Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

Sentinel lymph node biopsy is prognostic but not therapeutic for thick melanoma

Surgery, Available online 19 June 2015

Background

Sentinel lymph node biopsy (SNB) as a staging and therapeutic procedure in melanomas 1–4 mm in thickness has been investigated extensively, however, the clinical value of SNB in thick melanomas is poorly understood.

Methods

Patients undergoing operation for clinically node-negative melanoma >4 mm in depth between 2003 and 2010 were identified in the Surveillance Epidemiology and End Results registry. Two groups were constructed: one with a wide excision with SNB and the other with wide excision alone.

Results

A total of 4,571 patients with clinically node-negative, thick melanoma were identified. The median age was 71 years, 96.9% were white, and 64.3% were male. SNB was performed in 2,746 (60.1%) and was positive in 32.2%. Univariate analysis demonstrated SNB was associated with younger age (64 vs 75 years; P < .001) and extremity primaries (P < .0001). On logistic regression, advanced age (P < .001), female sex (P = .009), and location in the head and neck region (P < .001) were associated with observation. On log-rank analysis, improved 5-year disease-specific survival (DSS) was associated with SNB (65 vs 62%; P = .008), location in the extremity versus head and neck or trunk (67 vs 61.5 and 60.3%; P = .004), female sex (69 vs 61%; P < .001), and no ulceration (74 vs 54%; P < .001). On Cox regression analysis, advanced age (P < .001), male sex (P = .01), trunk location (P = .0001), and ulceration (P < .001) continued to be associated with DSS. SNB was not associated with survival (P = .20). SNB status was a robust predictor of survival; a negative SNB had a 5-year DSS of 75.3 versus 44.1% (P < .0001), with a positive node.

Conclusion

For patients with clinically node-negative, thick melanoma, SNB is a staging but not therapeutic procedure.

The utility of sentinel lymph node biopsy (SNB) in the management of cutaneous neoplasia has been investigated thoroughly. The concept of SNB for cutaneous neoplasia contends that each neoplasm has a sentinel lymph node (or nodes) that can be identified and that the sentinel lymph node will predict the status of the nodal basin. This concept was first used in melanoma, but applied subsequently to other cutaneous neoplasms, such as Merkel cell carcinoma, high-risk squamous cell cancer, apocrine adenocarcinoma, eccrine porocarcinoma, and malignant blue nevus.1, 2, 3, 4, 5, 6, 7, and 8 SNB has demonstrated consistently to have prognostic value, but the therapeutic value remains unclear.

The clinical utility of SNB has been defined most clearly for intermediate thickness (1–4 mm) melanoma. 9 Little controversy exists regarding the prognostic value of SNB in intermediate thickness melanoma. Patients with a negative SNB have a 5-year survival of 84–96.8% compared with 40–72.3% with a positive SNB.10, 11, 12, and 13 In the only prospective, randomized clinical trial examining SNB versus observation (no SNB), the MSLT-1 trial, Morton et al 14 found that the status of the SNB in patients with melanoma between 1.2–3.5 mm was the strongest predictor of survival.

In contrast, the therapeutic value of SNB for melanoma is less clear. In the MSLT-1 trial, the investigators found a significant improvement in disease-free survival but no benefit in overall or melanoma-specific survival at 5 and 10 years of follow-up (the primary study outcome).14 and 15 In a planned subset analysis, investigators compared patients in the SNB group with either a positive SNB or delayed recurrence with patients with a delayed recurrence in the observation arm (about 20% in both groups at 10 years). 14 The 10-year melanoma-specific survival was 62.1% for nodal disease in the SNB group and 41.5% for those with delayed recurrence in the observation group. 15 Morton et al 14 contend that the lack of significance in the overall trial was secondary to a dilution effect of the preponderance of node-negative patients and that the study was unpowered to detect an overall survival advantage of 3.2%. Consistent with the finding of the MSLT-1 trial, our group found that patients with intermediate thickness melanoma undergoing SNB had a modest but significant survival advantage when compared with observation. 13

As a result of conflicting data on the prognostic and therapeutic values of SNB for thick melanoma (>4 mm), the clinical utility in this group is less clear. Several studies and a Bayesian analysis reported no prognostic value to SNB for thick melanoma.16, 17, and 18 In contrast, multiple other studies have demonstrated a predictive value for SNB in thick melanoma.19, 20, 21, 22, 23, and 24 The therapeutic value of SNB for thick melanoma is also unclear. To better understand the clinical utility of SNB in thick melanoma, we reviewed data from the Surveillance Epidemiology and End Results (SEER) tumor registry, comparing patients who underwent SNB versus nodal observation.

