Interleukin-2 liposome inhalation therapy is safe and effective for dogs with spontaneous pulmonary metastases

Abstract

BACKGROUND

Systemic in vivo toxicity of interleukin-2 (IL-2) has been problematic. Antineoplastic activity of IL-2 has been modest. The authors have previously demonstrated the biologic activity and safety of aerosols of IL-2 liposomes in normal dogs. They now report objective regression of naturally occurring pulmonary metastases in dogs after 1 month of nebulized IL-2 liposome therapy.

METHODS

Dogs with pulmonary metastases (n = 7) and primary lung carcinoma (n = 2) were treated with aerosols of IL-2 liposomes. Response to therapy was monitored with serial chest radiographs. Effector populations, collected by bronchoalveolar lavage (BAL) and from heparinized whole blood, were assessed for cell type, immunophenotype, and tumor cytolytic activity. Immunogenicity of human IL-2 and human serum albumin (HSA) in dogs was assessed by immunofluorescence assay.

RESULTS

Two of four dogs with metastatic pulmonary osteosarcoma had complete regression of metastases; the regression remained stable for more than 12 and more than 20 months, respectively. One of two dogs with lung carcinoma had stabilization of disease for more than 8 months; the other had disease progression. Toxicity was minimal. BAL cell numbers increased more than fourfold (P = 0.01) and included significantly greater proportions and total numbers of eosinophils (P = 0.006) and lymphocytes (P = 0.008). Mean BAL effector lytic activity was significantly greater after 15 days of IL-2 liposome inhalation compared with pretreatment activity (P = 0.01); however, mean BAL lytic activity decreased after 30 days and was no longer significantly greater than pretreatment BAL lytic activity. No allergic reactions were associated with inhaled IL-2 liposome therapy. Canine antibodies against human IL-2 and HSA were detected in all dogs.

CONCLUSIONS

Pet dogs with naturally occurring pulmonary metastases and primary lung carcinomas accepted inhalation treatments easily. Nontoxic and effective treatment of pulmonary metastases of osteosarcoma is possible with nebulized IL-2 liposomes. Cancer 1997; 79:1409-21. © 1997 American Cancer Society.

The treatment outcomes for both pulmonary metastases and primary lung carcinoma have not changed significantly in more than 20 years despite intensification of chemotherapy and combination modality protocols.1-3 Effective and nontoxic alternative therapeutic approaches are needed. Interleukin-2 (IL-2) has induced antitumor effects in both human and animal tumors.4-7 The interaction of IL-2 with its receptor complex results in proliferation and activation of lymphocytes, monocytes, and macrophages.8-10 Antitumor effects of IL-2 are thought to be mediated by natural killer (NK) cells, and possibly other cytotoxic cells such as T-lymphocytes and macrophages.9, 11-13 Widespread use of IL-2 has been limited by its narrow therapeutic index, especially when administered in high dose intravenous protocols.14-16 Toxicities associated with intravenous IL-2 have included fever, pulmonary vascular leakage, weight gain and anasarca, malaise, rigors, azotemia, anemia, and thrombocytopenia.17, 18 Adverse effects of IL-2 are dependent on the dose, route of delivery, and formulation of IL-2.12, 19, 20

Nebulization of a number of therapeutic agents, including bronchodilators, antibiotics, chemotherapeutics, and immunomodulators, has been described.21-26 Aerosol therapy results in high pulmonary drug concentrations and relatively low systemic drug levels, thereby increasing the therapeutic index of the inhaled agents.27-29 Inhalation of free human IL-2 has been shown to be effective against tumors and nontoxic in both humans and in animal models.30-33

Liposomal formulations have been used to decrease toxicity, provide more favorable pharmacokinetics, and maximize tissue targeting of several agents.34-36 Liposomal formulations of IL-2 have been shown to increase the circulating half-life of IL-2, target tissues of the immune system (spleen, lymph nodes, bone marrow) and lungs, and decrease toxicity, including pulmonary vascular leakage.14, 37-39 Nebulized liposomal formulations of cytarabine, beclomethasone, and cyclosporine have increased pulmonary drug deposition and allow longer pulmonary retention after inhalation than aerosols of the free drug.28, 40-42

It has been the authors' interest to combine a liposomal formulation of IL-2 with aerosol delivery to provide an effective and nontoxic treatment for pulmonary metastases and primary lung carcinomas. They have previously demonstrated the safety and biologic activity of inhaled liposomal formulations of IL-2 in normal dogs.43 Dogs receiving aerosols of IL-2 liposomes had greater pulmonary immune activation (tumor cytolysis) than dogs receiving aerosols of free IL-2, empty liposomes, or saline. The inhalation of IL-2 liposomes did not result in significant activation of systemic immune effector cells. These biologic effects of IL-2 liposomes were achieved in the absence of significant toxicity.

