- Email: [email protected]
Sentinel Lymph Node Biopsy for Ocular Adnexal Melanoma: Experience in 30 Patients
Sentinel Lymph Node Biopsy for Ocular Adnexal Melanoma: Experience in 30 Patients Aaron Savar, MD,1 Merrick I. Ross, MD,2 Victor G. Prieto, MD, PhD,3,4 Doina Ivan, MD,3,4 Stella Kim, MD,1 Bita Esmaeli, MD1 Purpose: To report the findings on sentinel lymph node biopsy (SLNB) in 30 patients with ocular adnexal (conjunctival or eyelid) melanomas. Design: Prospective nonrandomized clinical trial. Participants: Thirty patients with diagnosis of eyelid or conjunctival melanoma. Methods: All patients with ocular adnexal melanoma who underwent SLNB at The University of Texas MD Anderson Cancer Center between December 2000 and July 2008 are the subject of this report. Sentinel lymph node biopsy was performed as previously described by our group, and patients were prospectively followed up. Main Outcome Measures: Findings on preoperative lymphoscintigraphy, SLNB, histopathologic examination of the primary tumor and sentinel lymph nodes (SNL), and nodal recurrence after SLNB. Results: Tumor sites were as follows: bulbar conjunctiva only, 14 patients; palpebral conjunctiva only, 8 patients; both bulbar and palpebral conjunctiva, 4 patients; and eyelid skin only, 4 patients. At least 1 SLN was identified in all patients. The median number of SLNs removed was 2 (range, 1–5). The most common basin sampled was the intraparotid (16 patients), followed by submandibular (level I) (11 patients), preauricular (9 patients), and superior cervical (level II) (6 patients). Five patients had SLN metastasis. Among the 25 patients with negative SLNB findings, there were 2 false-negative events. There were no false-negative events among patients treated during the last 4.5 years of the study. The mean Breslow thickness was 2.57 mm (range, 0.62–12 mm) among patients with negative SLNB and 4.86 mm (range, 2.0 –7.2 mm) among patients with positive SLNB findings (P ⫽ 0.055). Ulceration was present in 11 patients (39%): 4 (80%) of 5 patients with positive SLNB and 7 (28%) of 25 with negative SLNB, including both patients with false-negative results. The median time from SLNB to last contact was 2 years (range, 10 months to 6 years). Conclusions: Sentinel lymph node biopsy is effective for identifying nodal micrometastasis in patients with ocular adnexal melanoma and provides important prognostic information. The false-negative event rate in our series improved in the last 4 years, most likely because of a better technique and better patient selection for SLNB. We recommend consideration of SLNB for patients with intermediate-thickness ocular adnexal melanoma and those with ulceration. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2009;116:2217–2223 © 2009 by the American Academy of Ophthalmology.
Regional nodal status is the most important prognostic factor in patients with melanoma.1 Sentinel lymph node biopsy (SLNB) was first described for melanoma by Morton et al2 and has been shown to provide a disease-specific 5-year survival benefit for patients with intermediatethickness cutaneous melanomas.3 Although SLNB has become part of the standard of care for intermediate-thickness cutaneous melanomas, it has not gained wide acceptance in the management of ocular adnexal melanomas. Sentinel lymph node biopsy is based on the notion that certain tumors that spread through the lymphatics will metastasize to a first draining node in that region before spreading systemically. If the sentinel lymph node (SLN) is negative, the likelihood of involvement of other nodes or distant sites is low. If the SLN is positive, the patient requires further nodal resection or other treatment. Sentinel lymph node biopsy can prevent unnecessary and potentially morbid regional lymphadenectomy in patients with negative SLNs4; it can also identify patients with subclinical nodal © 2009 by the American Academy of Ophthalmology Published by Elsevier Inc.
metastases that would be missed with clinical or ultrasound examination alone. Sentinel lymph node biopsy for conjunctival melanoma was first reported by our group in 2001.5 Since then, Wilson et al6 have reported 2 cases of SLNB for ocular adnexal melanoma, and we have reported 8 additional cases.7 We report our expanded experience in 30 patients with ocular adnexal melanoma who underwent SLNB and report an updated SLNB positivity rate and the false-negative events for the technique.
Materials and Methods All patients who underwent SLNB for eyelid or conjunctival melanoma as part of a trial at The University of Texas MD Anderson Cancer Center between December 2000 and July 2008 are the subject of this report. Appropriate institutional review board approval was obtained for the conduct of this study. The eligibility criteria for SLNB were as follows: age ⱖ18 years; histologically documented invasive melanoma of ISSN 0161-6420/09/$–see front matter doi:10.1016/j.ophtha.2009.04.012
2217
Ophthalmology Volume 116, Number 11, November 2009
Figure 1. A, Preoperative lymphoscintigraphy in a patient with a conjunctival melanoma in the caruncle. B, Single photon emission computed tomography/computed tomography scan for the same patient shows the site of injection of technetium and C, drainage to a level II node.
the bulbar or palpebral conjunctiva or eyelid skin with Breslow equivalent thickness ⱖ1 mm; ability to undergo surgery under general anesthesia; chest x-ray, liver enzyme test, and head and neck computed tomography or magnetic resonance imaging negative for evidence of metastasis; sonography of the regional lymph nodes negative for evidence of metastasis; ability to give written informed consent; and, if of childbearing potential, a negative pregnancy test. In the early phases of the study, if the Breslow thickness for conjunctival tumors could not be reliably determined because of tangential cutting of the surgical specimens, the patient was allowed into the trial. In the later phases of the study, we excluded such patients unless the pathologist at
2218
The MD Anderson Cancer Center thought that the thickness was at least 1 mm even if the precise depth of invasion could not be determined. All patients underwent preoperative lymphoscintigraphy and intraoperative mapping. In addition, single photon emission computed tomography/computed tomography was performed in the last 6 patients (Fig 1A, B). This technique allows for higherresolution images than are produced by standard lymphoscintigraphy and permits better localization of the draining lymphatic basins. Approximately 1 hour before surgery, 0.2 ml (0.3 mCi) of technetium 99m-sulfur colloid was injected intradermally or subconjunctivally depending on the tumor location. A handheld
Savar et al 䡠 Sentinel Node Biopsy for Ocular Adnexal Melanoma gamma probe was used to identify SLNs transcutaneously. Nodes with radioactivity counts greater than twice the background level were considered positive. Sentinel lymph nodes were removed until the radioactivity counts in the nodal basin were less than twice the background level. The lymph nodes were sectioned at 2-mm intervals using the bread-loafing technique and entirely submitted for microscopic examination. A formalin-fixed, paraffin-embedded section was routinely stained with hematoxylin-eosin. If this initial section failed to reveal malignant cells, an additional hematoxylineosin–stained section was examined along with an immunohistochemical study using a pan-melanocytic cocktail (HMB-45, anti-tyrosinase, anti-MART1). Patients were considered to have had a false-negative event if there was no evidence of micrometastasis on SLNB but there was subsequent development of nodal disease during the follow-up period. For the purpose of this study, the false-negative event rate was defined as the number of false-negative events divided by the total number of patients with negative findings on SLNB. Variables assessed included age, gender, location of tumor, lymphoscintigraphy findings, number and location of SLNs biopsied, histopathologic findings for the primary tumor and the SLNs, postoperative complications, and length of follow-up. For continuous data, a t test was used to assess statistical significance.
