Twenty patients with clinical T1-T3 breast cancer at diagnosis who received NCT between October 2015 and July 2016 were prospectively enrolled from Yonsei Cancer Center, Severance Hospital, Yonsei University College of Medicine, Seoul, Korea. All patients had cytology-proven ALN metastases prior to NCT. Patients with a past history of allergic reaction to charcoal agents, inflammatory or T4 breast carcinoma, or stage IV cancer at diagnosis were excluded. At presentation, all patients with clinically enlarged ALNs or with radiologically suspicious nodal findings of loss of fatty hilum, cortical thickening > 3 mm, round shape, markedly hypoechoic cortex, or increased peripheral blood flow underwent ultrasound-guided fine needle aspiration cytology of the most suspicious or largest ALN to confirm the presence of metastasis.

Prior to NCT, the main breast tumor was marked with a metallic clip and then the cytology-proven metastatic ALN was tattooed with charcoal suspension (Charcotrace, Phebra, Sydney, Australia) through ultrasound-guidance by a radiologist (M.J.K.) who had more than 14 years of specialized experience in breast imaging. To minimize obstacles to pathologic examination, the charcoal was mainly injected into the capsular area of the metastasis-confirmed ALN and a needle-point injection track was made from the ALN to the skin of the axilla. A median dose of 0.5 mL (range, 0.2 to 0.8 mL) was used for node tattooing.

After four cycles of anthracycline followed by taxane-containing NCT, the responses of the primary tumor and nodes were comprehensively evaluated using clinical and radiological examinations, mainly with ultrasound or breast magnetic resonance images. For cases with human epidermal growth factor receptor 2 (HER2)‒positive disease, trastuzumab was incorporated into the NCT regimens. According to the Response Evaluation Criteria in Solid Tumors ver. 1.1, tumor response to NCT was considered as complete or partial response and non-response was defined with stable or progressive disease. The ALN response to NCT was defined with the disappearance of metastatic ALNs or no suspicious imaging findings in the axilla. Nodal non-response was considered when suspicious findings in the axilla remained after completion of NCT.

Pathologic stages were based on the AJCC Cancer Staging Manual, seventh edition [13]. The modified Scarff-Bloom-Richardson grading system was used for histologic grading. Tumors with ≥ 1% nuclear-stained cells by immunohistochemistry using core needle biopsy specimens were considered positive for estrogen receptor (ER) and progesterone receptor (PR) according to the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) guidelines [14]. HER2 staining was scored as 0, 1+, 2+, or 3+ according to the ASCO/CAP guidelines [15]. In cases with HER2 2+ results, silver in situ hybridization (SISH) was performed using the INFORM HER2 Dual ISH DNA Probe Cocktail Assay (Ventana Medical Systems, Tucson, AZ) with an automated slide stainer according to the manufacturer’s protocols. HER2 gene amplification was defined with a HER2 gene/chromosome 17 copy number ratio ≥ 2.0 or a HER2 gene/chromosome 17 copy number ratio < 2.0 with an average HER2 copy number ≥ 6.0 signals/cell according to the ASCO/CAP guidelines [15]. An immunohistochemistry score of 3+ or gene amplification by SISH was considered HER2-positive. During the study period, Ki-67 staining was mainly performed on post-treatment surgical specimens for residual carcinomas in our institution and was not investigated in the present study.

Based on ER, PR, HER2, and histologic grade findings, molecular phenotypes were categorized into the following four subgroups: luminal A (ER-positive and/or PR-positive, HER2-negative and grade 1 or 2), luminal B (ER-positive and/or PR-positive, HER2-negative and grade 3; or ER-positive and/or PR-positive, and HER2-positive), HER2-enriched (ER-negative, PR-negative, and HER2-positive), and triple-negative breast cancer (TNBC) (ER-negative, PR-negative, and HER2-negative).

