ORIGINAL ARTICLE. Reliability of Sentinel Lymph Node Mapping With Biopsy for Head and Neck Cutaneous Melanoma

Transcription

1 ORIGINAL ARTICLE Reliability of Sentinel Lymph Node Mapping With Biopsy for Head and Neck Cutaneous Melanoma Cecelia E. Schmalbach, MD; Brian Nussenbaum, MD; Riley S. Rees, MD; Jennifer Schwartz, MD; Timothy M. Johnson, MD; Carol R. Bradford, MD Objectives: To determine (1) the reliability of sentinel lymph node mapping with biopsy (SLNB) in head and neck cutaneous melanoma to accurately stage nodal basins and (2) the safety of SLNB in both the neck and parotid regions. Design: Retrospective cohort study with a median follow-up of 25 months. All patients had a minimum follow-up of 1 year. Setting: Academic medical center. Patients: Eighty evaluable patients diagnosed as having head and neck cutaneous melanoma and staged using SLNB. Interventions: Sentinel lymph nodes were identified using preoperative lymphoscintigraphy and a combination of intraoperative gamma probe and isosulfan blue dye. Patients with a SLN positive for melanoma underwent therapeutic lymphadenectomy followed by an evaluation for adjuvant therapies. Patients with a negative SLNB result were followed up clinically. Main Outcome Measures: Percentage of positive SLNs, regional recurrence in the setting of a negative SLNB result (false-negative rate), and procedure complications. Results: The mean Breslow depth was 2.35 mm. A SLN was identified in 77 (96.3%) of cases, with an average of 2.18 nodes per patient. Of the sentinel nodes identified, 74% were from the neck region. The remaining 26% were from the parotid basin. No facial nerve complications occurred. Of the patients, 14 (18%) were SLN positive for metastatic melanoma. The regional failure rate in the setting of a negative SLNB result was 4.5%. Conclusions: Sentinel lymph node mapping with biopsy is a reliable technique to diagnose regional spread from head and neck cutaneous melanoma. This procedure can be performed in both neck and parotid nodal basins with safety and accuracy similar to non head and neck sites. Arch Otolaryngol Head Neck Surg. 2003;129:61-65 From the Department of Otolaryngology Head and Neck Surgery (Drs Schmalbach, Nussenbaum, Johnson, and Bradford), Department of Surgery, Section of Plastic Surgery (Dr Rees and Johnson), and Department of Dermatology (Drs Schwartz and Johnson), University of Michigan, Ann Arbor. NUMEROUS QUESTIONS surround the management of head and neck cutaneous melanoma, especially with respect to the role for elective treatment of regional nodal basins. Prospective randomized trials have failed to demonstrate a survival benefit for patients with melanoma undergoing elective lymphadenectomy. 1,2 While nodal status is clearly recognized as the most significant prognostic factor for patients diagnosed with cutaneous melanoma, only 10% to 20% of patients present with occult lymph node metastasis. This risk for nodal metastasis increases as primary tumor thickness increases. Lesions less than 1 mm thick are associated with a less than 5% rate of regional metastasis, while lesions thicker than 4 mm are associated with a 30% to 50% rate of nodal involvement. 3,4 In an attempt to identify this small group of patients harboring occult nodal disease using a minimally invasive procedure, Morton et al 5 introduced sentinel lymph node mapping with biopsy (SLNB) for the evaluation of patients with trunk and extremity cutaneous melanoma. They demonstrated that the status of the SLN accurately represented the status of the entire nodal basin from which it was obtained. In doing so, SLNB provided a means of identifying patients with occult nodal metastasis who warranted therapeutic lymphadenectomy and adjuvant therapy, while sparing the remaining 80% of patients without regional disease the morbidity associated with formal lymphadenectomy. Recent multivariate analysis involving patients with stage I and II melanoma by Gershenwald et al 6 found the pathological status (positive or negative for 61

