Launching a novel preclinical infrastructure: comparative oncology trials consortium directed therapeutic targeting of TNFalpha to cancer vasculature

Abstract

Background: Under the direction and sponsorship of the National Cancer Institute, we report on the first pre-clinical trial of the Comparative Oncology Trials Consortium (COTC). The COTC is a novel infrastructure to integrate cancers that naturally develop in pet dogs into the development path of new human drugs. Trials are designed to address questions challenging in conventional preclinical models and early phase human trials. Large animal spontaneous cancer models can be a valuable addition to successful studies of cancer biology and novel therapeutic drug, imaging and device development.

Methodology/principal findings: Through this established infrastructure, the first trial of the COTC (COTC001) evaluated a targeted AAV-phage vector delivering tumor necrosis factor (RGD-A-TNF) to alphaV integrins on tumor endothelium. Trial progress and data was reviewed contemporaneously using a web-enabled electronic reporting system developed for the consortium. Dose-escalation in cohorts of 3 dogs (n = 24) determined an optimal safe dose (5x10(12) transducing units intravenous) of RGD-A-TNF. This demonstrated selective targeting of tumor-associated vasculature and sparing of normal tissues assessed via serial biopsy of both tumor and normal tissue. Repetitive dosing in a cohort of 14 dogs, at the defined optimal dose, was well tolerated and led to objective tumor regression in two dogs (14%), stable disease in six (43%), and disease progression in six (43%) via Response Evaluation Criteria in Solid Tumors (RECIST).

Conclusions/significance: The first study of the COTC has demonstrated the utility and efficiency of the established infrastructure to inform the development of new cancer drugs within large animal naturally occurring cancer models. The preclinical evaluation of RGD-A-TNF within this network provided valuable and necessary data to complete the design of first-in-man studies.

Figure 1. Schema representing the schedule for the dose escalation phase of RGD-A-TNF evaluation in dogs with spontaneous cancers.

(A) This study was structured as a dose-escalation using a modified Fibonacci design to govern dose escalation towards an MTD. (B) Three dogs were enrolled in the starting-dose cohort, and three dogs per cohort were enrolled there after for each of the five dose levels planned. Dogs were scheduled to receive RGD-A-TNF on day zero and to undergo definitive tumor resection 4 days later. This initial 4-day group was designed (i) to evaluate vector localization and TNFα expression within tumors and (ii) to verify that the tentative follow-up schedule of RGD-A-TNF administration at one-week dosing intervals was biologically appropriate. After a group of dogs (n = 18) were treated according to this schedule, an additional group was enrolled by equivalent inclusion criteria to receive RGD-A-TNF on the same day of definitive surgical resection. In this subset of dogs (n = 6), surgery was performed 4–6 hours post administration of RGD-A-TNF, this same-day experimental subset was designed to establish the acute selectivity of RGD-A-TNF trafficking and its tumor vascular localization.

Figure 2. RGD-A-TNF trafficking resulted in selective tumor endothelial cell localization and TNF α expression.

(A, B) RGD-A-TNF selectively targeted tumor-associated vasculature (arrows) and was absent from normal tissues at 4–6 hours (A) and at 4 days (B). Magnification, 400-fold; scale bar, 100 µM. Pre-treatment tumor biopsies, post-treatment tumor biopsies, and post-treatment normal tissues in dogs that received a single-dose of RGD-A-TNF double-stained with an anti-CD31 specific antibody plus an anti-bacteriophage specific antibody. Detection was performed with Alexa Fluor 488 (green, blood vessels), Alexa Fluor 594 (red, AAVP), and DAPI (blue, cell nuclei). (C) Pre-treatment tumor biopsies (day 0), post-treatment normal biopsies (day 4) and post-treatment tumor biopsies (day 4) were used for extraction of total RNA. RT-PCR was performed to measure transcript levels of human TNFα in quadruplicate. The Y-axis represents the relative TNFα expression levels in post-treatment normal biopsies and post-treatment tumor biopsies compared to pre-treatment tumor biopsies after normalization to GAPDH expression (Kruskal-Wallis Test, p = 0.0107). All data are presented as means±standard deviations. (D–F) Presence of RGD-A-TNF was evaluated in post-treatment (day 28) necropsy samples of tumor (D) and normal tissues (E, F). Tissues were stained for RGD-A-TNF as described earlier. RGD-A-TNF selectively targeted tumor-associated vasculature in post-treatment tumor samples (arrows). In contrast, the vector was not apparent in pre-treatment tumor samples or in post-treatment normal control necropsy samples (such as lung, liver, spleen or intestine) after serial administrations of RGD-A-TNF.

Figure 3. Schema representing the schedule for the multi-dose phase of RGD-A-TNF evaluation in dogs with spontaneous cancers.

This study was designed as an open label, multiple fixed-dose trial (i) to establish feasibility and (ii) to identify chronic and/or cumulative toxicity of repetitively administered RGD-A-TNF. Dogs received weekly doses (5×10¹² TU intravenously) of RGD-A-TNF. Anticancer activity of this agent was evaluated using RECIST criteria. The treatment received population included dogs that received at least four weekly doses (i.e., one cycle 1). This population consisted of 14 dogs. Dogs were permitted to receive additional therapy in subsequent cycles if there was evidence of either stable disease or tumor response. Tumor measurements were recorded every two weeks with full restaging every 28 days.

Figure 4. RGD-A-TNF administration resulted in objective tumor responses in dogs with spontaneous cancers.

A large primary soft tissue sarcoma on the flank of Dog #2.11 is shown to feature the potential magnitude of the tumor response. Prior to therapy (day 0), the tumor measured 12.3 cm in longest diameter. At day 28, after 4 weekly infusions of RGD-A-TNF, the tumor measured 8.2 cm in longest diameter (a 33% regression) and a RECIST-based partial response (PR). At day 56, after a total of 8 weekly systemic infusions of RGD-A-TNF, the tumor measured 1.85 cm in longest diameter prior to resection. Therefore, this response equated to an 85% regression from baseline and a continued clinical PR. Upon surgical resection of the residual lesion, no viable tumor was found and a pathological complete response was determined.