Since the introduction of the sentinel node (SN) procedure in breast cancer, many validation studies have confirmed the accuracy of the SN biopsy in predicting axillary node status.1–6 As compared with axillary lymph node dissection (ALND), SN biopsy causes less morbidity and provides the same staging information. As a consequence, ALND is nowadays abandoned in case of a negative SN biopsy, thereby avoiding unnecessary morbidity and costs associated with ALND.

However, sometimes the lymphatic mapping procedure fails because of insufficient or absent radioactive tracer uptake in the lymph nodes. Because a negative preoperative lymphoscintigraphy is predictive for failure of intraoperative SN identification, faint or nonvisualization of hotspots on lymphoscintigraphy (initial scan failure) frequently results in the need to perform ALND.7

After having experienced 14 patients with a negative preoperative lymphoscintigraphy in the first 3 years of our study, we started to use additional radiocolloid tracer injections as of May 2000, to avoid technical failure of the lymphatic mapping procedure.

The aim of this study is to determine the value of additional tracer injection to increase the technical success rate of the sentinel node procedure in case of a vague or negative lymphoscintigraphy and to identify factors that influence nonvisualization of hotspots.

Patients and Methods

From April 1997 to August 2007, after having received approval of the Local Ethical Committee, and after informed consent, a total of 1,208 consecutive patients with clinically node-negative operable primary breast cancer were included in a prospective study on SN biopsy. In phase I of this study (137 patients) SN biopsy was followed by completion axillary lymph node dissection in all cases. In phase II, after validation of the SN technique in our institute, completion axillary lymph node dissection was performed only in cases of tumor-positive axillary SN or unsuccessful SN procedure. From this ongoing prospective study we analyzed all consecutive patients who showed inadequate or absent visualization of hotspots on the initial lymphoscintigraphic images (Figs. 1–3).

Fig. 1
figure 1

Lymphoscintigraphy of patient with initial scan failure (a) and secondary scan failure after additional tracer injection (b)

Full size image

Our technique of SN biopsy has been described in detail elsewhere.2 The lymphatic mapping procedure consists of 370 MBq (10 mCi) TC-99 m-nanocolloid injected peritumorally or in the breast parenchyma surrounding the cavity of a previous excisional biopsy. All patients underwent preoperative lymphoscintigraphy following a mean interval of 16 h (range 12–18 h). In case of initial scan failure, additional radiocolloid tracer injections of 37 MBq (1 mCi) Tc-99 m-nanocolloid intracutaneously above the tumor were used as of May 2000, in order to increase the technical success rate of lymphoscintigraphy by obtaining secondary SN visualization. A second lymphoscintigraphy was performed 10–15 min after the additional tracer injection was given.

In relation to the mapping procedure, radiation exposure guidelines and doses were considered for both patients and personnel and were found to be well within legal safety limits, as published previously.2

In 2002 preoperative ultrasonography of the axilla was introduced in our hospital and became standard procedure in combination with fine-needle aspiration (FNA) cytology in case of clinically or radiologically suspicious axillary lymph nodes. Thus, patients treated before 2002 did not receive routine ultrasound investigation of the axilla.

During surgery, all axillary and nonaxillary SNs were pursued, as visualized by lymphoscintigraphy. Intraoperative identification of the SNs was based both on blue dye mapping and gamma probe detection. In all cases of a failed SN procedure, whether based on secondary scan failure or based on intraoperative SN identification failure, ALND was performed, which involved at least removal of all level I and II lymph nodes.

Histopathologic examination of the SN consisted of routine serial sectioning with hematoxylin and eosin (H&E) staining, followed by immunohistochemical (IHC) staining, whenever routine H&E staining did not reveal metastases.

Depending on the type of variable, chi-squared test or t-test was used to compare the characteristics of patients with successful lymphoscintigraphy and with initial scan failure.

Results

In all, 1,208 consecutive patients with clinically node-negative breast cancer were included in this study. Patient and tumor characteristics are listed in Table 1.