In this article, the treatment of radiographically measurable pulmonary metastases and primary lung carcinomas in dogs is described. The specific study objectives of this Phase I/II study included: 1) demonstration of the feasibility and safety of inhalation therapy in pet dogs; 2) evaluation of antitumor effects of inhaled IL-2 liposomes, and 3) analysis of changes in cellular immune effector populations in the lung and blood during inhalation therapy. Results suggest that significant antitumor responses can be generated without appreciable toxicity using aerosol delivery of IL-2 liposomes. These findings in spontaneous canine cancers may be relevant to the treatment of similar cancers in human patients.

MATERIALS AND METHODS

Canine Subjects

Nine pet dogs with radiographically measurable pulmonary metastases and primary lung carcinomas of one or both lungs were entered to this Phase I/II trial of inhaled IL-2 liposomes, at the University of Minnesota Veterinary Teaching Hospital (UMVTH), after written informed consent of the owners was obtained. Pretreatment evaluation was comprised of physical examination, complete blood count, serum biochemical profile, urinalysis, and thoracic radiographs. Dogs were presumed to have pulmonary metastases from prior history of histologically confirmed cancer (e.g., osteosarcoma), the radiographic appearance of pulmonary lesions (multiple pulmonary densities), or confirmation of pulmonary metastases at postmortem examination in the event of treatment failure. Selected cases had fine-needle biopsy of pulmonary metastatic lesions and/or concurrent extrapulmonary metastases. Dogs were presumed to have primary lung carcinomas based on radiographic appearance, the absence of identifiable extrathoracic primary cancer, fine-needle biopsy consistent with lung carcinoma, and/or postmortem examination in the event of treatment failure. All procedures involving animals were reviewed and approved by the Animal Care Committee at the University of Minnesota.

Nebulization (Aerosol) Treatment

IL-2 liposomes were kindly provided by Biomira USA, Inc. (formerly OncoTherapeutics Inc., Cranbury, NJ). This liposome preparation contains a synthetic lipid, dimyristoylphosphatidyl choline (DMPC) (Avanti Polar Lipids, Alabaster, AL), human serum albumin (HSA) (Baxter, Glenville CA), and human recombinant IL-2 of the natural sequence (Tecelleukin; specific activity 1.4 × 10⁷ IU/mg) (Hoffman LaRoche, Nutley, NJ). Method of preparation, size distribution of these multilamellar vesicles, and biologic activity in the CTLL-2 IL-2 bioassay have been previously reported.14, 36, 44

Nebulization of IL-2 liposomes, undertaken by the dog owners as previously described,43 involved the use of a Bunn 400A compressor (John Bunn Co., Happauge, NY) and Puritan Bennet Twin Jet Nebulizer (Puritan Bennet Corp., Carlsbad, CA) connected to a polyethylene rebreathing bag. The polyethylene bag was held over the dog's muzzle manually during 15-20 minute treatments. IL-2 liposomes (1 × 10⁶ IU of IL-2 in 20 mg of DMPC) were diluted in 3.5 mL of sodium chloride and then delivered to the dogs. The first four dogs were treated with 1 × 10⁶ IU IL-2 liposomes twice daily for 15 days and then 1 × 10⁶ IU IL-2 liposomes three times daily for 15 days. Dogs 5-9 were treated with 1 × 10⁶ IU IL-2 liposomes twice daily for 30 days.

Training of the dogs for nebulization involved one week of saline nebulization using positive reinforcement techniques (i.e., rewards for cooperative behavior). Dogs and owners were evaluated at the UMVTH at the completion of the training period to assess success of training for inhalation therapy.

Assessment of clinical toxicity associated with inhaled IL-2 liposomes was performed on Day 15, on the final day of inhalation therapy (Day 30), and 15 days after completion of therapy (Day 45). Complete physical examinations, hematologic studies (including red cell counts, leukocyte counts, platelet count, leukocyte differential, and red cell Coulter counter (Coulter, Hialeah, FL) indices, serum biochemistries (including determination of concentration of albumin, total protein, alanine transaminase, alkaline phosphatase, aspartate transaminase, total bilirubin, phosphorous, cholesterol, blood urea nitrogen, creatinine, potassium, sodium, and chloride) were performed. Blood samples were analyzed by the clinical diagnostic laboratories of the UMVTH.

Antitumor effects were assessed by comparison of thoracic radiographs taken at Days 0, 15, 30, 45, 60, 90, and then every 3 months thereafter. The number of pulmonary metastases or primary lung tumors, the individual tumor volume (volume was estimated by 4/3πr³; r = [radius] one-half of the greatest cross sectional dimension of the lesion) and an estimate of total pulmonary tumor burden (sum of individual tumor volumes) were taken from chest radiographs. History and physical examinations were used to detect and evaluate progression of extrathoracic metastases. The number and volume of extrathoracic metastasis was undertaken as above. Tumor responses were characterized using the following definitions: complete response (CR): disappearance of all clinical evidence of active tumor; partial response (PR): ≥50% decrease in the sum of the volumes of all measured lesions with no simultaneous increase in volume of any lesions or appearance of any new lesion; stable disease (SD): steady state or response of less than partial remission with no new lesions and no worsening of the clinical signs; and progressive disease (PD): unequivocal increase of at least 50% in the volume of any measurable lesion, or appearance of new lesions.