Materials and methods

Data source

The SEER tumor registry was used to gather patient data. The National Cancer Institute SEER program collects data on patient demographics, primary tumor site, tumor morphology, stage at diagnosis, and first course of treatment and provides information on cancer incidence and survival for patients in the United States. Currently, the SEER registry represents 28% of the United States’ population. 25

Patient selection

A query of the SEER registry using SEERStat 7.0 was performed for the diagnosis of melanoma from 2003 to 2010. Patient demographics, tumor characteristics, and treatment methods were identified. Patient demographics included age at diagnosis, race, patient sex, survival time in months, and vital status. Tumor characteristics were coded within either the SEER collection systems: historic stage A (1998–2008), collaborative stage (2004–2008), or extent of disease (1998–2003). Tumor characteristics included the extent of disease (local, regional, or distant), size of the primary tumor, ulceration, primary site, and nodal status. Treatment data included operations at the site of the primary and the type of nodal intervention (SNB, lymphadenectomy, or observation). The AJCC 7th Edition, Melanoma of the Skin staging criteria was used to derive a final pathologic TNM stage. 26

We excluded patients with clinically positive lymph nodes, patients <18 years of age, unknown stage, metastatic disease at presentation, unknown primary site, and tumors <4 mm depth of invasion. To access the effect of SNB in thick melanoma, we recreated 2 experiment arms similar to the MSLT-1 trial, but excluded those patients with stages T1 (≤1.0 mm), T2 (1.01–2.00 mm), and T3 (2.01–4.00 mm) disease (n = 61,781); 4,571 patients met the inclusion criteria. Patients were divided into the following 2 groups: wide local excision with SNB and wide local excision with nodal observation (meaning no SNB).

Statistical analysis

Univariate and multivariate descriptive and survival analyses were performed. Descriptive data are reported as mean values and ranges to summarize continuous variables or frequencies and proportions for categorical data. Univariate descriptive analyses were performed with Student's t test and Chi-square test as indicated. Logistic regression was performed for descriptive multivariate analysis. Kaplan–Meier curves and log-rank tests were used for univariate survival analysis, and a Cox regression was performed for multivariate survival analysis. Factors with a P value of ≤.20 were included in multivariate models. In both the univariate survival analysis and the Cox regression, disease (melanoma)-specific survival (DSS) was used to calculate survival. All data analysis and manipulation were performed using JMP Pro version 10.0.0; 2012 (SAS Institute Inc, Cary, NC).

Results

Demographics

Overall, 131,467 patients with melanoma were identified in the SEER tumor registry from 2003 to 2010; 4,571 with thick melanoma met the inclusion criteria. The median age was 71 years. The majority of patients were male (64.3%), white (96.9%), had ulcerated tumors (54.9%), that were located most commonly on the extremity (42.5%; Table I ).

Table I Patient demographics and characteristics, Surveillance Epidemiology and End Results 2003–2010

Variables Total population (n = 4,571), n (%) SNB (n = 2,746), n (%) Observation (n = 1,825), n (%) P value
Age (y)
 Median 71 65 79  
 Mean 68.6 64.1 75.4 <.0001
 Range 18–106 18–99 18–106  
Sex       .09
 Male 2,937 (64.3) 1,791 (65.2) 1,146 (62.8)  
 Female 1,634 (35.7) 955 (34.8) 679 (37.2)  
Race       .23
 White 4,430 (96.9) 2,657 (96.8) 1,773 (97.2)  
 African American 64 (1.4) 36 (1.3) 28 (1.5)  
 Other 77 (1.7) 53 (1.9) 24 (1.3)  
Ulceration       .96
 Present 2,510 (54.9) 1,507 (54.9) 1,003 (55.0)  
 Absent 2,061 (45.1) 1,239 (45.1) 822 (45.0)  
Primary site       <.0001
 Head and neck 1,325 (29.0) 607 (22.1) 718 (39.3)  
 Extremity 1,943 (42.5) 1,252 (45.6) 691 (37.9)  
 Trunk 1,303 (28.5) 887 (32.3) 416 (22.8)  

