Individualized Surgery: Gamma-Probe-Guided Lymphadenectomy in Patients with Clinically Enlarged Lymph Node Metastases from Melanomas

Between May 1998 and May 2011, 115 consecutive patients with clinically evident regional lymph node metastases from melanomas were treated at the University Medical Center in Göttingen. Of these, 25 had developed nodal recurrence following negative SLNB. There was no history of primary melanoma and no primary tumor could be located in 14 additional patients. The remaining 76 patients, who are the subject of the present study, had a known primary melanoma but no previous lymphoscintigraphy. In 29 patients, the primary melanoma was diagnosed at the same time as palpable metastases. Nodal recurrences after primary tumor excision were noted in 47 patients.

Magnetic resonance imaging of the head and neck were undertaken in patients with cervical metastases. All patients with inguinal nodal metastases received computed tomography scans in order to detect enlarged nodes in the lesser pelvis. The size of the lymphadenopathy was determined by preoperative ultrasound B-scans. In the majority of cases, fine-needle aspiration cytology was performed. In three patients who had undergone diagnostic metastasectomy, the size of the macrometastases was taken from the pathology report. Patients with clinical evidence of systemic metastases were not considered in the present study.

From November 2000, 34 selected patients received preoperative lymphoscintigraphy. In this group, two patients were actually referred for SLNB but macrometastasis >1 cm was diagnosed on preoperative ultrasound B-scans. The remaining 32 patients had clinically palpable disease.

Lymphatic mapping was deemed necessary: (1) if ambiguous lymphatic drainage from the primary melanoma site was conceivable or (2) if some kind of less radical lymph node dissection had been planned. In such cases, we aimed to excise all clinically unsuspicious but radioactive lymph nodes, even if they were situated beyond the borders of the previously designated node dissection field. All patients gave informed consent before undergoing gamma-probe-guided lymphadenectomy.

Patients with unknown primary melanoma sites as well as the false-negative cases after initially negative SLNB did not receive lymphatic mapping. Lymphatic mapping was also not carried out for patients with clinically evident metastases in two nodal basins, for patients with enlarged pelvic metastases, or for patients with grossly enlarged, fixed or matted nodes. Patients who had a primary tumor excision requiring reconstruction of the defect using skin flaps did not undergo lymphoscintigraphy. Patients with previously excised in-transit metastases were also not considered eligible for lymphatic mapping. We did include, however, patients with synchronous, surgically amenable in-transit disease. We also included three patients on whom diagnostic excision of the macrometastasis had been performed because we felt that bidirectional lymphatic drainage from the primary tumor site could still be detected.

Technically the lymphoscintigraphy did not differ from those applied on our patients with clinically unsuspicious lymph nodes.5,6 With a PICKER SX 100—a broad-view gamma camera equipped with a low-energy high-resolution collimator—preoperative dynamic lymphoscintigraphy utilizing dynamic acquisition during the first 30 min and static imaging after 1–2 h was performed on each of the patients. Approximately 18–24 h before the operation, 100 MBq of 99mTc-human albumin (Nanocoll; Nycomed Amersham Sorin) dissolved in a volume of 0.1–0.2 ml was injected into the dermis surrounding the primary melanoma or the biopsy scar. This relatively high dose was chosen to enhance the imaging process of the afferent lymphatics or deeply situated tumor-draining lymph nodes, e.g., iliac, subcostal or parasternal nodes. Static images were taken anteriorly and laterally 30 min and 2 h after injection.

Local excision with adequate safety margins was the standard treatment procedure of primary melanomas. Well-established standard surgical techniques of TLND were considered as standard of care (modified neck dissection, axillary dissection including nodal levels I–III with preservation of the pectoralis minor muscle, ilioinguinal dissection). Our surgical approaches of TLND have been previously described.8,9 An abdominoperineal rectum extirpation along with the enlarged paraproctic node metastases was performed on one patient.