Bronchoalveolar lavage (BAL) effector populations were collected from selected dogs (Dogs 1, 2, 3, 4, 6, and 7) using general anesthesia. Selection of dogs for bronchoscopy was based on the feasibility of repeated bronchoscopic examinations. Dogs were anesthetized with butorphanol premedication (0.4 mg/kg IM) (Torbugesic®, Fort Dodge, Fort Dodge, IA), thiopental (8 mg/kg) (Thiopental Sodium®, Gensia Laboratorie Irvine, CA) or propofol (Deprivan®, Zeneca Inc., Wilmington, DE) induction, and maintained on isoflurane (IsoFlo®, Solvay Animal Health, Mendota Heights, MN) as needed. BAL procedures averaged approximately 20 minutes per dog. Cells were harvested using 20 mL 0.9% saline lavages (total 160 mL) through a flexible fiberoptic endoscope (GIF-P2, Olympus, Lanexa, KS). Lavage fluid was held on ice, and cells were then concentrated by centrifugation at 400 g for 5 minutes. Coupage (forceful patting of the thoracic cage with a cupped hand) during the BAL procedure significantly increased the cellular yield of each collection.43 After concentration, BAL effectors were resuspended in cold media (RPMI 1640; Celox Co., Hopkins, MN) with 10% fetal calf serum (Sigma, St Louis, MO), 100 μg/mL sodium pyruvate, 100 μg/mL penicillin-streptomycin, and 2 mM L-glutamine (Sigma) passed through a 70-μm Cell Strainer™ (Becton Dickinson, Franklin Lakes, NJ) and washed twice in the same media. BAL effector viability was assessed by enumeration with trypan blue viability stain. Effectors were used for in vitro assays of tumor target cytolysis on the same day as collection and/or frozen in liquid nitrogen for later evaluation of surface cell marker expression.

BAL effector cytology was evaluated in dogs when BAL was undertaken. Cytospin slides of BAL fluid were stained with Wright-Giemsa stain and evaluated microscopically for morphology and differential cell count; 200 cells were evaluated per slide.

Effector cell surface markers were evaluated using mouse anticanine monoclonal antibodies, obtained from the laboratory of Dr. Peter Moore (University of California at Davis), recognizing canine CD3, CD4, and CD8.45 BAL cells were resuspended in 10% heat-inactivated goat serum and buffer (phosphate-buffered saline [PBS], 5mM ethylenediamine tetraacetic acid and 0.1% weight/volume sodium azide). Samples were incubated for 20 minutes on ice, as a blocking step. Cells were added at 3.5 × 10⁵ cells (33 μL) to U-bottom, 96-well plates. Primary monoclonal antibody culture supernatant fluid (10 to 25 μL) was added to respective wells and incubated for 30 minutes on ice. Plates were washed and cell pellets resuspended in 10 μL of 1:100 dilutions of secondary goat antimouse immunoglobulin (Ig)G-fluorescein isothiocyanate. Samples were incubated for 30 minutes on ice. Cell pellets were washed and cells were fixed in 200 μL of 1% paraformaldehyde in PBS, pH 7.4. Cells were then transferred to 12 mm × 75 mm polypropylene test tubes (Becton Dickinson, San Jose, CA) for analysis with Becton Dickinson FACScan equipped with Consort 30 software (Becton Dickinson, Mountainview, CA).

Peripheral blood mononuclear cell (PBMC) effector populations, separated from heparinized whole blood using techniques previously described,46 were harvested from selected dogs (Dogs 1, 2, 3, 4, 6, 7 and 9). Briefly, whole blood was diluted 1:1 with Hanks balanced salt solution without calcium or magnesium and centrifuged on an 8:3 Ficoll-Hypaque (Histopaque-1077; Sigma Diagnostics, St. Louis, MO) gradient at 400 g for 25 minutes. PBMC were then removed from the cellular interface and washed three times in cold media. PBMC effector viability was assessed by enumeration with trypan blue viability stain. Effectors were used for in vitro assays of tumor target cytolysis on the same day as collection.

