Aberrant lymphatic drainage and risk for melanoma recurrence after negative sentinel node biopsy in middle-aged and older men

Transcription

1 ORIGINAL ARTICLE Aberrant lymphatic drainage and risk for melanoma recurrence after negative sentinel node biopsy in middle-aged and older men Anthony H. Kaveh, BS, 1 Nicole M. Seminara, MD, 1 Melynda A. Barnes, MD, 2 Aaron J. Berger, MD, 3 Frank W. Chen, MD, 1 Mike Yao, MD, 1,2 Denise Johnson, MD, 4 Sean Parsa, MD, 4 Andrew Quon, MD, 5 Susan M. Swetter, MD, 1,6 John B. Sunwoo, MD 1,2 * 1 Department of Dermatology, Pigmented Lesion and Melanoma Program, Stanford University School of Medicine, Stanford, California, 2 Department of Otolaryngology, Division of Head and Neck Surgery, Stanford University School of Medicine, Stanford, California, 3 Department of Surgery, Division of Plastic Surgery, Stanford University School of Medicine, Stanford, California, 4 Department of Surgery, Stanford University School of Medicine, Stanford, California, 5 Department of Radiology, Molecular Imaging Program at Stanford, Stanford University, Stanford, California, 6 Dermatology Service, Veterans Affairs Palo Alto Health Care System, Palo Alto, California. Accepted 16 April 2015 Published online 14 July 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI /hed ABSTRACT: Background. Aberrant lymphatic drainage is believed to contribute to the high recurrence rate of head and neck melanomas. The purpose of this study was to identify the clinical significance of unexpected lymphatic drainage patterns. Methods. A single institution retrospective analysis was performed of middle-aged and older men (mean age, 66.2 years; range, years) who underwent successful lymphoscintigraphy with sentinel biopsy (SLNB) from 1997 through Node status, distribution, and recurrence were assessed comparing patients with expected and unexpected drainage patterns. Results. Sixty-six patients were identified with 55.8 months median follow-up (range, months). Unexpected sentinel drainage was associated with multiple basin drainage (p <.01) and greater recurrence after negative SLNB (p 5.03). Both groups had similar anatomic distribution, sentinel sampling, histopathologic characteristics, follow-up, and survival. Conclusion. Lymphatic drainage differing from expected patterns is associated with greater recurrence after negative SLNB in middle-aged and older men. VC 2015 Wiley Periodicals, Inc. Head Neck 38: E754 E760, 2016 KEY WORDS: recurrence, negative sentinel node biopsy, head and neck melanoma, lymphatic drainage, discordant INTRODUCTION Cutaneous melanoma is the deadliest form of skin cancer, and its incidence continues to climb, particularly in middle-aged and older men. It is predicted that the incidence of melanoma will exceed that of colon cancer in the United States by the early 2020s. 1 Melanomas arising from sites in the head and neck account for approximately 20% of cases, and these have been reported to have a worse prognosis compared with other sites on the trunk and extremities. 2,3 Lymphoscintigraphy-guided sentinel biopsy (SLNB) has become the established and preferred method for pathologic staging of regional s to assess for micrometastases. 4,5 Worldwide data support that sentinel status provides the most important prognostic information for melanoma recurrence and survival. The final analysis of the Multicenter Selective Lymphadenectomy Trial-I (MSLT-I) demonstrated that patients with intermediate thickness melanoma ( mm) and *Corresponding author: J. B. Sunwoo, Department of Otolaryngology, Division of Head and Neck Surgery, Stanford University School of Medicine, 801 Welch Road, 2nd Floor, Stanford, CA sunwoo@stanford.edu John B. Sunwoo and Susan M. Swetter contributed equally to this work. positive SLNB had a 10-year melanoma-specific survival rate of 62% compared to 85% in those with a negative SLNB. 6 In patients with thick melanomas (>3.5 mm), survival rates were 48% and 65%, respectively. In addition to prognostic information, sentinel status also provides important information for therapeutic decision-making, enabling early comprehensive lymphadenectomy of the draining nodal basins, which, in MSLT-1, was associated with an improvement in disease-specific survival. 6 Although the technique of SLNB has been shown in multiple studies to be feasible and reliable, 5,7 there is still a significant rate of regional recurrence, 8 especially in cases in which the primary melanoma is located in the head and neck region, 9,10 and this regional recurrence can be interpreted to be a false-negative finding at the time of the SLNB, with potentially poorer prognosis. 11 The reasons for the higher false-negative rate in SLNB of head and neck melanoma cases have been proposed to include the close proximity of primary head and neck tumors to regional nodal basins, multiple lymphatic drainage sites, and technically challenging surgical sites, particularly the parotid. 9 Head and neck melanomas are also more likely to manifest unexpected lymphatic drainage 12 because of their complex and rich lymphatics. The clinical significance of unexpected drainage remains unclear, because E754 HEAD & NECK DOI /HED APRIL 2016

