Stephanie L. Koonce*, Martin I. Newman
University of Minnesota Medical School, USA

*Corresponding author: Stephanie L. Koonce, University of Minnesota Medical School, Cleveland Clinic Florida, 2950 Cleveland Clinic Blvd., Weston, USA
Published: 23 Jun, 2017
Cite this article as: Koonce SL, Newman MI. A Systematic Review of Concordance between Indocyanine Green and 99m Technetium Sentinel Lymph Node Identification in Melanoma. Clin Oncol. 2017; 2: 1308.


Introduction: Radiocolloid 99mTechnetium with or without blue dyes is the most commonly employed method of identifying Sentinel Lymph Nodes (SLN) in melanoma staging. Indocyanine green (ICG) identification of SLN has been reported with equal or superior results and could avoid the use of a radioactive tracer. A systematic review and meta-analysis of the literature was performed.

Methods: A systematic review of the literature was performed identifying peer-reviewed articles examining the concordance between 99mTc and ICG in the identification of SLN in individuals undergoing SLN biopsy for melanoma. Only original study groups were included. The SLN false negative rate and identification rate were pooled according to radiocolloid, blue dye, ICG, or a combination.

Results: Between 1990 and 2016 a total of 14 studies were reported which met inclusion criteria. These studies included a total of 456 patients. The pooled SLN identification rate in patients using radiocolloid with or without blue dye was 1.94 (range 1.14-2.7). Using ICG the pooled SLN identification rate was 2.11 (range 1.1-2.67).

Conclusions: Systematic review of the literature demonstrates no significant difference in SLN identification rate between ICG and 99mTc in melanoma patients. A prospective, randomized controlled trial comparing ICG to 99mTc to more definitively establish ICG as an alternative modality for SLN identification is recommended.


Melanoma is the leading cause of skin cancer mortality and is diagnosed annually in 114,000 patients [1]. Melanoma metastasizes primarily via the lymphatic route with the first lymph nodes metastatic cells encounter being designated the Sentinel Lymph Nodes (SLN). Sentinel Lymph Node Biopsy (SLNB) is recommended by the National Comprehensive Cancer Network for staging of all melanomas with a breslow depth greater than 1 mm and may be considered in patients with melanoma depths less than 1 mm with high risk features [2].

The identification of SLN in melanoma patients is most commonly performed with a combination of blue dye and radiocolloid tracer. An injection of technetium 99m (99mTc) is instilled intradermally around the primary tumor site. Imaging follows, and intra-operative use of a handheld gamma probe allows localization, identification, and confirmation upon extirpation of the node. The blue dye is injected intradermally around the primary tumor site immediately prior to surgery. The blue dye and 99mTc act complementarily in identification of the SLN. SLN identification using these two methods has a reported success rate of 96-99% [3].

Indocyanine green (ICG) is a cyanine dye first tested in 1957 at the Mayo Clinic for used in human medicine [4]. The Federal Drug Administration has approved it for use in medical diagnostics. ICG binds tightly to plasma proteins, is hepatically microsomally metabolized, and has a half-life of 150- 180 seconds [5]. Side effects are rare and mostly minor (including flushing and sore throat).

A systematic review was performed on peer-reviewed articles to evaluate the rates of SLN identification with ICG versus radiocolloid, blue dye, or a combination of radiocolloid and blue dye. In addition, we report our technique for ICG identification of SLN in melanoma patients with SLN in known lymph node basins.


Figure 1
ICG is injected intradermally in four equal aliquots at the primary tumor site prior to incision.


Figure 2
Near infra-red (NIR) imaging is performed to obtain fluorescence imaging of the subcutaneous lymphatics.


Figure 3
Once the lymph node basin is identified, an access incision is performed The lymph node basin is then gently and bluntly probed while visualizing the soft tissue under NIR guidance.


