TO THE EDITOR: The article by Pelosi et al. (1) supports our previously published results (2) and those of others concerning the superiority of areolar type injections over subdermal/intradermal injections, as well as perilesional injections, in delivering activity to the sentinel node. A few additional points warrant mention.
Having dynamically monitored the results of 3 different injection methods performed sequentially on the same patient in a single imaging session, we noted that the efficiency of delivering activity to the sentinel node, that is, the percentage of the injected dose that appears to end up in the sentinel node, is much higher (>3–6 times higher) for areolar-cutaneous “junction” injections than for intradermal injections above the tumor (2,3). This superiority is even more dramatic when compared with perilesional injections: The activity delivered to the sentinel node is up to 50–100 times higher by areolar-cutaneous “junction” injections than by perilesional injections in select patients (2,3).
Most of the literature comparing areolar injections with dermal or perilesional injections is unclear on the exact details of areolar injection methods. Terms such as subareolar, periareolar, circumareolar, or just areolar are used. Exact location, depth of injection, and other factors are not sufficiently detailed in many articles, making comparisons or exact reproduction difficult. Questions about injecting into lactiferous ducts are not raised. Furthermore, the investigators do not generally quantify the efficiency of the various injection techniques, as we did (2). Upon review, most of these articles do not demonstrate either in figures or in numeric data an effect as dramatic as that of the areolar-cutaneous “junction” injection technique in delivering activity to the sentinel node, or the higher efficiency given the higher injection doses generally used. A similarly high efficiency is suggested in Figure 2 of Pelosi et al. (1) and probably reflects the similarity of their technique to our areolar-cutaneous “junction” injections as depicted in Figure 1.
Hybrid injection techniques, defined as combinations of perilesional and dermal or areolar injections, provide an option for centers wishing to visualize internal mammary and other extraaxillary nodes. The deep perilesional injection component of these hybrid injection techniques allows visualization of internal mammary and other extraaxillary nodes, whereas the areolar-cutaneous junction injection component provides the ability to generate extremely hot sentinel nodes, with their associated benefits (2–5).
When performing areolar injections, one should not overlook the unique image patterns that are produced. One occasionally sees patterns of lymphatic dilation that appear as focal concentrations of activity, or “pseudosentinel nodes,” which can persist for some time after injection. If not properly identified as pseudosentinel nodes by lymphoscintigraphy before the start of surgery, these could potentially mislead the surgeon, unnecessarily prolonging surgery and causing fruitless searching (2,3,6).
Another pattern, the “reverse echelon node,” has rarely been noted. Its presence requires that at least 2 hot nodes be removed to avoid potentially missing the true sentinel node (2,3,7). The reverse echelon node is along the lymphatic channel tributary supplied by the areolar injection—upstream of the point at which that tributary joins the main channel draining the perilesional injection site to the sentinel node. The reverse echelon node, because it is on its own tributary, is closer to and only drained to by the areolar injection. In contrast, the perilesional injection is initially on a different tributary before all tributaries merge to a common channel to the sentinel node. Nevertheless, the sentinel node visualized by the perilesional injection, downstream in a sense, was also always a node draining the areolar injection in our series (2). Theoretically, both nodes should be removed for maximal sensitivity.
Such pattern analysis is possible only if real-time monitoring of the serial imaging results of sequential, dissimilar injection methods is performed in a single imaging session (2,3). Unfortunately, this type of pattern analysis, along with attempts to quantify the results of the different injection methods, is nearly completely lacking in the literature.
Whether the injections have to be performed at the areolar-cutaneous junction site closest to the tumor or can be performed equally well at any other site around the areola is not clear. We choose to inject at the junction site closest to the tumor from a pragmatic standpoint, as probably did Pelosi et al. (1). Using deeper, higher-volume subareolar injections, Kern at times noted multiple pathways simultaneously exiting the areolar area, but most seemed to converge (8). In another, more recent, article, Maza et al. noted a 0% false-negative sentinel node rate as evidenced by follow-up axillary lymph node dissection in patients with disease proven by lumpectomy or core-needle breast biopsy who received subareolar injections at 8 or more sites around the areolar margin during lymphoscintigraphy (9). Drainage from the different areolar injection methods probably leads to the same sentinel nodes. However, their efficiency in doing so is different.