The 4,571 surgical patients with clinically node-negative thick melanoma were divided into 2 study cohorts: wide local excision and SNB (n = 2,746 [60.1%]) and wide local excision alone (observation; n = 1,825 [39.9%]). Patients undergoing SNB were found to have nodal metastasis in 32.2% of cases. On univariate analysis, there were significant differences between the observation and sentinel lymph node groups. The SNB patients were younger (65 vs 79 years old; P < .0001) with tumors located less commonly in the head and neck (P < .0001). The groups were similar in sex (P = .09), race (P = .23), and presence of ulceration (P = .96; Table I ). Older age (P < .0001), females (P = .009), and head and neck locations (P = .004) were independently associated with undergoing nodal observation on bivariate logistic regression ( Table II ).

Table II Logistic regression: Factors associated with undergoing nodal observation, Surveillance Epidemiology and End Results 2003–2010

Variables Odds ratio 95% CI P value
Age (y)     <.0001
 Per unit 1.05 1.05–1.06  
 Entire range 85.6 56.7–131.0  
Sex
 Male 1    
 Female 1.20 1.05–1.38 .009
Race
 White n/a    
 African American n/a    
 Other n/a    
Ulceration
 Present n/a    
 Absent n/a    
Primary site
 Extremity 1    
 Trunk 1.13 0.96–1.33 .14
 Head and neck 2.09 1.79–2.44 <.0001

Survival analysis

The study population had a median follow-up of 23 months (range, 0–95) and an overall median survival of 50 months. The 5-year DSS for the entire population was 64%. The median follow-up for the SNB group was 26 months (range, 0–95) and the observation 19 months (range, 0–95). The median overall survival was greater in the SNB group (66 vs 33 months; P < .001). The 5-year DSS was also greater (albeit small) for the SNB group compared with the observation group (65% vs 62%; P = .008). Improved 5-year survival was associated with female sex (69.1 vs 60.8%; P < .0001), nonulcerative lesions (74.3 vs 54.4%; P < .0001), extremity location (P = .0004), and again albeit small for performance of SNB (65 vs 61.7%; P = .008; Table III ). On the Cox proportional hazards model, SNB was no longer associated with improved survival ( Table IV ). Survival continued to be associated with age, sex, and presence of ulceration. In this model, those patients with trunk melanoma had an increased mortality as compared with extremity location (hazard ratio, 1.34; P = .001). There was also a trend toward diminished survival for head and neck melanomas (P = .06; Table IV ).

Table III Univariate Analysis: 5-year disease-specific survival, Surveillance Epidemiology and End Results 2003–2010

Variables 5-year survival (%) P value
Sex   <.0001
 Female 69.1  
 Male 60.8  
Race   .37
 Other 56.9  
 White 64  
 African American 60.4  
Ulceration   <.0001
 Absent 74.3  
 Present 54.4  
Primary site   .0004
 Head and neck 61.5  
 Extremity 67.4  
 Trunk 60.3  
Lymph node management   .008
 SNB 65  
 Observation 61.7  
Sentinel lymph node status   <.0001
 Positive 44.1  
 Negative 75  

SNB, Sentinel lymph node biopsy.

Table IV Multivariate analysis: 5-year disease-specific survival, Surveillance Epidemiology and End Results 2003–2010

Variables Hazard ratio 95% CI P value
Age
 Per year 1.01 1.00–1.01 <.0001
 Entire range 2.25 1.53–3.31  
Sex
 Female 1    
 Male 1.2 1.04–1.38 .01
Ulceration
 Absent 1    
 Present 2.14 1.87–2.46 <.0001
Primary site
 Extremity 1    
 Trunk 1.34 1.16–1.56 .0001
 Head and neck 1.17 0.99–1.37 .06
Lymph node management
 SNB 1    
 Observation 1.09 0.95–1.25 .20

SNB, Sentinel lymph node biopsy.

