| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
EDUCATION EXHIBIT |
Department of Radiology, Loyola Stritch School of Medicine, Loyola University Medical Center, Maywood, Illinois
Lymphoscintigraphy is a simple imaging technique that makes use of the principle that radiolabeled macromolecules introduced into the soft tissues around tumors are transported by the lymphatics and localized in lymph nodes that drain the region. The observation of Sherman and Ter-Pogossian (1) that colloidal Au-198, injected interstitially, could be seen in local lymph nodes on autoradiographs provided evidence that lymphoscintigraphy might be useful clinically. This technique was first widely used for evaluation of the lymphatic drainage in melanoma patients and has now become the primary method for determining the lymphatic drainage of breast tumors and locating the sentinel lymph node(s).
The vast majority of lymphoscintigraphy studies performed today are done on the breast, but the technique has been studied in several other tumors. It is useful in evaluating patients with melanomas, particularly those located in the trunk, where the lymphatic drainage can be unpredictable (2). Lymphoscintigraphy has also been studied to determine the lymphatic drainage of vulvar carcinoma, bladder carcinoma, and prostate carcinoma, and some investigators have also used it intraoperatively to locate sentinel lymph nodes in colon carcinoma (3–6). These other applications of lymphoscintigraphy were successful, but the technique was widely used only to evaluate patients with melanoma.
With most of the mentioned applications, there is a single injection technique that is used most often. For melanoma, multiple intradermal injections around the lesion (or the scar after excisional biopsy) is the preferred method. Perianal or perivulvar subcutaneous injections are used to visualize the inguinal and pelvic lymph nodes, which is important for evaluation of bladder, vulvar, or rectal tumors (5,7,8). These nodes are inconsistently visualized after injections of radiocolloid in the dorsum of the foot or injection of iodinated oil directly into the lymphatics of the feet in a lymphangiogram. Perilesional injections, along the mesenteric side of the bowel, are used to identify lymph nodes draining colon carcinoma. A handheld gamma probe is used to locate these nodes, and the entire procedure can be performed in the operating room (6). Transrectal prostatic injections have also been proposed to visualize the lymphatic drainage of the prostate gland (9). An excellent review of possible injection sites and the lymph node groups visualized with each technique has been prepared by Ege (10). This extensive list shows that there are few tumors or lymphatic pathways that have not been evaluated with lymphoscintigraphy.
Unlike the tumors discussed above, the complex lymphatic drainage of the breast has provided a unique challenge in demonstrating lymphatic tumor involvement by using lymphoscintigraphy. Lymphoscintigraphy was first used in breast cancer patients to evaluate the involvement of the internal mammary lymph nodes (11). This technique used subcutaneous subcostal injections, with the goal being to deposit the radiocolloid anterior to the posterior rectus sheath, near the origin of the parasternal lymphatics. Injections had to be performed on each side of the upper abdomen, in the midclavicular line (12). Although somewhat more difficult and certainly more invasive than the cutaneous and perilesional injections now used for breast lymphoscintigraphy, this technique can easily be mastered, and if successful, will show the lymph nodes of both internal mammary lymph node chains. Involvement of these lymph nodes is assumed if there is absent visualization of the lymph node chain on the ipsilateral side of the patient’s tumor (12). Radiation oncologists sometimes used this technique to determine the depth of the internal mammary lymph nodes. Lateral views of the chest allowed measurement of the distance of these nodes from the skin surface. Appropriate radiation dosimetry calculations then allowed optimal radiation delivery to these nodes (13). The difficulty of this technique, the occasional congenital absence of the internal mammary lymph node chain on one side, and the fact that few surgeons routinely sample these lymph nodes caused this technique to go out of favor.
Involvement of axillary lymph nodes changes the staging and prognosis of patients with breast cancer. Surgeons began to evaluate ways that they could better determine the location of these lymph nodes for easy biopsy. Initial attempts to locate axillary lymph nodes using intradigital injection sites often visualized multiple lymph nodes in the axilla, and it was impossible to determine which of these might also be draining the breast. Articles describing lymphoscintigraphy performed by injection of the radiocolloid in the breast showed that this could be a useful adjunct to surgical axillary lymph node dissection (14).
The preceding article by Krynyckyi et al (15) reviews the many scintigraphic techniques available to visualize the sentinel lymph node(s) in breast carcinoma. They give an excellent historical review of the various methods that have been attempted, giving both the advantages and pitfalls of each technique, as well as suggestions to help ensure their success. They also provide images to demonstrate the differences in these techniques. They conclude that in their experience, the dual-injection technique, what they describe as "lymphoboost," which includes both perilesional injections and injections around the areolar-cutaneous junction, is the most consistent method to find SNs. Their argument is that this technique visualizes the SNs much faster and produces nodes that are hotter compared to lymph nodes seen by using either a cutaneous or perilesional injection alone. They also emphasize that the value of including a perilesional injection in all patients is visualization of internal mammary SNs.
