Recent analysis of mortality from melanoma has shown that thin melanomas account for 20% and 25% of melanoma deaths in Australia and the United States, respectively, despite an overall survival rate of approximately 95% for patients with these cancers.1,2 Sentinel lymph node (SLN) biopsy is routinely recommended in patients with intermediate-thickness melanomas, but its role in thin melanomas (≤1 mm) is less well defined3,4; moreover, factors predictive of SLN positivity in this latter group have been variably reported. To specifically examine the prognostic significance of histologic subtype of thin melanomas for SLN metastasis, we reanalyzed these lesions in a cohort of patients originally created for a previous study of thin melanomas.5
Methods
The cohort for the previous study and the present reanalysis included 781 patients with thin primary cutaneous melanomas who underwent SLN biopsy from February 22, 1995, to June 27, 2011, at our institution. The SLN biopsy was performed for patients with thin melanomas on the basis of an individual patient’s melanoma risk factors and comorbidities, discussion of the risks and benefits of the biopsy procedure, and patient preferences. Predictors of SLN positivity in the previous study were found through multivariable logistic regression to be mitotic rate (odds ratio [OR], 7.01) and Clark level IV or V (OR, 3.45).5
Patient variables analyzed in the previous study were age and sex. Primary tumor characteristics included anatomic site, tumor thickness, Clark level, mitoses, tumor-infiltrating lymphocytes, regression, ulceration, lymphovascular invasion evident in hematoxylin-eosin–stained sections, and microsatellitosis. The following binary variables were used in the analyses: Clark level (II-III and unknown or IV-V), thickness (≤0.75 mm or 0.76–1 mm), tumor-infiltrating lymphocytes (present or absent), and mitoses (present or absent). For lesions in which an individual characteristic was not reported, the characteristic was recorded as unknown. Only patients with known mitotic rate data were included in the regression analyses (n = 698).5 Histologic subtype was not included in the previous analysis, thus providing the basis for this reanalysis.
This study was approved, with a waiver for patient informed consent, by the Institutional Review Board of the University of Pennsylvania.
Results
In our new prognostic model inclusive of histologic subtype, univariable analysis identified nodular melanoma (OR, 3.80) and acral lentiginous melanoma (ALM) (OR, 8.17) (P = .01 for both) as factors significantly associated with SLN positivity. By multivariable logistic regression analysis, ALM remained a factor significantly associated with SLN positivity (OR, 16.02; P = .004), as did increased Clark level (OR, 3.04; P = .02) and mitotic rate of 1mm2 or more (OR, 6.04; P = .01). Sentinel lymph node positivity was found in 2 of 10patients (20%)with ALM, 5 of 48 patients (10%)with nodular melanoma, 14 of 534 patients (3%) with superficial spreading melanoma, and none of 37 patients with lentigo maligna melanoma. Of the patients in the group with ALM, 9 of 10(90%)were non-Hispanic white and 1 of 10 (10%) was Hispanic white. Nine of the 10 ALM lesions were found on the foot and 1 ALM lesion was found on the hand. Thickness (OR, 2.22; P = .09) and nodular melanoma (OR, 2.48; P = .09) showed a trend toward, but did not reach, statistical significance in our model (Table).
Table.
Factors Associated With SLN Positivity: Logistic Regression Analysisa
| Characteristic | No. (%)
|
Univariable
|
Reduced Multivariable
|
|||
|---|---|---|---|---|---|---|
| Frequency | SLN Positivity | OR (95% CI) | P Value | OR (95% CI) | P Value | |
| Sex | ||||||
|
| ||||||
| Male | 389 (55.7) | 16 (4.1) | 1.28 (0.57–2.87) | .54 | ||
|
| ||||||
| Female | 309 (44.3) | 10 (3.2) | 1 [Reference] | |||
|
| ||||||
| Anatomic site | ||||||
|
| ||||||
| Axial | 403 (57.7) | 13 (3.2) | 0.73 (0.33–1.58) | .42 | ||
|
| ||||||
| Extremity | 295 (42.3) | 13 (4.4) | 1 [Reference] | |||
|
| ||||||
| Clark level | ||||||
|
| ||||||
| II–III and unknown | 374 (53.6) | 6 (1.6) | 1 [Reference] | 1 [Reference] | ||
|
| ||||||
| IV–V | 324 (46.4) | 20 (6.2) | 4.04 (1.60–10.17) | .003 | 3.04 (1.17–7.84) | .02 |
|
| ||||||
| Thickness, mm | ||||||
|
| ||||||
| 0.01–0.75 | 375 (53.7) | 7 (1.9) | 1 [Reference] | 1 [Reference] | ||
|
| ||||||
| 0.76–1.00 | 323 (46.3) | 19 (5.9) | 3.33 (1.40–8.16) | .01 | 2.22 (0.87–5.65) | .09 |
|
| ||||||
| Mitoses | ||||||
|
| ||||||
| Absent | 273 (39.1) | 2 (0.7) | 1 [Reference] | 1 [Reference] | ||