Results Thirty patients underwent SLNB for ocular adnexal melanoma during the study period. Eight of these patients were previously reported by our group.7 There were 19 women and 11 men. The median age was 62 years (range, 24 – 80 years). Fourteen patients (47%) had melanomas limited to the bulbar conjunctiva, 8 patients (27%) had melanomas limited to the palpebral conjunctiva, 4 patients (13%) had melanomas involving both the bulbar and palpebral conjunctiva, and 4 patients (13%) had melanomas limited to the eyelid skin. At least 1 SLN was identified in all 30 patients. The median number of SLNs removed was 2 (mean, 2.3; range 1–5). The nodal basins from which the SLNs were removed are listed in Table 1. The median number of basins sampled was 2 (mean, 1.6; range, 1–3). The nodal basin most commonly sampled was the intraparotid; the next most commonly sampled basins were the submandibular (level I), preauricular, and superior cervical (level II). In 16 patients (53%), findings on lymphoscintigraphy correlated with intraoperative findings in at least 1 nodal basin. Seven patients (23%) had correlation in all basins biopsied. In 14 patients (47%), there was no correlation between preoperative lymphoscintigraphy and intraoperative findings. In 1 patient, no SLNs were identified during preoperative lymphoscintigraphy, but an SLN Table 1. Nodal Basins Sampled Basin Head Preauricular Intraparotid Neck Level I Level II Level III Level IV Level V Level not specified
No. of Patients (%) 9 (30) 16 (53) 11 (37) 6 (20) 3 (10) 1 (3) 1 (3) 2 (7)
was identified intraoperatively. This was accomplished by a slight modification of the injection technique for technetium 99m-sulfur colloid in this patient on the day of surgery. The median length of follow-up from the time of SLNB to last contact was 2 years (range, 10 months to 6 years).
Patients with Positive Sentinel Lymph Node Biopsy Findings Five patients (16.67%) had histologic evidence of nodal metastasis on SLNB. The first patient was a 59-year-old man with melanoma of the temporal bulbar conjunctiva of the left eye with Breslow thickness of 2.1 mm. At the time of presentation to MD Anderson, his conjunctival tumor had already been resected with negative margins; thus, no further resection of his conjunctiva was done. Four SLNs were identified: 2 intraparotid, 1 upper jugular (level II), and 1 middle jugular (level III). Both intraparotid nodes were positive. One parotid node showed a focus of tumor measuring 2.0⫻0.5 mm with extracapsular extension. The other showed a focus of approximately 20 spindle-shaped cells within the nodal parenchyma that were confirmed by immunohistochemistry to be melanocytic. Both jugular nodes were negative for micrometastasis. After SLNB, the patient returned to his hometown for follow-up. A parotidectomy and completion neck dissection was performed, and adjuvant external-beam radiation therapy was administered to the involved nodal basins. The patient had 3 local recurrences on the conjunctival surface, which were believed to be in-transit metastases. He eventually developed liver metastasis. He was treated with chemotherapy (Temozolomide; Schering-Plough Corporation, Kenilworth, NJ) and died of metastatic disease 22 months after SLNB. The second patient was a 67-year-old man with recurrent melanoma of the right upper eyelid palpebral conjunctiva with Breslow thickness of 6 mm. The single SLN identified, a submandibular (level I) node measuring 1.7⫻1.1⫻0.4 cm, was positive, with tumor replacing most of the nodal tissue with no extracapsular extension. Completion neck dissection was recommended. The patient returned home for further care. He underwent neck dissection, and 1 additional positive node was found. He was treated with postoperative adjuvant radiation therapy to the neck. He developed local and regional recurrence approximately 8 months later, as well as widespread metastatic disease, and died 3 years after SLNB. The third patient was a 54-year-old man with melanoma of the left lower lid palpebral conjunctiva and lid margin with Breslow thickness of 3.8 mm. Two SLNs were removed. The first, from the parotid gland, was positive, with the focus of metastatic melanoma measuring 0.8⫻0.5 mm with no extracapsular extension. The second, from the midjugular region (level III), was negative. The patient underwent parotidectomy and neck dissection at MD Anderson, and no additional nodes were positive for melanoma. At the time of this report, he has been followed for 13 months without evidence of local or regional recurrence or metastasis. The fourth patient was a 71-year-old woman with melanoma of the left lower eyelid skin extending to the lid margin but not involving the palpebral conjunctiva (Fig 2A), with Breslow thickness of 7.2 mm and Clark level IV. One preauricular SLN was identified; it contained a 2.5⫻2.0 mm subcapsular and intraparenchymal focus of metastatic melanoma without extracapsular extension (Fig 2B). The patient subsequently underwent parotidectomy and completion neck dissection at MD Anderson. Two of the 27 additional nodes removed were positive, both in the parotid gland. She remained free of disease at last contact 10 months after SLNB. The fifth patient was a 71-year-old woman who was referred to MD Anderson after incomplete excision of a melanoma of the left lower eyelid palpebral conjunctiva with Breslow thickness of 5.2
2219
Ophthalmology Volume 116, Number 11, November 2009 mm (Fig 3). She underwent complete surgical resection of her eyelid and SLNB. Three SLNs were identified, and all 3 were positive. The first was in the preauricular area and contained tumor measuring 6⫻3.5 mm with extracapsular extension. The other 2 nodes were intraparotid, and each contained fewer than 50 tumor cells. The patient underwent parotidectomy and completion neck dissection, with 4 of the 45 additional nodes removed positive. Adjuvant postoperative external-beam radiation therapy was de-
Figure 3. This patient had previously undergone gross excision of a palpebral conjunctival melanoma at another institution. A conjunctival scar remaining in the lower eyelid was the only sign of the primary tumor location. Technetium Tc 99m was injected subconjunctivally around this scar, and the draining sentinel lymph nodes were biopsied and found to be histologically positive.
livered to the parotid and neck area. Four months after SLNB, she developed recurrent nodal disease within the radiation field and is currently considering systemic treatment options for stage III conjunctival melanoma.
