Diffusion weighted MR imaging (DWI) and ADC values in endometrial carcinoma.

OBJECTIVE To determine the sensitivity of MR imaging utilizing diffusion weighted imaging (DWI) in detection of endometrial cancer and to compare the Apparent Diffusion Coefficient (ADC) values of malignant and normal endometrium. STUDY DESIGN Cross-sectional analytical study. PLACE AND DURATION OF STUDY Radiology Department, Aga Khan University Hospital Karachi, from January 2007 to December 2009. METHODOLOGY Patients were defined as undergoing MRI for suspected endometrial malignancy, and the controls were female subjects who underwent MRI examination for indications other than endometrial malignancy. Studied variables included the signal characteristics of the endometrium and ADC values. The sensitivity of DWI for the detection of endometrial carcinoma was calculated using histopathology as the gold standard, and the ADC values of the endometrium in patients and controls were recorded. The mean ADC values were compared using two-sided t- test with significance at p < 0.05 at 95% confidence interval. RESULTS Out of 52 patients, 10 had hyperintense, 40 had hypointense and 2 had isointense endometrium on T2 weighted imaging. On DWI, 42 patients had hyperintense and 10 patients had hypointense endometrium. In comparison, 40 controls had hypointense endometrium on DWI and 12 had hyperintense endometrium. The mean ADC value for abnormal endometrium was 0.730±0.215 x10-3 mm2/sec and of normal endometrium was 1.265±0.305 x10-3 mm2/sec (p < 0.001). The sensitivity for detection of endometrial carcinoma on DWI was 77.77%. False positive cases were found to be due to secretory and hyperplastic endometrium. False negative findings were found in a few cases of adenocarcinoma, endometroid carcinoma and clear cell sarcoma. CONCLUSION In patients with suspected endometrial carcinoma, MRI examination utilizing DWI was a sensitive tool in detecting endometrial cancers with significantly lower ADC values for carcinoma in general. Physiological and histopathological variants may be responsible for a few false results.