite, phenylacetic and its cytotoxic effects are based on its ability to alkylate the nucleophilic portion of a DNA molecule through the formation of covalent bonds.¹³⁰ Chlorambucil causes ‘unwanted’ cross-linking of DNA, by forming adducts at the guanine-N7 position.¹³⁰ These ‘unwanted’ cross-linkages can occur within the same DNA strand (i.e. intrastrand cross-links) or with the opposite DNA strand (i.e. interstrand cross-links). This results in a defect of the DNA (e.g. mutation) which will then lead to cell death. Interstrand cross-links are most cytotoxic as they generate double-strand breaks.¹³⁰ Chlorambucil has high activity against lymphoid cells (e.g. B cells) and is considered a slow-acting drug (it may take ≤2 weeks for its therapeutic effects).⁶

Chlorambucil is a cytotoxic agent and therefore will potentiate other immunosuppressive drugs such as vincristine, doxorubicin and cisplastin. This may lead to severe myelosuppression if used together.

Pharmacokinetics and pharmacodynamics

In the literature, there are no PK studies for chlorambucil in dogs, and only one population PK study published for cats. In cats, it is rapidly absorbed from the GI tract, reaching a peak plasma concentration of 170 ng/mL within 15 min of oral administration, although a small secondary peak also was noted at four hours. At a dose of 2 mg/cat (0.28–0.74 mg/kg), the terminal half-life was 1.8 h, with no drug accumulation following two weeks of every-other-day dosing.¹³² The PK of chlorambucil in cats with immune-mediated dermatoses may differ from those reported in this study, however, as the cats in the population PK analysis were being treated for lymphoproliferative malignancies, including GI involvement, which may have affected absorption. The therapeutic concentration of chlorambucil in cats is unknown and additional information is needed on this drug.

Use for autoimmune dermatopathies in cats and dogs

In the cat, similar to CsA, chlorambucil is most often used as an adjunct or steroid-sparing agent for the treatment of PF. Indeed, in two retrospective studies, combination therapy of chlorambucil with GC is often required for maintenance of CR.^{31, 73} Chlorambucil is less frequently used in dogs for the treatment of autoimmune dermatoses. There are two case reports of the usage of chlorambucil in dogs for the treatment of vaccine-induced ischaemic dermatopathy¹³³ and canine eosinophilic granuloma,¹³⁴ both of which required concurrent therapy with ciclosporin and oral prednisolone. Table 4 summarises the usefulness of chlorambucil for selected autoimmune dermatoses in cats and dogs.

Adverse effect

Chlorambucil is generally well-tolerated in cats and dogs, although GI effects such as inappetence, vomiting and diarrhoea may be seen. However, cytotoxic myelosuppression can occur 7–14 days after initiation of treatment.⁶ Other less common adverse effects include reversible myoclonus¹³⁸ and Fanconi syndrome,¹³⁹ both of which were reported only in cats.

Mechanism of action

Mycophenolate mofetil exerts its immunosuppressive effect by inhibiting the formation of guanine nucleotides. Guanine is one of the four nucleotide bases that is needed for the formation of DNA. Guanine nucleotide is generated through two distinct pathways: the de novo pathway and the salvage pathway.^{99, 140} In the de novo pathway, the generation of guanine nucleotides requires the enzyme inosine monophosphate dehydrogenase (IMPDH). Mycophenolate mofetil, via its active metabolite MPA, inhibits the function of IMPDH, thereby reducing the number of guanine nucleotide formations available for DNA replication.⁹⁹ Lymphocytes are particularly affected by this drug because they rely solely on the de novo pathway for guanine nucleotide synthesis.^{99, 140} The result is decreased DNA production and cessation of the proliferation of both activated T and B lymphocytes, which subsequently stop autoantibody production.

Mycophenolate mofetil also affects other nonlymphocytic cells. It has been shown to inhibit the proliferation of fibroblasts, expression of cytokines and co-stimulatory receptors, and various adhesion molecules required for leucocyte chemotaxis and mobilisation to the target cells.¹⁴⁰

Pharmacokinetics and pharmacodynamics

The pharmacokinetics of MPA in dogs has been reported and show that oral absorption is highly variable, with maximum concentrations ranging from 380 to 5040 ng/mL and occurring at 45 min following a dose of 10 mg/kg MMF per os.¹⁴¹ Plasma concentrations of MPA decreased by 80% within eight hours of administration, suggesting a relatively high clearance and short half-life. In humans and, presumably, dogs, the elimination of MPA is mainly as the glucuronide conjugate. As cats are deficient in the glucuronyl transferase enzymes responsible for this reaction, there was concern that the drug could not safely be used in this species. However, there are several reports of MPA PK following MMF administration now published in cats. As with dogs, oral absorption is variable, yet as opposed to other species, the main metabolic route appears to be glucosidation.^{142, 143} Using a two-hour intravenous infusion model, conversion of MMF to MPA appears to be relatively slow in cats, and peak MPA concentrations do not occur until 0.75–1.5 h after stopping the MMF infusion.¹⁴² This is suggested as a potential reason for the high intra- and interindividual variability of MPA PK seen in cats, even with intravenous administration.

