- Email: [email protected]
Surgical management of aberrant sentinel lymph node drainage in cutaneous melanoma
Surgical management of aberrant sentinel lymph node drainage in cutaneous melanoma Kimberly A. Lieber, MD, Steven B. Standiford, MD, Boris W. Kuvshinoff, MD, and David M. Ota, MD, Columbia, Mo
Background. Sentinel lymph node (SLN) mapping by lymphoscintigraphy has changed the surgical management of regional lymph node metastases for melanoma. SLNs lying outside of traditional nodal basins are now being identified. Our hypothesis is that when preoperative lymphoscintigraphy identifies aberrant SLNs, these nodes should be excised and, if histologically positive, lymphadenectomy of the aberrant nodal basin should be performed. Methods. Patients with melanomas 1 mm or larger Breslow thickness and clinical stage N0M0 underwent lymphoscintigraphy and excision with SLN biopsy. Preoperative lymphoscintigraphy, intraoperative gamma probe, and intraoperative injection of isosulfan blue were performed to identify the SLN. Aberrant SLNs were defined as epitrochlear, supraclavicular, or popliteal nodes for extremity lesions and intramuscular nodes for truncal and head and neck lesions. Results. Thirty-two patients were entered into the protocol. Seven (22%) were found to have aberrant nodes. Five of 19 patients with extremity melanoma had an aberrant SLN; 2 of 13 patients with truncal and head and neck melanoma had an aberrant SLN. Conclusions. This study demonstrates that (1) aberrant SLNs are encountered with similar frequency for extremity and truncal lesions, (2) biopsy should be performed on aberrant SLNs with intraoperative lymph node mapping with the gamma probe and blue dye, and (3) lymphadenectomy of the aberrant region should be considered if the aberrant SLN is positive. (Surgery 1998;124:757-62.)
IDENTIFICATION OF REGIONAL LYMPH NODE drainage is of critical importance in the surgical management of cutaneous melanoma. Metastatic disease is most commonly found in the nodal basin that drains the primary lesion.1 Lymphoscintigraphy identifies sentinel lymph nodes on which biopsy can be performed and used to stage disease.2-4 Based on the histologic presence of metastases in the sentinel lymph node, surgeons can select patients who may benefit from lymph node dissection to remove regional microscopic metastases. Lymphoscintigraphy has expanded the definition of ambiguous cutaneous lymphatic drainage for lesions of the head, neck, and trunk. 4 Lymphatic drainage for lesions of the extremities, however, is considered to be well defined and without much variation. 5 Sentinel lymph node mapping and biopsy are used in the Presented at the Fifty-fifth Annual Meeting of the Central Surgical Association, Ann Arbor, Mich, March 5-7, 1998. Reprint requests: David M. Ota, MD, Ellis Fischel Cancer Center, University of Missouri Health Sciences Center, 115 Business Loop 70 W, Columbia, MO 65203. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0
11/6/90943
extremities to select patients for lymph node dissection who may benefit and reduce morbidity for those who likely would not benefit. When an axillary sentinel lymph node is found to have microscopic metastases, an axillary node dissection is recommended. Similarly, a superficial inguinal node dissection should be done if patients with lower extremity melanomas are found to have a histologically positive sentinel node biopsy specimen.6 As experience with the technique of lymphoscintigraphy increases, more patients are found to have unexpected or aberrant sentinel lymph nodes draining cutaneous melanoma of the extremities, as well as the head, neck, and trunk. Data will be presented supporting the use of lymphoscintigraphy for patients with primary melanoma of intermediate thickness, in any location. Aberrant sentinel lymph nodes found in the epitrochlear, popliteal, and supraclavicular lymph node basins will be described. Intramuscular sentinel lymph nodes, also referred to as in-transit nodes, and their treatment will also be discussed. MATERIAL AND METHODS Thirty-two consecutive patients with cutaneous SURGERY 757
758 Lieber et al
Surgery October 1998
Fig 2. Lymphoscintigraphy of patient with primary melanoma of posterior foot. Knee is seen in profile with radioactive hot spot (white) just proximal to joint. This represents popliteal lymph node. Fig 1. Lymphoscintigraphy of patient with primary melanoma in dorsum of hand. Technetium sulfur colloid was injected in left hand at injection site. Radioactive hot spots (white) are seen just proximal to elbow and in axilla. Radioactivity near elbow is in epitrochlear region.
