Hypoxia in human intraperitoneal and extremity sarcomas.
Evans SM., Hahn SM., Magarelli DP., Zhang PJ., Jenkins WT., Fraker DL., Hsi RA., McKenna WG., Koch CJ.
PURPOSE: The presence of hypoxia, measured by needle electrodes, has been shown to be associated with poor patient outcome in several human tumor types, including soft tissue sarcomas. The present report emphasizes the evaluation of hypoxia in soft tissue sarcomas based upon the binding of the 2-nitroimidazole drug EF5 (2-[2-nitro-1H-imidazol-1-yl]-N-(2,2,3,3,3-pentafluoropropyl) acetamide). EF5 has previously been shown to be predictive of radiation response in animal tumors and in in vitro studies. We have also previously reported studies of EF5 binding in human squamous cell tumors. Using fluorescent immunohistochemical techniques, we provide data on the presence and distribution of EF5 binding, as a surrogate for hypoxia, in human spindle cell tumors. METHODS AND MATERIALS: Patients with spindle cell tumors who were scheduled for tumor surgery were asked to participate in the Phase I trial of EF5. Approximately 48 h preoperatively, EF5 was administered i.v. at doses between 9 and 21 mg/kg. Binding in frozen sections of biopsied tissues was determined using monoclonal antibodies labeled with the green-excited, orange-emitting fluorescent dye, Cy3. Calibration studies were performed in vitro by incubating fresh tumor tissue cubes obtained from each patient with EF3 (an analog of EF5) under hypoxic conditions ("reference binding"). The goal of these calibration studies was to quantify the maximal binding levels possible in individual patient's tissues. The relationship between binding (in situ based on EF5 binding) and reference binding (in vitro based on EF3 binding) was determined. RESULTS: Eight patients were studied; 3 of these patients had gastrointestinal stromal tumors (GIST). The incubation of tumor tissue cubes in EF3 under hypoxic conditions demonstrated that all tumors bound drug to a similar extent. Reference binding showed a 3.2-fold variation in median fluorescence (113-356) on an absolute fluorescence scale, calibrated by a Cy3 dye standard. In situ binding in the brightest tumor section varied by a factor of 25.4 between the lowest and highest binding tumor (7.5-190.2). Heterogeneity of highest binding was greater between tumors than within individual tumors. A correspondence between EF5 binding and Eppendorf needle electrode studies was seen in the 5 patients with non-GISTs. CONCLUSION: Inter- and intratumoral heterogeneity of EF5 binding in spindle cell tumors has been documented. Patterns of binding consistent with diffusion limited hypoxia are present in human spindle cell neoplasms.