2 ABERRANT LYMPHATIC DRAINAGE AND MELANOMA RECURRENCE few studies have compared lymphatic anatomy to cancer biology and outcomes before and after intervention. 13,14 However, correlations between ambiguous, multiple lymphatic basin drainage from scalp primaries and regional recurrence after negative SLNB have been observed. 15 Age may also play a role in recurrence after a negative SLNB. The observations of decreasing SLNB positivity with age despite thicker melanoma burden 16 and higher false-negative SLNB rates 17 may be a result of agerelated declines in lymphatic flow and filtration, termed lymphatic dysfunction. 18 This phenomenon may affect the typical sequential progression through regional afferent lymphatics, bypassing first echelon nodes and resulting in unexpected drainage patterns. We hypothesized that unexpected (discordant) lymphatic drainage may contribute to a greater rate of recurrence after negative SLNBs in patients with head and neck melanoma. To evaluate this, we compared outcomes for middle-aged and older men (40 years of age and older) with discordant and concordant lymphatic drainage patterns to assess differences in regional nodal and distant recurrence rates after a negative SLNB. MATERIALS AND METHODS This study was approved by the Stanford University Institutional Review Board. A retrospective review of the Stanford Cancer Center Tumor Registry database was performed to identify patients who had histologic evidence of primary head and neck cutaneous melanoma and who had undergone SLNB staging at the time of their resection at Stanford University Medical Center between August 1997 and April Patients with conjunctival and mucosal melanomas were excluded from this study to limit pathobiological variability. 19 There are a number of studies that have demonstrated a strong association in melanoma between survival and sex (male vs female). These studies, including previous work by our group, 20 have shown that men have a significantly worse prognosis compared with women and that this observation is independent of tumor stage. Further, age has been associated with melanoma outcome and treatment-related factors, such as false-negative SLNB. 17 Thus, to limit variability in sex and risks conferred by adolescent and young adult age, this study was restricted to men aged 40 years and older, who comprised the majority of patients with head and neck melanoma in our database. Cases with operative reports lacking specific SLNB anatomic locations were excluded, resulting in removal of 45 cases. All pathology reports were reviewed for parameters of Breslow depth, mitotic index, ulceration, lymphovascular invasion (LVI), and perineural invasion (PNI) of the primary lesion. Complete histopathologic data were available for all patients except 12 cases that did not include mitotic index. Lymphatic drainage patterns were categorized as concordant or discordant based on expected anatomic patterns described by O Brien et al 23 and defined in Table 1. As with prior studies, any drainage outside of the predicted basins was categorized as discordant. 12,13 Midline lesions were considered concordant if drainage occurred in expected sites unilaterally or bilaterally. Primary anatomic sites included the scalp, face, ears, and neck. Sentinel TABLE 1. Primary site Predicted sites of metastatic spread. Forehead Face Ear Coronal scalp Posterior scalp Upper anterior neck Posterior neck Predicted site of drainage Parotid, levels I III Parotid, levels I III Parotid, levels I V Parotid, levels I V Occipital, levels II V Levels I IV Occipital, levels II V basins were categorized as parotid or neck, with the neck divided into the conventional 6 anatomic levels, as per standard surgical/anatomic definitions. 24 The lymphoscintigraphy procedure was performed by the Stanford Nuclear Medicine Division and utilized technetium 99-labeled sulfur colloid injection at the site of the primary melanoma. Multiple injections were made intradermally immediately around the melanoma site (perilesionally). Dynamic and static planar imaging was collected. In 46.8% of cases, single photon emission CT (SPECT)/CT imaging was also performed. In the operating room, 15 minutes before incision was performed, perilesional injection of isosulfan blue dye was performed intradermally in 69.4% of patients as an adjunctive aid for intraoperative identification of draining nodes. To localize nodes containing the radioactive colloid, a handheld gamma radiation probe (Neoprobe, Dublin, OH) was used to identify nodes containing 90% of the ex vivo radioactive count. Tissue was processed by the Stanford Dermatopathology Division using formalin-fixed paraffinembedded sections and analyzed with hematoxylin-eosin staining and immunohistochemical stains to identify melanoma-specific antigens, as per established histopathological standards in sentinel processing and histopathologic evaluation. 