Effects of hyperglycemia on oxygenation, radiosensitivity and bioenergetic status of subcutaneous RIF-1 tumors

Since tissue oxygen tension is a balance between delivery and consumption of oxygen, considerable effort has been directed at increasing the former and/or decreasing the latter. Techniques to decrease the rate of cellular oxygen consumption (increasing the distance oxygen can diffuse into tissues) include increasing glycolysis by administering supra-physiologic levels of glucose. We have examined the effect of hyperglycemia produced by intravenous glucose infusion on the tissue oxygenation and radiation response of subcutaneously implanted murine radiation induced fibrosarcomas (RIF-1). A 0.3 M glucose solution was delivered via tail vein injection according to a protocol that maintained glucose at a plasma concentration of 17+/-1 mM. The effect of this treatment on radiation response (clonogenic and growth delay studies), tumor oxygenation (needle electrode pO2 and 2-[2-nitro-1H-imidazol-1-yl]-N-(2,2,3,3,3-pentafluoropropyl) acetamide (EF5) binding), and tumor bioenergetics and pH (31P NMR spectroscopy) was examined. Systemic measurements included hematocrit and blood glucose and lactate concentrations. The results of these studies suggest that these subcutaneously implanted RIF-1 tumors are both radiobiologically and metabolically hypoxic and that intravenous glucose infusion is not an effective method of modifying this metabolic state.

Vasa vasorum hypoperfusion is responsible for medial hypoxia and anatomic remodeling in the newborn lamb ductus arteriosus

Postnatal constriction of the full-term ductus arteriosus produces hypoxia of the muscle media. This is associated with anatomic remodeling (including smooth muscle death) that prevents subsequent reopening. We used late-gestation fetal and neonatal lambs to determine which factors are responsible for the postnatal hypoxia. Hypoxia [measured by 2-(2-nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl) acetamide technique] and cell death (measured by terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling technique) were observed in regions of the constricted ductus wall within 4 h after delivery. Although there was a decrease in ductus luminal flow during the first 6 h after delivery (measured by Doppler transducer), the amount of oxygen delivered to the ductus lumen (3070 +/- 1880 micromol O2 x min(-1) x g(-1)) far exceeded the amount of oxygen consumed by the constricted ductus (0.052 +/- 0.021 micromol O2 x min(-1) x g(-1), measured in vitro). Postnatal constriction increased the effective oxygen diffusion distance across the ductus wall to >3x the limit that can be tolerated for normal tissue homeostasis. This was owing to both an increase in the thickness of the ductus (fetus, 1.12 +/- 0.20 mm; newborn, 1.60 +/- 0.17 mm; p < 0.01) and a marked reduction in vasa vasorum flow (fetus, 0.99 +/- 0.44 mL x min(-1) x g(-1); newborn, 0.21 +/- 0.08 mL x min(-1) x g(-1); p < 0.01). These findings suggest that hypoxic cell death in the full-term ductus is caused primarily by changes in vasa vasorum flow and muscle media thickness and can occur before luminal flow has been eliminated. We speculate that in contrast with the full-term ductus, the preterm ductus is much less likely to develop the degree of hypoxia needed for vessel remodeling inasmuch as it only is capable of increasing its oxygen diffusion distance to 1.3x the maximally tolerated limit.

Hypoxia and VEGF mRNA expression in human tumors

High expression of circulating plasma vascular endothelial growth factor (VEGF) in patients with cancer is an indicator of poor treatment response. Similarly, hypoxia in tumors, as measured by oxygen needle electrodes, has been found to predict for tumor-treatment failure. These two predictors may be related because hypoxia is a potent stimulator of VEGF expression in vitro. However, the demonstration of a relationship between hypoxia and VEGF in human tumors has, to date, been indirect or even negative. The purpose of this study was to test whether this unexpected result was caused by factors unique to human tumors, or whether the prior results could have been influenced by the known complexities of VEGF regulation. Therefore, we undertook a direct assessment of VEGF induction in human tumors using in situ hybridization and compared its distribution with that of hypoxia, as measured by the distribution of adducts of the hypoxia marker EF5. The distribution of both markers was assessed in relationship to the distribution of blood vessels, as measured by antibodies to CD31. Our hypothesis was that VEGF mRNA and hypoxia would colocalize, assuming that detectability of the former was not limiting. Four squamous cell carcinomas, three sarcomas and one glioblastoma multiforme were studied. When VEGF mRNA signal was detectable, its maxima colocalized with regional maxima of EF5 binding. The strongest levels of both signals were sometimes adjacent to regions of tissue necrosis. However, we were unable to predict absolute levels of EF5 binding based on absolute levels of VEGF mRNA. Conversely, for all tumors studied, regions with relatively low levels of EF5 binding had relatively low or undetectable VEGF mRNA. We found moderate EF5 binding in some keratinized cells but VEGF mRNA was not expressed by these differentiated cells. The paradigm that hypoxia and VEGF expression are linked in human tumors is supported by the data presented herein. A better understanding of the biology behind VEGF expression, including its modulation by hypoxia, is important for optimizing its use as a prognostic indicator and/or modulating its presence with biologic therapies.

