Return to PJ Online Home Page
The Pharmaceutical Journal Vol 264 No 7094 p666-673
April 29, 2000 Articles

Hypoxia and disease: opportunities for novel diagnostic and therapeutic prodrug strategies

By Rachel E. Airley, MRes, MRPharmS, Jane E. Monaghan, BSc, and Ian J. Stratford, PhD

This article discusses the causes of hypoxia in various disease states, and the role of hypoxia in the development of diagnostic imaging agents and novel treatments

A variety of pathological conditions exist where the affected tissues are hypoxic or exhibit a low oxygen tension. Primarily, the therapeutic consequences of hypoxia have been considered in the treatment of cancer. However, it is also an important factor where an inflammatory component exists, eg, chronic inflammatory bowel disease, rheumatoid arthritis and ischaemia/reperfusion injury. Moreover, hypoxia must be considered as a factor in diseases where oxygen tensions are likely to be insufficient due to poor respiratory function, eg, cystic fibrosis and chronic bronchitis, as well as to benign proliferative diseases like psoriasis, diabetic vasculopathies and epilepsy.
Advances in research have yielded a means of exploiting this feature, both as a means of imaging and diagnosing disease, as well as for therapy. Examples of therapeutic approaches that offer a means of specifically treating these conditions include the use of bioreductive drugs (ie, prodrugs that are activated only under conditions of hypoxia), gene therapy, and inhibitors of enzymes crucial to the survival of hypoxic cells. This article will discuss the causes of hypoxia in various disease states, and its role in the development of diagnostic imaging agents and novel treatments.

Hypoxia in solid tumours

Cells become hypoxic when, for a variety of reasons, oxygen demand exceeds oxygen supply. Hypoxia arises in solid tumours due to the rate of proliferation of their component cells. In normal tissue, an adequate oxygen and nutrient supply is maintained by structurally and functionally normal vasculature, which is able to expand and adapt to meet the needs of the growing tissue. Some tissues, such as skin and cartilage, are slightly less oxygenated. However, as a rule, normal tissues are sufficiently well vascularised to provide a partial oxygen pressure that exceeds 40mmHg.1 Figure 1 shows the range of partial oxygen pressures occurring in solid tumours as compared with various organs. An aggressively growing tumour will depend upon host tissue derived vascularisation. This vasculature is only sufficient to support selected parts of the tumour for a relatively short time, and as the tumour progresses, morphological and functional changes take place. Hypoxia within the tumour stimulates the production of vascular endothelial growth factor (VEGF), which in turn gives rise to tumour-derived angiogenesis.2,3 However, tumour derived vasculature is distinct from that of normal tissue in that vessels are unevenly distributed throughout the tumour, leading to avascular areas. In addition, tumour blood supply is further compromised as those vessels contained within the tumour are often elongated, dilated and twisted. These morphological changes, together with the accompanying rheological changes that occur as a result of the destabilisation of blood supply, can cause an increase in interstitial pressure, which further decreases perfusion.
In normal tissues, one strategy used to meet changes in oxygen demand is an adjustment of blood flow mediated by tumour vessel response. However, in chronically hypoxic tissue, tumour vessels are unable to respond efficiently to these changes as a result of their altered form. A summary of the factors that determine oxygen availability and its influence on oxygen distribution in tumour tissue is presented in Figure 2.4

Figure 1
Figure 1: Oxygen tension in solid tumours and normal tissues (adapted from reference 6)
Figure 2
Figure 2: Factors that determine the pO2 distribution in tumour tissue (adapted from reference 4)

Tumour hypoxia can occur in a chronic or acute form. The location, extent and the relative number of acutely or chronically hypoxic cells vary within and between individual tumours. However, in both chronic and acute hypoxia, oxygen deficiency initially develops in tissue distant from a functional tumour capillary and to the greatest extent at the venous end of the microvessel.5 A solid tumour that is chronically hypoxic shows a characteristic pattern (see Figure 3), whereby proliferating cells are situated at close proximity to a functional blood vessel and necrotic cells are situated in an outer layer. The cells that lie between the inner healthy cells and the outer, necrotic cells are chronically hypoxic and phenotypically adapted to these conditions.6 These cells are continually replaced by cells that are displaced from their position nearest to the blood vessel,7 and, in a variety of human and mouse tumours, have been found to lie in a layer situated at a distance of between 50 and 250µm.8 Anoxia, ie, a complete lack of oxygen, occurs in tumours with a radius in the region of 145µm or larger.1
Acute hypoxia occurs as a result of transient interruptions in blood flow through individual blood vessels within a tumour. The morphological changes in tumour vasculature are a possible cause, together with spontaneous vasomotion of upstream arterioles, an explanation substantiated by the experimental use of vasoactive agents such as angiotensin II and flunarizine to induce acute hypoxia.7

Figure 3
Figure 3: Pattern of viable, hypoxic and necrotic cells in a solid tumour (taken from reference 5, with permission)

Problems associated with tumour hypoxia

Cells exposed to a hypoxic environment are radio- and chemo-resistant. These resistant cells can remain viable after treatment, leading to the possibility of recurrence and metastasis on cessation of therapy. In addition, the adaptations necessary to maintain viability of these cells in hypoxic conditions may cultivate a more malignant phenotype - there is some evidence that hypoxia favours selection of aberrant p53 tumour suppresser gene9 and that exposure to hypoxia increases tumour growth rate and experimental metastatic potential.10 A correlation has also been found between hypoxia-induced tumour-derived angiogenesis, quantified by means of VEGF-induced areas of dense vascularisation, or "vascular hot spots", and poorer five-year survival and increased local recurrence in patients with cervical tumours.11 Cells cultured in anoxic conditions, ie, in the absence of oxygen, are approximately three times more resistant to the cytotoxic effects of radiation than fully aerated cells, radiosensitivity rapidly increasing with increasing oxygen concentration. The term used to express this effect is the oxygen enhancement ratio (OER), which is the ratio of radiation doses under hypoxic versus aerobic conditions to produce a similar biological effect.
Radioresistance occurs due to the deficiency of molecular oxygen, which acts as a radiosensitiser due to its action as an electron acceptor. As a consequence, the free radical formation that results in cell damage occurs to a much lesser extent under hypoxic conditions, leading to survival of these dangerous cells. Hypoxia also confers changes in radiosensitivity due to the proliferative, physiologic and metabolic changes associated with cells exposed to this adverse microenvironment, eg, DNA repair following radiotherapy may be augmented.12

