Ultrasonography of Osteosarcoma in Dogs
Sonography Atlas of Dogs

Other

Other adverse events observed in dogs enrolled in this study included: elevated creatinine kinase; otitis media; fever; urinary tract infection; epistaxis; stranguria; hyperglycemia; hypoglycemia; indolent ocular ulcer; seizure; shaking/panting; skin erythema; pyoderma; hyperkalemia; hypokalemia; hypoalbuminemia; hypercalcemia; ataxia; motor neuropathy; sensory neuropathy. These adverse events were believed to be unrelated to toceranib/oral therapy, and instead likely represent progression of disease or other co-morbid conditions.

Discussion

The benefits of amputation and adjuvant chemotherapy for dogs with OSA are well established in veterinary medicine, however most dogs will succumb to metastatic disease within 1 year of amputation and adjuvant chemotherapy, with 2 year survival rates reported in most studies of only 10–15% [8,46]. Despite numerous attempts to improve outcome, little progress has been made in extending survival for affected dogs. The purpose of this clinical trial was to determine whether the addition of metronomic chemotherapy using piroxicam and cyclophosphamide alone or in combination with toceranib would impact both disease free interval and overall survival in dogs with appendicular OSA following amputation and carboplatin chemotherapy. Results from this study demonstrate that piroxicam and cyclophosphamide metronomic therapy with toceranib did not improve median DFI, median OS or the 1-year survival rate in dogs with appendicular OSA over piroxicam and cyclophosphamide alone.

The median DFI and survival times reported here are similar to those reported for dogs that receive carboplatin chemotherapy after amputation [3,7]. Furthermore, the 1-year survival rates of 38.4% and 34.9% for toceranib-treated dogs and control dogs, respectively, are comparable to the 35.4% 1-year survival rate reported for carboplatin alone. Analogous results have been obtained with doxorubicin alone or in combination with carboplatin, cisplatin alone, and carboplatin given concurrently with gemcitabine [5,8,47–49]. A recent retrospective study compared multiple chemotherapy protocols used for canine appendicular OSA post-amputation. Differences in dose intensity and the chemotherapy protocol chosen did not significantly influence the reported median DFI and OS of 291 days and 284 days, respectively [8]. These data support the notion that alteration of therapeutic approaches to treat microscopic metastatic disease using currently available cytotoxic chemotherapeutics is unlikely to result in significant benefit for dogs with OSA.

Therapeutic approaches that target both the tumor itself as well as the tumor microenvironment are increasingly recognized as important components in regulating the metastatic process and several studies have investigated this approach. For example, studies in veterinary medicine have reported on safety and efficacy of metronomic chemotherapy protocols in the treatment of various tumors [9,10,50,51]. The combination of metronomic cyclophosphamide (10 mg/m2) and piroxicam (0.3 mg/kg) was shown to be well-tolerated and suggests that metronomic chemotherapy prolongs DFI for incompletely resected canine soft tissue sarcomas [9]. The use of metronomic cyclophosphamide has been also shown to decrease the number of circulating regulatory T-cells (Treg) and tumor microvessel density in tumor-bearing dogs when cyclophosphamide was administered at 15 mg/m2/day [11]. Unfortunately, the doses and regimens used in metronomic chemotherapy protocols are largely anecdotal, and predictive biomarkers of clinical benefit in dogs with OSA have not been thoroughly evaluated. In the present study cyclophosphamide and piroxicam were dosed at 10 mg/m2 and 0.3 mg/kg EOD, respectively. This therapeutic regimen was established based on the available literature at the beginning of the study, prior to demonstration of the immunomodulatory and antiangiogenic effects of higher doses of cyclophosphamide. Given the reported immunomodulatory and anti-angiogenic effects of cyclophosphamide used at higher doses, it is possible that the lower dose of cyclophosphamide in the present study and the use of EOD dosing resulted in a treatment regimen insufficient to significantly affect the tumor microenvironment, the number of circulating Tregs, or both.

While biologic activity of toceranib in metastatic canine appendicular osteosarcoma has been reported [37], the lack of improvement in DFI and OS in the present study suggests there may be different molecular drivers at different stages of disease in canine OSA. It is possible that while effects on tumor growth are observed in dogs with macroscopic metastatic pulmonary lesions following toceranib therapy, microscopic lesions may become resistant to therapy with in a short period of time, thereby negating any potential therapeutic value. Data generated from several murine models have suggested that VEGF inhibitors modulate the tumor microenvironment in a manner that can actually accelerate metastatic tumor growth [28,29,52]. There has therefore been concern that the use of VEGFR inhibitors in the setting of microscopic disease could actually promote a more aggressive phenotype. Importantly, this study demonstrated that the use of toceranib in dogs with microscopic metastatic OSA did not result in shortened survival times or aberrant patterns of metastasis that would indicate the induction of a more aggressive disease phenotype following treatment.

