Introduction
Cutaneous squamous cell carcinoma is a common type of skin cancer, with reported age-standardised rates of 77.3 and 34.1 per 100 000 person years for males and females, respectively, in the UK.Reference Venables, Autier, Nijsten, Wong, Langan and Rous1 There has been a steady increase in cutaneous malignant melanoma over the past decades. In the UK between 2016 and 2018, the average annual incidence rates of cutaneous malignant melanoma in females and males were 24.7 and 25.3 per 100 000 people, respectively.Reference Memon, Bannister, Rogers, Sundin, Al-Ayadhy and James2
Metastatic cutaneous SCC and cutaneous malignant melanoma can spread regionally to involve adjacent parotid lymph nodes or cervical lymph nodes and to distant sites. Lymph nodes are present in both the superficial lobe and deep lobe below the facial nerve.Reference Venables, Autier, Nijsten, Wong, Langan and Rous1,Reference Sönmez Ergün, Gayretli, Büyükpınarbaşılı, Yıldız, Gürses and Avşar3,Reference Pisani, Ramponi and Pia4 Pisani et al. examined lymph nodes in 18 total parotidectomy specimens. They found that 21 per cent of lymph nodes were in the deep lobe.Reference Pisani, Ramponi and Pia4 Studies evaluating parotid metastases from head and neck cutaneous SCC and cutaneous malignant melanoma found involvement of nodes that were both superficial and deep to the facial nerve.Reference Ahmed and Kelly5,Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6
The surgical management of the parotid gland by superficial parotidectomy or total parotidectomy in metastatic head and neck cutaneous SCC and cutaneous malignant melanoma is controversial. Some advise consideration of total parotidectomy in patients with evidence of superficial lobe metastases.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6,Reference Wertz, Durham, Malloy, Johnson, Bradford and McLean7 However, total parotidectomy may increase the risk of surgically related morbidity, most notably for facial nerve function with limited improvement in parotid area recurrence and survival.Reference Ruohoalho, Mäkitie, Aro, Atula, Haapaniemi and Keski-Säntti8–Reference Jecker, Hartwein and Töpfner11
The UK head and neck guidelines on metastatic head and neck cutaneous SCC advise resection of involved parotid tissue in case of parotid-positive disease, without specific mention of the extent of parotidectomy.Reference Newlands, Currie, Memon, Whitaker and Woolford12 The guidelines also recommend consideration of ipsilateral parotidectomy in parotid-negative disease with cervical involvement if the draining basin of the primary site is likely to pass through the parotid gland.Reference Newlands, Currie, Memon, Whitaker and Woolford12 Similar recommendations were given for head and neck metastatic melanoma.Reference Ahmed and Kelly5
The evidence to support the surgical management of the parotid gland for metastatic cutaneous malignant melanoma and cutaneous SCC of the head and neck is limited, particularly in regard to parotid gland management in patients with cervical nodal involvement without parotid involvement and the extent of parotidectomy in patients with only superficial parotid nodal involvement.
This systematic review aimed to appraise data from studies evaluating parotidectomy in metastatic cutaneous malignant melanoma and cutaneous SCC of the head and neck. The primary objectives were to examine parotid area recurrence, morbidity and survival in those undergoing superficial parotidectomy and total parotidectomy.
Materials and methods
The systematic review was conducted as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (‘PRISMA’) guidelines. The protocol was registered with Prospero (CRD42020217962).
Eligibility criteria
Primary interventional or observational studies evaluating parotidectomy in metastatic cutaneous SCC or cutaneous malignant melanoma of the head and neck were included. There was no restriction on publication date. Articles that were not available in the English language, had fewer than 10 patients, and those that were case reports, commentaries or letters, review articles, or animal studies were excluded. The outcome measures evaluated included regional recurrence, distant recurrence, surgical morbidity, quality of life and survival.
Information sources
PubMed, Embase, Central, ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform were systematically reviewed.
Search strategy
The following search terms were used in all search fields: (parotidectomy) AND (melanoma OR squamous cell carcinoma OR SCC OR squamous).
Study records
Two authors (WSL, OE) independently screened titles and abstracts from the initial search. The full text of records considered relevant were retrieved for further eligibility assessment. The reference list of included articles was screened and relevant full texts were assessed. All data collected by one author were checked for accuracy by another author. The data were collected using a standard data collection proforma.
