A Retrospective Study of 344 Patients with FIGO Stage IIB Cervical Carcinoma after Neoadjuvant Chemoradiation Therapy in Cervical Cancer: Pelvic Lymph Node Dissection Should Be Tailored

The aim of this study was to analyze the correlation between the CT/MRI diagnosed lymph node (Ln) involvement and postoperatively pathology Ln involvement, and investigate the relationship between the long-term survival outcome without and with pelvic Ln involvement, in FIGO stage IIB patients received preoperative chemoradiotherapy and radical surgery. Between January 2009 and November 2014, 344 patients with FIGO stage IIB cervical carcinoma were analyzed. All patients underwent pelvic CT with contrast and/or MRI scanning at diagnose. The criterion of lymph involvement in CT/MRI images was more than or equal to 0.8cm in short diameter. The patients were treated with 40-50Gy in 20-25 fractions for pelvic, with concurrent chemotherapy and got radical surgery with an interval 8-71 days (mean 26 days). The median follow-up time was 36 months (range 4-83m), and the median age was 48 years (range 25-72 years). At first, until to October 2015, the pelvic failure developed in ten patients (7 in vaginal residue, 3 in parametrium, no pelvic Ln recurrence). The distant metastasis was found in 26 patients. Thirty six patients died (33 patients died of cervical cancer; 3 died of other disease, one of lung cancer, one of lymphoma and one of acute myocardial infarction). Patients with pelvic Ln involvement in diagnosing CT/MRI images showed lower overall survival OS, distant metastatic free survival (DMFS) and similar pelvic failure free survival (PFFS), compared with those without Ln involvement (the 3-year OS 84.1% versus 94.4%, P = 0.007; the 3-year DMFS 87.8% versus 95.7%, P = 0.013; the 3-year PFFS 96.5% versus. 97.1%, P = 0.960). The second, 29 patients showed positive Ln in postoperative pathology (29/344, 8.4%). The patients with Ln involvement in diagnosing CT/MRI images showed higher incidence of Ln involvement in pathology than patients without Ln involvement on images, (25/150, 16.7% versus 4/194, 2.1%, P < 0.05). The last, there were a total of 51 metastasis nodes in 29 patients. The distribution of involved lymph nodes were as follows, external iliac nodes (16/51, 31.4%), internal iliac nodes (10/51, 19.6%), obturator nodes (10/51, 19.6%), Common iliac nodes (9/51, 17.6%), deep inguinal nodes (6/51, 11.8%). For the patients without Ln involvement in diagnosing CT/MRI images, the incidence of postoperatively pathological Ln involvement was very low (2.1%). It is valuable for discussing to omit the pelvic Ln dissection for this group of patients. For the patients with Ln involvement on diagnosing CT/MRI images, the rate of postoperatively pathological Ln involvement was 16.7%, and topographic distribution of involved Ln was not modified. So that, extensive lymphadenectomy still should be performed for this group of patients.