Results

In 1992, Morton and associates reported their initial experience with selective lymphadenectomy [24,25]. They were able to identify a blue-stained sentinel node in 194 (82%) of 237 regional lymphatic drainage basins (Table 1). All 223 patients underwent complete lymph node dissection regardless of the pathology of the sentinel node. Of these specimens, 40 (21%) contained metastases in at least one lymph node. In only 2 of 194 (1%) lymph node dissection specimens were nonsentinel nodes, the exclusive site of regional metastasis. These results are quite remarkable considering the fact that in most cases, preoperative lym-phoscintigraphy was not used and the kinetics of the blue dye had not been well defined [29]. Yet, this early report demonstrates the strength of selective lymph-adenectomy as a staging procedure.

Occult regional metastases were identified by both standard Hematoxylin & Eosin (H&E) staining and newer immunohistochemical techniques. Fifty-seven percent of nodal metastases were found using conventional techniques; the remainder were identified by immunohistochemical staining alone [24,25]. Using immunohistochemical staining techniques with an antiserum to S-100 protein, Cochran and associates had previously demonstrated that 29% of lymph nodes stained negative with H&E actually contained metastatic melanoma [30,31]. The 3338 lymph nodes excised in the selective lymphadenectomy patients were stained with the melanoma-specific murine monoclonal antibody NKI/C3 to confirm the presence of melanoma cells. They found few additional metastases with serial sectioning of the nodes as compared to just examining the bivalved faces. The role of additional sectioning of the sentinel nodes is unknown [32]. Whereas newer molecular biology techniques looking for specific gene sequences particu-

Table 1 Initial Experience with Selective Lymphadenectomy for Early-Stage Melanoma: Distribution of Metastases in Sentinel and Second-Tier Nodes

N

%

Total lymphadenectomies

237

100

Lymphadenectsmies with identified sentinel nodes

194

82

Lymphadenectsmies with tumor in nodes

40

21

Lymphadenectomies with tumor in sentinel nodes

38

20

Lymphadenectomies with tumor in second-tier nodes (exclusively)

2

<1

Total lymph nodes

3338

100

Total sentinel nodes

259

8

Sentinel nodes with tumor

47

18

Total second-tier nodes

3079

92

Total second-tier nodes with tumor (exclusively)

2

<0.1

lar to melanoma may further enhance the sensitivity of detecting metastases in the sentinel node [33-35], their role in the routine management of melanoma is also unknown.

Selective lymphadenectomy is a relatively difficult procedure, but its learning curve is steep. During his initial 58 cases, Morton identified only 81% of blue-stained sentinel nodes. However, during the next 58 cases, his rate of sentinel node identification increased to 96%, and it now approaches 100%. The surgeon with the most experience with the procedure achieved an early success rate of 96%, whereas the surgeon with the least experience had the lowest level of success: 72% (p < 0.01). The gradual improvement in the rate of sentinel node detection is partially based on the increased experience with the technique. We have found that the blue-stained afferent lymphatics and nodes can be difficult to identify. Most surgeons have little experience dissecting lymphatic channels prior to ever performing a selective lymphadenectomy. We have found that patients who have undergone wide excision of the primary with margins of more than 1.5 cm or have had any procedure that disrupts the lymphatic drainage are not candidates for selective lymphadenectomy [24,25,36].

The routine use of preoperative lymphoscintigraphy in all cases has played a significant role in decreasing the incidence of missed sentinel nodes. Glass and associates from the John Wayne Cancer Institute [28] recently reviewed their experience with lymphoscintigraphy using three commonly employed radiophar-maceuticals: 99mTc-labeled albumin colloid, 99mTc sulfur colloid, and 99mTc-HSA. They compared the three agents for their utility to identify the afferent lymphatics and sentinel lymph nodes. Using early (up to 30 min) images, the three agents were equally effective for identifying the sentinel nodes. On average, two lymph nodes were identified in each basin. When they delayed their images up to 4 h after injection of the radiopharmaceutical, the average number of nodes visualized did not change significantly. There appeared to be a wide variation in the number of lymph nodes seen on lymphoscintigraphy from patient to patient: 99mTc-labeled albumin colloid (range: 1-7), 99mTc sulfur colloid (range: 1-14), and 99mTc-HSA (range: 0-9).