Green fluorescent protein is a suitable reporter of tumor hypoxia despite an oxygen requirement for chromophore formation

The oxygen requirement for chromophore formation potentially limits the use of green fluorescent protein as a reporter under hypoxic conditions. In the light of this, the applicability of a hypoxia-responsive enhanced green fluorescent protein (EGFP)-based system to the measurement of tumor hypoxia was tested in human HT 1080 fibrosarcoma cells stably transfected with a destabilized EGFP vector containing the hypoxia-responsive 5HRE-hCMVmp promoter or, as a positive control, the strong constitutive CMV promoter. After various schedules of hypoxia and reoxygenation, EGFP fluorescence of live cells was assessed by flow cytometry, and protein levels were analyzed by Western blot. Fluorescence of CMV promoter positive control cells dropped to 38+/-5% of aerobic levels after 12 hours at <0.02% oxygen, but was unaffected by higher oxygen concentrations. Following 12 hours at <0.02% oxygen, cells transfected with the hypoxia-responsive vector exhibited maximum fluorescence after 4 hours of subsequent reoxygenation, reaching 68+/-2% of the levels in CMV promoter controls under aerobic conditions. With such reoxygenation, these cells exhibited a constant increase in fluorescence between 2% and <0.02% oxygen. EGFP chromophore formation is only affected by near-anoxic oxygen concentrations. The correlation of fluorescence and oxygen concentration is restored by a 4-hour reoxygenation period due to oxidation of pre-synthesized EGFP and a delayed increase in EGFP protein synthesis.

Hypoxia-inducible factor-1alpha is an intrinsic marker for hypoxia in cervical cancer xenografts

The hypoxia-inducible factor 1 (HIF-1) is known to induce the expression of several proteins linked to the maintenance of oxygen homeostasis, cellular energy metabolism, and tumor progression. Its alpha subunit (HIF-1alpha) is stabilized under hypoxic conditions and, therefore, might represent an intrinsic marker for tissue hypoxia. Here we report on the spatial relationship between HIF-1alpha and the nitroimidazole hypoxia marker EF5 in cervical carcinoma xenografts, and on their spatial relationship to tumor blood vessels. EF5 was administered to mice bearing ME180 and SiHa cervical cancer xenografts. Frozen tumor tissue sections, triple-stained for HIF-1alpha, the endothelial cell marker CD31, and EF5, were imaged using wide-field multiparameter immunofluorescence microscopy. Expression levels of EF5 and HIF-1alpha were similar in ME180 xenografts, but the percentage of tumor area stained with EF5 was significantly smaller than the percentage of HIF-1alpha-positive area in SiHa tumors. In both tumor types the EF5-HIF-1alpha overlap was statistically significant, thus confirming their spatial and temporal colocalization. Spatial distribution analysis of EF5 and HIF-1alpha is consistent with different pO2 value "thresholds" for EF5 binding and HIF-1alpha expression. Summarized, our results indicate that HIF-1alpha is a useful intrinsic marker for hypoxia in cervical carcinoma xenografts.

Hypoxic heterogeneity in human tumors: EF5 binding, vasculature, necrosis, and proliferation

We evaluated the levels and distribution of hypoxia in 31 human tumors using fluorescent immunohistochemical detection of binding by the 2-nitroimidazole, EF5. Hypoxia was found to be a heterogeneous property of human tumors. Necrosis was usually found adjacent to the highest level of binding in an individual patient's tumor. However, hypoxia often occurred without necrosis. In the group of tumors studied, the most common relationship between blood vessels (PECAM/CD31) and EF5 staining was consistent with diffusion-limited hypoxia; acute hypoxia occurred infrequently. Within a given patient's tumor, there was an inverse correlation between regions of proliferation (Ki-67) and regions of hypoxia. Again, however, when these parameters were examined in a group of patients, the absence of proliferation did not predict the presence of hypoxia. The relationships between hypoxia and other biologic endpoints are complex, but, within a given tumor's spatial relationships, they are in accord with known physiologic principles. Thus, our data emphasize that the relationships between hypoxia and other biologic parameters vary between patients. Necrosis, proliferation, and blood vessel distribution cannot predict the level or presence of hypoxia in an individual patient's tumor.