Figure 2. A, Patient with an eyelid melanoma extending to the eyelid margin. She had negative findings on magnetic resonance imaging and ultrasonography of her regional nodes but was found to have a (B) microscopically positive sentinel lymph node on sentinel lymph node biopsy (hematoxylin-eosin ⫻100).
2220
Figure 4. Histologic section of invasive melanoma with overlying ulceration and lacking epidermis (hematoxylin-eosin, ⫻200).
Savar et al 䡠 Sentinel Node Biopsy for Ocular Adnexal Melanoma Patients with Negative Sentinel Lymph Node Biopsy Findings Among the 25 patients with negative SLNB findings, 2 patients (8%) had a nodal recurrence (false-negative event). Both of these patients were among the first 10 patients who underwent SLNB for adnexal melanoma at our center; we did not observe any falsenegative events among patients treated during the last 4.5 years of the study. The details of these 2 patients were discussed in one of our previous publications.7 One additional patient with negative SLNB findings developed a local recurrence at multiple sites on her conjunctival surface and refused an orbital exenteration recommended at our center. She was treated at another center with plaque radiotherapy and experienced additional recurrences after that. She later died of widespread metastatic disease 30 months after SLNB. The remaining 22 patients remained free of disease at last follow-up, with a median follow-up time of 25 months (range, 10 months to 6 years).
Breslow Thickness and Ulceration Breslow thickness data were available for 26 patients (87%). The mean Breslow thickness was 2.57 mm (range, 0.62–12 mm) among patients with negative SLNB findings and 4.86 mm (range, 2.0 –7.2 mm) among patients with positive SLNB findings (P ⫽ 0.055). Data regarding the presence or absence of ulceration of the primary tumor on histopathologic examination were available for 28 patients (93%). Ulceration was present in 11 patients (39%) (Fig 4), including 4 (80%) of the 5 patients with positive SLNB findings and 7 (28%) of the 25 patients with negative SLNB findings, including both of the patients with false-negative results (P ⫽ 0.0618). Overall risk of nodal disease (i.e., either positive SNLB findings or false-negative SLNB findings) was significantly associated with ulceration (P ⫽ 0.0069).
Discussion Our study suggests that the results of SLNB provides important prognostic information for patients with ocular adnexal melanoma and can identify patients with early microscopic nodal metastasis, even as small as 0.5 mm or 20 cells. In this prospective trial, the overall SLN positivity rate was approximately 16%, and there were 2 false-negative events. Previous studies of intermediate-thickness cutaneous melanomas have shown similar rates of SLN positivity. Morton et al3 reported a rate of SLN micrometastasis of 16% in their randomized trial that included 764 SLNB procedures. Gomez-Rivera et al8 found that 20% of 113 patients with cutaneous melanoma of the head and neck had positive SLNs. Our false-negative event rate of 8% was slightly higher than previously reported rates for intermediate-thickness melanomas in other anatomic locations but close to rates reported for head and neck melanomas.9 Morton et al3 reported a false-negative rate of 3.4% among 764 patients undergoing SLNB. Their false-negative event rate as defined in our study was 4%. Schmalbach et al9 reported a false-negative rate of 4.5% (using the same definition as in our study). We have extensively discussed the potential reasons for false-negative events after SLNB for ocular adnexal melanoma in a previous publication.7 Both
of the false-negative events in our study occurred early in the study period. We believe that the false-negative event rate decreased because of our increased experience with the nuances of SLNB for ocular adnexal melanoma and improved selection of appropriate patients for SLNB. For example, 1 of the 2 false-negative events in our study occurred in a patient with a massive, rapidly growing and diffuse melanoma of 12-mm thickness involving the upper and lower eyelids and the bulbar and palpebral conjunctiva contiguously. For such massive tumors, the appropriate area of injection of technetium is difficult to determine and the thickness of the tumor makes distant-organ metastasis highly likely, making SLNB less useful. Today we would not perform SLNB on such a patient. Nevertheless, despite improved patient selection for SLNB, because of the 8% risk of false-negative events suggested in our study, we recommend that all patients with conjunctival or eyelid melanoma who have negative SLNB undergo long-term follow-up to rule out recurrence in the regional nodal basins. Another explanation for the improved false-negative results may be a shorter follow-up time for the most recent cases, but this is not likely to be the reason because most falsenegative events occurred within the first few months after SLN biopsy and each patient included in this report has had at least 10 months of follow-up. The goal of SLNB is to detect subclinical nodal metastasis and at the same time spare patients without nodal disease the morbidity of extensive nodal dissection.4 Four (80%) of the 5 patients with positive SLNs in our study had additional positive nodes found at the time of completion lymphadenectomy and therefore benefited from early identification by SLNB. Debate exists over whether the use of SLNB confers an absolute survival benefit when compared with observation.10 A change in survival is particularly difficult to demonstrate because the majority of patients undergoing SLNB will be expected to have no evidence of micrometastasis. However, in a recent randomized trial in which 1269 patients with an intermediate-thickness primary cutaneous melanoma were randomly assigned to SLNB or to observation, the patients who were found to have microscopically positive SLNs experienced a melanoma-specific survival benefit compared with the patients in the observation arm whose nodal disease was found by the wait-andwatch method.3 Breslow thickness is a well-known prognostic indicator for both cutaneous11 and conjunctival melanomas.12 Obtaining accurate histologic data for this purpose is especially important when dealing with delicate tissue such as conjunctiva and eyelid skin. Improper tissue handling and tangential sectioning can prevent the pathologist from making these measurements.13 Tuomaala and Kivelä14 analyzed 85 conjunctival melanomas and concluded that SLNB is most appropriate in lesions with Breslow thickness of at least 2 mm and for nonlimbal lesions. In our series, the thinnest tumor to have a positive SLN had a Breslow thickness of exactly 2.0 mm. However, it might still be useful to perform SLN biopsy for tumors less than 2.0 mm thick if they are associated with ulceration. The presence of ulceration may be another high-risk histologic feature and may predict nodal metastasis. In our series, development of nodal dis-
2221
