Targeting the lung: preclinical and comparative evaluation of anticancer aerosols in dogs with naturally occurring cancers.

Abstract

Pet dogs with naturally occurring cancers offer a novel opportunity for the study of both cancer biology and therapy. The following review will provide the rationale for the use of these spontaneous cancer models in translational research, particularly in the development of anticancer aerosols. A summary of work involving pet dogs with primary and metastatic cancers to the lung and the investigation of therapeutic chemotherapy and cytokine immunotherapy aerosols will be presented.

Full text

Page 1

Current Cancer Drug Targets, 2003, 3, 265-273265
1568-0096/03 $41.00+.00 © 2003 Bentham Science Publishers Ltd.
Targeting the Lung: Preclinical and Comparative Evaluation of Anticancer
Aerosols in Dogs with Naturally Occurring Cancers
Chand Khanna1 and David M. Vail2
1Comparative Oncology Program, Center for Cancer Research, National Cancer Institute, National Institutes of
Health, Bethesda, MD 20892, USA
2School of Veterinary Medicine and the Comprehensive Cancer Center, University of Wisconsin at Madison,
Madison, Wisconsin, USA
Abstract: Pet dogs with naturally occurring cancers offer a novel opportunity for the study of both cancer
biology and therapy. The following review will provide the rationale for the use of these spontaneous cancer
models in translational research, particularly in the development of anticancer aerosols. A summary of work
involving pet dogs with primary and metastatic cancers to the lung and the investigation of therapeutic
chemotherapy and cytokine immunotherapy aerosols will be presented.
INTRODUCTION
The significant anatomic and physiologic similarities
that exist between dogs and humans have been the basis for
the use of dogs in biomedical research for over 70 years.
Reports of the use of dogs in the field of cancer drug
development extend to the first form of systemic
chemotherapy, nitrogen mustard. [1] Dogs continue to be
used to define the safety profiles for novel cancer agents
destined for use in human phase I clinical studies. Similarly
the preclinical evaluation of aerosol-based therapies has a
long history in dogs. Dogs have been used to define particle
distribution to the lung following inhalation of therapeutic
aerosols. [2] These studies provided in vivo validation of
particle distribution and deposition hypotheses that are still
used to characterize inhaled aerosols based on particle size.
More recently, dogs have been used to assess the effects of
environmental exposure to inhaled radon gas [3]; acute and
chronic studies of inhaled allergens [4]; safety and activity of
bronchodilator therapy [5]; the direct carcinogenic effects of
inhaled cigarette smoke [6]; and the potential health risks
associated with second hand cigarette smoke.
To date the use of dogs in biomedical research has
largely focused on research animals, i.e. Beagle dogs. A
largely un-used model are pet dogs with naturally occurring
disease. The spontaneous development of disease in the pet
animal population provides significant opportunities for
translational research and human drug development. Studies
that include pet dogs included the evaluation of novel
treatments for infectious diseases the optimization of
techniques for wound healing, gene therapy to modulate
hereditary immune disorders, and the study of novel cancer
therapies [7-11].
*Address correspondence to this author at the Comparative Oncology
Program, Center for Cancer Research, National Cancer Institute, National
Institutes of Health, Bethesda, MD 20892; E-mail: khannac@mail.nih.gov
NATURALLY OCCURRING CANCERS IN PET
DOGS
Inbred rodent models and laboratory derived canine
populations have been the primary population of
experimental animals used for the preclinical development of
cancer therapeutics. Working with inbred populations in
controlled, artificial laboratory environments raises some
concern regarding the applicability of data as it relates to the
treatment of cancers in people. Many of these concerns may
be allayed through the study of naturally occurring tumors in
our companion animal population, (i.e. dog and cat pet
population). Companion animals with naturally occurring
tumors may provide an excellent opportunity to investigate
many aspects of malignancy from etiology to treatment.
Several aspects of companion animal disease make for
attractive comparative models. Companion animals share a
common environment with people and represent a more
natural outbred population. Exposure to environmental
carcinogens should, therefore, be similar to that of people.
Malignancies in companion animals develop spontaneously,
whereas experimental laboratory models utilize induced
tumors either through exposure to known carcinogens or
transplantation, often in the context on an
immunocompromized animal. The relative abundance of
cancer in companion animals provides a potentially large
population of cases for investigation. Over half of all
households in the United States include a companion
animal. This represents approximately 55 million dogs and
60 million cats at risk for developing cancers in the U.S.
[11] Being resistant to atherosclerosis-associated
cardiovascular disease, cancer is the number one cause of
death overall in dogs. In a necropsy series of 2000 dogs,
23% of all dogs, regardless of age, and 45% of dogs 10
years of age or older died of cancer. Estimates of age-
adjusted overall cancer incidence rates per 100,000
individuals/years at risk range from 243 to 381 for dogs and
156 to 264 for cats. [12] These rates are comparable to those
reported by the National Cancer Institute SEER program for
human beings (approximately 300 per 100,000).
Additionally, incidence rates for certain malignancies in
companion animals (e.g., canine osteosarcoma, non-