Yet more importantly, in more than half of the cases, we reviewed, more than one ''sentinel'' node was identified, often in basins or sites which would not be expected by strict anatomic definition [37]. Because the blue dye appears to travel along the same pathways as the radiopharmaceuticals, lymphoscintigra-phy is critical for the accuracy of the selective lymphadenectomy procedure.

Morton's initial experience with selective lymphadenectomy also demonstrated the technical differences between dissection in the groin, neck, and axillary basins [24,25]. Sentinel nodes were easier to identify in the groin (89% accuracy) and increasingly more difficult in the neck (81% accuracy) and axilla (78% accuracy). In 1993, Morton's group reported their experience with selective lymphadenectomy for melanoma of the lower torso and extremities that drained to the groin basin [38]. One hundred twenty-eight patients had selective lymph-adenectomy performed. Preoperative lymphoscintigraphy was used only for non-extremity primaries. Sentinel nodes were identified in 96% of the 51 patients who had complete groin dissections and in 98% of the next 77 patients having had a selective lymphadenectomy alone. In less than 1% of these patients, the sentinel node was free of disease while other (nonsentinel) nodes were involved. Even with this high rate of success with selective lymphadenectomy in this basin, we now employ preoperative lymphoscintigraphy in all cases. In 12% of selective lymphadenectomy procedures, lymphatic drainage was to two lymph node basins, including those rare primaries on the calf or foot that also drained to the popliteal region. In most cases, a single sentinel node was identified just inferior to the inguinal ligament. However, some of the sentinel nodes were located at the apex of the femoral triangle and occasionally two sentinel nodes were identified, usually on opposite sides of the femoral vein. Although lymphoscintigraphy may be considered unnecessary for some primaries on the lower extremities, the routine use of this procedure helps to identify the occasional aberrant lymph node in the groin or popliteal basins.

In 1993, Morton's group also reported their experience of selective lymphadenectomy for head and neck melanoma that drained to the cervical nodes [39]. All patients had preoperative cutaneous lymphoscintigraphy. At the time of surgery, blue dye alone was used to identify the sentinel nodes. The sentinel node was found in 71 of the 79 cervical drainage basins (90%). Most of the missed sentinel nodes were from the occipital, postauricular, or parotid basins, where the blue dye is difficult to identify. There were no regional recurrences in those patients with tumor-negative dissections during a mean follow-up of 27 months. Although preoperative lymphoscintigraphy was used in all cases, this early experience with selective lymphadenectomy demonstrated the intrinsic difficulty with the cervical basin. The lymphatic drainage from the head and neck is difficult to determine from the anatomical location of the primary [20]. As our own experience suggests, sentinel nodes in the midst of the parotid gland, deep in between the neck muscles or adjacent to the numerous facial veins, are difficult to identify. We have recently updated our experience with selective lymphadenectomy for early-stage melanoma of the head and neck employing both blue dye and radio-pharmaceutical. Bostick and associates reported on 117 patients undergoing selective lymphadenectomy with either blue dye alone (94 cases) or in combination with a radiopharmaceutical (23 cases) for probe-directed selective lymphadenectomy [40]. The accuracy rate for blue-dye-directed selective lymphadenectomy was 92% (only slightly better than our earlier report) but improved to 96% with the combination of blue dye and radiopharmaceutical. The probe was helpful for identifying sentinel nodes in the difficult sites such as the postauricular, occipital, and parotid basins. Ten percent of patients had drainage to two basins. Eighty-nine percent of the patients avoided complete neck dissection after undergoing a tumor-negative selective lymphadenectomy. There have been no regional lymph node recurrences over a median follow-up of 46 months (range: 1-125 months) [40]. Our improved accuracy rate probably relates not only to our increased experience with selective lymphadenectomy for this basin but also to the use of the hand-held gamma probe. Although the probe helped improve the accuracy rate of selective lymphadenectomy, the background radioactivity in adjacent lymph nodes in the cervical basin can lead to the removal of excessive (second-tier) nodes.

