• sentinel lymph nodes;
  • breast cancer;
  • lymphoscintigraphy;
  • internal mammary nodes


  1. Top of page
  2. Abstract
  6. Acknowledgements


Radiation to the internal mammary chain (IMC) may be indicated for breast carcinoma patients with positive axillary sentinel lymph nodes (SLNs) and lymphoscintigraphic evidence of drainage to the IMC. The purpose of this study was to identify predictors of IMC drainage in patients with positive axillary SLNs.


The records of 297 breast carcinoma patients with positive axillary SLNs and preoperative lymphoscintigraphy were reviewed between 1995 and 2002. Radiolabeled colloid was injected peritumorally with lymphoscintigraphy performed 30–60 minutes later. Drainage to the regional nodes of 279 patients was seen on lymphoscintigraphy. Associations among patient and tumor-related factors and drainage to the IMC were examined.


Drainage to the IMC on lymphoscintigraphy was seen in 63 patients (21%). IMC drainage only occurred in 4 patients, and 59 patients had both axillary and IMC drainage. The only variable that correlated with IMC drainage was tumor location (P = 0.017). Rates of drainage to the IMC were 14.1% for upper outer quadrant (n = 128), 16.7% for upper inner quadrant (n = 30), 31.6% for lower outer quadrant (n = 19), 42.9% for lower inner quadrant (n = 14), and 28.4% for central tumors (n = 88). IMC drainage rates differed significantly between upper and lower tumors (lower 36.4% vs. central 28.4% vs. upper 14.6%, P = 0.003) but not between medial and lateral tumors (medial 25.0% vs. central 28.4% vs. lateral 16.3%, P = 0.077).


Patients with tumors in the lower or central breast and positive axillary SLNs have increased incidence of drainage to the IMC. Preoperative lymphoscintigraphy can help to define the nodal basins at risk for harboring disease. Cancer 2005. © 2005 American Cancer Society.

It has been estimated that 30% of breast carcinoma patients with metastatic disease in the axillary lymph nodes also have disease involvement of the internal mammary chain (IMC).1 It is intuitive that patients with disease in this region may benefit from therapy directed to the IMC. However, surgical dissection of the IMC is not routinely performed because of the potential for significant morbidity and the failure of trials to demonstrate its efficacy.2 As an alternative to surgery, radiation treatment of the IMC has been studied, with mixed results. Some investigators have found that patients with medial or central tumors with positive axillary lymph nodes who undergo irradiation of the IMC have an increased survival rate and decreased tumor relapse rate compared with similar patients without IMC treatment.3–7 Other studies have demonstrated a lesser benefit or no benefit to irradiation of the IMC. Thus, irradiation of the IMC is currently an area of controversy.

One potential reason for mixed results concerning the value of IMC irradiation is that it is difficult to clearly define populations of breast carcinoma patients at highest risk for harboring IMC disease. Use of sonography or computed tomography to identify metastases within the IMC has not proved successful because these techniques lack sensitivity.8 Newer scanning techniques such as positron emission tomography also have limitations in detecting small volumes of microscopic disease. One recent advance in identifying patients at risk of IMC involvement is the use of lymphoscintigraphy in conjunction with sentinel lymph node (SLN) biopsy.9–11 Specifically, patients who have drainage to the IMC observed on lymphoscintigraphy and who are found to have a positive axillary SLN would be predicted to have a clinically relevant risk of disease in the IMC. Indeed, this hypothesis has been confirmed by a number of small studies in which IMC SLNs have been surgically removed.10, 12

Irradiation of the IMC increases the volume of normal lung and contralateral breast within the treatment fields. Therefore, irradiation of the IMC should be offered only to patients with a clinically relevant risk of disease in this region. As mentioned, one cohort for whom IMC irradiation is reasonable is a patient group with positive axillary SLNs who also have dual drainage to the axilla and IMC demonstrated on preoperative lymphoscintigraphy. However, many medical centers perform SLN surgery without using lymphoscintigraphy.