melanomas 1 mm or greater Breslow thickness and clinical stage N0M0 underwent wide local excision, lymph node mapping, and sentinel node biopsy from September 1994 to August 1997. Each patient was referred, already having tissue confirmation of the primary lesion (punch biopsy, wedge biopsy, or excisional biopsy), and each patient had undergone preoperative lymphoscintigraphy in the Department of Nuclear Medicine on the morning of the operation (Fig 1). Lymph node mapping was performed with 400 µCi unfiltered 99mTc-labeled sulfur colloid, which was injected in 4 aliquots surrounding the primary lesion. The Neoprobe 1000 hand-held gamma probe (Neoprobe Corp, Dublin, Ohio) was used during the operation to locate the sentinel lymph node and plan the excisional lymph node biopsy incision. Before skin preparation, isosulfan blue (Lymphazurin 1%; USSC, Norwalk, Conn) was injected around the primary lesion by the surgeon. Blue lymph nodes were excised as were any radioactive nodes that exhibited gamma counts significantly above background readings. Radioactive nodes were recognized by counts greater than 2 times the background. After excision, the lymph node bed was scanned to confirm that all radioactivity had been removed and counts had returned to background levels. Every sentinel lymph node was sent for permanent microhistologic evaluation consisting of step sectioning and
hematoxylin-eosin staining. Sentinel lymph nodes were considered aberrant if they were located outside of the traditional lymph node basin that would be included in a formal lymph node dissection for the respective primary lesion. Specifically, aberrant nodes included epitrochlear, supraclavicular, popliteal, or intramuscular lymph nodes. Based on the permanent histologic evaluation of the sentinel node, patients were advised regarding lymph node dissection and adjuvant therapy. RESULTS Thirty-two patients were enrolled in the database. Of these, 19 (60%) had cutaneous melanoma of the extremity and 13 (40%) had cutaneous melanoma of the head and neck or trunk. Two patients had recurrent melanoma. Mean Breslow thickness for all patients was 2.7 mm with a range from 0.8 to 7.0 mm. Sentinel lymph nodes were identified and biopsy was performed in all patients. Seven patients (22%) were found to have aberrant sentinel lymph nodes. Multiple sentinel lymph nodes were encountered in 15 (47%) of the patients studied, and the mean number of sentinel lymph nodes removed per patient was 2 (range 1 to 5). Positive sentinel lymph nodes were found in 4 patients (12.5%). Lymphadenectomy was performed in 3 of the 4 patients. Two of the positive sentinel lymph nodes were aberrant. One patient with a positive popliteal sentinel node also had 4 negative sentinel lymph nodes in the superficial inguinal basin (primary lesion was on the posterior sole of the right foot). This patient elected not to undergo a popliteal node dissection and had recurrent melanoma in the popliteal fossa. Another
Surgery Volume 124, Number 4
patient with a positive trapezius sentinel lymph node also had a positive posterior cervical node (primary lesion was located on the ipsilateral parietal scalp). This patient underwent radical neck dissection, with all nodes negative for metastatic melanoma, and has subsequently undergone resection for local recurrence but has no evidence of metastatic disease. Three of 9 patients with primary lesions of the upper extremity had lymph node drainage to nodes outside of the axilla. One patient with a primary subungual lesion of the right ring finger was found to have sentinel lymph nodes located in the epitrochlear lymph node basin and the ipsilateral axilla. Another patient whose primary lesion was located on the left elbow was found to have sentinel nodes in the bicipetal groove and the axilla. Final pathologic findings for both nodes were negative; however, the patient went on to have a recurrence in the ipsilateral axilla. A third patient with a primary lesion on the right proximal forearm had a sentinel node in the supraclavicular region and no sentinel nodes in the axilla. Two of 10 patients with lower extremity lesions had aberrant sentinel lymph nodes. Primary lesions were located on the sole of the foot and over the Achilles tendon. In both patients the aberrant sentinel lymph node was located in the popliteal nodal basin (Fig 2). The patient who underwent removal of a positive sentinel lymph node from the popliteal fossa who did not undergo popliteal node dissection was treated with α-interferon and changed to systemic chemotherapy when regional recurrence was confirmed. Fifteen percent of patients with primary lesions of the head, neck, and trunk were found to have aberrant sentinel lymph nodes. A patient with a primary lesion of the parietal scalp (mentioned above) was found to have 2 sentinel lymph nodes. One node was in the posterior cervical node chain (expected nodal basin); the other node was in the trapezius muscle (aberrant node). Both nodes were positive for metastatic disease. Radical neck dissection was performed and no further metastatic disease has been identified. Three sentinel lymph nodes were identified in another patient with recurrent melanoma of the low midback. Two of the sentinel nodes were located in the left axilla and the third was a node located very near the primary lesion (Fig 3). This patient continued to have negative nodes. DISCUSSION Lymphatic drainage of the extremity to the regional nodal basin is relatively more predictable
Lieber et al 759
Fig 3. Lymphoscintigraphy of patient with primary melanoma on left flank. Arms are raised over patient’s head. White spot marks left nipple (LT Nipple) anteriorly (ANT). Two small white spots are seen in region of left axilla. Technetium sulfur colloid was injected at large white spot (primary lesion). Small white spot representing intramuscular node is seen close to injection site.
than for the trunk. Unidirectional flow of lymphatics on the extremity implies that the sentinel lymph node for a cutaneous region will be located proximal to the primary lesion. Multidirectional flow exists on the trunk and head and neck.5 Sugarbaker and McBride7 expanded on Sappey’s anatomic description of lymphatic drainage of the trunk. Improved technology now expands the surgeon’s ability to know precisely where regional lymph node basins are located and where metastases are likely to be found. Lymphoscintigraphy further expands the area of ambiguous lymphatic drainage on the trunk, which extends like a belt and vest across the torso and shoulders and includes the head and neck.8 Regional lymph node basins for primary cutaneous lesions in the ambiguous zone are expected to be axillary, inguinal, or cervical. The ambiguity arises from an inability to predict anatomically which regional nodal basin will contain the sentinel node. Aberrant sentinel lymph nodes do not lie within expected lymph node basins and would be overlooked without lymphoscintigraphy. Aberrancy was encountered in 22% of patients in this series. Lymphoscintigraphy was performed for every patient, regardless of the location of the primary lesion. Aberrant nodes were discovered with similar frequency on the extremities as on the
760 Lieber et al
Surgery October 1998
Table I. Distribution of patients undergoing wide local excision for cutaneous melanoma Site
Patients (n)
Sex (F/M)
Age (yr)
Breslow thickness (mm)
SLNs (n)
4 9 9 10 32
2/2 2/7 4.5 9/1 17/15
46.8 (20-74) 47.9 (22-65) 54.7 (33-72) 55.7 (14-83) 52.1 (14-83)
1.77 (0.93-2.3) 2.31 (0.80-5.2) 3.38 (1.0-7.0) 2.80 (1.1-5.8) 2.70 (0.80-7.0)
2.8 (1-4) 2.0 (1-3) 2.1 (1-5) 1.5 (1-5) 2.0 (1-5)
Head and neck Trunk Upper extremities Lower extremeties Total
Age, thickness, and number of sentinel lymph nodes are expressed as mean (range). F, Female; M, male; SLNs, sentinel lymph nodes.