Characterization of the effects of antiangiogenic agents on tumor pathophysiology

A variety of strategies have been proposed to control tumor growth and metastasis by inhibiting tumor angiogenesis. To optimally combine such antiangiogenic approaches with conventional therapy, improved methods are needed to characterize the underlying pathophysiologic changes. The objective of the current work was to demonstrate the utility of a combination of recently developed immunohistochemical and image analysis techniques in quantitating changes in tumor vasculature and hypoxia. Murine MCa-35 mammary carcinomas were frozen after administration of two COX-2 inhibitors: meloxicam and celecoxib (Celebrex). Total blood vessels were visualized using anti-CD31 staining, perfused vessels by intravenous injection of DiOC7, and tumor hypoxia by EF5 uptake. Although both agents produced similar reductions in tumor volume compared with untreated tumors, varied effects on tumor vasculature and hypoxia were noted. Meloxicam reduced total vessel numbers significantly, whereas celecoxib had no effect. Both drugs substantially increased perfused vessel densities. Although mean hypoxic marker uptake was unchanged from matched controls, intratumor EF5 heterogeneities were significantly different between drugs. The results suggest that COX-2 inhibitors can have varying effects on tumor pathophysiology. Successful use of these drugs to enhance radiation response will likely require optimization of drug choice, dose schedule, and direct physiologic monitoring.

Pharmacokinetics of EF5 [2-(2-nitro-1-H-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl) acetamide] in human patients: implications for hypoxia measurements in vivo by 2-nitroimidazoles

OBJECTIVES

Pharmacokinetic studies were performed on the first 28 patients enrolled in a phase I trial to determine the ability of EF5 [2-(2-nitro-1-H-imidazolI-yl)-N-(2,2,3,3,3-pentafluoropropyl) acetamide] to detect hypoxia in human tumors in the absence of patient toxicity.

METHODS

EF5 was made in purified form and formulated for intravenous injection by the National Cancer Institute. After obtaining consent from the patients, EF5 was administered and blood samples were drawn at various times over approximately 48 h. For most patients it was possible to collect total urine at approximately 8-h intervals. EF5 in plasma and urine was analyzed by high-performance liquid chromatography.

RESULTS

EF5's plasma concentration followed a simple exponential decay following infusion. The plasma half-life was 11.7 +/- 2.6 h (+/- SD) and was not affected by drug dose (9 to 28 mg/kg), fractional urine recovery, patient weight or gender. Absolute plasma values suggested even biodistribution of the drug throughout the soft tissue with a volume of distribution equal to 0.56 l/ kg. Despite the relatively high lipid partition coefficient (logP = 0.6), EF5 was excreted primarily (up to 70%) via kidney clearance. No drug metabolites (e.g. retaining the 2-nitroimidazole chromophore) were detected in either plasma or urine. No toxicity was found at drug doses adequate to detect tumor hypoxia.

CONCLUSIONS

Currently held paradigms of 2-nitroimidazole metabolism (e.g. clearance rate and toxicity as affected by octanol/ water partition coefficient) are discussed. The results reported herein suggest that EF5 is biologically stable with predictable pharmacokinetics. EF5's consistent half-life and clearance properties will allow quantitative analysis of EF5 binding relative to tissue oxygen levels.

Intravascular HBO(2) saturations, perfusion and hypoxia in spontaneous and transplanted tumor models

Clinical trials utilizing strategies to manipulate tumor oxygenation, blood flow and angiogenesis are under way, although limited quantitative information exists regarding basic tumor pathophysiology. The current study utilized murine KHT fibrosarcomas, spontaneous mammary carcinomas and first-generation spontaneous transplants to examine heterogeneity in vascular structure and function, to relate these changes to the distribution of tumor hypoxia and to determine whether fundamental relationships among the different pathophysiological parameters exist. Three methods were included: (i) immunohistochemical staining of anatomical and perfused blood vessels, (ii) cryospectrophotometric measurement of intravascular oxyhemoglobin saturations and (iii) fluorescent detection of the EF5 hypoxic marker. While a distinct pattern of decreasing oxygenation with increasing distance from the tumor surface was observed for KHT tumors, striking intertumor variability was found in both spontaneous and first-generation transplants, with a reduced dependence on tumor volume. EF5 hypoxic marker uptake was also much more heterogeneous among individual spontaneous and first-generation tumors compared to KHT. Although mammary carcinomas demonstrated fewer anatomical blood vessels than fibrosarcomas, the proportion of perfused vessels was substantially reduced in KHT tumors, especially at larger tumor volumes. Vascular morphology, tissue histological appearance and pathophysiological parameters differed substantially between KHT tumors and both spontaneous and first-generation tumors. Such differences in vascular structure and function are also likely to correlate with altered response to therapies targeted to the vascular system. Finally, spontaneous differentiation status, tumor morphology, vascular configuration and function were well preserved in first-generation transplanted tumors, suggesting a close relationship between vascular development and function in early-generation transplants and spontaneous tumor models.

