The Relation of Visualization of Internal Mammary Lymph Nodes on Lymphoscintigraphy to Axillary Lymph Node Metastases in Breast Cancer

Introduction

The status of the axillary LN is one of the most important prognostic factors in women with breast cancer. Sentinel LN biopsy (SLNB) is used to provide staging information and to determine the need for axillary LN dissection in breast cancer patients. Many studies have already confirmed the validation of the SLNB for determining the status of regional LN.^2,3

As a result of the ease and accuracy of detecting the sentinel node, preoperative breast lymphoscintigraphy and intraoperative gamma probe localization have become standard practice in many institutions. ⁴ SLNB can prevent complete axillary LN dissection in patients in whom the sentinel LN is found to be free of metastasis. ⁵ Intraoperative lymphatic mapping and SLNB have been rapidly and widely adopted as oncologic equivalents to elective LN dissection. ⁶ However, the impact of the detection of internal mammary sentinel nodes during lymphoscintigraphy is still controversial in regard to staging and managing breast cancer. Sentinel LN mapping with lymphoscintigraphy often identifies non-axillary nodes, although the ability to identify these nodes does not imply a clinical benefit for biopsy or removal.^7–10

This study is designed to investigate whether the visible internal mammary LN of lymphoscintigraphy is related to the burden of axillary LN metastasis. We hypothesized that internal mammary nodal drainage on preoperative lymphoscintigraphy may identify breast cancer patients at risk for axillary LN metastasis through alternative drainage routes.

Material and Methods

Data analysis and statistics

The lymphoscintigraphy was visually categorized according to the following patterns: lymphatic flow, draining LN basin, and intensity of uptake. Images were interpreted by two trained nuclear medicine physicians. The anterior image obtained 5 min and 15 min was used in the present study. First, the lymphoscintigraphy was categorized according to internal mammary sentinel LN visualization. The positive internal mammary sentinel LN lymphoscintigraphy (IM-LPS) was defined when the first draining LN basin was the internal mammary area, regardless of delayed axillary LN visualization. Otherwise, the lymphoscintigraphy was categorized as negative IM-LPS. Second, the axillary lymphoscintigraphy (A-LPS) was visually graded into four groups regardless of internal mammary sentinel node visualization, according to the method suggested by Lee et al. (Fig. 1). ¹ Group I was defined when lymphoscintigraphy showed only focal sentinel LN activity at axillary area. Group II lymphoscintigraphy showed axillary sentinel LN activity with lymphatic flow activity to the axillary area. Group III showed only lymphatic channel to axillary area without definite axillary sentinel LN activity. In the case of no sentinel LN, activity was seen on lymphoscintigraphy, which was categorized as group IV.

OPEN IN VIEWER

FIG. 1.

The visual grading system of axillary lymphoscintigraphy pattern.

Chi-square test was used to compare proportions. The logistic regression was used to analyze the relationship between pathologic axillary LN metastasis and the following factors: the lymphoscintigraphy pattern, positive metastasis on F-18 FDG PET/CT, nuclear grade, tumor size, lymphovascular invasion, multifocal breast cancer, estrogen receptor (ER) status, and the number of negative sentinel LN (the number of negative sentinel LNs among dissected LNs). The tumor size and the number of negative sentinel LN were considered as continuous variables. The statistical analyses were performed using MedCalc® for windows version 12.0.4.0; p value less than 0.05 was regarded as significant.

Results

Discussion

The present study examined whether internal mammary visualization on preoperative lymphoscintigraphy with subareolar injection is associated with axillary LN metastasis in patients with breast cancer. We found that patients with internal mammary drainage have more axillary LN metastasis than those without internal mammary drainage.