It is important to note the proportional hazards assumption was tested in our final Cox regression models using the test statement in SAS Phreg Procedure, with several variables failing the time-dependent analyses as detailed by Kleinbaum and Klein. 27 Examples include age, t-stage, and sex; however, as suggested, 27 multiple methods should be used to access deviation from the proportional hazards assumption. In large samples, it is common for such assumptions to be violated owing to very low standard error estimates. Visual inspection of the Kaplan–Meier curves indicates that the potential deviations were minor and that Cox model estimates were likely conservative; a model that adjusts for such time dependencies would only illustrate greater differences in already identified predictors.

Prognostic value of SNB

A total of 883 patients (32.2%) undergoing SNB had a positive node. SNB was a robust predictor of survival. Median survival of patients who were sentinel node negative was 83 months and when metastatic disease was present in the node, survival was 36 months (P < .0001). Patients with a negative SNB had a 5-year DSS of 75%, whereas those with a positive SNB had a DSS of 44.1% (P < .0001; Fig ).

gr1

Figure Disease-specific survival for positive versus negative sentinel lymph node biopsy (SNB), Surveillance Epidemiology and End Results 2003–2010.

Discussion

The use of lymph node mapping and SNB has been investigated comprehensively in cutaneous neoplasia and is now the standard of care for intermediate thickness melanoma. 28 There is, however, controversy regarding clinical applicability in thick melanoma. Conflicting studies regarding prognostic value and limited data on the therapeutic value of SNB make this an important area of investigation. In this study, the largest study of SNB in clinically node-negative thick melanoma, we confirmed the prognostic value of SNB. We did not, however, document a survival benefit. This observation does not negate the clinical utility of this procedure, because SNB provides valuable prognostic information and may improve local regional control.

The therapeutic value of SNB for malignant melanoma is unclear. For patients with intermediate thickness melanoma, the MSLT-1 trial and other literature suggest strongly a small but clinically meaningful survival advantage.13, 14, and 15 Although at 10 years the DSS from the MSLT-1 failed demonstrate a survival advantage for SNB, the investigators postulated that the failure to document a survival advantage was secondary to a dilution effect of node-negative patients, and as such, the trial was underpowered to potentially identify an improvement in survival for the 20% of patients with nodal disease. 14 Retrospective large database studies with sufficient statistical power have been able to demonstrate a modest survival advantage to SNB for intermediate thickness melanoma.13 and 16

Few data are available about the therapeutic value of SNB in thick melanoma. Most of the studies reporting survival have been small, single institutional, and retrospective and failed to demonstrate a survival advantage.17 and 19 In a retrospective review of 5,840 patients with primary cutaneous melanoma, Van der Ploeg et al 16 found no improved DSS on multivariate for patients with melanoma >1 mm depth invasion. In the final report of the MSLT-1 trial, Morton et al 15 reports survival for 173 patients with thick melanoma randomized to SNB versus observation. They noted no difference in DSS (P = .78). 15 In this study, we reconstructed the study groups from the MSLT-1 trial using a greater number of patients with clinically node-negative, thick melanoma. In the observation group, the 5-year DSS was 61.7% versus 65% in the SLN group (P = .008), but no survival advantage for SNB was noted on multivariate regression.

The prognostic value of SNB in thick melanoma is unclear. Several early studies of SNB for thick melanoma failed to validate prognostic significant of nodal status. 29 Essner et al 29 reported no difference in survival between sentinel lymph positive and negative patients with thick melanoma. In 2011, a Bayesian analysis reaffirmed the lack of prognostic value of SNB status in thick melanoma. 18 In contrast, multiple studies have demonstrated that the status of the sentinel lymph node is predictive of survival with node-negative patients having a 5-year survival of 56–71% versus 26–48% with node-positive disease.30, 31, 32, 33, and 34 In a recent metaanalysis by Rondelli et al, 35 and colleagues SNB status was predictive of survival with a 71% 5-year survival if node(s) was negative for disease and 39% if positive. 35 Our findings are similar with a 5-year DSS of 75.4% for negative sentinel node patients versus 44.1 % in node-positive patients ( Fig ). In addition, as new immunotherapy regimens for melanoma are studied in the adjuvant setting, the status of SNB may provide important information for patient selection. 36

Substantial clinical nihilism exists for patients with thick melanoma. The survival in this group has been reported to be <50% at 5 years. 20 The results reported here are better with a median 5-year survival of 64%. These results are similar to those reported by Morton et al 15 for the MSLT-1 trial where 290 patients with thick melanoma had a 5-year survival rate of 67%.