The purpose of this investigation was to identify predictors of IMC drainage in patients with positive axillary SLNs. These data may be particularly useful in the design of radiation treatment fields at medical centers where lymphoscintigraphy is not routinely performed.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Patient Population

The authors of the current study reviewed the prospective SLN database at The University of Texas M. D. Anderson Cancer Center. This database includes records for all patients who have undergone technetium-99-labeled sulfur colloid lymphoscintigraphy scans within our institution. The Surveillance Committee (institutional review board) of our institution has approved this database and associated retrospective studies. The current study limited its review to patients who had preoperative lymphoscintigraphy, invasive breast carcinoma, and at least 1 positive axillary SLN, and who subsequently underwent a completion axillary lymph node dissection between 1995 and 2002. A total of 297 such patients was identified.

The endpoint of the current study was the presence of drainage to the IMC on preoperative lymphoscintigraphy. The authors investigated whether IMC drainage correlated with the following tumor characteristics: size, breast quadrant location, histologic type, modified Black nuclear grade, lymphovascular space invasion, estrogen and progesterone receptor status, HER-2/neu status, and clinical stage. Estrogen and progesterone receptor status were determined by immunohistochemical staining. HER-2/neu positivity was defined as either 3+ immunohistochemical staining or gene amplification noted on fluorescence in situ hybridization. Also investigated were the relation between IMC drainage and the number of positive SLNs, the amount of disease within positive SLNs, and the presence or absence of extranodal extension in positive SLNs.

Preoperative Lymphoscintigraphy

All patients underwent lymphoscintigraphy. A nuclear medicine physician injected 0.5–2.5 mCi of technetium-99m-labeled sulfur colloid into the breast parenchyma around the primary tumor. If a mass was not palpable, the injection was performed under mammographic or sonographic guidance. Planar images of the upper chest were obtained in multiple projections demonstrating tracer activity in regional lymph node areas. The first images were taken 30 minutes to 1 hour after injection of radiocolloid by placing the patient under a scintillation camera for 2–6 minutes. If no drainage was seen, images were obtained periodically up to 5 hours after injection of the radiocolloid.

Pathologic Evaluation of SLNs

The pathologic evaluation of SLNs at M. D. Anderson Cancer Center has been described elsewhere.13 Briefly, through the end of April 2000, SLNs were serially sectioned along their short axis at 2-mm to 3-mm intervals. The sections were embedded in paraffin blocks, and one level from each block was stained with hematoxylin and eosin. After April 2000, SLNs were macroscopically processed in a similar manner; however, each paraffin block was serially sectioned at 5-μm intervals to produce 10 levels. Levels 1, 2, and 4 were evaluated by routine hematoxylin-eosin staining, and Level 3 was analyzed for cytokeratin by immunohistochemical staining. Lymph nodes were considered positive if malignant cells were detected on either the hematoxylin-eosin–stained section or the immunohistochemically stained section. Single cytokeratin-positive cells were reported to be cytokeratin-positive cells of “uncertain biological significance” and were considered negative.

Statistical Methods

Descriptive statistics were used to assess the frequency distribution among the study population. Univariate analysis with the Chi-square test was used to evaluate whether associations existed between drainage to the IMC and the study variables. All analyses used a 2-tailed significance level of .05 and were performed using the SPSS 11.5 software package (SPSS Inc., Chicago, IL).


  1. Top of page
  2. Abstract
  6. Acknowledgements

Patient and Primary Tumor Data

The study population included 297 breast carcinoma patients with at least 1 positive SLN. Of these patients, 279 had drainage to the regional lymph node basins seen on lymphoscintigraphy. Dual drainage to the IMC and axilla was seen in 59 patients, and 4 patients had drainage to the IMC only.

The median age of patients was 53 years (range, 29–88 yrs). The most common tumor location was the upper outer quadrant (46%), and median tumor size was 1.7 cm (range, 0.1–9 cm). Lymphovascular space invasion was identified in 29% of tumors. The majority of tumors were estrogen receptor positive (80%), progesterone receptor positive (64%), and HER-2/neu negative (111 tumors of 140 in which HER-2/neu status was evaluated). Neoadjuvant chemotherapy was administered to 23% of patients. The majority of patients (68%) had only 1 positive SLN, and, in 49% of patients, the maximum size of the largest metastasis was between 2.1 and 10 mm. Extranodal extension was present in 23% of patients with positive SLNs.