SLNs were detected using a combination of radioisotopes and blue dye techniques except for one detected with a radioisotope alone. In brief, 0.5 mCi ^99mTc-Phytate (Korea Atomic Energy Research Institute, Daejeon, Korea) was diluted in 0.5 mL saline and injected into the subcutaneous layer of the areolar tissue in the direction of the main primary tumor at the Department of Nuclear Medicine on the day of surgery. In the operation room just before surgery was initiated, 0.8% indigo carmine dye (Carmine, Korea United Pharm Inc., Seoul, Korea) was injected into the periareolar area followed by a massage of the injection site. SLNs were excised and defined as hot and/or blue-colored nodes with a handheld gamma probe (Neoprobe Gamma Detection System, Neoprobe Corp., Dublin, OH) at gross visual inspection or with a radioactive count of ≥ 10% of the ex vivo count of the hottest node. During SLNB, clinically palpable cold or non-blue axillary nodal tissues were also excised and categorized as axillary sampling. Charcoal-tattooed ALN and track were removed during axilla surgery and included in the SLNB or axillary sampling procedures. SLNs and axillary sampling tissues were then sent to frozen section analysis. After SLNB and/or axillary sampling, the decision to perform further concomitant completion level I/II ALND was made taking into account patient’ consent, initial nodal burden, node responses during and after NCT, or frozen section results of SLNB and/or axillary sampling.

All excised ALNs including SLNs were subjected to routine pathological examinations, which included hematoxylin and eosin staining with or without immunohistochemistry for cytokeratin in accordance with previously published recommendations [16]. ALNs were considered positive if metastatic foci were > 0.2 mm and/or > 200 tumor cells (≥ ypN1mi) according to the AJCC Cancer Staging Manual, seventh edition [13]. Any nodes with isolated metastatic foci of ≤ 0.2 mm were considered negative [ypN0(i+)], but were not considered to have achieved a node pathologic complete response (pCR). Regarding primary tumor responses to NCT, the absence of in situ or invasive carcinomas (ypT0) or residual in situ carcinoma alone without invasive disease (ypTis) constituted a tumor pCR.

The detection of charcoal-tattooed node was investigated and its association with the detection of SLNs was analyzed. Detection failure of SLNs was defined as an inability to identify hot and/or blue nodes by lymphoscintigraphy or gamma probe with visual inspection. Sensitivity was calculated by dividing true positive (TP) findings by TP plus false-negative (FN) findings. FNR was defined as the proportion of patients with negative SLNs who subsequently had metastatic ALNs among patients with ≥ 1 positive lymph node. NPV was calculated by dividing true negative (TN) findings by TN plus FN findings. Accuracy was defined as the proportion of patients with TP or TN among patients with successful SLNB. The significances of differences in SLNB performance of subgroups and in clinicopathological parameters between groups were evaluated using the chi-square test or Fisher exact test. The Mann-Whitney U test was used to compare the median values of continuous numerical data. All statistical tests were two-sided, and p < 0.05 was considered statistically significant. SPSS ver. 23.0 (IBM Corp., Armonk, NY) and SAS ver. 9.4 (SAS Inc., Cary, NC) were used for analysis.

This study was approved by the Institutional Review Board (IRB) of Severance Hospital, Yonsei University Health System, Seoul, Korea (IRB No. 4-2015-0465) and registered to the Clinical Research Information Service (CRIS, http://cris.nih.go.kr/cris/index.jsp; No. KCT0002370), Korea. Written informed consent was obtained from all patients prior to study entry.

The mean age of 20 patients was 52.3±10.9 years (range, 34 to 70 years) at diagnosis. The mean size of the main primary tumor at presentation was 3.1±1.8 cm (range, 1.2 to 8.4 cm). Table 1 shows the clinicopathological characteristics and treatment patterns of the enrolled patients. No significant side effects or complications occurred during charcoal tattooing of metastatic ALN. After finishing the tattooing procedure, 16 patients (80%) underwent NCT on the very day that charcoal tattooing was performed and the time interval between charcoal tattooing and NCT ranged from 0 to 5 days. The median time interval between charcoal tattooing and surgery was 181 days (range, 165 to 197 days).

