POLAROGRAPHIC NEEDLE ELECTRODE MEASUREMENTS OF OXYGEN IN RAT PROSTATECARCINOMAS - ACCURACY AND REPRODUCIBILITY

Purpose: The oxygenation status of tumors may be important for predict ing tumor response to therapy. Previous studies with the (a)naplastic (R3327-AT) and well-differentiated (R3327-H) Dunning rat prostate tumo rs using indirect assays of tumor oxygenation indicated the relative h ypoxic and radioresistant nature of the anaplastic tumor. We now repor t direct measurements of oxygen in these tumors made with the pO(2) hi stograph to determine: (a) whether a significant difference in oxygena tion status could be detected between them; (b) whether sequential mea surements on the same tumor gave similar values; and (c) whether tumor oxygenation correlated with tumor volume. Methods and Materials: R332 7-AT and R3327-H tumors were grown in Fischer X Copenhagen rats to vol umes of 1.0-7.0 cm(3). Electrode measurements (100-200) were made in t umors in anesthetized animal along two parallel tracks. Repeat measure ments were made at 1-5 days along different parallel tracks. Oxygen pa rtial pressures of muscle tissue were measured and served as a normal tissue control. Statistical analyses were applied to determine whether tumor oxygen levels were different between the two tumor histologies, whether sequential measurements in the same tumor were reproducible, and whether tumor oxygenation correlated with tumor volume. Results: T he average median pO(2) of the well-differentiated (n = 15) and the an aplastic (n = 15) tumors was 6.0 mmHg (SE +/- 1.3) and 2.2 mmHg (SE +/ - 0.3), respectively. The average median pO(2) of normal rat muscle (n = 15) was 23.6 mmHg (SE +/- 2.0). These values represent highly signi ficant differences in oxygen concentration between the two tumors and rat muscle. The differences in average mean pO(2) values were also hig hly significant. Repeat measurements in the same tumors on different d ays gave average median values of 4.7 and 2.2 mmHg in the R3327-H (n = 15) and R3327-AT (n = 15) tumors, respectively. For these repeat meas urements, median pO(2) values decreased in 15 and increased in 15 tumo rs, and were not significantly different from the first measurements. The average differences observed in median pO(2) were 37% (SE +/- 7) a nd 58% (SE +/- 10) for the R3327-H and R3327-AT tumors, respectively. No significant correlation was observed between pO(2) levels and the t umor volumes investigated in this study. Conclusions: The median pO(2) values of the anaplastic I)Dunning tumors were significantly Lower th an those of the well-differentiated tumors (p < 0.001). Oxygen levels in both tumors were significantly lower than those measured in normal rat muscle (p < 0.00005). Repeat measurements of median pO(2) in the s ame tumors were not significantly different for either tumor model (p > 0.5). The changes observed in pO(2) distributions within individual tumors from day to day may indicate true dynamics of its oxygenation s tatus and/or the limits of electrode measurements, by sampling along o nly two insertion sites. The electrode measurements of pO(2) in these tumor models are reproducible and confirm previously detected oxygenat ion differences between the anaplastic and well-differentiated tumors.