Correlation between patient and tumor-related variables and drainage to the IMC.

Of 279 patients that had a clearly identified tumor location and drainage to the regional lymph node basins seen on lymphoscintigraphy, 63 had drainage to the IMC nodes. The univariate analyses of how the study variables correlated with drainage to the IMC are summarized in Table 1.

Table 1. Univariate Analyses of Multiple Potential Predictors of Drainage to Internal Mammary Chain (IMC) Lymph Nodes
VariableNo. (%) of patientsRate of drainage to IMC (%)P
  • MBNG: modified Black nuclear grade; LVSI: lymphovascular space invasion; ER: estrogen receptor; PR: progesterone receptor; SLN: sentinel lymph node; ENE: extranodal extension.

  • a

    Tumor size was not reported in 2 cases.

  • b

    MBNG was not reported in 2 cases.

  • c

    LVSI was not reported in 32 cases.

  • d

    ER status was not reported in 2 cases.

  • e

    PR status was not reported in 3 cases.

  • f

    HER-2/neu was not reported in 157 cases.

  • g

    Size of metastases was not reported in 15 cases.

  • h

    Amount of ENE was not reported in 25 cases.

Tumor quadrant location  0.017
 Upper outer128 (46)14.1 
 Upper inner30 (11)16.7 
 Lower outer19 (7)31.6 
 Lower inner14 (5)42.9 
 Central88 (31)28.4 
Tumor Sizea  0.241
 0–2 cm196 (66)23 
 2.1–5 cm90 (31)20 
 >5 cm9 (3)0 
Clinical stage  0.192
 IIA164 (55)23.8 
 IIB108 (37)20.4 
 IIIA15 (5)0.0 
 IIIB10 (3)20.0 
MBNGb  0.370
 I29 (10)31.0 
 II158 (53)20.9 
 III108 (37)18.5 
LVSIc  0.000
 Yes76 (29)22.4 
 No189 (71)5.8 
ER statusd  0.605
 Positive239 (81)20.9 
 Negative56 (19)21.4 
PR statuse  0.724
 Positive190 (65)20.0 
 Negative104 (35)23.1 
Her-2/neu statusf  0.238
 Positive29 (21)13.8 
 Negative111 (79)18.0 
No. of positive SLN(s)  0.285
 1201 (68)23.9 
 268 (23)20.6 
 317 (6)5.9 
 47 (2)0.0 
 >54 (1)0.0 
Size of metastasesg  0.146
 0–2 mm106 (38)22.6 
 2.1–10 mm147 (52)22.4 
 >10 mm29 (10)6.9 
Amount of ENEh  0.795
 0209 (77)23.0 
 0.1–1.9 mm46 (17)19.6 
 ≥2 mm17 (6)17.6 

There was a significant association between drainage to the IMC nodes and tumor location (P = 0.017). The other factors analyzed did not significantly correlate with drainage to the IMC.

Table 2 shows results of analyses with two different groupings by tumor location. Differences in IMC drainage rates were significant when tumors were grouped as lower, central, or upper. However, when tumors were grouped on the basis of medial versus lateral tumor location, differences in IMC drainage rates were not significant.

Table 2. Relation between Tumor Location and Drainage to Internal Mammary Chain (IMC) Lymph Nodes
VariableNo. (%) of patientsRate of drainage to IMC (%)P
Subgrouping 1  0.003
 Lower quadrants33 (12)36.4 
 Central breast88 (31)28.4 
 Upper quadrants158 (57)14.6 
Subgrouping 2  0.077
 Medial quadrants44 (16)25.0 
 Central breast88 (31)28.4 
 Lateral quadrants147 (53)16.3 