Generally, our approach was rather conservative in patients with increased general morbidity, with preexisting swellings of extremities or severe adiposity, with metastasis to more than one nodal basin, or in patients with preceding in-transit metastases. The latter group has been shown to have high risk of nodal basin recurrence despite thoroughly performed TLND.8 We distinguished between two types of less radical node dissections: (1) less radical but well-standardized dissections (selective neck dissection, dissection of axillary levels I–II only or exclusively inguinal dissection without pelvic dissection) and (2) even more limited, nonstandardized lymph node excision (LLND), which sometimes seemed to be appropriate for the reasons mentioned above. In our patients who underwent lymphatic mapping, every less radical operation included the removal of all radioactively labeled lymph nodes. During gamma-guided lymph node excision, a handheld gamma probe was used (Gamma Finder; W.O.M. World of Medicine AG, Ludwigsstadt, Germany). At the end of TLND, the borders of the node dissection area were generally checked for remaining radioactivity to ensure that all tumor-draining lymph nodes had been excised.

In cases displaying bidirectional lymphatic drainage, the radioactive lymph nodes located outside the clinically involved basin were excised in the same routine manner as in an SLNB. When micrometastasis in an additional nodal basin was diagnosed, no further TLND was performed.

Primary tumors were examined using routine histological methods. The SLNs excised from an additional nodal basin outside the clinically involved basin underwent step sections as previously described.10 Immunohistochemical staining was performed using the streptavidin–biotin complex method using alkaline phosphatase as the labeling enzyme and fast red chromogen as the substrate (detection kit K5005; Dako, Germany). The following antibodies were used: S-100 (clone S-100, dilution 1:3,000; Dako), HMB-45 (clone HMB45, dilution 1:200; Dako), MART-1 (clone A 103/M2-7C10/M2-9E3, dilution 1:200; Zymed, USA), and Pan-Melanoma Cocktail (clone HMB45þM2-7C10þM2-9E3þT311, dilution 1:300; Biocare Medical, USA) stained by an auto-immunostainer (Immunostar 80, Shandon Varistain 24-4, Germany).

Immediately after the surgical procedure, the lymph nodes were separated from the TLND specimen. Using the gamma probe it was possible to detect low amounts of radioactivity by touching the excised lymph nodes with the tip of the probe. Lymph nodes were considered radioactive whenever more than 4 counts were measured ex vivo in the absence of any background radiation.

Depending on their size, the lymph nodes from TLND specimens were sliced into two to four sections, each of which was embedded separately in paraffin. From each slice four microtome samples were produced and stained using hematoxylin and eosin (H&E), as well as immunohistochemical staining with anti-protein S-100 serum, MART-1, and anti-HMB-45.

Patient data including clinical and lymphoscintigraphic parameters, as well as histopathological results were entered routinely into an electronic database. For the present analysis descriptive statistics were applied. The chi-square test was used to compare the probabilities of metastatic involvement for radioactive and nonradioactive lymph nodes. The proportions of metastatic involvement of both radioactive and nonradioactive lymph nodes were also determined for each patient, and the means were compared using the t-test for dependent samples. The difference in the means was characterized with a 95 % confidence interval. Analyses of survival and relapse rates were performed using Kaplan–Meier estimates. The significance level was set at α = 5 %.

The characteristics of the patients displaying clinically enlarged metastases who were included in this study are summarized in Table 1. The majority of the patients undergoing gamma-guided lymphadenectomy (62 %) had a primary melanoma and clinically enlarged nodal metastases at the same time, whereas the majority of the patients without lymphatic mapping (81 %) had delayed node dissection of nodal recurrences. Due to this imbalance, the patients with lymphatic mapping tended to have more aggressive primary tumors. Furthermore, our selection criteria imply a lower nodal tumor burden in the group with gamma-guided surgery, although this cannot be statistically proven through the number of pathologically involved lymph nodes.

In the group without lymphoscintigraphy, six patients received some type of less radical lymphadenectomy: One patient had selective neck dissection; two patients with bilateral axillary metastases received axillary LLNDs (sparing the level III nodes, the lateral nodes, and the lymphatic on the ventral side of the major vessels); three patients received an inguinal lymphadenectomy without pelvic dissection.