Page 2

266 Current Cancer Drug Targets, 2003, Vol. 3, No. 4Khanna and Vail
Hodgkin’s lymphoma [NHL]) are higher than those observed
in people, which may provide a large population for study.
Conversely, certain neoplasms, such as hemangiosarcoma
and mast cell neoplasia, are extremely rare in humans but
abundant in companion animals, and allow meaningful
clinical data to be generated for tumor types with “orphan”
status in humans. [13-15] Table I presents a list of cancers
seen in cats and dogs that may be relevant models for the
study of the same human cancers. Tumors in companion
animals generally progress at a more rapid rate than their
human counterparts. This time-course is both long enough
to allow comparison of response times, but short enough to
ensure rapid collection of data. Companion animal cancers
more closely resemble human cancers than rodent models in
terms of size, cell kinetics, and biologic variables such as
hypoxia and clonal variation. [16-18] By virtue of their body
size, sample collection (i.e., serum, urine, cerebrospinal
fluid, multiple tissue samples), surgical interventions, and
imaging are more feasible than in rodent models.
Table I. Naturally Occurring Cancers Seen in Pet Animals
that may be Relevant Models of Human Cancer
Conditions
Cancer/Model Histology Species
Osteosarcoma Canine
Non-Hodgkin’s LymphomaCanine
Prostate Carcinoma Canine
Mammary CarcinomaCanine, Feline
MelanomaCanine
Lung CarcninomaCanine, Feline
Head and Neck Carcinoma Canine, Feline
Soft Tissue SarcomaCanine, Feline
Bladder Carcinoma Canine
Renal CystadenocarcinomaCanine
C-kit driven cancers
(canine mast cell tumor – Human GIST)1
Canine
Brain Tumors Canine
Lymphoma (Gut associated)
1 Similar juxtamembrane mutation in c-kit is seen in canine mast cell tumors and
human gastrointestinal stromal tumors. Canine mast cell tumors may represent an
important molecular model of a c-kit driven tumor.
Feline
In addition, an expanding animal rights movement is
making investigations with laboratory animals more
difficult. Provided that well-designed humane guidelines are
adhered to, clinical trials involving companion animals may
be more acceptable. Because the “standard of care” is not
established for many tumors encountered in the veterinary
profession, more latitude in prospective clinical trials is
allowable, and it is easier and morally acceptable to attempt
new and innovative treatment strategies. Importantly, such
latitude should not be abused, and present-day veterinary
institutions consistently use informed client consent and
institutional review boards to ensure study design and
ethical standards are maintained.
In general, clinical trials using veterinary patients can be
completed at significantly lower expense than similar human
clinical trials. Professional services, clinical pathology, and
diagnostic imaging, while of the highest quality, are of
lower cost in comparison to human clinical trials. Most
companion animal caregivers are highly committed, and are
actively seeking innovative and promising new therapies for
their companion’s cancer. As a whole, either through
personal experience with family members and friends, or
media coverage of available and leading-edge therapy, the
pet-owning public now demands the highest quality of care,
often only available through clinical trials. Compliance with
treatment and recheck visits is exceptional, and autopsy
compliance approaches 85%, significantly better than most
human clinical trials.
Finally, companion animal and human tumors share a
great deal in terms of etiopathogenesis and biology.
Mutations in many oncogenes and tumor suppressor genes
commonly seen in human cancer, such as p53. [19-28]
Rb,[21] Ras, [29-32] Myc , [19, 32] and bcl-2 [33, 34] have
been detected in a variety of canine and feline tumors.
Likewise, overexpression of telomerase [35-38] and matrix
metalloproteinases [39-42] have been detected in various
canine tumors. A variety of tyrosine kinase growth factors
and receptors have been detected in canine and feline tumors,
and thus they may serve as excellent models for the
preclinical development of small molecule inhibitors of
these growth factor pathways. Additionally, analogs of most
major angiogenic growth factors and their receptors exist in
dogs and cats, and elevations in serum or urine VEGF
and/or bFGF have been detected in several canine tumors.
[43-45]
EVALUATION OF ANTICANCER AEROSOLS IN
DOGS
Many of the advantages for the study of cancer in dogs
are particularly relevant to the evaluation of aerosol
approaches for the treatment of cancer. We have successfully
used pet dog cancer models to evaluate the delivery of
cytotoxic chemotherapy and cytokines by aerosol for the
treatment of primary cancers of the lung and cancers
metastatic to the lung. In both cases the use of the dog
models was essential for the development of each anticancer
aerosol therapy. A summary of preclinical studies undertaken
in dogs in the development of inhalation therapies are
presented below. The utility of dogs with naturally occurring
cancers in the translation and optimization of this therapeutic
modality to human cancer patients is emphasized.
DELIVERING
INHALATION
CHEMOTHERAPEUTICS BY
The concept of delivering chemotherapeutics by
inhalation for the treatment of lung cancer, either primary or
metastatic, has received little attention. The outcome for
treatment of primary and metastatic lung cancer in both
companion animals and people has not changed dramatically
in spite of the availability of new chemotherapeutic agents.
The reasons for treatment failure are diverse, but one
possibility may be the inability to deliver adequate drug

