Use of Recombinant Human Interferon Alpha-2a in the Management of a Dog With Epitheliotropic Lymphoma

Introduction

Epitheliotropic lymphoma is a well-recognized presentation of canine lymphoma that is characterized by infiltration of the skin by malignant T lymphoid cells.^1–3 Common clinical presentations are exfoliative erythroderma with pruritic erythema and scaling, mucocutaneous erythema, depigmentation or ulceration, and multiple or solitary nodules or plaques.⁴

In people, subclinical or undetectable systemic involvement occurs early in the disease course but does not become clinically apparent until later.⁵ The common clinical progression in people is scaling and pruritus, followed by erythema and ulceration, followed by the development of nodules and plaques. In dogs, any combination of these clinical manifestations may be seen on initial presentation.⁶ Infiltrative oral disease is common in dogs and often manifests as multiple erythematous plaques or nodules involving the gums or lips.⁷ Involvement of extracutaneous organs (such as lymph nodes, liver, spleen, and bone marrow) has also been described in dogs, as has the circulation of malignant T cells (analogous to Sézary’s syndrome).^8–10 Diagnosis is made by demonstration of classic histopathological findings, including Pautrier’s microabscesses, and confirmation of a T-cell immunophenotype.¹¹

Treatment for epitheliotropic lymphoma is palliative rather than curative in both canine and human patients.^4,13 Varieties of treatment protocols, including radiotherapy and systemic or topical chemotherapy, have been used with variable success in dogs.^12,14–19 The numerous treatment recommendations reflect the generally variable and frequently short-lived responses obtained with either single-agent or combination chemotherapy. Therapy generally provides clinical improvement for 1 to 5 months and median survival for up to 11 months.^7,19 Consequently, epitheliotropic lymphoma is considered to have a relatively poor prognosis compared with multicentric lymphoma.^4,20 In human medicine, recombinant human interferon alpha (rhIFN) has been exploited to treat numerous neoplastic conditions, including epitheliotropic lymphoma (mycosis fungoides).¹³ Recombinant human IFN-2a is used in both single-agent and multimodality protocols, and it has been proposed as the most effective single agent for treatment of mycosis fungoides.²¹

Although the immunomodulating and antiproliferative properties of IFN-2a are relatively well established in human medicine, the biological activity of IFN-2a in mycosis fungoides is incompletely understood. Derangements in cytokine expression play an important role in the pathogenesis of epitheliotropic lymphoma, so IFN-2a may act through manipulation of in vivo cytokine activity.²² Interferons are cytokines with both immunomodulating properties and the ability to influence cell proliferation. Interferons inhibit cell-cycle progression by partially blocking the steps from G₀ through the S phase and by inhibiting or potentiating the activity of other cytokines involved in cell-cycle regulation.²³

In people, mycosis fungoides arises from the T-helper 2 (Th2) subset of CD4-positive T cells, which produce interleukin-5 (IL-5), interleukin-4 (IL-4), and other cytokines. The efficacy of IFNmay be mediated via effects on these cytokines.^24,25 Indeed, IFN-mediated suppression of IL-5 activity has been documented in patients with Sézary’s syndrome and eosinophilia.²⁶ In addition, IFN inhibits IL-4 production in normal T cells and Sézary cells.²⁷ Less specifically, IFN-2a increases expression of the class I molecule on lymphocytes, and it has variable effects on both the activity of natural killer (NK) cells and the production of antibodies by B cells.^26,28

In contrast to the situation in people, most cases of epitheliotropic lymphoma in dogs derive from a CD8-positive T-cell lineage.^2,11 Accordingly, IFN-mediated suppression of IL-5 and IL-4 may be less relevant in dogs. The effects on interleukin-2 (IL-2) production and NK cytotoxi-city may play a greater role, given that human IFN has been shown to enhance both IL-2 production and NK cytotoxicity in dogs.²⁹

In veterinary medicine, recombinant interferon has been used primarily for its antiviral activity in the treatment of canine parvovirus.³⁰ Recombinant feline interferon-omega is most commonly used in this situation, and it is licensed for this use in the United Kingdom. In dogs, rhIFN has been used for its immunomodulatory properties in the treatment of immune-mediated or infectious ophthalmological and dermatological conditions.^31,33 Recombinant human interferon (rhIFN-2b) has also been used for the management of feline alimentary lymphoma that has epitheliotropic features.³⁴ In addition, a dosage regimen has been suggested for canine epitheliotropic lymphoma, but clinical response to rhIFN in either cutaneous lymphoma or other lymphoproliferative diseases of dogs has not been reported.³⁵

This case report describes the clinical response of a dog with epitheliotropic lymphoma that was treated with rhIFN-2a^a after relapsing on combination chemotherapy.