Data items
The following data were extracted from the manuscript: study characteristics (study type, year, country); participant characteristics (age, gender); parotid involvement of SSC or malignant melanoma; type of intervention (superficial or total parotidectomy, adjuvant radiotherapy or systemic therapy); and outcome measures (including local recurrence, survival, morbidity, quality of life). Superficial parotidectomy was defined as removal of parotid tissue superficial to the facial nerve. Total parotidectomy was defined as removal of all parotid tissue superficial and deep to the facial nerve.
Risk of bias in individual studies
Quality assessment was performed on included studies. A quality assessment tool developed by Murad et al. was used to assess case series or single arm studies.Reference Murad, Sultan, Haffar and Bazerbachi13
Data synthesis
A formal narrative synthesis was performed given the variability in extent of parotidectomy, neck dissection and use of adjuvant radiotherapy or chemotherapy. We described recurrence, complications and survival in individual studies and explored relationship in study design, outcomes and complication rates between studies.
Results
Figure 1 is a Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram that outlines the process of inclusion and exclusion of studies. In total, 12 studies were included in this systematic review.
Fig. 1. Preferred Reporting Items for Systematic Reviews and Meta-analyses (‘PRISMA’) flow diagram showing study selection.
Malignant melanoma
A total of 5 studies reported outcomes by extent of parotidectomy in metastatic cutaneous malignant melanoma (a total of 193 patients underwent superficial parotidectomy and 57 underwent total parotidectomy).Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6,Reference Wertz, Durham, Malloy, Johnson, Bradford and McLean7,Reference Berger, van Veen, Madu, van Akkooi, Vogel and Balm14–Reference Vaglini, Belli, Santinami and Cascinelli16 See Table 1 for individual study outcomes.
Table 1. Characteristics of studies reporting on superficial parotidectomy and total parotidectomy in patients with cutaneous malignant melanoma
SP = superficial parotidectomy; TP = total parotidectomy; HB = House–Brackmann
Superficial parotidectomy
Superficial parotidectomy was performed in cases where there was no pre-operative evidence of parotid nodal involvement (electively)Reference Berger, van Veen, Madu, van Akkooi, Vogel and Balm14 or as a therapeutic procedure with pre-operative evidence of parotid involvement.Reference Vaglini, Belli, Santinami and Cascinelli16
Berger et al. evaluated elective superficial parotidectomy in 40 patients with metastatic cutaneous malignant melanoma.Reference Berger, van Veen, Madu, van Akkooi, Vogel and Balm14 No patient had pre-operative evidence of parotid involvement of disease. All patients had cervical lymph node involvement and also underwent concomitant neck dissection. In their series, 73 per cent had post-operative radiotherapy. A total of 25 of 40 patients (62.5 per cent) had recurrence: local (5 per cent), regional (3 per cent) and distant (55 per cent). Post-operative histology identified 10 of 40 (25 per cent) with occult positive parotid nodes. Of these, 8 patients (80 per cent) developed recurrence (compared with 17 of 30 (56 per cent) who did not have positive parotid nodes on histology). Two patients (5 per cent) in this series developed permanent facial weakness affecting the temporal or buccal branch. The median disease-specific survival for the cohort was 2.1 years.Reference Berger, van Veen, Madu, van Akkooi, Vogel and Balm14
Another study evaluated superficial parotidectomy as an elective (n = 16) or therapeutic (n = 30) procedure in patients with metastatic cutaneous malignant melanoma.Reference Vaglini, Belli, Santinami and Cascinelli16 All patients had concomitant neck dissection. The authors did not report on the proportion of patients who underwent adjuvant treatment. Of the 16 patients who underwent elective superficial parotidectomy, 5 (31 per cent) had occult metastasis in the parotid lymph node. None of the patients had parotid area recurrence over a follow-up period of 4–19 years. The overall 5-year survival was 37 per cent.Reference Vaglini, Belli, Santinami and Cascinelli16
Martins et al. examined 12 patients with metastatic or adjacent parotid cutaneous malignant melanoma.Reference Martins, Souza, Souza and Lage15 Six patients (50 per cent) had clinically positive parotid nodes or neck nodes. All patients underwent superficial parotidectomy and neck dissection. There was no information provided on adjuvant treatments. They reported 1 case of locoregional recurrence (unclear from the reporting if this was parotid area or cervical), distant metastases in 50 per cent, and 25 per cent of patients had no evidence of disease at 27 months.Reference Martins, Souza, Souza and Lage15
Total parotidectomy
