Gynecologic radiotherapy fields defined by intraoperative measurements.

Whole-pelvis radiation therapy has been traditionally delivered through fields, the dimensions of which have been determined by convention and based largely on textbook anatomic landmarks. Since July 1986, 100 patients have had intraoperative retroperitoneal measurements carried out at the time of radical surgery in an effort to examine this anatomic basis for field dimensions. Structural measurements of the pelvic and paraaortic arterial branches were made in reference to the lumbosacral prominence to correlate with lymphatic pathways. The mean level of the aortic bifurcation was found to be 6.7 cm above the lumbosacral prominence. The mean level of the bifurcation of the common iliac artery was 1.7 cm above this reference point on the right and 1.4 cm above on the left. Both common iliac bifurcations were cephalad to the level of the lumbosacral prominence in 87% of patients. In only three patients were both bifurcations located below this level. Transverse pelvic dimension measurements demonstrated a width of 12.3 cm at the level of the obturator fossa and of 13.0 cm at the most inferior outside width of the external iliac arteries. To establish a simple external reference which could assist in defining radiotherapy field widths, the maximal separation of the femoral arteries at the level of inguinal ligaments was measured and averaged 14.6 cm. These data suggest that conventional fields frequently fail to correspond to true anatomic landmarks and that to optimally cover the lymphatics in radiotherapy, fields should, ideally, be based on intraoperative measurements. If such surgical guides are not available, we would suggest that standard whole-pelvis radiotherapy for cervical cancers should employ anterior and posterior fields with widths of at least 16 cm which will fully include the bifemoral separation. A superior border at the L4-L5 interspace is required to cover lymphatic pathways to the mid-common iliac nodal level. It may also be convincingly pointed out that the attachments of the uterosacral and cardinal ligaments are clearly posterior to the rectosigmoid, mandating lateral fields that should encompass the entire anterior sacral silhouette.