Adaptation to hypoxia

Viable hypoxic cells possess certain characteristics that enable them to survive the adverse conditions in which they are placed. The presence of hypoxia in mammalian tissue effects systemic and metabolic responses, such as increased ventilation, cardiac output and haemopoiesis, as well as biochemical changes, including an increase in anaerobic glycolysis and the level of protective stress proteins.
Changes occur in the expression of a variety of biochemicals, including cytokines and growth factors like erythropoietin (EPO) and vascular endothelial growth factor (VEGF), glycolytic enzymes and cellular redox regulators such as NADPH. Changes occur in the expression of various transcription factors, such as AP-1, NFkB, and HIF-1 (for review, see Wang and Semenza 199613).
The HIF-1 transcription factor HIF-1 (hypoxia-inducible factor 1) is of particular interest to us, and others investigating hypoxia, as it mediates the transcriptional activation of hypoxia-inducible genes in tumours. Studies have implicated the HIF-1 transcriptional response in the induction of glucose transporters, VEGF and certain glycolytic enzymes.14-16 This results in the up-regulation of processes advantageous to tumour growth, such as glucose transport and angiogenesis.
HIF-1 is part of a family of heterodimeric basic helix-loop-helix (bHLH) proteins, composed of two subunits, HIF-1a and HIF-1b (see Figure 4). The latter is alternatively designated the aryl hydrocarbon receptor nuclear translocator (ARNT) and is also involved in the response to drugs and environmental toxins. Expression of the mRNA of both components of HIF-1 has been detected in all organs. However, HIF-1b shows greater variation in its expression between organs. It has been noted that a transient up-regulation of HIF-1a occurs with a decrease in oxygen concentration from 21 per cent to 7 per cent,17 and further work has shown that levels of HIF-1a and HIF-1b are inversely related to O2 concentration, with a maximal response at 0.5 per cent O2. HIF-1 binding activity shows an increase that corresponds closely to the increase in HIF-1a, suggesting that HIF-1b is in excess, perhaps a reflection on the HIF-1b not being unique to the hypoxic response.18 HIF-1a is intrinsically unstable in normoxia, and so it is likely that activation of HIF-1 occurs upon post-translational stabilisation of the HIF-1a subunit.19 Dimerisation with HIF-1b induces a conformational change, allowing the complex to bind a DNA sequence termed the hypoxia responsive element (HRE).20 Work undertaken to investigate the processes involved in the expression of the hypoxia-inducible EPO gene has enabled the elucidation of a 5' TACGTGCT 3' HRE sequence.13 This acts as a common enhancer of hypoxia-inducible genes, the HIF-1 DNA binding complex having a role in the induction of hypoxia-responsive genes in non-EPO producing, as well as EPO producing cells.21 The characterisation of this DNA sequence has important implications for the development of gene therapy, where HRE driven genes administered to hypoxic tumour tissue can be used to induce expression of these diagnostic and therapeutic genes for cancer treatment.22

Figure 4
Figure 4: Binding of HIF-1 heterodimer to the 5' TACGTTGCT 3' hypoxia responsive element (HRE) can only take place upon dimerisation of HIF-1a and HIF-1b (ARNT) (adapted from reference 20)

Hypoxia and inflammation

The association between hypoxia and inflammation is well established, and experimental models of ischaemia and reperfusion injury have enabled the elucidation of hypoxia regulated expression of certain inflammatory mediators. Reperfusion injury is mediated by the accumulation of hypoxanthine and xanthine oxidase during ischaemic episodes, which occur as a result of the conversion of xanthine dehydrogenase to xanthine oxidase. The metabolism of hypoxanthine to xanthine by xanthine oxidase is oxygen dependent. Therefore, sufficient metabolism is hindered in ischaemic conditions. When reperfusion takes place, and normal oxygen tensions resume, xanthine oxidase is able to convert the accumulated hypoxanthine to xanthine with the generation of large amounts of superoxide ion, H2O2 and other oxygen radicals. These metabolites initiate lipid peroxidation of the cell membranes with the subsequent release of chemotactic agents, including arachidonic acid products such as leukotriene B4, thromboxane A2, and the interleukins IL-1 and IL-8. These chemotactic agents play a significant part in the migration, adhesion and activation of neutrophils, which are key players in the inflammatory process and have been implicated in the resulting tissue injury.23–25
The precise complement of chemotactic agents, and their regulation, is not fully understood. Hypoxia profoundly affects production of IL-2, IL-4 and interferon gamma (IFNg), but decreases the production of IL-10 by human monocytes.26 It is also known that hypoxia induces the production of the chemotactic agents IL-8, IL-1 and active IL-1a by macrophages and endothelial cells, both at the mRNA and the protein level. However, the duration of hypoxia, and the need for reoxygenation to achieve up-regulation and the downstream effects of these agents varies according to the experimental model that has been used.27–30 In addition, it must also be considered that hypoxia induces production of IL-6, which has been shown to possess anti-inflammatory properties such as the suppression of IL-1.31 Therefore, it could be said that the extent of inflammation is, in part, dependent upon the balance of the effects of hypoxic stimulation of particular interleukins.
Vasodilatation and increased vascular permeability accompany inflammation, which is a way of ensuring delivery of various leukocytes and their products. Important vasoactive substances include nitric oxide and the prostaglandins, production being mediated by the action of nitric oxide synthases (NOS) and cyclo-oxygenases (COX-1 and COX-2), respectively. Although it is known that hypoxia induces expression of COX-232 and is associated with maintenance of prostaglandin E2 in rheumatoid synovial fibroblasts,33 its effect on NOS expression is controversial. Unfortunately, different experimental models have provided conflicting results. One form of the enzyme, known as iNOS, is inducible by inflammatory mediators such as the interleukins, which are in turn hypoxia inducible. Endothelial nitric oxide synthase (eNOS) is also indirectly induced in hypoxic conditions, via an increase in intracellular calcium ions, which are involved in the activation of the enzyme.34
It is well established that polymorphonuclear leukocytes such as neutrophils have a detrimental, rather than the initially suggested beneficial effect on healing post myocardial infarction. Neutrophil-mediated endothelial injury takes place when the neutrophil binds via adhesion receptors to endothelial ligands such as the E-selectins (ELAM-1, CD62E) and leukocyte integrins such as ICAM-1, ICAM-2 and LFA-1. Cultured endothelial and muscle cells increase their adhesiveness for neutrophils under hypoxic conditions via an LFA-1 dependant mechanism.35 Further work has identified the existence of an adhesion ligand detectable by an antibody with affinity to an endothelial cell surface ligand termed HAL 1/13. Although there is no increase in HAL 1/13 density in response to hypoxia, the functionality of the molecule, ie, its contribution to leukocyte adherence, increases significantly in these conditions.36 The contribution of ICAM-1 and ICAM-2 to hypoxia-induced leukocyte adhesion is controversial. Whereas some studies have failed to find significant involvement,36,37 further work has demonstrated that a combination of hypoxia and lipopolysaccharide acts as a stimulus for ICAM-1 expression and ICAM-1 mediated leukocyte adhesion. However, this effect is not universal, occurring concomitantly with hypoxia induced expression of the nuclear transcription factor NF-kB.38 Anti-IL-1 antibodies suppress hypoxia induced leukocyte adherence, indicating a central role for IL-1.27 In addition, it has been demonstrated that hypoxia-induced IL-1 release leads to a reperfusion linked autocrine enhancement in the expression of adhesion molecules such as ICAM-1.30 E-Selectin production is stimulated in aortic endothelial cells by a combination of hypoxia and lipopolysaccharide, or hypoxia and tumour necrosis factor (TNF) possibly via a hypoxia induced decrease in c-AMP production.39 This is of particular significance, as it is has been shown that ischaemia stimulates TNF production in hepatic reperfusion injury.40
The activation of the neutrophil initiates the release of proteolytic enzymes and reactive oxygen metabolites into the extracellular space, culminating in the lysis of essential structural matrix proteins, increased microvascular permeability and oedema.23 Neutrophil-mediated tissue damage is limited during inflammation by apoptosis, or the programmed death of these cells. Hypoxia prolongs the survival of neutrophils,41 and has also been used experimentally to derive a population of macrophages that are also resistant to apoptosis, and will therefore perpetuate the production of inflammatory mediators, and conversely, the activation of these resistant neutrophils.42