Using an ex vivo mitogen stimulation assay in dogs, a proposed target therapeutic concentration of between 600 and 1000 ng/mL would result in a 50% inhibition in T-cell proliferation.¹⁴¹ These effects are not linearly increased at higher plasma MPA concentrations, yet chronic dosing (>14 days) does seem to cause a cumulative effect, with a more prolonged duration of inhibition during the dosing interval. Chronic dosing (every 8–12 h for eight days per os or two-hour infusion i.v. every 12 h for three days) in cats did not alter total peripheral blood mononuclear cell counts, nor alter CD4⁺ or CD8⁺ cell counts.^{143, 144} More information is still needed on the pharmacodynamics of this drug in cats, including the safety and efficacy of long-term dosing. TDM is available at the time of this writing for MMF, MPA and other metabolites using an immunoassay. Recommended sampling times include peak (1–2 h post-dosing) and trough concentrations.

In humans, concurrent administration of CsA, antacid drugs (e.g. omeprazole) and certain antibiotics (e.g. ciprofloxacin and amoxicillin/clavulanic acid) with MMF reduces the oral bioavailability of MMF,^145-147 yet this has not been recognised in animals.

Use for autoimmune dermatopathies in cats and dogs

Mycophenolate mofetil has been used to treat canine PF,^{148, 149} exfoliative cutaneous lupus erythematosus (ECLE),¹⁵⁰ epidermolysis bullosa acquisita,¹⁴⁹ VCLE¹⁴⁹ and MCLE.¹⁵¹ In all but the dog with ECLE,¹⁵⁰ MMF was used as an adjunct therapy with GC,^{148, 149} CsA¹⁴⁹ and GC combined with topical tacrolimus.¹⁵¹ Of these autoimmune dermatoses, by far, most dogs that were treated with MMF were dogs with PF. In an earlier retrospective study, six and two of nine dogs with PF achieved a CR and PR, respectively.¹⁴⁹ However, in a more recent retrospective study, the treatment outcome utilising combination therapy of MMF with GC for canine PF was not as favourable, as CR and PR were only achieved in two and four of eleven dogs, respectively, and none of the eleven dogs were successfully maintained in CR with MMF monotherapy.¹⁴⁸ In the dog with ECLE, monotherapy with MMF resulted in marked improvement within three weeks.¹⁵⁰ The variable response to oral MMF in dogs is due to a narrow therapeutic index and high inter- and intrapharmacokinetic variability.¹⁴¹ This further complicates the use of oral MMF to treat dogs with immune-mediated diseases because of the lack of data to define the therapeutic MFF levels in dogs. There are too few data available to make any conclusion on the usefulness of MMF, compared to other immunosuppressants, for the treatment of various autoimmune dermatoses. However, based on the information available (Table 5), MMF may produce a better clinical outcome as a steroid-sparing agent compared to being utilised as a sole therapy.

Adverse effects

The most common adverse effect reported in dogs receiving MMF is GI upset, particularly diarrhoea. In one recent retrospective study that included 127 dogs, 24% of dogs (n = 31) experienced GI adverse effects, of which diarrhoea was reported in 23 dogs.¹⁵⁵ In most dogs, these GI adverse effects resolve with discontinuation or dose reduction of MMF. In cats, doses of 10 mg/kg given orally every 12 h were well-tolerated, yet 15 mg/kg per os every 12 h caused mild and self-limiting diarrhoea in some cats, while 15 mg/kg given orally every eight hours caused more severe GI upset in all cats treated.¹⁴³

Mechanism of action

Janus kinases are intracellular, nonreceptor tyrosine kinases. They reside in the cytoplasm of cells and are attached to the intracellular/proximal portion of Type I and Type II cytokine receptors.^{156, 157} There are four family members of the JAKs—JAK1, JAK2, JAK3 and TYK2—and they play an important role in the transduction of cytokine-mediated signals through the JAK–signal transducers and activators of transcription (JAK–STAT) pathway. Upon binding of a ligand (e.g. cytokines and growth factors) to the receptor, the JAKs are activated and phosphorylate the intracellular domain of the receptor, thereby creating docking sites for the STAT molecules, which are signalling proteins. Once docked, the STAT molecules are further phosphorylated by the JAKs and then released back into the cytoplasm, where they form dimers with other phosphorylated STAT molecules. The dimerised STAT molecules translocate into the nucleus and bind with the DNA to transcript genes that regulate immunity, inflammation and haematopoiesis (Figure 3).^{156, 157}