25 Follow-up intervals were calculated as the time elapsed from diagnosis until the date of the last patient follow-up or death. Recurrence time was calculated as time from diagnosis until determination of recurrence by either biopsy or imaging in clinical context. Recurrence was designated as either regional nodal or distant metastasis without regional nodal recurrence (isolated distant). Because our study focused on recurrence in relation to SLNB status, local and in-transit recurrences proximal to the sentinel basin were not included. The outcomes evaluated in this study were the number of drained sentinel basins, sentinel involvement by metastasis, and disease recurrence. A recurrence after a negative SLNB was defined as a nodal regional or distant metastasis in a patient with a prior negative SLNB and without prior evidence of in-transit, nodal, or distant disease. Recurrence rate was defined as the ratio of patients with recurrence after a negative SLNB to all patients with negative SLNBs. Characteristics of the studied cohorts were compared using the 2-tailed Fisher exact test and the unpaired t test, defining p values <.05 as significant. Statistical analyses were performed with SAS software (SAS Institute, Cary, NC). Sensitivity analysis was performed for several subgroup characteristics, in which that specific subgroup was HEAD & NECK DOI /HED APRIL 2016 E755

3 KAVEH ET AL. TABLE 2. Characteristics of concordant and discordant lymphatic drainage groups. Population average Concordant sentinel Discordant sentinel p value No. of patients Average age, y Mean follow-up, mo Percentage alive at last follow-up 62.1% 62.0% 62.5% 1.0 Average Breslow depth, mm Prevalence LVI 7.6% 6.0% 12.5%.59 Prevalence PNI 13.6% 10.0% 25.0%.20 Average mitotic index, mitoses per mm Prevalence ulceration 24.2% 20.0% 37.5%.19 Tumor type Desmoplastic and spindle cell 19.7% 20.0% 18.8% 1 Lentigo maligna melanoma 19.7% 18.0% 25.0%.72 Nodular 9.1% 8.0% 12.5%.63 Superficial spreading 43.9% 44.0% 43.8% 1.0 Primary site Ear 24.2% 28.0% 12.5%.32 Face 37.9% 38.0% 37.5% 1.0 Neck 12.1% 8.0% 25.0%.090 Scalp 25.8% 26.0% 25.0% 1.0 Sentinel drainage basin Percent sentinel drainage to neck 77.3% 76.0% 81.3% 1.0 Percent sentinel drainage to parotid 33.3% 32.0% 37.5%.76 Average number of sentinel s harvested Percent multibasin sentinel drainage 25.8% 14.0% 62.5% Percent of positive SLNB 18.2% 22.0% 6.3%.27 Abbreviations: LVI, lymphovascular invasion; PNI, perineural invasion; SLNB, sentinel biopsy. removed from the analysis to assess the robustness of our results to potential confounders that may have otherwise explained observed findings. These subgroups included patients with and without SPECT/CT imaging, primary neck lesions, LVI, PNI, ulceration, incomplete histologic data, and patients with multiple basin drainage. RESULTS Patient overview We identified 111 male patients, aged 40 years and older, who had undergone surgical resection of a head and neck cutaneous melanoma with SLNB staging. This subset of patients accounted for 64% of all head and neck melanoma cases. Of these 111 patients, 66 had complete clinical and pathologic information for inclusion in the study. Patient and tumor characteristics of the 66 cases are presented in Table 2. The mean age of the study population was 66.2 years (range, years), with mean follow-up of 55.8 months (range, months). The majority of the melanoma cases were superficial spreading subtype (43.9%); the remainders were lentigo maligna melanoma (19.7%), desmoplastic and spindle cell melanoma (19.7%), nodular melanoma (10.6%), and unspecified (6.0%). Primary melanomas were located on the head in 87.9% of patients (25.8% on the scalp, 37.9% on the face, and 24.2% on the ears) and on the neck in 12.1% of patients. Mean Breslow depth was 3.54 mm, 24.2% were ulcerated, and mean mitotic index was 3.11/ mm 2. Rates of LVI and PNI were 7.6% and 13.6%, respectively. Sentinel s were identified in the parotid and cervical (neck) region in 33% and 77% of patients, respectively. Discordant drainage to sentinel s was observed in 24% of cases (Table 3). The average rate of sentinel positivity was 18.2%. Characteristics of cases with concordant versus discordant lymphatic drainage The study cases were categorized as having concordant or discordant lymphatic drainage from the primary site, 23 based on