The failure to visualize these internal mammary lymph nodes by using cutaneous injections was demonstrated by Roumen et al (16). They compared the results in patients who received both peritumoral and cutaneous injections at two different sittings. The internal mammary SNs were almost never seen with intradermal injections alone. If it is a priority of the surgeon and oncologist to visualize all possible SNs, whether they occur in the axilla or the internal mammary lymph node chain, then this LB technique should be used. However, sampling of the internal mammary lymph nodes is no longer widely practiced. Access to these lymph nodes involves much more complicated surgery with increased morbidity and mortality. With the exception of determining lymph node depth for radiation therapy, as discussed earlier, there is little need to visualize the internal mammary lymph nodes. After discussions with our surgeons, it was decided that visualizing SNs located in the internal mammary lymph node chain was not a priority. They preferred the consistent, rapid visualization of hot axillary SNs resulting from an intradermal injection.
When we first offered breast lymphoscintigraphy, we attempted perilesional injections, but we encountered some of the same problems described by the authors of this article. On several occasions, there were hematomas or seromas near the tumor, as a result of prior biopsy attempts, into which the radiocolloid was inadvertently placed. We were also unconvinced that we were placing the radiocolloid consistently in the proximity of the lesion, even when it was palpable. The perilesional injection technique is even more difficult if the lesion was initially found by mammography, with biopsy diagnosis only under mammographic guidance, and where there is no palpable mass. The presence of a localization guide wire also makes it difficult to perform a perilesional injection, as care must be taken to avoid movement of the end that is placed near the tumor. Multiple surgeons, with different operating times, as well as the scheduling constraints in mammography, means that the lymphoscintigraphy may be performed either before or after mammographic guide-wire localization.
For a short time we used intradermal injections, injecting in the skin surface closest to the site of the lesion. This technique is again difficult to use when the lesion is not palpable and is seen only at mammography. Selection of an injection site in these patients is imprecise. For these reasons, we have switched almost exclusively to an intradermal, periareolar technique. The radiocolloid is injected 1–2 cm from the areolar margin. Four to six intradermal injections of 1–1.5 mCi of filtered Tc-99m sulfur colloid in a total volume of 1 mL results in rapid visualization of a hot SN. In the rare cases where no SN is seen, we use some of the same techniques described in the article or reinject more of the radiocolloid. Patients who undergo injection in the afternoon prior to surgery are given 2–3 mCi. The intradermal technique results in more initial activity in the SN, helping to ensure there will be enough activity remaining at the time of surgery the next morning for the surgeon to identify it with the gamma probe.
The discomfort experienced by patients during lymphoscintigraphy can also be a factor in which technique is chosen. The authors of the article have found the use of a topical anesthetic on the skin, at the site of the cutaneous injections, to be effective (15). This may not be effective if the burning sensation some patients experience comes from a deep injection or the injection of larger volumes. In their experience with the lymphoboost technique, the rapid visualization of hot SNs in the axilla and/or in the internal mammary lymph nodes far outweighs the minimal added patient discomfort from their dual-injection technique. We have found that injecting only small volumes intradermally has seldom required a topical anesthetic. Very gentle massage of the injection site in a circular direction by the opposite hand can hasten the disappearance of the burning sensation some patients feel. The areolar area should be covered with a 4 x 4 gauze pad to avoid contamination of the skin surface by any activity that may leak out with the massage. This gauze must be removed before imaging. Most patients find the minor discomfort with this technique acceptable.
We routinely obtain anterior and lateral images of the breast that is injected, beginning 15 minutes after injection, being sure to include the axilla. The movement of activity with intradermal injections is so rapid that frequently an SN is seen at 15 minutes. We seldom have to go beyond 90 minutes before an SN is found, and usually the SN can be localized at 60 minutes. We also localize the SN with a handheld gamma probe. These probes are well collimated and highly directional. In most patients, the SN can be found by moving the probe over the surface of the skin. In large patients or if the SN cannot be found quickly by using this technique, a "hot" marker is used to locate the SN with the gamma camera. A single drop of any Tc-99m radionuclide on a small adhesive bandage that is then folded and placed in a rubber glove works well. The skin is marked at the location of highest activity. For deeper lesions, two marks on the skin surface placed at a 90° angle, one anteriorly and one laterally, allow the surgeon to locate the SN by triangulation. The number of SNs seen, their approximate location, along with the position of the patient’s arm and the angle and direction of the probe when performing the localization are then communicated to the surgeon.