Cytokine receptors can be grouped based on which JAKs are associated with the receptor complex. While cytokines involved in allergy, inflammation and pruritus bind receptor complexes that utilise JAK1, the binding of other cytokines or growth factors involved in haematopoiesis (e.g. erythropoietin) and innate immunity (e.g. IL-12) activate receptors which are associated with the pairing of JAK2/JAK2 or JAK2/TYK2.¹⁵⁶

Higher extra-label doses of oclacitinib are associated with immunosuppressive activity. Lymphocyte-enriched cells treated with 10 μM of oclacitinib (equivalent to 3–4 mg/kg twice daily) resulted in reduced secretion of IL-2, IL-15, IL-18 and IFN-γ, which inhibits the proliferation of T cells.¹⁵⁸ Higher doses of oclacitinib also have been shown to induce apoptosis of canine CD4⁺ and CD8⁺ T cells in vitro.¹⁵⁹ In a more recent study, oclacitinib prevented the generation (rather than inducing depletion) of regulatory T cells and the production of IL-10, which are important to maintain immune tolerance.¹⁶⁰ Taken together, higher doses of oclacitinib or doses used concurrently with another immunosuppressive drug can have immunosuppressive effects on the immune system.

Pharmacokinetics and pharmacodynamics

The pharmacokinetics of oclacitinib in the dog have been published.¹⁶¹ It is rapidly absorbed following oral administration with 89% bioavailability, reaching a peak concentration of 259 ng/mL within one hour of administration of label doses (0.4–0.6 mg/kg once daily). There are no significant effects of feeding, sex or breed (Beagles versus cross-breeds) on oral absorption. Importantly, PK were linear over a range of 0.6–3.0 mg/kg, with dose-proportional increases in C_max and AUC being reported. Twice-daily dosing of 0.6 mg/kg or once-daily dosing of 1.8 mg/kg, as used in the treatment of immune-mediated disease noted above, resulted in maximum concentrations of 328 and 1030 ng/mL, respectively, after 21 days of dosing.¹⁶¹ There is one small PK study in cats, which confirmed values similar to those found in the dog, with excellent bioavailability (87%).¹⁶² Absorption and elimination were faster in cats, compared to dogs, however, and the concentrations were more variable.

Use for autoimmune dermatopathies in cats and dogs

The first reported use of oclacitinib for the treatment of autoimmune dermatosis in veterinary medicine was in 2017 when a dog with autoimmune subepidermal blistering disease was successfully treated with oclacitinib monotherapy at the dosage of 0.5 mg/kg twice daily; a relapse occurred when the dose was tapered to once daily and resolved when it was increased to twice daily. Other autoimmune skin diseases that have been treated successfully with oclacitinib include PF in a cat,¹⁶³ a dog with PV,¹⁶⁴ and several variants of CLE (ECLE, MCLE and facial DLE) in seven dogs.¹⁶⁵ Interestingly, in all of these cases, oclacitinib monotherapy (dosages between 0.5 mg/kg twice daily and 1.8 mg/kg once daily in dogs; 1 mg/kg twice daily in a cat) was successful in inducing and maintaining CR; time-to-CR or significant improvement also was relatively short at between one and three weeks. Oclacitinib was also used successfully to treat immune-mediated dermatoses such as hyperkeratotic erythema multiforme,¹⁶⁶ ear tip ulcerative dermatitis¹⁶⁷ and ischaemic dermatopathy¹⁶⁸; the latter required a combination of oclacitinib (0.4–0.7 mg/kg twice daily) with oral GC.

Recent studies in canine CLE involving skin lesions transcriptomes^{150, 169, 170} showed strongly activated IFNαβ signalling via JAK–STAT with upregulation of CXCL10, ISG15 and S100, suggesting that these CLE variants represent a form of interferonopathy. Thus, treatment with oclacitinib could be an effective therapeutic option either as monotherapy or as a steroid-sparing agent. It is important to point out that the usage of oclacitinib for the treatment of autoimmune dermatopathies is considered as extra-label use in animals. Table 6 summarises the response to oclacitinib in dogs with selected autoimmune dermatopathies.

