Ultrasonography of Osteosarcoma in Dogs
Sonography Atlas of Dogs

Abstract

Histology shows minimal airway and lung toxicity from aerosol gemcitabine

Mild histological changes were present within the sections of the examined airways that could be directly attributed to aerosol gemcitabine. Four animals exhibited minimal to mild expansion of the submucosa of one or more of their larger conducting airways. The submucosa was thickened by loosely arranged highly vascular connective tissue, which formed short blunt papillary projections that protruded into the airway lumina. No similar lesions were noted among the nontreated animals. Alveoli remained normal in both aerosol-treated and aerosol naïve dogs (Fig. 1A and B). Seven animals, all of which had either pleural effusion or pleural metastases, had marked regionally extensive to generalized chronic proliferative villous pleuritis with prominent, congested vasculature. Six animals had pleural effusion and six animals had pleural metastases; however, only five (of 12) animals possessed both lesions concurrently. In one dog with pleural metastasis, the lungs remained free of pulmonary metastasis. Three dogs with pulmonary and or pleural metastasis developed hypertrophic osteopathy, which was identified antemortem and confirmed postmortem. Alveolar lumina immediately adjacent to metastatic foci were expanded by increased numbers of foamy alveolar macrophages, which frequently possessed abundant intracytoplasmic hemosiderin pigment, indicative of prior hemorrhage. Metastasis was identified in anatomical locations outside of the lungs. These locations included liver (n = 3), spleen (n = 3), kidney (n = 3), adrenal (n = 3), other bones (n = 3), eye (n = 1), and skin (n = 1).

Additional histopathological findings in the lungs (data not shown) included multifocal lymphoplasmacytic perivascular inflammation (n = 2), intravascular metastases (n = 3), mild chronic bronchitis (n = 2), peribronchial gland hyperplasia (n = 2), and acute focally extensive neutrophilic interstitial pneumonia (n = 1).

Fas expression and apoptosis in spontaneously arising canine osteosarcoma and pulmonary metastases

We examined Fas expression in both the primary tumor (Fig. 2a) and the metastatic foci (Fig. 2b and c) of spontaneously arising canine osteosarcoma. Canine osteosarcoma ranged in the expression of Fas as detected and measured by immunohistochemistry staining intensity (Fig. 2a). Fas staining intensity did not differ between the primary tumors obtained from aerosol gemcitabine-treated and -naïve dogs (p > 0.05; Fig. 2a, compare upper and lower rows). Interestingly, Fas staining intensity was decreased in pulmonary metastases (b) compared to the primary tumor (a), and this difference was statistically significant (p =  0.008; photomicrographs: compare a and b; scatter plot: compare Bone and Untreated). Importantly, the Fas staining intensity in pulmonary metastasis from aerosol gemcitabine-treated animals (c) was greater than that found in both the primary tumor (p = 0.025; photomicrographs: compare a and c; scatter plot: compare Bone and Aerosol Gemcitabine) and in the gemcitabine-naïve metastases (p = 0.0075; Fig. 2, photomicrographs: compare b and c; scatter plot, compare column untreated metastasis and gemcitabine).

The FAS/FASL pathway kills metastatic osteosarcoma cells by inducing apoptosis.^(41,48–51) To determine if the observed increase in Fas expression resulted in an increase in death of tumor cells by apoptosis, we performed the TUNEL assay (Fig 3). Minimal TUNEL staining was identified in untreated pulmonary metastasis (a). In contrast, TUNEL staining was increased in osteosarcoma cells in the pulmonary metastasis obtained from aerosol gemcitabine treated dogs (b), and this difference was significant (p = 0.028).

Discussion

The development of pulmonary metastases is the most common form of treatment failure in both dogs⁽¹⁾ and humans⁽²⁾ with osteosarcoma. Effective treatment for lung metastases remains a clinical challenge. Therefore, the development of aerosol chemotherapy, which targets delivery to the respiratory system where metastases are most likely to be found, may provide a noninvasive method of therapy for pulmonary metastasis. In this study, we evaluated the novel delivery method of gemcitabine as an aerosol to dogs harboring gross metastasis of osteosarcoma in an effort to determine the preclinical safety and to assess the role of the Fas/FasL pathway in this model system. This is believed to be the first in vivo investigation of the antitumor effects of aerosol gemcitabine in spontaneously arising osteosarcoma in any species.

