Current Pharmaceutical Design - Volume 8, Issue 15, 2002

Quinone based bioreductive drugs have, potentially, a very versatile use in cancer chemotherapy. They can be activated by DT-diaphorase and hence can be used to target tumour types rich in this O2- independent reductase enzyme. Small molecular modifications can substantially reduce specificity for DTdiaphorase and under these circumstances the quinones become much less toxic in air but retain their potent cytotoxic effects under hypoxic conditions. Our understanding of the reductive (bio) chemistry of indolequinones, in particular, has subsequently allowed us to develop a platform technology where almost any therapeutic entity can potentially be delivered, selectively, to hypoxic tumours. Antiangiogenic approaches are currently receiving a substantial amount of attention and this review brings their development into context in view of the hypoxia dependence for neovascularization. Lastly, the use of bioreductive drugs when combined with hypoxia-mediated gene therapy is described. Such an approach provides a unique dual level of specificity for targeting hypoxic tumours and potentially can provide substantial therapeutic benefit.

The expression of drug metabolising cytochrome P450s (CYPs) notably 1A, 1B, 2C, 3A, 2D subfamily members have been identified in a wide range of human cancers. Individual tumour types have distinct P450 profiles as studied by detection of P450 activity, identification of immunoreactive CYP protein and detection of CYP mRNA. Selected P450s, especially CYP1B1, are overexpressed in tumours including cancers of the lung, breast, liver, gastrointestinal tract, prostate, bladder. Several prodrug anti-tumour agents have retrospectively been identified as P450 substrates for which tumour CYP activation may hitherto have been underestimated. Those in clinical use include prodrug alkylating agents (cyclophosphamide, ifosphamide, dacarbazine, procarbazine), Tegafur, a prodrug fluoropyrimidine, methoxymorphylinodoxorubicin, a metabolically activated anthracycline, as well as flutamide and tamoxifen, two non-steroidal hormone receptor antagonists that are significantly more active following CYP-hydroxylation. More exciting is the prospect of developing new agents designed to be selectively dependent on tumour CYP activation. This can be illustrated with P450 activation of the 2-(4-aminophenyl)benzothiazoles exclusively in CYP1A1 inducible tumours. Also of interest is the bioreductive antitumour prodrug AQ4N, a CYP3A substrate that is activated to a cytotoxic metabolite specifically in hypoxic tumour regions.

Nitroreductases that metabolise aromatic nitro groups to hydroxylamines are attractive as enzymes for GDEPT because of the very large electronic change that this metabolism generates, providing an efficient “switch”that can be exploited to generate potent cytotoxins. While nitroreductase enzymes are widespread, nearly all the work using these in GDEPT has been with the nfsB gene product of Escherichia coli, an oxygen-insensitive flavin mononucleotide nitroreductase (NTR). Four classes of prodrugs for NTR have been described, dinitroaziridinylbenzamides, dinitrobenzamide mustards, 4-nitrobenzylcarbamates and nitroindolines. While some quinones are excellent substrates for NTR, none have been identified as potential GDEPT prodrugs. The most widely studied prodrug used for GDEPT in conjunction with NTR is the dinitroaziridinylbenzamide CB 1954. This shows high selectivity (>1000-fold) in cell lines transfected with NTR, has potent and long-lasting inhibition of NTR-transfected tumours in mice, and is in Phase I trial in conjunction with virally-delivered NTR enzyme. The related mustard SN 23862 has similar selectivity and superior bystander effects in animal models. Nitrobenzyl carbamates of a variety of cytotoxic amines (including aniline mustards, enediynes, duocarmycin analogues, pyrrolobenzodiazepines and the antitumour antibiotics doxorubicin, actinomycin D and mitomcyin C) are metabolised efficiently by NTR to the hydroxylamines, that fragment to release the amines. Nitroindoline derivatives of duocarmycins also show moderate selectivity for NTR-transfected cell lines in culture.