Our breast surgeons currently use a combination of both a dye localization technique and radionuclide lymphoscintigraphy. Isosulfan blue (Lymphazurin 1%; United States Surgical, Norwalk, Conn) is injected in the subcutaneous tissues around the areola. This is done in the operating room after the patient is anesthetized, as this material moves too rapidly to be injected at the same time as the radiocolloid. The surgeon dissects along the blue lymphatics until a lymph node is found. The activity within the lymph node is determined with the handheld gamma probe. The SN will have significantly higher activity than the surrounding tissues. The combination of a blue and hot lymph node identifies the SN. With this technique, SNs may be found even when the scintigraphic images fail to show any activity in the axilla. We have demonstrated a 99.5% success rate in finding an SN using the combination of radionuclide lymphoscintigraphy and isosulfan blue dye injection, despite the fact that either of the techniques alone may be unsuccessful (17).
The techniques described by Krynyckyi et al (15) give radiologists and nuclear medicine physicians many choices when performing lymphoscintigraphy of the breast. This article describes these techniques and the results that can be expected from each technique. The choice of which technique will work best at an institution will depend on several factors. The imaging physicians should discuss the possible options with the referring surgeons, oncologists, and radiation oncologists who will be caring for these patients. There is variation among surgeons as to what information they require from a lymphoscintigraphy study. If the breast surgeon is comfortable with routinely identifying the SN(s) found only within the axilla and does not need the information about the presence of SNs in the internal mammary lymph node chain, then any of the single injection cutaneous methods would be acceptable. If the surgeon is willing to explore the internal mammary lymph nodes or if the oncologist or radiation oncologist always wants to know whether or not there is drainage to the internal mammary lymph nodes, particularly in tumors found in the inner quadrants, then a technique that includes injecting the radiocolloid into the deep tissues surrounding the lesion should be used. In a large institution with many surgeons, this may require offering several injection techniques, and this can be problematic. Discussing technique options with the surgeons and oncologists prior to performing lymphoscintigraphy studies on the breast will ensure that the referring physicians have the information they require and patients receive the best possible care.
 |
References |
|---|
 Top References References |
|---|
Department of Radiology, Mount Sinai Medical Center, New York, New York
We appreciate the added dimensions to SN harvesting that Dr Dillehay eloquently discusses. On average, the different intradermal injection techniques will demonstrate the same nodes in the vast majority of cases and will generally all produce relatively hot nodes compared with perilesional injections. Four years ago, we were using simultaneous perilesional and intradermal injections of radiocolloid above the tumor (1,2). However, our later results with sequential perilesional, intradermal over the tumor, and areolar-cutaneous junction injections performed in the same patient at one sitting demonstrated that the areolar-cutaneous junction injections, which we dubbed LymphoBoost, produced unquestionably hotter nodes than even cutaneous injections over the lesion by a factor of about three (3). The injections at the areolar-cutaneous junction (or other similar areolar injections) could easily be performed by themselves and in the majority of cases produce good results. For these reasons, we have debated dropping perilesional injections altogether and performing areolar-cutaneous junction lymphoboost injections exclusively, as also mentioned by Dr Dillehay. However, we chose to retain perilesional injections because of several reasons:
1. The presence of drainage to internal mammary nodes might alter radiation therapy decisions if surgeons are presented with the information. We completely agree that the internal mammary nodes are not removed in the majority of clinical practices, and our own surgeons also do not harvest internal mammary nodes routinely. They do, however, want to know if they appear during a study and have indicated that their presence will often justify external-beam therapy to the region. It is interesting that the newest revision of the classification for breast cancer staging, the sixth TNM classification, defines internal mammary node disease as N1–N3, depending on size and/or concurrent axillary involvement, as compared to only N3 in the fifth edition. This suggests that the internal mammary nodes detected with lymphoscintigraphy still have a completely undefined role that is not trivial enough to completely ignore (4,5).
2. A small fraction of patients exhibit a partial discongruence pattern, the reverse echelon node discussed in the article. The reverse echelon node, as defined by us, is most likely a common feature of all dermal injections, especially those injections performed around the areola, which are, in some cases, further away from the axilla than the primary lesion itself. These reverse echelon nodes would clinically be a problem (false negative) only if surgeons harvested just the first or hottest "sentinel node" they came across and ignored "echelon nodes" downstream. In fact, even with perilesional injections alone, the hottest node is not the node with disease in up to 28% of cases, as noted in one study, and multiple nodes if displaying activity should always be harvested (6).