Thom et al. evaluated the role of total parotidectomy and neck dissection in 23 patients with metastatic cutaneous malignant melanoma.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6 All patients had nodal disease involving the superficial parotid. A total of 14 patients (61 per cent) had adjuvant radiotherapy (n = 12) or chemotherapy (n = 2). Three patients (13 per cent) had separate metastases to both the superficial and deep lobe on post-operative histology. The facial nerve was preserved in 21 patients (91 per cent) following total parotidectomy. No patient developed parotid area recurrence during a median follow up of 30.6 months; however, 6 patients (26 per cent) developed cervical recurrence and 11 patients (48 per cent) had distant metastases. All 3 patients with deep lobe involvement developed recurrence in the neck within 10 months of surgery.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6
Superficial and total parotidectomy
One study compared the rate of parotid recurrence in patients who had undergone superficial parotidectomy or total parotidectomy for metastatic cutaneous malignant melanoma.Reference Wertz, Durham, Malloy, Johnson, Bradford and McLean7 They included a total of 129 patients; 34 underwent total parotidectomy and 95 had superficial parotidectomy. All patients had metastasis to the superficial parotid gland. There was no standardised protocol for the extent of parotidectomy. The indication for total parotidectomy was based on individual surgeon's discretion; however, it was generally performed when there was pre-operative or intra-operative evidence of deep lobe involvement. Therefore, the outcomes of superficial parotidectomy and total parotidectomy are not directly comparable given the selection bias. The use of adjuvant treatment was similar in both groups. There was no difference in facial nerve function post-operatively. Superficial parotidectomy was associated with a significantly higher rate of parotid bed recurrence compared with total parotidectomy (13 per cent vs 0 per cent).Reference Wertz, Durham, Malloy, Johnson, Bradford and McLean7
Squamous cell carcinoma
In the studies that evaluated cutaneous SCC (n = 8), 326 patients underwent superficial parotidectomy, and 220 patients underwent total parotidectomy (Table 2).Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6,Reference Cassisi, Dickerson and Million17–Reference Xie, Wang, Zhang and Liu23
Table 2. Characteristics of studies reporting on superficial parotidectomy and total parotidectomy in patients with cutaneous squamous cell carcinoma
* External auditory canal squamous cell carcinoma. TP = total parotidectomy; SP = superficial parotidectomy; SCC = Squamous cell carcinoma; HB = House–Brackmann
Total parotidectomy
Thom et al. evaluated the outcomes of therapeutic total parotidectomy and ipsilateral neck dissection in 42 patients with metastatic cutaneous SCC.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6 They found that 31 (74 per cent) patients had metastasis that involved the superficial lobe only and 11 (26 per cent) patients had metastasis that involved both superficial and deep lobes. A total of 9 patients (21 per cent) had adjuvant chemoradiotherapy and 27 patients (64 per cent) had radiotherapy alone. Patients were followed up for a median of 36.4 months: 3 (7 per cent) developed parotid area recurrence, 2 (5 per cent) had cervical recurrence and 9 (21 per cent) had distant metastases. There was no difference in 5-year locoregional control rates between patients without (88 per cent) and those with (78 per cent) deep parotid lobe involvement. In their univariate analysis, neck metastasis and N2 disease was associated with spread to deep parotid nodes. Involvement of the deep lobe of the parotid was associated with significantly higher rates of distant metastases, disease recurrence, death from disease and death from all causes.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6
Superficial and total parotidectomy
The indication for superficial parotidectomy was metastatic cutaneous SCC to the parotid gland in the absence of clinical, radiological or intra-operative evidence of deep lobe involvement,Reference Hirshoren, Ruskin, McDowell, Magarey, Kleid and Dixon19 or patients with cervical lymph node involvement with a primary tumour within the area of drainage of the parotid gland.Reference Kampel, Dorman, Horovitz, Warshavsky, Gutfeld and Muhanna20
The following indications for total parotidectomy were reported: evidence of metastasis to deep parotid nodes,Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6 facial nerve involvement during pre-operative assessment or identification of deep lobe involvement intra-operatively.Reference Goh, Bova and Fogarty18 Seven studies reported outcomes of superficial parotidectomy and total parotidectomy in patients with metastatic head and neck cutaneous SCC (Table 2).Reference Cassisi, Dickerson and Million17–Reference Xie, Wang, Zhang and Liu23
