Ultrasonography of Osteosarcoma in Dogs
Sonography Atlas of Dogs

Toceranib phosphate1 is a receptor tyrosine kinase inhibitor (TKI) active against the split kinase family of tyrosine kinases including Kit, vascular endothelial growth factor receptor-2, platelet-derived growth factor receptor, and Flt-3.[1] Toceranib and structurally similar TKIs such as sunitinib demonstrate potent antitumor and antiangiogenic effects in mouse tumor models and in humans with certain malignancies.[2-4] Toceranib is currently approved for the treatment of mast cell tumors in dogs. The drug also appears to have biological activity against other tumors in dogs such as mammary carcinoma, soft tissue sarcoma, and anal gland adenocarcinoma.[1, 5]

In addition to their cytotoxic effects on tumor cells, recent data suggest that TKIs may also have immunomodulatory effects, particularly on key immune cells recruited by tumors such as myeloid-derived suppressor cells (MDSCs) and regulatory T-lymphocytes (Treg). These cell populations play central roles in immune suppression, and reversal of their effects is critically important to the success of cancer immunotherapy. For example, in a mouse tumor model, treatment with sunitinib decreased MDSC accumulation, prevented Treg development, and improved the efficacy of a tumor immunotherapy protocol.[6] In humans with renal cell carcinoma, sunitinib treatment appeared to reverse peripheral immune dysfunction, primarily by stimulation of a type-1 (Th1) T-cell-mediated immune response. Treg isolated from these patients demonstrated diminished suppressive function against autologous effector T cells.[7] Although the mechanisms for these effects are not completely understood, sunitinib and other TKIs appear to modulate Treg indirectly, largely through suppressive effects on MDSCs and tumor-associated macrophages.[8] Therefore, there is considerable interest in the immunomodulatory properties of TKIs and in their potential combination with cancer immunotherapy.

Because of their central role in tumor-induced immune suppression and correlation with a poorer prognosis in many human tumors, Treg represent an attractive therapeutic target. Although the clinical relevance of Treg is not yet known in dogs with cancer, Treg are present in increased numbers in the blood and regional lymph nodes of dogs with various malignancies and may be associated with outcome.[9, 10] Similar to sunitinib, the alkylating chemotherapy agent cyclophosphamide (CYC) also is known to selectively decrease Treg and inhibit their function in mice and humans with cancer. In these species the effects of CYC on Treg occur through both direct and indirect effects of the drug.[11-13] We recently demonstrated that daily oral administration of low-dose CYC to dogs with soft tissue sarcoma led to a selective decrease in the absolute number and percentage of Treg in peripheral blood, suggesting that CYC may have similar immunomodulatory effects on canine Treg.[14]

Low-dose CYC and TKI treatment are attractive treatment approaches for companion animals with cancer. This is largely because of the ease of oral administration of these drugs compared to conventional chemotherapy protocols. Combination treatment with low-dose CYC and toceranib has not been previously described, nor have the effects of toceranib on Treg and other T-lymphocyte subsets been determined. Therefore, we designed a prospective study to evaluate the effects of toceranib, administered as a single agent and then in combination with low-dose CYC, on circulating Treg numbers and percentages in dogs with various malignancies. In addition, we sought to determine the safety of combination toceranib and metronomic CYC treatment. We hypothesized that administration of toceranib would selectively decrease Treg numbers in the blood of dogs with cancer and that the addition of CYC would further enhance this decrease. We also expected the drug combination to be well tolerated in dogs with advanced disease.

Materials and Methods

In vitro Expansion of Treg

Canine lymphocytes were activated and cultured in IL-2 and TGF-β by means of a previously described protocol that expands a population of CD4+FoxP3+ T cells with phenotypic and functional properties consistent with Treg.[20] Briefly, PBMC were obtained from 3 healthy dogs and cultured at 2 × 10⁶/mL in media7 containing 10% fetal bovine serum and 1% each of penicillin/streptomycin and L-glutamine in 24-well plates. The cells were stimulated with recombinant human IL-2 and TGF-ß8 and Concanavalin A9 (ConA) for 5 days at 37°C in 5% CO₂. After 4 days of culture, toceranib was added to some of the cultures at either 10 or 100 ng/mL. These concentrations were chosen based on the range achievable in the serum of dogs receiving the drug.[22] Twenty-four hours later, the cells were harvested, stained with CD4-FITC and Foxp3-PE, and analyzed by flow cytometry as described above.

