Survival analysis in head and neck melanoma after negative sentinel lymph node biopsy: A seer-based population study

Abstract

Background

Cutaneous malignant melanoma (CMM) is one of the most aggressive skin tumors. Sentinel lymph node biopsy (SLNB) is an important test before thorough treatment of melanoma. The aim of this study was to investigate cancer-specific survival (CSS) in patients with head and neck CMM after negative SLNB and to analyze predictors of decreased survival.

Methods

Based on the Surveillance, Epidemiology and End Results (SEER) database, a study was conducted using data from patients with head and neck CMM after negative SLNB. The demographic, clinical, and pathological characteristics of the case population were analyzed. Cox univariate, Kaplan–Meier analysis, and multivariate Cox regression models were used to explore predictors of decreased survival; propensity score matching (PSM) analysis was used to reduce confounding bias, and outcomes were compared between the wide margin excision and narrow margin excision groups.

Results

A total of 1597 confirmed head and neck CMM patients with SLNB-negative were found. A Breslow>4.0 mm was the highest independent risk predictor for patients (HR 3.82, 95% CI 2.04-7.16, P < .001), and significant risk independent predictors also included a high mitotic rate >4 (HR 1.54, 95% CI 1.06-2.25, P = .023). Age< 60 years old was a significant survival predictor (HR 0.56, 95% CI .37-.85, P = .007), and not scalp and neck CMM were also important factors for longer survival (auricle skin: HR .51, 95% CI .29-.90, P = .02; unspecified parts of face: HR .59, 95% CI .40-.87, P = .007). After harmonizing baseline data by PSM, it was found that the extent of surgical resection did not affect patient survival.

Conclusion

This study analyzed the risk factors affecting CSS in patients with CMM of the head and neck region with SLNB-negative and observed a statistically significant difference in the prognosis of patients with CMM in different aesthetic subunits of the head and neck region. Close clinical follow-up for this population is necessary, and periodic medical examinations should be carried out.

Keywords

head and neck melanoma predictor cancer-specific survival prognosis

Introduction

Cutaneous malignant melanoma (CMM) is a malignant tumor derived from neural crest melanocytes that is highly malignant and prone to lymphatic metastasis.^1,2 For CMM patients with clinically negative lymph node metastasis, sentinel lymph node biopsy (SLNB) is the most important and accurate prognostic factor in the pathological evidence of lymph node metastasis.^3-5 Sentinel lymph node biopsy has high application value in the auxiliary diagnosis of malignant melanoma, which can early determine the lymphatic metastasis of tumor patients, accurately assess the stage of malignant melanoma, predict prognosis and formulate follow-up treatment plans.⁶ However, the lymphatic drainage of the head and neck is the most complex and highly variable, usually involving multiple lymph node drainage regions, and the location of the sentinel lymph node group is relatively close to the primary tumor, so the implementation of SLNB is more difficult, and there is also the possibility of false-negative SLNB.^7-9 A false-negative SLNB was defined as a negative SLNB that recurred in the sampled lymph node region after several years of follow-up. True-negative SLNB was defined as negative SLNB without regional recurrence in previously sampled lymph node basins. According to statistics, up to 21.0% of negative SLNB patients may develop metastasis in the draining lymph node area during the follow-up period.¹⁰ Studies have recommended lymph node dissection in patients with SLNB-negative at regional recurrence.¹¹

It is currently recognized that the factors affecting the survival of patients with CMM involve age, sex, tumor Breslow thickness, ulceration, and mitotic rate. Other studies have shown that the status of sentinel lymph nodes is also the strongest predictor of survival in patients with intermediate-thickness melanoma.^12,13 However, the multicenter selective lymphadenectomy trial-I (MSLT-I) demonstrated that SLNB-based management extended disease-free survival in all patients or in patients with medium-thick melanoma compared with clinical lymph node observation, and the management also prolonged melanoma-specific survival in patients with lymph node metastases.¹⁴

There are currently few studies on prognostic factors in patients with CMM after negative SLNB. Therefore, the purpose of this retrospective study was to analyze the predictors of tumor-specific survival in patients with clinically node-negative head and neck CMM after SLNB and to evaluate the surgical benefit of patients with different surgical resection extents after propensity score matching (PSM) analysis.