Cardiac implications of hypoxia

The outcome after a myocardial infarction is profoundly limited by the incidence of reperfusion injury, increasing the chance of arrhythmias, myocardial stunning and infarct size.43 Paradoxically, hypoxia regulates growth, proliferative capacity and collagen type-1 production in cardiac fibroblasts, and therefore may play a part in the post-infarct remodelling of the cardiac collagen matrix.44
As well as inflammation, myocardial injury occurs as a result of cardiomyocyte apoptosis.45 In fact apoptosis has been shown to represent the major independent form of myocyte cell death during, and immediately after infarction.46 Hypoxia induces apoptosis, this phenomenon having been demonstrated in many cell types.47 A study using rat myocardium has shown that apoptosis occurring during 45 minutes of ischaemia advances substantially during 3 hours' reperfusion. In an effort to elucidate the regulation of hypoxia-induced apoptosis in the myocardium, experiments have been undertaken to examine levels of certain pertinent proteins. For example, co-localisation of apoptotic cells occurs with the cysteine protease caspase 3, which belongs to a family of enzymes common in various models of mammalian programmed cell death.48 There is also evidence for a role for the enzyme c-Jun kinase, which is induced upon reoxygenation.49 The role of p53, which is a protein central to the regulation of tumour cell apoptosis, is dependent on other tumour microenvironmental factors such as pH.50 Apoptosis found in hypoxic cardiomyocytes has been found to be concomitant with raised p53 activity, indicating a critical role for p53 activated pathways.51 However, it has been demonstrated that hypoxia-induced apoptosis seen in acute myocardial infarction also occurs in the absence of p53.52
Diagnosis of acute myocardial infarction is facilitated by raised levels of the enzymes lactate dehydrogenase (LDH) and creatinine kinase (CK), levels of which increase in response to hypoxia/reoxygenation, along with a corresponding increase in levels of nitric oxide.53 It is logical that the vasodilatory effect of nitric oxide contributes to the inflammatory response, ultimately leading to reperfusion injury. The production of this chemical is catalysed by the enzyme nitric oxide synthase (NOS). There are three types. Perhaps the most pertinent when considering cardiac ischaemia is eNOS, which is produced constitutively by blood vessel endothelia, and is responsible for the synthesis of nitric oxide or, as it was previously known, endothelium-derived relaxing factor (EDRF). The two other forms are cytokine-inducible NOS (iNOS), which is calcium independent and important in the immunogenic response of T-lymphocytes, macrophages and reticulo-endothelial cells, and nNOS, which is the neuronal form of nitric oxide synthase.34
Although it is known that nitric oxide is released from cardiomyocytes in response to hypoxia/reoxygenation,53 a decreased production of endothelial nitric oxide by both acute and chronic hypoxia has been observed in pulmonary arteries.54 Likewise, experiments undertaken to elucidate the effect of acute or chronic ischaemia/hypoxia on the expression of these enzymes in the cardiopulmonary system has failed to provide an equivocal answer, providing evidence for both up- and down-regulation of these enzymes in hypoxic conditions.55–59 The diverse role of nitric oxide as a haemodynamic agent, acting both locally and systemically on a variety of tissues in response to different degrees of hypoxia, precludes a full understanding of the subject. However, further investigation of the nitric oxide pathways will uncover useful therapeutic targets. These include a role for nitric oxide stimulation in the protection of transplanted vascular organs from the effects of ischaemia/reperfusion,60 and the use of gene therapy where eNOS could be produced under hypoxic conditions to treat pulmonary arterial hypertension.61 There is also believed to be a link between the cardio-protective effects of ACE inhibitors, such as ramipril, and its stimulation of nitric oxide production.62