Study Patients

A total of 7 male and 8 female dogs were enrolled in this prospective clinical trial. Two of the male dogs and 1 of the female dogs were intact and the remaining dogs were neutered. Breeds included 2 Labrador Retrievers, 1 Rottweiler, 1 Boxer, 1 Saint Bernard, 1 Norwegian Elkhound and 1 Miniature Schnauzer, with the remainder being mixed-breed dogs. The median age was 10 (range, 2–13) years and median body weight was 28.9 (range, 11.3–51) kg. Nine of the 15 dogs received 1 or more medications for pain during the study period (including carprofen, tramadol, gabapentin, amantadine or some combination of these) and 6 dogs received maropitant or metronidazole as needed to manage nausea or diarrhea, respectively.

Tumor types included osteosarcoma of the appendicular or axial skeleton in 3 dogs and 1 dog, respectively, soft tissue sarcoma in 4 dogs, and 1 each of hemangiosarcoma, lymphangiosarcoma, malignant histiocytosis, transitional cell carcinoma (TCC) of the urethra, thyroid carcinoma, anal sac adenocarcinoma, maxillary carcinoma, and carcinosarcoma of the perineum. One dog had 2 separate tumors (a urethral TCC and a soft tissue sarcoma of the body wall). At the time of study initiation, 12 of the 15 dogs had metastatic disease to local lymph nodes, lung, liver, spleen, bone, or some combination of sites. For the 13 dogs that were enrolled through at least the first 4 weeks of the 8-week study period, the mean toceranib dosage was 2.67 (range, 2.67–2.79) mg/kg and the mean CYC dosage was 15.46 (range, 13.71–16.62) mg/m². Of these dogs, 6 had stable disease throughout the 8-week study period whereas 7 were determined to have progressive disease at either the 6- or 8-week study visit. Two dogs were removed from the study before beginning CYC treatment because of rapidly progressive disease and therefore received toceranib alone. All 15 dogs have either died or been euthanized because of clinical signs attributable to progression of their disease.

Discussion

Several important findings emerged from these studies. First, we found that administration of toceranib was associated with a significant and selective decrease in the number and percentage of Treg in the blood of dogs with cancer. When a metronomic dosage of CYC was combined with toceranib, Treg numbers remained low although serum concentrations of the proinflammatory cytokine IFNγ increased significantly, suggesting potential amelioration of the immunosuppressive effects of Treg. In addition, the combination of toceranib and low-dose CYC was well tolerated in dogs with advanced neoplasia. Although tumor efficacy was not a study endpoint, 6 of 15 dogs had stable disease for the 8-week study period. Therefore, toceranib combined with metronomic CYC may represent an attractive treatment option for dogs with cancer based on potential antitumor and immunomodulatory effects.

Toceranib has demonstrated therapeutic activity in dogs with mast cell tumors and several other tumor types. The data presented herein indicate that decreasing numbers of circulating Treg may be another mechanism of action of toceranib. Although preliminary, our in vitro data suggest that Treg modulation may occur through indirect rather than direct effects of the drug because, under conditions known to expand the Treg subset from activated lymphocytes, we were unable to detect a difference in Treg percentages between cultures that contained toceranib and those that did not. This result is consistent with those of studies in human cancer patients in which clinically relevant doses of sunitinib decreased peripheral Treg but had no effect on the ability of either Treg or other T effector cells to expand in vitro despite periods of drug exposure as long as 14 days.[7] Unfortunately, drug exposure times longer than 24 hours were not assessed in our study; therefore more delayed effects on Treg viability cannot be ruled out. In addition, this study was limited by the use of 2 markers (CD4 and FoxP3) to identify the canine Treg population; inclusion of the recently developed canine-specific anti-CD25 antibody as a third marker might have increased the accuracy of Treg identification.[25, 26] Because some dogs received other medications during the course of this clinical trial, we also cannot exclude the possibility that other drugs might also impact Treg numbers or interact with toceranib's effects on this T cell subset. This might be particularly relevant to the 4 dogs that received carprofen during the study period, because data from mouse tumor models suggest that inhibition of cyclooxygenase-2 also may suppress Treg function.[27]