Table 1 Comparison of characteristics of patients (N = 1208) with successful lymphoscintigraphy and with initial scan failure
Full size table

Initial Scan Failure

In 93 patients, the results of preoperative lymphoscintigraphy were classified as initial scan failure: in 52 patients, because lymphoscintigraphy showed no axillary hotspots at all (negative lymphoscintigraphy), and in 41 patients because the initial SN visualization was considered inadequate (i.e., too faint to allow successful intraoperative SN retrieval). The mean age of these 93 patients was 65.8 years, compared with 58.5 years for the remaining group of 1,115 patients with clear hotspots on initial lymphoscintigraphy (p < 0.0001).

Within the first 3 years of the study 14 patients (group A, Table 2) with negative initial lymphoscintigraphy did not receive additional tracer injection and consequently ALND was performed in all cases. Of these 14 patients, 10 were found to have positive lymph nodes, with 5 of them having more than ten involved nodes. The remaining four patients had a negative SN procedure, using the blue dye technique, which was confirmed by completion ALND. No axillary recurrences were seen in these four patients within a median follow-up of 80 months.

Table 2 Patient and tumor characteristics of patients with initial scan failure (N = 93)
Full size table

As of May 2000, in case of initial scan failure, an additional tracer injection of 1 mCi Tc-99 m-nanocolloid intracutaneously was routinely used.

Secondary Scan Failure

In five patients (group B, Table 2) with initial scan failure, following additional tracer injection, repeated lymphoscintigraphy still did not show hotspots (Fig. 1a, b). Mean age of these patients was 66.2 years. Successful SN harvesting was possible in only one patient, because a (tumor-positive) palpable lymph node was found intraoperatively.

Axillary dissection was done in all five patients and showed a tumor-positive axilla in four (80%), of whom three patients showed massive tumor infiltration of almost all axillary lymph nodes. The only node-negative patient in this subgroup had no axillary recurrence during a follow-up period of 49 months.

Secondary SN Visualization

In 33 patients (group C, Table 2) with initial scan failure, we noted secondary SN visualization on repeated lymphoscintigraphy as a direct result of additional tracer injection (Fig. 2a, b). In these patients mean age was 64.7 years. Secondary lymphoscintigraphy clearly showed one or more axillary hotspots, which subsequently could be harvested in all cases, showing a positive SN in 17 patients (52%), all of whom underwent ALND. Massive nodal tumor burden, with ten or more tumor-positive nodes, was present in three patients only (9%). In 15 patients the SN was found to be tumor-negative and no ALND was performed. No axillary recurrences were noted during a median follow-up of 27 months in these patients. One patient (5.5%, 1/18) in this subgroup had a false-negative SN procedure. A tumor-positive lymph node was found during subsequent simple mastectomy. She was given axillary radiotherapy. No axillary recurrence was seen thereafter within a 60-month follow-up period.

Fig. 2
figure 2

Lymphoscintigraphy of patient with initial scan failure (a), but secondary SN visualization after additional tracer injection (b)

Full size image

Enhanced Lymphoscintigraphy

In 41 patients (group D, Table 2) with initial scan failure, in which initial lymphoscintigraphy showed only a faint hotspot, judged to be insufficient to allow successful SN harvesting, additional tracer injection led to increased radioactive uptake in the same hotspot (Fig. 3a, b), rather than visualizing additional hotspots. The mean age of these patients was 67.0 years. Additional hotspots as a result of additional tracer injection were seen in only five patients (12%). The increased radioactive uptake facilitated SN harvesting, which was successful in all cases. A tumor-positive SN was found in 16 patients (39%), which was not different from the group of 1,115 patients with clear hotspots on initial lymphoscintigraphy (39%). Only one patient (2.4%) had massive tumor infiltration in the axilla, showing all axillary lymph nodes to be tumor-positive. No completion ALND was performed in the remaining 25 patients with a negative SN biopsy. Among these 25 patients again no axillary recurrences were found during a median follow-up of 41 months.

Fig. 3
figure 3

Lymphoscintigraphy of patient with faint hotspot (a), but clear hotspot after additional tracer injection (b)

Full size image

Discussion

Because of its high sensitivity to detect nodal metastatic disease and its minimally invasive nature, SN biopsy has become the standard of care for staging early invasive breast cancer, thereby limiting axillary dissection to patients with axillary metastases and sparing node-negative patients the morbidity of axillary dissection. However, sometimes the lymphatic mapping procedure tends to fail, because of inadequate or even absent radioisotope uptake in the SN. This urges the surgeon to perform a complete axillary lymph node dissection, which might have been avoided had the lymphatic mapping been successful.