Combined prostaglandin and nitric oxide inhibition produces anatomic remodeling and closure of the ductus arteriosus in the premature newborn baboon

After birth, the full-term ductus arteriosus actively constricts and undergoes extensive histologic changes that prevent subsequent reopening. These changes are thought to occur only if a region of intense hypoxia develops within the ductus wall after the initial active constriction. In preterm infants, indomethacin-induced constriction of the ductus is often transient and is followed by reopening. Prostaglandins and nitric oxide both play a role in inhibiting ductus closure in vitro. We hypothesized that combined inhibition of both prostaglandin and nitric oxide production (with indomethacin and N-nitro-L-arginine (L-NA), respectively) may be required to produce the degree of functional closure that is needed to cause intense hypoxia. We used preterm (0.67 gestation) newborn baboons that were mechanically ventilated for 6 d: 6 received indomethacin alone, 7 received indomethacin plus L-NA, and 16 received no treatment (control). Just before necropsy, only 25% of control ductus and 33% of indomethacin-treated ductus were closed on Doppler examination; in contrast, 100% of the indomethacin-plus-L-NA-treated ductus were closed. Control and indomethacin-treated baboons developed negligible-to-mild ductus hypoxia (EF5 technique). Similarly, there was minimal evidence of ductus remodeling. In contrast, indomethacin-plus-L-NA-treated baboons developed intense hypoxia in regions where the ductus was most constricted. The hypoxic muscle strongly expressed vascular endothelial growth factor, and proliferating luminal endothelial cells filled and occluded the lumen. In addition, cells in the most hypoxic regions were undergoing DNA fragmentation. In conclusion, preterm newborns are capable of remodeling their ductus, just like the full-term newborn, if they can reduce their luminal blood flow to a point that produces intense ductus wall hypoxia. Combined prostaglandin and nitric oxide inhibition may be necessary to produce permanent closure of the ductus and prevent reopening in preterm infants.

Differential effects of buthionine sulphoximine in hypoxic and non-hypoxic regions of human cervical carcinoma xenografts

BACKGROUND AND PURPOSE

Recently we reported increased glutathione (GSH) levels in hypoxic regions of ME 180 and SiHa cervical cancer xenografts. Since this association might act synergistically to protect from radiotherapy, we examined the differential effects of the GSH depleting agent buthionine suiphoximine (BSO) in relation to tumor oxygenation.

MATERIALS AND METHODS

The nitroimidazole EF5 was used to label tumor hypoxia. GSH levels were determined in cryostat sections using a sensitive HPLC assay and in parallel sections using fluorescence image analysis. Using a dual-labeling method, GSH levels were determined selectively in hypoxic and non-hypoxic tumor regions.

RESULTS

GSH levels were higher in hypoxic than in non-hypoxic regions of cervical carcinoma xenografts. Treatment with BSO produced a more pronounced GSH depletion in regions of hypoxia, resulting in similar post-treatment levels in hypoxic and non-hypoxic areas.

CONCLUSIONS

BSO effectively depletes GSH in hypoxic microregions of tumors. These findings suggest a potential role for BSO as an adjunct to radiotherapy in cervical cancer patients.

Hypoxia in radiation-induced blood-spinal cord barrier breakdown

The vascular endothelial cell is believed to be a major target cell of radiation-induced injury to the central nervous system. Dysfunction of the blood-brain barrier is associated with radiation-induced white matter lesions. The aim of this study was to determine the role of hypoxia in radiation-induced blood-brain barrier disruption. Adult rats were irradiated with graded single doses of 0-22 Gy to the cervical spinal cord. At various times up to 28 weeks after radiation, blood-spinal cord barrier (BSCB) permeability was assessed using immunohistochemistry with antialbumin antibody and gamma counting of (99m)Tc-diethylenetriamine pentaacetic acid. Expression of vascular endothelial growth factor (VEGF) was assessed using immunohistochemistry and in situ hybridization. Hypoxia was assessed using two 2-nitroimidazole markers, [(125)I]iodoazomycin arabinodise and 2-(2-nitro-1H-imidazol-l-yl)-N-(2,2,3,3,3,-pentafluoropropyl) acetamide (EF5), with binding in the rat spinal cord measured using gamma counting and immunohistochemistry, respectively. In the nonirradiated rat spinal cord, there was no evidence of BSCB disruption or VEGF expression. After 16-22 Gy, there was a dose-dependent increase in albumin staining and (99m)Tc-diethylenetriamine pentaacetic acid activity beginning at 16 weeks, consistent with barrier breakdown. A similar dose-dependent increase in white matter astrocytes that showed immunoreactivity and in situ hybridization signals for VEGF was observed. No increase in VEGF-positive cells was observed in gray matter. By 20 weeks after 20-22 Gy, animals developed white matter necrosis associated with diffuse albumin staining. Irradiated rat spinal cord showed a dose (16-22 Gy)- and time-dependent (16-20 weeks after 22 Gy) increase in [(125)I]iodoazomycin arabinodise accumulation compared to nonirradiated controls. A similar pattern of dose- and time-dependent EF5 immunoreactivity was also observed in white matter. Areas of EF5 expression and VEGF in situ signals colocalized with areas of albumin immunoreactivity. It is concluded that there is a dose-dependent temporal and spatial association of hypoxia, VEGF up-regulation, and radiation-induced BSCB dysfunction. Hypoxia may provide the signal for VEGF up-regulation and perpetuate endothelial permeability damage in the central nervous system after ionizing radiation.

