Oncology
Spontaneously occurring canine cancer as a relevant animal model for developing novel treatments for human cancers
2021 年 3 巻 2 号 p. 51-59
詳細
2021 年 3 巻 2 号 p. 51-59
The most common cause of death among adult or older dogs is malignancy. Dogs spontaneously develop malignant tumors with age, and the incidence of malignant tumors in dogs is higher than that in humans. Basic research on treatment of tumors is generally conducted using syngeneic or immunodeficient mice, which have some shortcomings as a model for human tumors that develop over a long period of time, while interacting with the host immune system. On the other hand, naturally occurring canine tumors develop under immunocompetent conditions and may be suitable for preclinical studies, especially for the development of therapies that affect host immunity. In this review, some of the canine tumors and immunotherapies that have been implemented thus far will be discussed, with the intent of helping establish cancer treatment research, using dogs with naturally occurring tumors, for translational studies in humans.
● Naturally occurring canine malignancies can be excellent spontaneous animal models of human cancer.
● Unlike mouse tumor models, canine tumors have some similarities to human tumors, as they grow over time in interaction with host immunity.
● Novel immunotherapies used to treat canine cancers, as a translational research animal model, may provide useful information for treatment of cancer in humans.
In pet veterinary medicine, the annually increasing incidence of malignant tumors is of great concern [1]. In particular, dogs develop tumors at an incidence of more than 1,000 per 100,000 dog-year [2], which is a higher rate than in humans. As in humans, the incidence of tumors increases with age, and malignant tumors account for approximately one-half to one-third of all deaths, especially in older dogs.
Tumors in dogs have some interesting characteristics. The frequency of each type of tumor differs from that observed in humans [3]. In humans, epithelial tumors such as breast, lung, and colon cancers are common, and sarcomas are relatively rare, whereas in dogs, lymphomas and sarcomas are relatively common, rather than epithelial tumors [3]. Another characteristic is that the incidence and types of malignant tumors vary considerably by breed [3,4,5]. As shown in Table 1, the occurrence of certain tumors is significantly more frequent in certain breeds of dogs [6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22]. These trends sometimes vary from country to country, and the fact that the incidence varies with genetic background even within the same breed suggests that the occurrence of tumors in dogs itself strongly reflects a genetic background.
| Tumor | Canine breeds with high incidence | References |
|---|---|---|
| Osteosarcoma | Irish Wolfhound, Rottweiler, Great Dane, Greyhound, Saint Bernard | [7] |
| Histiosarcoma | Bernese Mountain Dog, Flat-Coated Retriever | [8,9,10] |
| Hemangiosarcoma | Boxer, German Shepherd, Golden Retriever | [11,12,13,14,15] |
| Mast cell tumor | Boxer, Bulldog, Bullmastiff, Boston Terrier, Golden Retriever, Staffordshire Bull Terrier | [16, 17] |
| Lymphoma | Boxer, Bulldog, Bullmastiff, Golden Retriever | [15, 18, 19] |
| Cutaneous melanoma | Schnauzer | [20] |
| Oral melanoma | Chowchow, Golden Retriever, Poodle | [20, 21] |
| Bladder cancer | Beagle, Scottish Terrier, Shetland Sheepdog, Wire Hair Fox Terrier, West Highland White Terrier | [22] |
The clinical and histopathological similarities and differences between canine tumors and their human counterparts have long been investigated in comparative oncology [23]. Recently, the development of next-generation sequencing techniques has made it possible to perform transcriptome and exome analyses of canine tumors, and there have been an increasing number of reports of molecular comparisons between counterparts of canine and human cancers [24]. For example, the phosphatase and tensin homolog deleted on chromosome 10 (PTEN) gene mutations and the activation of the phosphatidylinositol-3 kinase (PI3K) signaling pathway have both been observed in human angiosarcoma and canine hemangiosarcoma [25, 26]. Diffuse large B-cell lymphoma (DLBCL) is the most common type of lymphoma in dogs and humans, and transcriptome analysis of canine DLBCL revealed that the activation of the Nuclear factor κB (NK-κB) is similar to the process associated with activated B cell subtypes in human DLBCL [27].
