Increasing levels of hypoxia in prostate carcinoma correlate significantlywith increasing clinical stage and patient age - An Eppendorf pO(2) study

BACKGROUND. The purpose of this study was to analyze the extent of hypoxia in prostate carcinoma tumors using the Eppendorf pO(2) microelectrode and c orrelate this with pretreatment characteristics and prognostic factors. METHODS, Custom-made Eppendorf pO(2) microelectrodes were used to obtain pO (2) measurements from the pathologically involved region of the prostate la s determined by the pretreatment sextant biopsies) as well as from a region of normal muscle for comparison. Each set of measurements comprised approx imately 100 separate readings of pO(2), for a total of 10,804 individual me asurements. Fifty-five patients with localized prostate carcinoma were stud ied: Forty-one patients received brachytherapy implants, and 14 patients un derwent radical prostatectomy. The pO(2) measurements were obtained in the operating room by using a sterile technique under spinal anesthesia for the brachytherapy group and under general anesthesia for the surgery group. Th e Eppendorf histograms were recorded and described by the median pO(2), mea n pO(2), and percentage < 5 mm Hg and < 10 mm Hg. A multivariate mixed-effe cts analysis for the prediction of tumor oxygenation was performed and incl uded the following covariates: type of tissue (prostate vs. muscle), type o f treatment (implant vs. surgery) and/or anesthesia (spinal vs. general), p rostate specific antigen level, disease stage, patient age and race, tumor grade, tumor volume, perineural invasion, and hormonal therapy. RESULTS. Due to differences in patient characteristics and the anesthesia e mployed, control measurements were obtained from normal muscle tin all but two patients). This internal comparison showed that the oxygen measurements from the pathologically involved portion of the prostate were significantl y lower (average median pO(2), 9.9 mm Hg) compared with the measurements no rmal muscle (average median pO(2), 28.6 mm Hg; P < 0.0001). A multivariate, linear, mixed analysis demonstrated that, among all of the patients, the s ignificant predictors of oxygenation were tissue (prostate Es. muscle) and anesthesia (spinal ts. general) or treatment (implant vs. surgery). Among t he brachytherapy (spinal anesthesia) patients, the significant predictors o f pO(2), were tissue type, disease stage, and patient age. There were no si gnificant predictors of oxygenation in the surgical [general anesthesia] gr oup. CONCLUSIONS. This study, employing in vivo electrode oxygen measurements, d emonstrated that hypoxia exists in prostate carcinoma tumors. A dramatic ef fect of anesthesia was observed, likely due to modulation of polarography i n the presence of fluorine. Within the group of brachytherapy (spinal anest hesia) pa patients, increasing levels of hypoxia (within prostatic tissue) correlated significantly with increasing clinical stage and patient age, Mo re patients will be accrued to this prospective study to further correlate the oxygenation status in prostate carcinoma tumors with known prognostic f actors and, ultimately, treatment outcome. (C) 2000 American Cancer Society .