Utero-ovarian anastomosis: histopathologic correlation after uterine artery embolization with or without ovarian artery embolization.

PURPOSE

To study utero-ovarian anastomosis at angiography and its histologic effect on patients who were treated with uterine artery embolization (UAE) with or without ovarian artery embolization (OAE) for symptomatic uterine leiomyomata.

MATERIALS AND METHODS

Four hundred patients (mean age, 43.6 years +/- 6.34) underwent UAE at the authors' institution from June 1998 to May 2005. Eight of the 400 patients underwent hysterectomy with removal of at least one adnexae after UAE. Five patients received tris-acryl gelatin microspheres and three received polyvinyl alcohol particles. Two patients also underwent OAE with gelatin sponges. Specimens from 16 fallopian tubes and 12 ovaries were reviewed. Histologic slides were prepared and reviewed by two pathologists who were blinded to the angiographic findings. The presence of utero-ovarian anastomoses at angiography, the histologic features of adnexa, the presence of particles in the adnexa, and the size and location of the particles were studied.

RESULTS

Utero-ovarian anastomosis was present at angiography in three of the eight patients (38%) and five of the 16 adnexa (31%). Particles were present within the fallopian tube or ovary in all patients who demonstrated utero-ovarian anastomoses at angiography. When utero-ovarian anastomoses were identified bilaterally, particles were found in both adnexae. In cases with particles in the adnexa, the adnexal tissues were histologically viable without evidence of ischemic changes or infarction. Particles were not present in the ovary of patients without utero-ovarian anastomosis at angiography.

CONCLUSION

The angiographic finding of a utero-ovarian anastomosis during UAE appears to correlate with particle embolization in the fallopian tube or ovary. Histologically normal fallopian tubes and ovaries can be expected after UAE with microsphere particles with and without OAE with gelatin sponges.