Table II. Distribution of location of sentinel lymph nodes by site of primary lesion Primary site Upper extremity (9) Lower extremity (10) Trunk (9) Head and neck (4)
Expected node
Aberrant node
Axilla (6) Groin (8) Axilla or groin (8) Cervical (3)
Supraclavicular (1) Epitrochlear (1) Intramuscular (1) Popliteal (2) Intramuscluar (1) Intramuscular (1)
Numbers in parentheses are numbers of patients.
trunk or head and neck (26% vs 15%). Gamma probe guidance during the operation enables the surgeon to plan precisely where to excise these nodes that would otherwise be difficult to locate through a small incision.9 Blue dye aids in identification of the sentinel lymph node.10 Isosulfan blue has a relatively short washout period, however, and a blue node could not be found in all cases.11 Aberrant sentinel lymph nodes have been described in the literature. Smith et al12 performed a retrospective chart review of 240 patients with melanoma of the upper extremity from 1955 to 1970. The measured end point of their study was clinical involvement of the epitrochlear lymph nodes. Seven (18%) of 39 patients with primary lesions distal to the elbow had epitrochlear metastases. They conclude that early dissection of clinically suspicious epitrochlear nodes improves patient outcome. Lymphoscintigraphy was not widely available for the patients of that study. Tanabe13 recently described the technique of epitrochlear lymphadenectomy and advocates selective lymph node dissection based on the status of the sentinel node. Regarding lymphatic drainage of the lower extremity, lesions of the sole and heel are most likely to drain into the popliteal nodal basin.14 Karakousis15 has described a technique for dissection of the popliteal lymph nodes when metastatic disease is present or strongly suspected. Although metastases to popliteal nodes are considered rare for cutaneous melanoma, 2 of 10 patients with lower extremity lesions in this series had sentinel lymph nodes in the popliteal fossa on
lymphoscintigraphy. One of these patients went on to have recurrence in this area. This unfortunate experience suggests that lymphadenectomy should be considered whenever and wherever a positive sentinel lymph node is detected. To locate these nodes, the combined modalities of lymphoscintigraphy, blue dye, and intraoperative gamma detection is important. Without the preoperative lymphoscintigraphy, unexpected nodal basins can be overlooked. Aberrant sentinel lymph nodes in locations not associated with any nodal basin (eg, muscle nodes) are easily excised with the gamma probe during the operation. Therefore we recommend the use of lymphoscintigraphy for every patient with a cutaneous melanoma 1 mm or greater in thickness, regardless of location. Biopsy should be performed in all sentinel lymph nodes, and when a positive sentinel node is identified, dissection of the regional nodal basin should be considered. REFERENCES 1. Balch CM, Ross MI. General principles of regional lymphadenectomy. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S, editors. Cutaneous melanoma. 2nd ed. Philadelphia: JB Lippincott; 1992. p 340-4. 2. Morton DL, Wen D, Wong JH, Economou JS, Cagle LA, Storm FS, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392-9. 3. Thompson JF, McCarthy WH, Bosch CMJ, O’Brien CJ, Quinn MJ, Paramaesvaran S, et al. Sentinel lymph node status as an indicator of the presence of metastatic melanoma in regional lymph nodes. Melanoma Res 1995;5:255-60. 4. Berger DH, Feig BW, Podoloff D, Norman J, Cruse CW, Reintgen DS, Ross MI. Lymphoscintigraphy as a predictor of lymphatic drainage from cutaneous melanoma. Ann Surg Oncol 1997;4:247-51.
Surgery Volume 124, Number 4
5. Balch CM, Lamki LM, Logic JR. Defining lymphatic drainage patterns with cutaneous lymphoscintigraphy. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S, editors. Cutaneous melanoma. 2nd ed. Philadelphia: JB Lippincott; 1992. p 368-74. 6. Balch CM, Milton GW, Cascinelli N, Sim FH. Elective lymph node dissection: pros and cons. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S, editors. Cutaneous melanoma. 2nd ed. Philadelphia: JB Lippincott; 1992. p 346-64. 7. Sugarbaker EV, McBride CM. Melanoma of the trunk: the results of surgical excision and anatomic guidelines for predicting nodal metastasis. Surgery 1976;80:22-30. 8. Norman J, Cruse CW, Espinosa C, Cox C, Berman C, Clark R, et al. Redefinition of cutaneous lymphatic drainage with the use of lymphoscintigraphy for malignant melanoma. Am J Surg 1991;162:432-7. 9. Krag DN, Meijer SJ, Weaver DL, Loggie BW, Harlow SP, Tanabe KK, et al. Minimal access surgery for staging malignant melanoma. Arch Surg 1995;130:654-8. 10. Pijpers R, Borgstein PJ, Meijer S, Hoekstra OS, van Haattum LH, Teule GJJ. Sentinel node biopsy in melanoma patients: dynamic lymphoscintigraphy followed by intraoperative gamma probe and vital dye guidance. World J Surg 1997;21:788-93. 11. Albertini JJ, Cruse CW, Rapaport D, Wells K, Ross M, DeConti R, et al. Intraoperative radiolymphoscintigraphy improves sentinel lymph node identification for patients with melanoma. Ann Surg 1996;223:217-24. 12. Smith TJ, Sloan GM, Baker AR. Epitrochlear node involvement in melanoma of the upper extremity. Cancer 1983;51:756-60. 13. Tanabe KK. Lymphatic mapping and epitrochlear lymph node dissection for melanoma. Surgery 1997;121:102-4. 14. Haagensen CD, Feind CR, Herter FP, Slanetz CA Jr, Weinberg JA, editors. The lymphatics in cancer. Philadelphia: WB Saunders; 1972. 15. Karakousis CP. The technique of popliteal lymph node dissection. Surg Gynecol Obstet 1980;151:421-3.