There are limitations inherent to an analysis of SEER registry data. These data are retrospective and rely on accurate reporting. One such example, as discussed by Gimotty et al, 37 includes an error in the coding algorithm of the SEER registry with regard to patients who had undergone SNB and subsequently completion lymphadenectomy. Procedures performed before 2003 were coded in a hierarchical manner, such that if a patient had undergone completion lymphadenectomy after an SNB, only the completion lymphadenectomy was coded. This issue has been remedied in patients coded in 2003 and beyond. To avoid errors secondary to coding issues, we limited our population to patients diagnosed after 2002. Data regarding disease recurrence is unavailable within the SEER registry; therefore, we were unable to assess the disease-free survival. Tumor-related prognostic factors such as vascular invasion, tumor vascularity, and mitotic rate 38 are also unavailable in the SEER registry. Another limitation may lie in the selection of Cox regression models to examine the data. The proportional hazards assumption is very common, and if there are interactions between time and covariates studied here, that it is possible that these estimates are conservative.

In conclusion, SNB in thick melanoma has value as a staging procedure, but is not associated with improved DSS. The prognosis for patients with thick melanoma is significantly better than traditionally thought, and SNB status further delineates tumor biology. One could expect that patients with a negative SNB will have 5-year survival of >70% and expect those with node-positive disease to have a 5-year survival of <40%. Neither our data nor any other series demonstrates a survival benefit to SNB in thick melanoma. Although the SEER tumor registry does not allow us to fully examine local control, other investigators report improved local control for patients with thick melanomas undergoing a SNB. Given these findings, it would be reasonable to consider SNB as an acceptable standard of care for patients with thick melanoma because SNB provides prognostic and potential therapeutic benefits.

