Oncolytic viruses are gaining ground as an alternative therapy for treating cancer in veterinary oncology. Indeed, several studies involving adenovirus, myxoma virus, sendai virus, reovirus and vesicular stomatitis virus have been conducted in dogs and cats [28–36]. Promising results have been reported by intratumoral or intravenous routes on dogs and cats bearing cancers. At present, none of them evaluated an oncolytic VACV allowing intratumoral production of chemotherapy drug. Our study provides a description of the maximal tolerated dose, clinical toxicities and viral shedding after intramuscular administration of TG6002 with oral 5-FC in healthy immune-competent dogs. As TG6002 is engineered from VACV, evaluation of both safety and shedding profiles of this attenuated virus remains essential. Thus, to provide more reliable information about tolerance, healthy dogs with a competent immune system have been selected for this study.
Data collected in this preclinical study indicate that administration of TG6002 was well tolerated for all dogs. MTD was not reached even at the highest tested dose of 5 × 107 PFU/kg. Only transient weight losses for all dogs and digestive disorders for one dog out of seven were observed. One transient history of hyperthermia was noticed for one dog receiving TG6002 at the lowest dosage. During the second part of the study, a decrease of white blood cell counts was observed at day 16 for all dogs. Hematological changes were transient and values remained within the interval range. Decrease of white blood cell counts was suspected to be induced by viral infection. Indeed, similar variations have been described in laboratory beagles receiving intravenous administrations of an oncolytic VACV encoding CD40 ligand [37]. Slight leukopenia has been also reported in human trials with intratumoral oncolytic VACV but remained rare [38, 39]. However, as 5-FU can induced bone marrow depression, it would have been interesting to evaluate serum concentrations of 5-FU in our study.
Others adverse events reported in laboratory beagles, receiving intravenous administration of an oncolytic VACV encoding CD40 ligand, included transient grade 1 fever (n = 1/2) and grade 3 seizure (n = 1/2) [37]. Additionally, only a grade 1 increase in alkaline phosphatase (n = 1/2) and a low albumin level (grade 1) (n = 1/2) were noticed without hematological or urinary abnormalities [37]. In human trial adverse events secondary to intratumoral or intravenous administration of oncolytic VACV included fever, rigors, abdominal pain, nausea, vomiting, tiredness and headache [19, 20, 39].
Infection with VACV is generally characterized by the development of cutaneous and mucocutaneous pock lesions [17]. No pock lesions were observed throughout the study. Indeed, human patients diagnosed with cancer receiving attenuated oncolytic VACV have been reported to develop these lesions [10, 19, 21, 22]. Intradermal injection of VTK-79, a TK-deleted VACV, has also been reported to induce the development of small nodules in laboratory beagles [24].
TG6002 has a tumor-selective viral replication induced by the double TK-RR deletion of the TG6002 genome [11]. No pock lesions or other major clinical abnormalities were observed in our study. The lack of clinical adverse events confirms the safety profile of TG6002. However, the use of healthy dogs in our study can limit the assessment of adverse events. Indeed, the dogs in this study did not have cancer, viral load could be probably higher in dogs with tumors that allow viral amplification. Thus, tolerance and viral shedding will have to be evaluated on neoplastic dogs.
In our study, a transient mild weight loss was noticed for all dogs, which may be induced by virus administration. Indeed, weight loss was observed for both dogs receiving TG6002 alone or with 5-FC. One dog developed gastrointestinal signs which can be related to 5-FC or TG6002.
