Ten-year experience on the surgical approach of parotid pleomorphic adenoma

Introduction

Pleomorphic adenomas (PAs) are benign tumours with heterogenous clinical manifestations and histological characteristics (1). They are the most common salivary gland tumour, comprising as much as 60 percent of all salivary gland neoplasms, and are the most common benign parotid gland tumour (2). PA predominantly involves the superficial lobe of the parotid gland, in which most of the parotid parenchyma is found (3). Recurrence is common, and ‘metastasis’, although histologically benign, has been well described in the literature (4). There is also a risk of malignant transformation (carcinoma ex-PA) with an incidence described to be 1.5% within the first 5 years and up to ~10% at 15 years (1,4-6). Therefore, the gold standard management for these tumours remains to be surgical excision; however, the optimal surgical approach remains unclear. With parotid tumour resection first described by Bernard in 1823, enucleation described by Senn in 1895, and superficial parotidectomy further defined by Janes and Bailey in the 20^th century, surgical excision is a mainstay in the treatment of PA (6).

The aetiology of PA remains ill-defined (1). Histologically, PAs are diverse with heterogenous appearance of their epithelium. Sub-types include myoepithelial, epithelial or paucicellular PAs, with predominantly eosinophilic and hyalinised stroma that is occasionally characterised by chondroid, fibrotic, or myxoid stromal changes (4,7). PAs have discontinuous lobular margins caused by ‘melting’ of myoepithelial cells from the ducts into the stroma that facilitate infiltration of pseudopodia through the capsule and herniation of the tumour (6,7). These features facilitate dissemination of tumour into the vasculature which can lead to an increased risk of recurrence as well as give rise to benign metastasising PA (4,6).

Several classification and nomenclature systems exist for defining parotid surgery. For the purposes of the study, we have classified each approach into the following groups:

Extracapsular dissection (ECD) involves complete en bloc removal of the tumour and its capsule, along with a healthy margin of tissue 2–3 mm without planed dissection of the facial nerve trunk (2). Complete superficial parotidectomy (CSP) comprises dissection of all main branches of the facial nerve with the excision of the entire superficial lobe of the parotid, while partial superficial parotidectomy (PSP) comprises partial dissection of the facial nerve trunk and its branches in the superficial lobe nearest the tumour, taking a margin of tissue around the neoplasm (1,8). In contrast, TP involves removing both deep and superficial lobes of the parotid with full dissection and sparing of the facial nerve branches (1,8).

Parotid surgery is associated with several notable complications, which are influenced in-part by locoregional anatomy. The facial nerve is liable to injury given its intimate association with the gland leading to temporary or permanent paresis or paralysis (3). Autonomic dysfunction of the auriculotemporal nerve leads to Frey’s syndrome, in which gustatory stimuli precipitate local skin erythema and diaphoresis (3). Injury or sacrifice to the great auricular nerve (GAN) may result in earlobe paraesthesia (3). Other complications including first bite syndrome, salivary fistula and sialocele are also well described within the literature.

Recurrence of disease can also occur and is linked to morphological characteristics of the tumour including hypocellular and chondromyxoid tumours, incomplete tumour capsule, tumour size and presence of pseudopodia and satellite nodules (6,9,10). Other factors including deep lobe involvement, previous recurrence and decreased preoperative facial nerve function also contribute to an increased risk of recurrence (3,11). Surgical variables including surgical approach (in particular enucleation) resection margins, intra-operative tumour spill or capsule breach have also been demonstrated to lead to an increased risk of recurrence (9). Typically, recurrence is detected 7 to 10 years after the index procedure (6).

Overall, our study aims to analyse the relationships between surgical approach and its impact on surgical outcomes including recurrence and facial nerve outcomes. Furthermore, we aim to investigate the impact of tumour variables including tumour size and location in regard to these surgical outcomes.

To the author’s knowledge, this is the one of the largest series to date on PAs in Australia. As a high-volume head and neck unit, the strength of the data lies in the large sample size and comprehensive retrospective time period of 10 years.

