Objective: The purpose of this study was to determine the factors that are related to the performance of lymph node assessment and its impact on prognosis in ovarian cancer.
Study design: This was a retrospective analysis of stage IIIC/IV epithelial ovarian cancer in patients who had undergone primary surgery between 1994 and 1998. Simple statistics and univariate and multivariable analysis were performed.
Results: Two hundred nineteen patients met the inclusion criteria; lymph node assessment was performed for 93 of these patients (41%). Sixty-one patients (65.5%) underwent complete pelvic and para-aortic lymphadenectomy, and 32 patients (34.5%) underwent a more limited lymph node sampling. In patients with residual disease >1 cm, lymph node assessment was an independent predictor of outcome. In this same subgroup, lymphadenectomy appeared to be superior to lymph node sampling (5-year overall survival, 50% (lymphadenectomy) vs 33% (lymph node sampling) vs 29% (no lymph node assessment); P = .01). Considering survival of the subgroup who underwent lymph node assessment, we observed a significantly worse outcome for those with lymphatic involvement (5-year overall survival, 31.5% [positive for nodal metastases] vs 54% [negative for nodal metastases]; P = .003). Although multiple factors were correlated with the decision to perform lymph node assessment in univariate analysis, only the surgeon (P < .001), low residual disease (P = .004), American Society of Anesthesiology 1 or 2 (P = .004), and the absence of carcinomatosis (P = .0002) were independent factors in the multivariable analysis. Further, if lymph node assessment was performed, the decision to do lymphadenectomy versus lymph node sampling was associated independently with the surgeon (P < .001), low residual disease (P < .001), and patient age of <65 years (P < .001).
Conclusion: Removal of obviously involved lymph nodes in patients with residual disease near 1 cm and lymphadenectomy for patients with complete or near complete resection of abdominal disease appears to be justified. A lack of standard recommendation in advanced ovarian cancer results in wide variations that are based on individual preference in addition to logical factors.