After having experienced 14 patients with a negative preoperative lymphoscintigraphy in the first 3 years of our study, we started to use additional radiocolloid tracer injections as of May 2000, to avoid technical failure of the lymphatic mapping procedure. The use of additional tracer injection has been previously reported by Cserni et al. in 20 patients.8

Several factors have been identified to influence lymphatic mapping. Besides technical factors, such as tracer volume, tracer dose, site of tracer injection, and timing of lymphoscintigraphic imaging, other factors such as prior breast surgery and upper lateral tumor location, which might hamper lymphatic mapping due to shine through, have been reported to influence lymphatic drainage patterns. However, there are three major reasons for initial scan failure: extensive nodal tumor infiltration, increased age, and increased body mass index (BMI).

Extensive Nodal Tumor Infiltration

Axillary hotspot visualization is negatively influenced by extensive nodal involvement.2,9–14 This can be explained by nodal tracer uptake physiology: the radioactive tracer is bound to colloid, which is phagocytosed by macrophages within the normal lymph tissue of the SN. If the SN, or the afferent lymph tracts, show massive tumor infiltration and/or extranodal growth, the lymph flow might be blocked, or there might not be enough functional lymph tissue left to phagocytose the radiocolloid. Thus, if a SN is completely replaced by tumor or the afferent lymph tract is blocked by extensive lymphatic tract infiltration, initial lymphoscintigraphy can fail to show any hotspot. Additional tracer injection under these circumstances might result in secondary scan failure, in which case ALND is mandatory.

Lymphatic tumor burden does not seem to affect blue dye uptake because, in contrast to the uptake of radiocolloid, which is based on sufficient functional lymph tissue, blue dye uptake is a process of passive diffusion through the lymphatic system.11

Massive nodal tumor and/or lymphatic tract infiltration was present in 57.1% of 14 patients (group A) early in our study, in which a negative initial lymphoscintigraphy was not followed by additional tracer injection, and in 60% of 5 patients (group B), in whom additional tracer injection, after initial scan failure, did not result in secondary SN visualization.

If additional tracer injection does result in secondary SN visualization, as was seen in 33 patients (group C) in our study, it can be argued whether this visualized hotspot is in fact the true SN, or might be an alternative lymph node which, based on rerouting of lymph flow, is erroneously considered to be the true SN.13,15,16 If so, rerouting of lymph flow during lymphatic mapping would ultimately result in increased false-negative rates.17

However, of 33 patients with secondary SN visualization resulting from additional tracer injection, only 1 patient (5.5%) had a false-negative SN biopsy, whereas 15 patients with a negative SN after additional tracer injection, who consequently had no ALND, showed no axillary recurrences within 29 months of follow-up. These results suggest that SN biopsy after additional tracer injection does not compromise its accuracy, as was also pointed out by others.18

Internal Mammary (IM) Basin Drainage

The incidence of IM metastasis is correlated with axillary nodal involvement.19 It seems intuitive that a large axillary tumor burden can result in blockage of lymph flow to the axilla, consequently leading to redirection of lymph flow to the IM chain.13 However our results, like those of other investigators, show that extensive nodal infiltration of the axilla is not correlated with increased lymph drainage to the IM lymph nodes, since IM hotspots were noted in only 5.4% (5/93) of patients with initial scan failure, versus 21.2% in patients with a successful initial lymphoscintigraphy (Table 1).11 This seemingly unexpected result is probably explained by the fact that patients with initial scan failure had a significantly higher mean age and BMI, both of which are inversely correlated to IM lymph drainage.

Palpable Lymph Nodes

An axillary lymph node with gross tumor involvement might consequently be enlarged, and thus is more likely to be detected by routine physical examination and/or axillary ultrasound.

However, because a tumor-loaded lymph node also can be the same size as a tumor-negative lymph node, physical examination of the axillary lymph nodes in order to detect nodal metastases has been shown to be inaccurate.20,21 Lymph nodes that can be palpated intraoperatively through the incision of the SN biopsy and are felt to be suspicious should be harvested and regarded as SN.2,22,23

Axillary Ultrasound

Routine preoperative axillary ultrasound is advocated by several investigators.9,20,24 Combined with ultrasound-guided FNA of enlarged or suspicious lymph nodes, preoperative ultrasound can detect metastatic disease, hence avoiding an unnecessary or potentially unsuccessful SN biopsy.

