Sentinel lymph node biopsy for breast cancer patients using fluorescence navigation with indocyanine green

Sentinel lymph node biopsy (SLNB) is useful to predicts accurate pathological nodal staging and detect the existence of axillary lymph node metastasis. The major problem of axillary dissection is that it produces various morbidities such as serome, infection, numbness and lymphadema of the arm [1, 2]. Since April 2010, this technique has been covered by the national health insurance in Japan, promoting its wide use in clinical settings. False-negative of SLNB is a critical problem, as it may lead to inadequate adjuvant therapy, which can negatively affect a patient's prognosis. Currently, three methods are available to identify sentinel lymph nodes: the dye method, the gamma probe method, and a combination of both [3‐14]. Although the dye method has several benefits including ease of use, cost effectiveness and safety, it has been pointed out that the detection rate is low compared to the gamma probe method [3‐14]. The avoidance of radioisotope is also beneficial especially at institutes where the use of radioisotope is limited. Reports have indicated that the percentage of false-negative SLNB results could be less than 5% in experienced hands if dual-agent techniques are used [3‐16]. Meanwhile, the lymphoscintigraphy with a radioisotope cannot clearly visualize the lymphatic drainage pathway. The photodynamic eye (PDE) can visualize the lymphatic drainage pathway clearly and demonstrate the accurate location of sentinel lymph node real-time in the operating room. In our study, we performed SLNB using the indocyanine green (ICG) fluorescence technique to determine whether this can overcome the shortcomings of the dye method, and analyzed breast lymphatic drainage pathways.

Many studies have been conducted on SLNB, since the publication by Krag et al. on the method using radioisotope in 1993 [10, 11], and a report by Giuliano et al. on the dye method in 1994 [12, 13]. Veronesi et al suggested the possibility that a total dissection of the axillary lymph nodes is not always necessary, and a large scale RCT was carried out [1, 2, 20‐29]. These reported successful lymphatic mapping in 92% using a combination of a dye and a gamma probe method. They found that SLNB using dye alone allowed the identification in approximately 70-80% of patients and that adding a radioisotope improved the detection rate. It is therefore certain that the combination of dye and radioisotope will become more commonly used in SLNB [3‐16].

It is important, therefore, that the accuracy of SLNB be secured by each surgeon and each institute. Currently, two types of tracers, dye and radioisotope, are used to detect the sentinel lymph nodes in breast cancer patients, and a combined use of the two tracers has yielded a high diagnostic value [3‐14].

We performed the SLNB using the ICG fluorescence measurement technique in order to overcome some drawbacks inherent in the dye method (Table 6). At our department, the identification rate of sentinel lymph nodes is 100%. This is a result similar to or better than those reported by other researchers, which were obtained by the gamma probe method or the combined method that uses the gamma probe and the dye [3‐14]. Thus, it can be said that SLNB using ICG fluorescence could be very useful.

We usually inject ICG into the subareola and periareola. Because the mammary gland is embryologically derived from the ectoderm, it is thought that the mammary gland tissue and the subcutaneous tissue share common lymphatic vessels, which lead to the relevant lymph nodes [30‐34]. Lymphatic flow originating in the mammary gland tissues first moves into the subareolar lymphatic vessels, and then drains to the axillary lymph nodes. Therefore, the injection of ICG into the subareola and periareola allows us to identify the same sentinel lymph nodes as those identified with the peritumonal injection. Some studies [30‐34] attempted to clarify whether injection of mapping agents into the subareolar breast tissue results in the identification of the same sentinel lymph nodes as injection into intradermal sites overlying the breast tumor. It would be simpler if the injection could be performed in the same manner for all patients regardless of tumor location or palpability. Other studies [30‐34] have shown that a dermal or subdermal injection of mapping agents results in more reliable sentinel lymph node identification than a peritumonal injection. It has been hypothesized that the lymphatic drainage pathway of the entire breast flows to the same few sentinel lymph nodes. No lymphatic drainage pathway was observed to the parasternal lymph node and the supraclavicular lymph node. It was reported that the lymphatic drainage pathway in the breast parenchyma flowed toward the subareolar lymphatic plexus and drained in to the axilla [30‐34]. In the present study, a lymphatic drainage pathway from the areolar area was detected in almost all patients.

In our study, ICG, which was injected into the subareola and periareola, emitted a high level of fluorescence at the injection location, then promptly moved into the lymphatic vessels, and finally spread further through the lymphatic flow courses. The PDE infrared camera system has an ability to capture fluorescence up to a depth of 1 cm from the skin surface. In patients with a low body mass index (BMI), it was possible to directly observe the fluorescent lymph nodes in the axillary region because their subcutaneous fat layer was thin.

