Impact of sentinel lymph node biopsy on staging of early cervical cancer: Results of a prospective, multicenter study

Abstract

CRA5506 Background: 10% to 15% of patients with pN0 early cervical cancer experience recurrences. This may be related either to nodes missed by the dissection or located outside the dissection field or to failed diagnosis of node metastases. The objective of this study was to measure the benefits from sentinel node (SN) detection in terms of nodes collected from unusual territories and of detected micrometastases and isolated tumor cells (ITCs). Methods: 145 patients who had stage Ia1-Ib1 epidermoid cancer or adenocarcinoma or adenosquamous cancer were included in a multicenter study (January 2005 - June 2007). Noninclusion criteria were age<18 years, pregnancy, and previous treatment. SNs were identified by combined technetium and blue-dye labeling in the pelvic and para-aortic territories. Slices were cut 200 μm apart. At each level, HES staining and labeling with anti-cytokeratin antibodies (AE1-AE3) were performed. SNs in an unusual territory were defined as SNs outside the ilio-obturator region. ITC was defined as size <0.2 mm, micrometastasis as size 0.2 to 2mm, and macrometastases as size >2 mm. The study was funded by the French National Institute of Cancer and reviewed by an IRB. Results: 17 patients were excluded for major protocol deviations, leaving 128 patients for the per protocol analysis. One or more SNs were detected in 98.4% of patients (95%CI, 94.4 to 99.9%). The 430 detected SNs were located as follows: external iliac, 80.5 %; common iliac, 8.6%; presacral and paraaortic, 5.5%; and parametrial, 4.9 %. SN detection identified at least one SN in an unusual territory in 48/128 (37.5%) patients. There were 26 positive SNs in 21(16.4%) patients of whom 8 (38%) had macrometastases, 7 (33%) micrometastases, and 6 (29%) ITCs. Of these 26 nodes, 7 (27%) were detected only by immunohistochemistry (6/128 patients: 4.6%). There was no false-negative No node metastases were found in 104/128 (81.2%) patients. Conclusions: SN detection supplied additional information in 39.8% of patients (51/128), either showing that drainage occurred via unusual pathways or detecting cancer spread via immunohistochemistry. Node dissection could have been avoided in the 104/128 patients with negative nodes, potentially decreasing treatment-associated morbidity. No significant financial relationships to disclose.