In this study, routine preoperative ultrasonography of the axilla, which was introduced in our institute in 2002, was performed in only 43 of the 93 patients with initial scan failure. Moreover, of 48 patients with a tumor-positive axilla in this subgroup, only 15 patients had a preoperative ultrasound examination of the axilla. In one of these patients, ultrasound showed a suspicious lymph node, which was tumor-negative on FNA. Among 15 patients who showed extensive nodal involvement or extranodal tract invasion in relation to initial scan failure (Table 2), preoperative axillary ultrasound was performed in only two patients.

Thus, within this study we were not able to define the diagnostic yield of preoperative axillary ultrasound in relation to initial scan failure.

However, in retrospect, it seems clear that a substantial number of our patients with initial scan failure would not have needed a lymphatic mapping procedure had routine preoperative axillary ultrasonography been introduced earlier in our institute.

Increased Age

With the loss of estrogen levels in postmenopausal women, the breast parenchyma is partially replaced by fatty tissue. This process of fatty degeneration results in a substantial decrease of lymphatic capillaries, normally confined to the breast parenchyma, which in turn is correlated with decreased radioactive uptake in the SN. Thus, increased age is inversely correlated with the ability to visualize axillary hotspots and internal mammary hotspots.13,25–29

Likewise, our study showed that the mean age of patients with initial scan failure was significantly higher when compared with patients with a successful initial lymphoscintigraphy (65.8 years versus 58.5 years, p < 0.0001).

The inverse effect of increased age on the technical success rate of lymphoscintigraphy becomes even more obvious when looking at the results of 41 patients (group D) in our study with initial scan failure showing only very faint hotspots (Fig. 3), necessitating additional tracer injection for successful SN retrieval. In this subgroup massive nodal tumor infiltration was nearly absent (one patient) and therefore no causal factor for initial scan failure. Subsequently, the increased age in this subgroup (mean 67 years)is the main factor, besides BMI, accountable for non-visualization of hotspots (Table 2).

Body Mass Index (BMI)

BMI is also recognized as a factor that affects lymphoscintigraphic visualization of sentinel nodes.26,27,30 As is true for increased age, patients with high BMI might have sparse lymphatic capillaries and more fatty tissue in their breast parenchyma.

In our study, mean BMI was 29.0 kg/m2 for the patients with initial scan failure and 25.5 kg/m2 for a random sample of 100 patients with clearly visible hotspots on the initial scan. The difference was highly significant (p < 0.0001). This result confirms the findings from other investigators regarding the effect of BMI on the success rate of lymphoscintigraphy.

Previous Excisional Biopsy

Previous excisional biopsy is reported to result in change of lymphatic drainage patterns and scan failure.4,13,31,32 In a previous study, we reported the results of 88 patients, in which SN biopsy with completion ALND was performed following previous excisional biopsy.33 Initial scan failure was seen in only four patients (4.5%) and no false-negative SN procedures were noted. In the present study 199 patients had a previous excisional biopsy. Of these, 8.0% showed initial scan failure as compared with 7.6% in patients in whom breast cancer was diagnosed by core biopsy (Table 1). Therefore, in our opinion, excisional biopsy prior to SN biopsy is not correlated with scan failure.

Conclusion

Sentinel node (SN) biopsy has become the standard of care in the treatment of breast cancer. To prevent unnecessary ALND, the results of preoperative lymphoscintigraphy should be optimal. To avoid technical failure of lymphatic mapping, we evaluated the results and accuracy of additional intracutaneous radiocolloid tracer injections of 37 MBq (1 mCi) in patients with initial scan failure. A negative initial lymphoscintigraphy, if followed by additional tracer injection, resulted in secondary SN visualization in 87% of these patients, thus enabling accurate (false-negative rate 5.5%) SN biopsy. Both age and extensive nodal tumor infiltration as well as BMI adversely affect SN visualization.

In case additional tracer injection does not result in (secondary) SN visualization, ALND is mandatory, because of the high risk of positive lymph nodes, many times with massive tumor infiltration of the axilla.

If initial SN visualization is faint, this is mainly correlated with increased age and high BMI, rather than large tumor burden. In these patients additional tracer injection facilitates an accurate SN procedure.