Materials and methods

Collection of baseline information

We performed a population-based cross-sectional analysis according to the National Cancer Institute Surveillance, Epidemiology, and End Results (SEER). We identified head and neck CMM by selecting the year of diagnosis from 2010 to 2015, the International Classification of Diseases in Oncology, Third Edition (ICD-0-3) histology code 8743. Follow-up information showed patients at diagnosis aged ≥18 years old, sex, race, and primary site (the primary sites included were skin of the lip (C44.0), eyelid (C44.1), external ear (C44.2), skin of other and unspecified parts of face (C44.3), and skin of scalp and neck (C44.4)). All cases had complete AJCC TNM staging and melanoma mitotic rate records, and all underwent auxiliary examinations, such as SLNB. Patients with unknown or unspecified TNM stage, survival time, survival status, and surgical status were excluded from this study. We also excluded N1-N3, NX in AJCC TNM staging and distant metastasis in TNM staging M1. The survival status and survival time of each patient were confirmed. And narrow margin excision (NME) standard codes were 30, 31, 32, 33 and wide margin excision (WME) codes were 45, 46,47.

Statistical analysis

Hazard ratios (HR) and 95% confidence intervals (CI) were estimated using Cox regression models. Cancer-specific mortality (CSS) was used as the event. Univariate at P < .05 was incorporated into multivariate analysis to predict independent prognostic factors, and survival curves were drawn using the Kaplan–Meier method. Then, all significant variables in the univariate analysis were extracted, and patients treated with WME and NME were randomly matched according to the propensity matching score (PSM) principle, and their survival outcomes were compared. R (version 3.6.1, https://www.r-project.org/) and IBM SPSS (version 22.0) were used for statistical analysis.

Results

Population statistics

A total of 1597 patients with CMM were collected from the Monitoring, Epidemiology, and Results Database of the National Cancer Institute, and they had negative SLNB cases for AJCC melanoma segmentation data. The median CSS of the confirmed population was 58 months. In terms of population statistics, age ≥60 years old and males accounted for the majority of all patients, 64.9% and 76.3%, respectively; 43.5% of cases originated from scalp and neck; in the clinical pathological characteristics of patient populations, approximately 57.8% of patients without CMM ulceration; approximately 12.5% of patients with Breslow> 4.0 mm, 26.5% Mitotic Rate> 4. The specific demographic, clinical, and pathological characteristics are shown in Table 1.

Table 1.

Baseline demographic and clinical characteristics.

	Alive (N = 1457)	Dead (N = 140)	Overall (N = 1597)
Age
<60	532 (36.5%)	29 (20.7%)	561 (35.1%)
≥60	925 (63.5%)	111 (79.3%)	1036 (64.9%)
Sex
Female	356 (24.4%)	22 (15.7%)	378 (23.7%)
Male	1101 (75.6%)	118 (84.3%)	1219 (76.3%)
Race
Black	2 (.1%)	0 (0%)	2 (.1%)
Other	14 (1.0%)	2 (1.4%)	16 (1.0%)
White	1441 (98.9%)	138 (98.6%)	1579 (98.9%)
Ulceration
No	853 (58.5%)	70 (50.0%)	923 (57.8%)
Yes	604 (41.5%)	70 (50.0%)	674 (42.2%)
Breslow
≤1.0 mm	520 (35.7%)	21 (15.0%)	541 (33.9%)
>1.0-2.0 mm	519 (35.6%)	35 (25.0%)	554 (34.7%)
>2.0-4.0mm	258 (17.7%)	45 (32.1%)	303 (19.0%)
>4.0 mm	160 (11.0%)	39 (27.9%)	199 (12.5%)
Mitotic rate
<1	258 (17.7%)	13 (9.3%)	271 (17.0%)
1∼4	844 (57.9%)	59 (42.1%)	903 (56.5%)
>4	355 (24.4%)	68 (48.6%)	423 (26.5%)
Summary stage
Localized	1323 (90.8%)	109 (77.9%)	1432 (89.7%)
Regional	134 (9.2%)	31 (22.1%)	165 (10.3%)
Primary site
Skin of lips	6 (.4%)	0 (0%)	6 (.4%)
Eyelid	14 (1.0%)	1 (.7%)	15 (.9%)
Auricle skin	244 (16.7%)	14 (10.0%)	258 (16.2%)
Unspecified parts of face	584 (40.1%)	39 (27.9%)	623 (39.0%)
Scalp and neck	609 (41.8%)	86 (61.4%)	695 (43.5%)
CSS (Month)
Mean (SD)	62.6 (23.1)	34.9 (17.9)	60.2 (24.0)
Median [Min, Max]	61.0 [1.00, 107]	33.0 [7.00, 78.0]	58.0 [1.00, 107]

