A Clinical Trial of AdNRGM Plus CB1954 in Prostate Cancer

Background and Rationale:

Locally relapsed prostate cancer presents an attractive target for gene therapy because the tumour remains localised to the prostate or to the surrounding tissue for a long phase of the disease. Also the delivery of gene therapy vectors to the prostate is straightforward using either ultrasound-guided, trans-rectal injection or adapting methods developed for brachytherapy, to deliver multiple, template-guided trans-perineal injections to achieve saturation coverage of the prostate. Several gene therapy and gene-immunotherapy strategies have been tested in patients with locally relapsed or metastatic prostate cancer, including Virus Directed Enzyme Prodrug Therapy (VDEPT), Conditionally Replicating Adenoviruses (CRAds) and GMCSF based immune stimulation. These treatments were well tolerated and showed some evidence of anti-tumour activity.

The nitroreductase/CB1954 VDEPT system:

The nitroreductase/CB1954 system employs E. coli nitroreductase (NTR) as the prodrug activating enzyme, and CB1954 as the prodrug to be activated. To exploit this system a an E1-E3 deleted, replication-defective human adenovirus (Ad5) designated CTL102 which contains the NTR gene under the control of the cytomegalovirus (CMV) promoter was previously generated. CTL102 was able to infect cancer cells and to induce NTR expression in vitro, in addition NTR expressing cells are sensitised to CB1954. Experiments carried out in animal models demonstrated that NTR expression could be detected in prostate cancer xenografts injected with CTL102 and that growth of the xenografts was inhibited by intratumoural injection of CTL102 followed 2 days later by intra peritoneal administration of CB1954.

CB1954 prodrug conversion and clinical development:

CB1954 [5-(aziridin-1-yl)-2,4-dinitrobenzamide] is a weak alkylating agent that is reduced by the rat nitroreductase enzyme DT diaphorase to a potent, bifunctional alkylating agent (5-(aziridin-1-yl)-4-hydroxylamino-2-nitrobenzamide) that cross-links DNA and is able to kill both dividing and non-dividing cells. The human and murine forms of DT diaphorase are deficient in this reaction. A nitroreductase enzyme encoded by the nfsB gene of E. coli was found to perform the same reduction 60 to 100-fold faster than the rat DT diaphorase, and this is the enzyme (NTR) encoded in CTL102 and AdNRGM.

A phase I pharmacokinetic trial of CB1954 defined its maximum tolerated dose (MTD), dose-limiting toxicities and plasma concentration-time profile. Ten cohorts of three patients each were treated with CB1954 at dose levels of 3mg/m2 to 37.5 mg/m^2. No marrow suppression, nephrotoxicity or alopecia was observed. Dose limiting toxicities (one grade 4 diarrhoea and two grade 2 hepatic toxicity) were seen at 37.5 mg/m^2; other toxicities included nausea, vomiting, anorexia and fatigue. Side effects recorded at 24 mg/m^2 were grade 3 nausea and fatigue, therefore this dose was recommended as the standard dose for future clinical trials. A clouding of the lens of the eye, also known as cataract, occurred in a few animals given the CB1954 drug. The Investigators believe this change is highly unlikely to occur in humans and none of the patients who were treated with this drug in previous clinical trials has suffered from this complication.

Prostate cancer trial:

A phase I/II clinical trial of CTL102/CB1954 was completed in men with prostate cancer. In the first stage of the trial 20 patients who were scheduled for radical prostatectomy received CTL102 alone, via intraprostatic injection, prior to surgery. Objectives of this part of the study were safety, tolerability and level of NTR expression. Immunohistochemistry demonstrated NTR staining of the glandular epithelium of the tumour and of the normal prostate tissue at all dose levels which were used (1x10^10 to 1x10^12 virus particles). Expression of NTR was found in 30-50% of the prostate specimen slides and appeared to increase when using higher vector doses or multiple injections, however there was no statistically significant relationship between virus dose and extent of NTR expression. Gene therapy was well tolerated, although one patient who received 5x10^10 vp had a DLT (transient bilirubin increase). The cohort was therefore expanded but no other DLT was observed and dose escalation could be resumed. (The DLT was subsequently attributed to a post-operative myocardial infarction.) Other adverse events were transient grade 3 lymphopaenia (3 patients) and grade 2 hepatic enzymes increase (4 patients). There were no treatment-related serious adverse events.

Our data show that the intraprostatic injection of CTL102 followed by intravenous CB1954 is safe and well tolerated11. CTL102 was able to induce NTR expression and the CTL102/CB1954 treatment provided early indications of activity. However, the overall anti-tumour efficacy of this approach was sub-optimal, with only two patients showing a PSA decrease >50% and viable tumour cells still present in post-treatment biopsies. It is likely that the relatively small fraction of tumour cells expressing NTR, and their localised distribution was responsible for the limited anti-tumour activity of CTL102/CB1954. In order to increase efficacy the current clinical trial has incorporated two significant improvements:

1. To achieve better distribution of the vector throughout the prostate the vector will be administered using an adaptation of the technique of prostate brachytherapy, via multiple template-guided trans-perineal injections.

2. To increase efficacy VDEPT with GMCSF based immune stimulation, using an E1,E3-deleted, replication defective adenovirus (AdNRGM) expressing both NTR and human GMCSF will be combined.

The proposed dose levels for AdNRGM are 10^10, 3x10^10, 10^11, 3x10^11, 10^12 vp while the prodrug CB1954 will be given at a standard dose of 24 mg/m^2. In vitro experiments allow us to estimate that 1 x 10^10 vp of AdNRGM is likely to result in production of 5 - 25 µg GMCSF/24 h. Clinical trials of GMCSF have used single doses of 500 µg, or repeated doses of up to 250 µg/m^2. The Investigators consider that a starting dose of 1 x 10^10 vp therefore provides a sufficient safety margin for GMCSF production. The fixed dose or 24 mg/m^2 CB1954 is the dose recommended following an initial dose-escalation trial of CB1954 alone.