  1. Top of page
  2. Abstract
  6. Acknowledgements

The status of internal mammary lymph nodes on pathologic evaluation is an important determinant of prognosis for patients with breast carcinoma.14, 15 Currently, there is no ideal diagnostic method to evaluate the status of the internal mammary nodes — most current imaging modalities have low sensitivity in detecting small deposits of metastatic disease. The low sensitivity of computed tomography and sonography in determining which patients have IMC involvement contributes to ambiguity over which subgroup of patients may benefit from treatment of the IMC. Preoperative lymphoscintigraphy may be an improved method of selecting patients for IMC treatment, in that it helps to discriminate patients whose tumors drain solely to the axilla from those whose tumors also drain to the IMC. However, in a number of medical centers that perform SLN surgery, lymphoscintigraphy is not performed. When it is used, lymphoscintigraphy has been shown to successfully identify the lymphatic drainage patterns in up to 90% of breast carcinoma patients.16 In addition, lymphoscintigraphy identifies drainage to the IMC in 15–44% of patients, depending on the location of the primary tumor and the site of injection of the isotope. (Data are summarized in Table 3.)

Table 3. Tumor Location and Frequency of Drainage to the Internal Mammary Chain (IMC) Nodes on Lymphoscintigraphy
AuthorTumor locationNo. of casesRate of drainage to IMC (%)
Shahar (current study)Total27922.5
 Inner quadrant/central10524
 Inner quadrant/central3212

The purpose of the current study was to identify patients at high risk for pathologically involved internal mammary nodes and a subset of these patients who may benefit from treatment of the IMC. Authors of the current study intentionally examined patients with positive axillary lymph nodes as studies have shown an increased risk of disease involvement of the IMC in patients with positive axillary nodes compared with patients with negative axillary nodes.12, 17, 18 The current study, in agreement with others, suggests that the location of the tumor is a significant predictor of lymphatic drainage to the IMC. In the current study, the rate of drainage to the IMC was higher for lower tumors (36.4%) and central tumors (28.4%) than for upper tumors (14.6%). Other studies that used peritumoral radiocolloid injection and lymphoscintigraphy found similar results (Table 3); however, the majority report an increased risk of drainage from inner-quadrant and centrally located tumors rather than from lower tumors. The reason for this discrepancy may be different patient characteristics and experimental methods.

The overall identification rate of axillary SLNs on lymphoscintigraphy is as high as 88–90%.14, 19 Although drainage to the IMC was present in 23% of patients, the current study authors did not sample the IMC; thus, the current study documents only drainage to the IMC, not metastasis to the IMC. Biopsy of the IMC is more difficult than axillary node biopsy and usually requires an extra incision in patients undergoing breast conservation therapy. This carries the risk of pain, bleeding, and pneumothorax from entry into the pleura. Klauber-DeMore et al.20 reviewed seven studies that included IMC lymphatic mapping with lymphoscintigraphy and biopsy of the internal mammary SLNs on the basis of lymphoscintigraphy. These studies, combined with 3 others (summarized in Table 4), show that 21% of patients who underwent internal mammary SLN biopsy on the basis of lymphoscintigraphy findings had metastases in the IMC nodes.

Table 4. Results of Lymphoscintigraphy and Sentinel Lymph Node Biopsy of the Internal Mammary Chain (IMC)
AuthorYrType of injectionTime to scanNo. of ptsPts with IMC drainage No. (%)Percentage rate of pathologic IMC involvement % (No. pts)
  1. pts: patients; TAC: technetium 99m-human albumin colloid; TSC: technetium 99m-sulfur colloid; TNC: technetium 99m-nanocolloid; TTC: technetium 99m-tin colloid; PT: peritumoral; ID: intradermal; NA: not available.

Snider3819981 mCi TSC PT45–310 min481 (2)0 (0/1)
van der Ent39199910 mCi TAC PT + blue dyeNA7024 (34)33 (5/15)
Winchester4019991 mCi TSC PT1–4 hr, 80 pts; 16–20 hr, 100 pts18020 (11)15 (3/20)
Jansen41200040–60 MBq TAC PT30 min and 4 hr11317 (15)18 (3/17)
Johnson3620001 mCi TSC PT2 hr8010 (12.5)30 (3/10)
Noguchi4220003 mCi TSC IP + 4 mL blue dyeNA415 (12)20 (1/5)
Zurrida4320005–10 MBq TAC PT (ID if tumor superficial)20 min & 3 hr37615 (4)7 (1/15)
Van der Ent44200110 mCi TNC PT16 hr (range 12–18 hr)25665 (25.3)27 (11/41)
Carcoforo10200210 MBq TNC PT17 hr14327 (18.9)25 (4/20)
Ogawa4520031.5 mCi TTC PT4 hr4915 (31)18 (2/11)
Total   1356199 (14.7)21.3 (33/155)