Methods

All patients with cytological or histopathological diagnosis of PA at a single Australian tertiary level head and neck centre, Royal Brisbane and Women’s Hospital (RBWH), Brisbane, Australia over a 10-year period [2009–2019] were included in the retrospective analysis. Data were collected from the patients’ medical records and medical integration systems (iEMR, the Viewer, ORMIS, AUSLAB). Each operation report was assessed and classified into the relevant surgical approach. All data collected were anonymised and stored in a secure electronic system and a standardised data extraction template with predefined outcomes and classifications was used for data extraction. Patients who presented to RBWH with recurrence had details of their original surgery from the referring centre extracted and included in the data analysis. Patients with incomplete medical records were excluded. All operations were either performed or supervised by an experienced consultant surgeon who was a member of the Royal Australian College of Surgeons (RACS). The study is reported according to the STROBE reporting guidelines (available at https://www.theajo.com/article/view/10.21037/ajo-24-42/rc).

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and approved by the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (Ref No. LNR/2020/QRBW/64972). Because of the retrospective nature of the research, the requirement for informed consent was waived.

Outcomes

Surgical outcomes were collected including prevalence of recurrence, facial nerve injury (transient or permanent), Frey’s syndrome, capsule breach, malignant transformation, adverse surgical outcomes including haematoma, salivary fistula, seroma, GAN numbness, discomfort, abscess/infection and sialocele. Patients with transient facial nerve dysfunction were followed for up to 12 months; if function had not fully returned then these patients were classified as a permanent palsy.

Secondary outcomes including location of tumour, tumour size and resection margins were also collected and analysed.

An analysis was performed on patients with parotid PAs who were classified according to their surgical approach as defined by the following:

• ECD: dissection of parotid mass surrounded by a cuff of healthy parotid tissue without planed dissection and identification of the facial nerve trunk;
• PSP: dissection of parotid mass surrounded by a cuff of healthy parotid tissue with dissection and identification of the facial nerve trunk;
• CSP: complete dissection of the superficial lobe of parotid with dissection of all branches of facial nerve;
• SDP: dissection of parotid mass within the deep lobe and selectively removed;
• TP: total removal of the parotid gland.

A comparison of the five main surgical modalities was performed evaluating the impact of surgical approach on recurrence, facial nerve injury and other adverse surgical outcomes. Additional data was collected on potential confounders including use of facial Nerve Integrity Monitor (NIM) (intra-operative facial nerve monitor) and a multivariate logistic regression was performed on facial paresis, facial paralysis and surgical approach with results adjusted accordingly.

Statistical analysis

Logistic regression was used to evaluate the relationship between surgical approach and surgical outcomes. Results have been expressed as odds ratio (OR) and 95% confidence interval (CI). Statistically significant results are defined by a P value <0.05. STATA version 15 was used for statistical analysis.

Results

Epidemiology, location and distribution of PA

A total of 338 patients were diagnosed with PA over a period of 10 years at a single tertiary centre. The median age at diagnosis was 50 years old (interquartile range, 37–64 years old). More females were affected than males (61% vs. 39%). Distribution of the location of PA is listed in Table 1, with 83.2% of our patients having a PA of the parotid gland (Table 1).

Distribution of parotid PAs

Of the total number of patients (N=338), 283 were diagnosed with PA of the parotid gland. Of these 283 patients, 82% (n=232) were found in the superficial lobe, 14% (n=40) in the deep lobe, 1% (n=3) in accessory and 3% (n=8) affecting both superficial and deep lobes (Figure 1).

Figure 1 Distribution of parotid pleomorphic adenomas by subsite.

The median tumour size of parotid PA was 23 mm. Table 2 outlines the tumour size with each parotidectomy approach. There was no significant association found between tumour size and subsequent recurrence or other adverse surgical outcomes.

Surgical outcomes

After excluding patients due to incomplete records, of the 283 patients with parotid PA, 242 patients were included for full analysis of surgical outcomes with details of the surgical approach listed in Table 3.

The prevalence of temporary facial nerve paresis was 18% (n=44). Six percent patients had permanent facial nerve paresis (n=15). The prevalence of Frey’s syndrome was 3% (n=7). The prevalence of ‘other adverse surgical outcomes’ including great auricular paraesthesia, seroma, sialocele and more are outlined in Figure 2.

