RECONSTRUCTIVE SURGERY

Sentinel Lymph Node Biopsy for Melanoma Is There a Correlation of Preoperative Lymphatic Mapping With Sentinel Lymph Nodes Harvested? Kristen Ann Hudak, MD,* Kevin E. Hudak, MS, MD,Þ and William W. Dzwierzynski, MD* Background: Nodal status is the most significant prognostic factor in melanoma. No study has examined the relationship between lymphoscintigraphy, F probe counts, harvested nodes, and nodal status. Methods: Two-hundred sixty two patients were identified who underwent sentinel lymph node biopsy for melanoma between 2001 and 2010. Clinicopathologic and treatment information was collected. The number of lymph nodes and basins demonstrated on lymphoscintigraphy was compared to those at surgery. F Probe counts were compared. Results: Median age was 54.5 years (range, 18Y90 years) with 52.3% male. Average Breslow depth was 2.0 (1.9) mm; 99.6% of lymphoscintigraphy studies identified at least 1 basin, 80% showed only 1 (range, 0Y4). Lymphoscintigraphy identified on average 1.5 (0.9) sentinel nodes and 31% with secondary node. Surgery excised on average 2.6 (1.4) nodes involving 1.2 (0.5) basins; 17.6% had a positive sentinel lymph node. There was no difference in the sum or average of F counts between positive and negative sentinel lymph node groups (P = 0.2, P = 0.5). When comparing lymphoscintigraphy and surgical excision, the correlation of lymphatic basins was r = 0.67 and of lymph node numbers was r = 0.33. Conclusions: Lymphoscintigraphy should be used to identify the proper lymphatic basins for a sentinel node procedure, however, the removal of nodes must continue until the background count is less than 10%. The correlation of lymph node number identified on lymphoscintigraphy to surgical excision is weak. F Probe counts cannot be used to differentiate positive from negative nodes and the positive lymph node is not always the hottest node. Key Words: melanoma, lymphoscintigraphy (Ann Plast Surg 2015;74: 462Y466)

M

elanoma is a rapidly increasing cancer, with an estimated 76,250 new diagnoses and 9180 deaths in 2012.1 With nodal status being the most significant prognostic factor in patients with melanoma, adequate staging is important.2Y7 NCCN guidelines recommend sentinel lymph node biopsy for patients with thickness greater than 0.75 mm or those with high risk features such as ulceration, mitotic figures, lymphovascular invasive, or Clark level 4 or 5 (invasion of reticular dermis or subcutaneous tissues).5 To identify the sentinel lymph node or first lymph node to receive lymphatic drainage from a tumorVlymphoscintigraphy is performed preoperatively. This involves 4 to 8 intradermal injections

Received February 15, 2013, and accepted for publication, after revision, June 21, 2013. From the Departments of *Plastic Surgery, and †General Surgery, Medical College of Wisconsin, Milwaukee, WI. Conflicts of interest and sources of funding: This work was supported by resources of the Department of Plastic Surgery, Medical College of Wisconsin. Reprints: William W. Dzwierzynski, MD, Department of Plastic Surgery, Medical College of Wisconsin, 8700 Watertown Plank Rd, Milwaukee, WI 53226-3595. E-mail: [email protected]. Copyright * 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0148-7043/15/7404-0462 DOI: 10.1097/SAP.0b013e3182a1e544

462

www.annalsplasticsurgery.com

of technetium-99m sulfur colloid around the primary tumor. A F detection camera takes dynamic radiographs to identify the sentinel lymph nodes or first node to receive drainage. This also identifies any secondary nodes that receive drainage through a sentinel node as well as intercalated or in-transit nodes (eg, around the knee or elbow for an extremity lesion).3,6,8,9 There is concern that lymphatic mapping is not highly reliable in predicting number of nodes removed and understanding how to use the results of this study is often difficult. No study has examined the relationship between lymphatic mapping results and nodes harvested, or has looked at the relationship of F probe counts and lymph node pathology. The purpose of this study was to examine a large cohort of consecutive patients at Froedtert Hospital who underwent sentinel lymph node biopsy for melanoma to evaluate the correlation between preoperative 99mTc-sulfur colloid lymphatic mapping results and the number of lymph nodes harvested as well as looking at the relationship of F probe counts and lymph node pathology.