The duocarmycins and (+)-CC-1065 are amongst the most potent antitumour antibiotics discovered to date and yet have not progressed into the clinic. The natural products are extremely stable to nucleophilic attack until bound to their DNA target and are not substrates for any other biological nucleophile. The mechanism for this target activation of the duocarmycins is discussed with relation to both an acid-catalyzed activation and a binding-induced conformational change leading to ground state destabilization. It is suggested that targeting of the duocarmycins to their site of action in a tumour may be more important than introducing systemically-activated prodrugs as the natural product itself can be considered to be a type of prodrug, activated only on binding to its targets. Methods that have been used to target CC-1065 and the duocarmycins are reviewed as well as efforts towards systemically activated prodrugs. A simple analysis of the approaches that could be taken to vary the structure for targeting is suggested.

The selective activation of a relatively non-toxic prodrug by an enzyme present only in the tumour should enhance the drug concentration at the tumour site and result in a better anti-tumour effect and a reduction in systemic toxicity as compared to conventional chemotherapy. β-Glucuronidase is such an enzyme. It is normally expressed in the lysosomes of cells. In larger tumours, however, high levels of the enzyme are present in necrotic areas. Several glucuronide prodrugs have been synthesised that can be activated by β- glucuronidase. They are relatively non-toxic due to their hydrophilic nature, which prevents them from entering cells and thus from contact with lysosomal β-glucuronidase. The main problem of glucuronide prodrugs for clinical use is their fast renal clearance. Special attention should be paid to the development of new less hydrophilic prodrugs with slower clearance, as this would result in a prolonged exposure to β- glucuronidase at the site of the tumour and a reduction of the amount of prodrug needed. A number of interesting anthracyclin-based glucuronide prodrugs have been synthesised and have shown favourable therapeutic effects compared to treatment with the parent drug. The tumoural levels of β-glucuronidase can even be enhanced by two-step approaches, in which exogenous enzyme is targeted to the tumour by an antibody (ADEPT) or by the gene encoding the enzyme in transduced tumour cells (GDEPT). The ADEPT and GDEPT approaches in combination with glucuronide prodrugs have shown enhanced efficacy in experimental tumour models. Further improvement of ADEPT and GDEPT is warranted to optimise the tumour uptake and retention of antibody-enzyme fusion proteins and the efficiency and safety of current gene delivery methods. In conclusion, it is clear that glucuronide prodrugs hold promise for future use in the treatment of cancer in patients as monotherapy. Enhancement of the therapeutic effects of glucuronide prodrugs, also in patients with small tumour lesions, may possibly be achieved by techniques that target β-glucuronidase specifically to the site of the tumour.

Several commonly used cancer chemotherapeutic prodrugs, including cyclophosphamide and ifosfamide, are metabolized in the liver by a cytochrome P450 (CYP)-catalyzed prodrug activation reaction that is required for therapeutic activity. Preclinical studies have shown that the chemosensitivity of tumors to these prodrugs can be dramatically increased by P450 gene transfer, which confers the capability to activate the prodrug directly within the target tissue. This P450 gene-directed enzyme prodrug therapy (P450 GDEPT) greatly enhances the therapeutic effect of P450-activated anti-cancer prodrugs without increasing host toxicity associated with systemic distribution of active drug metabolites formed by the liver. The efficacy of P450 GDEPT can be enhanced by further increasing the partition ratio for tumor:liver prodrug activation in favor of increased intratumoral metabolism. This can be achieved by co-expression of P450 with the flavoenzyme NADPH-P450 reductase, which increases P450 metabolic activity, by localized prodrug delivery, or by the selective pharmacologic inhibition of liver prodrug activation. P450 GDEPT prodrug substrates are diverse in their structure, mechanism of action, and optimal prodrug-activating P450 gene, they include both established and investigational anticancer prodrugs, as well as bioreductive drugs that can be activated by P450 / P450 reductase in a hypoxic tumor environment. Several strategies may be employed to achieve the tumor-selective gene delivery that is required for the success of P450 GDEPT, these include the use of tumortargeted cellular vectors and tumor-selective oncolytic viruses. Overall, P450-based GDEPT presents several important, practical advantages over other GDEPT strategies that should facilitate the incorporation of P450 GDEPT into existing cancer treatment regimens. A recent report of clinical efficacy in a P450-based phase I / II gene therapy trial for pancreatic cancer patients supports this conclusion.