preoperative lymphoscintigraphic evidence of drainage outside of the expected draining nodal basins (Table 1). The proportion of cases in which SPECT/CT was utilized at the time of the lymphoscintigraphy was similar in both the concordant and discordant groups (44.7% and 53.3%, respectively; p 5.77). There was no significant difference in rate of positive SLNB in patients with and without SPECT/CT imaging (20.7% and 9.1%, respectively; p 5.28). In addition, the use of intraoperative blue dye injection to aid with the identification of the sentinel was similar between the concordant and discordant groups (72.3% and 60.0%, respectively; p 5.52). Distribution of the melanoma primary sites (scalp, face, ears, and neck) and the tumor histopathologic characteristics were similar in both the concordant (n 5 50) and discordant (n 5 16) groups. There were no significant differences observed in Breslow depth, mitotic index, LVI, PNI, ulceration, or tumor subtype between the 2 groups. The discordant group seemed to have a higher prevalence of primary neck lesions (25.0% in the E756 HEAD & NECK DOI /HED APRIL 2016

4 ABERRANT LYMPHATIC DRAINAGE AND MELANOMA RECURRENCE TABLE 3. Characteristics of patients with discordant sentinel drainage. No. of sentinel s Sentinel basins Unexpected drainage Breslow depth, mm LVI PNI MI Ulceration Type Primary site Case no No No 0 No LMM Left cheek 0/3 Left neck levels II, V Neck level V No No 1 No Superficial spreading Right ear 0/3 Right parotid, right supraclavicular Supraclavicular neck No No 1 No LMM Left ear 0/2 Left posterior neck Posterior neck No Yes 0 No Desmoplastic Left posterior neck 0/4 Left parotid tail, left supraclavicular Parotid tail No No 3 No Superficial spreading Left temple 0/6 Left neck, left supraclavicular neck Supraclavicular neck 6 4 No No 3 Yes Nodular Left postauricular neck 0/5 Left Submandibular 7 4 Yes No 1 Yes Nodular Right postauricular neck 0/10 Right Submandibular 8 4 No No 3 Yes Superficial spreading Left postauricular neck 0/1 Left neck level I Neck level I 9 5 No Yes 7 No LMM Right nose 0/3 Bilateral submental and Left-sided drainage Inferolateral neck 10 7 Yes Yes 1 No LMM Right forehead 0/3 Right preparotid, inferolateral neck, 11 8 No No No Desmoplastic Right posterior scalp 1/1 Superficial parotid Parotid No No 4 No Superficial spreading Right posterior scalp 0/2 Right preauricular Preauricular No No 1 Yes Superficial spreading Left posterior scalp 0/2 Bilateral neck level V Right-sided drainage No No 5 Yes Superficial spreading Right cheek 0/9 Right neck levels II and V Neck level V Postauricular neck No No 1 Yes Superficial spreading Left temple 0/4 Left neck level IIb, postauricular neck, and parotid No Yes 1 No Desmoplastic Right scalp 0/8 Bilateral suboccipital and subclavicular Left-sided drainage Abbreviations: LVI, lymphovascular invasion; PNI, perineural invasion; MI, mitotic index (per mm 2 ); LMM, lentigo maligna melanoma. discordant group vs 8.0% in the concordant group), LVI (12.5% in the discordant group vs 6.0% in the concordant group), PNI (25.0% in the discordant group vs 10.0% in the concordant group), and ulceration (37.5% in the discordant group vs 20.0% in the concordant group). Although none of these differences reached statistical significance, we recognized this may be due to the small sample size. This potential bias was addressed with subsequent sensitivity analyses. Several important clinical differences between the 2 groups were observed. The mean age in the discordant group was 9 years older than that of the concordant group (p 5.008). In addition, although the rates of sentinel drainage to the parotid (32.0% in the concordant group vs 37.5% in the discordant group) and the neck (76.0% in the concordant group vs 81.3% in the discordant group) were similar between the 2 groups, patients with discordant drainage had a significantly higher rate of multiple lymphatic basin drainage (14.0% in the concordant group vs 62.5% in the discordant group; p <.001). As such, there were, on average, a greater number of sentinel s excised among the cases in the discordant group (mean of 4.1 nodes in the discordant group vs 3.3 in the concordant group). Despite this greater number of sampled nodes, the discordant cases tended to have a lower rate of SLNB positivity (6.3%) compared with those in the concordant group (22.0%). Higher rates of recurrence after negative sentinel lymph node biopsy in patients with discordant drainage Although the overall survival (OS) was comparable between both concordant and discordant groups (62.0 vs 62.5%, respectively), we hypothesized that the greater incidence of multiple lymphatic basin drainage and lower rate of SLNB positivity observed in the discordant group may have affected the recurrence rate despite negative SLNBs. An examination