2 Table 1. Distribution of Head and Neck Cutaneous Melanoma Location No. (%) of Cases Cheek 16 (20) Scalp 16 (20) Neck 13 (16) Auricle 13 (16) Forehead 5 (6) Temple 4 (5) Mentum 4 (5) Preauricular 4 (5) Cutaneous lip 2 (3) Nose 1 (1) Postauricular 1 (1) Eyelid 1 (1) metastasis) of the SLN to be the most important prognostic factor for recurrence and survival. While SLNB has a defined role in the evaluation of cutaneous melanoma of the trunk and extremities, several questions remain unanswered with respect to its application in the head and neck region. 7 The complexity of the head and neck lymphatic system has caused concern surrounding the reliability of SLNB to accurately reflect the status of the entire nodal basin. The popularity of SLNB in this region has also been limited by technical difficulties, 8 concern surrounding damage to vital structures such as the facial nerve, 8 and the necessity for nuclear medicine staff as well as pathologists who specialize in SLNB technique. The objective of this retrospective cohort study was to determine the reliability of SLNB for regional staging of head and neck cutaneous melanoma. METHODS Approval for this study was granted by the University of Michigan Medical School Institutional Review Board for Human Subject Research, Ann Arbor. This retrospective cohort study included 87 patients (7 of whom were lost to follow-up) treated for head and neck cutaneous melanoma who were staged using SLNB by the senior authors (C.R.B. and R.S.R.). Patients were identified through a query of the prospective University of Michigan Melanoma Database from April 1998 through December All patients who had histologically proven melanoma with a minimum Breslow depth of 1.00 mm (or 1.00 mm in the setting of other adverse prognostic variables such as ulceration, tumor extension to the deep margin, extensive regression, or young age) in the clinical absence of regional or distant metastasis were counseled for SLNB. Patients who underwent the procedure were included in the analysis. All patients were initially evaluated with a comprehensive consult by the University of Michigan Multidisciplinary Melanoma Clinic. Patients who underwent prior melanoma excision with wide margins or prior neck surgery were excluded due to the decreased accuracy in identifying the true SLN. Only patients with a minimum follow-up of 1 year were included in the analysis. All patients underwent preoperative lymphoscintigraphy to determine the number, location, and laterality of nodal basins at risk for metastatic disease. The lymphoscintigraphy was performed 2 to 4 hours prior to surgery using techniques previously described. 9 Specifically, a mean dose of 2.3 µci (range, µci) of technetium Tc 99m sulfur colloid (CIS-US Inc, Bedford, Mass) was injected intradermally into the 4 quadrants surrounding the primary melanoma lesion. Planar imaging was performed 15 to 30 minutes following injection (E.CAM; Siemens, Hoffman Estates, Ill). Intraoperative lymphatic mapping with isosulfan blue dye (1% Lymphazurin; Hirsch Industries Inc, Richmond, Va) was performed using previously described techniques. 5 Approximately 1 ml of dye was injected into the intradermal layer surrounding the primary melanoma lesion. Following wide local excision of the primary lesion, nodal basins were evaluated for increased radioactivity using a handheld gamma probe (Navigator GPS; RMD Instruments, Watertown, Mass). A 1- to 3-cm incision was made overlying the areas of increased radioactivity. A preauricular incision was used for SLNB in the parotid region. Sentinel lymph nodes were identified using a combination of gamma probe and blue dye. Each SLN was individually dissected from surrounding tissue. It was the preference of one of the 2 surgeons to use continuous facial nerve monitoring for SLNB within the parotid nodal basin (Viking; Nicolet Instrument Corporation, Madison, Wis). The staging procedure was considered complete when all nodal basins had minimal background radioactivity relative to the primary lesion and sentinel nodes. All SLNs were sent for histologic evaluation using permanent sections. Histologic evaluation included serial sectioning and staining with hematoxylin-eosin (H&E). Special immunohistochemical staining for S100, melan-a, and HMB-45 were performed for all SLNs that were negative on H&E staining. Patients with a SLN positive for metastatic melanoma underwent therapeutic lymphadenectomy within 2 weeks of the biopsy. Counseling for adjuvant interferon alfa-2 was performed in the setting of a positive lymph node. In addition, adjuvant radiation was recommended for patients when metastatic disease involved 3 or more lymph nodes or when extracapsular spread was identified. The remaining patients with a negative SLNB result were followed up clinically. The University of Michigan Melanoma Database was used to define the population demographics. Main outcome measures included the percentage of positive SLNs, regional recurrence in the setting of a negative SLNB result (false-negative