Introduction

Sentinel lymph node (SLN) biopsy has become a standard procedure in early breast cancer detection, not only in western countries but also in Japan [1], and large-scale studies have not shown the efficiency of axillary lymph node dissection (ALND) on the outcomes of patients with SLN-negative breast cancer [2]. Moreover, the American College of Surgeons Oncology Group’s (ACOSOG) Z0011 trial suggested that avoidance of an ALND may be permitted even for SLN-positive patients [3]. The American Society of Clinical Oncology’s (ASCO) guideline recommendations published in 2014 stated that women without SLN metastases should not receive an ALND and that women with one to two metastatic SLNs planning to undergo breast-conserving surgery with whole-breast radiation should not undergo ALND, in most cases [4]. Breast cancer clinical practice guidelines currently available in Japan recommended an SLN biopsy using a combination of dye and radioisotope (RI) methods based on several reports [5–7].However, RI has the disadvantages of requiring a radioactive facility, causing exposure to radiation, and a high cost. In contrast, the blue dye method is easy to perform, even though SLN identification rate is reported to be lower when compared to the RI method, and some patients experience an allergic reaction [8].

Indocyanine green (ICG) in plasma emits light at the wavelength of 760 nm and produces fluorescence at 830 nm, which are visualized as a real-time image with a photodynamic eye (PDE) camera. Therefore, the ICG fluorescence method is suited for intraoperative SLN biopsy [9–13]. Since ICG alone is not as sensitive a dye for color perception [14], we have occasionally had difficulty in identifying an SLN using a dye method without fluorescence. Moreover, fluorescence is more sensitive than RI [15], and the ICG can spread further within the lymphatic basin because of its low molecular weight [12]. Therefore, the ICG fluorescence method has led to too many SLNs or difficulty in identifying SLNs due to frequent leaks in the ICG. To solve these problems, the patent blue was used in combination with the ICG as a dye. The patent blue seems to be better as a dye for color perception than ICG alone [12, 15]. The present study was performed to demonstrate the efficiency of ICG fluorescence and the blue dye method, and we describe the follow-up results of patients who underwent SLN biopsy alone using this method.

Patients and methods

We have performed SLN biopsy using the combination of ICG fluorescence and the blue dye (patent blue) method in primary breast cancer patients with a T1–3 tumor without clinical axillary lymph node metastasis (N0) and distant metastasis (M0) at the Kaizuka City Hospital from March 2007 to February 2014. Patients who had preoperative systemic therapy (e.g., chemotherapy or endocrine therapy) were excluded. Administration of the patent blue was approved by the local ethics committees at our hospital. Well-trained surgeons performed the SLN biopsy according to the standard procedure.

Sentinel lymph node (SLNs) were defined as: axillary lymph nodes, which are blue-stained and/or have positive ICG fluorescence, as detected with a PDE. Para-SLNs were defined as surgically removed swollen axillary lymph nodes that contained neither ICG fluorescence nor blue dye inside. Since swollen nodes other than blue nodes may be considered similar to SLNs in a wide sense of the sentinel node concept, we resected the swollen para-SLNs, if any.

During the procedure an agent, which was a mixture of 1.0 mL of 0.5 % ICG as a source of fluorescence and 2.0 mL of 1 % patent blue as a blue dye, was injected into the subareolar area after administering general anesthesia and sterilizing the operation site. After 3.5–5.0 min of breast massage, the lymphatic flow pathway was traced with a PDE camera (a charge-coupled device; Hamamatsu Photonics Co., Hamamatsu, Japan) (Fig. 1a). An appropriate skin incision was made at the point where the lymphatic flow drained into the axillary area, and then the patent blue and/or ICG stained lymph nodes were removed (Fig. 1b). The resected lymph nodes were examined with a PDE and were classified as SLNs or para-SLNs according to ICG fluorescence and/or blue dye detection (Fig. 1c). SLN metastasis was detected with intraoperative hematoxylin and eosin-based histopathologic examination of sliced 2 mm sections of frozen tissues.

