A prospective randomized trial comparing colorimetric and fluorescent imaging for pelvic sentinel lymph node mapping in endometrial cancer

Abstract

Objectives: To prospectively compare sentinel lymph node (SLN) mapping using isosulfan blue (ISB) with indocyanine green (ICG) using near-infrared imaging in patients with apparent uterine-confined endometrial cancer (EC). Our aims were to: (1) test the null hypothesis of no difference between ISB detection of SLN versus ISB + ICG in a randomized clinical trial, and (2) determine sensitivity and negative predictive value for detection of lymph node (LN) metastasis. Methods: Following US Food and Drug Administration and institutional review board approvals, 200 endometrial cancer patients were enrolled (September 2012 to January 2015) with 5 surgeons. Patients were randomized using SAS version 9.2, with 10% cases allocated to ISB alone. ISB (2 mL) and ICG (2 mL; 1 mg/mL) were injected into the cervix, 2 to 5 mm below the mucosa. All patients underwent robotic-assisted hysterectomy with pelvic ± aortic lymphadenectomy. Retroperitoneal spaces were opened and ISB results recorded on all 200 cases. Envelopes were opened and near-infrared imaging was used per randomization cards on 180 subjects. Results: A total of 180 cases were mapped with ISB + ICG (group A), and 20 cases were randomized to ISB alone (group B). Mean age was 64.5 ± 8.4 years and body mass index was 33 ± 8 kg/m2. Histologies included endometrioid G1 (43%), G2 (30%), G3 (7%), and type II (20%). Pathology characteristics included lymphovascular space invasion (31.5%), DOI more than 50% (27.5%), lesion size greater than 2 cm (77%). Mean time from dye injection to opening of spaces was 10.3 ± 6 minutes and time to complete mapping with LN removal was 22.4 ± 9.4 minutes. Operating room time was 136 ± 36 minutes, total LN were 23.1 ± 10.7. Group A ISB detection of SLN was not different from that of control group B (P = .35). SLN detection for ISB + ICG group A (n = 180) versus all ISB (n = 200) were as follows: bilateral—83.9% versus 40%; unilateral—12.2% versus 36%; and none—3.9% versus 24%, P < .001). Median SLN per case in group A was 2 (range, 0–4). Positive SLNs were found in 39 group A (21.7%) and 2 group B cases (10%). There was 1 false-negative SLN (97.5% sensitivity, 99.3% negative predictive value, 2.5% false-negative rate). SLN was the only positive LN in 25 (62.5%) of 40 node-positive cases. Isolated tumor cells were found in 11 (28.2%) of 39 SLNs compared with 2 (14.3%) of 14 non-SLN metastases. ISB improved the ICG detection of SLNs (3 bilateral, 10 unilateral, 1 LN metastasis). No allergic reactions to either dye were seen. Conclusions: ISB + ICG and near-infrared imaging detected more SLN and LN metastases than ISB alone in this phase III trial. SLN mapping with ICG +ISB had excellent sensitivity for detection of metastasis.