DISCUSSION Dr Edward G. Mansour (Cleveland, Ohio). You are bringing to our attention a very important observation that is essential to the successful implementation of lymphoscintigraphy and sentinel lymph node biopsy with intraoperative mapping in the management of malignant melanoma 1 mm or greater in thickness. Your findings clearly demonstrate the need of a well-informed and trained radiologist in nuclear medicine. The radiologist needs to work closely with a trained surgeon in scanning multiple draining sites and identifying the first-appearing sentinel lymph node, by using dynamic rapid sequence imaging every 30 seconds until the sentinel lymph becomes evident. This usually occurs after 5 to 7 minutes. It is also important to have the technical ability to obtain views that determine the lymph node in a 3-dimensional reconstructed static picture. The spot overlying the lymph node is then tattooed. You describe 2 melanomas on the sole of the foot and the Achilles tendon that drained to the popliteal region. I believe this area is the primary drainage basin for these sites as described by Haagensen. The identification of
Lieber et al 761
inguinal lymph nodes in a patient with a popliteal sentinel lymph node implies rapid transit of the radioactive material to the secondary basin, which is not a sentinel lymph node site. Similarly, the epitrochlear area is not an unusual drainage basin for the distal upper extremity. The patient with a parietal scalp melanoma manifested a sentinel lymph node in the posterior cervical region and a second lymph node in the “trapezius muscle.” It is important to define clearly whether the so-called trapezius lymph node was truly within the muscle or along its edge, consistent with a deep occipital lymph node, as is often encountered in a posterior neck dissection. Realizing that the “aberrant” draining areas to the epitrochlear, popliteal, and trapezius regions are truly the primary drainage basins of the melanoma sites described, I wonder whether you might consider the term uncommon rather than aberrant. In our series of 27 consecutive lymphoscintigraphies, intraoperative mapping, and sentinel lymph node biopsies, we encountered 2 patients with intramuscular sentinel lymph nodes associated with trunkal lesions and a third patient with an epitrochlear lymph node in a distal upper extremity melanoma. The 3 patients had secondary axillary drainage sites on which we elected not to perform a biopsy. Sentinel lymph nodes in the 3 cases were negative. I fully agree with your conclusions. I would like to reemphasize the need for adequate training for the surgeon and radiologist by attending accredited training sessions in this subject, to obtain the most accurate information and to minimize or eliminate the false-negative rate of the sentinel lymph node biopsy. I have 3 questions: (1) Do you recommend biopsy of an identified axillary or inguinal lymph node in a patient who underwent biopsy of a sentinel lymph node in either the epitrochlear or popliteal region, respectively? (2) What was the average elapsed time between the injection of the radioactive material and the surgical procedure? In your patient with the proximal forearm melanoma who had a supraclavicular sentinel lymph node, could the radioactivity have cleared out of the axilla? (3) Realizing the small number in your series, the rate of positive sentinel lymph nodes is 12.5%, lower than the rate observed in other series (20% to 25%). What are the pathologic techniques used to examine the sentinel lymph node? Are you using multiple sections of the lymph node with immunohistochemical staining for S100 and monoclonal antibodies? Dr Lieber. In response to your first question, these are sample photographs that were taken from a more dynamic imaging process. So, to be honest, we chose the best pictures to show. I would agree with you that drainage to the popliteal basin is normal and should be considered in the analysis of sentinel node mapping. Upper extremity lesions from the ulnar aspect of the hand have a greater likelihood of draining to the epitrochlear region and the posterior foot can drain to the popliteal basin, as Dr Mansour pointed out. The node in the parietal scalp lesion was in the body of the trapezius muscle; it was well beyond the edge of the muscle.
762 Lieber et al
Regarding your second question, biopsy should be performed on a radioactive node seen in the axilla or the inguinal region even if a popliteal or epitrochlear node is found. I do not think we can ignore these, especially if we do not know whether these nodes are in sequence or if they have 2 separate lymphatic patterns. The current sunbelt melanoma trial chaired by Dr Kelly McMasters encourages biopsy of epitrochlear/popliteal sentinel nodes, as well as axillary/inguinal sentinel nodes. The supraclavicular node that I mentioned was an unexpected finding because the lesion was an upper arm lesion very near the elbow. In response to your third question, our pathologists do not routinely use S-100 or any immunohistologic staining. We have not been able to convince them to do this. Dr Hiram C. Polk, Jr (Louisville, Ky). Because you made such a broad endorsement of lymphoscintigraphy and sentinel lymph node biopsy, I am sure you can give us the data on costs and charges that are extra in the operating room and radiology as a result of this procedure. Dr Lieber. Professional and technical fees for lymphoscintigraphy are $820.00. Dr Vernon K. Sondak (Ann Arbor, Mich). I want to expand on that, because your conclusion that every patient should undergo preoperative lymphoscintigraphy 1 day before the operation would add dramatically to the cost and inconvenience of lymphatic mapping. Clearly, as has been pointed out, some of your drainage is not aberrant but just unsuspected and revealed by the radioactive tracers you used. We use different agents used to do lymphoscintigraphy (ie, preoperative imaging of the direction of lymphatic flow, human serum albumin and intraoperative lymphatic mapping, and unfiltered sulfur colloid). Both are labeled with technetium, but there are differences in the cost and sensitivity that make each one appropriate for their respective roles.
Surgery October 1998
In many cases we have abandoned doing preoperative lymphoscintigraphy with technetium-labeled albumin in favor of simply having our nuclear medicine colleagues call us at the time they inject the technetium sulfur colloid immediately before the surgical procedure. In nuclear medicine they do exactly what Dr Mansour indicated: image the patient from 5 minutes to 35 minutes after the injection and tell us how many sentinel nodes there are and where they are located. During the operation, we then confirm with our gamma detector that we can identify the hot spot in the relevant area. In doing so, you should not be encountering unsuspected lymphatic drainage areas that you cannot find in the operating room, even if you have not done preoperative lymphoscintigraphy with labeled albumin. Preoperative lymphoscintigraphy without albumin is still indicated in our opinion for the patients with midline truncal lesions with ambiguous drainage. If we find such a patient to have lymphatic drainage to 3 or 4 basins, we believe it is an insurmountable clinical challenge to find all the sentinel nodes and therefore exclude such patients from consideration for intraoperative lymphatic mapping and selective lymphadenectomy. However, I think to advocate preoperative lymphoscintigraphy routinely as a separate procedure for patients with extremity lesions or even truncal lesions that are off to one side of the midline seems to be overkill. Dr Lieber. This was a surprising result that sentinel nodes were found in unusual locations, especially on the extremities. We expected those on the trunk, as you mentioned. If the surgeon is willing to look vigorously through the extremity with the gamma probe, I believe that these unusual locations can be identified without lymphoscintigraphic scanning. Preoperative scanning makes it easier to uncover any unexpected sentinel nodes.