A comprehensive systematic review of published literature over the years 1990-2016 was conducted using the search terms: “melanoma”, “sentinel lymph nodes”, “lymphatic mapping”, “lymph node mapping”, “radiocolloid”, “technetium”, “ICG”, and “indocyanine” in MEDLINE and EMBASE. The search was then supplemented with the references of the selected articles. Inclusion criteria included original study population and intraoperative identification of sentinel lymph node using ICG, radiocolloid with or without blue dye, or a combination of techniques. Studies were required to report the number of patients in whom SLNB were attempted and the number who were successfully mapped with each technique. Studies that did not report SLN identification rate were excluded. Data from all included studies was analyzed with Review Manager 5.3 (Cochrane Review 2014). Data was verified between two reviewers and discrepancies were settled by consensus discussion. Quality of the studies was evaluated using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool. Each item on the QUADAS tool was recorded as “yes”, “no”, or “unclear”.

Outcomes were noted as measures of test performance: the proportion of lymph nodes successfully mapped with ICG and the proportion of lymph nodes successfully mapped with 99mTc. Distributions of covariates were evaluated and summary measures of central tendency and variability were estimated. Summary measures of all outcome measures across studies were estimated by the Mantel- Haenszel method. As significant heterogeneity was observed for most outcomes, a random effects model was utilized with each study weighted by the inverse of its variance. Summary effect estimates hypothesis testing was based on the z-statistic with 95% CIs provided for individual studies, as well as the summary overall effect estimate.

Method of ICG SLN identification
Although methods vary among surgeons, our methods are outlined here for reference. This method is most easily employed when the sentinel lymph node is expected to reside in a specific lymph node basin (e.g.: axilla or groin) though pre-operative knowledge of the nodal basin is not required. After intubation, the patient is prepped and draped in the normal sterile fashion. Using a 30 gauge needle and a 1 cc syringe, a total of 0.10 ml of 500 uM ICG is injected intradermally in four equal aliquots at the primary tumor site prior to incision (Figure 1) Larger lesions may require the injection of a second aliquot. It should be noted that injecting intradermal quantities of ICG in excess of this amount might overwhelm the lymphatic basin resulting in a complete “white out” when visualized under near infrared (NIR) imaging. After ICG injection, the surgeon should wait at least ten minutes before excising the tumor or manipulating the lymphatic channels to allow the ICG to travel to the SLN. Near infrared (NIR) imaging is performed to obtain fluorescence imaging of the subcutaneous lymphatics (Figure 2). Once the lymph node basin is identified, an access incision is performed (Figure 3). The lymph node basin is then gently and bluntly probed while visualizing the soft tissue under NIR guidance. In our experience, the SLN is easily visualized using this method. Once identified in situ, it is excised. Confirmation of the node fluorescence is made with NIR after the excision (Figure 4 and 5).


Figure 4
Sentinel lymph node fluorescence after the excision.


Figure 5
Confirmation of the node fluorescence is made with NIR after the excision.


Figure 6
PRISMA diagram [25].


Figure 7
Forest plot of comparison ICG to 99mTc.


The original literature search returned 3418 references. After excluding abstracts that did not meet inclusion criteria or were unable to be translated into English, 23 abstracts were extracted for full text review (Figure 6). Of these, 14 met criteria for full analysis (Table 1).

Years of publication of analyzed studies ranged from 2009 to 2016. After aggregation of all data, 456 patients underwent SLNB for melanoma with indocyanine green. In addition to the ICG, 449/456 patients also had technetium radioactive tracer used as a control for SLN identification; 263/456 also had blue dye used as a second control for SLN identification. In studies that reported gender, 45.4% (139/306) were female. Follow-up duration averaged 5.6 months (range 3.2-9 months). The primary tumor was localized on the extremities, trunk, or head/neck in 50.6%, 35.9%, and 13.5% respectively. Mean tumour depth was 2.22 mm (range 0.9-3.73 mm). A blue dye was also used for SLN mapping in 263 of the 456 patients. Blue dyes utilized included methylene blue, nonvital blue dye, and isosulfan blue. The mean number of SLN removed identified by ICG was 2.11 and 1.94 for 99mTc with or without blue dye. No significant difference was found between identification rates of SLN with ICG versus 99mTc with or without blue dye (p <0.00001) (Figure 7). Pooled false negative rate for ICG was 2.8% and 5.3% for 99mTc.


