Detection of sentinel lymph nodes in patients with endometrial cancer undergoing robotic-assisted staging: A comparison of colorimetric and fluorescence imaging

Abstract

Objective

To retrospectively compare results from lymphatic mapping of pelvic sentinel lymph nodes (SLN) using fluorescence near-infrared (NIR) imaging of indocyanine green (ICG) and colorimetric imaging of isosulfan blue (ISB) dyes in women with endometrial cancer (EC) undergoing robotic-assisted lymphadenectomy (RAL). A secondary aim was to investigate the ability of SLN biopsies to increase the detection of metastatic disease.

Methods

Thirty-five patients underwent RAL with hysterectomy. One mL ISB was injected submucosally in four quadrants of the cervix, followed by 0.5 mL ICG [1.25 mg/mL] immediately prior to placement of a uterine manipulator. Retroperitoneal spaces were dissected for colorimetric detection of lymphatic pathways. The da Vinci^® camera was switched to fluorescence imaging and results recorded. SLN were removed for permanent analysis with ultra-sectioning, H&E, and IHC staining. Hysterectomy with RAL was completed.

Results

Twenty-seven (77%) and 34 (97%) of patients had bilateral pelvic or aortic SLN detected by colorimetric and fluorescence, respectively (p = 0.03). Considering each hemi-pelvis separately, 15/70 (21.4%) had “weak” uptake of ISB in SLN confirmed positive with fluorescence imaging. Using both methods, bilateral detection was 100%. Ten (28.6%) patients had lymph node (LN) metastasis, and 9 of these had SLN metastasis (90% sensitivity, one false negative SLN biopsy). Seven of nine (78%) SLN metastases were ISB positive and 100% were ICG positive. Twenty-five had normal LN, all with negative SLN biopsies (100% specificity). Four (40%) with LN metastasis were detected only by IHC and ultra-sectioning of SLN.

Conclusions

Fluorescence imaging with ICG detected bilateral SLN and SLN metastasis more often than ISB, and the combination resulted in 100% bilateral detection of SLN. Ultra-sectioning/IHC of SLN increased the detection of lymph node metastasis.

Introduction

Lymphatic mapping for assessment of sentinel lymph nodes (SLN) is an accepted practice for breast, melanoma, and vulvar cancers [1], [2], [3] with the primary goal to reduce morbidity of a complete lymphadenectomy. A secondary goal is to improve detection of metastatic disease with pathology protocols that utilize ultra-sectioning of SLN and immunohistochemical (IHC) staining [4]. In 2008, a consensus panel of experts reported that sentinel node assessment in endometrial cancer was worthy of further investigation; however, there was insufficient data to comment on feasibility or benefits [5]. Since then, investigators have reported on their experiences with pelvic lymphatic mapping using colorimetric imaging of blue dyes [isosulfan blue (ISB), patent blue, and methylene blue] and/or radiocolloid mapping with Technetium-99 (Tc-99). Bilateral detection of pelvic lymph nodes is reported in 66 to 86% of cervix and endometrial cancer cases [6], [7], [8], [9]. Furthermore, Roy et al. [9] reported a 7.8% increase in SLN detection utilizing both ISB and Tc-99 compared to ISB alone, achieving a 90.6% bilateral detection rate in patients with cervical cancer. However, radiocolloid mapping suffers from difficulties associated with coordinating injection times in the radiology suite relative to operating times when imaging is desired, variability of operators' ability to interpret the radioactive signal intra-operatively, cost, and patient concerns with injection of radioactive pharmaceuticals.

Recently, other medical dyes that fluoresce in light at the near-infrared (NIR) spectrum (700–900 nm) using laparoscopic imaging systems have been reported for use in lymphatic mapping of gastric, breast, rectal, cervical, and endometrial cancers [10], [11], [12], [13], [14]. Indocyanine green (ICG) is the most clinically useful agent for NIR lymphatic mapping [15], and has been used clinically for two decades with an excellent safety profile. The risk of allergic reactions with ICG has been estimated 1 per 42,000 uses [16]. The da Vinci^® NIR fluorescence imaging system is FDA cleared for vascular imaging and is useful for confirming patency of vascular anastamoses in cardiovascular surgery [17]. The robotic NIR system has also been used for partial nephrectomy by revealing photopenic tumor relative to the surrounding normal renal parenchyma [18]. Rossi et al. [14] recently described preliminary results with NIR fluorescence imaging used with robotic-assisted surgery in patients with cervical and endometrial cancers, recommending that a 1 mg dose of ICG was most efficacious for lymphatic mapping.

In this study, we sought to retrospectively compare the ability of fluorescence imaging of ICG and standard colorimetric analysis of ISB dyes for the detection of SLN in small cohort of women with endometrial carcinoma who underwent robotic-assisted lymphadenectomy (RAL). A secondary aim of this study was to investigate the ability of SLN mapping to increase the detection of metastatic disease by comparing SLN ultra-sectioning and IHC to traditional hematoxylin and eosin (H&E) staining results.