Background. Sentinel lymph node (SLN) mapping by lymphoscintigraphy has changed the surgical management of regional lymph node metastases for melanoma. SLNs lying outside of traditional nodal basins are now being identified. Our hypothesis is that when preoperative lymphoscintigraphy identifies aberrant SLNs, these nodes should be excised and, if histologically positive, lymphadenectomy of the aberrant nodal basin should be performed. Methods. Patients with melanomas 1 mm or larger Breslow thickness and clinical stage N0M0 underwent lymphoscintigraphy and excision with SLN biopsy. Preoperative lymphoscintigraphy, intraoperative gamma probe, and intraoperative injection of isosulfan blue were performed to identify the SLN. Aberrant SLNs were defined as epitrochlear, supraclavicular, or popliteal nodes for extremity lesions and intramuscular nodes for truncal and head and neck lesions. Results. Thirty-two patients were entered into the protocol. Seven (22%) were found to have aberrant nodes. Five of 19 patients with extremity melanoma had an aberrant SLN; 2 of 13 patients with truncal and head and neck melanoma had an aberrant SLN. Conclusions. This study demonstrates that (1) aberrant SLNs are encountered with similar frequency for extremity and truncal lesions, (2) biopsy should be performed on aberrant SLNs with intraoperative lymph node mapping with the gamma probe and blue dye, and (3) lymphadenectomy of the aberrant region should be considered if the aberrant SLN is positive. (Surgery 1998;124:757-62.)
IDENTIFICATION OF REGIONAL LYMPH NODE drainage is of critical importance in the surgical management of cutaneous melanoma. Metastatic disease is most commonly found in the nodal basin that drains the primary lesion.1 Lymphoscintigraphy identifies sentinel lymph nodes on which biopsy can be performed and used to stage disease.2-4 Based on the histologic presence of metastases in the sentinel lymph node, surgeons can select patients who may benefit from lymph node dissection to remove regional microscopic metastases. Lymphoscintigraphy has expanded the definition of ambiguous cutaneous lymphatic drainage for lesions of the head, neck, and trunk. 4 Lymphatic drainage for lesions of the extremities, however, is considered to be well defined and without much variation. 5 Sentinel lymph node mapping and biopsy are used in the Presented at the Fifty-fifth Annual Meeting of the Central Surgical Association, Ann Arbor, Mich, March 5-7, 1998. Reprint requests: David M. Ota, MD, Ellis Fischel Cancer Center, University of Missouri Health Sciences Center, 115 Business Loop 70 W, Columbia, MO 65203. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0
11/6/90943
extremities to select patients for lymph node dissection who may benefit and reduce morbidity for those who likely would not benefit. When an axillary sentinel lymph node is found to have microscopic metastases, an axillary node dissection is recommended. Similarly, a superficial inguinal node dissection should be done if patients with lower extremity melanomas are found to have a histologically positive sentinel node biopsy specimen.6 As experience with the technique of lymphoscintigraphy increases, more patients are found to have unexpected or aberrant sentinel lymph nodes draining cutaneous melanoma of the extremities, as well as the head, neck, and trunk. Data will be presented supporting the use of lymphoscintigraphy for patients with primary melanoma of intermediate thickness, in any location. Aberrant sentinel lymph nodes found in the epitrochlear, popliteal, and supraclavicular lymph node basins will be described. Intramuscular sentinel lymph nodes, also referred to as in-transit nodes, and their treatment will also be discussed. MATERIAL AND METHODS Thirty-two consecutive patients with cutaneous SURGERY 757
758 Lieber et al
Surgery October 1998
Fig 2. Lymphoscintigraphy of patient with primary melanoma of posterior foot. Knee is seen in profile with radioactive hot spot (white) just proximal to joint. This represents popliteal lymph node. Fig 1. Lymphoscintigraphy of patient with primary melanoma in dorsum of hand. Technetium sulfur colloid was injected in left hand at injection site. Radioactive hot spots (white) are seen just proximal to elbow and in axilla. Radioactivity near elbow is in epitrochlear region.
melanomas 1 mm or greater Breslow thickness and clinical stage N0M0 underwent wide local excision, lymph node mapping, and sentinel node biopsy from September 1994 to August 1997. Each patient was referred, already having tissue confirmation of the primary lesion (punch biopsy, wedge biopsy, or excisional biopsy), and each patient had undergone preoperative lymphoscintigraphy in the Department of Nuclear Medicine on the morning of the operation (Fig 1). Lymph node mapping was performed with 400 µCi unfiltered 99mTc-labeled sulfur colloid, which was injected in 4 aliquots surrounding the primary lesion. The Neoprobe 1000 hand-held gamma probe (Neoprobe Corp, Dublin, Ohio) was used during the operation to locate the sentinel lymph node and plan the excisional lymph node biopsy incision. Before skin preparation, isosulfan blue (Lymphazurin 1%; USSC, Norwalk, Conn) was injected around the primary lesion by the surgeon. Blue lymph nodes were excised as were any radioactive nodes that exhibited gamma counts significantly above background readings. Radioactive nodes were recognized by counts greater than 2 times the background. After excision, the lymph node bed was scanned to confirm that all radioactivity had been removed and counts had returned to background levels. Every sentinel lymph node was sent for permanent microhistologic evaluation consisting of step sectioning and