rate), and facial nerve injury. RESULTS Eighty-seven patients treated between April 1998 and December 2000 met the inclusion criteria for this study. Seven patients were lost to follow-up. Of the remaining 80 evaluable patients, 54 (68%) were men and 26 (32%) were women. The median patient age was 55 years (range, 7-86 years). The mean Breslow depth was 2.35 mm (range, mm). Melanoma subtypes included superficial spreading (32%), unclassified (21%), nodular (20%), lentigo maligna (19%), desmoplastic (5%), spindle cell (1%), polypoid melanoma (1%), and neurotropic (1%). Fourteen percent of the lesions were ulcerated. The distribution of the primary melanoma lesions are listed in Table 1. Using the combined techniques of lymphoscintigraphy, intraoperative gamma probe, and isosulfan blue dye, a SLN was found in 77 (96%) of the 80 cases performed. The average number of SLNs identified per patient was 2.18 (range, 1-7 nodes). Of the 168 SLNs, 74% were identified in neck nodal basins. The remaining 26% were found within the parotid bed. Of the 30 patients with SLNs that drained to the parotid basin, 28 (93%) underwent successful SLNB. One patient underwent success- 62

3 Table 2. Distribution of Sentinel Lymph Node Biopsy (SLNB) Results According to Breslow Depth* Patients Total, No Positive SLNB result, No. (%) 2 (20) 4 (12) 5 (22) 3 (25) *One patient did not have Breslow depth reported. ful SLNB in a neck nodal basin but required a superficial parotidectomy because of the deep location of a second SLN within the superficial lobe. This parotid SLN was identified in the surgically removed lobe using the gamma probe. The hot node was sent separately for histologic review and was found to be negative for metastatic disease. A second patient experienced significant intraoperative bleeding from the parotid tissue. The procedure was aborted due to the associated risk to the facial nerve, and the patient was followed up clinically. This patient remains free of disease at a follow-up interval of 36 months. There was minimal morbidity related to this procedure. No anaphylactic reactions occurred following injection of isosulfan blue dye. There were no cases of cranial nerve damage, and all patients had normal postoperative cranial nerve function, including the facial nerve. No damage to vital neck structures was reported. Of the patients, 14 (17.5%) had metastatic melanoma identified using SLNB. A breakdown of SLNB results according to Breslow depth is provided in Table 2. All of these patients subsequently underwent therapeutic lymphadenectomy. Specifically, the neck nodal basins were treated with a modified radical neck dissection sparing the sternocleidomastoid muscle, internal jugular vein, and spinal accessory nerve. A posterolateral neck dissection was completed when clinically indicated. Lymphadenectomy of the parotid nodal basin entailed a superficial parotidectomy. The median follow-up for the positive SLN group was 25 months (range, months). The remaining 66 (82.5%) of patients with a negative SLNB result were followed up clinically. Median follow-up for the negative SLN group was 25 months (range, months). During this follow-up interval, 8 (12%) of 66 patients developed recurrent disease. The distribution of recurrences is summarized in Table 3. Three patients had a recurrence of isolated regional disease in a previously mapped nodal basin. Thus, the regional failure rate in the setting of a negative SLNB result, also referred to as the false-negative rate, was 4.5% at a median follow-up interval of 25 months. COMMENT Breslow Depth, mm There are 4 reasons that necessitate the use of SLNB for accurate regional staging of cutaneous melanoma. 10 First, the technique provides important prognostic information to the physician and patient in guiding subsequent treatment options. Second, it helps to identify patients harboring nodal metastases, who then may benefit from Table 3. Location of First Recurrence for 8 Patients in Negative Sentinel Lymph Node Group Primary Tumor Site (Breslow Depth, mm) Lateral neck (5.0) Auricle (1.12) Scalp (3.12) Auricle (1.3) Scalp (2.09) Scalp (2.46) Scalp (3.7) Cheek (2.1) Area of Recurrence Satellite/in-transit Satellite/subcutaneous Skin and pulmonary metastasis Lung and liver metastasis Lung metastasis *Case contributing to false-negative rate of sentinel lymph node mapping with biopsy. early therapeutic lymphadenectomy. Third, it identifies patients who are candidates for adjuvant treatment including interferon alfa-2 and