References

  • 1 J.E. Allen, L.B. Stolle. Utility of sentinel node biopsy in patients with high-risk cutaneous squamous cell carcinoma. Eur J Surg Oncol. 2015;41:197-200
  • 2 S.D. Kachare, J.H. Wong, N.A. Vohra, E.E. Zervos, T.L. Fitzgerald. Sentinel lymph node biopsy is associated with improved survival in Merkel cell carcinoma. Ann Surg Oncol. 2014;21:1624-1630 Crossref
  • 3 K.L. Hollowell, S.C. Agle, E.E. Zervos, T.L. Fitzgerald. Cutaneous apocrine adenocarcinoma: defining epidemiology, outcomes, and optimal therapy for a rare neoplasm. J Surg Oncol. 2012;105:415-419 Crossref
  • 4 S.D. Kachare, S.C. Agle, Z.P. Englert, E.E. Zervos, N.A. Vohra, J.H. Wong, et al. Malignant blue nevus: clinicopathologically similar to melanoma. Am Surg. 2013;79:651-656
  • 5 J.T. Krediet, M. Beyer, K. Lenz, C. Ulrich, B. Lange-Asschenfeldt, E. Stockfleth, et al. Sentinel lymph node and risk factors for predicting metastasis in cutaneous squamous cell carcinoma. Br J Dermatol. 2015;172:1029-1036
  • 6 M. Barnes, A. Hestley, D.R. Murray, G.W. Carlson, D. Parker, K.A. Delman. The risk of lymph node involvement in malignant cutaneous adnexal tumors. Am Surg. 2014;80:270-274
  • 7 I. Stoffels, T. Poeppel, C. Boy, S. Mueller, F. Wichmann, J. Dissemond, et al. Radio-guided surgery: advantages of a new portable gamma-camera (Sentinella) for intraoperative real time imaging and detection of sentinel lymph nodes in cutaneous malignancies. J Eur Acad Dermatol Venereol. 2012;26:308-313 Crossref
  • 8 C.W. Brown Jr., L.C. Dy. Eccrine porocarcinoma. Dermatol Ther. 2008;21:433-438
  • 9 R.M. Kauffmann, S.L. Chen. Workup and staging of malignant melanoma. Surg Clin North Am. 2014;94:963-972
  • 10 D.L. Morton, J.F. Thompson, A.J. Cochran, N. Mozzillo, R. Elashoff, R. Essner, et al. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med. 2006;355:1307-1317 Crossref
  • 11 J.E. Gershenwald, W. Thompson, P.F. Mansfield, J.E. Lee, M.I. Colome, C.H. Tseng, et al. Multi-institutional melanoma lymphatic mapping experience: the prognostic value of sentinel lymph node status in 612 stage I or II melanoma patients. J Clin Oncol. 1999;17:976-983
  • 12 Z.I. Nowecki, P. Rutkowski, A. Nasierowska-Guttmejer, W. Ruka. Sentinel lymph node biopsy in melanoma patients with clinically negative regional lymph nodes–one institution's experience. Melanoma Res. 2003;13:35-43 Crossref
  • 13 S.D. Kachare, J. Brinkley, J.H. Wong, N.A. Vohra, E.E. Zervos, T.L. Fitzgerald. The influence of sentinel lymph node biopsy on survival for intermediate-thickness melanoma. Ann Surg Oncol. 2014;21:3377-3385
  • 14 D.L. Morton. Overview and update of the phase III Multicenter Selective Lymphadenectomy Trials (MSLT-I and MSLT-II) in melanoma. Clin Exp Metastasis. 2012;29:699-706 Crossref
  • 15 D.L. Morton, J.F. Thompson, A.J. Cochran, N. Mozzillo, O.E. Nieweg, D.F. Roses, et al. Final trial report of sentinel-node biopsy versus nodal observation in melanoma. N Engl J Med. 2014;370:599-609 Crossref
  • 16 A.P. van der Ploeg, L.E. Haydu, A.J. Spillane, M.J. Quinn, R.P. Saw, K.F. Shannon, et al. Outcome following sentinel node biopsy plus wide local excision versus wide local excision only for primary cutaneous melanoma: analysis of 5840 patients treated at a single institution. Ann Surg. 2014;260:149-157
  • 17 S. Ribero, S. Osella-Abate, M. Sanlorenzo, E. Balagna, R. Senetta, M.T. Fierro, et al. Sentinel lymph node biopsy in thick-melanoma patients (n=350): what is its prognostic role?. Ann Surg Oncol. 2015;22:1967-1973
  • 18 A.R. Rhodes. Prognostic usefulness of sentinel lymph node biopsy for patients who have clinically node negative, localized, primary invasive cutaneous melanoma: a Bayesian analysis using informative published reports. Arch Dermatol. 2011;147:408-415 Crossref
  • 19 N.G. Fairbairn, G. Orfaniotis, M. Butterworth. Sentinel lymph node biopsy in thick malignant melanoma: a 10-year single unit experience. J Plast Reconstr Aesthet Surg. 2012;65:1396-1402 Crossref