Cox univariate and multivariate survival analysis

Cox univariate analysis found that age, sex, Breslow>2.0 mm, mitotic rate>4, summary stage, and scalp and neck lesions were associated with CSS in patients (Table 2). According to the CSS survival curve drawn by the Kaplan–Meier method (shown in Figure 1), the 5-year CSS of the entire study cohort was 90.9%; the 5-year CSS of patients aged ≥ 60 years old was 88.2%, which was significantly lower than that of patients aged < 60 years (95.4%). The 5-year CSS of male patients was 89.9%; compared with Breslow>4.0 mm and >2.0-4.0 mm head and neck SLNB-negative patients, the 5-year survival rate (75.6% and 83.5%), the CSS of the other two groups was significantly improved (5 annual survival rates were 94% and 96.5%); in those with mitotic rate >4, the 5-year survival rate was 82.0%; in patients with CSS in the scalp and neck, the 5-year survival rate was 87.5%; and the 5-year CSS (80.4%) for regional metastatic lesions had a worse prognosis than local lesions (92.1%).

Table 2.

Univariate analysis for patients with head and neck CMM after negative SLNB.

	Univariate analysis
	HR 95%CI	P-value
Age (<60 vs ≥60)	2.20 (1.46-3.42)	<.001
Sex (Female vs Male)	1.73 (1.11-2.85)	.027
Race		.79
Black	NA	—
Other	Reference	—
White	0.67 (0.18-4.30)	.59
Ulceration (Yes vs No)	1.41 (0.99-2.00)	.062
Breslow		<.001
≤1.0 mm	Reference
>1.0-2.0 mm	1.67 (0.97-2.95)	.07
>2.0-4.0 mm	4.32 (2.55-7.54)	<.001
>4.0 mm	6.04 (3.49-10.73)	<.001
Mitotic rate		<.001
<1	Reference
1∼4	1.39 (0.77-2.68)	.298
>4	3.80 (2.12-7.33)	<0.001
Summary stage (Localized vs Regional)	2.81 (1.79-4.30)	<.001
Primary site		<.001
Auricle skin	Reference
Skin of lips	NA	—
Eyelid	1.24 (0.07-6.92)	.84
Unspecified parts of face	1.16 (0.64-2.16)	.63
Scalp and neck	2.46 (1.42-5.60)	<.001
CSS (Month) median [Min, Max]	0.94 (0.93-0.95)	<.001

Figure 1.

(A–F) Cancer specific survival curve for patients with head and neck CMM after negative SLNB based on age (A), sex (B), Breslow (C), Mitotic rate (D), summary stage (E) and Primary Site (F). Mitotic rate A: <1, Mitotic rate B: 1∼4, Mitotic rate C: >4.

To further investigate the prognostic factors after negative SLNB, we constructed a multivariate Cox regression model of CSS (Figure 2), and the results showed that Breslow>4.0 mm was the highest independent risk predictor for patients (HR 3.82, 95% CI 2.04-7.16, P < .001). Significant independent risk predictors also included Breslow > 2.0-4.0 mm (HR 3.04, 95% CI 1.74-5.31, P < .001) and mitotic rate > 4 (HR 1.54, 95% CI 1.06-2.25, P = .023). Age< 60 years old was a significant survival predictor (HR .56, 95% CI 0.37-0.85, P = .007), and not scalp and neck CMM were also important factors for longer survival (auricle skin: HR .51, 95% CI .29-.90, P = 0.02; unspecified parts of face: HR .59, 95% CI .40-.87, P = .007).

Figure 2.

Cox multivariate analysis for patients with head and neck CMM after negative SLNB.

Surgical survival analysis

Among the 1597 patients (PSM) who received surgery (before PSM), higher Breslow scores, high mitotic rates and scalp and neck lesions were mainly concentrated in the main WME group (Table 3). The survival time between the two groups was significantly different (P = .012). To avoid age, gender and other variables on patients' adverse effects on CSS of patients, we carried out the clinical baseline of patients with two surgical methods. After matching the PSM, the clinical baseline data between patients in the WME and NME groups were not significantly different (P = .3), and the patient population with WME did not show longer survival (Figure 3).