hematoxylin-eosin staining. Sentinel lymph nodes were considered aberrant if they were located outside of the traditional lymph node basin that would be included in a formal lymph node dissection for the respective primary lesion. Specifically, aberrant nodes included epitrochlear, supraclavicular, popliteal, or intramuscular lymph nodes. Based on the permanent histologic evaluation of the sentinel node, patients were advised regarding lymph node dissection and adjuvant therapy. RESULTS Thirty-two patients were enrolled in the database. Of these, 19 (60%) had cutaneous melanoma of the extremity and 13 (40%) had cutaneous melanoma of the head and neck or trunk. Two patients had recurrent melanoma. Mean Breslow thickness for all patients was 2.7 mm with a range from 0.8 to 7.0 mm. Sentinel lymph nodes were identified and biopsy was performed in all patients. Seven patients (22%) were found to have aberrant sentinel lymph nodes. Multiple sentinel lymph nodes were encountered in 15 (47%) of the patients studied, and the mean number of sentinel lymph nodes removed per patient was 2 (range 1 to 5). Positive sentinel lymph nodes were found in 4 patients (12.5%). Lymphadenectomy was performed in 3 of the 4 patients. Two of the positive sentinel lymph nodes were aberrant. One patient with a positive popliteal sentinel node also had 4 negative sentinel lymph nodes in the superficial inguinal basin (primary lesion was on the posterior sole of the right foot). This patient elected not to undergo a popliteal node dissection and had recurrent melanoma in the popliteal fossa. Another
Surgery Volume 124, Number 4
patient with a positive trapezius sentinel lymph node also had a positive posterior cervical node (primary lesion was located on the ipsilateral parietal scalp). This patient underwent radical neck dissection, with all nodes negative for metastatic melanoma, and has subsequently undergone resection for local recurrence but has no evidence of metastatic disease. Three of 9 patients with primary lesions of the upper extremity had lymph node drainage to nodes outside of the axilla. One patient with a primary subungual lesion of the right ring finger was found to have sentinel lymph nodes located in the epitrochlear lymph node basin and the ipsilateral axilla. Another patient whose primary lesion was located on the left elbow was found to have sentinel nodes in the bicipetal groove and the axilla. Final pathologic findings for both nodes were negative; however, the patient went on to have a recurrence in the ipsilateral axilla. A third patient with a primary lesion on the right proximal forearm had a sentinel node in the supraclavicular region and no sentinel nodes in the axilla. Two of 10 patients with lower extremity lesions had aberrant sentinel lymph nodes. Primary lesions were located on the sole of the foot and over the Achilles tendon. In both patients the aberrant sentinel lymph node was located in the popliteal nodal basin (Fig 2). The patient who underwent removal of a positive sentinel lymph node from the popliteal fossa who did not undergo popliteal node dissection was treated with α-interferon and changed to systemic chemotherapy when regional recurrence was confirmed. Fifteen percent of patients with primary lesions of the head, neck, and trunk were found to have aberrant sentinel lymph nodes. A patient with a primary lesion of the parietal scalp (mentioned above) was found to have 2 sentinel lymph nodes. One node was in the posterior cervical node chain (expected nodal basin); the other node was in the trapezius muscle (aberrant node). Both nodes were positive for metastatic disease. Radical neck dissection was performed and no further metastatic disease has been identified. Three sentinel lymph nodes were identified in another patient with recurrent melanoma of the low midback. Two of the sentinel nodes were located in the left axilla and the third was a node located very near the primary lesion (Fig 3). This patient continued to have negative nodes. DISCUSSION Lymphatic drainage of the extremity to the regional nodal basin is relatively more predictable
Lieber et al 759
Fig 3. Lymphoscintigraphy of patient with primary melanoma on left flank. Arms are raised over patient’s head. White spot marks left nipple (LT Nipple) anteriorly (ANT). Two small white spots are seen in region of left axilla. Technetium sulfur colloid was injected at large white spot (primary lesion). Small white spot representing intramuscular node is seen close to injection site.
than for the trunk. Unidirectional flow of lymphatics on the extremity implies that the sentinel lymph node for a cutaneous region will be located proximal to the primary lesion. Multidirectional flow exists on the trunk and head and neck.5 Sugarbaker and McBride7 expanded on Sappey’s anatomic description of lymphatic drainage of the trunk. Improved technology now expands the surgeon’s ability to know precisely where regional lymph node basins are located and where metastases are likely to be found. Lymphoscintigraphy further expands the area of ambiguous lymphatic drainage on the trunk, which extends like a belt and vest across the torso and shoulders and includes the head and neck.8 Regional lymph node basins for primary cutaneous lesions in the ambiguous zone are expected to be axillary, inguinal, or cervical. The ambiguity arises from an inability to predict anatomically which regional nodal basin will contain the sentinel node. Aberrant sentinel lymph nodes do not lie within expected lymph node basins and would be overlooked without lymphoscintigraphy. Aberrancy was encountered in 22% of patients in this series. Lymphoscintigraphy was performed for every patient, regardless of the location of the primary lesion. Aberrant nodes were discovered with similar frequency on the extremities as on the
760 Lieber et al
Surgery October 1998
Table I. Distribution of patients undergoing wide local excision for cutaneous melanoma Site
Patients (n)
Sex (F/M)
Age (yr)
Breslow thickness (mm)
SLNs (n)
4 9 9 10 32
2/2 2/7 4.5 9/1 17/15
46.8 (20-74) 47.9 (22-65) 54.7 (33-72) 55.7 (14-83) 52.1 (14-83)
1.77 (0.93-2.3) 2.31 (0.80-5.2) 3.38 (1.0-7.0) 2.80 (1.1-5.8) 2.70 (0.80-7.0)
2.8 (1-4) 2.0 (1-3) 2.1 (1-5) 1.5 (1-5) 2.0 (1-5)
Head and neck Trunk Upper extremities Lower extremeties Total
Age, thickness, and number of sentinel lymph nodes are expressed as mean (range). F, Female; M, male; SLNs, sentinel lymph nodes.
