Measurement of hypoxia using the comet assay correlates with preirradiation microelectrode pO2 histography in R3327-AT rodent tumors

Polarographic determination of tumor oxygenation by Eppendorf histography is currently under investigation as a possible predictor of radiotherapy outcome. Alternatively, the alkaline comet assay has been proposed as a radiobiological approach for the detection of hypoxia in clinical tumor samples. Direct comparisons of these methods are scarce. One earlier study with different murine tumors could not establish a correlation, whereas a weak correlation was reported for a variety of human tumors. Considering the different end points and spatial resolution of the two methods, a direct comparison for a single tumor entity appeared desirable. Anaplastic R3327-AT Dunning prostate tumors were grown on Copenhagen rats to volumes of 1-6 cm(3). Eppendorf histography (100-200 readings in 5 parallel tracks) for 8 different tumors revealed various degrees of oxygenation, with median pO(2) values ranging from 1.1 to 23 mmHg. Within 5 min after an acute exposure to 8 Gy (60)Co gamma rays, tumors were excised from killed animals and rapidly cooled to limit repair, and a single cell suspension was prepared for use with the comet assay. The resulting comet moment distributions did not exhibit two subpopulations (one hypoxic and the other aerobic), and a hypoxic fraction could not be calculated. Instead, the average comet moment distribution was taken as a parameter of overall strand break induction. Corresponding experiments with tumor cells grown in vitro allowed us to derive the relationship between the oxygen enhancement ratio (OER) for the average comet moment and oxygen partial pressure (Howard-Flanders and Alper formula). The validity of this relationship was inferred for cells exposed in situ, and the convolution of a pO(2) distribution with the formula of Howard-Flanders and Alper yielded an array of expected OER values for each tumor. The average expected OER correlated well with the average comet moment (r = 0.89, P < 0.01), and the in situ comet moment distributions could be predicted from the Eppendorf data when 50% repair was taken into account, assuming a 5-min damage half-life. The findings confirm the potential of interstitial polarography to reflect radiobiologically relevant intracellular oxygenation, but also underscore the confounding influence of differences in repair that may occur when cells are prepared from irradiated tissues for use with the comet assay.