We recently completed a pilot study evaluating the effects of low-dose CYC on circulating Treg numbers in dogs with soft tissue sarcoma.[14] In that study, we found that CYC administered orally at dosages of either 12.5 or 15 mg/m²/day resulted in a significant decrease in the number of circulating Treg within a 28-day period of time. Therefore, we expected that the addition of CYC at 15 mg/m²/day to toceranib treatment would elicit a further decrease in circulating Treg numbers compared to that achieved with toceranib alone. This notion also is consistent with data from preclinical studies that demonstrate the cooperative anti-angiogenic effects of TKIs and CYC.[28, 29] However, although Treg were significantly decreased after 6 weeks of combination treatment compared to pre-study values (Fig 2), they were not different from day 14 when only toceranib had been administered. Although it is possible that combination treatment was required to achieve the low value at day 56, it could also be true that prolonged toceranib treatment alone might result in a similar number. Indeed, a weakness of this study was the absence of a group of dogs that received single agent toceranib treatment over the same 8-week period. Such a group might have been helpful in assessing chronic drug effects on Treg numbers. Other factors, such as drug interactions and the effects of tumor type or stage on Treg, also may have obscured detection of any synergistic activities of the 2 drugs.

Decreases in peripheral Treg numbers frequently are associated with improved cell-mediated immune responses in humans with cancer and in mouse tumor models. In many clinical trials of humans, these responses are monitored with concentrations of key Th1 cytokines such as IFN-γ. Therefore, we hypothesized that the decrease in Treg numbers observed in the dogs of this study would correspond to an increase in serum concentrations of IFN-γ. Unexpectedly, despite the sharp decrease in Treg numbers, the concentration of IFN-γ after 2 weeks of toceranib treatment was similar to pretreatment concentrations. In fact, a significant increase from baseline concentrations was not observed until 2 weeks after toceranib and CYC were being given concurrently. Interestingly, the inverse relationship between the increased concentration of IFN-γ and decreased numbers of Treg was significant only after combination treatment had been given for 6 weeks, at the time when Treg numbers were at their lowest. One possible explanation for these results is that sufficiently low numbers of Treg are required before suppressed Th1 immune responses can recover. Given that CYC has been shown to promote the secretion of IFN-γ from CD8+ T cells, it is also possible that CYC improves nonregulatory T cell function in dogs by stimulating secretion of IFN-γ from Th1 T cells or through activation of other cells such as dendritic cells.[30, 31] Mechanistic studies examining the effects of both drugs on the function of Treg and other T cell subsets in dogs will be required to gain a better understanding of their immunomodulatory properties.

Toxicity of combined CYC and toceranib was characterized as being mild (grade 1) or moderate (grade 2) in all the dogs of this study. The majority of gastrointestinal AE occurred during the first 2 weeks, at a time when only toceranib was being administered. This is consistent with the results of other studies with toceranib in which gastrointestinal toxicities predominate.[32] Only 1 dog developed mild neutropenia during our study. Although combination of toceranib with other drugs such as vinblastine has been limited by neutropenia, the results of our previous work with CYC at 15 mg/m²/day suggested that bone marrow toxicity from combination treatment was unlikely. In other studies evaluating metronomic dosing of CYC, sterile hemorrhagic cystitis (SHC) has been reported in as many as 10–22% of dogs.[33, 34] Although the duration of CYC administration in our study was short (6 weeks), clinical and laboratory signs consistent with SHC were not observed.

In conclusion, we found that the combination of toceranib and low-dose CYC was safe in dogs with cancer. In addition, administration of toceranib resulted in a selective decrease in the number and percentage of Treg in the peripheral blood. Our data suggest that the addition of CYC may further modulate canine Treg, or potentially other immune cells, by stimulating secretion of IFN-γ. Because of the small number of dogs in this study, these results must be considered preliminary, but they support additional studies to better characterize the antitumor and immunomodulatory effects of combination toceranib and low-dose CYC treatment.