radiation. Fourth, accurate regional staging enables the identification of a homogeneous population of patients for enrollment into clinical trials. Regional metastasis is recognized as an important prognostic factor. Without accurate pathologic staging, stratification is impossible and results of clinical trials will remain inconsistent and difficult to interpret. Fortunately, SLNB provides a minimally invasive means of regional staging. While SLNB is routinely performed for cutaneous trunk and extremity melanoma, 7 the role of this procedure remains uncertain for head and neck cutaneous melanoma. The main concern surrounds the reliability of the SLN to accurately predict the disease status of the entire nodal basin within this region. The interlacing network of cervical lymphatic vessels is often deemed watershed in nature. The complexity of this lymphatic system was demonstrated by O Brien et al 11 who reported a 34% discordance between the clinical prediction of lymphatic drainage and lymphoscintigraphy findings in 97 cases of head and neck cutaneous melanoma. Our study demonstrated that the complexity of the head and neck lymphatic system does not preclude the use of SLNB for staging of cutaneous melanoma. Sentinel lymph node mapping with biopsy in the head and neck region accurately predicts the status of the nodal basin, with 14 (17.5%) of 80 patients identified with a positive SLNB result and 3 (4.5%) of 66 patients developing regional recurrence following a negative SLNB result. A review of other institutional experiences in the use of SLNB for head and neck cutaneous melanoma is presented in Table 3. Studies were included only if information specific to the head and neck sentinel nodes could be ascertained. The 17.5% rate of SLN positivity and the 4.5% false-negative rate reported in our study compares favorably with the success of SLNB achieved in other anatomic sites. 12,13 Reported regional recurrence rates following a negative SLNB result in the head and neck region were quite variable, ranging from 0% to 25% (Table 4). Follow-up time is likely one reason for this variability. Two thirds of recurrences from cutaneous melanoma are expected to occur within 3 years following diagnosis. 22 The 63

4 Table 4. Previous Reports of SLNB for Head and Neck Cutaneous Melanoma Source No. of Patients Positive SLNB Result, % Regional Failure Rate Following a Negative SLNB Result, % Median Follow-up, mo Treatment Modality O Brien et al, LSG + Dye Wells et al, Dye ± LSG + GP* Alex et al, LSG + GP ± Dye Jansen et al, LSG + GP + Dye Carlson et al, LSG + GP ± Dye Rasgon, LSG + GP + Dye Jacobs et al, LSG + GP Patel et al, LSG + GP ± Dye Chi et al, LSG + GP Abbreviations: Dye, vital blue dye; GP, intraoperative gamma probe; LSG, lymphoscintigraphy; plus sign, with; plus/minus sign, with or without; SLNB, sentinel lymph node mapping with biopsy. *Of the patients, 38% underwent mapping with vital blue dye alone. Of the patients, 52% underwent mapping with all 3 treatment modalities. Mean follow-up reported. Vital blue dye used in a few sporadic cases. Of the patients, 85.7% underwent mapping with all 3 treatment modalities. regional recurrence rate of 0% reported by Wells et al 14 included a follow-up interval of only 11.6 months. The 1.9% failure rate reported by Patel et al 20 included a follow-up interval beginning 1 month following SLNB. Our median follow-up of 25 months (range, months) was moderate in duration. Repeat analysis at the 5- to 10-year follow-up interval is warranted to confirm our initial findings. The experience of the surgeon may also account for variability, and a 30-case learning curve has previously been suggested. 7 Studies with higher regional failures compared with our reported 4.5% often involved smaller cohorts of only 20 to 30 patients (Table 4). At our institution, most SLNBs for head and neck melanoma are performed by the 2 senior surgeons. Although retrospective studies often reveal higher failure rates at the beginning of a surgeon s experience, our study did not show that the learning curve contributed to regional failures. 