OXYGEN-TENSION DISTRIBUTIONS ARE SUFFICIENT TO EXPLAIN THE LOCAL RESPONSE OF HUMAN BREAST-TUMORS TREATED WITH RADIATION ALONE

Purpose: Several factors are known to influence the probability of tum or control after radiation. These include tumor oxygen tension distrib ution, glutathione content, intrinsic radiation sensitivity, rate of r epopulation, tumor size, physician skill, etc. The relative impact of oxygen on human tumor response is unknown. The purpose of this analysi s is to determine to what extent the observed shape of the radiation r esponse curve for human tumors can be predicted by the tumor oxygenati on status. Methods and Materials: The radiation dose response curve fo r patients treated with radiation alone for breast cancer was calculat ed based on pooled data. Tumor control rates as a function of radiatio n dose were fitted to a probit curve. Twenty-two women with breast can cer in Mainz (Germany) and at Stanford University had pO2 measurements made of their tumors. An average of 87 +/- 58 (range 21 to 300) measu rements were made from each patient. Hypoxia was assumed to be a purel y dose modifying factor with a maximum oxygen enhancement ratio of 2.5 . Assuming patients are treated with daily radiation doses of 2 Gy, th e breast cancer alpha/beta ratio is 10 Gy, tumors have a mean of 10(8) stem cells, and using the linear quadratic formula for modelling surv iving fraction, it was possible to estimate tumor control probability. Results: Tumor oxygenation was an extremely important modifier of the shape of the dose response curve and alone was sufficient to account for the slope of the observed dose response curve for human breast car cinoma. Tumor size distribution had a smaller effect on the shape and the slope of the dose response curve. Two models of radiation induced reoxygenation were tested, one that allowed full reoxygenation to the baseline state between the daily radiation fractions and another with no reoxygenation between fractions. The clinical data fell between the se two models in accordance with the expected incomplete reoxygenation between treatments. Conclusion: The results support the conclusion th at in human breast carcinoma, oxygen tension distribution is a critica l modifier of radiation treatment response.