As mentioned above, dogs have many naturally occurring tumors, and the incidence of tumors in different breeds of dogs varies significantly. Therefore, analysis of the genetic background of each breed may be useful for analyzing rare tumors in humans. In addition, as will be discussed later, diagnostic devices and treatment methods can be used in the ways similar to those for human patients with cancer, as opposed to the difficulties associated with mouse models. Because the immune system of dogs is similar to that of humans, tumor development occurs in the presence of the action of host anti-tumor immunity, leading to tumor heterogeneity. Dogs also have the appropriate body size for use in investigating new therapies; therefore, pharmacokinetic and pharmacodynamic analyses are easier to perform. Furthermore, their short lifespan makes it possible to evaluate the long-term effects of new cancer treatments in a shorter period than in humans. All these factors make canine cancer an ideal animal model for human cancer studies.
For dogs with malignant tumors, veterinarians can perform examinations similar to those available for detection of human cancer, and the advances in medical care for these pets have been remarkable [28]. In veterinary hospitals, for example, blood tests are used to assess the general condition of the animal, and imaging tests such as X-ray, ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET) scans are used to reveal the presence of tumors [29]. Tumor samples can be collected through biopsies, and histopathologic techniques can be performed to make definitive diagnoses. Tumor samples may then be tested for genetic mutations to determine if molecularly targeted drugs are suitable for the treatment of the tumor [30], or the analysis of cell surface antigens can be done using a flow cytometer to determine the presence of different tumor subtypes [31].
As in humans, three main therapies have been utilized for treating dogs with malignant tumors: surgery, radiation therapy, and anticancer drug therapy. For localized tumors, in addition to surgery, radiotherapy is an effective treatment, and the number of facilities that can provide treatment using devices such as linear accelerators is gradually increasing. For tumors that have metastasized or are particularly sensitive to anti-cancer drugs, such as hematological cancers, anti-cancer drug therapy is often the treatment of choice. One of the major differences between people and dogs is that dogs have a shorter life span, so owners often do not choose aggressive treatments that aim for a complete cure, as people do; rather, owners choose anti-tumor therapies that prioritize maintaining the animal’s quality of life. In addition, unlike human tumors, it is difficult to detect tumors in animals at an early stage, and when they are detected, they are often at an advanced stage, so the three major therapies alone are often insufficient for completely curing cancer in many cases. Therefore, there is an urgent need for new therapies to supplement or augment existing therapies.
Syngeneic mouse models, genetically modified mouse models, and xenograft mouse models using immunodeficient mice have been used for a long time to analyze the mechanisms of tumorigenesis and to develop therapeutic methods, as these models have revealed many tumor mechanisms [32]. In particular, the contribution of individual genes to tumorigenesis has not been clarified in many cases without the use of genetically modified mouse models. However, the development of therapeutic drugs using mouse models has its limitations, wherein the success rate is yet to reach 10%, which is a major problem [33]. In recent years, with the success of immunotherapy, the importance of host immunity in tumor development and growth has become clear, as exemplified through the concept of immunoediting [34]; thus, the limitations of using mouse models have become more apparent. In particular, xenograft mouse models use immunodeficient mice, making it difficult to study the interaction of cancer cells with the host immune system. Although syngeneic mouse models retain their immune system, they do not reflect the conditions of human tumorigenesis because they use uniform tumor cell lines to induce tumor formation within a short period of time.