14,23 Following completion of the recommended number of cases that represents the learning curve, 1 regional failure involved the 19th case for one surgeon, while the remaining 2 failures involved the 43rd and 45th case for the second surgeon. This observation reflects the technical challenges involved in performing SLNB. O Brien et al 11 reported that 33% of patients with head and neck cutaneous melanoma who underwent preoperative lymphoscintigraphy had drainage to lymph nodes within the parotid bed. Our distribution was similar, with 30 (26%) of 80 patients mapping to this region. We successfully staged 28 (93.3%) of these 30 patients using SLNB. Although potential injury to the facial nerve from SLNB has led some authors to advocate superficial parotidectomy over the mapping procedure, 8 we successfully removed 39 nodes from 28 parotid basins without facial nerve injury. Similar results were achieved by both senior surgeons, regardless of the use of continuous facial nerve monitoring. Concern has also been expressed that SLNB causes inflammation and fibrosis, which could place the facial nerve at increased risk when reoperation is required to definitively treat the parotid basin in the setting of a positive SLN. 8 In our study, patients with a positive parotid SLN underwent a superficial parotidectomy as a subsequent procedure, and normal facial nerve function was achieved postoperatively in all cases. Our findings are consistent with other reports demonstrating that SLNB can reliably and safely be performed within the parotid nodal basin. 24,25 While the surgeon s experience and technical skill are both vital to the success of SLNB, we attribute our overall low regional failure rate to a team effort involving surgeons, dermatologists, nuclear medicine staff, and pathologists. Appropriate patient selection is imperative because patients with regional or distant metastatic disease or previous surgical disruption of the lymphatic system are not candidates for SLNB. An experienced nuclear medicine staff is necessary because inappropriate administration of the radioactive tracer can lead to shine through, which will render the intraoperative gamma probe useless. Communication with the nuclear medicine team is helpful not only in interpreting the lymphoscintigram, but also to ensure that the appropriate lesion is mapped. Patients with melanoma often present with multiple lesions. It is imperative that only the invasive melanoma is injected with radioactive colloid. One regional failure in this study occurred in a patient who presented for simultaneous treatment of adjacent melanoma and Bowen s disease of the scalp. Subsequent communication with the nuclear medicine team revealed that both lesions were injected, which may have adversely affected the accuracy of SLNB. Lastly, the pathologist plays a critical role in the success of SLNB. Wagner et al 26 reported the mean tumor volume in SLNs positive for metastatic melanoma to be only 4.7 mm 3. Joseph et al 27 reported identification of only 73% of metastatic SLNs using standard H&E staining. Therefore, occult lymphatic metastasis from cutaneous melanoma can be difficult to detect and warrants rigorous pathological analysis including serial sectioning, special immunohistochemical study when indicated, and interpretation by an experienced pathologist. Sentinel lymph node mapping with biopsy provides the pathologist with a limited number of nodes to thoroughly evaluate. There- 64

5 fore, the histologic analysis of SLNs is more thorough and complete compared to the evaluation of the entire lymphadenectomy specimen. 13 CONCLUSIONS Many questions surround the treatment of head and neck cutaneous melanoma. This retrospective study addresses the reliability and safety of SLNB to accurately determine the presence of occult regional spread for head and neck cutaneous melanoma. With a median follow-up of 25 months, 17.5% of patients had a positive SLNB result and only 4.5% of patients developed regional recurrence following a negative SLNB result. The procedure was performed with equivalent safety in both the neck and parotid nodal basins. There were no reported complications of facial nerve weakness or injury to other vital structures. This documented accuracy reported in the setting of minimal morbidity indicates that SLNB is a reliable procedure for regional staging of head and neck cutaneous melanoma. Given that the SLN status is the most important prognostic factor for patients with melanoma, 6 the accuracy of SLNB in the head and neck region demonstrated in the present study is quite promising. Accepted for publication June 27, This study was presented at the annual meeting of the American Head and Neck Society, Boca Raton, Fla, May 11, Corresponding author: Carol R. Bradford, MD, Department of Otolaryngology Head and Neck Surgery, University of Michigan, 1500 E Medical Center Dr, 1904 Taubman Center, Ann Arbor, MI ( cbradfor@umich.edu). REFERENCES 1. Balch CM, Soong SJ, Bartolucci AA, et al. Efficacy of an elective regional lymph node dissection of 1 to 4 mm thick melanomas for patients 60 years of age and younger. Ann Surg. 1996;224: Cascinelli N, Morabito A, Santinami M, MacKie RM, Belli F. Immediate or delayed dissection of regional nodes in patients with melanoma of the trunk: a randomized trial. Lancet. 