Table 3.

Baseline demographic and clinical characteristics before and after matching the PSM.

	Before PMS		P-value	After PMS		P-value
	NME (N = 1006)	WME (N = 591)	P-value	NME (N = 322)	WME (N = 238)	P-value
Age			.228			.719
<60	365 (36.3%)	196 (33.2%)		66 (20.5%)	45 (18.9%)
≥60	641 (63.7%)	395 (66.8%)		256 (79.5%)	193 (81.1%)
Sex			.131			.095
Female	251 (25.0%)	127 (21.5%)		43 (13.4%)	45 (18.9%)
Male	755 (75.0%)	464 (78.5%)		279 (86.6%)	193 (81.1%)
Race			.494			1
Black	2 (.2%)	0 (0%)
Other	11 (1.1%)	5 (.8%)		2 (.6%)	1 (.4%)
White	993 (98.7%)	586 (99.2%)		320 (99.4%)	237 (99.6%)
Ulceration			.534			.511
No	575 (57.2%)	348 (58.9%)		182 (56.5%)	142 (59.7%)
Yes	431 (42.8%)	243 (41.1%)		140 (43.5%)	96 (40.3%)
Breslow			<.001			.171
≤1.0 mm	382 (38.0%)	159 (26.9%)		84 (26.1%)	47 (19.7%)
>1.0-2.0 mm	343 (34.1%)	211 (35.7%)		98 (30.4%)	83 (34.9%)
>2.0-4.0 mm	170 (16.9%)	133 (22.5%)		70 (21.7%)	63 (26.5%)
>4.0 mm	111 (11.0%)	88 (14.9%)		70 (21.7%)	45 (18.9%)
Mitotic rate			<.001			.155
<1	187 (18.6%)	84 (14.2%)		58 (18.0%)	33 (13.9%)
1∼4	583 (58.0%)	320 (54.1%)		155 (48.1%)	107 (45.0%)
>4	236 (23.5%)	187 (31.6%)		109 (33.9%)	98 (41.2%)
Summary stage			.45			.47
Localized	907 (90.2%)	525 (88.8%)		269 (83.5%)	205 (86.1%)
Regional	99 (9.8%)	66 (11.2%)		53 (16.5%)	33 (13.9%)
Primary site			.001			.619
Skin of lips	3 (.3%)	3 (.5%)		2 (.6%)	1 (.4%)
Eyelid	12 (1.2%)	3 (.5%)		3 (.9%)	2 (.8%)
Auricle skin	168 (16.7%)	90 (15.2%)		51 (15.8%)	30 (12.6%)
Unspecified parts of face	423 (42.0%)	200 (33.8%)		131 (40.7%)	90 (37.8%)
Scalp and neck	400 (39.8%)	295 (49.9%)		135 (41.9%)	115 (48.3%)
CSS (Month)			.667			.628
Mean (SD)	60.0 (23.2)	60.5 (25.4)		50.1 (23.4)	51.1 (25.8)
Median [Min, Max]	57.0 [1.00, 107]	60.0 [3.00, 107]		48.0 [1.00, 106]	49.0 [3.00, 107]

Figure 3.

Total survival curves for patients with the WME and NME groups. (A): before matching the PSM; (B): after matching the PSM.

Discussion

For CMM patients without clinically positive lymph node evidence, SLNB is the most important pathological evidence of lymph node metastasis. It can detect occult lymph node metastasis in some lymph node-negative patient groups and has become the most accurate auxiliary test for evaluating the prognosis of CMM patients. The study found that age ≥60 years old, scalp and neck CMM, Breslow >2.0 mm and mitotic rate >4 were associated with the prognosis of patients with head and neck CMM after negative SLNB.

Although CMM can occur anywhere on the body, melanoma at the head and neck site is more likely to exhibit aggressive clinicopathological features and has a poorer prognosis and higher recurrence rate than melanoma at other skin sites.¹⁵ However, the head and neck can be aesthetically divided into more than ten subunits: scalp, forehead, orbital area, ear and periauricular area, cheeks, nose and neck. Different aesthetic subunits have certain differences in the prognosis of CMM.¹⁶ The data of this study were analyzed by Cox multivariate regression analysis, suggesting that negative SLNB patients with scalp and neck CMM had a lower CSS rate, while patients with auricular skin and unspecified parts of the face had higher survival rates.