Adverse effects

Owing to the MoA, higher doses of oclacitinib could lead to impaired T-cell proliferation and increased risk of infection. Indeed, during the initial animal safety study, oclacitinib induced papillomas in 12-month-old dogs, and led to the development of bacterial pneumonia and generalized demodicosis in 6-month-old dogs (package insert; Apoquel, Zoetis). In a retrospective study of 53 dogs with AD treated with prolonged twice-daily administration of oclacitinib (0.4–0.6 mg/kg twice daily), pyoderma, gastrointestinal signs and otitis externa were the most common adverse effect reported; only three dogs developed mild neutropaenia.¹⁷² In an feline immunodeficiency virus (FIV)-positive cat, treatment with oclacitinib (1 mg/kg twice daily) for feline atopic skin syndrome resulted in fatal disseminated toxoplasmosis; the clinical signs developed five months after initiation of this therapy.¹⁷³ In cats, doses of 1–2 mg/kg every 12 h for 28 days were shown to be safe and well-tolerated, although the higher doses did result in gastrointestinal signs in two of 10 cats.¹⁷⁴ Long-term studies with a large number of cats have not been performed and therefore, the long-term safety of oclacitinib in cats is as yet unknown. At present, oclacitinib is not licensed for use in cats.

Mechanism of action

BTKi can be classified into two types: reversible inhibitors and irreversible inhibitors. Irreversible BTKi have been associated with more adverse effects as they may bind to off-target sites (e.g. endogenous thiols, which are important to protect cells from ionising radiation^{180, 181}), while reversible BTKi may be safer yet less efficacious, depending on the degree of selectivity, clearance and maintenance of target inhibition.¹⁸² More recently, a new hybrid BTKi that can reversibly bind to BTK in a covalent manner (i.e. reversible covalent BTKi) has been developed with increased potency, prolonged duration of action and fewer off-targets effects (and, thus, fewer adverse effects).¹⁸³ One example of this new hybrid BTKi is rilzabrutinib (PRN1008).¹⁸²

Receptors that activate BTK are antigen receptors that include BCRs, growth factor and cytokines receptors, chemokine receptors and integrins.¹⁸⁴ In B cells, attachment of ligands to BCR leads to activation of BTK, which triggers several downstream signalling cascades, including the PI3K-ALT pathway, PLC, PKC and NFκB, which are important for B-cell survival, proliferation and differentiation to plasma cells that produce autoantibodies.^{175, 184} BTK is also involved in downstream signalling vital for neutrophil recruitment such as macrophage antigen-1 (MAC-1).¹⁸⁵

Altogether, BTK inhibition causes a blockade of different downstream signalling pathways related to the development of autoimmunity and B-cell malignancies (Figure 3).

Pharmacokinetics and pharmacodynamics

The PK/PD of these drugs are not available in dogs and cats at the time of writing.

Use for autoimmune dermatopathies in cats and dogs

The use of BTKi for the treatment of autoimmune dermatoses in veterinary medicine is still in its infancy and off-label. In 2020, two studies involving a total of 13 dogs investigated the efficacy and safety of two BTKis, PRN473 and PRN1008 (rilzabrutinib), for the treatment of PF.^{186, 187} Treatment outcomes were evaluated based on a modified canine version of a validated human pemphigus disease activity index (cPDAI) and classified as good, fair or poor response.¹⁸⁷ At the end of the study period (20 weeks), five, three and two dogs had a good, fair and poor response, respectively. In all dogs, lesions reduction could be seen within 2 weeks. Interestingly, in both studies, there was no sustained depletion of B-cell counts over the 20-week period and the authors commented that this does not mean that the studied BTKi does not induce pan-B-cell deficiency.¹⁸⁷ There were only two dogs where the frequency of administration of BTKi was able to be reduced to every other day without resulting in a relapse.¹⁸⁷

Adverse effects

Adverse effects reported from the study involving dogs with PF treated with PRN473 included immune-mediated polyarthritis (n = 1; Week 12), mast cell tumour (n = 1; Week 4), peripheral lymphadenopathy (n = 2) and pancreatitis (n = 1).¹⁸⁷ Of the four dogs that received PRN1008 for canine PF, only one dog had pyometra, and increased ALT and aspartate aminotransferase.¹⁸⁶ In tolerability studies using an unrelated BTKi (G278), hepatotoxicity was reported in dogs.¹⁸⁸

It is clear that more studies are needed to gather data on the safety profiles of BTKis in dogs, yet it is very important to be aware that the sample sizes for both studies in canine PF were small and it cannot be concluded that all of the adverse effects were a direct result of the BTKis administered. For instance, peripheral lymphadenopathy in two dogs resolved spontaneously while still receiving PRN472,¹⁸⁷ while the one dog that developed pyometra was also an eight-year-old intact female dog.¹⁸⁶