METHODS In an institutional review board-approved study, a retrospective chart review was conducted on consecutive patients who underwent a sentinel lymph node biopsy for melanoma from May 1, 2001, to December 31, 2010, at Froedtert Memorial Lutheran Hospital. A search of billing records by Current Procedural Terminology (38500, 28510, 28520, and 28525) was conducted and identified 262 patients. All the patients were treated at the same institution by the Department of Plastic Surgery, under the same senior author (W.W.D.). We excluded any patient younger than 18 years. We included any patient undergoing sentinel lymph node biopsy for a primary melanoma skin lesion. At our institution, lymphoscintigraphy for melanoma is typically performed the day of surgery. This involves 4 injections totaling 800 KCi of filtered Tc-99m sulfur colloid in 0.8 mL administered around the outer border of the known melanoma. Images are obtained at 30 s/frame for 10 minutes, followed by a high count planar image of all potential lymph node drainage basins at 10 minutes. The visualized nodes are marked on the skin and the patient is then sent to surgery for probe localization. At the time of surgery, 1.5 mL of methylene blue is injected intradermally around the melanoma. The visualized blue dye and a F probe counter are used in combination to aid in lymph node identification. Lymph node excision continues until the background count is less than 10% of the hottest node. Data abstracted were stored using Microsoft Excel (Microsoft Corporation, Redmond, Wash) containing 65 fields for clinical data entry including demographics, primary melanoma site, pathologic characteristics of primary melanoma including Breslow depth, presence of ulceration, presence of mitotic figures, lymphatic mapping results, and pathologic studies of the lymph nodes. Descriptive statistics were completed on the demographic data, pathological characteristics of the primary melanoma, lymphatic mapping results, F probe counts, and sentinel lymph node pathology. Paired 2-sided Student t test and Pearson correlation coefficients Annals of Plastic Surgery

& Volume 74, Number 4, April 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Annals of Plastic Surgery

& Volume 74, Number 4, April 2015

were completed on the number of lymph nodes identified on lymphatic mapping versus surgical excision as well as the basin identified on lymphatic mapping versus explored surgically. Two-sided Student t tests were completed on the F probe count sum and average versus lymph node status.

RESULTS Of the 262 patients, the median age was 54.5 years (range, 18Y90 years) with 52.3% male. Ninety eight percent were white with 1.15% African American (Table 1). The average Breslow depth of biopsy was 2.0 (1.9) mm, 42% were shave biopsies with tumor present at the base (Tx). After the Tx lesions, T2 lesions (between 1.01 and 2 mm thickness) were the most common in 25.6% of specimens. Ulceration was present in 28% of specimens. Mitotic figures were not routinely recorded until 2009 but present in 16.8% of specimens (Table 1). Preoperative lymphatic mapping studies were reviewed and 99.6% of lymphatic mapping studies identified at least 1 lymphatic basin. There was only 1 study where no sentinel lymph node was identified. Eighty percent of studies showed 1 nodal basin (range, 0Y4). The lymphoscintigraphy mapping identified an average of 1.5 (0.9) sentinel nodes and 31% of studies had at least 1 secondary node. At surgery on average 2.6 (1.4) lymph nodes were excised (P G 0.0001) involving 1.2 (0.5) basins (P = 0.97) (Table 2). The

Lymphoscintigraphy for Melanoma

TABLE 2. Results of Lymphatic Mapping Study and Surgical Exploration Lymphatic Mapping

Number (%)

Lymph nodes basins Mean SD Sentinel lymph nodes Mean SD Secondary nodes present Secondary nodes Mean SD

