Two- and 3-dimensional MRI comparison of levator ani structure, volume, and integrity in women with stress incontinence and prolapse.

OBJECTIVE The aim of this study was to identify imaging markers for genuine stress incontinence and pelvic organ prolapse by using magnetic resonance imaging and reconstructed 3-dimensional models. STUDY DESIGN Thirty women were studied, 10 with prolapse, 10 with genuine stress incontinence, and 10 asymptomatic volunteers. Axial and sagittal T1 and T2 weighted pelvic magnetic resonance scans were obtained with the patient in the supine position. Source images were measured to determine levator hiatus height, bladder neck to pubococcygeal line, levator plate angle, and perineal descent at rest and maximum Valsalva. Manual segmentation and surface modeling was applied to build 3-dimensional models of the organs. The 3-dimensional models were measured to determine levator muscle volume, shape and hiatus width, distance between symphysis and levator sling muscle, posterior urethrovesical angle, bladder neck descent, and levator plate angle. RESULTS The 3 groups of subjects were comparable in age, parity, and body mass index. In the control, genuine stress incontinence, and prolapse groups, the menopausal rate was 40%, 60%, and 55% (P =.7). In the same order, significant mean 2-dimensional measures were: resting bladder neck descent of 24, 17, and 3 mm (P <.005), straining levator plate angle of -4.3, -11.5, and -31 degrees (P =.01), straining levator hiatus height of 48.5, 51.1, and 65.3 mm (P <.005), and straining perineal descent of 17.2, 22.5, 27.2 mm (P =.02). Similarly ordered mean 3-dimensional parameters showed levator volumes of 32.2, 23.3, and 18.4 cm(3) (P <.005); hiatus widths of 25.7, 34.7, and 40.3 mm (P <.005); left levator sling muscle gaps of 15.6, 20.3, and 23.8 mm (P =.03), right levator sling muscle gaps of 15.6, 22.5, and 30.8 mm, (P = 0.003), and levator shape (90%, 40%, and 20% dome shaped; P <.005). CONCLUSION Both 2-dimensional magnetic resonance images and 3-dimensional models yield findings that differ among asymptomatic subjects compared with those with genuine stress incontinence and prolapse. Our 3-dimensional data demonstrate a statistically significant continuum in levator volume, shape, and integrity across groups of asymptomatic, genuine stress incontinence, and prolapse subjects.