3. An even rarer pattern of complete discongruence (outside the internal mammary basin) was noted by us, where completely different, independent, axillary "sentinel nodes" were seen with perilesional injections as compared to dermal-type injections, including lymphoboost. This pattern was seen in no more than 2% of cases and could be used as a rough estimate of a potential increase in the false-negative rate from this factor alone, if one was to elect not to perform perilesional injections.
4. Finally, in very rare instances, the perilesional injection produces nodes that are hotter than those with lymphoboost injections, especially when intervening lumpectomy scars are present or the areolar-cutaneous junction has been surgically distorted.
Aside from these issues, the inclusion of dynamic imaging and the perilesional injections as a "quality control" feature was, and continues to be, valuable to us in assessing for unique patterns of drainage to SNs and in assuring that the lymphoboost injections at the areolar-cutaneous junction do indeed delineate the same nodes in the vast majority of patients, while performing a better job of it. We agree with Dr Dillehay that if only lymphoboost injections were performed without perilesional injections, any negative effects to patients would most likely be in a fairly small minority, although definitely not zero, and these similar issues most likely also apply to all other types of areolar and dermal injections. The nodes would be basically as hot as without perilesional injections in most cases. This efficiency is suggested by others using similar-type areolar injections, as noted in our article and since its writing (7,8). Nevertheless, on the basis of the available data in the literature and corresponding examples of results from figures, the very specifically defined areolar-cutaneous junction injections, as described by us, appear to generate hotter nodes than any of the wide range of different areolar injection techniques described.
Is the pattern of partial or complete discongruence solely a feature of lymphoboost injections? Most likely not, and it is reported by us simply because we were the first, to our knowledge, to sequentially perform all the different injections in one sitting and to simultaneously monitor the results dynamically. Others used blue dye or performed injections on different days, making the type of direct comparisons we made difficult if not impossible (9). Any type of areolar injection, and most likely dermal injections, would also be susceptible to these various issues.
The ultimate test of long-term benefit will require the completion of large ongoing outcome trials (10). In the meantime, recent studies from the Mayo Clinic present data that look promising. In over 680 patients who underwent sentinel lymph node biopsy (SLNB) alone who were node negative, there was one axillary recurrence when the patients were followed up for over 2 years (11). Patients also experienced lower levels of lymphedema, arm pain, and seroma formation by at least 20%–25% compared to an axillary lymph node dissection (ALND) group. In another study, where 516 patients were randomized into equal groups for SLNB and ALND and followed up for a mean of 46 months, there were fewer unfavorable events related to recurrence of disease in the SLNB group compared to the ALND group (12). Again, there were lower levels of morbidity at both 6 and 24 months after surgery for the SLNB group.
Not all studies demonstrate a benefit of morbidity reduction for SLNB. When short-term effects of SLNB versus ALND were examined, no difference in morbidity could be found (13). Unlike most other studies that examined differences in morbidity, this study did not utilize the benefits of lymphoscintigraphy, and perhaps, by lacking the detailed road map lymphoscintigraphy provides, the surgeons would dissect the patients’ tissues in a more extensive manner to find the SN, leading to an increase in morbidity.
The role that SLNB and lymphoscintigraphy have in treating breast cancer patients appears to be expanding, as suggestions have been made that they are also appropriate in patients with multicentric and multifocal invasive breast cancers (14). In addition, it has been shown that the accuracy of SLNB after neoadjuvant chemotherapy is similar to that in patients who have not undergone chemotherapy and primary surgery (15–17). Patients with large primary tumors, ductal carcinoma in situ (with or without microinvasion), pregnancy, or implants and those who have undergone excisional biopsy are also appropriate for SLNB in the proper circumstances (18–20). Lymphoscintigraphy results are also highly reproducible, further lending credence to the technique in general (21).
Several additional points warrant mention. Discomfort is minimal or completely lacking in a majority of patients when we include anesthetic mixed with the Tc-99m sulfur colloid in the syringe in addition to using topical anesthetic. The appropriate manipulation of the raw data when displaying and printing the images is critical in delineating faint nodes and lymphatic tracts in the optimal way. In addition to the adjustment of standard upper and lower levels, additional functions entailing predisplay processing of data are usually required to enhance low-count data relative to the injection site data. Also, a nonlinear display response curve that maps the postprocessed data to the gray scale in a way that suppresses high-count data to prevent injection site "blooming" is usually required. Finally, we wholeheartedly agree with the comments by Dr Dillehay concerning communicating on technique and the plan of action with surgeons, which is extremely important.
|
References |
|---|
|
TopReferencesReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOGRAPHICS | RADIOLOGY | RSNA JOURNALS ONLINE |