Opportunities for therapy in hypoxic conditions

The role of hypoxia in solid tumours, cardiac ischaemia and the aetiology of inflammation and reperfusion injury has been discussed in detail, and therefore extends to conditions such as rheumatoid and other arthritides, chronic inflammatory bowel disease and skin conditions such as eczema and psoriasis. Inflammation and, therefore, hypoxia feature strongly in the mechanisms producing vasculopathies, such as vasculopathy following transplantation,63 diabetic inflammatory vasculopathy64 and diabetic neuropathy.65 Chronic respiratory diseases such as chronic bronchitis, cystic fibrosis and tuberculosis cause inflammation, and it is logical that the accompanying decrease in lung function leads to a hypoxemic state.
Reperfusion injury, aside from its impact on cardiac function, also leads to cerebral, hepatic, renal66 and intestinal complications such as hypoxia induced necrotising enterocolitis.67 Hypoxia induced oxyradical formation has been implicated in the evolution of atherosclerotic lesions,68 and decreased blood flow to the extremities, with the ensuing local hypoxia, can lead to ulceration.69 Meanwhile, in the area of sports medicine, repetitive activity resulting in a deficiency of nutrients and tissue hypoxia can precede tendon degeneration.70
Bioreductive drugs are currently being developed as a means of selectively targeting hypoxic tumour cells. Hypoxia directed drug delivery systems might offer the potential for selective treatment in a whole variety of conditions manifesting chronic hypoxia and inflammation, which are routinely treated using glucocorticoids and/or non-steroidal anti-inflammatory drugs and other therapeutic strategies. This offers the advantage of tissue specificity with reduced toxicity.

Bioreductive drugs

Bioreductive drugs were initially designed to diagnose the presence of, and be toxic towards, the hypoxic cells in tumours. There are four main classes of bioreductive drugs currently undergoing clinical or preclinical evaluation in the treatment and prognostication of solid tumours. These are:

For these agents to act as therapeutic or diagnostic agents in hypoxic conditions, a reductive activation must take place (Figure 5).71
Figure 5
Figure 5: The hypoxia-mediated activation of bioreductive drugs to give cytotoxic or diagnostic entities.
Complete reduction only occurs in hypoxia as the first reductive step is inhibited by oxygen

The formation of the one-electron reduced intermediate is in equilibrium with its reoxidation by molecular oxygen, which results in "futile cycling" and the production of a superoxide radical, thought to be responsible for the aerobic toxicity seen in bioreductive cancer agents undergoing clinical trials. The position of this equilibrium depends upon the concentration of the drug, cellular oxygen tension and its redox potential. Consequently, prodrugs vary in the extent of hypoxia necessary for their bioreduction.72 Activation of these prodrugs is catalysed by various families of reductases, the most important to be considered being the cytochrome P450 reductase and cytochrome P450 family, which are the focus of much research into bioreductive drug metabolism.73,74 A scheme for the activation of the bioreductive drug tirapazamine is shown in Figure 6, which involves a hypoxia-dependent one-electron reduction to an intermediate tirapazamine radical. Hydrogen abstraction from DNA in the cell results in the formation of the 2-electron reduction product SR 4317, which is not toxic to normoxic or hypoxic cells. The intermediate tirapazamine radical, however, acts as the cytotoxic species, inducing the formation of single DNA strand breaks.72

Figure 6
Figure 6: The metabolic activation of tirapazamine induced by hypoxia. The metabolites SR4317 and SR4330 produced by a 2-electron reductive process are not toxic, whereas the intermediate tirapazamine radical is thought to act as a DNA strand breaker (taken from reference 75, with permission)

Clinical evaluation of bioreductive drugs

Hypoxic cell populations are resistant to both radiotherapy and chemotherapy. The aim of bioreductive chemotherapy is not only to overcome this resistance, but also to exploit this feature by using agents that selectively kill these cells, rendering the tumour more responsive to other therapies, as well as reducing the chance of metastasis and relapse. In addition, it is hoped that another clinical benefit will be a reduced necessity for repeated treatment that may entail a poor response rate with an increased toxicity to the patient.
Clinical trials so far have investigated the role of bioreductive drugs in multimodality therapy, ie, in combination with radiotherapy or chemotherapy. Experimental systems have shown that hypoxia-selective drugs target hypoxic cell populations while either cytotoxic agents or radiotherapy target the surviving aerobic cell population. For example, the antitumour effectiveness of tirapazamine and RSU1069, a 2-nitroimidazole bioreductive drug, has been shown to be effective when used in combination with single high doses of X-rays. The radiation will preferentially kill aerobic cells, tumour response being dependent upon the extent of the remaining hypoxic cell population. The administration of a bioreductive drug immediately after irradiation reduces the number of hypoxic cells and improves tumour response.72 Fractionated radiotherapy in conjunction with bioreductive drugs is also effective, as it allows rehypoxiation, ie, reversion to hypoxic status between fractionations, which provides the conditions necessary for bioactivation of the drug.76
Bioreductive drugs have been shown to potentiate the response to other antitumour agents. Nitroimidazoles potentiate alkylating agents in a mechanism that involves modulation of thiol levels, inhibition of repair and alteration of alkylating pharmacokinetics, eg, misonidazole used with melphalan.76 Clinical trials of tirapazamine in combination with cisplatin are currently in phase III. The observed potentiation effect is schedule dependent, with maximum tumour cell kill seen when tirapazamine is administered approximately 2.5 hours before cisplatin.77 This potentiation appears to be related to the DNA repair capacity of a particular cell type. Evidence supporting this notion comes from a pair of ovarian cell lines, one of which is resistant to cisplatin (due to increased repair capacity). Killing of this latter line is not potentiated by tirapazamine, whereas the wild type cell line from which it is derived is 600 times more sensitive to the tirapazamine/cisplatin combination.78 Phase I clinical trials have defined the maximum tolerated dose of tirapazamine as 390mg/m2, alone and in combination with 75mg/m2 cisplatin. The combination shows no significant enhancement of systemic toxicity,77 the major side effects of tirapazamine being acute reversible hearing loss and muscle cramps.79 Phase I/II trials have shown this combination to be effective in malignant melanoma, metastatic and recurrent cervical, lung, bladder, and head and neck cancers. Phase III CATAPULT (cisplatin and tirapazamine against platinum alone in previously untreated lung tumours) trials conclude that tirapazamine in combination with cisplatin prolongs survival in patients with advanced non small cell lung cancer. One year survival is significantly improved and tumour response rate is doubled in comparison to cisplatin alone.80 Trials assessing the use of tirapazamine in combination with radiation, other agents such as vinorelbine, paclitaxel, fluorouracil, and various other two- and three-agent combinations are also taking place.