Studies of cytotoxicity against tumors, using a 4-hour ⁵¹chromium (Cr) release assay, was performed as previously described6, 47 on selected dogs (Dogs 1, 2, 4, and 6). Briefly, a coculture (4 hours in 5% CO₂ at 37 °C) of effectors (PBMC or BAL leukocytes) and a ⁵¹Cr-labeled (5 mCi/mL ⁵¹Cr) (Amersham Co., Arlington Heights, IL) canine thyroid adenocarcinoma tumor target (CTAC) (kindly provided by Dr. Stuart Helfand, University of Wisconsin at Madison, College of Veterinary Medicine) was undertaken in 96-well V-bottom plates. NK-92,48 a human NK cell line provided by Dr. H. G. Klingeman (Vancouver, British Columbia, Canada) was used as a positive control effector cell to determine day-to-day assay variation in lytic sensitivity of tumor target cells. Serial dilutions of effectors at effector to target ratios of 100:1; 33:1; 11:1; and 3.7:1 were performed in triplicate. A beta emission counter (LKB 1216, Rackbeta, Turku, Finland) was used to measure ⁵¹Cr release from 100 μL of culture supernatant fluid (taken from each well) in 3 mL of scintillation fluid (Cytocint; ICN, Costa Mesa, CA). Percent specific cytolysis was calculated as follows:

• % cytolysis = [(counts per minute [cpm] effector and target coculture - cpm spontaneous)/ (cpm maximum - cpm spontaneous)] × 100

Maximum cpm was determined by ⁵¹Cr release from incubated target cells lysed with 100 μL lytic detergent (0.4% hexacecyl-tri-methyl ammonium bromide) (Sigma). Spontaneous cpm was determined by measurement of ⁵¹Cr release from an incubated target population in assay media.

Canine antibodies directed against human recombinant human IL-2 and HSA at Day 0, 15, 30, and 45 of inhalation of IL-2 liposomes were evaluated by immunofluorescence assay (IFA). One hundred μL of IL-2 (25 μg/mL) (Roussel Uclaf, Paris, France), or HSA (50 μg/mL) (American Red Cross, Philadelphia, PA) in PBS were placed in each well of a 96-well microtiter plate and incubated at 4 °C for at least 2 hours. The IL-2 and HSA were removed from the microtiter plate wells and 250 μL of 2% solution of dry milk in PBS were added to each well and incubated for 15 minutes at 4 °C. The milk solution was removed from the wells and they were washed three times with 250 μL of PBS. Serum or plasma of dogs were diluted in 0.2% dry milk in PBS and 100 μL were added to the wells and incubated for 1 hour at 4 °C. Normal dog serum/plasma or serum/plasma of dogs prior to treatment were used as controls. The samples were removed and the wells were washed 3 times with 250 μL of PBS. A diluted secondary biotinylated rabbit antibody recognizing canine IgG (1/500, Sigma) was added to each well and incubated for 30 minutes at 4 °C. The antibody was removed and wells were washed 3 times with 250 μL of PBS. Europium-labeled streptavidin (1/1000, diluted in 0.2% milk) (Wallac, Gaithersburg, MD) was added to the wells for 10 minutes at room temperature. It was then removed and the wells were washed 3 times with 250 μL of PBS. One hundred μLs of Enhance™ solution (Wallac) was added to the wells and incubated at least 5 minutes at room temperature while shaking. The fluorescence of europium counts per second (CPS) was measured by time-resolved 1234 Delfia Research Fluorometry (Wallac). The sensitivity range of this IFA for IgG is between 10 and 0.001 μg/mL.

Statistical Analysis

Descriptive statistics and comparisons of differences between means of data sets using Student's t test for paired or unpaired data were calculated by InStat™ software (Macintosh version; Graph Pad Software, San Diego, CA). Statistically significant differences in data sets were defined by a P value <0.05. Statistical trends in data set comparisons were defined by a P value of <0.20 and >0.05.

RESULTS

Canine Subjects

Nine dogs were entered to this Phase I/II trial of inhaled IL-2 liposomes. Two dogs had primary lung carcinomas and seven had pulmonary metastases. Of the dogs with pulmonary metastases, four had primary extremity osteosarcomas (Dogs 1, 2, 5, and 9), one had a primary mammary carcinoma (Dog 3), one had a primary digital melanoma (Dog 4), and one had a soft tissue fibrosarcoma (Dog 5). Table 1 includes the World Health Organization stage49 of dogs at the time of entry to this trial.