Page 3

Targeting the LungCurrent Cancer Drug Targets, 2003, Vol. 3, No. 4 267
Fig. (1). Inhalation of paclitaxol in dogs every 2 weeks (x 6) resulted in greater than 50% volume regression of a mammary gland
carcinoma pulmonary metastasis in a dog. (A) A lateral thoracic radiograph taken just prior to initiation of inhalation therapy. (B)
Lateral thoracic radiograph 14 days following a single treatment with paclitaxel by the inhaled route. Reprinted with permission from
Hershey, et al., Inhalation chemotherapy for macroscopic primary or metastatic lung tumors: proof of principle using dogs with
spontaneously occurring tumors as a model. Clin Cancer Res, 1999. 5(9): p. 2653-2659.
concentrations to the tumor site using systemic
administration. Pulmonary delivery of antineoplastic drugs
offers the theoretical advantage of achieving high local
pulmonary concentrations of drug while minimizing
systemic exposure. Thus the possibility exists to optimize
local action of chemotherapeutics with significantly lower
overall dose and fewer systemic side effects. Efficacy of
locoregional application of chemotherapy has been
demonstrated in a variety of cancers including liver, bladder,
and ovarian cancer. Isolated lung perfusion has been used
with limited success in a small number of patients, but is
unlikely to be widely used because it requires surgery. [46]
Few studies exist documenting the feasibility of
delivering antineoplastic agents by inhalation. Tatsumura et
al. evaluated inhaled administration of 5-fluoruracil (5-FU)
in dogs and achieved high concentrations of drug in trachea,
hilar bronchi, and regional lymph nodes within 2 hours of
treatment. [47] He also reported responses in 6 of 10 patients
with non-small cell lung cancer given aerosolized 5-FU.
Preclinical studies, performed in normal rodents and
dogs, have demonstrated the feasibility of delivering
chemotherapeutic agents by the inhalation route without
serious toxicity. [48] As part of an effort to determine the
safety and efficacy of pulmonary delivered antineoplastic
drugs, a study using pet dogs with primary and metastatic
tumors to lung was recently completed by our group. [48] In
this proof of principle trial twenty-eight privately-owned
dogs with spontaneously occurring primary or metastatic
lung tumors received newly developed proprietary
formulations of either paclitaxol (PTX) or doxorubicin
(DOX) by inhalation once every 2 weeks using an aerosol
device specially designed to capture all exhaled and fugitive
drug. The targeted dose delivery to the pulmonary tree was 3
mg of DOX and 40 mg of PTX per inhalation therapy event
and the amount of each drug delivered was controlled by the
duration (in minutes) of inhalation exposure based on the
dog’s stabilized minute volume of respiration under light
anesthesia. A total of 6 biweekly therapies were scheduled
per patient unless additional therapy was warranted based on
continued response.
Tumor regression was achieved in 25% of dogs with
measurable lung tumors and the delivery of efficacious doses
of PTX and DOX did not result in side effects normally
associated with systemic administration of these drugs. Five
of 25 dogs had partial responses and 1 dog had a complete
response to therapy (Fig. 1). Five of 6 responding animals
had metastatic sarcoma (4 were treated with DOX and 1 with
PTX) and 1 animal had metastatic mammary carcinoma
(treated with inhalational PTX). Three of the six responders
were dogs with pulmonary metastatic osteosarcoma (OSA)
and one dog demonstrated partial regression of metastatic
hilar lymphadenopathy, suggesting uptake of inhaled drug
into regional lymph nodes. Similar observations have been
made by Tatsumura et al., who demonstrated high
concentrations of 5-FU in hilar lymph nodes in dogs
undergoing inhaled administration of the drug. [47]
Similarly, exposure of the lung to DOX by isolated
perfusion was capable of achieving significantly higher drug
levels in lymph nodes and lung. [46, 49] These responses in
dogs with OSA are particularly noteworthy for two reasons.
First, systemic chemotherapy is virtually ineffective for