Individualisation of bioreductive chemotherapy in cancer

It is now acknowledged that the severity and advancement of disease, as well as the predicted response, should be considered when designing a treatment regimen for an individual patient. Tumour response to both bioreductive agents and radiation is modulated by the hypoxic fraction, ie, the proportion of hypoxic cells in a tumour. This varies spatially and temporally between individual tumours.
In order to individualise treatment with bioreductive drugs, it is desirable to obtain an accurate prediction of the hypoxic fraction of cells, or the extent of hypoxia, so that the drug may be rationally applied. Various methods of measurement, based on nitroimidazole hypoxia markers, have been investigated.81–83 However, any method used must account for cell populations that are transiently or acutely hypoxic, as well as those that are chronically hypoxic. Ideally, samples should be obtained at biopsy, or incorporate a non-invasive, non-toxic means of detection. This would be especially convenient as hypoxic fraction could be predicted from the same sample as other prognostic indicators, potentially reducing the number of necessary procedures. Methods that have achieved the greatest success so far include direct measurement of pO2, using a polarographic oxygen microelectrode, and bioreductive chemical markers.
Oxygen electrodes enable direct measurement of oxygen tension within tissue by means of a 300µm probe that contains a membranised gold microcathode 12µm in diameter. This is held at –700mV relative to a silver chloride cathode. To be an accurate predictor, the electrode must be capable of detecting the low oxygen tensions that are representative of radiobiological hypoxia. The spatial and temporal heterogeneity of pO2 values necessitates a rapid response time, this arrangement recording 90 per cent of available oxygen in 500ms.84 The value of oxygen electrodes has been demonstrated when measuring intra- and inter-tumour variability, though the extent of the variability also shows that a minimum of measurements are necessary to gain a true prediction of tumour oxygenation.4 Oxygen electrodes have been used successfully to measure tumour hypoxia in various sites such as cervix, breast, head and neck (for review see Stone85). However, the technique has obvious limitations, in that it is an invasive tool and necessitates a surgical procedure.
The use of bioreductive drugs as markers of hypoxia is possible due to the formation of intracellular adducts derived from the covalent binding of reduced metabolites to cellular components. These intracellular adducts can be imaged non-invasively using several methods, including PET (positron emission tomography),86 or by analysis of biopsy samples by flow cytometry and immunohistochemical techniques.82 The most prolific markers tested to date are the nitroimidazoles, particularly the 2-nitroimidazoles, some examples of which are shown in Figure 7. For example, NITP and pimonidazole, whose bound adducts are detected immunologically, whereas SR4554 can be successfully detected by fluorine-magnetic resonance spectroscopy. In contrast to their therapeutic counterparts, bioreductive markers can be used at much lower doses, which provides much optimism that these agents will be available for diagnosis not only of tumours but of non-malignant diseases as well.87
Figure 7
Figure 7: Examples of 2-nitroimidazoles used as bioreductive markers of hypoxia

While the assessment of the level of hypoxia and/or the fraction of radiobiologically hypoxic cells in tumours is an important aspect for the rational application of bioreductive drugs, it will also be aided by a knowledge of the levels of cellular reductases which can activate these prodrugs under hypoxic conditions. As mentioned previously, one of the most important of these is P450 reductase. It has been shown that the cellular levels of this enzyme are important for the activation and hypoxic toxicity of tirapazamine, nitroheterocyclics and indolequinones.88,89 Thus it may be particularly advantageous to target these drugs towards tumours that are not only shown to be hypoxic but also have high levels of this reductase.71 Information on both of these requirements for the individualisation of bioreductive drug therapy may come from the use of the bioreductive drug "markers" of hypoxia detailed above, although quantitative relationships between the level of hypoxia, reductase expression and invasive and non-invasive measurement of the appropriate marker needs to be established.
Intrinsic markers of hypoxia offer the advantage of providing a non-invasive means of detection that can be carried out without the addition of exogenous agents at the time of biopsy. In oncology, the hypoxia-dependent product VEGF has been shown to exist in a variety of cancers.90 VEGF expression has been associated with areas of high vascular density (vascular hot spots)11 and has been identified as an adverse prognostic indicator. There is evidence to suggest that VEGF expression is related to the overall level of hypoxia in the tumours,91 but its relationship to pO2 measurements and the binding of the bioreductive hypoxia markers remains to be established. Other genes that could act as surrogates to predict the level of hypoxia might include platelet-derived endothelial cell growth factor (PD-ECGF) - which is a hypoxia-regulated enzyme - thymidine phosphorylase,92 and some glucose transporters, but as with VEGF their use still needs to be validated. Similarly, there is much current research directed towards the measurement of the levels of the transcription factor HIF-1 (hypoxia inducible factor 1). This heterodimeric protein regulates the expression of many hypoxia-regulated genes and may provide an even more precise measure of hypoxia in tumours and other diseased tissues.93

Conclusion

There is no doubt that hypoxia is present in solid tumours and is a feature of other debilitating diseases. This is providing a robust basis on which to develop novel diagnostic and therapeutic strategies for the treatment of these diseases. However, as we enter the new millennium, the call will be for greater individualisation of treatment. Pharmacists will have a central role in the identification of those patients most likely to respond to a given treatment, thus allowing intelligent clinical trial design and, later, dose adjustment and the avoidance of unnecessary treatment. The development of hypoxic markers, be it extrinsic or intrinsic, will help us meet this challenge.

ACKNOWLEDGMENTS Rachel Airley is supported by a CRISP (Collaborative Research Studentship Programme for Pharmacy) award from the Royal Pharmaceutical Society.

