Richard F. Schneider

Measuring hypoxia and predicting tumor radioresistance with nuclear medicine assays

Tumor cells at low oxygen tension are relatively radioresistant. The hypoxic fraction of individual tumors before, during and after radiotherapy is likely to have prognostic value but its diagnosis still awaits an accurate and acceptable assay. The recent indications that hypoxia can also induce the expression of specific genes and promote a more aggressive tumor phenotype makes its diagnosis even more important. Over 15 years ago, misonidazole, an azomycin-based hypoxic cell radiosensitizer, was found to link covalently to cellular molecules at rates inversely proportional to intracellular oxygen concentration. The use of bioreducible markers to positively label zones of viable hypoxic cells within solid tumors and to predict for tumor radioresistance was proposed. Several hypoxic markers have now been identified and their selective binding within tumors has been measured by both invasive and non-invasive assays. Research from our laboratory has emphasized both mechanistic and preclinical studies associated with nuclear medicine procedures for measuring tumor hypoxia and predicting tumor radioresistance. This report updates radiation oncologists about the status of nuclear medicine hypoxic marker research and development as of mid-1997. While several potential imaging agents have been identified, their testing and validation in appropriate human tumors will require focused research efforts by individual academic departments and, possibly, by clinical trials performed through cooperative groups. Since the prediction of hypoxia in individual tumors could strongly impact radiotherapy treatment planning, the radiation oncology research community is best positioned to execute the validation studies associated with these markers.

The synthesis and radiolabelling of novel markers of tissue hypoxia of the iodinated azomycin nucleoside class.

Seven second-generation hypoxic markers of the iodinated azomycin nucleoside class have been synthesized and tested for hypoxia marking activity with tumor cells in vitro and in vivo. β-D-lodoazomycin galactoside (IAZG) and β-D-iodoazomycin xylopyranoside (IAZXP) demonstrated superior hypoxia marking properties relative to IAZA because of their higher water solubilities, rapid plasma clearance rates from tumor-bearing mice and maximum tumor/blood (T/B) and tumor/muscle (T/M) ratios. Our studies with animal tumor models show that T/B or T/M ratios of these markers determined by scintigraphy or planar imaging can predict for the relative degree of tumor hypoxia and for tumor radioresistance.

PRECLINICAL ASSESSMENT OF HYPOXIC MARKER SPECIFICITY AND SENSITIVITY

PURPOSE In the search for a sensitive, accurate, and noninvasive technique for quantifying human tumor hypoxia, our laboratory has synthesized several potential radiodiagnostic agents. The purpose of this study was to assess and compare the hypoxic marking properties of both radioiodinated and Tc-99m labeled markers in appropriate test systems which can predict for in vivo activity. MATERIALS AND METHODS Preclinical assessment of hypoxic marker specificity and sensitivity employed three laboratory assays with tumor cells in vitro and in vivo. Radiolabeled marker uptake and/or binding to whole EMT-6 tumor cells under extremely hypoxic and aerobic conditions was measured and their ratio defined hypoxia-specific factor (HSF). Marker specificity to hypoxic tumor tissue was estimated from its selective avidity to two rodent tumors in vivo, whose radiobiologic hypoxic fractions (HF) had been measured. The ratios of % injected dose/gram (%ID/g) of marker at various times in EMT-6 tumor tissue relative to that in the blood and muscle of scid mice were used to quantify hypoxia-specific activity. This tumor in this host exhibited an average radiobiologic HF of approximately 35%. As well, nuclear medicine images were acquired from R3327-AT (HF approximately =15%) and R3327-H (no measurable HF) prostate carcinomas growing in rats to distinguish between marker avidity due to hypoxia versus perfusion. RESULTS The HSF for FC-103 and other iodinated markers were higher (5-40) than those for FC-306 and other Tc-99m labeled markers. The latter did not show hypoxia-specific uptake into cells in vitro. Qualitative differences were observed in the biodistribution and clearance kinetics of the iodinated azomycin nucleosides relative to the technetium chelates. The largest tumor/blood (T/B) and tumor/muscle (T/M) ratios were observed for compounds of the azomycin nucleoside class in EMT-6 tumor-bearing scid mice. These markers also showed a 3-4 x higher uptake into R3327-AT tumors relative to the well-perfused R3327-H tumors. While both FC-306 and CERETEC rapidly distributed at unique concentrations to different tissues, their avidity to EMT-6 and R3327-AT tumors did not correlate with tumor HF. CONCLUSIONS The halogenated azomycin nucleosides with the lowest lipid/water partition coefficient values were found to yield the optimal hypoxia-specific signal in these animal tumors. Our Tc-99m-labeled azomycin chelates showed little or no hypoxia-specific uptake and had in vivo biodistribution and clearance kinetics similar to those of CERETEC, a perfusion agent with no known hypoxic binding activity.

Structure Distribution Relationship in the Design of Tc-99m Radiopharmaceuticals

The origin of Quantitative Structure Activity Relationship (QSAR) in medicinal chemistry and drug design can be traced back to the year 1870 when Crum-Brown and Fraser (1) advanced the idea that for any drug, Biological Response (activity), BR= f (S) where f (S) is a function of chemical structure. They suggested that it should be possible to develop a mathematical formulation of QSAR by making changes in the chemical structure and relating these changes to the biological response or activity of the drugs so designed. The real barrier at that time was to define changes of chemical structure in numerical terms.