1.2 0.5 1.5 0.9 81 (30.9) 0.6 1.1

Surgery

Number (%)

Lymph node basins Mean SD Lymph nodes Mean SD Lymph node positive

1.2 0.5 2.6 1.4 46 (17.6)

TABLE 1. Patient and Primary Biopsy Characteristics Number (%) Patient characteristics Age, y Median Range Sex Male Female Race White African American Hispanic Unknown Biopsy characters Breslow depth, mm Mean SD T stage Tx T1 T2 T3 T4 Unknown Ulceration Yes No Unknown Mitotic figures Yes No Unknown

54.5 18Y90 137 (52.3) 125 (47.7) 257 (98.1) 3 (1.15) 1 (0.38) 1 (0.38)

2.01 1.9

number of sentinel lymph nodes identified on lymphatic mapping and excised surgically correlated in 32.8% of cases (r = 0.33, P G 0.0001). The lymphatic basins identified on lymphoscintigraphy and those explored surgically correlated in 85.1% of cases (r = 0.67, P G 0.0001) (Fig. 1). F Probe numbers were documented and available to review for 86% of patients. Ninety four percent of the patients had documentation that excision continued until background counts were less than 10%. Pathology demonstrated 17.6% of patients had a positive sentinel lymph node. Table 3 demonstrates the relationship of the average and sum of F probe counts for each patient separated by their sentinel lymph node status. There was no difference in the sum or average of F counts per patient between positive and negative sentinel lymph node groups (P = 0.2 and P = 0.5). Further analysis demonstrated that 32 patients had a positive sentinel lymph node among multiple removed nodes and removing only the hottest node from this group of patients would have missed the positive sentinel lymph node in 3 patients.

DISCUSSION 110 (42.0) 31 (11.8) 67 (25.6) 28 (10.7) 17 (6.5) 9 (3.4) 73 (28.0) 138 (52.9) 50 (19.2) 44 (16.8) 16 (6.1) 202 (77.1)

* 2015 Wolters Kluwer Health, Inc. All rights reserved.

Lymph node status has a significant effect on prognosis as a positive node decreases 5-year survival to 20% to 70%.1 When studies examined those patients undergoing elective lymph node dissections, they found only 20% had positive nodal disease, which created an increased morbidity for many patients.6 The Multicenter Selective Lymphadenectomy Trial I randomly assigned 1269 patients with melanoma greater than 1 mm thick to either SLNB or nodal observation. Interim analysis demonstrates those patients who underwent observation with later clinical nodal disease had poorer 5-year survival compared to those with positive lymph nodes in the SLNB group (52.4% vs 72.3%).2,4,5 The description of a sentinel node was first referenced in the 1860s; however, the term was coined in 1960 by Gould to describe the first lymph node in the nodal basin to which a cancer would travel. A sentinel lymph node biopsy involves injection of technetium-99m sulfur colloid around the primary tumor intradermally. A F detection www.annalsplasticsurgery.com

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

463

Annals of Plastic Surgery

Hudak et al

& Volume 74, Number 4, April 2015

FIGURE 1. Correlation of sentinel lymph nodes and lymphatic basins. Correlation of the number of sentinel lymph nodes and lymphatic basins identified on lymphoscintigraphy with surgical excision.