In contrast to using artificial rodent models for tumor development, canine tumors develop spontaneously with age but in a shorter period of time than humans and with constant interaction with host the immune system. The diagnostic and therapeutic devices used in canine models are similar to those used in humans, and as dogs are larger than rodents, it would be easier to perform hematological and imaging monitoring of various clinical parameters during treatment. It is difficult to obtain sufficient amounts of blood samples for therapeutic monitoring in rodents or to monitor changes in tumor tissue in individual animals over a period of time through continuous biopsies. Dogs have a shorter lifespan than humans, so long-term monitoring of novel treatments can be performed in a relatively short period of time. In addition, malignant tumors that are considered rare cancers due to their low incidence in humans often have a relatively high incidence in dogs. For example, osteosarcoma is the most frequent tumor that occurs in the bones of both humans and dogs [35]. It is classified as a rare cancer in humans due to its incidence; however, its incidence in dogs is 10–50 times higher than that in humans [35]. There are also some differences between canine and human cancers, such as in osteosarcoma, where its incidence increases with age in large dogs with long legs, while in humans it is generally diagnosed in children and young adults, although some common pathogeneses have been observed [36]. Hence, many new treatments for osteosarcoma in humans have been used in trials for osteosarcoma in dogs [37]. Furthermore, in the United States, the Morris Animal Foundation, QuadW Foundation, and National Cancer Institute’s Comparative Oncology Trial Consortium are collaborating on the Osteosarcoma Project, which aims to find novel therapies to suppress metastasis using naturally occurring canine osteosarcoma. Recently, this project reported a multicenter, randomized, prospective clinical trial focused on the use of a mammalian target of rapamycin (mTOR) inhibitor, sirolimus in 324 dogs with osteosarcoma treated with amputation and chemotherapy. However, the results showed no advantage with the use of sirolimus, but this is a good example of a preclinical clinical trial for drugs aimed to treat osteosarcoma [38].
The main advantage of dogs as an animal model for naturally occurring cancers is that they have a complete immune system and they can develop malignant tumors spontaneously over time, just as humans do. Therefore, they are suitable for the development of new drugs as animal models with naturally occurring tumors rather than artificially created tumors in mice. Since dogs have an immune system like humans, analyses of immune cells can be carried out. Immunohistochemistry and flow cytometry have been used to analyze cells in the tumor microenvironment, which are particularly important for tumor progression. The tumor microenvironment is composed of many types of cells, including adipocytes, adipose-derived mesenchymal stem cells, cancer-associated fibroblasts, as well as soluble factors and lymphatics, in addition to the infiltrated immune cells. Gene expression in breast cancer cells is modulated by cancer-associated fibroblasts in canines and humans [39, 40]. The production of IL-8 in canine hemangiosarcoma and its effects on the tumor microenvironment, including angiogenesis, is important [41] and has been suggested to be similar in human tumors [42, 43]. Elevation of cyclooxygenase 2 (Cox-2), which induces the production of vascular endothelial growth factor (VEGF) [44], has been observed in canine and human prostate carcinoma and transitional cell carcinoma [45, 46]. Nonsteroidal anti-inflammatory drugs (NSAIDs), such as piroxicam and deracoxib, have been established as anti-tumor drugs in canine transitional cell carcinoma [47, 48]. As for immune cells, there have been many reports on the relationship between the percentage of cytotoxic T cells and the number of regulatory T cells in tumor tissues and disease prognosis [49,50,51], and the increased number of myeloid-derived suppressor cells (MDSCs) in the peripheral blood of tumor-bearing dogs has been reported [52,53,54]. Analysis of macrophages in tumor tissues from mammary gland tumors and osteosarcoma has also been performed [55,56,57]. Regarding the inflammatory response in tumors, they can be classified into immunological hot tumors and cold tumors in human cancer [58], and this inflammatory response type is known to be related to their response to immune checkpoint blockade therapy [59]. This immunological classification has also been applied to canine osteosarcoma [60].
Historically, as the treatment of malignant tumors in dogs has been developed by following treatment patterns in humans, various treatments using the immune system have been also attempted in dogs. This follows the expectation of the effects of immunotherapy in humans, although there are no immunotherapies that have shown significant effects in dogs thus far. Cytokine therapy in human using interleukin-2 (IL-2) and interferon-α (IFN-α) was approved for the treatment of melanoma and renal cell carcinoma, despite its limited efficacy and strong side effects [61]. However, to date, no veterinary cytokine preparations have been approved for the treatment of canine tumors. Although recombinant canine IFN-γ has been approved for the treatment of canine atopic dermatitis in Japan [62] and was used off-label with an expectation of exerting antitumor effects, there have been no associated reports of clinical antitumor effects, probably because the dose used was not sufficient [63]. On the other hand, the amino acid sequence homology of cytokines among different species is relatively high, and there are many examples of human cytokines acting on animals in vitro. In particular, recombinant human IL-2 instead of canine IL-2 preparations has long been used to stimulate T cells in dogs. Therefore, although there are some reports of clinical trials in which human IL-2 was used in place of canine IL-2 and showed some efficacy in spontaneously occurring cancers in dogs [64,65,66,67], it is not commonly used, probably owing to concerns regarding its long-term safety due to immunogenicity.