Of the 34 patients with gamma-guided lymphadenectomy, 26 received a standard TLND (8 of them had lymphatic drainage to ectopic SLNs). The remaining eight patients received some form of less radical lymphadenectomy: One patient had axillary dissection of the first two levels only; three patients received an exclusively inguinal lymphadenectomy, and five had nonstandardized LLNDs. Of the eight patients with less radical procedures, seven had ectopic SLNs. Some patients had more than one reason for restricting the procedure, such as in-transit metastases, significantly increased general morbidity, or bilateral node excision. Importantly, all radioactive lymph nodes were excised in all less radical procedures. The percentages of patients undergoing less radical procedures in the groups with and without lymphatic mapping were 24 and 14 %, respectively. As a result, the mean number of excised lymph nodes was lower in the lymphatic mapping group (14 versus 20 nodes).

The greatest diameter of the largest metastasis ranged from 1.1 to 6.5 cm (median 3.0 cm). Lymphoscintigraphy detected at least one radioactive node in 30 of the 34 clinically involved basins. All but one patient displayed at least one radioactive lymph node; the average was five radioactive nodes per patient. In the surgical specimens, the largest macrometastasis was radioactive in 24 of the 31 cases available for analysis. The median number of excised radioactive lymph nodes was 5 (range 0–15).

Most of the primary tumor sites were suggestive for bivalent lymphatic drainage; 16 were situated near the midline of the body, whereas 15 were located in a border region between two ipsilateral nodal basins. Lymphatic drainage to a second (ectopic) nodal basin included the axilla (n = 8), the parasternal nodes (n = 1), the groin (n = 4), the lesser pelvis (in the absence of drainage to the ipsilateral groin, n = 1), and the popliteal fossa (n = 1). Of the 15 additional nodal basins detected through lymphoscintigraphy, 14 were surgically explored. The corresponding SLNs were successfully excised in 13 patients (Table 2).

SLNs situated in a separate (ectopic) nodal basin were involved with micrometastasis in six cases (four patients with simultaneous excision of primary melanoma and lymph node metastases and two patients with delayed lymph node dissection for a palpable recurrence).

The metastatic disease was restricted to the radioactive nodes in 15 of the 29 patients undergoing standardized TLND. Patients with metastases exclusively within radioactive nodes had a significantly lower number of lymph node metastases as compared with patients with negative nonradioactive nodes (2.9 ± 1.9 versus 5.0 ± 4.3 node metastases, P = 0.03). In five patients (one of whom did not display any radioactive nodes), the metastasis was restricted to the nonradioactive nodes. Overall, we excised 172 radioactive lymph nodes; 71 (41.2 %) were determined to be pathologically positive. Of the 321 nonradioactive lymph nodes excised, 55 (17.1 %) were pathologically positive. The proportion of tumor involvement for radioactive and nonradioactive lymph nodes was also calculated for each patient. The resulting mean probabilities were 44.5 and 16.9 %, respectively [P = 0.00002, difference 27.6 % (95 % confidence interval 12.3–43.1 %)]. Thus, the radioactively labeled nodes carried a significantly higher risk of being tumor-involved.

Local recurrence after lymphadenectomy was defined as any evidence of recurrent disease within the surgical basin harboring the macrometastasis, including relapses after generalized metastasis. The recurrence and survival rates for the patients with and without lymphoscintigraphy are shown in Fig. 1. The two groups are not directly comparable because our selection criteria appeared to favor a higher nodal tumor burden in the control group. However, a local recurrence rate of about 10 % in the lymphoscintigraphy group seems to be an acceptable result for a group with exclusively palpable nodes and a mean Breslow thickness of 5.8 mm. Of the patients who underwent some kind of less radical gamma-guided dissection, two recurred within the node dissection field. One of them had initially received excision of in-transit metastases and LLND of cervical macrometastases. In this patient, the disease recurred with the same metastasis pattern. The second patient had a superficial inguinal node dissection plus excision of contralateral pelvic SLNs. This disease recurred within the scar of the inguinal dissection simultaneously with hepatic metastasis. We did not observe any recurrences in any of the 13 ectopic basins. However, one patient diagnosed with initial lymphatic drainage to only one axilla showed recurrence in the contralateral axilla. Thus, in the group with gamma-guided lymphadenectomy, 20.6 % of the patients showed metastasis to more than one nodal basin.

In the group without lymphoscintigraphy, 9.5 % of the patients showed metastasis to two different nodal sites: One patient had bilateral cervical metastasis; two patients had bilateral axillary metastases. A further patient with a primary melanoma of the calf and enlarged inguinal and iliac metastases developed an isolated recurrence in the fossa poplitea.