The majority of breast tumors drain to the axillary nodes, either alone or in combination with other sites. Exclusive drainage to the IMC is relatively rare. In the current study, only four patients had drainage to the IMC without evidence of axillary drainage. Other studies have reported exclusive internal mammary lymphatic drainage in 3–7% of patients.10

Lymphoscintigraphy has been used in breast carcinoma treatment for many years but was more recently of interest to aid surgeons in selecting the SLN.21 Some studies state that lymphoscintigraphy adds little information to that obtained with intraoperative lymphatic mapping and that its routine use is not justified.22–24 However, lymphoscintigraphy scans provide information that can be used by other physicians who are also treating these patients. As a result of the lymphoscintigraphy findings, patient prognosis may be altered, and the subsequent therapeutic approach may be modified. This is particularly true when SLNs are identified in extraaxillary sites, such as the IMC. Estourgie et al. reported that 22 (17%) of 130 patients with IMC drainage had tumor positive nodes when internal mammary sentinel nodes were harvested as part of the surgical procedure.25 Stage migration was seen in all patients who had tumor positive IMC nodes, and there was a change in management reported in 38 (29%) of 130 patients. Of particular interest in patient management is that radiation oncologists can define radiation portals on the basis of the route of lymphatic drainage. At our institution, we consider irradiation of the IMC in patients with evidence of drainage to the IMC seen on lymphoscintigraphy who also have positive SLNs in the axillary basin.

Although there is no standard therapy for disease involving the IMC, randomized studies showed that radiation therapy is of benefit to high-risk patients, specifically patients with inner-quadrant or centrally located tumors.26–29 Stemmer et al.30 found that inclusion of the IMC in the radiation field after surgical resection and high-dose chemotherapy was associated with increased disease-free survival and a trend toward increased overall survival in high-risk breast carcinoma patients. The benefits and risks of IMC irradiation are currently under investigation by the Radiotherapy and Breast Cancer Groups of the European Organization for Research and Treatment of Cancer. These groups initiated a large Phase III randomized multicenter trial to investigate the value of adjuvant irradiation of the internal mammary and medial supraclavicular lymph node chains in patients with Stages I–III localized breast carcinoma with medially or centrally located tumors and/or axillary lymph node involvement.

A number of limitations can cause inconsistency in lymphoscintigraphic results. First, the radiocolloid and vital blue dye injection needs to be of sufficiently large volume around the primary site so that breast lymphatics are able to transport the mapping agents. The radiocolloid should not be injected into the tumor because high intratumoral pressure makes it difficult to inject a sufficient amount of radiocolloid. Second, mapping agent type and administration technique are critical for successful lymphatic mapping. The ideal radiocolloid would have a particle size and stability allowing rapid migration into the lymphatics with deposition and concentration in the SLN. Third, variations in the injection technique can produce mixed results. Paganelli31 found that an SLN was observed in the IMC in 65.6% of patients with inner-quadrant tumors after deep injection and in 2.1% of patients after superficial injection. Although the injection method does not influence the ability to identify an axillary SLN, intraparenchymal injection may be better for internal mammary SLN identification.

In conclusion, lymphoscintigraphy is a safe and valuable tool for assessment of lymphatic drainage patterns in patients with breast carcinoma. Because some centers do not perform lymphoscintigraphy, it is useful to know which factors predict drainage to the IMC so that the subset of patients at risk can be treated. The current study found that patients with a positive axillary SLN, lower-quadrant and centrally located tumors were associated with the greatest risk for drainage to the IMC. Elective irradiation of the IMC in breast carcinoma patients with lower-quadrant or central tumors and pathologically proven axillary lymph node involvement is a reasonable treatment option.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors are grateful to Stephanie Deming for her critical reading of the manuscript.


  1. Top of page
  2. Abstract
  6. Acknowledgements