Rachel Airley and Jane Monaghan are graduate students and Ian Stratford is professor of pharmacy in the Experimental Oncology Group, School of Pharmacy and Pharmaceutical Sciences, University of Manchester, Oxford Road, Manchester M13 9PL. Correspondence to Miss Airley

References

1. Nias AHW. An introduction to radiobiology (2nd edition). London: J. Wiley & Sons, 1998.
2. Waleh NS, Brody MD, Knapp MA, Mendinca HL, Lord EM, Koch CJ et al. Mapping of vascular endothelial growth factor-producing hypoxic cells in multicellular tumour spherioids using a hypoxia-specific marker. Cancer Res 1995;55: 6222-6.
3. Shweiki D, Neeman M, Itin A, Keshet E. Induction of vascular endothelial growth factor expression by hypoxia and by glucose deficiency in multicell spheroids: implications for tumour angiogenesis. Proc Natl Acad Sci USA 1995;92:768-72.
4. Vaupel, Kelleher, Gunderoth (editors). Tumour oxygenation. No 24. Functionanalyse Biologischer Systeme. 1995.
5. Vaupel P, Kallinowski F, Okunieff P. Blood flow, oxygen and nutrient supply, and metabolic environment of human tumours: a review. Cancer Res 1989;49:6449-65.
6. Hall EJ. Radiobiology for the radiologist. Philadelphia:Lippincott, 1988
7. Horsman MR, Overgaard J.The oxygen effect. In: Steel GG (editor). Basic clinical radiobiology (2nd edition). London: Arnold, 1997:132-40.
8. Sutherland RM. Cell and environment interactions in tu-mour microregions: the multicell spheroid model. Science 1988;248:177-84.
9. Graeber TG, Osmanian C, Jacks T, Housman DE, Koch CJ, Lowe SW et al. Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature 1996; 379:88-91.
10. Stackpole CW, Groszek L, Kalbag SS. Benign to malignant B16 melanoma progression induced in two stages in vitro by exposure to hypoxia. J Natl Cancer Inst 1994;86:361-7.
11. Cooper RA, Wilks DP, Logue JP, Davidson SE, Hunter RD, Roberts SA et al. High tumour angiogenesis is associated with poorer survival in carcinoma of the cervix treated with radiotherapy. Clin Cancer Res 1998;4:2795-800.
12. Rockwell S. Use of hypoxia-directed drugs in the therapy of solid tumours. Sem Oncol 1992;19(4 Suppl 11):29-40.
13. Wang GL, Semenza GL. Oxygen sensing and response to hypoxia by mammalian cells. Redox Report 1996;2:89-96.
14. Maxwell PH, Dachs GU, Gleadle JM, Nicholls LG, Harris, AL, Stratford IJ et al. Hypoxia-inducible factor-1 modulates gene expression in solid tumours and influences both angiogenesis and tumour growth. Proc Natl Acad Sci USA 1997; 94:8104-9.
15. Semenza GL, Jiang B, Leung SW, Passantino R, Concordet J, Maire P et al. Hypoxia-responsive elements in the aldolase A, enolase 1 and lactate dehydrogenase A gene promoters contain essential binding sites for hypoxia-inducible factor-1. J Biol Chem 1996;271:32529-37.
16. Ebert BL, Gleadle JM, O'Rourke JF, Bartlett SM, Poulton J, Ratcliffe PJ. Iso-enzyme-specific regulation of genes involved in energy metabolism by hypoxia: similarities with the regulation of erythropoietin. Biochem J 1996;313:809-14.
17. Weiner CM, Booth G, Semenza GL. In vivo expression of mRNA's encoding hypoxia-inducible factor-1. Biochem Biophys Res Commun 1996;225:485-8.
18. Jiang B, Semenza GL, Bauer C, Marti HH. Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am J Physiol 1996;271(4 pt 1):C1172-80.
19. Huang LE, Arany Z, Livingston DM, Bunn HF. Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its a subunit. J Biol Chem 1996;271:32253-9.
20. Kallio PJ, Pongratz I, Gradin K, McGuire J, Poellinger L. Activation of hypoxia-inducible factor 1a: post-transcriptional regulation and conformational change by recruitment of the ARNT transcription factor. Proc Natl Acad Sci USA 1997;94:5667-72.
21. Wang GL, Semenza GL. General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. Proc Natl Acad Sci USA 1993;90:4304-8.
22. Dachs GU, Patterson AV, Firth JD, Ratcliffe PJ, Townsend KMS, Stratford IJ et al. Targeting gene expression to hypoxic tumour cells. Nature Med 1997;3:515-20.
23. Welbourne CRB, Goldman G, Paterson IS, Valeri CR, Shepro D, Hechtman HB. Pathophysiology of ischaemia reperfusion injury: central role of the neutrophil. Br J Surg 1991;78: 651-5.
24. Granger DN. Role of xanthine oxidase and granulocytes in ischaemia-reperfusion injury. Am J Physiol 1988;255(Suppl): H1269-75.
25. Sekido N, Mukaida N, Harada A, Nakanishi I, Watanabe Y, Matsushima K. Prevention of lung reperfusion injury in rabbits by a monoclonal antibody against interleukin-8. Nature 1993;365:654-7.
26. Naldini A, Carraro F, Silvestri S, Bocci V. Hypoxia affects cytokine production and proliferative responses by human peripheral mononuclear cells. J Cell Physiol 1997;173:335-42.
27. Koga S, Ogawa S, Kuwabara K, Brett J, Leavy JA, Ryan J et al. Synthesis and release of interleukin 1 by reoxygenated human mononuclear phagocytes. J Clin Invest 1992;90:1007-15.
28. Metinko AP, Kunkel SL, Standiford TJ, Strieter RM. Anoxia-hyperoxia induces monocyte-derived interleukin-8. J Clin Invest 1992;90:791-8.
29. Karakarum M, Shreeniwas R, Chen J, Pinsky D, Yan SD, Anderson M et al. Hypoxic induction of interleukin-8 gene expression in human endothelial cells. J Clin Invest 1994;93: 1564-70.
30. Shreeniwas R, Koga S, Karakarum M, Pinsky D, Kaiser E, Brett J et al. Hypoxia-mediated induction of endothelial cell interleukin 1a. J Clin Invest 1992;90:2333-9.