Case Report

A vaccinated, 8-year-old, intact male, mixed-breed dog weighing 10.4 kg was referred to the University of Liverpool Small Animal Hospital for evaluation of a cutaneous preputial mass. The mass had appeared 2 weeks before the dog’s presentation at the authors’ clinic. Treatment by the referring veterinary surgeon consisted of dexamethasone^b (0.2 mg/kg subcutaneously [SC] once), prednisolone^c (0.5 mg/kg per os [PO] q 24 hours), marbofloxacin^d (2 mg/kg PO q 24 hours), and clavulanate potentiated amoxicillin^e (12.5 mg/kg PO q 12 hours). The mass decreased in size after initial treatment by the referring veterinarian, but it enlarged again within 7 days. In fact, the mass was larger upon referral than at first presentation. The dog had been otherwise clinically well.

Upon referral, the dog was bright and in good body condition. Gross thickening of the skin was seen along the entire prepuce. Extensive ulceration of the dorsal, lateral, and ventral preputial surfaces, extending to the preputial orifice, was noted. These ulcers were poorly demarcated and had multiple ulcerated, coalescing, plaque-like lesions adjacent to them and extending over the right peripreputial and inguinal regions [Figure 1]. A fleshy, pedunculated, well-circumscribed “skin tag” was located on the right muzzle. Superficial lymph nodes were palpably normal. No further abnormalities were detected on physical examination.

Hematology revealed mild neutrophilia (15.4 × 10⁹/L, reference range 3 to 12 × 10⁹/L), moderate lymphopenia (0.62 × 10⁹/L, reference range 1.2 to 3.8 × 10⁹/L), and eosinopenia (<0.01 × 10⁹/L, reference range 0.1 to 1.3 × 10⁹/L), consistent with a stress hemogram and/or recent corticosteroid therapy. No morphological abnormalities in either the red- or white-cell series were noted on blood-smear examination. Serum biochemical analysis showed mildly elevated alkaline phosphatase (ALP) activity (176 IU/L, reference range 0 to 100 IU/L) and alanine transferase (ALT) activity (71 IU/L, reference range 7 to 50 IU/L), consistent with recent steroid therapy.

Fine-needle aspirates taken from the muzzle mass were consistent with a naevus. Fine-needle aspirates from the preputial lesions revealed a round cell tumor, suggesting either lymphoid neoplasia or a poorly differentiated mast cell tumor.

An incisional biopsy of the preputial mass [Figure 2] revealed extensive lymphocytic infiltration. Multiple neoplastic cells observed in the epidermis and epithelium of the hair follicles were consistent with Pautrier’s microabscesses. Immunohistochemical analysis identified numerous CD3-positive CD79a-negative cells, which supported a diagnosis of T-cell epitheliotropic lymphoma.

Lateral and dorsoventral thoracic radiographs, lateral abdominal radiographs, and abdominal ultrasonography did not reveal abnormalities. Multiple, ultrasound-guided, fine-needle aspirates from the spleen and liver showed no evidence of neoplastic infiltration. Bone marrow aspiration was not performed.

Chemotherapy was initiated using a multidrug protocol consisting of high-dose cyclophosphamide^f (250 mg/m² intravenously [IV] q 3 weeks), vincristine^g (0.75 mg/m² IV q 7 days), and prednisolone^c (20 mg/m² PO q 24 hours) (COP protocol).³⁶ Additional treatment included ranitidine^h (2 mg/kg PO q 12 hours) and sucralfateⁱ (0.05 g/kg PO q 8 hours), which were continued throughout treatment as pro-phylaxis against corticosteroid-induced gastrointestinal ulceration. Treatment resulted in a partial remission (defined as a reduction in tumor volume by >50%) after the first week of therapy. Mild gastrointestinal toxicity was noted 48 to 72 hours after the initial administration of cyclophosphamide. By the 4th week of therapy, new lesions were noted adjacent to the prepuce.