Influence of hydralazine administration on oxygenation in spontaneous and transplanted tumor models

PURPOSE

To examine the effects of hydralazine on vascular perfusion and hypoxia in spontaneous vs. first generation and long-term transplanted murine tumor models.

METHODS AND MATERIALS

Total anatomic blood vessels were quantified using image analysis of CD31 stained frozen sections, perfused vessels by i.v. injection of fluorescent DiOC(7), and tumor hypoxia was measured using the EF5 hypoxia marker. KHT sarcomas, spontaneous mammary carcinomas, and first generation transplants of the spontaneous tumors were evaluated before and after i.p. administration of 5 mg/kg hydralazine.

RESULTS

Although anatomic and perfused vessel spacings were similar among untreated tumors, response to hydralazine varied widely among the three tumor models. In KHT tumors, perfused vessel numbers decreased significantly at 30 min post-hydralazine, then recovered somewhat by 60 min. First-generation transplants showed a less substantial decrease in perfused vessels following hydralazine, which tapered off slightly by 60 min. Finally, spontaneous tumors had only a modest decrease in perfused vessel numbers, with complete recovery at 60 min. Although response of individual tumors varied widely, overall hypoxic marker uptake was significantly increased in both KHT and first generation tumors, and slightly reduced in the spontaneous tumors.

CONCLUSION

Response to hydralazine varies substantially between transplanted and spontaneous tumor models. Results suggest that increased tumor pressure may be a critical factor in tumor response to hydralazine, possibly explaining tumor volume dependent variations.

Hypoxia in human intraperitoneal and extremity sarcomas

PURPOSE

The presence of hypoxia, measured by needle electrodes, has been shown to be associated with poor patient outcome in several human tumor types, including soft tissue sarcomas. The present report emphasizes the evaluation of hypoxia in soft tissue sarcomas based upon the binding of the 2-nitroimidazole drug EF5 (2-[2-nitro-1H-imidazol-1-yl]-N-(2,2,3,3,3-pentafluoropropyl) acetamide). EF5 has previously been shown to be predictive of radiation response in animal tumors and in in vitro studies. We have also previously reported studies of EF5 binding in human squamous cell tumors. Using fluorescent immunohistochemical techniques, we provide data on the presence and distribution of EF5 binding, as a surrogate for hypoxia, in human spindle cell tumors.

METHODS AND MATERIALS

Patients with spindle cell tumors who were scheduled for tumor surgery were asked to participate in the Phase I trial of EF5. Approximately 48 h preoperatively, EF5 was administered i.v. at doses between 9 and 21 mg/kg. Binding in frozen sections of biopsied tissues was determined using monoclonal antibodies labeled with the green-excited, orange-emitting fluorescent dye, Cy3. Calibration studies were performed in vitro by incubating fresh tumor tissue cubes obtained from each patient with EF3 (an analog of EF5) under hypoxic conditions ("reference binding"). The goal of these calibration studies was to quantify the maximal binding levels possible in individual patient's tissues. The relationship between binding (in situ based on EF5 binding) and reference binding (in vitro based on EF3 binding) was determined.

RESULTS

Eight patients were studied; 3 of these patients had gastrointestinal stromal tumors (GIST). The incubation of tumor tissue cubes in EF3 under hypoxic conditions demonstrated that all tumors bound drug to a similar extent. Reference binding showed a 3.2-fold variation in median fluorescence (113-356) on an absolute fluorescence scale, calibrated by a Cy3 dye standard. In situ binding in the brightest tumor section varied by a factor of 25.4 between the lowest and highest binding tumor (7.5-190.2). Heterogeneity of highest binding was greater between tumors than within individual tumors. A correspondence between EF5 binding and Eppendorf needle electrode studies was seen in the 5 patients with non-GISTs.

CONCLUSION

Inter- and intratumoral heterogeneity of EF5 binding in spindle cell tumors has been documented. Patterns of binding consistent with diffusion limited hypoxia are present in human spindle cell neoplasms.

Effects of radiation on tumor intravascular oxygenation, vascular configuration, development of hypoxia, and clonogenic survival

The underlying physiological mechanisms leading to tumor reoxygenation after irradiation have elicited considerable interest, but they remain somewhat unclear. The current study was undertaken to determine the effects of a single dose of 10 Gy gamma radiation on both tumor pathophysiology and radiobiologically hypoxic fraction. Immunohistochemical staining and perfusion markers were used to quantify tumor vasculature, uptake of the hypoxia marker EF5 to assess the distribution of hypoxia, and intravascular HbO(2) measurements to determine oxygen availability. Tumor radiosensitivity was measured by a clonogenic assay. At 24 h postirradiation, oxygen availability increased, perfused vessel numbers decreased, EF5 uptake decreased, and the radiobiologically hypoxic fraction was unchanged. Together, these results demonstrate that tumor hypoxia develops at an increased distance from perfused blood vessels after irradiation, suggesting a decrease in oxygen consumption at 24 h. By 72 h postirradiation, all physiological parameters had returned to the levels in volume-matched, nonirradiated controls. These studies clearly show that single measures of either tumor oxygenation or vascular structure are inadequate for assessing the effects of radiation on tumor clonogenicity. Although such direct measurements have previously proven valuable in predicting tumor response to therapy or oxygen manipulation, a combination of parameters is required to adequately describe the mechanisms underlying these changes after irradiation.

