Lymphatic drainage patterns of head and neck cutaneous melanoma: does primary melanoma site correlate with anatomic distribution of pathologically involved lymph nodes?

Abstract

The aim of this study was to analyse patterns of metastatic spread from cutaneous head and neck melanoma, which are said to be highly variable. The medical records of 145 patients with pathologically proven metastatic melanoma were reviewed retrospectively. The location of pathologically positive lymph nodes was compared with clinically predicted spread, and patients with metastatic disease in areas outside of predicted drainage patterns were considered aberrant. There were 33 curative and 73 elective neck dissections. 21 of 77 patients undergoing parotidectomy had positive results for metastases. Clinical prediction proved to be correct in 33 of 45 cases (73.3%). Two patients with lateralized melanomas were initially seen with contralateral metastases. Six of 45 patients (13.3%) developed contralateral metastases after neck dissection. Patients with clinical involvement of the parotid gland were at high risk of occult neck disease (40%). Patients undergoing neck dissection for primaries originating in face, forehead, coronal scalp, periauricular area, and upper neck should be considered for parotidectomy. Patients with posterior scalp and posterior neck primaries should be considered for selective neck dissection in conjunction with posterior lymphadenectomy. In patients with coronal scalp and periauricular primaries, a complete neck dissection including parotidectomy is the recommended approach.

Treatment of melanoma of the head and neck is a complex issue. The behaviour of this malignancy is aggressive, and it has an overall worse prognosis than that of other skin sites.

Histological involvement of regional lymph nodes by metastases of malignant melanoma in the head and neck region is a very poor prognostic sign, and it appears to carry an even more ominous prognosis than histological involvement of axillary or inguinal lymph nodes in draining melanomas of the trunk or extremities. Accurate localization of pathologically involved cervical lymph nodes is crucial in the treatment of these patients.

Some studies have reported that patterns of lymphatic drainage from a given area of skin in patients with melanoma are highly variable from patient to patient and that accurate clinical prediction of lymphatic pathways is not possible. This study is design to determine the accuracy of clinically predicting the distribution of pathologically involved lymph nodes related to the primary melanoma site.

Materials and methods

The medical records of patients with primary cutaneous head and neck melanomas, who were referred to the Department of Maxillofacial Surgery of the University Hospital Dubrava, University of Zagreb School of Medicine, from 1 January 1985 to 31 December 2005, were reviewed retrospectively. Patients with primary melanomas of the head and neck who underwent primary surgical treatment were included. The exclusion criteria were: patients who did not undergo primary surgical treatment; patients with primary cutaneous melanoma localized outside of the head and neck region; patients who had been previously excised, had local recurrence, or neck metastasis; and patients with external ear melanomas. Amongst 145 patients with cutaneous head and neck melanomas, 39 patients underwent wide local excision due to their thickness (<1 mm), whilst 106 patients underwent neck dissection.

All cases were re-classified according to the AJCC melanoma staging system. Primary tumour mitotic rate has been excluded from the survival analysis as a potential predictive factor because it was not a factor used in earlier classifications and data on it are incomplete and unreliable. The harvested lymph nodes were evaluated using serial sectioning with haematoxylin-eosin (H-E) staining. Immunohistochemical staining with antibodies for S-100p, HMB-45 and MART-1 was used for patients when H-E did not reveal evidence of metastatic disease since this procedure has 10–30% higher sensitivity for identifying micrometastases compared with conventional H-E staining. The location of pathologically positive lymph nodes was compared with clinically predicted spread, and all patients with metastatic disease in areas outside of predicted drainage patterns (including drainage to contralateral lymph node of neck) were considered aberrant. Follow-up intervals were calculated in months from the date of first treatment at the authors’ department to the date of last follow-up or death.

Statistical methods

In this retrospective, single-institution study, the clinical features, including age, gender, histological type, and localization of the tumours were obtained from clinical records. Patient characteristics were recorded on MS Excel ® spreadsheets and analysed using MedCalc statistical software (Version 11.3.1 © 1993-2010. MedCalc Software bvba, Broekstraat 52, 9030 Mariakerke, Belgium). Fisher’s exact test was used to determine the level of significance for categorical variables. The Kaplan–Meier statistical method was used to generate survival curves. Differences in survival were analysed using a log-rank test. Values of P < 0.05 were considered statistically significant.

Statistical methods

In this retrospective, single-institution study, the clinical features, including age, gender, histological type, and localization of the tumours were obtained from clinical records. Patient characteristics were recorded on MS Excel ® spreadsheets and analysed using MedCalc statistical software (Version 11.3.1 © 1993-2010. MedCalc Software bvba, Broekstraat 52, 9030 Mariakerke, Belgium). Fisher’s exact test was used to determine the level of significance for categorical variables. The Kaplan–Meier statistical method was used to generate survival curves. Differences in survival were analysed using a log-rank test. Values of P < 0.05 were considered statistically significant.

Results

Over the 21-year period, 145 patients with head and neck cutaneous melanomas were surgically treated. There were 106 neck dissections, amongst which 33 were curative and 73 were elective procedures. Radical neck dissections were performed as well as elective and curative treatments (14 versus 12), whilst the selective were used almost exclusively as an elective (40 versus 1). More than two-thirds of elective selective neck dissections were supraomohyoid neck dissection (34/40), and the only curative selective dissection was performed for regions 2–5. Modified radical neck dissections were performed more often as elective procedures (12 versus 7), whilst extended radical neck dissections were used more often as curative treatments (10 versus 6). There were 4 bilateral neck dissections (one bilateral modified radical neck dissections, two modified radical neck dissections with contralateral radical neck dissection, and one radical neck dissection with contralateral selective neck dissection). The types of neck dissections performed are summarized in Table 1 .

