Objective To establish the accuracy of sentinel lymph node (SLN) detection in early cervical cancer.

Materials and Methods Sentinel lymph node detection was performed prospectively over a 6-year period in 86 women undergoing surgery for cervical carcinoma by the combined method (Tc-99m and methylene blue dye). Further ultrastaging was performed on a subgroup of 26 patients who had benign SLNs on initial routine histological examination.

Results The SLN was detected in 84 (97.7%) of 86 women by the combined method. Blue dye uptake was not seen in 8 women (90.7%). Sentinel lymph nodes were detected bilaterally in 63 women (73.3%), and the external iliac region was the most common anatomic location (48.8%). The median SLN count was 3 nodes (range, 1–7). Of the 84 women with sentinel node detection, 65 also underwent bilateral pelvic lymph node dissection, and in none of these cases was a benign SLN associated with a malignant non-SLN (100% negative predictive value). The median non-SLN count for all patients was 19 nodes (range, 8–35). Eighteen patients underwent removal of the SLN without bilateral pelvic lymph node dissection. Nine women (10.5%) had positive lymph nodes on final histology. One patient had bulky pelvic nodes on preoperative imaging and underwent removal of the negative bulky malignant lymph nodes and a benign SLN on the contralateral side. This latter case confirms the unreliability of the SLN method with bulky nodes. The remaining 8 patients had positive SLNs with negative nonsentinel lymph nodes. Fifty-nine SLNs from 26 patients, which were benign on initial routine histology, underwent ultrastaging, but no further disease was identified. Four patients (5%) relapsed after a median follow-up of 28 months (range, 8–80 months).

Conclusion Sentinel lymph node detection is an accurate and safe method in the assessment of nodal status in early cervical carcinoma.