Previously, aerosol gemcitabine chemotherapy has been used in experimental murine models of osteosarcoma^(41,42,48) and in vitro in a lung cancer model.⁽⁴³⁾ As had been the case in murine^(41,42,48) and baboon studies,⁽⁴³⁾ aerosol gemcitabine was well tolerated in our cancer-bearing dogs. Monitoring of subjective (appetite, attitude, activity levels) and objective (body weight, CBC, and SBA) parameters identified only a single episode of Grade I neutropenia, which resolved without intervention and did not recur. Because these dogs were not trained respiratory research animals, we chose to measure arterial blood gas and alveolar–arterial gradients as a surrogate end point for pulmonary function. We identified no decrease in pulmonary function as a result of aerosol gemcitabine. Taken together, these findings confirmed that aerosol gemcitabine is systemically well tolerated, and suggested that local toxicity (manifest as changes in pulmonary function or radiographic appearance) would also be minimal. To more directly assess local toxicity, tissues from aerosol-treated dogs were evaluated by two veterinary pathologists (T.C., D.W.W.). Multiple sections of tissue from all levels of the respiratory tree, including the proximal and distal trachea, main stem bronchi, terminal airways, and alveolar sacs (Fig. 1A and B) were evaluated for evidence of toxicity associated with inhalational exposure to gemcitabine chemotherapy. There was no evidence of a direct toxicological effect associated with aerosol gemcitabine treatment in tumor-free regions of the lung. Although the submucosal expansion present within four of the animals was minimal, no similar lesion was noted among the nontreated animals. There was no apparent association between these findings and total cumulative dose of aerosol gemcitabine (i.e., duration or number of treatments or 25-mg vs. 50-mg group).

The remaining histopathological findings are considered incidental or due to progression of the metastatic process. The most notable adverse lesion was that of moderate to marked intrapulmonary and pleural hemorrhage secondary to necrosis of the metastases; although it is possible that aerosol gemcitabine was responsible for this finding, the progression of the disease was considered a more likely etiology. These clinicopathological and pathological findings confirmed and extended the murine studies^(41,42) in the tumor-bearing dog. This study demonstrated that aerosol gemcitabine is well tolerated both by local tissues and by the body as a whole.

The role of the Fas/FasL pathway in the metastatic potential of osteosarcoma in murine models has been investigated.^(41,50,51) In the primary tumor cells, cell surface Fas expression was present but was lost in the pulmonary metastases.⁽⁴¹⁾ Studies have demonstrated that Fas-transfected tumor cells express higher levels of Fas on the cell surface and are rendered more sensitive to Fas-induced cell death than parental lines; as a result, these Fas expressing cells produce significantly fewer pulmonary “metastatic” nodules than nontransfected cells.⁽⁵⁰⁾ The Fas/FasL system was also noted to be corrupted in pulmonary osteosarcoma metastasis collected from human surgical specimens, where 32% were weakly positive and 60% of the specimens were negative for Fas staining.⁽⁴⁸⁾ These data suggested that loss of Fas cell surface expression may aid osteosarcoma in the evasion of local lung resistance and the establishment of metastatic pulmonary nodules. We were interested in determining if this pathway was also involved in our canine model.

As was the case in mouse and human, the canine primary tumors expressed cell surface Fas, which was decreased or lost in the chemotherapy-naive metastatic foci. In contrast, in response to treatment of dogs with aerosol-gemcitabine there was an increase in Fas cell surface staining intensity in the metastatic pulmonary foci (Fig. 2) and an increase in apoptosis as measured by TUNEL (Fig. 3). Other nucleoside analogues such as nelarabine⁽⁵²⁾ and cytarabine⁽⁵³⁾ or other anticancer drugs at therapeutic concentrations such as doxorubicin, methotrexate, cisplatin, and etoposide^(53,54) have also been shown to increase Fas expression.