Ophthalmology Volume 116, Number 11, November 2009 ease (positive SLN or false-negative event) was statistically significantly associated with ulceration. Although there was a trend toward SLN positivity among patients with ulceration, this difference was not statistically significant. In our study, 47% of patients had SLNs present in both the head (parotid and preauricular regions) and the neck (level I–V nodes, i.e., below the jaw). Lim et al15 looked at the pattern of nodal metastasis in 12 patients with conjunctival melanoma and found that temporal lesions tended to metastasize to the preauricular nodes, whereas nasal lesions tended to metastasize to the submandibular nodes. Preoperative lymphoscintigraphy and SLNB data from our study did not corroborate this distribution. Preoperative lymphoscintigraphy allows one to determine the location of the SLN before surgery. At least 1 SLN was identified in all but 1 patient in our study during preoperative lymphatic mapping. In the 1 patient with no drainage identified during preoperative lymphoscintigraphy, drainage to SLNs was seen intraoperatively. This was accomplished by modifying the technique for injection of technetium slightly on the day of surgery to ensure drainage from the ocular surface or the eyelid. This is particularly important when the thin eyelid skin is injected, because the injection has to be intradermal for ideal drainage to the SLNs. Single photon emission computed tomography/ computed tomography has become a standard part of the preoperative lymphatic mapping procedure at the MD Anderson Cancer Center and can provide far more anatomic detail than standard lymphoscintigraphy. Patients with conjunctival or eyelid melanoma who have positive SLNB findings should be referred for completion node dissection and parotidectomy if the parotid basin is involved. If additional positive nodes are found during parotidectomy and neck dissection, the patient should be considered for adjuvant radiation therapy delivered to the nodal basins. The presence of extracapsular extension increases the risk of nodal recurrence.16 It has been recommended that such patients, as well as those with multiple positive nodes, be considered for adjuvant irradiation of the involved nodal basins after nodal dissection.17 The technical aspects and safety profile of SLNB have been extensively described in our previous publications.18,19 To date, we have encountered only 3 cases of temporary weakness of the marginal mandibular branch of the facial nerve, which have resolved spontaneously. There have been no other side effects attributable to SLNB. In our experience, the procedure has proven to be safe, takes 30 to 90 minutes, and can be combined with the primary resection of the conjunctival or eyelid mass or, when referral to our center takes place after resection of the primary ocular tumor, done as a secondary procedure within a few weeks after the ocular tumor resection. In conclusion, to our knowledge, this is the largest reported series of SLNB for ocular adnexal melanoma. Despite this, the absolute number of cases is small, limiting our ability to draw strong conclusions with respect to a survival benefit from SLNB. However, we can say that SLNB is a safe procedure and can identify nodal micrometastasis in patients with ocular adnexal melanoma. The SLNB positivity rate is similar to rates reported for intermediate-thickness
2222
cutaneous melanomas in all other anatomic sites. The falsenegative event rate in our series was higher than that reported in other anatomic sites but has improved in the last 4 years, we believe as a result of improved technique and better patient selection. We recommend consideration of SLNB for patients with intermediate-thickness ocular adnexal melanoma, especially if ulceration is identified.
References 1. 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;19:3622–34. 2. Morton DL, Wen DR, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392–9. 3. Morton DL, Thompson JF, Cochran AJ, et al., MSLT Group. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med 2006;355:1307–17. 4. van Akkooi AC, Bouwhuis MG, van Geel AN, et al. Morbidity and prognosis after therapeutic lymph node dissections for malignant melanoma. Eur J Surg Oncol 2007;33:102– 8. 5. Esmaeli B, Eicher S, Popp J, et al. Sentinel lymph node biopsy for conjunctival melanoma. Ophthal Plast Reconstr Surg 2001;17:436 – 42. 6. Wilson MW, Fleming JC, Fleming RM, Haik BG. Sentinel node biopsy for orbital and ocular adnexal tumors. Ophthal Plast Reconstr Surg 2001;17:338 – 44. 7. Ho VH, Ross MI, Prieto VG, et al. Sentinel lymph node biopsy for sebaceous cell carcinoma and melanoma of the ocular adnexa. Arch Otolaryngol Head Neck Surg 2007;133: 820 – 6. 8. Gomez-Rivera F, Santillan A, McMurphey AB, et al. Sentinel node biopsy in patients with cutaneous melanoma of the head and neck: recurrence and survival study. Head Neck 2008;30: 1284 –94. 9. Schmalbach CE, Nussenbaum B, Rees RS, et al. Reliability of sentinel lymph node mapping with biopsy for head and neck cutaneous melanoma. Arch Otolaryngol Head Neck Surg 2003;129:61–5. 10. De Wilt JH, van Akkooi AC, Verhoef C, Eggermont AM. Detection of melanoma micrometastases in sentinel nodes— the cons. Surg Oncol 2008;17:175– 81. 11. Breslow A. Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970;172:902– 8. 12. Tuomaala S, Eskelin S, Tarkkanen A, Kivelä T. Populationbased assessment of clinical characteristics predicting outcome of conjunctival melanoma in whites. Invest Ophthalmol Vis Sci 2002;43:3399 – 408. 13. Esmaeli B, De Martelaere S, Ozkan Y, et al. Surgical specimen handling for conjunctival melanoma: implications for tumor thickness determination and sentinel lymph node biopsy [letter]. Ophthal Plast Reconstr Surg 2006;22:238 –9. 14. Tuomaala S, Kivelä T. Metastatic pattern and survival in disseminated conjunctival melanoma: implications for sentinel lymph node biopsy. Ophthalmology 2004;111:816 –21. 15. Lim M, Tatla T, Hersh D, Hungerford J. Patterns of regional head and neck lymph node metastasis in primary conjunctival malignant melanoma. Br J Ophthalmol 2006;90:1468 –71. 16. Lee RJ, Gibbs JF, Proulx GM, et al. Nodal basin recurrence following lymph node dissection for melanoma: implications
Savar et al 䡠 Sentinel Node Biopsy for Ocular Adnexal Melanoma for adjuvant radiotherapy. Int J Radiat Oncol Biol Phys 2000; 46:467–74. 17. Houghton A, Coit D, Bloomer W, et al. NCCN melanoma practice guidelines. National Comprehensive Cancer Network. Oncology (Williston Park) 1998;12:153–77. 18. Amato M, Esmaeli B, Ahmadi MA, et al. Feasibility of preoperative lymphoscintigraphy for identification of sen-
tinel lymph nodes in patients with conjunctival and periocular skin malignancies. Ophthal Plast Reconstr Surg 2003;19:102– 6. 19. Nijhawan N, Ross MI, Diba R, et al. Experience with sentinel lymph node biopsy for eyelid and conjunctival malignancies at a cancer center. Ophthalmic Plast Reconstr Surg 2004;20: 291–5.
