Limits of lymphoscintigraphy for sentinel node biopsy in women with endometrial cancer.

Objective: Lymph node status in endometrial cancer is a major prognostic factor. Sentinel lymph node (SLN) biopsy using radiocolloid and blue dye labeling has emerged as an alternative to systematic lymphadenectomy. This technique requires a preoperative lymphoscintigraphy. The aim of this study was to evaluate the limits of day-before preoperative lymphoscintigraphy to SLN biopsy.

Methods: Between July 2002 and March 2007, 38 patients with endometrial cancer underwent laparoscopic SLN procedure using radiocolloid and blue dye. Those with early-stage I endometrial cancer (35 patients) underwent a SLN procedure followed by systematic pelvic lymphadenectomy and a hysterectomy with bilateral salpingo-oophorectomy while those with presumed stage IIB on MR imaging (3 patients) underwent a radical hysterectomy. Omentectomy and paraaortic lymphadenectomy were also performed for women with clear cell or serous papillary carcinoma (5 patients). The SLN identification rates and false-negative rates were studied.

Results: The detection rate of lymphoscintigraphy was 84.5% (32/38), with 1.9 nodes per patient. Eight of 17 patients (47%) with unilateral sentinel lymph node on lymphoscintigraphy had bilateral SLNs at surgery and three of 15 patients (20%) with bilateral SLN on lymphoscintigraphy had unilateral SLN at surgery. The correlation was poor (kappa=0.266). When categorized in <2 and > or =2 sentinel nodes, the correlation between lymphoscintigraphic and surgical SLN mapping was moderate (kappa=0.33).

Conclusion: Our results demonstrated the low correlation between day-before lymphoscintigraphy and surgical SLN mapping raising issues of its usefulness and cost-effectiveness in routine practice.