Necrosis, however, is not a hallmark of Fas-mediated cell killing. The observed central necrosis with viable peripheral tumor cells and no inflammatory response has been observed in lesions resected following salvage chemotherapy or tumor necrosis factor (TNF) treatment.^(55,56) In contrast to those findings, we found an increased number of inflammatory cells immediately adjacent to the metastatic foci. This inflammatory response could be secondary to the increased incidence and severity of intratumoral necrosis noted within the treated animals. Previous studies in mice have evaluated the histopathological changes in metastases after only 4–12 treatments (2–6 weeks of twice weekly treatment).^(41,42) In the human studies pulmonary inflammation was either not evaluated⁽⁴⁸⁾ or the metastatic tissues were removed after a single dose of salvage chemotherapy.⁽⁵⁶⁾ In contrast, our patient dogs received a median of 28 treatments (14 weeks of twice weekly treatment). Therefore, it is possible that with prolonged therapy the initial Fas-mediated apoptosis of tumor cells in the metastatic foci is eclipsed by progressive destruction of the metastatic nodule so that the predominant pathological aspect within the foci would be necrotic with attendant inflammation, as was seen in our dogs. As was the case in the local tissues, there was no apparent association between these findings and total cumulative dose of aerosol gemcitabine (i.e., duration or number of treatments or 25-mg vs. 50-mg group). The study was not powered to identify an association between dose and degree of response.

It is also possible that both/either the gemcitabine chemotherapy itself and/or the gemcitabine-induced upregulation of Fas in the FasL-rich environment of the lung resulted in the marked necrosis of the metastasis. The central necrosis could also be a direct effect of gemcitabine on the tumor neovasculature.⁽⁵⁷⁾ Previous studies in the baboon have demonstrated marked variability in gemcitabine deposition in the lung both between therapy sessions and between subjects.⁽⁴³⁾ We did not measure plasma levels or deposition of gemcitabine or the accumulation of gemcitabine triphosphate (the active metabolite) in the various areas of the lung in these patients because all had last received therapy more than 2 days prior to euthanasia and because the levels would likely be extremely low. Variability in pulmonary deposition could also account for the marked variability in both the observed, FAS expression, TUNEL staining, and necrosis. Objective measure of the deposition of aerosolized gemcitabine in the dog lung in normal dogs is being considered in an effort to be able correlate degree of biological response with any local toxicity or tumor response.

Metastatic osteosarcoma carries a poor long-term prognosis in the dog. The median time (days) to metastasis [disease-free interval (DFI); 223] and the overall survival (OS; 289) of approximately 700 dogs with osteosarcoma in relapse for which no treatment was performed can be estimated (66 days ± 44) from the literature. Our cohort of 21 dogs had a median DFI (139), survival in relapse (79), and OS (262), which were similar. Although aerosol gemcitabine demonstrates FAS-mediated antitumor properties against metastatic canine osteosarcoma, none of these dogs had radiographic resolution of their disease and these survival numbers underscore our realization that this novel delivery will need to be combined with other modes of therapy to maximize survival both in this relevant large animal model and in children with metastatic osteosarcoma.

The metastatic cascade is a complex process requiring multiple steps that necessitate the perfect union of host, tumor, and local microenvironmental factors. We acknowledge the important and necessary complementary roll played by local FasL in the eradication of metastatic osteosarcoma. In gld mice (that lack normal FasL function), aerosol gemcitabine had no effect in eradicating metastases from the lung. In the case of our dogs, we identified metastatic foci outside of the lung in the relatively FasL poor locations such as liver, kidneys, and skin. Although Fas was not specifically evaluated in these extrapulmonary foci, their presence speaks to the need for additional treatment strategies in combination with aerosol gemcitabine in the management of metastatic osteosarcoma.

In conclusion, the aerosol method of gemcitabine administration was very well tolerated regionally and systemically by our canine patients. The results from this pivotal translational study in a highly relevant immune competent large-animal tumor model of human osteosarcoma support the notion that aerosol gemcitabine may be useful against pulmonary metastasis of osteosarcoma. Although there were no cures, because there was a demonstrated histological tumor response, additional studies that evaluate the aerosol route of administration of gemcitabine as a form of therapy in metastatic pediatric osteosarcoma combined with other metastasis-effective, FasL-inducing treatments such as IL-12 or ifosfamide; or immunity modulators, such liposome encapsulated muramyl tripeptide, are warranted.

Acknowledgments

This work was funded by a Center for Companion Animal Health Grant 06-07 (C.O.R.) and National Institute of Health Grant CA42992 (E.S.K.). C.O.R. would like to thank Charlotte Marra for her generous support, and is indebted to the guardians of these beloved companions for their tireless effort; without their dedication, this work would not have been possible. P.M.A. acknowledges Research Support from the University of Texas M.D. Anderson Cancer Center, and is grateful for additional research support from the Thorrison, Rosen, Belvo, Haynie, Meares, Silberblatt, Brodie, Jager, and Lash families.

Author Disclosure Statement

No conflicts of interest exist.

References

Share this article / Teilen Sie diesen Artikel