After recombinant human IL-2 became available, antigen-nonspecific T cell therapies, such as lymphokine-activated killer cells (LAK) and CD3-activated T cell (CAT) cell therapies, became available for human use in the 1980s [68]. CAT therapy has also been used in cancer-bearing dogs for some time; however, there is no clear evidence that they are effective. However, there is one report that showed autologous lymphocyte therapy activated with artificial antigen-presenting cells prolonged PFS and OS in dogs with lymphoma after remission with chemotherapy [69]. Therefore, we cannot deny the possibility that autologous lymphocyte therapy may be effective in preventing relapse in some human cancers [70, 71], depending on how it is used. Furthermore, it was recently reported that the use of activated lymphocytes along with autologous tumor cell-derived vaccines after leg amputation in dogs with osteosarcoma not only prolonged survival but also resulted in complete remission of metastatic disease in one case [72].
Since the identification of cancer antigens in 1991 [73], many therapeutic studies on cancer vaccines, especially peptide-based cancer vaccines, have been conducted in human medicine [74]. However, peptide-based cancer vaccines have not been used in dogs because not enough of antigens targeted by the vaccine in each type of tumor have been identified. Moreover, analysis of the dog leukocyte antigen (DLA) in each breed, which is important for the identification of peptides useful for each dog, has not been routinely implemented. To overcome this disadvantage, a plasmid DNA vaccine capable of expressing the full length of a tumor antigen, which can be used without DLA and peptide analysis, was developed. Among the few identified antigens in canine tumors, tyrosinase was selected as a target antigen for oral malignant melanoma (OMM), as well as for human and mouse melanoma. Subsequently, therapeutic agents targeting tyrosinase were approved as a treatment against OMM in dogs, prior to approval for usage in humans [75]. OnceptTM is a plasmid DNA xenogeneic vaccine that uses the human tyrosinase gene. The proof of concept of this vaccine was shown in mice in 1998 [76], followed by clinical trials in many canine cases of OMM [77, 78]. The results of these trials led to the approval of the vaccine in 2010 as an adjunct therapy to surgery and radiation therapy for stage II and III canine OMM. Many follow-up trials have been conducted since, but their efficacy remains controversial [79]. This is one example in which clinical trials for drugs using naturally occurring canine tumors were used to precede approval for eventual use in humans, but this DNA vaccine has not been approved for use in humans [80]. Following the success of OnceptTM, a DNA vaccine using the chondroitin sulfate proteoglycan 4 (CSPG4) gene, which encodes a cancer antigen, has been reported to be effective in canine OMM [81, 82], but it is yet to be commercially available.
In recent years, immune checkpoint inhibitor (ICI) antibody therapy and chimeric antigen receptor T (CAR-T) cell therapies have been attracting attention as new immunotherapies for human cancers, with several approved molecules and indicators and thousands of clinical trials being conducted worldwide [83, 84]. These could also hold potential for treating dogs, but in practice, canine cancer biology is less progressed, compared to human medicine, although research progress is continuously being made.