of overall recurrence in our study population revealed a significantly higher rate of recurrence after negative SLNBs among the cases in the discordant drainage group compared with those in the concordant drainage group (33.3% vs 7.7%, respectively; p 5.03; Figure 1; Tables 4 and 5), with mean time to recurrence of 30.5 months (range, months). When the group was examined for isolated nodal recurrence or isolated distant metastatic recurrence, the discordant drainage group had a higher incidence in both categories after a negative SLNB when compared to the concordant drainage group (Figure 1). Although these rates were not individually statistically significant (isolated nodal recurrence p 5.12; isolated distant recurrence p 5.18), the subgroup sample sizes were small, and a 50% larger sample size was estimated to be needed to show significance. Nevertheless, in aggregate, there was a significantly higher rate of recurrence after negative SLNBs observed in the discordant drainage group. To evaluate the robustness of these findings, several sensitivity analyses were performed. These sensitivity analyses specifically addressed the higher rates of ulceration, LVI, PNI, and neck primary site in the discordant group. Patients with tumors having these traits were individually removed from the comparison between HEAD & NECK DOI /HED APRIL 2016 E757

5 KAVEH ET AL. TABLE 4. Rates of recurrence after negative sentinel node biopsy. Recurrence rates Recurrence type Population average Concordant sentinel Discordant sentinel p value Regional 9.3% 5.1% 20.0%.12 Isolated distant 5.6% 2.6% 13.3%.18 Any (excludes local) 14.8% 7.7% 33.3%.03 discordant and concordant patients. In addition, patients with missing reports of mitotic index were removed. In all analyses, patients with discordant drainage still exhibited increased recurrence rates after negative SLNBs to levels within 25% of what was observed in the initial comparison. Given the strong association between discordant lymphatic drainage and multiple lymphatic basin drainage (Table 2), the recurrence rate in the subgroup of patients with multiple lymphatic basin drainage was compared with patients with single basin drainage. The multiple lymphatic basin drainage population was found to have a significantly higher recurrence rate after negative SLNBs than the single nodal basin drainage population (p <.05), consistent with previous reports of an association between multiple lymphatic basin drainage and false-negative SLNB in truncal melanomas. 27 To determine whether discordancy represented an independent association with higher incidence of recurrence after negative SLNB, sensitivity analysis was performed in which patients with multiple lymphatic basin drainage were removed from our evaluation of discordant and concordant patients. Our findings of increased recurrence after negative SLNBs remained within 15% of the original analysis. DISCUSSION Cutaneous melanomas of the head and neck have higher rates of recurrence after negative SLNBs compared with those of the trunk and extremities. In a systematic review of 3442 patients with head and neck melanomas examined in 32 published studies between 1990 and 2009, the mean false-negative rate for nodal recurrence was 20.4%. 9 This contrasts with reported false-negative rates of 3% to 11% when considering all sites. 17,28,29 Location of a melanoma primary lesion on the head and neck region is an independent predictor of recurrence with an odds ratio of 3.02 (95% confidence interval ) reported in a study of 515 patients with a negative SLNB. 26 The reasons for this increased rate of recurrence are not clear but may be related to the abundant lymphatic drainage and other biologic factors associated with this region. Lymphatic spread from cutaneous malignancies of the head and neck often have unpredictable and highly variable patterns, and this is one of the rationales for using lymphoscintigraphy to guide nodal sampling. However, from our observations and that of others, 13,30 32 it is clear that the majority of patients have drainage patterns that follow predicted ( concordant ) patterns (Table 1), and that there is a minor but substantial group of patients who have drainage patterns that are outside of these predicted patterns ( discordant ). The principal finding of our study is that patients with discordant patterns of drainage (seen on preoperative lymphoscintigraphy) have higher incidences of recurrence after negative SLNBs than patients with concordant patterns. Overall, the incidence of concordant drainage in our series (76%) was consistent with the reported range of 57% to 92%. 