A few previous studies reported a correlation between decreased LN uptake and increased risk of axillary involvement.^1,11,12 A reasonable explanation for this phenomenon would be that the lymphatics become progressively infiltrated with tumor cells, blocking the passage of radiotracers. ¹¹ This explanation was supported by a recent study conducted by Lee et al. ¹ The authors used the grading system that reflects the degree of impairment of lymphatic transport. They found that, as the grade increases, the radioactivity of the upstream lymphatic vessels increases while that of the sentinel LN decreases. The major explanation of this phenomenon was that a higher grade pattern is related to a greater lymphatic tumor burden, and this may be related to an increased likelihood of nonsentinel LN metastasis. Moreover, there is a case report about nondetection of sentinel LN with lymphoscintigraphy as a result of massive malignant invasion in a LN, in a patient with malignant melanoma. ¹³

We used a grading system similar to that of Lee et al. ¹ with the Tc-99m phytate as a radiotracer instead of Tc-99m human serum albumin. The Tc-99m phytate colloid can be formed in situ in the blood by interacting of the phytate ligand with calcium content in serum. ¹⁴ We found that the large number of patients showed axillary LN metastasis in higher grade of lymphoscintigraphy. However, in grade IV, the proportion was slightly decreased compared to that of group II and III. There may be a false negative rate secondary to shine through from injection site because we used only anterior view of image without lateral or oblique view.

This finding could be extended to the explanation of the correlation of positive IM-LPS with axillary LN metastasis. Among the patients with positive IM-LPS, 56.3% (18/32) of patients showed axillary LN metastasis, whereas 31.4% (124/369) of patients with negative IM-LPS were diagnosed with pathologic axillary LN metastasis. In general, internal mammary sentinel LN drainage is found in up to approximately 25% of patients during preoperative lymphoscintigraphy.^9,15,16 The internal mammary LN is usually visualized simultaneously with axillary LN, and the incidence of isolated lymphatic drainage to internal mammary LN has been estimated to be 1%–8%.^10,16–19 Comparable to previous reports, 8.0% of our patients showed internal mammary sentinel LN uptake on lymphoscintigraphy and isolated IM-LPS drainage was found in 3.24% (13/401) of patients.

The lymphatic drainage of the breast is poorly understood. Several investigators have shown discordant drainage patterns from injections of localizing agents into the subareolar plexus, subdermal breast tissue, and the deep breast parenchyma.^{18, 20–22} Generally it is shown that with deeper injections, a higher visualization rate of extra-axillary drainage is seen compared with superficial injections.^23,24 The intraparenchymal injections consistently visualize a more diverse pattern of lymph flow. Specifically, the internal mammary chain and supraclavicular nodes are commonly seen after intraparenchymal injection, and injection in different quadrants of the breast also show different patterns of localization. ²¹ In the present study, Tc-99m phytate was injected in the periareolar area in all patients interadermally, therefore we suppose that there was no influence on uptake in internal mammary basin on lymphoscintigraphy according to injection site.

The uptake of internal mammary LN on lymphoscintigraphy does not represent metastasis in the internal mammary LN basin. ⁹ The previous study also suggested that predominant drainage to the internal mammary LN area is not associated with internal mammary LN involvement and long-term outcome, even in the theoretical overestimation of internal mammary LN containing micro or macro metastasis at the time of diagnosis. ²⁵ In the present study, internal mammary LN metastasis was compared to F-18 FDG PET/CT results. F-18 FDG PET study is known to be effective in detecting metastasis in LN of patients with breast cancer (sensitivity 94%). ²⁶ We observed that only 2 patients with positive IM-LPS showed internal mammary LN metastasis on F-18 FDG PET/CT. In this case, the internal mammary visualization on lymphoscintigraphy could be supposed as sentinel LN. However, there were large metastatic tumor burden in axillary basin in 2 patients (Fig. 2). This finding supports the theory of an internal mammary pathway as an altered lymphatic flow result of interrupted axillary flow due to metastastatic tumor cells.

OPEN IN VIEWER

FIG. 2.

The patients with positive internal mammary lymphoscintigraphy showing large metastatic burden in ipsilateral axillary lymph node area on F-18 FDG PET/CT.

Recently, new imaging agents have been used to image the lymphatics and SLN with MRI and optical techniques, and new fluorescent nanoparticles such as upconverting nanocrystals are potential future agents. ²⁷ Also, in addition to these imaging agents, a bimodal nanoparticle contrast agent should be attempted for visualization of SLN on breast cancer in the future. ²⁸

Conclusion

This study suggests that IM-LPS and A-LPS patterns might be useful to show the burden of axillary LN metastasis in breast cancer patients, as an indicator of altered lymphatic pathways.