Table 1. Tumor Responses Associated with Inhaled IL-2 Immunotherapy in Dogs

Dog	Diagnosis	Day 0	Day 15	Day 30	Day 45	Response
IL-2: interleukin-2; CR: complete response; PR: partial response; SD: stable disease; PD: progressive disease. a World Health Organization stage49 assigned to dogs at the time of entry to Phase I/II trial of inhaled interleukin-2 liposomes. b Tumor burden (cm^3): calculated from the sum of individual tumor volumes (4/3πr^3, in which r: tumor greatest dimension/2 from lateral thoracic radiograph). c Number of tumors: enumerated from lateral thoracic radiographs. No attempt was made to count > 50 individual metastases. d Complete response (regression) or prescapular lymph node metastasis was documented at Day 60. e Development of melanoma metastases to the right kidney represented progressive disease and resulted in death.
1	Osteosarcoma metastases (TIIb-resectedM1M1-lung and lymph node)a
	Pulmonary tumor burdenb	4.6	0.6	0	0	CR > 600 days
	Pulmonary tumor numberc	1	1	0	0
	Lymph node tumor burden	121	57.4	37.9	13.7d	CR > 600 days
2	Osteosarcoma metastases (TIIb-resectedM1-lung)
	Pulmonary tumor burden	4.2	14.7	23.4	51.7	PD
	Pulmonary tumor number	1	2	3	3
3	Mammary carcinoma metastases (T2-resectedN0M1-lung)
	Pulmonary tumor burden	>500	>500	>500	Unable to count	PD
	Pulmonary tumor number	34	48	50
4	Digital melanoma metastases (TresectedN0M1-lung)
	Pulmonary tumor burden	28.5	31.5	29.0	Dead	PDe
	Pulmonary tumor number	32	42	33
5	Fibrosarcoma metastases (T3-resectedN0M1-lung
	Pulmonary tumor burden	3.0	5.0	13.5	Dead
	Pulmonary tumor number	4	8	17		PD
6	Osteosarcoma metastasis (TIIbM1-lung)
	Pulmonary tumor burden	5.4	5.1	0	0	CR
	Pulmonary tumor number	6	3	0	0	> 360 days
7	Primary lung (T1N0M0)
	Tumor burden	18.8	14.1	14.1	14.1	SD
	Tumor number	1	1	1	1	> 240 days
8	Primary lung (T1N0M0)
	Tumor burden	8.2	12.8	22.4	26.5	PD
	Tumor number	1	1	1	1
9	Osteosarcoma metastasis (TIIbM1-lung)
	Pulmonary tumor burden	8.2	8.2	10.3	14.1	PD
	Pulmonary tumor number	1	1	1	1

Training of dogs for nebulization was feasible and was completed in <1 week in all subjects. All dogs became cooperative during inhalation treatments and could be easily treated by their owners. Each nebulization treatment with IL-2 liposomes (4.0 mL) was completed in approximately 20 minutes.

Toxicity of Inhaled IL-2 Liposomes

There were no significant adverse reactions associated with inhaled IL-2 liposomes twice daily or three times daily in any pet dogs. Dogs 2, 4, and 6 had mild coughs immediately after aerosolization treatments. Complete blood counts, serum biochemical profiles, and urinalyses on Day 0, Day 15, Day 30, and Day 45 revealed no significant abnormalities in Dogs 1-5. Dogs 6-9 were evaluated with clinical examinations only and had no significant abnormalities. No dog developed fever, weight gain, or tachypnea while on IL-2 liposome aerosol therapy.

Antitumor Responses

Responses to therapy with inhaled IL-2 liposomes are presented in Table 1. Two dogs with metastatic osteosarcoma had CR of all metastatic lesions after inhalation of IL-2 liposomes. Dog 1 had a CR of not only a lung metastasis but also a prescapular lymph node metastasis documented by biopsy. This was an eight-year-old female labrador with an osteosarcoma of the proximal humerus. Initial therapy included amputation of the tumor-bearing limb followed by adjuvant cisplatin (Bristol-Myers Oncology, Evansville, IN) chemotherapy. After a 3-month disease free interval, a pulmonary metastatic nodule in the right caudal lung field and regional prescapular lymphadenopathy were noted. Fine-needle biopsy of the lymph node revealed malignant sarcoma cells (Fig. 1). Treatment with inhaled IL-2 liposomes was associated with a CR of the pulmonary metastatic nodule and prescapular lymph node within 60 days of initiation of inhalation therapy. At last follow-up, this CR had been stable for >600 days.

Figure 1. Metastatic invasion of malignant sarcoma cells to the regional (prescapular) lymph node of Dog 1. Photomicrograph (×400) of lymph node fine-needle biopsy demonstrates numerous large round malignant cells (resembling osteoblasts) consistent with a metastatic sarcoma. The lymph node volume at Day 0 was 121 cm³. Complete regression of the lymph node metastasis occurred 30 days after completion of inhalation therapy. This dog with a complete response, including regression of a pulmonary metastasis, had, at last follow-up, been stable for more than 600 days.

The second responder (Dog 6) also had lung metastases from an extremity osteosarcoma. This dog was a nine-year-old male golden retriever with an osteosarcoma of the right accessory carpal bone. Therapy of this primary lesion was comprised of palliative radiation therapy (800 centigray on Days 0, 7, and 21). Four months after successful palliative radiation therapy, reevaluation revealed several pulmonary metastatic lesions. Treatment with inhaled IL-2 liposomes was associated with a CR in the pulmonary metastatic disease after 30 days of inhaled IL-2 liposomes. Figure 2 illustrates radiographic regression of the pulmonary metastases in this dog. At last follow-up, this CR had been stable for >360 days. The accessory carpal bone in this dog has undergone complete lysis since entry to this study. Follow-up biopsy has revealed persistence of osteosarcoma in the soft tissue of the carpus.