13,30 32 Although discordant lymphatic drainage can result from iatrogenic disruption of lymphatic channels after wide local excision with wide margins and skin graft or rotation flap closure, 33 none of the subjects in our study had prior surgery around the area of the primary tumors. Thus, the discordant lymphatic drainage seems to be an inherent biologic or anatomic characteristic of the individual patients and/or tumors. In the head and neck, discordant lymphatic drainage may be a consequence of the region s dense and complicated lymphatics and vasculature In addition, primary dysfunction of this complex lymphatic anatomy in older patients may disrupt the orderly sequence of lymph node progression, resulting in sampling of incorrect drainage basins. Age-related declines in function, termed senescence 37 and intranodal shunting, 38 have been proposed to be factors involved in reducing the potential for trapping metastatic cells and facilitating skip metastases that may be observed as discordant drainage. It is important to note that there was, overall, a greater number of sentinel nodes sampled per patient in the discordant group, and, thus, the higher rate of recurrence after negative SLNBs observed in the FIGURE 1. Rates of recurrence after negative sentinel biopsy (SLNB). Patients were categorized as having lymphatic drainage from the primary site that was either concordant or discordant, based on established and previously described definitions. Higher rates of recurrence after negative SLNB were observed in the discordant drainage group. *p E758 HEAD & NECK DOI /HED APRIL 2016

6 ABERRANT LYMPHATIC DRAINAGE AND MELANOMA RECURRENCE TABLE 5. Characteristics of patients with regional nodal or distant recurrence after negative sentinel node biopsy. Recurrence interval, mo Recurrence site Breslow depth No. of sentinel s Type Ulceration MI LVI PNI Unexpected drainage Sentinel lymph node location Case no. Primary site 1 Right forehead Right parotid None 2 Spindle cell Yes 4 No Yes 9 Lungs Left posterior scalp Posterior left neck None 4 Desmoplastic No 0 No No 6 Left posterior neck 7.6 No 2 No No 3 Left parotid 42.4 None 4 Superficial spreading 3 Forehead Left parotid, right neck 4 Left temple Left supraclavicular Supraclavicular neck 6 Superficial No 3 No No 3.5 Left parotid 25.2 neck, left neck spreading 5 Left postauricular neck Left Submandibular 5 Nodular Yes 3 No No 4 Bone Left-sided drainage 3 LMM No 7 No Yes 5 Lung, adrenal, femur, CNS 6 Right side of nose Bilateral submental and Inferolateral neck 3 LMM No 1 Yes Yes 7 Right parotid, CNS Right forehead Right inferolateral neck, preparotid, 13.4 Yes 1 No No 18 Left level II, lungs, liver Postauricular neck 4 Superficial spreading 8 Left temple Left postauricular neck, parotid, and level IIb Abbreviations: MI, mitotic index (per mm 2 ); LVI, lymphovascular invasion; PNI, perineural invasion; LMM, lentigo maligna melanoma; CNS, central nervous system. discordant group does not seem to be the result of technical issues in node sampling. Rather, the discrepancy reflected in the lower incidence of positive SLNBs in the discordant group (6.3%) versus that in the concordant group (22%) and the higher recurrence rate in the discordant group (33%) versus that in the concordant group (7.7%) is hypothesized to be related to the possible biologic explanations discussed above. Furthermore, the patterns and locations of failure (Table 5; Figure 1) indicate that a significant proportion of the failures in the discordant group were distant recurrences, suggesting an underlying difference in the biologic process of spread in this group. In our study, we also observed a higher incidence of multiple lymphatic basin drainage in the discordant group. The complex anatomy of the head and neck is believed to contribute to an overall higher rate of multiple lymphatic basin drainage in this region. 39 In addition, collateral lymphatic formation in response to paracrine lymphangiogenic effects from tumor-induced factors may accentuate this and may explain new lymphatic channels that result in discordant drainage patterns. 40,41 Although the clinical significance of multiple lymphatic basin drainage in head and neck melanoma is unclear, 42 multiple lymphatic basin drainage at truncal sites may portend a worse prognosis, 39 and we note the significant overlap of discordancy and multiple lymphatic basin drainage in our series. Nonetheless, sensitivity analysis preserved our findings after removing cases of multiple lymphatic basin drainage, suggesting discordant drainage to be an independent factor driving recurrence after a negative SLNB. Age has been proposed to be a factor in false-negative SLNBs. In the Sunbelt Melanoma Trial, 17 there was an age-related increase in false-negative SLNBs that was observed. In our study, the discordant drainage group was older than the concordant drainage group. However, the patients with recurrences after negative SLNBs had similar ages, indicating that this was not a