In Morton's initial experience, the axilla was the most difficult of basins in which to identify the sentinel lymph node. The anatomy of the axilla prevents the nuclear medicine physician from marking the site of the sentinel node with the patient under the gamma camera, even with the patient positioned for surgery. With a hand-held gamma detection probe, this is considerably easier. We have also found that drainage patterns to the axilla vary greatly, especially from primaries on the chest and back.

In order to improve on the accuracy rate and diminish the learning curve for selective lymphadenectomy, we explored the technique of selective lymphad-enectomy directed by a radiopharmaceutical and a gamma detection probe [41]. A gamma-probe-guided directed biopsy was first performed using the combination of blue isosulfan dye injected intraoperatively with 0.5-1.0 mL or 18.5 MBq (0.5 mCi) of 99mTc-HSA (Amersham) at the primary site. A hand-held gamma counter (Neoprobe 1000, Neoprobe Corp., Dublin, OH) was used to follow the radioactive tracer to the regional basin. Morton's group tested this approach initially in 30 melanoma patients. Thirty-four lymph node basins were identified by preoperative lymphoscintigraphy. At least one sentinel node was identified in each basin. The blue dye identified 36 sentinel nodes, and the gamma probe detected all 36 nodes plus an additional 6 nodes. Overall, blue-stained sentinel lymph nodes had a roughly twofold higher radioactive count rate than adjacent nonblue nodes, and up to an eightfold higher count rate than the lymph basin background. Although none of the sentinel nodes contained metastatic disease, this study demonstrated the utility of the hand-held gamma counter to help identify blue-stained sentinel nodes and the close concordance between the findings from blue dye and the radiopharmaceutical. One of the difficulties with the combined technique is the logistics of injecting a radiopharmaceutical in the operating room.

Bostick and associates recently reviewed the John Wayne Cancer Institute experience with gamma-probe-guided selective lymphadenectomy in 100 lymph node basins from 87 patients [42]. All patients underwent preoperative lympho-scintigraphy with one of the three radiopharmaceuticals commonly used in the United States. Selective lymphadenectomy was performed with either concurrent injection of blue dye and 99mTc-labeled human serum albumin or the 99mTc sulfur colloid injected up to 4 h prior to the operative procedure. One hundred thirty-

six blue-stained and radioactive lymph nodes and eight additional non-blue-stained but ''hot'' nodes were removed in 98 lymph node basins (success rate 98%) (Table 2). A hand-held gamma probe was used to determine the radioactive counts over the blue nodes, adjacent nonblue nodes and an irrelevant background site. Ninety-two percent of the blue-stained lymph nodes had an in vivo count to background ratio of more than 2, and 87% had in vivo count ratios exceeding 3. Seventeen sentinel nodes from 15 basins contained metastases: 16 were located with blue dye and gamma probe and 1 was found with blue dye alone. None of the tumor-positive lymph nodes was identified with the gamma probe alone. Using the definition of a radioactive sentinel node as having an in vivo count ratio of more than 2 compared to the background, a success rate of 85% would be achieved. When the sentinel-node-to-background ratio was increased to a minimum of 3 to improve the specificity of the technique, the success rate decreased to 78% (Fig. 3). The concordance between the two techniques was not 100%. Not all blue-stained lymph nodes have an elevated count ratio and, conversely, not all nodes with an elevated count ratio are blue. In fact, when the in vivo count ratios for all the blue-stained lymph nodes were examined, a wide variation ranging from less than 1 to over 100 was noted. Similar results were observed when the ex vivo count ratio of the nodes was examined, suggesting that the

Table 2 Results from Blue-Dye-Directed and from Radiopharmaceutical-Directed Selective Lymphadenectomy: Lymph Node Basins and Sentinel Lymph Nodes

Lymph node basins Total

Sentinel node identified Sentinel node blue

Sentinel node blue and radioactive (ratio > 2) Sentinel node blue and radioactive (ratio > 3) Sentinel node blue only (ratio <2) Sentinel node radioactive only (ratio > 2) Sentinel node radioactive only (ratio > 3)

Sentinel nodes Total Blue

Radioactive (ratio > 2) Radioactive (ratio > 3) Blue only (ratio <2) Blue and radioactive (ratio > 2) Blue and radioactive (ratio > 3) Radioactive only (ratio > 2)

144 136 132 125 12 124 117 8

In Vivo Ratio

Figure 3 In vivo count ratios of blue-stained sentinel lymph nodes from 100 lymph basins. In vivo count ratios were calculated by determining the counts overlying the blue-stained sentinel nodes and comparing these counts to an irrelevant background site. Of sentinel nodes, 85% had an in vivo count ratio of more than 2 and 78% had a ratio exceeding 3.