1998;351: Balch CM, Soong SJ, Gershenwald JE, et al. Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol. 2001;19: Singluff CL, Stidham KR, Ricci WM, Stanley WE, Seigler HF. Surgical management of regional lymph nodes in patients with melanoma. Ann Surg. 1994;219: Morton DL, Wen DR, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg. 1992;127: Gershenwald JE, Thompson W, Mansfield PF, et al. Multi-institutional melanoma lymphatic mapping experience: the prognostic value of sentinel lymph node status in 612 stage I or II melanoma patients. J Clin Oncol. 1999;17: Morton DL, Thompson JF, Essner R, et al. Validation of the accuracy of intraoperative lymphatic mapping and sentinel lymphadenectomy for early-stage melanoma. Ann Surg. 1999;230: Eicher SA, Clayman GL, Myers JN, Gillenwater AM. A prospective study of intraoperative lymphatic mapping for head and neck cutaneous melanoma. Arch Otolaryngol Head Neck Surg. 2002;128: Uren RF, Howman-Giles RB, Shaw HM, Thompson JF, McCarthy WH. Lymphoscintigraphy in high risk melanoma of the trunk: predicting draining node groups, defining lymphatic channels and locating the sentinel node. J Nucl Med. 1993; 34: McMasters KM, Reintgen DS, Ross MI, et al. Sentinel lymph node biopsy for melanoma: controversy despite widespread agreement. J Clin Oncol. 2001;19: O Brien CJ, Uren RF, Thompson JF, et al. Prediction of potential metastatic sites in cutaneous head and neck melanoma using lymphoscintigraphy. Am J Surg. 1995;170: Gershenwald JE, Colome MI, Lee JE, et al. Patterns of recurrence following a negative sentinel lymph node biopsy in 243 patients with stage I or II melanoma. J Clin Oncol. 1998;16: Cascinelli N, Belli F, Santinami M, et al. Sentinel lymph node biopsy in cutaneous melanoma: the WHO Melanoma Program experience. Ann Surg Oncol. 2000; 7: Wells KE, Rapaport DP, Cruse CW, et al. Sentinel lymph node biopsy in melanoma of the head and neck. Plast Reconstr Surg. 1997;100: Alex JC, Krag DN, Harlow SP, et al. Localization of regional lymph nodes in melanomas of the head and neck. Arch Otolaryngol Head Neck Surg. 1998;124: Jansen L, Koops HS, Nieweg OE, et al. Sentinel node biopsy for melanoma in the head and neck region. Head Neck. 2000;22: Carlson GW, Murray DR, Greenlee R, et al. Management of malignant melanoma of the head and neck using dynamic lymphoscintigraphy and gamma probeguided sentinel lymph node biopsy. Arch Otolaryngol Head Neck Surg. 2000; 126: Rasgon BM. Use of low-dose technetium Tc 99m sulfur colloid to locate sentinel lymph nodes in melanoma of the head and neck: preliminary study. Laryngoscope. 2001;111: Jacobs IA, Chevinsky AH, Swayne LC, Magidson JG, Britto EJ, Smith TJ. Gamma probe-directed lymphatic mapping and sentinel lymphadenectomy in primary melanoma: reliability of the procedure and analysis of failures after long-term followup. J Surg Oncol. 2001;77: Patel SG, Coit DG, Shaha AR, et al. Sentinel lymph node biopsy for cutaneous head and neck melanomas. Arch Otolaryngol Head Neck Surg. 2002;128: Chi DH, Dorofi DB, Desper E, Levine PA. Lymphoscintigraphy and sentinel node sampling for head and neck melanoma. Presented at: American Academy of Otolaryngology Head and Neck Surgery Foundation Annual Meeting; September 11, 2001; Denver, Colo. 22. Reintgen DS, Vollmer R, Tso CY, Seigler HR. Prognosis for recurrent stage I malignant melanoma. Arch Surg. 1987;122: Jansen L, Nieweg OE, Peterse JL, Hoefnagel CA, Valdes Olmos RA, Kroon BBR. Reliability of sentinel lymph node biopsy for staging melanoma. Br J Surg. 2000; 87: Wells KE, Stadelmann WK, Rapaport DP, Hamlin R, Cruse CW, Reintgen D. Parotid selective lymphadenectomy in malignant melanoma. Ann Plast Surg. 1999; 43: Ollila DW, Foshag LJ, Essner R, Stern SL, Morton DL. Parotid region lymphatic mapping and sentinel lymphadenectomy for cutaneous melanoma. Ann Surg Oncol. 1999;6: Wagner JD, Davidson D, Coleman JJ, et al. Lymph node tumor volumes in patients undergoing sentinel lymph node biopsy for cutaneous melanoma. Ann Surg Oncol. 1999;6: Joseph E, Brobeil A, Glass F, et al. Results of complete lymph node dissection in 83 melanoma patients with positive sentinel nodes. Ann Surg Oncol. 1998;5: