LYMPHATIC SPREAD IN STAGE-IB AND STAGE-II CERVICAL-CARCINOMA - ANATOMY AND SURGICAL IMPLICATIONS

Objective: To determine the frequency and topography of pelvic and par a-aortic node involvement in cervical carcinoma and to identify the ap propriate level far resection of the lymphatic chains. Methods: Betwee n 1985 and 1994, 421 women with stage Ib or II cervical carcinoma were treated by surgery in combination with irradiation. Each underwent a radical hysterectomy with systematic pelvic and para-aortic lymphadene ctomy. Results: A median of 34 lymph nodes were removed per patient. T he overall frequency of lymph node involvement was 26%, and the freque ncy of para-aortic metastases was 8%. The frequency of lymph node meta stasis was associated significantly with stage (chi(2) = 7.8; P < .02) , tumor size (chi(2) = 14.8; P < .001), and patient age (chi(2) = 5.9; P < .05). The frequency of para-aortic involvement was below 3% in pa tients with small tumors (under 2 cm). When pelvic nodes were involved , the obturator group was concerned in 76 cases (18%) and the external iliac group in 48 patients (11%). When Fara-aortic nodes were involve d, the left para-aortic chain was the most frequently concerned (23 pa tients [5%]). In eight of these patients, nodal involvement was found only above the level of the inferior mesenteric artery. Among 106 pati ents with pelvic positive nodes, 28 (26%) also had para-aortic metasta tic nodes. Conclusion: Para-aortic lymphadenectomy should remove all o f the left para-aortic chain (inframesenteric and supramesenteric) and so should be performed up to the level of the left renal vein. Accord ing to the low frequency of para-aortic involvement when tumor size is below 2 cm, such a procedure could be avoided in patients with small tumors. (C) 1998 by The American College of Obstetricians and Gynecolo gists.