L-asparaginase^j (400 IU/kg intramuscularly [IM], once) was administered at week 5, without response. Intensification of chemotherapy using an additional cyto-toxic agent (epirubicin or cytosine arabinoside) was declined by the owners at week 6, but palliative therapy with cyclophosphamide^k (50 mg/m² PO q 4 days) and pred-nisolone was continued (20 mg/kg m² PO q 48 hours).

At week 7, oral cyclophosphamide was discontinued, and rhIFN-2a (4.8 × 10⁴ IU/kg SC q 24 hours) was initiated. The rhIFN-2a dosage was based on the lowest effective dosage reported in the human literature.²¹ Prednisolone was continued as above. Partial remission (as previously defined) was noted within 4 days of initial rhIFN-2a administration. Chlorpheniramine^l (0.2 mg/kg PO q 12 hours) was added to the regimen at this point as a symptomatic treatment for self-trauma; it was discontinued at week 9. The preputial lesion remained erythematous with superficial scale but continued to regress. By week 10, the lesions were reduced to an area of partially depigmented, palpably normal skin along the prepuce and abdominal wall [Figure 3].

The authors observed no clinically significant adverse reactions to rhIFN-2a therapy. A paroxysmal episode of tachypnea during week 8 spontaneously resolved and did not coincide with any alteration in treatment.

Hematological monitoring was performed at each visit throughout therapy (every 7 to 21 days). Mature neutrophilia (range 12 to 24 × 10¹²/L) was the only intermittent abnormality detected, and this may have been secondary to corticosteroid administration. Routine biochemical testing performed during weeks 7 to 25 revealed variably elevated ALP (range 60 to 146 IU/L) and ALT (range 20 to 110 IU/L).

Asymptomatic gingival and buccal mucosal erythemas were noted at week 17, and biopsies confirmed epitheliotropic lymphoma in these sites. At this time, preputial biopsies from the alopecic but otherwise grossly normal skin confirmed microscopic residual tumor. Chemotherapy with cytosine arabinoside^m (100 mg/m² SC q 24 hours for 4 days) was initiated, with prednisolone and rhIFN-2a continued as previously described. No response was noted to this second round of chemotherapy. Epirubicinⁿ (30 mg/m², IV infusion once) was administered at week 21, but no improvement occurred. Epirubicin therapy resulted in moderate gastrointestinal toxicity that required hospitalization. No further chemotherapy was given, although prednisolone was continued.

The oral lesions remained stable, but by week 22 a new subcutaneous nodule had developed on the right lateral shoulder. Cytology of a fine-needle aspirate confirmed a neoplastic lymphoid population, consistent with progressive disease. At this stage, no gross recurrence of the preputial lesion was seen.

At week 24, the owners reported the dog’s inappetence and oral discomfort from a sudden progression of the oral lesions. Hypofractionated megavoltage radiotherapy (36 Gy in four fractions of 9 Gy at 7-day intervals, delivered using a 4-MeV linear accelerator) was delivered as palliative therapy for the oral lesions. Radiotherapy was completed at week 29, when the dose of rhIFN-2a was increased (5.3 × 10 ⁴ IU/kg SC q 24 hours).

The dog was examined again at week 33. The lesion on the right shoulder was no longer apparent, and the oral mucosa appeared grossly normal. It was not possible to determine if the oral lesions had responded to the radiation therapy, the increased dosage of rhIFN-2a, or both. However, patchy, nonpruritic alopecia was now evident throughout the coat. The rhIFN-2a dosage was again increased to 6.4 × 10⁴ IU/kg (q 24 hours SC), which resulted in stabilization of disease for another 12 weeks.