Figure 2. Complete regression of a pulmonary metastases from a skeletal osteosarcoma in Dog 6 treated for 30 days with inhaled interleukin-2 (IL-2) liposomes. (A) A lateral thoracic radiograph taken at Day 0 demonstrates 6 pulmonary metastases (closed arrowheads) with a total pulmonary tumor burden of 5.4 cm³. (B) Lateral thoracic radiograph after 15 days of inhaled IL-2 liposome immunotherapy demonstrates a reduction in the number of metastases (from 6 to 3 metastases [closed arrowheads]) and the pulmonary tumor burden (5.1 cm³). (C) Lateral thoracic radiograph after 30 days of inhaled IL-2 liposome immunotherapy demonstrates a complete regression of pulmonary metastases. This dog with a complete regression of pulmonary metastases had at last follow-up been stable for more than 360 days.

Dog 7, a 13-year-old black labrador dog with a primary lung carcinoma, had SD for 280 days, before PD. The primary lung nodule had decreased in size after the initiation of therapy (17.8 cm³ to 14.1 cm³) and no new lesions were documented on follow-up pulmonary radiographs until 280 days. All other dogs had progression of their pulmonary metastases or primary lung carcinomas and were euthanized as a result of pulmonary or extrapulmonary disease.

Cytological Evaluation of BAL Effectors

Cytological evaluation of BAL effectors revealed a significant increase in the total number of cells and the mean total number of macrophages, eosinophils, and lymphocytes collected by bronchosopic lavage 15 days and 30 days after initiation of inhalation therapy (Fig. 3).

Figure 3. Inhaled interleukin-2 (IL-2) liposomes significantly increase both total bronchoalveolar lavage (BAL) cell counts and the proportion of BAL eosinophils and lymphocytes. BAL leukocyte effectors collected from Dogs 1, 2, 3, 4, 6, and 7 after 15 and 30 days of treatment with nebulized IL-2 liposomes were significantly increased over pretreatment counts. Increase in cellular effectors on Day 15 compared with pretreatment (Day 0): total effector population, P = 0.01; macrophages, P = 0.04; eosinophils, P = 0.006; and lymphocytes, P = 0.008. Increase in cellular effectors on Day 30 compared with pretreatment (Day 0): total effector population, P = 0.04; macrophages, P = 0.06; eosinophils, P = 0.02; and lymphocytes, P = 0.01.

Changes in Effector Lymphocyte Cell Surface Marker Expression

Figure 4 illustrates the changes in BAL lymphocyte immunophenotype during inhaled IL-2 liposome therapy. Inhalation of IL-2 liposomes increased expression of CD3 on BAL lymphocytes after 30 days of inhalation. Expression of CD4 and CD8 on BAL lymphocytes was unchanged on Days 15 and 30 when compared with Day 0; however, a shift in the CD4 to CD8 ratio was apparent.

Figure 4. Bronchoalveolar lavage (BAL) lymphocyte expression of CD3 was significantly increased after inhalation of interleukin-2 (IL-2) liposomes. Assessment of BAL lymphocyte surface marker expression was undertaken in Dogs 1, 2, 3, 4, and 6. CD3 expression on BAL lymphocytes increased from 35% at Day 0 to 45% (P = 0.17) and 57% (*P = 0.03) after 15 and 30 days of inhalation, respectively. Expression of CD4 on BAL lymphocytes was 21% on Day 0, 30% on Day 15 (P = 0.27), and 29% on Day 30 (P = 0.46). CD8 expression was unchanged on Days 15 and 30 when compared with Day 0.

Cytolytic Activity of Cellular Immune Effectors before and after Aerosolized IL-2 Liposomes (⁵¹Cr Release Assays)

Increased tumor cytolysis by BAL effectors was observed after 15 days of inhalation of IL-2 liposomes (Fig. 5). Despite continued IL-2 liposome treatment, mean BAL cytolytic activity appeared to decrease by Day 30 and was no longer significantly greater than pretreatment cytolysis (P = 0.15). No significant differences were observed in BAL cytolytic activity in dogs treated twice daily for 15 days and then three times daily for 15 days compared with dogs treated twice daily for 30 days (data not shown). No significant tumor target cytolysis was observed by PBMC effectors before, after 15 days, or after 30 days of inhalation (data not shown).

Figure 5. Inhaled interleukin-2 (IL-2) liposomes induce tumor cytolytic activity in bronchoalveolar lavage (BAL; pulmonary) leukocyte effectors. Fresh BAL leukocyte effectors from dogs (Dogs 1, 2, 4, and 6) treated with nebulized IL-2 liposomes demonstrate significant increases in killing of canine tumor target (canine thyroid adenocarcinoma [CTAC]) after 15 days of inhalation compared with pretreatment cytolysis. Cytolytic activity of canine (BAL) leukocytes was determined by cytotoxic release of ⁵¹chromium (4-hour coculture) from tumor targets. Inhalation of IL-2 liposomes for 15 days resulted in significantly increased mean BAL lytic activity against the CTAC tumor target compared with pretreatment lytic activity (Day 0 vs. Day 15, P = 0.01). After inhalation of IL-2 liposomes for 30 days, the mean lytic activity of BAL effectors against the CTAC tumor target decreased compared with Day 15 levels. Day 30 mean BAL lytic activity was not significantly increased compared with pretreatment lytic activity (Day 0 vs. Day 30, P = 0.25). The lytic activity of canine BAL effectors stimulated with 1000 IU/mL IL-2 was 8% using identical culture conditions.43

Immunogenicity of Inhaled Human IL-2 Liposomes in Dogs

Allergic reactions were observed not seen in any of the dogs treated with inhaled IL-2 liposomes. All dogs treated with inhaled liposomes containing human IL-2 and HSA developed high levels of antibodies directed against human IL-2 and HSA, as measured by IFA (Fig. 6).