Although it has already been 20 years since the success of the first antibody drug in humans, the development of antibody drugs for dogs has been very slow, wherein the first canine antibody drug, an anti-canine interleukin-31 (IL-31) canonized antibody for the treatment of canine atopic dermatitis, launched in 2017 [85]. The reason for the delay in the development of antibody drugs for dogs is that antibodies are biological products and are highly species-specific; thus, it is difficult to use human preparations for dogs, as is the case with anticancer drugs such as chemotherapy drugs or small molecule inhibitors. Therefore, antibody drugs unique to dogs must be developed. In addition, the market size for veterinary drugs is extremely small compared to that for human drugs; hence, veterinary pharmaceutical companies have corresponding limitations in drug development. Although there are no antibody drugs for the treatment of canine tumors in the market, many groups, including academic laboratories and veterinary drug companies, are developing antibody drugs for canine tumors. In particular, several groups have reported the development of anti-dog programmed death-1 (PD-1) and anti-dog programmed death ligand1 (PD-L1) antibodies [86,87,88,89,90]. In each of these reports, it is clear that anti-dog PD-1 and anti-dog PD-L1 antibodies enhance T cell activation and that PD-L1 is expressed in a variety of canine tumors, such as malignant melanoma, nasal carcinoma, squamous cell carcinoma, osteosarcoma, mammary adenocarcinoma, and lymphoma [90,91,92]. IFN-γ has also been shown to enhance PD-L1 expression in canine tumor cell lines [86, 91, 93]. However, only our group [94] and Maekawa et al. [95, 96] have reported clinical trials using these antibody drugs in cancer-bearing dogs. The first clinical trial for this indication was conducted by the Hokkaido University group. In their pilot report [95], partial remission was observed in one of seven canine OMM cases and one of two undifferentiated sarcoma cases treated with anti-dog PD-L1 chimeric antibodies. Recently, they reported on a clinical trial wherein antibodies were administered to 30 dogs with stage IV OMM [96]. They found complete remission in 1 of 13 dogs that had ‘measurable diseases’ at baseline, and the resolution of lung metastases in 4 of 17 dogs with lung metastases that were classified as ‘non-measurable lesions’ at baseline, according to the canine response evaluation criteria for solid tumors (RECIST) v1.0. In our pilot clinical trials [94], we administered anti-dog PD-1 chimeric or canonized antibodies to 30 dogs with advanced-stage tumors who had completed standard therapy and had no other treatment options. Among the 30 dogs included, 15 had stage IV OMM and other types of tumors. Of the 15 dogs with OMM, 4 showed an objective response, and the median survival time of all 15 dogs was extended as compared with our historical control group, including the dogs that received conventional therapies. Although these reports mainly focused on OMM and found efficacy in some studies, we still need multicenter randomized clinical trials to prove the efficacy of ICIs in this disease. Notably, these clinical trials indicate that canine ICI therapy has the potential to treat cancer-bearing dogs, even at an advanced stage. It has been pointed out that canine OMM has different characteristics from human cutaneous melanoma [97], but it seems to be a good animal model of mucosal melanoma, a subtype of human melanoma categorized as a rare tumor due to its low incidence. Because of the high incidence of OMM in dogs compared to mucosal melanoma in humans, the evaluation of ICI therapy combined with other novel therapies could be implemented in dogs in the future. As described previously, DNA vaccines for canine OMM are commercially available; therefore, concurrent therapies using DNA vaccines and ICI could be another treatment option, which will provide valuable information for treating mucosal melanoma in human patients.