At week 45, an ulcerated cutaneous lesion on the lateral aspect of the right pinna was noted [Figure 4], as well as generalized subcutaneous swelling ventral to the mandible. Cytology of a fine-needle aspirate from this lesion again demonstrated a neoplastic round cell population with lymphoid morphology. At this stage, an additional increase in interferon dosage or rescue therapy (lomustine) was declined because of financial constraints. The dog was distressed by the progressive, pruritic lesions, and it was euthanized at the owners’ request. Postmortem examination was declined.

Discussion

Although canine epitheliotropic lymphoma has been widely reported, no standard therapy currently exists. Treatment has been based on standard chemotherapy regimens used to treat other lymphoproliferative diseases or on extrapolation from treatment of human epitheliotropic lymphoma or mycosis fungoides.^11,13–17

Both localized and systemic treatments have been described for epitheliotropic lymphoma in dogs. Local therapies include surgical excision, topical retinoids or mechlorethamine (nitrogen mustard), and fractionated radiation therapy.^4,14–18 The most established local treatment is radiation therapy for oral lesions.^19,38 Systemic treatments include cyclophosphamide, vincristine, (cytosine arabinoside), and prednisolone (CO[A]P) protocols; liposome-encapsulated doxorubicin; polyethylene glycol (PEG)-conjugated asparaginase; and 1-(2-chloroethyl)-3-cyclohexyl-l-nitroso-urea (CCNU; lomustine). Responses to these various systemic regimens are frequently poor,^4,20 although median remissions of up to 11 months have been reported.³⁷

In the current case, localized treatment of the initial preputial and peripreputial lesions was excluded as a sole treatment option. This decision was based on the low sensitivity of tumor staging and on the location, extent, and systemic behavior of the lesions. Systemic chemotherapy combined with localized radiation was considered, but the location and extent of lesions meant that there was significant risk of deep-tissue side effects from hypofractionated megavoltage radiotherapy (the only radiation therapy regimen available).

The owners initially elected simple COP combination chemotherapy. Other agents (including retinoids) were discussed when the dog relapsed, but the owners were concerned about the risk of toxicity, as well as the delayed onset of clinical response associated with such therapies.^4,19 Liposome-encapsulated doxorubicin or PEG-asparaginase may have been an appealing option for this case, because these preparations may result in more effective penetration of tumor vasculature with potentially reduced systemic toxicity compared with other chemotherapy protocols. Unfortunately, liposome-encapsulated doxorubicin and PEG-asparaginase are unavailable in the United Kingdom. Therapy with unconjugated preparations of these drugs was unsuccessful in achieving remission in the current case.

Recombinant human IFN-2a provided rapid and complete remission of the preputial and peripreputial lesions (as determined by the absence of gross disease) within 3 weeks of initiating treatment. The remission and survival obtained (45 weeks) is similar to or greater than that obtained with conventional chemotherapy.^7,19

Combination therapy makes it difficult to apportion drug efficacy. However, the rhIFN-2a therapy appeared to be primarily responsible for remission in the current case. First, resolution of gross disease was chronologically related to the start of interferon therapy. Moreover, previous therapy was clearly ineffective, resulting in tumor progression shortly after treatment began.

Prednisolone was the only other cytotoxic agent used consistently throughout treatment. Prednisolone may have had cytoreductive benefits when used alone early in the course of the disease, but this is based solely on unsubstantiated reports from the owner. Furthermore, the rapid expansion of lesions while prednisolone was being administered before referral, as well as the relapse observed during chemotherapy with prednisolone, suggest that any pred-nisolone effect was short-lived. It is also possible that initial steroid therapy (before combination chemotherapy was started) may have actually had an adverse effect on the response to COP therapy.⁴

Synergistic effects of prednisolone and rhIFN-2a therapies have not been described in the human literature, and the concurrent use of prednisolone and rhIFN-2a is not a recognized treatment for mycosis fungoides. In the current case, the response obtained after initiation of rhIFN-2a was not likely due to either prednisolone or the combination of prednisolone and interferon. Nevertheless, the potential for prednisolone to increase or decrease the efficacy of rhIFN-2a cannot be excluded. Indeed, concurrent pred-nisolone therapy might have delayed relapse by delaying production of neutralizing antibody against rhIFN-2a, which could have been the ultimate cause of treatment failure. Support for this is found in human medicine, where concurrent photodynamic (e.g., psoralen plus ultraviolet A light [PUVA]) or retinoid therapy is used synergistically with interferon to decrease the proportion of patients developing neutralizing antibodies.^39,40