31. Yan SF, Tritto I, Pinsky D, Liao H, Huang J, Fuller G et al. Induction of interleukin (IL-6) by hypoxia in vascular cells. J Biol Chem 1995;270:11463-71.
32. Ji YS, Xu Q, Scmedtje JF Jr. Hypoxia induces high-mobility-group protein I (Y) and transcription of cyclooxygenase-2 gene in human vascular endothelium. Circ Res 1998;83:295-304.
33. McGough KA, Jackson JR, Minton JA, Marshall LA, Jacobs RS, Winkler JD. Inflammatory PGE2 production is maintained during hypoxia in rheumatoid synovial fibroblasts. Inflamm Res 1997;46(Suppl 2):S147-8.
34. Chinje EC, Stratford IJ. Role of nitric oxide in growth of solid tumours: a balancing act. Essays Biochem 1997;32:61-72.
35. Ginis I, Mentzer SJ, Faller DV. Oxygen tension regulates neutrophil adhesion to human endothelial cells via an LFA-1 dependent mechanism. J Cell Physiol 1993;157:569-78.
36. Ginis I, Mentzer SJ, Li X, Faller DV. Characterization of a hypoxia-responsive adhesion molecule for leukocytes on human endothelial cells. J Immunol 1995;155:802-10.
37. Yoshida N, Granger DN, Anderson DC, Rothlein R, Lane C, Kveietys PR. Anoxia/reoxygenation induced neutrophil adherence to cultured endothelial cells. Am J Physiol 1992; 262 (Suppl):H1891-7.
38. Zund G, Uezono S, Stahl GL, Dzus AL, Mcgowan FX, Hickey PR et al. Hypoxia enhances induction of endothelial ICAM-1; role for metabolic acidosis and proteasomes. Am J Physiol 1997;273(Suppl):C1571-80.
39. Zund G, Nelson DP, Neufeld EJ, Dzus AL, Bischoff J, Mayer JE et al. Hypoxia enhances stimulus-dependent induction of E-selectin on aortic endothelial cells. Proc Natl Acad Sci USA 1996;93:7075-80.
40. Colletti LM, Remick DG, Burtch GD, Kunkel SL, Strieter RM, Campbell DA. Role of tumour necrosis factor-a in the pathophysiologic alterations after hepatic ischaemia/reperfusion injury in the rat. J Clin Invest 1990;85:1936-43.
41. Hannah S, Mecklenburgh K, Rahman I, Bellingan GJ, Greening A, Haslett C et al. Hypoxia prolongs neutrophil survival in vitro. FEBS Lett 1995;372:233-7.
42. Yun JK, McCormick TS, Villabona C, Judware RR, Espinosa MB, Lapetina EG. Inflammatory mediators are perpetuated in macrophages resistant to apoptosis induced by hypoxia. Proc Natl Acad Sci USA 1997;94:13903-8.
43. Grech ED, Jackson MJ, Ramsdale DR. Reperfusion injury after acute myocardial infarction. BMJ 1995;310:477-8.
44. Agocha A, Lee HW, Eghbali-Webb M. Hypoxia regulates basal and induced DNA synthesis and collagen type 1 production in human cardiac fibroblasts: effect of transforming growth factor-beta-1, thyroid hormone, angiotensin II and basic fibroblast growth factor. J Mol Cell Cardiol 1997;29: 2233-44.
45. Gottlieb RA, Burleson KO, Kloner RA, Babor BM, Engler RL. Reperfusion injury induces apoptosis in rabbit cardiomyocytes. J Clin Invest 1993;94:1621-8.
46. Kajstura J, Cheng W, Reiss K, Clark WA, Sonnenblick EH, Krajewski S et al. Apoptotic and necrotic cell deaths are independent contributing variables of infarct size in rats. Lab Invest 1996;74:86-107.
47. Tanaka M, Ito H, Adachi S, Akimoto H, Nishikawa T, Kasajima T et al. Hypoxia induces apoptosis with enhanced expression of Fas antigen mRNA in cultured neonatal rat myocytes. Circ Res 1994;75:426-33.
48. Black SC, Huang JQ, Rezaiefar P, Radinovic S, Eberhart A, Nicholson DW et al. Co-localization of the cysteine protease caspase-3 with apoptotoc myocytes after in vivo myocardial ischaemia and reperfusion in the rat. J Mol Cell Cardiol 1998; 30:733-42.
49. Laderoute KR, Webster KA. Hypoxia/reoxygenation stimulates Jun-kinase activity through redox signaling in cardiac myocytes. Circ Res 1997;80:336-44.
50. Schmaltz C, Hardenbergh PH, Wells A, Fisher DE. Regulation of proliferation-survival decisions during tumor cell hypoxia. Mol Cell Biol 1998;18:2845-54.
51. Long X, Boluyt MO, Hipolito ML, Lundberg MS, Zheng JS, O'Neill L et al. p53 and the hypoxia-induced apoptosis of cultured neonatal rat cardiac myocytes. J Clin Invest 1997;99: 2635-43.
52. Bialik S, Greenen DL, Sasson IE, Cheng R, Horner JW, Evans SM et al. Myocyte apoptosis during acute myocardial infarction in the mouse localises to hypoxic regions but occurs independently of p53. J Clin Invest 1997;100:1363-72.
53. Wei L, Zhu H. The influence of hypoxia/reoxygenation on production and release of nitric oxide from cultured newborn rat cardiomyocytes. Chinese Med J 1996;76:697-99.
54. Shaul PW, Wells LB, Horning KM. Acute and prolonged hypoxia attenuate endothelial nitric oxide production in rat pulmonary arteries by different mechanisms. J Cardiovasc Pharmacol 1993;22:819-27.
55. Kacimi R, Long CS, Karliner JS. Chronic hypoxia modulates the interleukin-1 beta-stimulated inducible nitric oxide synthase pathway in cardiac myocytes. Circulation 1997;96: 1937-43.
56. LeCras TD, Xue C, Rengasamy A, Johns RA. Chronic hypoxia upregulates endothelial and inducible NO synthase gene and protein expression in rat lung. Am J Physiol 1996; 270(1 Pt 1):L164-70.
57. McQuillan LP, Leung GK, Marsden PA, Kostyk SK, Kourembanas S. Hypoxia inhibits expression of eNOS via transcriptional and post-transcriptional mechanisms. Am J Physiol 1994;267(5 pt 2):H1921-7.
58. Resta TC, Gonzales TC, Dail WG, Sanders TC, Walker BR. Selective up-regulation of arterial endothelial nitric oxide synthase in pulmonary hypertension. Am J Physiol 1997;272 (2 pt 2):H806-13.