By far the main goal of sentinel lymph node biopsy, as compared with traditional axillary dissection, is to reduce morbidity while maintaining sensitivity. The hotter sentinel node provided by the areolar-cutaneous “junction” injections (and similar injections described by Pelosi et al. (1)) assists in morbidity reduction for several reasons. It allows easier detection with the handheld probe at the skin surface, which should assist with the targeted approach by allowing a more direct path to the sentinel node target. This should reduce morbidity through a reduction in the extent of dissection. With a hot node, triangulated skin marking is facilitated, which can also guide the surgeon in determining where to make the initial incision, especially important in obese patients. In obese patients, a hot node offsets the negative effects of attenuation. In patients scheduled for surgery the day after lymphoscintigraphy, the negative effects of decay can be offset with a hotter node from the start.
Nevertheless, given the prominent lymphatic tracks that can arise with the areolar injection techniques we have noted here, we suggest dynamic monitoring and multiple views, including the standing/sitting position, to best map out what is really happening in the patient. In our opinion, not striving for a hotter node and not performing lymphoscintigraphy with triangulated skin markings, but simply depending on intraoperative probe detection alone, as practiced by some centers, goes against the very goal of morbidity reduction that sentinel lymph node biopsy promises.
REFERENCES
REPLY: Our study (1) validated the periareolar (PA) injection technique and underlined some of its reported advantages over the subdermal/peritumoral technique. In particular, we found the following: the sentinel lymph node (SLN) identification rate was significantly higher for PA injection of tracers (labeled nanocolloid or blue dye) than for subdermal/peritumoral injection; at lymphoscintigraphy, the number of late images necessary to visualize the SLN was significantly reduced (20% for PA injections vs. 39.5% for subdermal/peritumoral injections); and with the PA injection technique, the need for image-guided injection was bypassed for patients with nonpalpable tumors.
The PA injection technique is easy to perform and simpler than the other mentioned techniques. In a volume of 0.5 mL, we inject, as a single aliquot, 20–40 MBq of 99mTc-labeled Nanocoll (Nycomed Amersham Sorin S.r.l.). As shown in Figure 1 of our paper (1), the nanocolloid is injected subdermally at the PA site (1–2 mm from the areolar-cutaneous “junction” (2)), at the level corresponding to the tumor. Then, to aid clearance of radiocolloids, a gentle massage is performed.
The injection of such a low volume (0.5 mL) of labeled nanocolloid (instead of sulfur colloid) allows the procedure to be completed in a few minutes, without using anesthetic or causing discomfort to the patient.
In their letter, Kim et al. suggest, first, a new injection technique at the areolar-cutaneous junction to increase the efficiency of delivering activity to the sentinel node (2,3) and, second, lymphoscintigraphic “dynamic monitoring and multiple views … to best map out what is really happening in the patient” after tracer injection. Their goal is to minimize morbidity from SLN biopsy in patients with early breast cancer.
We thank Kim et al. for their letter and agree with them that, theoretically, their suggested solutions would improve the accuracy of SLN biopsy. However, as reported for different studies in which axillary lymph node dissection was performed after SLN biopsy, the false-negative rate of SLN biopsy in patients with early breast cancer is rather low, ranging from 5% to 10% (4–6). Kim et al. did not report an improved false-negative rate from using the multiple-injection technique and dynamic monitoring of radiotracer distribution. In addition, the increased number and volume of injections, and the use of a multiple-view dynamic acquisition for lymphatic mapping, increases patient discomfort and lengthens the procedure. Therefore, we believe that further studies on larger patient populations are mandatory to quantify the real improvement achievable with the technique suggested by Kim et al.
To visualize the axillary SLN in patients with early breast cancer, PA injection of 20–40 MBq of labeled nanocolloid in a low, 0.5-mL, volume, is suggested.