Final permanent pathology of breast and ALNs is presented in Table 1. Preoperative imaging studies of three patients with initially suspicious non-axillary regional lymph nodes showed no radiological evidence of non-axillary regional metastasis after completion of NCT. Therefore, when pathologic results of the breast and axilla were considered, both tumor and nodal pCR were determined in six patients (30.0%). There was no statistical difference in the nodal pCR rate according to the type of axillary surgery (ALND 41.7% vs. axillary sampling 62.5%, p=0.650) (Table 1).

One or more hot and/or blue SLNs were detected in all patients (detection rate, 100%). The median number of SLNs and total ALNs (SLNs+non-sentinel ALNs) retrieved was 3.0 (range, 1 to 12) and 10.0 (range, 4 to 18), respectively. The median number of non-sentinel ALNs excised was 4.5 (range, 0 to 15). The number of pathologically retrieved nodes in two patients who underwent SLNB alone was four and 10, respectively. During axillary surgery, charcoal-tattooed ALN and track were easily identified by visual inspection and excised in all patients (detection rate of tattooed node, 100%). According to the type of axillary surgery, the median number of SLNs was not different (n=3 in each group, p=0.624), but the median number of non-sentinel ALNs was significantly larger in the ALND group; therefore, the median number of total ALNs removed was significantly higher in the ALND group (ALND, 13 [range, 7 to 18] vs. axillary sampling, 4.5 [range, 4 to 10]; p < 0.001).

Table 2 compares SLNB results of removing hot and/or blue SLNs with the final ALN status. In addition to hot and/or blue SLNs, charcoal-tattooed node was calculated together as the modified SLNB. The diagnostic performance of the modified SLNB showed the highest values as follows: sensitivity, 100%; FNR, 0.0%; NPV, 100%; and accuracy, 100%.

Overall diagnostic performance of SLNB is presented in Table 3 including subgroups. When more SLNs were excised, the sensitivity, NPV, and accuracy improved and FNR decreased, although statistical significance was not demonstrated. Similarly, wider axillary surgery or favorable nodal response was associated with improved diagnostic performance of SLNB. Better performance was exhibited in patients with luminal A or B subtypes than in those with non-luminal (HER2-enriched or TNBC) subtypes with borderline significance.

Charcoal-tattooed ALN was among the hot and/or blue SLNs in 15 patients (75.0%) but removed during cold and non-blue axillary nodes sampling in five cases (25.0%). Among 15 patients with concordance between the tattooed node and SLNs, eight (53.3%) showed residual metastatic disease in the charcoal-tattooed ALN. The other seven patients (46.7%) were determined to have negative conversion of the tattooed node and finally achieved nodal pCR. Among five patients with discordance between the tattooed ALN and SLNs, three (60.0%) showed negative sentinel and tattooed nodes and finally achieved nodal pCR. Of the remaining two patients, one 61-year-old patient showed that three SLNs detected were negative, but the charcoal-tattooed non-sentinel ALN showed residual metastasis (Fig. 1). In the other 59-year-old patient, only one SLN was retrieved. Frozen section and permanent pathology of the SLN confirmed the presence of a few atypical cells and isolated tumor cells with a maximal diameter of 100 μm [ypN0(i+)], respectively. Subsequently, ALND was performed and metastatic carcinomas in four out of 12 ALNs were found including micrometastasis (450 μm) in the tattooed ALN.

Regarding the intensity order of the uptaken radioisotope, the tattooed node was the hottest in eight patients, the second hottest in three, and the third hottest in four. When clinicopathological characteristics of patients with concordant SLNs and tattooed node were compared to those of patients with discordant findings, no statistically significant parameters were identified (Table 4).