Sentinel lymph node biopsy is an established diagnostic procedure in melanoma staging to detect subclinical lymph node metastases. A survival benefit has been demonstrated in patients who underwent lymphadenectomy for intermediate thickness melanoma with positive SLN compared to patients who were observed with lymph node dissection once nodal disease was evident [6]. SLNB is an important staging tool, but has not been shown to improve disease specific survival among all patients. However, among patients with melanoma thickness of 1.2 to 3.5 mm, SLNB is associated with improvement in distant metastasis free survival compared to patients with similar thickness melanoma who are initially observed and subsequently develop nodal metastases [2]. Over the past 18 years, identification of SLN has improved to an overall rate greater than 95% [3]. Reported false negative rate is approximately 5-10% [6,9- 22]. False negatives may occur for myriad reasons including nonlocalization of the SLN. ICG was found to have a lower false negative rate than 99mTc in this review.

Blue dyes have been associated with allergic reactions including urticaria, hives, skin rash, and severe anaphylactic shock [21]. The rate of anaphylaxis was reported as 0.5% to isosulfan blue in a series of 2392 SLN with an overall allergic reaction incidence of 1.6% [23]. Blue dye injection can also cause permanent skin tattooing and stains the surrounding tissue obscuring the surgical field. It is currently contraindicated in pregnancy. 99mTc requires an additional clinic or hospital visit, exposure to radiation, a frequently painful injection, and not insignificant cost. Radiotracer mapping is limited in most developing countries, and production constraints with possible shortage of 99mTc are predicted [22].

Indocyanine green is a fluorescent dye visible with near infrared light that represents an alternative to 99mTc and blue dyes. The side effect profile is minimal with one series reporting 0.15% mild reactions and 0.05% severe reactions [4,5,24]. The small molecular size of ICG compared to the radiocolloids allows for uptake in lymphatics that may be partially blocked from tumor, previous surgery, or trauma. This could explain why ICG may decrease the false negative SLN rate by detecting additional SLN. Fujisawa and Jain noted a higher average number of SLN removed with ICG compared to 99mTc, but other studies noted no significant difference [7-10].

Optimal dosage of ICG for SLN identification has been investigated using dose-escalating protocols. A comparison of 600, 800, 1000, and 1200 uM of ICG: HSA (albumin bound ICG) for SLN identification found 600uM to be the optimal dose [11]. A lower dose study comparing 100, 250, and 500uM ICG:HSA determined 500uM provided the best SLN identification [12]. This suggests that dosage should be in the 500 uM-600 uM range.

The surgeon is able to follow lymphatic drainage in real time at time of ICG injection. The fluorescent nodes themselves, however, are only visible after the skin incision is made, thus the nodal basin should be known prior to injection particularly if the patient has a high BMI, the suspected nodal basin is deep, or there is variable lymphatic drainage [13,14].

Systematic review of the literature demonstrates SLN identification rate is similar between ICG and 99mTc in melanoma patients. Given the predicted shortage and high cost of 99mTc, ICG appears to be a viable option for SLN identification in expected lymph node basins at facilities with limited or no access to radiotracer lymphoscintigraphy. The review suggests that a hybrid approach using both ICG and 99mTc might be the best of both worlds by producing the lowest false negative rate and allowing pre-operative identification of the nodal basin of interest. A hybrid tracer of ICG and 99mTc is being studied which would allow basin identification pre-operatively but would not obviate the issues with 99mTc [15,16]. ICG is well tolerated with a reasonable side effect profile. Further study with a prospective, randomized controlled trial comparing ICG to 99mTc is recommended to more definitively establish ICG as an alternative modality for SLNB in select cases.