Recently, administration of toceranib at 2.75 mg/kg EOD to cancer-bearing dogs was associated with a significant decrease in circulating Tregs, possibly through indirect immunomodulatory mechanisms [39]. Doses of toceranib below the label dose (ranging from 2.4–2.9 mg/kg EOD) still result in target inhibition, supporting the use of 2.75 mg/kg in the present study [37]. Therefore, the lack of improvement in DFI and OS reported here is unlikely secondary to the use of a lower dose of toceranib than the current label dose. Although VEGFR inhibition was not specifically assessed in this study, upregulation of plasma VEGF concentrations following toceranib treatment has been previously demonstrated in dogs receiving 2.4–2.9 mg/kg EOD, consistent with effective VEGFR2 inhibition [37].

Adverse events necessitating administration of chlorambucil in place of cyclophosphamide occurred in 8 dogs, with all but one of these due to sterile cystitis. The overall incidence of cyclophosphamide-induced cystitis is low in the veterinary literature, and increased risk has been associated with a higher cumulative cyclophosphamide dose and higher dose intensity [53]. As there were a small number of dogs in each group that developed cystitis (n = 2 control, n = 5 toceranib), and it is likely coincidence that a higher number of toceranib-treated dogs developed cystitis. The use of metronomic chlorambucil has been associated with some antitumor activity in a variety of spontaneous canine tumors, and decreases in numbers of circulating Tregs compared to baseline has been documented in dogs receiving chlorambucil [51]. While the immunomodulatory effects of metronomic cyclophosphamide and chlorambucil have been reported, the impact of chlorambucil administration in this study is not known. However, given the small number of dogs that went on to receive this drug in place of cyclophosphamide, it was unlikely to significantly impact outcome in either group.

Multiple investigations have reported on the prognostic significance of OSA occurring in the proximal humerus [3,7,54–57]. While decreased survival times were noted in dogs with tumors located in the proximal humerus in the current study, the difference in OS was not significant compared to other tumor locations. This may have been due to the small number of dogs that enrolled in the study with proximal humerus disease (n = 31). Alternatively, it is possible that the metronomic therapy had an effect on outcome in this population to improve survival. Future controlled studies would be necessary to determine the benefits of metronomic chemotherapy in this subset of patients.

Toceranib-treated dogs had a higher adverse event profile compared to control dogs, although the adverse events reported were consistent with those expected based on prior published studies. It is important to note that in the current study, dogs experiencing adverse events while on toceranib were effectively managed with drug holidays, concomitant medications and/or dose reductions. The frequency of adverse events was lower than that reported in a recent clinical trial of dogs with various solid tumors receiving toceranib at doses of 2.4–2.9 mg/kg [37]. In this study, the adverse event profile was far superior to that associated with the label dose of toceranib (3.25 mg/kg) but both adequate drug exposure and biologic activity were maintained [37]. Despite the inherent difficulties in comparing findings from various studies, the dose of toceranib utilized in the present study (2.75 mg/kg) was better tolerated compared to the label dose, allowing improved adherence to the treatment protocol [36,37].

While necropsy was not required in the current study and the detection of metastatic disease was largely dependent on clinical findings and thoracic radiographs, it is unlikely that gross metastatic disease in other locations was missed in many dogs. The incidence of gross OSA metastasis detectable with abdominal ultrasound has been previously reported, with abdominal metastasis identified in 0–1.7% of dogs that do not have evidence of pulmonary metastatic disease [58,59]. Additionally, the reported rate for detecting unsuspected osseous metastasis with nuclear scintigraphy in dogs with appendicular OSA at the time of initial presentation (prior to amputation) is 7.8% [60]. Thus, while the lack of necropsy may have contributed to an under-reporting of the true incidence of metastasis, thoracic radiography is considered a reliable diagnostic procedure to detect metastatic disease for most dogs following amputation and chemotherapy, with extrathoracic metastases usually identified by virtue of clinical signs (i.e., lameness for bone metastasis).

It is possible that the use of compounded therapeutics could have impacted our results. The Apothecary Shoppe (now Avella Specialty Pharmacy) was chosen as the supplier of compounded cyclophosphamide and piroxicam as this pharmacy is certified by the Pharmacy Compounding Accreditation Board. However it is recognized that compounding may have resulted in inadvertent overdosing or under dosing, resulting in increased toxicity or sub-therapeutic drug concentrations in some patients, respectively.

A potential weakness of the current study is that the metronomic treatment regimen began after completion of carboplatin chemotherapy. This may have given the remaining microscopic metastatic tumor cells time to develop resistance to multiple therapies, thus negating any impact of the subsequent treatment. While the use of metronomic cyclophosphamide and piroxicam administered concurrently with carboplatin for dogs with appendicular osteosarcoma was reported to be well-tolerated in another study [50], no improvement in survival time was noted with concurrent administration. Nevertheless, it may be warranted to determine if concurrent use of toceranib in combination with metronomic chemotherapy and carboplatin significantly impacts DFI and OS.