Footnotes and Financial Disclosures Originally received: December 9, 2008. Final revision: April 1, 2009. Accepted: April 3, 2009. Available online: September 19, 2009.
4
Manuscript no. 2008-1465.
1
Section of Ophthalmology, Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, Houston, Texas. 2
Department of Surgical Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas. 3
Departments of Pathology, The University of Texas MD Anderson Cancer Center, Houston, Texas.
Department of Dermatology, The University of Texas MD Anderson Cancer Center, Houston, Texas. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Bita Esmaeli, MD, FACS, Section of Ophthalmology, Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Unit 1445, Houston, TX 77030. E-mail: besmaeli@ mdanderson.org.
2223
Regional nodal status is the most important prognostic factor in patients with melanoma.1 Sentinel lymph node biopsy (SLNB) was first described for melanoma by Morton et al2 and has been shown to provide a disease-specific 5-year survival benefit for patients with intermediatethickness cutaneous melanomas.3 Although SLNB has become part of the standard of care for intermediate-thickness cutaneous melanomas, it has not gained wide acceptance in the management of ocular adnexal melanomas. Sentinel lymph node biopsy is based on the notion that certain tumors that spread through the lymphatics will metastasize to a first draining node in that region before spreading systemically. If the sentinel lymph node (SLN) is negative, the likelihood of involvement of other nodes or distant sites is low. If the SLN is positive, the patient requires further nodal resection or other treatment. Sentinel lymph node biopsy can prevent unnecessary and potentially morbid regional lymphadenectomy in patients with negative SLNs4; it can also identify patients with subclinical nodal © 2009 by the American Academy of Ophthalmology Published by Elsevier Inc.
metastases that would be missed with clinical or ultrasound examination alone. Sentinel lymph node biopsy for conjunctival melanoma was first reported by our group in 2001.5 Since then, Wilson et al6 have reported 2 cases of SLNB for ocular adnexal melanoma, and we have reported 8 additional cases.7 We report our expanded experience in 30 patients with ocular adnexal melanoma who underwent SLNB and report an updated SLNB positivity rate and the false-negative events for the technique.
Materials and Methods All patients who underwent SLNB for eyelid or conjunctival melanoma as part of a trial at The University of Texas MD Anderson Cancer Center between December 2000 and July 2008 are the subject of this report. Appropriate institutional review board approval was obtained for the conduct of this study. The eligibility criteria for SLNB were as follows: age ⱖ18 years; histologically documented invasive melanoma of ISSN 0161-6420/09/$–see front matter doi:10.1016/j.ophtha.2009.04.012
2217
Ophthalmology Volume 116, Number 11, November 2009
Figure 1. A, Preoperative lymphoscintigraphy in a patient with a conjunctival melanoma in the caruncle. B, Single photon emission computed tomography/computed tomography scan for the same patient shows the site of injection of technetium and C, drainage to a level II node.
the bulbar or palpebral conjunctiva or eyelid skin with Breslow equivalent thickness ⱖ1 mm; ability to undergo surgery under general anesthesia; chest x-ray, liver enzyme test, and head and neck computed tomography or magnetic resonance imaging negative for evidence of metastasis; sonography of the regional lymph nodes negative for evidence of metastasis; ability to give written informed consent; and, if of childbearing potential, a negative pregnancy test. In the early phases of the study, if the Breslow thickness for conjunctival tumors could not be reliably determined because of tangential cutting of the surgical specimens, the patient was allowed into the trial. In the later phases of the study, we excluded such patients unless the pathologist at
2218
The MD Anderson Cancer Center thought that the thickness was at least 1 mm even if the precise depth of invasion could not be determined. All patients underwent preoperative lymphoscintigraphy and intraoperative mapping. In addition, single photon emission computed tomography/computed tomography was performed in the last 6 patients (Fig 1A, B). This technique allows for higherresolution images than are produced by standard lymphoscintigraphy and permits better localization of the draining lymphatic basins. Approximately 1 hour before surgery, 0.2 ml (0.3 mCi) of technetium 99m-sulfur colloid was injected intradermally or subconjunctivally depending on the tumor location. A handheld
Savar et al 䡠 Sentinel Node Biopsy for Ocular Adnexal Melanoma gamma probe was used to identify SLNs transcutaneously. Nodes with radioactivity counts greater than twice the background level were considered positive. Sentinel lymph nodes were removed until the radioactivity counts in the nodal basin were less than twice the background level. The lymph nodes were sectioned at 2-mm intervals using the bread-loafing technique and entirely submitted for microscopic examination. A formalin-fixed, paraffin-embedded section was routinely stained with hematoxylin-eosin. If this initial section failed to reveal malignant cells, an additional hematoxylineosin–stained section was examined along with an immunohistochemical study using a pan-melanocytic cocktail (HMB-45, anti-tyrosinase, anti-MART1). Patients were considered to have had a false-negative event if there was no evidence of micrometastasis on SLNB but there was subsequent development of nodal disease during the follow-up period. For the purpose of this study, the false-negative event rate was defined as the number of false-negative events divided by the total number of patients with negative findings on SLNB. Variables assessed included age, gender, location of tumor, lymphoscintigraphy findings, number and location of SLNs biopsied, histopathologic findings for the primary tumor and the SLNs, postoperative complications, and length of follow-up. For continuous data, a t test was used to assess statistical significance.