Page 4

268 Current Cancer Drug Targets, 2003, Vol. 3, No. 4Khanna and Vail
canine OSA once gross pulmonary metastases develop. [50]
Secondly, all 3 dogs with OSA exhibiting responses to
inhaled drug had received prior systemic chemotherapy; two
of whom demonstrated responses to inhaled DOX after
having received systemic DOX prior to development of their
pulmonary metastases.
Administered doses of drugs were sufficient to result in
regression of tumors, without associated toxicity. Acute side
effects of myelosuppression, nausea, and vomiting
frequently associated with intravenous administration of
PTX and DOX were not observed in any dog.
Cardiotoxicity, a cumulative dose-limiting effect of
intravenous administration of DOX, was not observed in any
dog receiving DOX inhalation alone. These findings are
consistent with the results of distribution of 14C-DOX
following intravenous or inhalation administration in normal
beagle dogsA. Levels of 14C in lung were demonstrated to
be 15-20 times higher following inhalation during the first
24 hours after dosing. By contrast, uptake of 14C in
systemic circulation and heart was significantly lower
following inhalation administration.
Acute local pulmonary effects of DOX were observed in
nearly 50% of the dogs and consisted of an intermittent,
nonproductive cough. These side effects were not dose
limiting in any animal, usually self-limiting and only rarely
required antitussive therapy. This side effect is likely related
to the direct effect of DOX on pulmonary tissues, although
the primary cancer and irritation from the endotracheal tube
may be contributing factors. Histologic examination of dogs
at necropsy revealed mild or moderate pneumonitis,
multifocal interstitial fibrosis or alveolar histiocytosis that
was not clinically evident. In contrast, doses up to 90
mg/m2 of paclitaxel could be safely delivered through this
route with no local toxicity. Mild systemic leukopenia was
observed only at the highest paclitaxel dose levels.
The alternate-week dosing schedule used in this trial was
chosen based on the cyclic dosing of cytotoxic drugs given
intravenously, in which myelosuppression is generally dose-
limiting. However, for inhalational chemotherapy, where
loco-regional responses and toxicity may be dose-limiting, a
schedule based on non-pulmonary drug effects may not be
appropriate. Further evaluation of dose and schedule are
warranted.
As the toxicities of inhalational chemotherapy are for the
most part locoregional, it can be theorized that concomitant
use of chemotherapeutics by both the inhaled route and the
intravenous route would not add to overall toxicity and may
enhance efficacy. To test this hypothesis 17 additional pet
dogs with splenic hemangiosarcoma (HSA) were treated with
concurrent inhalational and systemic chemotherapy. [45]
Splenic hemangiosarcoma (HSA) in the dog is an aggressive
malignancy with a high potential for widespread metastasis,
including abdominal viscera and the lungs. Two weeks
following splenectomy for their primary disease, dogs
received systemic DOX and cyclophosphamide at their
known maximally tolerated doses every 3 weeks for a total
A Zutshi, A., Townsend, R.W., Moutvic, R., Trigg, N., Watts, J.K., and