In Vivo Ratio

Figure 3 In vivo count ratios of blue-stained sentinel lymph nodes from 100 lymph basins. In vivo count ratios were calculated by determining the counts overlying the blue-stained sentinel nodes and comparing these counts to an irrelevant background site. Of sentinel nodes, 85% had an in vivo count ratio of more than 2 and 78% had a ratio exceeding 3.

radiopharmaceuticals alone can be misleading for sentinel node dissection. Both the radiopharmaceuticals were found to give similar count ratios for radio-lymphoscintigraphy and led to surgical excision of similar numbers of lymph nodes. At our center, we have little difficulty with performing lymphoscintigra-phy and selective lymphadenectomy on the same day, but, logistically, this approach can be difficult. We are currently examining the use of 99mTc sulfur colloid administration for next-day selective lymphadenectomy.

One of the theoretical advantages of selective lymphadenectomy over conventional staging of the regional lymph nodes is the potential reduction in morbidity and cost compared to elective lymph node dissection. In 1997, Essner and associates reported their updated experience with selective lymphadenectomy for patients with melanoma of the torso and lower extremities that drained to the groin [43]. Fifty-two patients all underwent selective lymphadenectomy followed by complete groin dissection. Only a select portion of the patients underwent lymphoscintigraphy. A subsequent group of 114 consecutive patients all under went lymphoscintigraphy followed by selective lymphadenectomy, with complete groin dissection only if the sentinel node contained metastasis. Three patients had bilateral drainage patterns. The average primary tumor thickness was 2.08 mm. The average number of lymph nodes removed for the 114 patients undergoing selective lymphadenectomy alone was 4.2 ± 6.5, and for those undergoing selective and complete node dissection, it was 13.0 ± 6.6 (p < 0.05). Seventy-three percent of patients having a selective lymphadenectomy alone had only one lymph node excised. The frequency of tumor positive dissections was similar in the two groups. Eighty percent of patients had one tumor-positive lymph node and additional tumor-positive lymph nodes were found in 20% of cases during completion lymph node dissection. We found no difference in survival between patients treated by either of the two approaches. Patients with tumor-negative dissections had a 5-year survival of 89.5 ± 5% following selective lymphadenectomy alone and 93 ± 3% after combined selective and complete node dissection. To determine the differences in costs of the two operative procedures, we compared both the length of in-patient stays and the total estimated hospital costs. Patients undergoing selective lymphadenectomy had an average hospital stay of less than 1 day. Patients treated by complete groin dissection had an average hospital stay of 6.05 days (p = 0.01) (Fig. 4). We included in our

# patients

100-

# patients

100-

Figure 4 Length of hospital stay following selective lymphadenectomy (SLND) or complete groin dissection. Hospital stays were determined for patients treated by either selective lymphadenectomy or complete groin dissection. Most patients undergoing selective lymphadenectomy had less than 1 day of hospitalization. Patients with skin grafts usually required longer hospitalization.

Figure 4 Length of hospital stay following selective lymphadenectomy (SLND) or complete groin dissection. Hospital stays were determined for patients treated by either selective lymphadenectomy or complete groin dissection. Most patients undergoing selective lymphadenectomy had less than 1 day of hospitalization. Patients with skin grafts usually required longer hospitalization.

calculation the hospital, nuclear medicine services, and pathology services (Table 3). Although the total costs will vary from institution to institution, our results represent the relative cost differences between these procedures at our community cancer center. When selective lymphadenectomy was performed under local anesthesia in our clinic, the cost was only $1,018. When the same procedure was performed through the out-patient operating room with participation of an anesthesiologist (either general or regional anesthesia), the costs increased to $7,150. Both superficial and deep groin dissections were even more expensive, as the costs increased substantially with patient admission to the hospital. The costs of the pathology evaluation were higher for selective-lymphadenectomy-treated patients only because of the routine use of immunohistochemical staining of the lymph nodes. Although this study focused on patients having selective lymphade-nectomy of the groin basin, we would expect similar cost savings for patients having selective lymphadenectomy of the axillary or cervical basins. This study did not focus on the morbidity of the two operative procedures, but we expect that patients treated by the less extensive selective lymphadenectomy would have a lower risk of complications than those who also underwent the more radical dissection [44,45].