Hirshoren et al. reported outcomes in patients with metastatic cutaneous SCC who had superficial parotidectomy (n = 65) and total parotidectomy (n = 13).Reference Hirshoren, Ruskin, McDowell, Magarey, Kleid and Dixon19 The indication for superficial parotidectomy was metastatic cutaneous SCC to the parotid in the absence of clinical, radiological or intra-operative evidence of deep lobe involvement (otherwise a total parotidectomy was performed).Reference Sweeny, Zimmerman, Carroll, Schmalbach, Day and Rosenthal22 All patients had parotid lymph node involvement at presentation. A neck dissection was performed on 62 per cent (40 of 65) of patients who had superficial parotidectomy and 46 per cent (6 of 13) of patients who had total parotidectomy. Patients were followed up for a minimum of five years. Regional recurrence was observed in 11 patients (17 per cent) who underwent superficial parotidectomy and 4 patients (31 per cent) who underwent total parotidectomy.Reference Hirshoren, Ruskin, McDowell, Magarey, Kleid and Dixon19
Moore et al. also evaluated the outcomes in 40 patients with metastatic head and neck cutaneous SCC.Reference Moore, Weber, Prieto, El-Naggar, Holsinger and Zhou21 All patients had evidence of metastatic disease involving the parotid gland, cervical nodes or both at presentation; of these patients, 22 underwent superficial parotidectomy and 9 underwent total parotidectomy. The indication for extent of parotidectomy was not reported. A total of 26 patients (65 per cent) had concomitant neck dissection, and 37 patients (92.5 per cent) underwent post-operative radiotherapy. The authors reported that the extent of parotidectomy and facial nerve sacrifice did not significantly affect any survival outcome or time to recurrence.Reference Moore, Weber, Prieto, El-Naggar, Holsinger and Zhou21
Goh et al. evaluated 67 patients with metastatic cutaneous SCC who underwent superficial parotidectomy (n = 53) and total parotidectomy (n = 14); of these patients, 41 patients (61 per cent) presented with parotid node metastasis and 26 patients (39 per cent) presented with both parotid and cervical lymph node metastasis.Reference Goh, Bova and Fogarty18 The indication of total parotidectomy was pre-operative or intra-operative facial nerve involvement. A total of 54 patients (81 per cent) had concomitant ipsilateral neck dissection. All patients had adjuvant radiotherapy, and three patients had additional chemotherapy. Follow up ranged between 2 and 5 years. Patients who underwent superficial parotidectomy had a 10-year disease-specific survival of 17 per cent (8 of 47) compared with 0 per cent (0 of 14) for patients who underwent total parotidectomy.Reference Goh, Bova and Fogarty18
Kampel et al. evaluated 67 patients with advanced cutaneous SCC who underwent superficial parotidectomy (n = 27), total parotidectomy (n = 20), neck dissection (n = 14) and excision of primary tumour only (n = 6).Reference Kampel, Dorman, Horovitz, Warshavsky, Gutfeld and Muhanna20 A total of 35 patients (52 per cent) had parotid involvement at presentation and subsequently underwent parotidectomy, including superficial parotidectomy (n = 19) and total parotidectomy (n = 16). The extent of parotidectomy was at individual surgeon’s discretion. There was no difference in survival between total parotidectomy or radical parotidectomy compared with superficial parotidectomy. However, comparisons between parotidectomy and no parotidectomy showed improved overall 5-year survival in patients who underwent parotidectomy (59 per cent vs 18 per cent).Reference Kampel, Dorman, Horovitz, Warshavsky, Gutfeld and Muhanna20 However, this was not statistically significant in multivariable analysis.
Sweeney et al. evaluated 218 patients who underwent superficial parotidectomy (n = 110) and total parotidectomy (n = 108) for metastatic cutaneous SCC.Reference Sweeny, Zimmerman, Carroll, Schmalbach, Day and Rosenthal22 The indication for extent of parotidectomy was not reported; 81 per cent had concomitant neck dissection, and 48 per cent had adjuvant radiotherapy. Survival was not affected by the extent of parotidectomy and was similar irrespective of parotid lymph node involvement. However, overall survival decreased when cervical node involvement was present.Reference Sweeny, Zimmerman, Carroll, Schmalbach, Day and Rosenthal22
Two studies with very small cohorts (n = 15 and 21) also reported outcomes by extent of parotidectomy in patients with head and neck cutaneous SCC with parotid area involvement (see Table 2).Reference Cassisi, Dickerson and Million17,Reference Xie, Wang, Zhang and Liu23 Table 3 shows the quality assessment of the included studies. The median quality score was 3 out of 5.
Table 3. Quality assessment of included studies
* Maximum score = 5
Discussion
This systematic review reported outcomes in patients with metastatic cutaneous SCC and cutaneous malignant melanoma who underwent parotidectomy as part of their surgical management.