Imondi, A.R. Disposition of inhaled (IH) 14C-Doxorubicin (14C-DOX) in
dogs. Proc Amer Assoc Canc Res, 1999. 40: p. 416-421
of 4 treatments. Concurrently, DOX was administered via
the inhalation route to a targeted dose of 3 mg as described
above on the same day as systemic chemotherapy. No
increase in systemic toxicity grades (hemataologic,
gastrointestinal or cardiac) were observed following
combination therapy. While median survivals were superior
to a similar group of historical control dogs treated with an
identical systemic chemotherapy protocol without inhaled
chemotherapy [51], a randomized prospective trial would be
necessary to further evaluate the efficacy of combined
therapy. These results do suggest, however, that inhalation
chemotherapy combined with systemic chemotherapy may
offer some improvement in efficacy over systemic
chemotherapy alone without adding significant treatment
related toxicity. These studies, undertaken in dogs with
naturally occurring cancer have established the potential
activity and safety of this therapeutic approach for human
cancer patients.
INHALED CYTOKINE IMMUNOTHERAPY IN
DOGS
Cytokines including, interleukin-2, interleukin-12, GM-
CSF and others have demonstrated profound antitumor
activity in human and animal cancers [52-56]. A limitation
to the use of many cytokines, including interleukin-2 in
patients has been the development of dose dependent
toxicity [55, 57, 58]. Toxicities associated with intravenous
IL-2 have included pulmonary vascular leakage (most
common dose limiting toxicity), fever, weight gain and
anasarca, malaise, rigors, azotemia, anemia, and
thrombocytopenia [59, 60]. Adverse effects of IL-2 are
dependent on the dose, route of delivery, and formulation of
IL-2 [57, 61]. Novel approaches to reduce the toxicity of IL-
2 may increase the therapeutic index of this central cytokine
of the cell-mediated immune system and allow greater
application to the treatment of cancer. Given that a primary
problem in the management of cancer patients is the
development of pulmonary metastases the evaluation of the
aerosol route of delivery of cytokine based immunotherapy
held the promise of increased local delivery of interleukin-2
to limit toxicity and potentially improve local
biodistrtibution of IL-2 to the lung following inhalation.
Huland et al have demonstrated the clinical activity inhaled
IL-2 for patients with metastatic renal cancer [62, 63]. The
optimization of this therapeutic approach from the
standpoint of dose, regimen (qd, bid, tid) and aerosol
formulation (free IL-2 or liposomal IL-2) was necessary. To
answer these questions we initiated a series of preclinical
studies that included the study of pet dogs with naturally
occurring cancers.
In unpublished data we used a mildly immunogenic
murine pulmonary metastases model, the MCA 106
pulmonary sarcoma model, to evaluate the potential anti-
tumor activity of inhaled IL-2 and IL-2 liposomes. Groups
of 10 mice were then injected with 5 x 105 tumor cells to
the tail vein. On the day of tail vein injection, mice were
treated in groups of 10 in a plastic inhalation chamber with
aerosols of 1.0 x 106 IU free IL-2 (n=10), 1.0 x 106 IU IL-2
liposomes (n=10), and saline alone (n=10) twice daily.
Survival of mice treated with aerosols of IL-2 (free or