On the other hand, patients with a high BMI usually have a thick subcutaneous fat layer, making it difficult or impossible to track the lymphatic drainage pathway toward the axilla by fluorescence. It seems that in 7 patients (2%) failure to detect a lymphatic drainage pathway was unavoidable, as all 7 patients were obese (BMI > 30). BMI has been shown to correlate with an increased incidence of failure in identifying a sentinel lymph node. The specific causes of mapping failure remain uncertain, but some contributing factors have been observed [35‐39]. Using inadequate volume of injection, inadequate dispersion, inappropriate timing, and inadequate massage may weaken the ability of the dyes and radioisotopes to reach the nodal basin. Fatty tissue around the lymph nodes may also cause decreased flow to the lymphatic basin. Successfully lymphatic drainage pathway mapped patients had a mean BMI of 22.3 kg/m² (range 17.9 to 33 kg/m²).

However, for these obesity patients (BMI > 30), it was possible to easily identify the lymph nodes and lymphatic vessels embedded in the adipose tissue by incising the subcutaneous fat in the axillary incision site. The intensity of fluorescence can be adjusted if required. It is also possible to distinguish between the fluorescence from the lymph nodes or lymphatic vessels and the fluorescence from ICG that happens to attach somewhere. If the fluorescence is strong, and can determine whether the lymph nodes or lymphatic vessels, the deposit of a ICG is also possible to adjust the high and low emission intensity of the device. None of these obesity 7 patients were not failed to identify sentinel lymph node and not had false negative sentinel lymph nodes.

More sentinel lymph nodes were detected by ICG fluorescence imaging compared with other methods (Table 4) [1‐14, 17, 18]. The uptake of the localization agents into the lymphatic vessels is dependent on the interstitial pressure [19, 30‐39]. Our data from past experiences suggest that ICG fluorescence could detect a wide range of sentinel lymph nodes. ICG in the lymph node can be illuminated by fluorescence with very high sensitivity even if it does not appear green. When we detect many harvested sentinel lymph nodes, the following can be suggested according to several differences between the dye and radioisotope: (1) the molecular weight and the chemical structure, (2) the intensity of binding of some kinds of protein to the tracers, (3) the injection time to tracers, and (4) the sensitivity of methods [3, 4]. The small blue dye crystals passively rely on prevailing fluid dynamics for visualization of the lymph node as fluorescence is emitted. Our experiments clearly revealed that extensive sentinel lymph nodes in the axilla could be detected only with the ICG fluorescence method. In our department, 20 times diluted ICG is used. The intensity level of fluorescence obtained from this ICG is appropriate, allowing us to perform the SLNB in an effective and efficient manner. It was possible to identify the irregularities, which may appear around the fluorescence emission when some lymphatic vessels are damaged, by adjusting the emission intensity of fluorescence. It is also possible to distinguish the lymph node from the lymphatic vessel by adjusting the fluorescence emission intensity. (The lymph nodes illuminate in a spotty pattern, whereas the lymph vessels illuminate in a linear or streaky pattern.)

We observed two main lymphatic drainage pathways draining from the areola in the direction of the axilla: one was directly toward the axilla and the other was toward and through the outside of the mammary gland. Regarding the number of lymphatic vessels, it was observed in many patients that several lymphatic vessels running in the same direction join together to form the lymphatic vessel network. In some cases, it flows into multiple sentinel lymph nodes through multiple lymphatic vessels (Figure 1, 2).

When the presence of lymph node metastasis was observed in intraoperative biopsy, a total dissection of the axillary lymph nodes was performed. However, additional surgery is not usually performed for patients in which micrometastasis is detected by the postoperative HE staining of permanent preparation but other lymph nodes removed as part of a group sentinel are metastasis-negative. Currently, postoperative tumor recurrence in the axillary lymph nodes has not been observed (Table 5). The clinical significance of micrometastasis and ITC (isolated tumor cell) has not been established. These were false negative cases, resulting from the fact that metastatic lesions in the slice for intraoperative biopsy were too small and it was only possible to detect such lesions using permanent preparation. Any metastatic legion in the 8 patients who showed false negative results for metastasis was very small, less than 2 mm, which prevented detection by intraoperative biopsy. However, some reports have suggested that there might be an association between micrometastasies in the lymph nodes and distant metastasis [27‐29]. The prognostic significance of a micrometastasis in terms of metastatic risk is controversial [1, 2, 20‐29].

At present, no standard guidelines are available for dealing with micrometastasies in the sentinel lymph nodes. Among the patients who underwent a total dissection because metastasis had been detected in a single sentinel lymph node, no patient developed metastasis in other organs or parts other than the sentinel lymph nodes. If we are able to accurately identify the patients with metastasis only in the sentinel lymph node(s), it would be possible to reduce the need for further surgery or unnecessary total dissections.

The fluorescence method allows us to see how the lymphatic drainage pathway flows from the areola toward other parts in real-time during surgery, as well as clearly understand how the lymphatic vessels run from the areola to the axilla. With the dye method, sometimes it is difficult or impossible to visually determine whether the extracted lymph nodes are sentinel lymph nodes or not. Meanwhile, with the fluorescence method, this can be done relatively easily by checking the existence of fluorescence in the extracted lymph nodes. This method is a technique that enables us to easily identify the sentinel lymph nodes during surgery with low invasiveness and no exposure to radiation, thus providing a safe and very useful way to detect the axillary sentinel lymph nodes in patients with breast cancer, in additon to improving the identification rate.