Fig. 1
figure 1

a The lymphatic flow pathway of the indocyanine green (ICG) fluorescence is traced with a photodynamic eye (PDE) camera. b After an appropriate skin incision is made at the point where the lymphatic flow drained into the axilla, the patent blue and/or ICG stained lymph nodes are removed. c The resected lymph nodes are examined with a PDE and are classified according to the ICG fluorescence and/or blue dye detection. The sentinel lymph node (SLN) have ICG fluorescence and/or blue dye. The para-SLNs have no ICG fluorescence or blue dye inside, even if ICG fluorescence and/or blue dye in the surrounding tissue are observed

Full size image

The estrogen receptor (ER) and progesterone receptor (PgR) of the main breast tumor were evaluated with immunohistochemistry (IHC), and tumors with ≥1 % positively stained tumor cells were classed as positive. The HER2 status was considered positive if the IHC was 3+ or if the fluorescence in situ hybridization (HER2 to chromosome 17) was >2.2.

Results

From March 2007 to February 2014, we performed SLN biopsy in 714 patients with early breast cancer using a combination of ICG fluorescence and the blue dye method. By using this combination method, SLNs were identified in 711 of 714 patients (detection rate, 99.6 %). One of three patients who were not identified with an SLN had a history of previous surgery in the axillary area. The average number of SLNs was 2.4 (range 1–7 nodes), and the average of para-SLNs was 0.4 (range 0–5 nodes). No adverse reactions because of hypersensitivity were found.

Among 711 patients, 107 had SLN involvement in the final pathological reports. Of those, 94 patients had macrometastases of SLNs, 9 had micrometastases of SLNs, and 4 patients had isolated tumor cells (ITCs) of SLNs.

Ninety-nine patients had SLN and/or para-SLN involvement according to intraoperative pathological examination; thus, they underwent complete ALND. Of 99 patients with ALND, 46 had only SLN involvement, 50 had both SLN and non-SLN involvement, and 3 had only non-SLN involvement including para-SLN (Table 1). Two of three patients whose SLNs were not identified underwent ALND. One patient preferred not to undergo ALND.

Table 1 Detection rate and results of intraoperative pathological examination with the combination method
Full size table

In 612 patients with negative SLNs and para-SLNs, according to intraoperative pathological examination, 148 were diagnosed with ductal carcinoma in situ (DCIS) based on the resected specimen. Patients who underwent ALND, those with DCIS, and those whose SLNs were unidentified were excluded. Finally, 464 patients with invasive carcinoma who did not undergo ALND were included in the follow-up analysis. The patient’s characteristics are shown in Table 2. These patients had negative SLN metastasis according to intraoperative pathological reports, but two had macrometastases of SLNs, seven had micrometastases of SLNs, and two had ITCs of SLNs according to the final reports (Table 3). Adjuvant therapy was given according to St. Gallen’s recommendations or the ASCO guidelines (Table 2).

Table 2 Patients and tumor characteristics
Full size table
Table 3 Final pathological results of SLN-negative cases with intraoperative examination
Full size table

During a median follow-up of 38 months (range 4.4–87.7), 2 of 464 patients (0.4 %) developed axillary lymph node (ALN) recurrence and 11 of 464 patients (2.4 %) developed distant recurrences. However, no patient had ipsilateral breast tumor recurrence or both ALN and distant recurrences.

One patient with ALN recurrence was 45 years old at the time of the primary surgical treatment. She had a T2 tumor and underwent mastectomy; tamoxifen was administered as adjuvant therapy. The pathological reports were as follow: invasive ductal carcinoma, pT2, no involvement of 1 SLN and 2 para-SLNs, grade 2, ER positive, PgR positive, and lymphovascular invasion (LVI) positive. After a disease-free interval of 50 months, ALN recurrence occurred. Another patient with an ALN recurrence was 67 years old at the time of the primary surgical treatment. She had a T1 tumor and underwent mastectomy; letrozole was administered as adjuvant therapy. The pathological reports were as follows: invasive ductal carcinoma, pT1, no involvement of three SLNs, grade 3, ER positive, PgR positive, and LVI positive. After a disease-free interval of 24 months, ALN recurrence occurred. Both patients were successfully treated with salvage axillary dissection, and they have had no locoregional or distant recurrence after ALND (Table 4).