Figure 6. Dogs treated with inhaled interleukin-2 (IL-2) liposomes developed immunologically active antibodies against human recombinant IL-2 and human serum albumin (HSA). Using an immunofluorescent assay, canine antibodies directed against both IL-2 and HSA were detected in serum collected during and after inhaled IL-2 liposome immunotherapy. The levels of anti-HSA and anti-IL-2 antibodies were similar in responding and nonresponding dogs.

DISCUSSION

Use of IL-2 in cancer immunotherapy has been limited by systemic toxicity. In this study, the authors demonstrated that twice daily inhalation of IL-2 liposomes for 30 days resulted in the complete regression of naturally occurring pulmonary metastases in 2 dogs. No toxicity was observed. After inhaled IL-2 liposome immunotherapy, a significant increase in the number of pulmonary effectors and the individual effector cytolytic activity was observed. These results suggest that inhaled IL-2 liposomes may provide a safe, effective, and convenient therapy for pulmonary metastatic disease. The feasibility of training pet dogs, with naturally occurring cancers, to undergo inhalation therapy will allow future studies of inhaled biotherapy in dogs. Such studies may be highly relevant in the treatment of human cancers.

The most impressive antitumor responses observed in this study occurred in dogs with pulmonary metastases from osteosarcoma (Dogs 1 and 6). CRs against measurable pulmonary metastases of osteosarcoma have not been previously described in dogs. In one report, 45 dogs with radiographically evident pulmonary metastases of osteosarcoma were treated with single agent chemotherapy.50 Responses were limited to a PR of pulmonary metastases, which were stable for only 21 days. Canine osteosarcoma is a highly relevant model of osteosarcoma observed in adult and pediatric patients and of pulmonary micrometastatic disease in general.51 The histologic classification of canine osteosarcoma is essentially identical to human osteosarcoma with a high grade osteoblastic subclassification given to >90% of canine tumors.51 Osteosarcoma patient profiles are also similar in dogs and humans, with male gender predilection and large patient size, although canine patients are more commonly middle-aged to elderly.51 Appendicular metaphyseal locations for primary tumors are most common in dogs, with a predictable pattern of metastatic spread to the lungs.51 Despite no evidence of metastatic disease at the time of initial surgery, >90% die of pulmonary metastases in approximately 4 months.52, 53 Chemotherapy has increased the disease free interval and survival of dogs after amputation or limb salvage, extending the median survival times to 9.5-14 months.54, 55 Nevertheless, metastases to the lungs continue to be the most common site of failure and cause of death.

Immunotherapy, alone and in combination with chemotherapy, after primary tumor control has been investigated as a treatment modality for canine osteosarcoma. Recent work by MacEwen et al., with liposomal muramyl tripeptide phosphatidyl ethanolamine. (MTP-PE), has demonstrated improved survival over surgery alone56 and synergy when MTP-PE liposomes were used with chemotherapy.57 These studies in dogs with spontaneous osteosarcoma have provided valuable preclinical information on the use of MTP-PE for human trials.3, 57

The extent of pulmonary tumor burden in dogs at the initiation of inhalation therapy may have greatly influenced the treatment responses.58 In this Phase I/II trial, even dogs with advanced disease were eligible for therapy. Many of the dogs not responding to therapy had large tumor burden in the lungs that may have dictated the outcome of therapy to a greater extent than the sensitivity of the tumor type to inhaled IL-2.

The increase in the total immune effector population collected by bronchoscopy and increase in lytic activity of immune effectors were similar to the cellular changes that the authors have previously described in normal dogs treated with inhaled IL-2 liposomes.43 The combination of increased BAL effector numbers and individual BAL effector cytolysis suggests a biologically effective increase in total pulmonary antitumor immune activity. The increase in lytic activity of BAL effectors after inhalation therapy, demonstrated by the 4-hour ⁵¹Cr release assay, was statistically significant but relatively small. This modest increase in lytic activity was observed despite the expected cellular kinetics of target lysis by the BAL effector population (predominately macrophages). An 18-hour ⁵¹Cr release assay has yielded a greater difference in BAL effector lytic activity in IL-2 treated versus untreated dogs.43