59. Prabhaker NR, Pieramici SF, Premkumar DR, Kumar GK, Kalaria RN. Activation of nitric oxide synthase gene expression by hypoxia in central and peripheral neurons. Brain Res Mol Brain Res 1996;43:341-6.
60. Pinsky DJ, Oz MC, Koga S, Taha Z, Broekman MJ, Marcus AJ et al. Cardiac preservation is enhanced in a heterotopic rat transplant model by supplementing the nitric oxide pathway. J Clin Invest 1994;93:2291-7.
61. Setoguchi Y, Tamaki Y, Oka M, Kira S. Transfer of endothelial nitric oxide synthase gene in the purpose of gene therapy for pulmonary arterial hypertension. Nippon Rinsho 1996; 54:369-76.
62. Hartman, JC. The role of bradykinin and nitric oxide in the cardioprotective action of ACE inhibitors. Ann Thorac Surg 1995;60:789-92.
63. Hayry P. Chronic rejection: risk factors, regulation and possible sites of therapeutic intervention. Transplant Proc 1998; 30:2407-10.
64. Krendel DA. Diabetic inflammatory vasculopathy (letter and comment). Muscle Nerve 1997;20:520-1.
65. Dyck PJ, Giannini C. Pathologic alterations in the diabetic neuropathies of humans: a review. J Neuropathol Exp Neurol 1996;55:1181-93.
66. Edelstein CL, Ling H, Schrier RW. The nature of renal cell injury. Kidney Int 1997;51:1341-51.
67. Koloske AM. A unifying hypothesis and prevention of necrotizing enterocolitis. J Pediatr 1990;117(Suppl):S68-74.
68. Crawford DW, Blankenhorn DH. Arterial wall oxygenation, oxyradicals, and atherosclerosis. Atherosclerosis 1991;89:97-108.
69. Vanhoute PM, Corcaud S, De Montrion C. Venous disease: from pathophysiology to quality of life. Angiology 1997;48: 559-67.
70. Jarvinen M, Jozsa L, Kannus P, Jarvinen TL, Kvist M, Leadbetter W. Histopathological findings in chronic tendon disorders. Scand J Med Sci Sports 1997;7:86-95.
71. Stratford IJ, Workman P. Bioreductive drugs in the next millenium. Anti-cancer Drug Des 1998;13:519-28.
72. Stratford, IJ. Bioreductive drugs in cancer therapy. Br J Radiol 1992;(Suppl 24):128-36.
73. Patterson AV, Saunders MP, Chinje EC, Patterson LH, Stratford IJ. Enzymology of tirapazamine metabolism: a review. Anticancer Drug Des 1998;13:541-73.
74. Joseph P, Jaiswal AK, Stobbe CC, Chapman JD. The role of specific reductases in the intracellular activation and binding of 2-nitroimidazoles. Int J Radiation Biol Phys 1994;29:351-5.
75. Brown JM, Wang L. Tirapazamine: laboratory data relevant to clinical activity. Anticancer Drug Des 1998;13:529-39.
76. Brown JM. Targeting bioreductive agents to tumours: is it necessary to manipulate blood flow? Int J Rad Biol 1991;60: 231-6.
77. Dorie MJ, Brown JM. Tumour-specific, schedule-dependent interaction between tirapazamine (SR4233) and cisplatin. Cancer Res 1993;53:4633-6.
78. Monaghan JE, Stratford IJ. Overcoming cisplatin resistance with the bioreductive drug tirapazamine. Br J Cancer 1998;78 (Suppl 1):P41.
79. Doherty N, Hancock SL, Kaye S, Coleman CN, Shulman L, Marquez C et al. Muscle cramping in phase I clinical trials of tirapazamine (SR4233) with and without radiation. Int J Radiat Oncol Biol Phys 1994;29:379-82.
80. Pawel JV, Roemeling RV. Survival benefit from Tirazone and cisplatin in advanced non-small cell lung cancer (NSCLC) patients: final results from the international phase III CATAPULT I trial. Proc Am Soc Clin Oncol 1998; 1749.
81. Hodgkiss RJ, Wardman P. The measurement of hypoxia in tumours. Br J Radiol 1992;(Suppl 24):105-10.
82. Hodgkiss RJ. Use of 2-nitroimidazoles as bioreductive markers for tumour hypoxia. Anticancer Drug Des 1998;13:687-702.
83. Raleigh JA, Calkins-Adams DP, Rinker LH, Ballenger CA, Weissler MC, Fowler WC. Cancer Res 1998;58:3765-8.
84. Cruickshank GS, Rampling R. Peri-tumoural hypoxia in human brain: pre-operative measurement of tissue oxygen tension around malignant brain tumours. Acta Neurochir 1994; 60:375-7.
85. Stone HB, Brown JM, Philips TL, Sutherland RM. Oxygen in human tumours: correlation between methods of measurement and response to therapy. Rad Res 1993;136: 422-34.
86. Chapman JD, Coia LR, Stobbe CC, Engelhardt EL, Fenning MC, Schneider RF. Prediction of tumour hypoxia and radioresistance with nuclear medicine markers. Br J Cancer 1996;74(Suppl 27):S204-8.
87. Blake DR, Stevens CR, Sahinoglu T, Ellis G, Gaffney K, Edmonds S et al. Biochem Soc Trans 1997;25:812-6.
88. Patterson AV, Saunder M, Chinjie EC, Talbot DC, Harris AL, Stratford IJ. Overexpression of P450 reductase confers enhanced sensitivity to both tirapazamine and RSU1069. Br J Cancer 1997;76:1338-47.
89. Saunders M, Jaffar M, Patterson AV, Phillips RM, Harris AL, Stratford IJ. The importance of P450 reductase for determining the sensitivity of human tumour cells to the indolequinone E09. Biochem Pharmacol. In press.
90. Takahashi Y, Kitadai Y, Bucana CD, Cleary DR Ellis LM. Expression of vascular endothelial growth factor and its receptor, KDR, correlates with vascularity, metastasis and proliferation of human colon cancer. Cancer Res 1995;55: 3964-8.
91. Shwiki D, Itin A, Soffer D, Keshet E. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature 1992;359:843-5.
92. Griffiths L, Dachs GU, Bicknell R, Harris AL, Stratford IJ. The influence of oxygen tension on the expression of platelet-derived endothelial cell growth factor/thymidine phosphorylase in human breast tumour cells grown in vitro and in vivo. Cancer Res 1997;57:570-2.
93. Semenza GL. Expression of hypoxia-inducible factor 1: mechanisms and consequences. Biochem Pharmacol 2000;59: 47-53.