Only five studies reported the outcomes of the extent of parotidectomy in patients with cutaneous malignant melanoma. The effect of parotidectomy was generally a secondary objective. In addition, four of these studies had no control group. Only one study directly compared outcomes between superficial parotidectomy and total parotidectomy in patients with metastatic cutaneous malignant melanoma.Reference Wertz, Durham, Malloy, Johnson, Bradford and McLean7 The authors found a significantly higher rate of parotid area recurrence in patients who had superficial parotidectomy compared with total parotidectomy; there was no difference in post-operative facial function.Reference Wertz, Durham, Malloy, Johnson, Bradford and McLean7 However, their analysis is limited by selection bias as the choice of extent of surgery was at the surgeons’ discretion and influenced by tumour characteristics.
The outcomes following superficial parotidectomy and total parotidectomy for cutaneous SCC were reported in seven studies. There are conflicting findings: some found higher regional recurrenceReference Hirshoren, Ruskin, McDowell, Magarey, Kleid and Dixon19 and reduced survival in total parotidectomy.Reference Goh, Bova and Fogarty18,Reference Xie, Wang, Zhang and Liu23 Other authors found no difference in survival between superficial parotidectomy and total parotidectomy.Reference Kampel, Dorman, Horovitz, Warshavsky, Gutfeld and Muhanna20 Two studies reported that the survival outcomes were not affected by the extent of parotidectomy; however, there were no specific data or survival analyses available to support these conclusions.Reference Moore, Weber, Prieto, El-Naggar, Holsinger and Zhou21,Reference Sweeny, Zimmerman, Carroll, Schmalbach, Day and Rosenthal22
A possible explanation for the fact there was no difference between superficial parotidectomy and total parotidectomy is that most patients have adjuvant radiotherapy. This may treat occult metastases in the deep lobe in patients who undergo superficial parotidectomy. Comparisons of outcomes were limited by the small sample sizes and selection biases.
Some authors recommend total parotidectomy in cutaneous malignant melanoma with nodal metastases in the parotid gland, even if considered to be only affecting nodes superficial to the facial nerve.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6 The rationale is to clear all parotid lymph nodes and reduce regional recurrence rates from occult positive deep lobe disease. In the case of a positive cervical node without parotid involvement, the rationale is to clear a possible occult node in the parotid gland: if the primary cutaneous disease is at a site thought to drain through the parotid gland (i.e. the scalp, forehead and lateral aspect of face (temple, cheek, periauricular)).Reference Vauterin, Veness, Morgan, Poulsen and O'Brien24 In the setting of metastatic cutaneous malignant melanoma, one argument for total parotidectomy would be the high rate (13 per cent) of occult nodal disease reported in the deep lobe.Reference Thom, Moore, Price, Kasperbauer, Starkman and Olsen6
The difference in morbidity between superficial parotidectomy and total parotidectomy is important for decision-making. Although most small series report no difference between superficial parotidectomy and total parotidectomy with regards to facial weakness in our review,Reference Moore, Weber, Prieto, El-Naggar, Holsinger and Zhou21,Reference Sweeny, Zimmerman, Carroll, Schmalbach, Day and Rosenthal22 systematic reviews have found higher rates of paresis of 6.75 per cent and 15 per cent in superficial parotidectomy and total parotidectomy, respectively, for benign tumours. Furthermore, the mean rate of facial paralysis was 0.8 per cent and 4.4 per cent after superficial parotidectomy and total parotidectomy, respectively.Reference Fol HA, Beheiri and Zaqri25
There are a number of limitations to this systematic review. Generally, the extent of parotidectomy (superficial vs total) was not comparable because the extent of parotidectomy was often based on the pre-operative extent of disease; therefore, it was limited by selection bias. In addition, the heterogeneity in practices and study methods prevented quantitative summation and analysis.
Based on the findings of this systematic review, there is a scope for a well-designed multicentre study with a prespecified treatment protocol, which can address the following controversies: superficial parotidectomy or total parotidectomy or active monitoring of the parotid area for patients with head and neck cutaneous SCC or cutaneous malignant melanoma with no clinical evidence of parotid node involvement but with cervical involvement, and superficial parotidectomy or total parotidectomy in patients with head and neck cutaneous SCC or cutaneous malignant melanoma with pre-operative evidence of superficial parotid nodal involvement. This would help identify the relationship between the extent of parotidectomy against functional and oncological outcomes. Variables such as tumour stage, neck dissection and adjuvant treatment would also need to be controlled.
Conclusion
The association between extent of parotidectomy and outcomes in cutaneous malignant melanoma and cutaneous SCC remains unknown because of lack of data, varied practices, selection biases and small sample sizes. This systematic review highlights the need for well-designed studies to direct better care.
Competing interests
None declared
