Determining the volume of prostatic carcinoma: value of MR imaging with an external-array coil

Abstract

Objective: The size of prostatic carcinomas, as determined by estimations of the tumor volume, has a direct correlation with the histologic grade and stage. Therefore, accurate information about tumor volume can play an important role in determining appropriate treatment and establishing a patient's prognosis. Accordingly, we performed a study to assess the accuracy of MR imaging with external-array coils and the fast spin-echo technique for determining tumor volume in patients with prostatic cancer.

Subjects and methods: In a prospective study, 20 patients with biopsy-proved prostatic cancer had MR imaging before radical prostatectomy. A 1.5-T system, a prototype external-array coil, and fast spin-echo technique were used to obtain axial T2-weighted (4000/120 [TR/TE]) MR images throughout the prostate and seminal vesicles. Contiguous 5-mm sections were obtained with a 256 x 256 or 256 x 192 matrix, 24-cm field of view, four excitations, and fat suppression. The volumes of cancer obtained with MR imaging and surgical specimens were determined independently. The areas of tumor were determined from individual axial sections by summing and multiplying by the section thickness.

Results: Seventeen of the 20 cancers were detected on MR images. A significant correlation between the volume of the tumor as determined with MR imaging and measurements of the specimens was observed (r = .81, p < .001). Four transition zone tumors were detected as zones of rather homogeneous decreased intensity. One important source of error for volumes determined with MR images was a limitation in the histologic specificity of MR imaging for the delineation of cancer; in some cases benign changes or normal prostates were interpreted as cancer. Another limitation was the presence of changes after biopsy that concealed portions of 10 of the 17 tumors visualized.

Conclusion: The accuracy of the MR technique used was not sufficient for reliable prediction of tumor volume. The lack of specificity of T2-weighted MR imaging for cancer detection and the presence of changes caused by biopsy were major sources of error.