[18F]-EF5, a marker for PET detection of hypoxia: synthesis of precursor and a new fluorination procedure

There is a great deal of clinical and experimental interest in determining tissue hypoxia using non-invasive imaging methods. We have developed EF5, 2-(2-nitro-1[H]-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl)-acetamide, with both invasive and non-invasive hypoxia detection in mind. EF5 and other 2-nitroimidazoles are used to detect hypoxia, because the rate of their bioreductive metabolism is inversely dependent on oxygen partial pressure. Such metabolism leads to the formation of covalent adducts within the metabolizing cells. Previously, we have described the invasive detection of these adducts by highly specific monoclonal antibodies after tissue biopsy. In this report, we demonstrate the synthesis of 18F-labeled EF5, [18F]-2-(2-nitro-1[H]-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl)-acetamide, in greater than 10% yield by direct fluorination of the newly synthesized precursor 2-(2-nitro-1[H]-imidazol-1-yl)-N-(2,3,3-trifluoroallyl)-acetamide by [18F]-F2 in trifluoroacetic acid. Our objective was to optimize the electrophilic fluorination of the fluorinated alkene bond with fluorine gas, a new method of 18F-labeling of polyfluorinated molecules. Previous biodistribution studies in mice have demonstrated uniform access of EF5 to all tissues with bioelimination dominated by renal excretion. When [18F]-EF5 was injected into a rat followed by urine collection and analysis, we found no detectable metabolism to other radioactive compounds. Thus, [18F]-EF5 should be well suited for use as a non-invasive hypoxia marker with detection using positron emission tomography (PET).

Detection of hypoxia in human squamous cell carcinoma by EF5 binding

Localization and quantitation of 2-nitroimidazole drug binding in low pO2 tumors is a technique that can allow the assessment of hypoxia as a predictive assay. EF5 [2-(2-nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl) acetamide] is such a drug, and it has been shown to be predictive of radiation response in rodent tumors. Using fluorescence immunohistochemical techniques, we provide data on the presence, distribution, and levels of EF5 binding as a surrogate for hypoxia in human head and neck and uterine cervix squamous cell cancers (SCCs). Six patients with SCC were studied. Four patients had head and neck tumors, and two had uterine cervix cancers. The incubation of fresh tissue cubes in EF3 under hypoxic conditions ("reference binding") demonstrated that all tumors were capable of binding drug, and that this binding varied by a factor of 2.9-fold (174.5-516.1) on an absolute fluorescence scale. In the five patients treated at the lowest drug doses (9 mg/kg), in situ binding was quantitatable. For all six patients, the maximum rate of in situ binding varied by a factor of 6.7 between the lowest and highest binding tumor (24.8-160.3) on an absolute fluorescence scale. In tumors with high binding regions, intratumoral heterogeneity was large, extending from minimal fluorescence (<1%) up to 88.6% of reference binding. In tumors with minimal binding, there was little intratumoral heterogeneity. These studies demonstrate substantial heterogeneity of in situ binding between and within individual squamous cell tumors.

Hypoxia and necrosis in rat 9L glioma and Morris 7777 hepatoma tumors: comparative measurements using EF5 binding and the Eppendorf needle electrode

PURPOSE

The purpose of this study was to assess the presence of tumor hypoxia using two independent techniques: binding of the 2-nitroimidazole EF5 and Eppendorf needle electrode measurements. The distribution of tumor hypoxia was assessed with respect to tumor necrosis in corresponding histological studies.

METHODS AND MATERIALS

Each of several rats bearing a subcutaneous 9L glioma or Morris 7777 hepatoma tumor was given EF5 i.v. to a final, whole-body concentration of 100 microM. About 2.5 h later, each rat was anesthetized, and needle electrode measurements were made in the tumor along 1-5 tracks (30-200 individual measurements). At 3 h post-EF5 injection, the tumor was excised and frozen. Frozen sections were analyzed for the presence and distribution of binding of EF5 and necrosis using immunohistochemical techniques followed by staining with hematoxylin and eosin (H&E). The histochemical analysis and electrode readings in similar regions of the tumor were compared.

RESULTS

Electrode measurements were taken at 0.4-mm intervals along one-dimensional tracks, whereas EF5 binding measurements from tissue sections contained two-dimensional information at high spatial resolution ( approximately 2.5 micro). The EF5 measurements showed greater spatial heterogeneity than did the electrode measurements. In tumor regions with minimal necrosis, needle tracks with relatively high pO(2) readings were usually found to contain relatively low EF5 binding, and vice versa. Because EF5 binding is inversely related to tissue pO(2), this result was expected. The expected inverse correlation of the two techniques was most disparate in necrotic tumor regions (confirmed by H&E staining), where needle electrode measurements showed low to zero pO(2) values, but little or no EF5 binding was found.