5-FC is an antifungal agent, mainly used against strains of Cryptococcus and Candida. 5-FC penetrates fungal cells where it is deaminated by cytosine deaminase to 5-FU. 5-FU acts as an antimetabolite by competing with uracil, thereby interfering with pyrimidine metabolism and eventually RNA and protein syntheses. It is thought that 5-FU is converted to 5-fluoro-2’-deoxyuridylate that inhibits thymidylate synthesis and ultimately DNA synthesis. In human, 5-FC is rapidly absorbed with a bioavailability amounts to 76–89% [40]. It penetrates well into the most body sites and is eliminated by about 90% by glomerular filtration with a half-life of 3 to 4 hours [41, 42]. A dose-dependent bone marrow depression (anemia, leukopenia, thrombocytopenia) and an increase of hepatic enzymes are reported [43]. Adverse events reported in dogs also include toxic epidermal necrolysis on the scrotum, the nasal planum, lips and eyelids [44, 45]. These cutaneous adverse events resolve following discontinuation of 5-FC [44]. No cutaneous lesion nor bone marrow depression induced by 5-FC were observed in our study. Studies suggested 5-FC conversion by the microorganisms in the intestinal microflora leading to 5-FU exposure [46, 47]. It would have been interesting to evaluate serum concentrations of 5-FC and 5-FU and the presence of 5-FC and 5-FU in the feces. Otherwise, due to the gastrointestinal signs reported in human trials, oncolytic VACV toxicity could also be suspected. Even if no macroscopic bowel lesion was observed during the necropsy, histological analysis would have been interesting.
Viral shedding detection also acts as an essential part in the environmental risk assessment for this novel therapy. Viral shedding has been previously evaluated in dogs after the intradermal administration of the TK-deleted VACV VTK-79 [24]. Even if dogs with induced pock lesions were in close contact with sentinel dogs, none of these control animals developed VACV antibodies suggesting the absence of shedding [24]. In a study with intravenous oncolytic VACV administrations in healthy dogs, the viral load, detected by qPCR, declined quickly in blood samples during the 4 hours after infusion and viral DNA was not detected in feces, saliva and urine samples collected at 1, 2, or 4 days after virus administration [37]. In human clinical trials, viral genome has been also detected in the blood of patients after intratumoral administrations of oncolytic VACV [19, 48]. In the present study, neither infectious virus nor viral genome copies were detectable in blood, urine, saliva and feces. Due to limits associated with cell culture, feces samples could not be analysed by infectious titer assay. The absence of viral shedding can be explained by the characteristics of the viral vector and the route of administration. Indeed, due to the deletion of TK and RR genes, the virus is only able to replicate in dividing cells [11]. Thereby, intramuscular administrations of viral vector in healthy dogs were not expected to be associated with replication of the vector. TG6002 was designed for intratumoral injection, and this route of administration might promote the replication of the vector. Thus, viral burden could probably be higher in dogs with tumors that allow viral amplification. In this context, tolerance and viral shedding will also be evaluated on neoplastic dogs.
Previous studies evaluating oncolytic potency of TG6002 in cell lines and xenograft models have shown promising results [11, 13]. However, murine xenograft models show several limits including the lack of immune system, their size and the non-spontaneous origin of the cancer. Highly relevant animal models are needed to assess the efficacy of treatment prior to human trials. In contrast to rodents, cancers arising in dogs have several commonalities with human cancers. Indeed, dogs share the same environment with their owners, their immune system is intact, and cancer progression is spontaneous resulting in similar complexity, clonality, and immune suppression as seen in man [49]. In addition, the same cancer-associated genes and histological features have been found in both species for several cancers like urothelial carcinoma or mammary carcinoma [50, 51]. Thereby, it is now well recognized that dogs with spontaneous cancer serve as a good model for several human cancers [52]. A previous study has already established the oncolytic potency of TG6002 with 5-FC in canine mammary tumor explants [13]. Indeed, histological analyses of ex vivo canine mammary adenocarcinoma explants cultured with TG6002 and 5-FC, allowed assessment of tumor necrosis and conversion of 5-FC into 5-FU [13]. Considering dog as a relevant model in oncology in one hand, and promising in vitro results combine to the safety profile of TG6002 on the other hand, intratumoral injections of 5 × 107 pfu/kg of TG6002 can be considered for a clinical trial in dogs with spontaneous tumors. Indeed, the individual dog will benefit from effective treatment and the results are expected to help other dogs. Finally, the findings are expected to advance treatments in human oncology.