Table II. Distribution of location of sentinel lymph nodes by site of primary lesion Primary site Upper extremity (9) Lower extremity (10) Trunk (9) Head and neck (4)
Expected node
Aberrant node
Axilla (6) Groin (8) Axilla or groin (8) Cervical (3)
Supraclavicular (1) Epitrochlear (1) Intramuscular (1) Popliteal (2) Intramuscluar (1) Intramuscular (1)
Numbers in parentheses are numbers of patients.
trunk or head and neck (26% vs 15%). Gamma probe guidance during the operation enables the surgeon to plan precisely where to excise these nodes that would otherwise be difficult to locate through a small incision.9 Blue dye aids in identification of the sentinel lymph node.10 Isosulfan blue has a relatively short washout period, however, and a blue node could not be found in all cases.11 Aberrant sentinel lymph nodes have been described in the literature. Smith et al12 performed a retrospective chart review of 240 patients with melanoma of the upper extremity from 1955 to 1970. The measured end point of their study was clinical involvement of the epitrochlear lymph nodes. Seven (18%) of 39 patients with primary lesions distal to the elbow had epitrochlear metastases. They conclude that early dissection of clinically suspicious epitrochlear nodes improves patient outcome. Lymphoscintigraphy was not widely available for the patients of that study. Tanabe13 recently described the technique of epitrochlear lymphadenectomy and advocates selective lymph node dissection based on the status of the sentinel node. Regarding lymphatic drainage of the lower extremity, lesions of the sole and heel are most likely to drain into the popliteal nodal basin.14 Karakousis15 has described a technique for dissection of the popliteal lymph nodes when metastatic disease is present or strongly suspected. Although metastases to popliteal nodes are considered rare for cutaneous melanoma, 2 of 10 patients with lower extremity lesions in this series had sentinel lymph nodes in the popliteal fossa on
lymphoscintigraphy. One of these patients went on to have recurrence in this area. This unfortunate experience suggests that lymphadenectomy should be considered whenever and wherever a positive sentinel lymph node is detected. To locate these nodes, the combined modalities of lymphoscintigraphy, blue dye, and intraoperative gamma detection is important. Without the preoperative lymphoscintigraphy, unexpected nodal basins can be overlooked. Aberrant sentinel lymph nodes in locations not associated with any nodal basin (eg, muscle nodes) are easily excised with the gamma probe during the operation. Therefore we recommend the use of lymphoscintigraphy for every patient with a cutaneous melanoma 1 mm or greater in thickness, regardless of location. Biopsy should be performed in all sentinel lymph nodes, and when a positive sentinel node is identified, dissection of the regional nodal basin should be considered. REFERENCES 1. Balch CM, Ross MI. General principles of regional lymphadenectomy. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S, editors. Cutaneous melanoma. 2nd ed. Philadelphia: JB Lippincott; 1992. p 340-4. 2. Morton DL, Wen D, Wong JH, Economou JS, Cagle LA, Storm FS, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392-9. 3. Thompson JF, McCarthy WH, Bosch CMJ, O’Brien CJ, Quinn MJ, Paramaesvaran S, et al. Sentinel lymph node status as an indicator of the presence of metastatic melanoma in regional lymph nodes. Melanoma Res 1995;5:255-60. 4. Berger DH, Feig BW, Podoloff D, Norman J, Cruse CW, Reintgen DS, Ross MI. Lymphoscintigraphy as a predictor of lymphatic drainage from cutaneous melanoma. Ann Surg Oncol 1997;4:247-51.
Surgery Volume 124, Number 4
5. Balch CM, Lamki LM, Logic JR. Defining lymphatic drainage patterns with cutaneous lymphoscintigraphy. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S, editors. Cutaneous melanoma. 2nd ed. Philadelphia: JB Lippincott; 1992. p 368-74. 6. Balch CM, Milton GW, Cascinelli N, Sim FH. Elective lymph node dissection: pros and cons. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S, editors. Cutaneous melanoma. 2nd ed. Philadelphia: JB Lippincott; 1992. p 346-64. 7. Sugarbaker EV, McBride CM. Melanoma of the trunk: the results of surgical excision and anatomic guidelines for predicting nodal metastasis. Surgery 1976;80:22-30. 8. Norman J, Cruse CW, Espinosa C, Cox C, Berman C, Clark R, et al. Redefinition of cutaneous lymphatic drainage with the use of lymphoscintigraphy for malignant melanoma. Am J Surg 1991;162:432-7. 9. Krag DN, Meijer SJ, Weaver DL, Loggie BW, Harlow SP, Tanabe KK, et al. Minimal access surgery for staging malignant melanoma. Arch Surg 1995;130:654-8. 10. Pijpers R, Borgstein PJ, Meijer S, Hoekstra OS, van Haattum LH, Teule GJJ. Sentinel node biopsy in melanoma patients: dynamic lymphoscintigraphy followed by intraoperative gamma probe and vital dye guidance. World J Surg 1997;21:788-93. 11. Albertini JJ, Cruse CW, Rapaport D, Wells K, Ross M, DeConti R, et al. Intraoperative radiolymphoscintigraphy improves sentinel lymph node identification for patients with melanoma. Ann Surg 1996;223:217-24. 12. Smith TJ, Sloan GM, Baker AR. Epitrochlear node involvement in melanoma of the upper extremity. Cancer 1983;51:756-60. 13. Tanabe KK. Lymphatic mapping and epitrochlear lymph node dissection for melanoma. Surgery 1997;121:102-4. 14. Haagensen CD, Feind CR, Herter FP, Slanetz CA Jr, Weinberg JA, editors. The lymphatics in cancer. Philadelphia: WB Saunders; 1972. 15. Karakousis CP. The technique of popliteal lymph node dissection. Surg Gynecol Obstet 1980;151:421-3.