Age is a highly significant and powerful predictor of outcome in patients with cutaneous melanoma; although it lacks a significant predictive role for lymph node involvement, recurrence, and metastasis, increasing age may enhance the aggressive histological features of CMM, leading to poorer survival.^17,18 This study also confirmed that age ≥60 years old is an important factor affecting the prognosis of CSS in negative SLNB patients, which is consistent with the data of an earlier study.¹⁹ Although it is known that CSS rates are lower in male CMM patients, this study did not demonstrate an important relationship with survival outcomes in negative SLNB patients, which is different from some studies.

The American Joint Committee on Cancer (AJCC) TNM included tumor thickness, ulceration, mitotic rate and lymph node status in the risk assessment of recurrence and metastasis in patients with localized melanoma.²⁰ Multiple studies have confirmed that thin and medium-thick tumors (Breslow thickness ≤ 4.0 mm) are favorable prognostic factors for survival in CMM patients.²¹ This study confirms that there is a close relationship between tumor thickness and CSS in SLNB-negative patients. The features of the 7th edition AJCC T staging for CMM are related not only to overall tumor thickness and the presence of ulcers but also to the value of mitotic rates in lesions <1 mm thick (T1).²² The mitosis rate is an important predictor of sentinel lymph node metastasis in T1a melanoma.¹⁸ Increased tumor mitotic rates reduce 5-year overall survival and cancer-specific survival in patients with head and neck melanoma.²³ Although in the 8th edition of the AJCC Handbook, the tumor mitotic rate was deleted as a staging criterion, this study still recorded the data of the specific negative SLNB mitotic rate. Cox multivariate regression analysis found that a mitotic rate>4 was associated with SLNB-negative. A high mitotic rate is also an important CSS risk prediction factor for patients.

Surgery and systemic tumor control are important treatments for prolonging patient survival. In the absence of local organ invasion and tumor metastasis, surgical resection should be the first choice. However, extensive local excision, sentinel lymph node biopsy, and completion of lymph node dissection and reconstruction require special consideration of the location of the head and neck melanoma and the proposed margins and surrounding structures involved.²⁴ Although there are few controlled prospective studies evaluating WME in CMM, it has been shown that NME may lead to high recurrence rates in CMM patients, but larger WME does not confer oncological benefit.¹⁰ Our study further discussed the influence of tumor surgical resection extent on the prognosis of negative SLNB patients. We first performed a PSM analysis of patients with different resection extents by age, sex, and CMM pathology to reduce the effect of treatment selection bias on outcomes by matching clinical baseline data. The results of the study confirmed that WME and NME have no significant effect on CSS in patients with head and neck CMM after negative SLNB.

Limitations

This study has some limitations that must be addressed. First, treatment factors such as radiotherapy and chemotherapy were not considered in this study; second, as this was a retrospective review, there may be some selection bias and record entry errors; third, this study lacks specific information on patient relapse, so more prospective randomized studies are needed.

Conclusion

This study analyzed the risk factors affecting CSS in negative SLNB patients with CMM in the head and neck regions and observed a statistically significant difference in the prognosis of patients with CMM in different anatomical subunits of the head and neck regions. Close clinical follow-up for this population is necessary, and periodic medical examinations should be carried out.

Footnotes

Author contributions

Shulian Li, MS carried out the development or design of the methodology, statistical analysis, and original draft. Jun Tang MD designed the methodology and wrote original draft. Lanlan Wang. MS helped critically revise the draft for the design of the methodology and data curation. Juxiang Gou MS and Wei You MS participated in the methodology, data curation, and validation. Qiang Ji designed the conceptualization and carried out the article and editing. All authors have checked and approved the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics

We acknowledge the SEER database for providing their platforms and contributors for uploading their meaningful datasets. SEER belongs to public databases. The patients involved in the database obtained ethical approval. Users can download relevant data for free for research and publish relevant articles. Our study is based on open data, so there are no ethical issues or other conflicts of interest.

ORCID iD

Qiang Ji

References

Erdmann

Sigler

Uslu

, et al. Risk factors for regional and systemic metastases in patients with sentinel lymph node-negative melanoma. Anticancer Res. 2018;38(11):6571-6577.