Because selective lymphadenectomy has virtually replaced elective lymph node dissection for staging the regional lymph nodes, we chose to compare the therapeutic value of selective lymphadenectomy to elective lymph node dissection. In 1998, Essner and associates performed a method pair statistical analysis to compare the outcome of 534 early-stage melanoma patients [46-48]. Half were treated by selective lymphadenectomy and the other half by elective lymph node dissection. Patients were matched from our computer-assisted database by age (54% over 50 years of age), gender (63% male), site of the primary (49% extremity, 36% on the trunk, and 15% on the head and neck), and thickness of the primary (7% <0.75 mm, 42% 0.75-1.5 mm, 43% 1.51-4.0 mm, and 8%

Table 3 Estimated Hospital Costs for Patients Undergoing Groin Dissections (U.S. Dollars)

Nuclear

Hospital medicine Pathology Total

Out-patient clinic SLNDa 833

Out-patient ORb SLND 6,965

In-patient OR superficial groin dissections 8,274

In-patient OR superficial and deep groin 11,929 dissections

54 54

131 131 85 170

1,018 7,157 8,359 17,999

a SLND = selective lymph node dissection. b OR = operating room.

>4 mm). Patients treated by selective lymphadenectomy only had a complete lymphadenectomy if the sentinel node contained metastasis. Overall, the incidence of lymph node metastases was not different between selective lymphadenectomy (15.7%) and elective lymph node dissection (12%) groups. Yet, the incidence of nodal metastases in patients with intermediate thickness primaries (1.51-4.0 mm) was significantly higher when treated by selective lymphadenectomy (23.7%) instead of elective lymph node dissection (12.2%) (p = 0.025). Although unexplained, this difference may relate to the improved staging accuracy of selective lymphadenectomy over elective lymph node dissection. Selective lymphadenectomy and elective lymph node dissection carried equivalent 5-year rates of disease-free survival (79 ± 3.3% and 84 ± 2.2%, respectively, p = 0.25) and overall survival (88 ± 3.0% and 86 ± 2.1%, respectively, p = 0.98).

Groups of 225 selective lymphadenectomy and 235 elective lymph node dissection patients had tumor-negative dissections. Overall 5-year survival was identical for the two treatments (p = 0.37). Twenty-six selective lymphadenectomy patients (11.5%) and 35 elective lymph node dissection patients (14.9%) have recurred over a median follow up of 45 and 169 months, respectively. Eleven of the recurrences after selective lymphadenectomy were confined to the dissected basin. We have since reexamined the original pathology of these sentinel nodes. In four (36%) cases, metastases were located in the sentinel node. The true surgical false-negative sentinel node rate is 3%. Among the 35 patients to have recurrences following elective lymph node dissection, five were in the dissected lymph node basin. The dissected basin recurrence rate (2%) was the same as in selective lymphadenectomy.

Fifteen of the 26 (58%) selective lymphadenectomy and 12 of the 35 (34%) elective lymph node dissection recurrences occurred within 2 years of lymph node dissection. We calculated the yearly probability of recurrence following the two techniques. Seven percent of selective lymphadenectomy patients recurred within 2 years compared to 5.1% of elective lymph node dissection patients. Although we have a relatively short follow-up for selective-lymphadenectomy-treated patients, these results demonstrate that selective lymphadenectomy and elective lymph node dissection are therapeutically equivalent procedures. We anticipate that longer follow-up of our selective-lymphadenectomy-treated patients will help to determine the true recurrence rates following a tumor-negative dissection [49].

Was this article helpful?

0 0
How To Prevent Skin Cancer

How To Prevent Skin Cancer

Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.

Get My Free Ebook


Post a comment