CONCLUSION

The two methods compared in this study operate in fundamentally different ways and provide substantially different information. EF5 binding provided detailed spatial information on the distribution of hypoxia in viable tumor tissue. There was no EF5 binding in necrotic tumor tissue because cells in such tissue were unable to metabolize the drug. In contrast, output from the needle electrode method appeared to represent a "track-average" tissue pO(2) and did not distinguish between extreme hypoxia and either macroscopic or microscopic necrosis. At the present time, the importance of tumor necrosis in determining treatment response is unknown. However, our data suggest that the Eppendorf needle electrode technique will overestimate the presence of hypoxia. Both techniques are potentially limited by sampling errors in tumors with heterogeneous distributions of hypoxia.

Enhancement of tumor perfusion and oxygenation by carbogen and nicotinamide during single- and multifraction irradiation

Numerous experimental and clinical studies have been completed regarding the effects of carbogen and nicotinamide on tumor oxygenation and radiosensitivity. The current study incorporates three physiological measurement techniques to further define spatial variations in oxygen availability and development of hypoxia after single- and multifraction irradiation in KHT murine fibrosarcomas. Distances to anatomical and perfused blood vessels were measured using immunohistochemical and fluorescent staining, intravascular oxygen levels were determined cryospectrophotometrically, and tumor hypoxia was quantified using uptake of EF5, a marker of hypoxia. Carbogen, nicotinamide, and the combination of both all increased intravascular oxygen availability compared to controls. While nicotinamide had no effect on the number of perfused blood vessels in nonirradiated tumors, carbogen produced a substantial closing of vessels. After a single dose of 4 Gy, only the combination of nicotinamide and carbogen produced significant improvements in oxygen availability, while numbers of perfused vessels were significantly increased for nicotinamide, unchanged for the combination of nicotinamide and carbogen, and significantly decreased for carbogen. After 4 x 4-Gy fractions, oxygen availability was increased substantially with the combination of nicotinamide and carbogen, somewhat with carbogen, and not at all with nicotinamide. Tumor oxygenation changes were estimated by EF5/Cy3 intensity distributions, which demonstrated that manipulative agents could produce disparate effects on tumor hypoxia when combined with either single- or multifraction irradiation.

Decreased brain infarct following focal ischemia in mice lacking the transcription factor E2F1

E2F1+/- mice subjected to 2 h middle cerebral artery occlusion developed an infarct of 77.0 +/- 3.2 mm3 (mean +/- s.e.m., n = 15) in the ischemic hemisphere after 24 h reperfusion. A significantly smaller infarct of 58.8 +/- 4.8 mm3 (n = 15; p < 0.01) was found in E2F1-/- animals. Both deficient and normal mice had similar cerebral angioarchitecture and intra-ischemic decreases in regional blood flow. Similar areas of hypoxia in both groups of ischemic animals were demonstrated directly by immunohistochemical detection of nitroimidazole adducts. It was concluded that all animals received the same ischemic insult, yet the subsequent damage was different in the mutant mice. This is the first indication that the E2F1 gene plays a role in ischemic death of post-mitotic neurons.

Detection of hypoxic cells with the 2-nitroimidazole, EF5, correlates with early redox changes in rat brain after perinatal hypoxia-ischemia

The hypoxia-dependent activation of nitroheterocyclic drugs by cellular nitroreductases leads to the formation of intracellular adducts between the drugs and cellular macromolecules. Because this covalent binding is maximal in the absence of oxygen, detection of bound adducts provides an assay for estimating the degree of cellular hypoxia in vivo. Using a pentafluorintated derivative of etanidazole called EF5, we studied the distribution of EF5 adducts in seven-day-old rats subjected to different treatments which decrease the level of oxygen in the brain. EF5 solution was administered intraperitoneally 30 min prior to each treatment. The effect of acute and chronic hypoxia on EF5 adduct formation (binding) was studied in the brain of newborn rats exposed to global hypoxia (8% O2 for 30, 90 or 150 min) and in the brain of chronically hypoxic rat pups with congenital cardiac defects (Wistar Kyoto). The effect of combined hypoxia-ischemia was investigated in rat pups subjected to right carotid coagulation and concurrent exposure to 8% O2 for 30, 90 or 150 min. Brains were frozen immediately at the end of each treatment. Using a Cy3-conjugated monoclonal mouse antibody (ELK3-51) raised against EF5 adducts, hypoxic cells within brain regions were visualized by fluorescence immunocytochemistry. Brains from controls or vehicle-injected animals showed no EF5 binding. Notably, brains from animals which were chronically hypoxemic as a result of congenital cardiac defects also showed no EF5 binding. A short exposure (30 min) to hypoxia or to combined hypoxia-ischemia resulted in increased background stain and few scattered cells with low-intensity immunostaining. Acute hypoxia exposure of at least 90-150 min, which in this age animal does not result in frank cellular damage, produced patchy areas of low- to moderate-intensity fluorescence scattered throughout the brain. In contrast, 90-150 min of hypoxia-ischemia was associated with intense immunofluorescence in the hemisphere ipsilateral to the carotid occlusion, with a pattern similar to that reported previously for the histological damage seen in this model. This study provides a sensitive method for the evaluation of the level of oxygen depletion in brain tissue after neonatal hypoxia-ischemia at times much earlier than any method demonstrates apoptotic or necrotic cell death Since the level of in vivo formation of macromolecular adducts of EF5 depends on the degree of oxygen depletion in a tissue, intracellular EF5 binding may serve as a useful marker of regional cellular vulnerability and redox state after brain injury resulting from hypoxia-ischemia.