Table 1
Type of neck dissections performed.
Neck dissection Elective ND Curative ND Total
ERND 6 10 16
RND 14 12 26
MRND 12 7 19
1. SAN 9 5
2. SAN + IJV 2 1
SAN + IJV + SCM 1 1
SND 40 1 41
MRNDBL 0 1 1
MRND + RNDCL 1 1 2
RND + SNDCL 1–3 0 1 1
Total 73 33 106
ERND, extended radical neck dissection; RND, radical neck dissection; MRND, modified radical neck dissection; SAN, spinal accessory nerve; IJV, internal jugular vein; SCM, sternocleidomastoid muscle; SND, selective neck dissection; MRNDBL, bilateral modified radical neck dissection; RNDCL, contralateral radical neck dissection; SNDCL, contralateral selective neck dissection.

Figure 1 shows the Kaplan–Meier survival curves of patients with an ulcerated melanoma compared with patients with a non-ulcerated melanoma. Comparison of survival curves has shown significant difference with respect to survival ( P = 0.048). Age was considered a discontinuous variable and was divided into three arbitrary groups: 29 patients were aged 21–40 years; 33 patients were aged 41–60 years, and 44 patients were over 61 years old. The age groups failed to show any significant difference with respect to survival ( P = 0.624) ( Fig. 2 ). Gender was analysed by log-rank test comparison of survival curves. Cumulative survival at 5 years for the entire group was 61.3%. The gender groups differed significantly with respect to survival ( P = 0.042) ( Fig. 3 ). Lymph node involvement significantly decreased survival rates ( P < 0.01). Figure 4 shows comparative melanoma-specific survival amongst patients compared to nodal status. Analysis of influence on survival of negative neck compared with nodal disease in the parotid gland and neck and with nodal disease in the parotid lymphatics or neck alone has shown significant difference in survival rates ( P < 0.01). When all node-positive ( n = 45) and node-negative ( n = 51) patients were considered as two different groups, the cumulative survival was 44% and 87% at 5 years, respectively ( P < 0.01).

Fig. 1
Survival curves comparing patients with an ulcerated and patients with a nonulcerated melanoma. Solid line, nonulcerated melanoma ( n = 68); dashed line, ulcerated melanoma ( n = 28). Log-rank test, P = 0.048.

Fig. 2
Kaplan–Meier curves showing overall survival of patients of different ages with head and neck cutaneous melanoma. Solid line, ages 21–40 years; dashed line, ages 41–60 years; dotted line, 61 years and over. Log-rank test, P = 0.624.

Fig. 3
Kaplan–Meier curves showing cumulative melanoma-specific survival for 106 patients and survival of patients separated by gender. Solid line, men ( n = 56); dashed line, women ( n = 50). Log-rank test, P = 0.042.

Fig. 4
Kaplan–Meier survival curves of patients compared to nodal status. Solid line, negative neck ( n = 51); dashed line, positive neck ( n = 21); dotted line, positive parotid gland ( n = 8); dot-dashed line, positive parotid and neck ( n = 12). Log-rank test, P < 0.01.

Extracapsular extension (ECE) of melanomas was identified in 9 of 45 (20%) patients with histological positive nodes. There was no significant difference in survival between patients with ECE compared with patients without ECE ( P = 0.956). Lymph node metastases were found in 84.8% (28 of 33) of curative neck dissections, and occult metastases were identified in 17.8% (13 of 73) of elective neck dissections. In 77 patients, a parotidectomy was carried out in conjunction with a neck dissection. Amongst the elective procedures, there were 34 superficial and 28 total parotidectomies. The therapeutic procedures comprised 9 superficial and 6 total parotidectomies. 21 patients undergoing parotidectomy as a part of regional lymphadenectomy had involvement of the parotid lymphatics (11 of 15 in therapeutic parotidectomy and 10 of 62 in elective procedures). Amongst 15 patients with clinical metastatic melanoma of the parotid, 9 patients (60%) had pathological involvement of cervical lymph nodes (6 elective and 3 curative neck dissections). Mean follow up was 56.3 months (range 0–243 months). There was one postoperative death involving a 72-year-old man who suffered myocardial infarction and died 6 days after surgical therapy. As adjuvant treatment, the majority of patients received interferon ( n = 23), followed by chemotherapy ( n = 18), chemotherapy with radiation ( n = 3), radiation alone ( n = 1), and radiation with immunotherapy ( n = 1) and they were all included in the study group.

After eliminating patients with histological negative necks ( n = 51), patients with nodal metastases from an unknown primary melanoma site ( n = 6) and patients with primaries outside the head and neck region with regional metastases ( n = 4), 45 patients were eligible for this study. Clinical characteristics of these patients are summarized in Table 2 . The study group included 28 men and 17 women with a median age of 49 years (range 21–75 years).

Table 2
Clinical characteristics of 45 patients with nodal metastases from head and neck primary melanomas.
No. (%)
Male 28 (62.2)
Female 17 (37.8)
Age Years
Median 49
Range 21–75
Primary melanoma site No (%)
Face and forehead 16 (35.6)
Coronal scalp and periauricular area 16 (35.6)
Posterior scalp, upper and lower posterior neck 6 (13.3)
Upper neck 6 (13.3)
Lower anterior neck 1 (2.2)
Tumour thickness mm
Mean 4.2
Range 1–10
Only gold members can continue reading. Log In or Register to continue

Jan 26, 2018 | Posted by in Oral and Maxillofacial Surgery | Comments Off on Lymphatic drainage patterns of head and neck cutaneous melanoma: does primary melanoma site correlate with anatomic distribution of pathologically involved lymph nodes?
Premium Wordpress Themes by UFO Themes