Results Thirty patients underwent SLNB for ocular adnexal melanoma during the study period. Eight of these patients were previously reported by our group.7 There were 19 women and 11 men. The median age was 62 years (range, 24 – 80 years). Fourteen patients (47%) had melanomas limited to the bulbar conjunctiva, 8 patients (27%) had melanomas limited to the palpebral conjunctiva, 4 patients (13%) had melanomas involving both the bulbar and palpebral conjunctiva, and 4 patients (13%) had melanomas limited to the eyelid skin. At least 1 SLN was identified in all 30 patients. The median number of SLNs removed was 2 (mean, 2.3; range 1–5). The nodal basins from which the SLNs were removed are listed in Table 1. The median number of basins sampled was 2 (mean, 1.6; range, 1–3). The nodal basin most commonly sampled was the intraparotid; the next most commonly sampled basins were the submandibular (level I), preauricular, and superior cervical (level II). In 16 patients (53%), findings on lymphoscintigraphy correlated with intraoperative findings in at least 1 nodal basin. Seven patients (23%) had correlation in all basins biopsied. In 14 patients (47%), there was no correlation between preoperative lymphoscintigraphy and intraoperative findings. In 1 patient, no SLNs were identified during preoperative lymphoscintigraphy, but an SLN Table 1. Nodal Basins Sampled Basin Head Preauricular Intraparotid Neck Level I Level II Level III Level IV Level V Level not specified
No. of Patients (%) 9 (30) 16 (53) 11 (37) 6 (20) 3 (10) 1 (3) 1 (3) 2 (7)
was identified intraoperatively. This was accomplished by a slight modification of the injection technique for technetium 99m-sulfur colloid in this patient on the day of surgery. The median length of follow-up from the time of SLNB to last contact was 2 years (range, 10 months to 6 years).
Patients with Positive Sentinel Lymph Node Biopsy Findings Five patients (16.67%) had histologic evidence of nodal metastasis on SLNB. The first patient was a 59-year-old man with melanoma of the temporal bulbar conjunctiva of the left eye with Breslow thickness of 2.1 mm. At the time of presentation to MD Anderson, his conjunctival tumor had already been resected with negative margins; thus, no further resection of his conjunctiva was done. Four SLNs were identified: 2 intraparotid, 1 upper jugular (level II), and 1 middle jugular (level III). Both intraparotid nodes were positive. One parotid node showed a focus of tumor measuring 2.0⫻0.5 mm with extracapsular extension. The other showed a focus of approximately 20 spindle-shaped cells within the nodal parenchyma that were confirmed by immunohistochemistry to be melanocytic. Both jugular nodes were negative for micrometastasis. After SLNB, the patient returned to his hometown for follow-up. A parotidectomy and completion neck dissection was performed, and adjuvant external-beam radiation therapy was administered to the involved nodal basins. The patient had 3 local recurrences on the conjunctival surface, which were believed to be in-transit metastases. He eventually developed liver metastasis. He was treated with chemotherapy (Temozolomide; Schering-Plough Corporation, Kenilworth, NJ) and died of metastatic disease 22 months after SLNB. The second patient was a 67-year-old man with recurrent melanoma of the right upper eyelid palpebral conjunctiva with Breslow thickness of 6 mm. The single SLN identified, a submandibular (level I) node measuring 1.7⫻1.1⫻0.4 cm, was positive, with tumor replacing most of the nodal tissue with no extracapsular extension. Completion neck dissection was recommended. The patient returned home for further care. He underwent neck dissection, and 1 additional positive node was found. He was treated with postoperative adjuvant radiation therapy to the neck. He developed local and regional recurrence approximately 8 months later, as well as widespread metastatic disease, and died 3 years after SLNB. The third patient was a 54-year-old man with melanoma of the left lower lid palpebral conjunctiva and lid margin with Breslow thickness of 3.8 mm. Two SLNs were removed. The first, from the parotid gland, was positive, with the focus of metastatic melanoma measuring 0.8⫻0.5 mm with no extracapsular extension. The second, from the midjugular region (level III), was negative. The patient underwent parotidectomy and neck dissection at MD Anderson, and no additional nodes were positive for melanoma. At the time of this report, he has been followed for 13 months without evidence of local or regional recurrence or metastasis. The fourth patient was a 71-year-old woman with melanoma of the left lower eyelid skin extending to the lid margin but not involving the palpebral conjunctiva (Fig 2A), with Breslow thickness of 7.2 mm and Clark level IV. One preauricular SLN was identified; it contained a 2.5⫻2.0 mm subcapsular and intraparenchymal focus of metastatic melanoma without extracapsular extension (Fig 2B). The patient subsequently underwent parotidectomy and completion neck dissection at MD Anderson. Two of the 27 additional nodes removed were positive, both in the parotid gland. She remained free of disease at last contact 10 months after SLNB. The fifth patient was a 71-year-old woman who was referred to MD Anderson after incomplete excision of a melanoma of the left lower eyelid palpebral conjunctiva with Breslow thickness of 5.2
2219
Ophthalmology Volume 116, Number 11, November 2009 mm (Fig 3). She underwent complete surgical resection of her eyelid and SLNB. Three SLNs were identified, and all 3 were positive. The first was in the preauricular area and contained tumor measuring 6⫻3.5 mm with extracapsular extension. The other 2 nodes were intraparotid, and each contained fewer than 50 tumor cells. The patient underwent parotidectomy and completion neck dissection, with 4 of the 45 additional nodes removed positive. Adjuvant postoperative external-beam radiation therapy was de-
Figure 3. This patient had previously undergone gross excision of a palpebral conjunctival melanoma at another institution. A conjunctival scar remaining in the lower eyelid was the only sign of the primary tumor location. Technetium Tc 99m was injected subconjunctivally around this scar, and the draining sentinel lymph nodes were biopsied and found to be histologically positive.
livered to the parotid and neck area. Four months after SLNB, she developed recurrent nodal disease within the radiation field and is currently considering systemic treatment options for stage III conjunctival melanoma.
Figure 2. A, Patient with an eyelid melanoma extending to the eyelid margin. She had negative findings on magnetic resonance imaging and ultrasonography of her regional nodes but was found to have a (B) microscopically positive sentinel lymph node on sentinel lymph node biopsy (hematoxylin-eosin ⫻100).
2220
Figure 4. Histologic section of invasive melanoma with overlying ulceration and lacking epidermis (hematoxylin-eosin, ⫻200).