camera takes dynamic delayed radiographs. This procedure is especially helpful in areas of unclear lymphatic drainage. In surgery, a F counter is used to identify the radioactive nodes and all nodes that have F probe count greater than 10% of the hottest node are excised. Removing only the hottest or most radioactive node misses 20% of sentinel nodes.3,6,8,9 In our study, when we further evaluated patients with multiple nodes removed, we found that 3 of these 32 patients would have failed to identify the positive sentinel lymph node if only the hottest node was removed. Most surgeons add an intraoperative blue dye to identify the afferent lymphatics and sentinel node. The combination of both procedures to identify the sentinel node results in a 96% to 99% nodal identification rate.3,6,8Y10 Most studies report an approximately 5% false-negative rate.3,6,8,9 However, it is shown that the location of the sentinel node alters identification rate, with 99.3% in the groin, 95.3% in the axilla, and only 84.5% in the neck.10,11 In the groin, the lymphatics are essential 2-dimensional with the sentinel node usually the node closest to the lesion for lower extremity melanomas. However, the axilla and neck present a more complex 3dimensional structure. Melanoma of the head and neck has multiple drainage patterns that are unpredictable and varied. Drainage patterns include the 7 different levels of the neck as well as intraparotid nodes in up to 30% of cases.10 Drainage in the axilla can include any of the 3 levels of the axilla. The literature demonstrates removal of an average of 2.1 to 2.3 sentinel node; our study removed 2.6 lymph nodes on average.6,8,12 Intraoperatively, it is important to examine the popliteal and epitrochlear region with the F probe counter. Uren et al13 demonstrated 8% of calf and foot melanomas had popliteal drainage and 16% of forearm and hand melanomas had epitrochlear drainage. Our study demonstrated involvement of the popliteal basin in 4% of lower extremity lesions

TABLE 3. F Probe Counts

Average F probe counts Sum of F probe counts

Sentinel Lymph Node Positive

Sentinel Lymph Node Negative

P

5054.8 12,011.0

4384.4 9486.9

0.49 0.20

Comparing the average and sum of F probe per sentinel lymph node procedure demonstrated no difference with regard to lymph node status.

464

www.annalsplasticsurgery.com

and of the epitrochlear basin in 6.5% of upper extremity lesions. Failure to detect and remove all sentinel nodes results in insufficient nodal staging and may increase risk of recurrence.6,8,12 Carlson et al10 demonstrated a positive lymph node in 17.6% of sentinel lymph node biopsies, which is the same as our rate of 17.6%. NCCN guidelines recommend a patient with a positive sentinel lymph node undergo completion lymphadenectomy. The Multicenter Selective Lymphadenectomy Trial 2 is currently looking at outcomes in patients with positive sentinel nodes undergoing completion lymphadenectomy versus observation.2,5 However, previous studies show a completion lymphadenectomy identifies additional involved nodes in 15% to 28% of patients and the number of nodes involved has prognostic value.2,14 Lymphatic mapping studies can often be difficult to interpret. For example, 1 patient had no sentinel node identified on imaging, however, 1 sentinel lymph node was found with surgical exploration with the aid of blue dye. Another patient had an even more difficult study to interrupt on first glance (Fig. 2) as it showed 1 sentinel lymph node with 4 secondary lymph nodes along the same chain. Further examination demonstrates a single track from the tumor to the first lymph node and the remaining lymph nodes branch out from this sentinel node. This patient had 4 lymph nodes, which had a background count more than 10% of the highest count, and these were removed surgically. Lymphoscintigraphy can clearly identify 2 separate lymphatic drainage basins as demonstrated in Figure 2. There were 2 patients who had a nonsentinel lymph node removed that was positive for metastatic disease. The first patient had an abnormal appearing lymph node that was neither hot nor blue removed and the second patient had a lymph node removed due to a blue tract coursing toward it. A large tumor burden may obstruct the lymphatic channels preventing the passage of the blue dye and technetium into the lymph node and impeding sentinel lymph node identification. Lymphoscintigraphy identifies lymphatic drainage patterns directing the physicians to the proper basin for removal of sentinel lymph nodes; however, it is often difficult for surgeons to understand how to use these data. The lymphoscintigram serves as a guide as where to begin the surgical exploration. The number of sentinel nodes excised does not necessarily equal the number identified on mapping studies, and an extremely high F probe count does not indicate a positive sentinel lymph node. The number of sentinel lymph nodes identified on lymphoscintigraphy compares weakly to the * 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Annals of Plastic Surgery

& Volume 74, Number 4, April 2015

Lymphoscintigraphy for Melanoma

FIGURE 2. Left, A, Lymphoscintigraphy identifies 1 sentinel lymph node and 4 secondary nodes. Surgical exploration led to the removal of 4 lymph nodes. Right, B, Lymphoscintigraphy identifies 2 distinct tracts to a sentinel lymph node in the right axilla and right groin for a trunk melanoma.