Page 5

Targeting the Lung Current Cancer Drug Targets, 2003, Vol. 3, No. 4 269
Fig. (2). Pet dogs were easily trained to receive nebulization therapy. Nebulization was accomplished using a standard compressor
and a Puritan Bennet Twin Jet nebulizer connected to a polyurethane rebreathing bag. Nebulized interleukin-2 formulations filled the
rebreathing bag and allowed inhalation therapy of awake dogs. Treatments were performed twice daily for 20 minutes. Treatment was
administered under active ventilation to limit secondary inhalation by the human handler.
liposomal) was significantly prolonged compared to the
saline treated mice.
The murine model confirmed the potential antitumor
activity of inhaled IL-2; however, it was not possible to
reliably predict delivered dose and inhaled dose, investigate
intermediate longitudinal endpoints of immunologic
activation, and define relevant toxicities associated with
inhalation. The dog was thus considered to be an appropriate
animal (model) for further investigation of the safety and
biological activity of inhaled IL-2 liposomes. The potential
to investigate antitumor activities associated with inhaled
IL-2 cytokine immunotherapy in dogs with intact immune
responses and relevant tumor-host microenvironment was a
particularly attractive component of the use of the dog.
Initial studies in normal research dogs defined the
feasibility, biological activity and pulmonary biodistribution
of aerosols of IL-2 delivered to dogs [64]. A major difference
between the evaluation of inhalation chemotherapy (as
discussed above) and inhalation cytokine therapy is the
regimen of therapy. Based on work with inhaled IL-2 in
human patients, multiple daily treatments with cytokine
were considered to be necessary over long periods of time
(days to months). This repeated delivery regimen required
the training of dogs to accept inhalation while awake rather
than though closed capture anesthesia. Using a polyurethane
re-breathing bag that covered the dog’s muzzle, normal
research dogs were trained to receive repeated inhalation
treatments (Fig. 2). Within 7 days all dogs were easily
trained to receive inhalation treatments that lasted 15-30
minutes. We later found this same delivery mechanism to be
effective in the training of pet owners to deliver cytokines to
their cancer bearing pet dogs. Most of the delivered aerosol
entered the lung following oral inhalation (i.e. during mouth
breathing); however, the re-breathing bag was associated
with aerosol dose loss that resulted from particle trapping in
the rhinarium of the dog. These losses are thought to be
slightly greater than upper airway losses associated with
nebulization in human subjects. Alternative strategies have
been employed for the delivery of aerosols to non-
anesthetized spontaneously respiring (awake) dogs. Most
techniques have included tracheostomy implants that could
be used to deliver aerosols by nebulizer or metered dose
inhaler [65-67]. Such techniques are best suited for
experimental dogs and may not be as useful for aerosol
delivery to pet dogs whose owners may not be comfortable
with such surgical procedures.
A significant advantage of the use of dogs to assess the
biological activity of inhaled cytokine immunotherapy was
the fact that longitudinal collection and analysis of
bronchoalveolar leukocyte effectors before and after exposure
to aerosols was possible. The collection of such biological
samples would be difficult in other animal models of
inhalation and not easy to justify in human clinical trials.
These biological endpoints allowed the optimization of
inhalation regimen, IL-2 dose, and selection of the most
active IL-2 formulation for delivery. Biological studies in