Savar et al 䡠 Sentinel Node Biopsy for Ocular Adnexal Melanoma Patients with Negative Sentinel Lymph Node Biopsy Findings Among the 25 patients with negative SLNB findings, 2 patients (8%) had a nodal recurrence (false-negative event). Both of these patients were among the first 10 patients who underwent SLNB for adnexal melanoma at our center; we did not observe any falsenegative events among patients treated during the last 4.5 years of the study. The details of these 2 patients were discussed in one of our previous publications.7 One additional patient with negative SLNB findings developed a local recurrence at multiple sites on her conjunctival surface and refused an orbital exenteration recommended at our center. She was treated at another center with plaque radiotherapy and experienced additional recurrences after that. She later died of widespread metastatic disease 30 months after SLNB. The remaining 22 patients remained free of disease at last follow-up, with a median follow-up time of 25 months (range, 10 months to 6 years).
Breslow Thickness and Ulceration Breslow thickness data were available for 26 patients (87%). The mean Breslow thickness was 2.57 mm (range, 0.62–12 mm) among patients with negative SLNB findings and 4.86 mm (range, 2.0 –7.2 mm) among patients with positive SLNB findings (P ⫽ 0.055). Data regarding the presence or absence of ulceration of the primary tumor on histopathologic examination were available for 28 patients (93%). Ulceration was present in 11 patients (39%) (Fig 4), including 4 (80%) of the 5 patients with positive SLNB findings and 7 (28%) of the 25 patients with negative SLNB findings, including both of the patients with false-negative results (P ⫽ 0.0618). Overall risk of nodal disease (i.e., either positive SNLB findings or false-negative SLNB findings) was significantly associated with ulceration (P ⫽ 0.0069).
Discussion Our study suggests that the results of SLNB provides important prognostic information for patients with ocular adnexal melanoma and can identify patients with early microscopic nodal metastasis, even as small as 0.5 mm or 20 cells. In this prospective trial, the overall SLN positivity rate was approximately 16%, and there were 2 false-negative events. Previous studies of intermediate-thickness cutaneous melanomas have shown similar rates of SLN positivity. Morton et al3 reported a rate of SLN micrometastasis of 16% in their randomized trial that included 764 SLNB procedures. Gomez-Rivera et al8 found that 20% of 113 patients with cutaneous melanoma of the head and neck had positive SLNs. Our false-negative event rate of 8% was slightly higher than previously reported rates for intermediate-thickness melanomas in other anatomic locations but close to rates reported for head and neck melanomas.9 Morton et al3 reported a false-negative rate of 3.4% among 764 patients undergoing SLNB. Their false-negative event rate as defined in our study was 4%. Schmalbach et al9 reported a false-negative rate of 4.5% (using the same definition as in our study). We have extensively discussed the potential reasons for false-negative events after SLNB for ocular adnexal melanoma in a previous publication.7 Both
of the false-negative events in our study occurred early in the study period. We believe that the false-negative event rate decreased because of our increased experience with the nuances of SLNB for ocular adnexal melanoma and improved selection of appropriate patients for SLNB. For example, 1 of the 2 false-negative events in our study occurred in a patient with a massive, rapidly growing and diffuse melanoma of 12-mm thickness involving the upper and lower eyelids and the bulbar and palpebral conjunctiva contiguously. For such massive tumors, the appropriate area of injection of technetium is difficult to determine and the thickness of the tumor makes distant-organ metastasis highly likely, making SLNB less useful. Today we would not perform SLNB on such a patient. Nevertheless, despite improved patient selection for SLNB, because of the 8% risk of false-negative events suggested in our study, we recommend that all patients with conjunctival or eyelid melanoma who have negative SLNB undergo long-term follow-up to rule out recurrence in the regional nodal basins. Another explanation for the improved false-negative results may be a shorter follow-up time for the most recent cases, but this is not likely to be the reason because most falsenegative events occurred within the first few months after SLN biopsy and each patient included in this report has had at least 10 months of follow-up. The goal of SLNB is to detect subclinical nodal metastasis and at the same time spare patients without nodal disease the morbidity of extensive nodal dissection.4 Four (80%) of the 5 patients with positive SLNs in our study had additional positive nodes found at the time of completion lymphadenectomy and therefore benefited from early identification by SLNB. Debate exists over whether the use of SLNB confers an absolute survival benefit when compared with observation.10 A change in survival is particularly difficult to demonstrate because the majority of patients undergoing SLNB will be expected to have no evidence of micrometastasis. However, in a recent randomized trial in which 1269 patients with an intermediate-thickness primary cutaneous melanoma were randomly assigned to SLNB or to observation, the patients who were found to have microscopically positive SLNs experienced a melanoma-specific survival benefit compared with the patients in the observation arm whose nodal disease was found by the wait-andwatch method.3 Breslow thickness is a well-known prognostic indicator for both cutaneous11 and conjunctival melanomas.12 Obtaining accurate histologic data for this purpose is especially important when dealing with delicate tissue such as conjunctiva and eyelid skin. Improper tissue handling and tangential sectioning can prevent the pathologist from making these measurements.13 Tuomaala and Kivelä14 analyzed 85 conjunctival melanomas and concluded that SLNB is most appropriate in lesions with Breslow thickness of at least 2 mm and for nonlimbal lesions. In our series, the thinnest tumor to have a positive SLN had a Breslow thickness of exactly 2.0 mm. However, it might still be useful to perform SLN biopsy for tumors less than 2.0 mm thick if they are associated with ulceration. The presence of ulceration may be another high-risk histologic feature and may predict nodal metastasis. In our series, development of nodal dis-
2221