Table 4 Patients with ALN recurrence
Full size table

Discussion

Although the RI method for SLN biopsy has a high SLN detection rate, it requires a radioactive facility, stringent safety controls, and has a high cost [8]. In contrast, the ICG fluorescence method requires little skill, and the cost of the agent and device is much lower than in the RI method [12, 14]. Therefore, the use of this method has been increasing recently, especially in community hospitals in Japan.

The ICG fluorescence method is advantageous for determining the skin incision line in the axilla area of the SLN biopsy, because fluorescence of the lymphatic flow is traced with a PDE camera simultaneously. When the lymphatic flow is traced with a PDE camera, the lighting is decreased to visualize the ICG fluorescence clearly, but it is difficult to perform the SLN biopsy simultaneously in dim light [9]. Additionally, SLN detection is awkward without fluorescence under surgical light, because ICG is less sensitive than other blue dyes with regards to color perception. In former studies, the SLN biopsy using ICG without fluorescence had a low detection rate. Motomura et al. [14, 16] reported that the SLN detection rate was 73.8 % with ICG alone without fluorescence. In our experience of performing SLN biopsy with ICG alone without fluorescence between January 2003 and July 2007, the detection rate was 65.8 % (25/38 patients). We also performed SLN biopsies with dye (the patent blue) alone from August 2006 to March 2007, and the detection rate was 80.8 % (21/26 patients). Therefore, a high identification rate may be achieved by combining the patent blue with ICG which is excellent for color perception. Hojo et al. [17] showed a high identification rate with a combination of ICG fluorescence and the patent blue method. Wishart et al. [18] demonstrated that the combination of ICG and the blue dye method had the highest nodal sensitivity in a study using a three tracer (RI, blue dye, and ICG). The SLN detection rate in our study is consistent with these reports. However, we could not distinguish ICG fluorescence only-positive SLNs from both ICG fluorescence and blue dye-positive SLNs because of contamination. It was not proven that the combination method is superior to ICG fluorescence only. In fact, we had an elevated detection rate when using the combination method. This may easily detect deep SLN because of the advantages of ICG fluorescence and blue dye.

However, the prognosis after performing SLN biopsies using a combination of ICG fluorescence and the blue dye (patent blue) method is not yet clear. ACOSOG’s Z0010 trial was a large multicenter study that evaluated locoregional recurrence after a negative SLN biopsy with RI and/or blue dye method. The study demonstrated that the 5-year incident rate for local recurrence, regional recurrence, and distant recurrence were 2.4, 0.5, and 2.8 %, respectively [19]. In the current study, we observed the outcome of patients undergoing SLN biopsy using the combination method. At a median follow-up of 38 months (range 4.4–87.7), there were only 2 (0.4 %) ALN recurrences, 11 (2.4 %) distant recurrences, no local recurrence, and no patient among the SLN-negative patients with both ALN and distant recurrences.

It appears to be meaningless to statistically analyze the predictive factors related to the ALN recurrences, because the numbers of these events were too small. However, it may be noteworthy that both the patients with ALN recurrences received mastectomy, endocrine therapy, and no radiation therapy as adjuvant therapy; LVI positive in the pathological examinations was common. Schlembach et al. [20] noted that radiation therapy associated with breast-conserving surgery irradiated the entire axillary dissection field site in nearly all patients. ACOSOG’s Z0011 trial suggested that no axillary-specific irradiation is likely for treating a significant portion of the axilla [3]. Thus, no radiation therapy and LVI may be associated with ALN recurrence.

Compared with ACOSOG’s Z0011 trial, a limitation of our study was that the follow-up period was short. Considering that cases undergoing breast conservative surgery following radiation therapy were fewer in comparison with ACOSOG’s Z0010 trial (56.7 vs. 92.2 %), the results of our study may be similar to that of ACOSOG’s Z0010 trial [3, 19].

However, there may be a possible disadvantage to using ICG fluorescence and the blue dye method, because the two kinds of dye may increase shock. Fortunately, no adverse reactions due to hypersensitivity were observed in this study.

In conclusion, it is suggested that a combination of ICG fluorescence and the blue dye (patent blue) method can be substituted for a combination of RI and the blue dye method; therefore, it may be possible to avoid using RI. However, a large prospective study is needed to confirm this efficacy.