factor in our results. Although our study showed an increased rate of recurrence after negative SLNBs in patients with discordant drainage, impact on OS remains unclear. Parrett et al 43 found no significant differences in OS between truepositive and false-negative SLNBs in a retrospective study of 365 patients with head and neck melanoma. The final report of the MSLT-1 trial, 6 however, showed patients with false-negative SLNBs to have similar melanoma-specific survival as patients with lymphadenectomy delayed until identification of palpable nodes, a group with significantly lower melanoma-specific survival than patients with true-positive SLNBs. The Sunbelt Melanoma Trial 17 showed significant reduction in diseasefree survival, but not in OS, in patients with falsenegative SLNBs compared with those with true-positive SLNBs. Neither of these 2 prospective studies, however, focused exclusively on head and neck melanoma cases. As with any observational study, potential confounders must be addressed. Our study focused on a high-risk demographic, selected from the same source population, with minimal variability across tumor characteristics. Both groups showed nonsignificant differences in wellestablished risk factors, including Breslow depth, mitotic HEAD & NECK DOI /HED APRIL 2016 E759

7 KAVEH ET AL. index, PNI, and LVI. The Sunbelt Melanoma Trial 17 did identify ulceration as a risk for false-negative SLNB, but our findings remained robust to sensitivity analyses that excluded patients with ulceration, as well as those with PNI, LVI, incomplete histologic data, and neck primary sites. Although the rates of regional nodal recurrence between the discordant and concordant groups did not reach statistical significance, this was likely a limitation of our sample size and still suggests a more aggressive tumor biology driving greater overall recurrence. Larger studies are warranted to better investigate this relationship. In summary, lymphatic drainage on preoperative lymphoscintigraphy that differs from predicted drainage patterns is associated with increased incidence of recurrence after negative SLNBs in cutaneous melanoma arising in the head and neck regions in middle-aged and older men. This population may benefit from more vigilant followup, including intensified postoperative surveillance with imaging modalities, such as ultrasonography. Acknowledgment The authors thank Terri Owen, Stanford Cancer Registrar, for her assistance in collecting the data used for this study. REFERENCES 1. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM. Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States. Cancer Res 2014;74: Lachiewicz AM, Berwick M, Wiggins CL, Thomas NE. Survival differences between patients with scalp or neck melanoma and those with melanoma of other sites in the Surveillance, Epidemiology, and End Results (SEER) program. Arch Dermatol 2008;144: Terakedis BE, Anker CJ, Leachman SA, et al. Patterns of failure and predictors of outcome in cutaneous malignant melanoma of the scalp. JAm Acad Dermatol 2014;70: Balch CM, Morton DL, Gershenwald JE, et al. Sentinel node biopsy and standard of care for melanoma. J Am Acad Dermatol 2009;60: Schmalbach CE, Bradford CR. Is sentinel biopsy the standard of care for cutaneous head and neck melanoma? Laryngoscope 2015;125: Morton DL, Thompson JF, Cochran AJ, et al. Final trial report of sentinelnode biopsy versus nodal observation in melanoma. N Engl J Med 2014; 370: Schmalbach CE, Lowe L, Teknos TN, Johnson TM, Bradford CR. Reliability of sentinel biopsy for regional staging of head and neck Merkel cell carcinoma. Arch Otolaryngol Head Neck Surg 2005;131: Manca G, Rubello D, Romanini A, Mariani G. False-negative sentinel biopsy in melanoma patients. Nucl Med Commun 2014;35: de Rosa N, Lyman GH, Silbermins D, et al. Sentinel node biopsy for head and neck melanoma: a systematic review. Otolaryngol Head Neck Surg 2011;145: Chao C, Wong SL, Edwards MJ, et al. Sentinel biopsy for head and neck melanomas. Ann Surg Oncol 2003;10: Caraco C, Marone U, Celentano E, Botti G, Mozzillo N. Impact of falsenegative sentinel biopsy on survival in patients with cutaneous melanoma. Ann Surg Oncol 2007;14: Leong SP, Achtem TA, Habib FA, et al. Discordancy between clinical predictions vs lymphoscintigraphic and intraoperative mapping of sentinel lymph node drainage of primary melanoma. Arch Dermatol 1999;135: Lin D, Franc BL, Kashani Sabet M, Singer MI. Lymphatic drainage patterns of head and neck cutaneous melanoma observed on lymphoscintigraphy and sentinel biopsy. Head Neck 2006;28: de Wilt JH, Thompson JF, Uren RF, et al. Correlation between preoperative lymphoscintigraphy and metastatic nodal disease sites in 362 patients with cutaneous melanomas of the head and neck. Ann Surg 2004;239: Davis Malesevich MV, Goepfert R, Kubik M, Roberts DB, Myers JN, Kupferman ME. Recurrence of cutaneous melanoma of the head and neck after negative sentinel biopsy. Head Neck [Epub ahead of print]. 