The increase in the proportion of eosinophils and lymphocytes collected by BAL after inhaled IL-2 liposomes was similar to changes observed in normal dogs; however, the increase in the lymphoid population was not as dramatic in tumor-bearing dogs.43 Such changes in the BAL cellular populations are similar to those observed in human patients treated with aerosols of free human IL-2.59 Increases in peripheral eosinophil counts are commonly reported after subcutaneous and intravenous delivery of IL-2.60 BAL eosinophilia may be a similar cytokine (direct or indirect) effect associated with this route of delivery or a reaction to the inhaled foreign human IL-2. Studies involving the inhalation of human IL-2 in the guinea pig suggest that the recruitment of eosinophils by IL-2 is a secondary cytokine effect associated with a stimulated cell-mediated immune system.61 The role of the eosinophil in the antitumor response should not be dismissed, because the eosinophil's ability to act as an antitumor effector has been described.62 Furthermore, it has been suggested that the eosinophil may be an important effector in IL-2 treated patients.60, 63

Studies of surface cell marker expression may provide some insight into the cellular mechanisms responsible for IL-2-mediated tumor regression. Determination of pulmonary NK cell number after inhalation therapy would be of interest. Unfortunately, the authors are not aware of antibodies reactive against canine NK cells. The increased lymphocyte numbers and increase in the relative expression of CD3 and CD4 is consistent with the hypothesis that IL-2 may stimulate specific cell-mediated immune responses against metastases.12 Naturally occurring specific tumor-reactive T cells may be stimulated to proliferate and expand, by a vaccine adjuvant effect, through the local delivery of IL-2, by liposomes, to bronchial-associated lymphoid centers.12, 38, 64 This proliferation and immune center intercommunication may explain the regression of the peripheral lymph node metastasis in Dog 1.

High pulmonary and low systemic drug levels of aerosolized therapy coupled with the pulmonary depot effects of liposomal formulations may explain the efficacy and lack of toxicity associated with this route and formulation of IL-2. The most severe toxicities associated with IL-2 have been observed in high dose intravenous protocols, in which IL-2 has been used at the maximally tolerated dose.12, 65 The importance of route of delivery and formulation of administered IL-2 has been previously reviewed.19 In the current study the combined strategies of liposomal formulation with local delivery via inhalation was very well tolerated by dogs and was easily undertaken by pet owners. Lack of toxicity was also observed in normal dogs treated with inhaled IL-2, IL-2 liposome, and empty liposome formulations.43 Studies in normal dogs have demonstrated the advantage of IL-2 liposomes in stimulating pulmonary immune antitumor activity compared with inhaled free IL-2 and inhaled empty liposomes.43 In human patients with metastatic renal cell carcinoma, inhalation of free IL-2 has been associated with objective regression of pulmonary metastases without significant toxicity.31, 59 Treatment in this human clinical trial with inhaled free IL-2 was given five times daily. The high degree of biologic effectiveness of inhaled IL-2 liposomes given twice daily to dogs may be translated into a more convenient schedule for future human clinical trials.

Human patients treated with human IL-2 sometimes develop antibodies against the recombinant protein. Thus, the development of antibodies in dogs against human IL-2 was not unexpected. Canine IL-2 shares considerable sequence homology with the human protein. However, canine IL-2 is clearly different and has approximately twice the molecular weight of human IL-2.66 All dogs treated with inhaled IL-2 liposomes developed antibodies against human IL-2 and HSA as measured by IFA. It is probable that the development of antibodies is in part responsible for the decline in immune effector lytic activity observed at Day 30. Canine neutralizing antibodies against human IL-2 were detected in these dogs after 30 days of inhalation therapy (unpublished data). The presence of canine antibodies to human IL-2 did not induce clinically evident immunosuppression in any dogs. In future canine trials using human IL-2, the development of neutralizing antibodies against human IL-2 by Day 30 should be considered.

The findings of this study suggest that inhaled IL-2 liposome immunotherapy in pet dogs is both efficacious and safe. This therapy was convenient for pet owners, well tolerated by dogs, and resulted in significant antitumor responses in dogs with metastatic osteosarcoma. The therapeutic implications for human patients with metastatic osteosarcoma, other IL-2 responsive cancers metastatic to the lung (such as renal cell carcinoma), and primary lung carcinoma remain to be determined.

Acknowledgements

Biomira USA, Inc., kindly provided, as a gift, interleukin-2 liposomes for this study. Dr. Mary Neville (employed by Biomira USA, Inc., previously OncoTherapeutics, Inc.) was involved in the analysis of canine antibodies to human serum albumin and human interleukin-2. All samples were coded when submitted to Dr. Neville.

The authors acknowledge the assistance of Maria Crowley, Jennifer McCarty, and Lynn Lawrence in the care and evaluation of the study dogs. They would also like to thank Drs. Daniel Saltzman, Ed Sunbery, Mircea Popescu, Mark Sorenson, and Keith Skubitz for encouragement and helpful discussions during these studies. They are especially thankful to the dog owners for their time and efforts in aerosol administration.