Characterization of [18F]fluoroetanidazole, a new radiopharmaceutical for detecting tumor hypoxia

Fluorinated derivatives of etanidazole are being explored as probes for tumor hypoxia. Our research group has synthesized [18F]fluoroetanidazole (FETA) and now reports the oxygen dependency of binding to cells in vitro, the biodistribution of the tracer in tumor-bearing mice and the analysis of metabolites in their plasma and urine.

METHODS

Four cultured rodent cell lines (V79, 36B10, EMT6 and RIF1) were incubated with [18F]FETA for various times under graded O2 concentrations. We also compared the biodistributions of [18F]FETA and [18F]fluoromisonidazole (FMISO) at 2 and 4 h postinjection in C3H mice bearing KHTn tumors (130-430 mg). Reverse-phase high-performance liquid chromatography was used to distinguish metabolites from parent drugs in urine and plasma of mice injected with [18F]FETA or [18F]FMISO.

RESULTS

In cells labeled in vitro, O2 levels of 600-1300 ppm inhibited binding by 50% relative to uptake under anoxic conditions (<10 ppm). These inhibitory values are not statistically different from those reported for [18F]FMISO in the same cell lines (700-1500 ppm). In the biodistribution studies, uptake in heart, intestine, kidney and tumor was similar for both tracers 4 h after injection, whereas retention of [18F]FETA in liver and lung was significantly lower. Less uptake of [18F]FETA in liver suggests that this nitroimidazole is metabolized less than [18F]FMISO. The brain-to-blood ratios indicate that [18F]FETA readily crosses the blood-brain barrier. High-performance liquid chromatography of urine demonstrated that 10% of [18F]FETA-derived activity was in metabolites at 2 h postinjection, with 15% in metabolites by 4 h; comparable values for [18F]FMISO were 36% and 57%, respectively.

CONCLUSION

We conclude from these data that [18F]FETA holds promise as a new hypoxia tracer in patients, having oxygen dependency of binding similar to [18F]FMISO in vitro and displaying less retention in liver and fewer metabolites in vivo.

VEGF deprivation-induced apoptosis is a component of programmed capillary regression

The pupillary membrane (PM) is a transient ocular capillary network, which can serve as a model system in which to study the mechanism of capillary regression. Previous work has shown that there is a tight correlation between the cessation of blood flow in a capillary segment and the appearance of apoptotic capillary cells throughout the segment. This pattern of cell death is referred to as synchronous apoptosis (Lang, R. A., Lustig, M., Francois, F., Sellinger, M. and Plesken, H. (1994) Development 120, 3395–3404; Meeson, A., Palmer, M., Calfon, M. and Lang, R. A. (1996) Development 122, 3929–3938). In the present study, we have investigated whether the cause of synchronous apoptosis might be a segmental deficiency of either oxygen or a survival factor. Labeling with the compound EF5 in a normal PM indicated no segmental hypoxia; this argued that oxygen deprivation was unlikely to be the cause of synchronous apoptosis. When rat plasma was used as a source of survival factors in an in vitro PM explant assay, inhibition of vascular endothelial growth factor (VEGF) all but eliminated the activity of plasma in suppressing apoptosis. This argued that VEGF was an important plasma survival factor. Furthermore, inhibition of VEGF in vivo using fusion proteins of the human Flk-1/KDR receptor resulted in a significantly increased number of capillaries showing synchronous apoptosis. This provides evidence that VEGF is necessary for endothelial cell survival in this system and in addition, that VEGF deprivation mediated by flow cessation is a component of synchronous apoptosis.

Co-localization of hypoxia and apoptosis in irradiated and untreated HCT116 human colon carcinoma xenografts

HCT116 human colon carcinoma xenografts were grown in nude mice. Frozen sections of control and irradiated tumors were stained and analysed for the distribution and extent of hypoxia and apoptosis. Tissue oxygen partial pressure was measured by immunohistochemical staining of hypoxia-dependent metabolites of the 2-nitroimidazole EF5. Apoptosis was assessed using a commercial kit which stains damaged DNA. Although the apoptosis stain was unlikely to exclude other forms of cell death (necrosis, pyknosis) all staining was found to associate with regions of near anoxia.