|
| ||||||
| Present | 425 (60.9) | 24 (5.6) | 8.11 (1.90–34.60) | .005 | 6.04 (1.34–27.32) | .01 |
|
| ||||||
| TIL | ||||||
|
| ||||||
| Absent | 191 (27.4) | 7 (3.7) | 0.98 (0.40–2.36) | .59 | ||
|
| ||||||
| Present or unknown | 507 (72.6) | 19 (3.7) | 1 [Reference] | |||
|
| ||||||
| Regression | ||||||
|
| ||||||
| Present | 146 (20.9) | 4 (2.7) | 0.68 (0.23–2.00) | .48 | ||
|
| ||||||
| Absent or unknown | 552 (79.1) | 22 (4.0) | 1 [Reference] | |||
|
| ||||||
| Ulcerationb | ||||||
|
| ||||||
| Present | 27 (3.9) | 0 | ||||
|
| ||||||
| Absent or unknown | 671 (96.1) | 26 (3.9) | ||||
|
| ||||||
| NM, ALM, neither NM nor ALMc | ||||||
|
| ||||||
| ALM | 10 (1.4) | 2 (20.0) | 8.17 (1.62–41.09) | .01 | 16.02 (2.46–104.15) | .004 |
|
| ||||||
| NM | 48 (6.9) | 5 (10.4) | 3.80 (1.35–10.67) | .01 | 2.48 (0.85–7.20) | .09 |
|
| ||||||
| Neither | 640 (91.7) | 19 (3.0) | 1 [Reference] | 1 [Reference] | ||
Abbreviations: ALM, acral lentiginous melanoma; NM, nodular melanoma; OR, odds ratio; SLN, sentinel lymph node; TIL, tumor-infiltrating lymphocytes.
N = 698. Analysis included only patients for whom the mitotic rate was known.
Ulceration was not present in the tumor of any patient with a positive finding on SLN biopsy and therefore could not be included.
Histologic status was coded as ALM/NM/neither because ALM and NM were significant predictors of SLN positivity in univariable analysis.
Discussion
The reanalysis found that the histologic subtype of ALM as well as mitoses and Clark level IV-V were independent predictors of SLN positivity. This finding is limited, however, by the relatively small number of patients (n = 10) with ALM in the cohort on which the reanalysis was done. The finding of ALM as a predictor of SLN metastasis should be further confirmed in other studies. Although the rate of SLN positivity was low for the study cohort of patients with thin melanomas (3.7%), a subset of patients can be identified with appreciable rates of nodal metastasis, who can be stratified for risk by the Clark level, mitotic rate, and histologic subtype of their melanoma. Further study of these factors can help guide clinical decision-making in patients with thin melanomas.
Acknowledgments
Funding/Support: Dr Chu is supported in part by the Dermatology Foundation Dermatopathology Research Career Development Award. Mr Marek is supported in part by the National Center for Advancing Translational Sciences of the National Institutes of Health under award TL1TR000138. Drs Bartlett and Karakousis are supported in part by grants P50-CA093372, P30-CA016520, and P50-174523 from the National Cancer Institute.
Footnotes
Conflict of Interest Disclosures: None reported.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Author Contributions: Dr Chu and Mr Marek had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Bartlett, Karakousis, Chu.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Marek.
Critical revision of the manuscript for important intellectual content: Ming, Bartlett, Karakousis, Chu.
Statistical analysis: Marek, Bartlett.
Obtained funding: Marek, Bartlett, Chu.
Administrative, technical, or material support: Marek, Ming, Bartlett, Chu.
Study supervision: Ming, Karakousis, Chu.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
References
- 1.Criscione VD, Weinstock MA. Melanoma thickness trends in the United States, 1988–2006. J Invest Dermatol. 2010;130(3):793–797. doi: 10.1038/jid.2009.328. [DOI] [PubMed] [Google Scholar]
- 2.Whiteman DC, Baade PD, Olsen CM. More people die from thin melanomas (≤1 mm) than from thick melanomas (>4 mm) in Queensland, Australia. J Invest Dermatol. 2015;135(4):1190–1193. doi: 10.1038/jid.2014.452. [DOI] [PubMed] [Google Scholar]
- 3.Karakousis GC, Gimotty PA, Czerniecki BJ, et al. Regional nodal metastatic disease is the strongest predictor of survival in patients with thin vertical growth phase melanomas: a case for SLN staging biopsy in these patients. Ann Surg Oncol. 2007;14(5):1596–1603. doi: 10.1245/s10434-006-9319-y. [DOI] [PubMed] [Google Scholar]
- 4.Wright BE, Scheri RP, Ye X, et al. Importance of sentinel lymph node biopsy in patients with thin melanoma. Arch Surg. 2008;143(9):892–899. doi: 10.1001/archsurg.143.9.892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bartlett EK, Gimotty PA, Sinnamon AJ, et al. Clark level risk stratifies patients with mitogenic thin melanomas for sentinel lymph node biopsy. Ann Surg Oncol. 2014;21(2):643–649. doi: 10.1245/s10434-013-3313-y. [DOI] [PMC free article] [PubMed] [Google Scholar]