DISCUSSION Dr Edward G. Mansour (Cleveland, Ohio). You are bringing to our attention a very important observation that is essential to the successful implementation of lymphoscintigraphy and sentinel lymph node biopsy with intraoperative mapping in the management of malignant melanoma 1 mm or greater in thickness. Your findings clearly demonstrate the need of a well-informed and trained radiologist in nuclear medicine. The radiologist needs to work closely with a trained surgeon in scanning multiple draining sites and identifying the first-appearing sentinel lymph node, by using dynamic rapid sequence imaging every 30 seconds until the sentinel lymph becomes evident. This usually occurs after 5 to 7 minutes. It is also important to have the technical ability to obtain views that determine the lymph node in a 3-dimensional reconstructed static picture. The spot overlying the lymph node is then tattooed. You describe 2 melanomas on the sole of the foot and the Achilles tendon that drained to the popliteal region. I believe this area is the primary drainage basin for these sites as described by Haagensen. The identification of
Lieber et al 761
inguinal lymph nodes in a patient with a popliteal sentinel lymph node implies rapid transit of the radioactive material to the secondary basin, which is not a sentinel lymph node site. Similarly, the epitrochlear area is not an unusual drainage basin for the distal upper extremity. The patient with a parietal scalp melanoma manifested a sentinel lymph node in the posterior cervical region and a second lymph node in the “trapezius muscle.” It is important to define clearly whether the so-called trapezius lymph node was truly within the muscle or along its edge, consistent with a deep occipital lymph node, as is often encountered in a posterior neck dissection. Realizing that the “aberrant” draining areas to the epitrochlear, popliteal, and trapezius regions are truly the primary drainage basins of the melanoma sites described, I wonder whether you might consider the term uncommon rather than aberrant. In our series of 27 consecutive lymphoscintigraphies, intraoperative mapping, and sentinel lymph node biopsies, we encountered 2 patients with intramuscular sentinel lymph nodes associated with trunkal lesions and a third patient with an epitrochlear lymph node in a distal upper extremity melanoma. The 3 patients had secondary axillary drainage sites on which we elected not to perform a biopsy. Sentinel lymph nodes in the 3 cases were negative. I fully agree with your conclusions. I would like to reemphasize the need for adequate training for the surgeon and radiologist by attending accredited training sessions in this subject, to obtain the most accurate information and to minimize or eliminate the false-negative rate of the sentinel lymph node biopsy. I have 3 questions: (1) Do you recommend biopsy of an identified axillary or inguinal lymph node in a patient who underwent biopsy of a sentinel lymph node in either the epitrochlear or popliteal region, respectively? (2) What was the average elapsed time between the injection of the radioactive material and the surgical procedure? In your patient with the proximal forearm melanoma who had a supraclavicular sentinel lymph node, could the radioactivity have cleared out of the axilla? (3) Realizing the small number in your series, the rate of positive sentinel lymph nodes is 12.5%, lower than the rate observed in other series (20% to 25%). What are the pathologic techniques used to examine the sentinel lymph node? Are you using multiple sections of the lymph node with immunohistochemical staining for S100 and monoclonal antibodies? Dr Lieber. In response to your first question, these are sample photographs that were taken from a more dynamic imaging process. So, to be honest, we chose the best pictures to show. I would agree with you that drainage to the popliteal basin is normal and should be considered in the analysis of sentinel node mapping. Upper extremity lesions from the ulnar aspect of the hand have a greater likelihood of draining to the epitrochlear region and the posterior foot can drain to the popliteal basin, as Dr Mansour pointed out. The node in the parietal scalp lesion was in the body of the trapezius muscle; it was well beyond the edge of the muscle.
762 Lieber et al
Regarding your second question, biopsy should be performed on a radioactive node seen in the axilla or the inguinal region even if a popliteal or epitrochlear node is found. I do not think we can ignore these, especially if we do not know whether these nodes are in sequence or if they have 2 separate lymphatic patterns. The current sunbelt melanoma trial chaired by Dr Kelly McMasters encourages biopsy of epitrochlear/popliteal sentinel nodes, as well as axillary/inguinal sentinel nodes. The supraclavicular node that I mentioned was an unexpected finding because the lesion was an upper arm lesion very near the elbow. In response to your third question, our pathologists do not routinely use S-100 or any immunohistologic staining. We have not been able to convince them to do this. Dr Hiram C. Polk, Jr (Louisville, Ky). Because you made such a broad endorsement of lymphoscintigraphy and sentinel lymph node biopsy, I am sure you can give us the data on costs and charges that are extra in the operating room and radiology as a result of this procedure. Dr Lieber. Professional and technical fees for lymphoscintigraphy are $820.00. Dr Vernon K. Sondak (Ann Arbor, Mich). I want to expand on that, because your conclusion that every patient should undergo preoperative lymphoscintigraphy 1 day before the operation would add dramatically to the cost and inconvenience of lymphatic mapping. Clearly, as has been pointed out, some of your drainage is not aberrant but just unsuspected and revealed by the radioactive tracers you used. We use different agents used to do lymphoscintigraphy (ie, preoperative imaging of the direction of lymphatic flow, human serum albumin and intraoperative lymphatic mapping, and unfiltered sulfur colloid). Both are labeled with technetium, but there are differences in the cost and sensitivity that make each one appropriate for their respective roles.
Surgery October 1998
In many cases we have abandoned doing preoperative lymphoscintigraphy with technetium-labeled albumin in favor of simply having our nuclear medicine colleagues call us at the time they inject the technetium sulfur colloid immediately before the surgical procedure. In nuclear medicine they do exactly what Dr Mansour indicated: image the patient from 5 minutes to 35 minutes after the injection and tell us how many sentinel nodes there are and where they are located. During the operation, we then confirm with our gamma detector that we can identify the hot spot in the relevant area. In doing so, you should not be encountering unsuspected lymphatic drainage areas that you cannot find in the operating room, even if you have not done preoperative lymphoscintigraphy with labeled albumin. Preoperative lymphoscintigraphy without albumin is still indicated in our opinion for the patients with midline truncal lesions with ambiguous drainage. If we find such a patient to have lymphatic drainage to 3 or 4 basins, we believe it is an insurmountable clinical challenge to find all the sentinel nodes and therefore exclude such patients from consideration for intraoperative lymphatic mapping and selective lymphadenectomy. However, I think to advocate preoperative lymphoscintigraphy routinely as a separate procedure for patients with extremity lesions or even truncal lesions that are off to one side of the midline seems to be overkill. Dr Lieber. This was a surprising result that sentinel nodes were found in unusual locations, especially on the extremities. We expected those on the trunk, as you mentioned. If the surgeon is willing to look vigorously through the extremity with the gamma probe, I believe that these unusual locations can be identified without lymphoscintigraphic scanning. Preoperative scanning makes it easier to uncover any unexpected sentinel nodes.