FIGURE 3.

number surgically excision. There is no difference in F counts between positive and negative lymph node groups. The removal of lymph nodes should be continued until the background count is less than 10%, especially because the sentinel node is not always the hottest node.

Department of Population Health, and the Clinical Translational Science Institute of Southeast Wisconsin.

ACKNOWLEDGMENTS The authors thank Daniel Eastwood, Medical College of Wisconsin Biostatistics Department, for the statistical assistance. This service is supported by funds from the Division of Biostatics,

2. Gershenwald JE, Ross MI. Sentinel-lymph-node biopsy for cutaneous melanoma. N Engl J Med. 2011;364:1738Y1745.

* 2015 Wolters Kluwer Health, Inc. All rights reserved.

REFERENCES 1. Howlader N, Noone AM, Krapcho M, et al., eds. SEER Cancer Statistics Review, 1975Y2009. Bethesda, MD: National Cancer Institute.

3. Johnson T, Sondak V, Bichakjian C, et al. The role of sentinel lymph node biopsy for melanoma: evidence assessment. J Am Acad Dermatol. 2005;54: 19Y27.

www.annalsplasticsurgery.com

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

465

Annals of Plastic Surgery

Hudak et al

4. Leong S, Zuber M, Ferris R, et al. Impact of nodal status and tumor burden in sentinel lymph nodes on clinical outcomes of cancer patients. J Surg Oncol. 2001;103:518Y530. 5. Melanoma: NCCN Clinical Practice Guidelines in Oncology. National Comprehensive Cancer Network. May 3, 2011. Accessed July 12, 2011. 6. Stadelmann W. The role of lymphatic mapping and sentinel lymph node biopsy in the staging and treatment of melanoma. Clin Plast Surg. 2010;37:79Y99. 7. Younghoon RC, Chiang MP. Epidemiology, staging (new system), and prognosis of cutaneous melanoma. Clin Plast Surg. 2010;37:47Y53. 8. Veenstra H, Wouters M, Kroon B, et al. Less false-negative sentinel node procedures in melanoma patients with experience and proper collaboration. J Surg Oncol. 2011;1Y4. 9. Alazraki N, Glass E, Castronovo F, et al. Procedure guideline for lymphoscintigraphy and the use of intraoperative gamma probe for sentinel

466

www.annalsplasticsurgery.com

10.

11. 12.

13. 14.

& Volume 74, Number 4, April 2015

lymph node localization in melanoma of intermediate thickness. J Nucl Med. 2002;43:1414Y1418. Carlson G, Murray D, Lyles R, et al. Sentinel lymph node biopsy in the management of cutaneous head and neck melanoma. Plast Reconstr Surg. 2005;115:721Y728. Morton D, Cochran A, Thompson J, et al. Sentinel node biopsy for early stage melanoma. Ann Surg. 2005;242:302Y313. Wagner JD, Corbett L, Park MH, et al. Sentinel lymph node biopsy for melanoma: experience with 234 consecutive procedures. Plast Reconstr Surg. 2000;105:1956Y1966. Uren RF, Howman-Giles R, Thompson JF. Patterns of lymphatic drainage from the skin in patients with melanoma. J Nucl Med. 2003;44:570Y582. Dzwierzynski WW. Complete lymph node dissection for regional nodal metastasis. Clin Plast Surg. 2010;37:113Y125.

* 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Sentinel lymph node biopsy for melanoma: is there a correlation of preoperative lymphatic mapping with sentinel lymph nodes harvested?

Nodal status is the most significant prognostic factor in melanoma. No study has examined the relationship between lymphoscintigraphy, γ probe counts,...
2MB Sizes 0 Downloads 10 Views