Leong

Pissas

Scarato

, et al. The lymphatic system and sentinel lymph nodes: conduit for cancer metastasis. Clin Exp Metastasis. 2022;39(1):139-157.

Narang

Hue

Bingmer

, et al. Sentinel lymph node biopsy guideline concordance in melanoma: Analysis of the National Cancer Database. J Surg Oncol. 2021;124(4):669-678.

Joyce

McInerney

Joyce

, et al. A review of sentinel lymph node biopsy for thin melanoma. Ir J Med Sci. 2015;184(1):119-123.

Faries

Testori

AAE

Gershenwald

. Sentinel node biopsy for primary cutaneous melanoma. Ann Oncol. 2021;32(3):290-292.

Marchetti

Bartlett

. Sentinel lymph node biopsy in cutaneous melanoma-where do we stand? JAMA Dermatol. 2021;157(10):1159-1160.

Echanique

Ghazizadeh

Moon

, et al. Head & neck melanoma: A 22-year experience of recurrence following sentinel lymph node biopsy. Laryngoscope Investig Otolaryngol. 2021;6(4):738-746.

Egger

. The role of clinical prediction tools to risk stratify patients with melanoma after a positive sentinel lymph node biopsy. Ann Surg Oncol. 2021;28(8):4082-4083.

Cheriyan

Wernberg

Urquhart

. Head and neck melanoma. Surg Clin North Am. 2014;94(5):1091-1113.

10.

Torres Maldonado

Tam

Han

St John

. Head and neck melanoma: What is the goal for margins? Laryngoscope. 2022;132(5):916-917.

11.

Lee

Huynh

Teng

, et al. Predictors and survival impact of false-negative sentinel nodes in melanoma. Ann Surg Oncol. 2016;23(3):1012-1018.

12.

Kachare

Brinkley

Wong

, et al. The influence of sentinel lymph node biopsy on survival for intermediate-thickness melanoma. Ann Surg Oncol. 2014;21(11):3377-3385.

13.

Santos

FMD

Silva

Pedron

, et al. Association between tumor-infiltrating lymphocytes and sentinel lymph node positivity in thin melanoma. An Bras Dermatol. 2019;94(1):47-51.

14.

Morton

Thompson

Cochran

, et al. Final trial report of sentinel-node biopsy versus nodal observation in melanoma. N Engl J Med. 2014;370(7):599-609.

15.

Al-Qurayshi

Hassan

Srivastav

, et al. Risk and survival of patients with head and neck cutaneous melanoma: National perspective. Oncology. 2017;93(1):18-28.

16.

Derebasinlioglu

. Distribution of skin cancers of the head and neck according to anatomical subunit. Eur Arch Otorhinolaryngol. 2022;279(3):1461-1466.

17.

Tas

Erturk

. Patient age and cutaneous malignant melanoma: Elderly patients are likely to have more aggressive histological features and poorer survival. Mol Clin Oncol. 2017;7(6):1083-1088.

18.

Shannon

Wood

Straker

, et al. Age and mitogenicity are important predictors of sentinel lymph node metastasis in t1a melanoma. Ann Surg Oncol. 2021;28(9):4777-4779.

19.

Egger

Stepp

Callender

, et al. Outcomes and prognostic factors in superficial spreading melanoma. Am J Surg. 2013;206(6):861-867; discussion 867-8.

20.

Balch

Gershenwald

Soong

, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009;27(36):6199-6206.

21.

El Sharouni

Aivazian

Witkamp

, et al. Association of histologic regression with a favorable outcome in patients with stage 1 and stage 2 cutaneous melanoma. JAMA Dermatol. 2021;157(2):166-173.

22.

Ogata

Namikawa

Takahashi

Yamazaki

. A review of the AJCC melanoma staging system in the TNM classification (eighth edition). Jpn J Clin Oncol. 2021;51(5):671-674.

23.

Babajanian

Tamaki

Bordeaux

, et al. Clinical significance of tumor mitotic rate and lack of epidermal attachment in melanoma of the head and neck. Head Neck. 2018;40(8):1691-1696.

24.

Olla

Tufaro

Neumeister

. Extirpative considerations of melanoma of the head and neck. Clin Plast Surg. 2021;48(4):659-668.