16. Bleicher RJ, Essner R, Foshag LJ, Wanek LA, Morton DL. Role of sentinel lymphadenectomy in thin invasive cutaneous melanomas. J Clin Oncol 2003;21: Scoggins CR, Martin RC, Ross MI, et al. Factors associated with falsenegative sentinel biopsy in melanoma patients. Ann Surg Oncol 2010;17: Conway WC, Faries MB, Nicholl MB, et al. Age-related lymphatic dysfunction in melanoma patients. Ann Surg Oncol 2009;16: Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report on cutaneous and noncutaneous melanoma: a summary of 84,836 cases from the past decade. The American College of Surgeons Commission on Cancer and the American Cancer Society. Cancer 1998;83: Gamba CS, Clarke CA, Keegan TH, Tao L, Swetter SM. Melanoma survival disadvantage in young, non-hispanic white males compared with females. JAMA Dermatol 2013;149: Austin MT, Xing Y, Hayes Jordan AA, Lally KP, Cormier JN. Melanoma incidence rises for children and adolescents: an epidemiologic review of pediatric melanoma in the United States. J Pediatr Surg 2013;48: Joosse A, Collette S, Suciu S, et al. Superior outcome of women with stage I/II cutaneous melanoma: pooled analysis of four European Organisation for Research and Treatment of Cancer phase III trials. J Clin Oncol 2012; 30: 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: Robbins KT, Medina JE, Wolfe GT, Levine PA, Sessions RB, Pruet CW. Standardizing neck dissection terminology. Official report of the Academy s Committee for Head and Neck Surgery and Oncology. Arch Otolaryngol Head Neck Surg 1991;117: National Comprehensive Cancer Network. Melanoma (version ). Available at: Accessed August 29, Jones EL, Jones TS, Pearlman NW, et al. Long-term follow-up and survival of patients following a recurrence of melanoma after a negative sentinel biopsy result. JAMA Surg 2013;148: McHugh JB, Su L, Griffith KA, et al. Significance of multiple lymphatic basin drainage in truncal melanoma patients undergoing sentinel lymph node biopsy. Ann Surg Oncol 2006;13: Morton DL, Cochran AJ, Thompson JF, et al. Sentinel node biopsy for early-stage melanoma: accuracy and morbidity in MSLT-I, an international multicenter trial. Ann Surg 2005;242: ; discussion Gershenwald JE, Colome MI, Lee JE, et al. Patterns of recurrence following a negative sentinel biopsy in 243 patients with stage I or II melanoma. J Clin Oncol 1998;16: Pathak I, O Brien CJ, Petersen Schaeffer K, et al. Do nodal metastases from cutaneous melanoma of the head and neck follow a clinically predictable pattern? Head Neck 2001;23: Fincher TR, O Brien JC, McCarty TM, et al. Patterns of drainage and recurrence following sentinel biopsy for cutaneous melanoma of the head and neck. Arch Otolaryngol Head Neck Surg 2004;130: Klop WM, Veenstra HJ, Vermeeren L, Nieweg OE, Balm AJ, Lohuis PJ. Assessment of lymphatic drainage patterns and implications for the extent of neck dissection in head and neck melanoma patients. J Surg Oncol 2011; 103: Kelemen PR, Essner R, Foshag LJ, Morton DL. Lymphatic mapping and sentinel lymphadenectomy after wide local excision of primary melanoma. J Am Coll Surg 1999;189: Thompson JF, Uren RF, Shaw HM, et al. Location of sentinel s in patients with cutaneous melanoma: new insights into lymphatic anatomy. J Am Coll Surg 1999;189: Wagner JD, Park HM, Coleman JJ III, Love C, Hayes JT. Cervical sentinel biopsy for melanomas of the head and neck and upper thorax. Arch Otolaryngol Head Neck Surg 2000;126: Fadaki N, Li R, Parrett B, et al. Is head and neck melanoma different from trunk and extremity melanomas with respect to sentinel status and clinical outcome? Ann Surg Oncol 2013;20: Ahmadi O, McCall JL, Stringer MD. Does senescence affect number and morphology? A systematic review. ANZ J Surg 2013;83: Murakami G, Taniguchi I. Histologic heterogeneity and intranodal shunt flow in s from elderly subjects: a cadaveric study. Ann Surg Oncol 2004;11(3 Suppl):279S 284S. 39. McMasters KM, Reintgen DS, Ross MI, et al. Sentinel biopsy for melanoma: how many radioactive nodes should be removed? Ann Surg Oncol 2001;8: Jimenez RE, Panageas K, Busam KJ, Brady MS. Prognostic implications of multiple lymphatic basin drainage in patients with truncal melanoma. J Clin Oncol 2005;23: Shayan R, Achen MG, Stacker SA. Lymphatic vessels in cancer metastasis: bridging the gaps. Carcinogenesis 2006;27: Federico AC, Chagpar AB, Ross MI, et al. Effect of multiple-nodal basin drainage on cutaneous melanoma. Arch Surg 2008;143: ; discussion Parrett BM, Kashani Sabet M, Singer MI, et al. Long-term prognosis and significance of the sentinel in head and neck melanoma. Otolaryngol Head Neck Surg 2012;147: E760 HEAD & NECK DOI /HED APRIL 2016