Ophthalmology Volume 116, Number 11, November 2009 ease (positive SLN or false-negative event) was statistically significantly associated with ulceration. Although there was a trend toward SLN positivity among patients with ulceration, this difference was not statistically significant. In our study, 47% of patients had SLNs present in both the head (parotid and preauricular regions) and the neck (level I–V nodes, i.e., below the jaw). Lim et al15 looked at the pattern of nodal metastasis in 12 patients with conjunctival melanoma and found that temporal lesions tended to metastasize to the preauricular nodes, whereas nasal lesions tended to metastasize to the submandibular nodes. Preoperative lymphoscintigraphy and SLNB data from our study did not corroborate this distribution. Preoperative lymphoscintigraphy allows one to determine the location of the SLN before surgery. At least 1 SLN was identified in all but 1 patient in our study during preoperative lymphatic mapping. In the 1 patient with no drainage identified during preoperative lymphoscintigraphy, drainage to SLNs was seen intraoperatively. This was accomplished by modifying the technique for injection of technetium slightly on the day of surgery to ensure drainage from the ocular surface or the eyelid. This is particularly important when the thin eyelid skin is injected, because the injection has to be intradermal for ideal drainage to the SLNs. Single photon emission computed tomography/ computed tomography has become a standard part of the preoperative lymphatic mapping procedure at the MD Anderson Cancer Center and can provide far more anatomic detail than standard lymphoscintigraphy. Patients with conjunctival or eyelid melanoma who have positive SLNB findings should be referred for completion node dissection and parotidectomy if the parotid basin is involved. If additional positive nodes are found during parotidectomy and neck dissection, the patient should be considered for adjuvant radiation therapy delivered to the nodal basins. The presence of extracapsular extension increases the risk of nodal recurrence.16 It has been recommended that such patients, as well as those with multiple positive nodes, be considered for adjuvant irradiation of the involved nodal basins after nodal dissection.17 The technical aspects and safety profile of SLNB have been extensively described in our previous publications.18,19 To date, we have encountered only 3 cases of temporary weakness of the marginal mandibular branch of the facial nerve, which have resolved spontaneously. There have been no other side effects attributable to SLNB. In our experience, the procedure has proven to be safe, takes 30 to 90 minutes, and can be combined with the primary resection of the conjunctival or eyelid mass or, when referral to our center takes place after resection of the primary ocular tumor, done as a secondary procedure within a few weeks after the ocular tumor resection. In conclusion, to our knowledge, this is the largest reported series of SLNB for ocular adnexal melanoma. Despite this, the absolute number of cases is small, limiting our ability to draw strong conclusions with respect to a survival benefit from SLNB. However, we can say that SLNB is a safe procedure and can identify nodal micrometastasis in patients with ocular adnexal melanoma. The SLNB positivity rate is similar to rates reported for intermediate-thickness
2222
cutaneous melanomas in all other anatomic sites. The falsenegative event rate in our series was higher than that reported in other anatomic sites but has improved in the last 4 years, we believe as a result of improved technique and better patient selection. We recommend consideration of SLNB for patients with intermediate-thickness ocular adnexal melanoma, especially if ulceration is identified.
References 1. 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;19:3622–34. 2. Morton DL, Wen DR, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392–9. 3. Morton DL, Thompson JF, Cochran AJ, et al., MSLT Group. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med 2006;355:1307–17. 4. van Akkooi AC, Bouwhuis MG, van Geel AN, et al. Morbidity and prognosis after therapeutic lymph node dissections for malignant melanoma. Eur J Surg Oncol 2007;33:102– 8. 5. Esmaeli B, Eicher S, Popp J, et al. Sentinel lymph node biopsy for conjunctival melanoma. Ophthal Plast Reconstr Surg 2001;17:436 – 42. 6. Wilson MW, Fleming JC, Fleming RM, Haik BG. Sentinel node biopsy for orbital and ocular adnexal tumors. Ophthal Plast Reconstr Surg 2001;17:338 – 44. 7. Ho VH, Ross MI, Prieto VG, et al. Sentinel lymph node biopsy for sebaceous cell carcinoma and melanoma of the ocular adnexa. Arch Otolaryngol Head Neck Surg 2007;133: 820 – 6. 8. Gomez-Rivera F, Santillan A, McMurphey AB, et al. Sentinel node biopsy in patients with cutaneous melanoma of the head and neck: recurrence and survival study. Head Neck 2008;30: 1284 –94. 9. Schmalbach CE, Nussenbaum B, Rees RS, et al. Reliability of sentinel lymph node mapping with biopsy for head and neck cutaneous melanoma. Arch Otolaryngol Head Neck Surg 2003;129:61–5. 10. De Wilt JH, van Akkooi AC, Verhoef C, Eggermont AM. Detection of melanoma micrometastases in sentinel nodes— the cons. Surg Oncol 2008;17:175– 81. 11. Breslow A. Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970;172:902– 8. 12. Tuomaala S, Eskelin S, Tarkkanen A, Kivelä T. Populationbased assessment of clinical characteristics predicting outcome of conjunctival melanoma in whites. Invest Ophthalmol Vis Sci 2002;43:3399 – 408. 13. Esmaeli B, De Martelaere S, Ozkan Y, et al. Surgical specimen handling for conjunctival melanoma: implications for tumor thickness determination and sentinel lymph node biopsy [letter]. Ophthal Plast Reconstr Surg 2006;22:238 –9. 14. Tuomaala S, Kivelä T. Metastatic pattern and survival in disseminated conjunctival melanoma: implications for sentinel lymph node biopsy. Ophthalmology 2004;111:816 –21. 15. Lim M, Tatla T, Hersh D, Hungerford J. Patterns of regional head and neck lymph node metastasis in primary conjunctival malignant melanoma. Br J Ophthalmol 2006;90:1468 –71. 16. Lee RJ, Gibbs JF, Proulx GM, et al. Nodal basin recurrence following lymph node dissection for melanoma: implications
Savar et al 䡠 Sentinel Node Biopsy for Ocular Adnexal Melanoma for adjuvant radiotherapy. Int J Radiat Oncol Biol Phys 2000; 46:467–74. 17. Houghton A, Coit D, Bloomer W, et al. NCCN melanoma practice guidelines. National Comprehensive Cancer Network. Oncology (Williston Park) 1998;12:153–77. 18. Amato M, Esmaeli B, Ahmadi MA, et al. Feasibility of preoperative lymphoscintigraphy for identification of sen-
tinel lymph nodes in patients with conjunctival and periocular skin malignancies. Ophthal Plast Reconstr Surg 2003;19:102– 6. 19. Nijhawan N, Ross MI, Diba R, et al. Experience with sentinel lymph node biopsy for eyelid and conjunctival malignancies at a cancer center. Ophthalmic Plast Reconstr Surg 2004;20: 291–5.
Footnotes and Financial Disclosures Originally received: December 9, 2008. Final revision: April 1, 2009. Accepted: April 3, 2009. Available online: September 19, 2009.
4
Manuscript no. 2008-1465.
1
Section of Ophthalmology, Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, Houston, Texas. 2
Department of Surgical Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas. 3
Departments of Pathology, The University of Texas MD Anderson Cancer Center, Houston, Texas.
Department of Dermatology, The University of Texas MD Anderson Cancer Center, Houston, Texas. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Bita Esmaeli, MD, FACS, Section of Ophthalmology, Department of Head and Neck Surgery, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Unit 1445, Houston, TX 77030. E-mail: besmaeli@ mdanderson.org.
2223