Antibody drugs produced for humans cannot, in principle, be used for other species such as dogs, due to the potential immunogenicity and cross-reactivity of the antibodies themselves, so the problem that research on dogs is less progressed compared to research on humans is significant. There are examples of early proof-of-concept studies in the use of human antibody drugs for the treatment of canine tumors, but immunogenicity issues still need to be considered. As mentioned above, the control of regulatory T cells in the tumor microenvironment is very important in the treatment of malignant tumors [98], and it is clear that this can enhance anti-tumor immunity in mice [99]. However, there are no established therapies to control regulatory T cells in patients with cancer [100]. The importance of regulatory T cells in the tumor tissues for the prognosis of canine tumors has been implicated previously [51,52,53], and there is a recent report that human antibody drugs may control regulatory T cells in dogs [101]. An anti-human CCR4 antibody drug (mogamulizumab) was developed for the treatment of human adult T cell leukemia [102] because the CCR4 molecule was originally expressed on leukemia cells of adult T-cell leukemia and this antibody killed the tumor cells. Maeda et al. [103] used this human antibody drug and found that mogamulizumab also bound to the CCR4 molecule on the regulatory T cells of dogs [101], which in turn reduced the number of regulatory T cells in cancer-bearing dogs. Based on these data, they showed that mogamulizumab was effective in shrinking tumors and prolonging life in dogs with transitional cell carcinoma of the bladder. The use of humanized antibodies in dogs also raises the issue of immunogenicity, but this proof of concept that the modulation of the immune system is effective in controlling naturally occurring tumors in dogs has the potential to progress into a translational study for humans because no therapy has yet been established to reduce number of regulatory T cells in patients with cancer. In addition to anti-CCR4 antibody therapy, the possibility of other therapies to deplete regulatory T cells, such as antibody therapies targeting other surface molecules or inhibitors of signaling molecules, has been proposed [100]. Preclinical studies in dogs would provide useful information for therapies yet to be tested in clinical trials.
CAR-T cell therapy, an effective new immunotherapy, uses immune cells. CAR-T cells are recombinant autologous T cells introduced with a single-chain variable fragment of an antibody against the antigens specifically expressed on tumor cells and the signaling domain of T cells, and their tumor specificity and potent cytotoxic activity make them a highly effective therapeutic agent, especially for hematological cancers. Although several types of CAR-T cell therapies have already been approved in humans [104], there are still no approved CAR-T cell therapies for dogs. However, several groups, including our own, have already reported on how to create canine CAR-T cells with some efficacy in vitro [105] and in vivo [106,107,108]. To maximize the effect of CAR-T cell therapy in patients, CAR-T cells must be administered after myelosuppression using strong anticancer drugs and radiation to eliminate the immune cells. However, there are several barriers to their common use in medicine. While chemotherapy and radiotherapy are routinely used in canine cancer treatment, procedures that require complete myeloablation are not common, because the control of severe infections is not as easy in dogs as it is in humans. Another concern is related to the Convention on Biological Diversity. Dogs administered with CAR-T cells produced in the conventional way using lentiviruses or retroviruses are classified as Living Modified Organisms, and need to be kept in a limited facility, at least until the safety of virus excretion can be guaranteed, in Japan. These are difficult limitations for translational research in the future. Nevertheless, the use of cancer-bearing dogs as a preclinical model for CAR-T therapy could make a significant contribution to human clinical trials. Currently, the efficacy of CAR-T therapy is observed mainly in hematological tumors in humans, and it is necessary to investigate methods to overcome this limitation, especially in solid tumors. In particular, as the number of patients with rare tumors, such as sarcoma, is small, proof-of-concept research may be accelerated using a large number of dogs with equivalent tumors.
As discussed in this review, the use of cancer-bearing dogs as a spontaneous tumor model for the development of human cancer therapies has potential to be a viable approach in some cases. The development of new drugs in human medicine generally takes a long time. In contrast, new treatments can be adapted relatively early as an advanced therapy for dogs with similar target tumors. This can be very beneficial for dogs and their owners who are affected by the types of tumors that have no established standard of care nor multiple treatment options. In addition, as mentioned earlier, in cancers that are rare in humans but occur frequently in dogs, the use of a new treatment in dogs prior to humans may reduce development costs by obtaining successful proof-of-concept studies in a shorter period of time. Despite these advantages, however, there are also some limitations. The number of researchers in canine oncology is limited, so there are not enough research tools, and in many cases, as in the case of ICIs and CAR-T cell therapy, the human medical field is more advanced. To solve this problem, human cancer researchers should pay more attention to canine cancer research and conduct research on cancer treatment together with canine cancer researchers. In the US, as shown by the Osteosarcoma Project, studies on canine cancer being carried out in collaboration with human cancer researchers have been conducted for quite some time, mainly by the National Cancer Institute [109, 110]. It is conceivable that the global implementation of such initiatives could also eventually lead to the promotion of human cancer research.
The authors have nothing to disclose.
