Journal of Ultrasound in Medicine, Vol 9, Issue 11 637-644, Copyright © 1990 by American Institute of Ultrasound in Medicine
Transvaginal sonography of postmenopausal ovaries with pathologic correlation
A. C. Fleischer, M. S. McKee, A. N. Gordon, D. L. Page, D. M. Kepple, J. A. Worrell, H. W. Jones 3rd, L. S. Burnett and A. E. James Jr
Department of Radiology, Vanderbilt University Medical Center, Nashville, TN 37232-2675.
The sonographic appearance of 67 ovaries in 34 postmenopausal women who
underwent preoperative transvaginal sonography (TVS) was correlated to
findings on pathologic examination. Both ovaries were detected by TVS in
60% of the women examined; in 85%, at least one ovary was detected. The
size of the normal, sonographically visualized postmenopausal ovary was 2.2
+/- 0.7 cm in transverse, 1.2 +/- 0.3 cm in anteroposterior, and 1.1 +/-
0.6 cm in longitudinal axes, with an average volume of 2.6 +/- 2.0 cm3. The
average size of ovaries that were not detected by TVS was 0.7 x 0.4 cm
(range, 0.3 to 1.3 cm); most of these (five of six) were atrophic on
pathologic exam. The difference between actual and sonographically measured
size was negligible (TVS overestimated by 0.3 cm). Four simple cysts that
ranged from 0.5 to 3.5 cm were found by TVS and confirmed pathologically,
as were three benign serous cystadenomas that ranged from 2.5 to 3.5 cm,
one 3 x 6-cm tubal carcinoma, and one 1 X 4-cm paratubal cyst. TVS missed a
6-cm dermoid, a 2.5-cm cystadenoma, a 0.8-cm Sertoli cell tumor, and a
0.5-cm fibrothecoma that were nonpalpable but that were found on pathologic
examination. None of the missed lesions were palpable preoperatively. The
positive predictive value was 94% for detection of an ovarian mass; the
negative predictive value for exclusion of an ovarian lesion was 92%. It is
concluded that TVS can accurately delineate the ovaries in most, but not
all, postmenopausal women and that only rarely will pathologic lesions not
be detected by TVS.