  • 20 I. White, J. Fortino, B. Curti, J. Vetto. Clinical impact of sentinel lymph node biopsy in patients with thick (>4 mm) melanomas. Am J Surg. 2014;207:702-707 Crossref
  • 21 T. Hinz, H. Ahmadzadehfar, A. Wierzbicki, T. Hoeller, J. Wenzel, H.J. Biersack, et al. Sentinel lymph node status as most important prognostic factor in patients with high-risk cutaneous melanomas (tumour thickness >4.00 mm): outcome analysis from a single institution. Eur J Nuclear Med Mol Imag. 2012;39:1316-1325 Crossref
  • 22 M.G. Rughani, M.C. Swan, T.S. Adams, A. Marshall, R. Asher, O.C. Cassell, et al. Sentinel node status predicts survival in thick melanomas: the Oxford perspective. Eur J Surg Oncol. 2012;38:936-942 Crossref
  • 23 N. Mozzillo, E. Pennacchioli, S. Gandini, C. Caraco, A. Crispo, G. Botti, et al. Sentinel node biopsy in thin and thick melanoma. Ann Surg Oncol. 2013;20:2780-2786 Crossref
  • 24 T. Gambichler, L. Scholl. Sentinel lymph node status is the strongest independent predictor for disease free and overall survival in patients with thick melanoma. J Eur Acad Dermatol Venereol. 2013;27:1454-1456 Crossref
  • 25 Surveillance, Epidemiology, and End Results Program. SEER research data (1973–2009). (National Cancer Institute, Bethesda, MD, 2011)
  • 26 S.B. Edge, D.R. Byrd, C.C. Compton, A.G. Fritz, F.L. Greene, A. Trotti. AJCC cancer staging manual. (Lippincott Raven, Philadelphia, 2010)
  • 27 Y. Kubo, L.R. Sterling, P.S. Parfrey, K. Gill, K.W. Mahaffey, I. Gioni, et al. Assessing the treatment effect in a randomized controlled trial with extensive non-adherence: the EVOLVE trial. Pharm Stat. 2015;14:242-251
  • 28 D.G. Coit, J.A. Thompson, R. Andtbacka, C.J. Anker, C.K. Bichakjian, W.E. Carson 3rd, et al. Melanoma, version 4.2014. J Natl Compr Cancer Netw. 2014;12:621-629
  • 29 R. Essner, M.H. Chung, R. Bleicher, E. Hsueh, L. Wanek, D.L. Morton. Prognostic implications of thick (> or = 4-mm) melanoma in the era of intraoperative lymphatic mapping and sentinel lymphadenectomy. Ann Surg Oncol. 2002;9:754-761
  • 30 J.E. Gershenwald, P.F. Mansfield, J.E. Lee, M.I. Ross. Role for lymphatic mapping and sentinel lymph node biopsy in patients with thick (> or = 4 mm) primary melanoma. Ann Surg Oncol. 2000;7:160-165 Crossref
  • 31 C.R. Ferrone, K.S. Panageas, K. Busam, M.S. Brady, D.G. Coit. Multivariate prognostic model for patients with thick cutaneous melanoma: importance of sentinel lymph node status. Ann Surg Oncol. 2002;9:637-645
  • 32 R. Gutzmer, I. Satzger, K.M. Thoms, B. Volker, C. Mitteldorf, A. Kapp, et al. Sentinel lymph node status is the most important prognostic factor for thick (> or = 4 mm) melanomas. J Dtsch Dermatol Ges. 2008;6:198-203 Crossref
  • 33 L. Vermeeren, F.W. van der Ent, P.S. Sastrowijoto, K.W. Hulsewe. Thick melanoma: prognostic value of positive sentinel nodes. World J Surg. 2009;33:2464-2468 Crossref
  • 34 C. Gajdos, K.A. Griffith, S.L. Wong, T.M. Johnson, A.E. Chang, V.M. Cimmino, et al. Is there a benefit to sentinel lymph node biopsy in patients with T4 melanoma?. Cancer. 2009;115:5752-5760 Crossref
  • 35 F. Rondelli, M.C. Vedovati, C. Becattini, G.M. Tomassini, S. Messina, G. Noya, et al. Prognostic role of sentinel node biopsy in patients with thick melanoma: a meta-analysis. J Eur Acad Dermatol Venereol. 2012;26:560-565 Crossref
  • 36 E. Minutilli, C. Feliciani. Adjuvant therapy for resected stage III melanoma patients: high-dose interferon-alpha versus ipilimumab combined with kinases inhibitors. Tumori. 2012;98:185-190
  • 37 P.A. Gimotty, F. Yoon, R. Hammond, P. Rosenbaum, D. Guerry. Therapeutic effect of sentinel lymph node biopsy in melanoma remains an open question. J Clin Oncol. 2009;27:4236-4238 Crossref
  • 38 J. Homsi, M. Kashani-Sabet, J.L. Messina, A. Daud. Cutaneous melanoma: prognostic factors. Cancer Control. 2005;12:223-229

Footnotes

a The Brody School of Medicine, East Carolina University, Greenville, NC

b Division of Surgical Oncology, Department of Surgery, East Carolina University, Greenville, NC

Reprint requests: Timothy L. Fitzgerald, MD, Brody School of Medicine, 4S24, 600 Moye Boulevard, Greenville, NC 27834.