The authors suspect that treatment failure in the current case resulted from resistance to rhIFN-2b. Resistance to rhIFN-2b is a multifactorial process.21 However, increased levels of resistance in people are generally seen in antibody-positive patients, and the presence of neutralizing antibodies has been shown to be variably correlated with response in human mycosis fungoides.^21,41 This has also been documented in cats. Loss of clinical response to parenteral rhIFN-2b in feline alimentary epitheliotropic lymphoma was associated with development of anti-IFN-2b antibodies, which is a pattern that mirrors experimental results in cats receiving rhIFN-2b.34,42 The use of an autologous (in this case, canine) recombinant IFN-2a preparation might be expected to reduce or eliminate the development of neutralizing antibodies. However, human patients treated with recombinant human IFN-2a can also develop neutralizing antibodies. Interestingly, studies have shown that the presence of neutralizing antibodies in people can be reversed with continued treatment.⁴³ Measurement of anti-rhIFN-2a antibodies in this dog may have revealed a similar association between relapsing disease, production of anti-rhIFN-2a antibody, and reduction in treatment efficacy.

In the case reported here, recurrent and new cutaneous lesions responded favorably to increased dosage of rhIFN-2a. In human medicine, dose escalation also commonly enhances efficacy, but toxicity prevents use of daily doses above 18 million international units.^21,44,45 The maximum tolerated dosage in dogs is unknown.

In this case, the oral lesions most likely resolved as a result of the localized radiation therapy, rather than the concurrent interferon dosage escalation. However, the dosage escalation possibly contributed to control of the oral lesions, because the increased dosage seemed to improve other lesions that had developed during initial interferon therapy.

The usefulness of rhIFN-2a in human mycosis fungoides is limited more by the development of toxicity than by production of neutralizing antibody.^41,45–47 Human patients may develop a chronic cough (with or without shortness of breath) as a result of therapy.⁴⁵ In the current case, episodes of tachypnea were seen early in therapy, but these resolved rapidly with no further episodes (even after dosage escalation). Reversible myelosuppression is reported in human medicine;⁴⁵ however, it did not occur in this dog. Dose-limiting toxicity was not apparent in this dog at the doses prescribed, which is consistent with previous reports.^30–32 However, the dosages used were low, with escalation prohibited by the owner’s financial constraints rather than by toxicity concerns.

The potential differences in immunophenotype between canine and human epitheliotropic lymphoma may prevent mechanistic extrapolation from human models. Therefore, further canine-specific studies will be required to determine the efficacy of IFN-2a in dogs. In the current case, the authors were unable to perform further immunohistochemistry beyond identification of CD3-positive cells. Further immunophenotyping and monitoring of neutralizing antibody titers may assist in understanding the efficacy of rhIFN-2a in canine epitheliotropic lymphoma.

Conclusion

This report documents the role of IFN-2a in managing a case of canine epitheliotropic lymphoma that did not respond to standard COP therapy. Therapy with IFN-2a resulted in 10 weeks without apparent disease, as well as decreased disease progression after dosage escalation. These findings suggest that IFN-2a may prove useful for management of canine epitheliotropic lymphoma, a disease that is frequently refractory to conventional chemotherapy. Other advantages of this therapy include minimal toxicity and the potential for outpatient treatment. A phase I clinical trial in dogs would help to confirm the efficacy that the authors observed. Such a trial could be followed by double-blinded, placebo-controlled studies to further investigate efficacy. Such studies should also assess the development of antibodies to IFN-2a, the relationship between antibodies and the development of resistance, and the response to dose escalation.

Acknowledgments

The authors thank Gareth Thomas and Dr. Tim Nuttal for interpretation of the histopathology; Annette Kerins for her assistance with imaging; Sue Murphy at the Animal Health Trust, Newmarket, for administering the radiation therapy; and the nursing and ancillary staff at Liverpool Small Animal Hospital.