Figure 2 Prevalence of adverse surgical outcomes excluding facial nerve paralysis following parotidectomy. GAN, great auricular nerve.

Recurrence

Of the 242 patients with parotid PA, 14 presented to RBWH with recurrence. Details of their original surgery were retrieved and recorded. The median time to recurrence was 13 years (range, 2–47 years). The minimum follow-up period was 3 months and the maximum period was up to 47 years.

Patients with capsule breach intraoperatively had greater odds of recurrence (OR =6.42, 95% CI: 2.03–20.34, P<0.01).

Further, patients undergoing ECD had greater odds of recurrence than those with PSP (OR =74.00, 95% CI: 5.83–938.80, P<0.01). There was no significant association found between histological subtype or resection margin with recurrence.

Parotid surgical approach and facial nerve outcomes

Facial nerve dysfunction was defined as temporary (paresis) or permanent (palsy). See Table 4 for a logistic regression analysis of variables for facial nerve palsy. ECD had greater odds of permanent facial nerve palsy in comparison to patients who underwent a PSP (OR =25.00, 95% CI: 2.56–243.07, P<0.01). Those who underwent a TP also demonstrated an increased odds of facial nerve palsy (OR =17.86, 95% CI: 3.51–90.91, P<0.01). After mutual adjustment, patients undergoing PSP was the superior surgical approach in regard to minimising facial nerve palsy in comparison to TP, SDP and ECD (OR =13.04, 95% CI: 1.84–92.55, P<0.01; OR =9.33, 95% CI: 1.12–77.87, P<0.05; OR =29.52, 95% CI: 2.43–358.47, P<0.01), respectively. There was no significant difference demonstrated when comparing PSP with CSP and facial nerve dysfunction. There was no association between surgical approach and temporary facial nerve paresis.

Other adverse surgical outcomes

The most common other adverse surgical outcome was GAN paraesthesia followed by seroma (Figure 2). There was no significant association found between surgical approach and these other adverse surgical outcomes.

Location of tumour and facial nerve outcomes

Patients with tumours involving both lobes of the parotid gland or the accessory lobe had greater odds of facial palsy compared to those with tumours affecting the superficial lobe only (OR =28.71, 95% CI: 3.52–234.45, P<0.01 and OR =47.86, 95% CI: 9.49–241.25, P<0.001, respectively).

Discussion

Surgical resection is considered the gold standard treatment for parotid PA; however, the optimal method of surgical approach remains a topic of contention. Minimising adverse surgical outcomes such as facial nerve palsy and recurrence is of particular interest amongst surgeons as these can carry significant morbidity. Over the decades there has been shifts in the trends of approaching parotid PA; from the now disregarded enucleation, to complete superficial or TP to now limited parotid surgery of ECD and PSP (12-14).

Limited parotid surgery in the forms of ECD and PSP in which a cuff of parotid tumour is taken alongside the tumour is now the favoured approach amongst surgeons rather than CSP in which the whole superficial lobe is removed. This trend towards limited parotid surgery is supported in the literature by studies demonstrating superiority regarding facial nerve outcomes and recurrence (3,15,16).

A systematic review and meta-analysis by Foresta et al. (3) encompassing 123 studies from 1950–2011 concluded that the rates of PA recurrence, facial nerve paralysis, and Frey’s syndrome were higher in patients who underwent SP compared to ECD. Similarly, a systematic review and meta-analysis of 14 cohort studies by Xie et al. (16) comprising 3,194 patients comparing outcomes following SP and ECD revealed that ECD was superior to SP; with significantly lower rates of facial nerve palsy (transient and permanent) and Frey’s syndrome in patients undergoing ECD in comparison to those who underwent superficial parotidectomy. They did not find a significant difference between surgical approach and rates of recurrence [risk ratio =0.71 (95% CI: 0.40–1.27)] (16).

However, these studies did not delineate between the types of superficial parotidectomy approaches (i.e., CSP vs. PSP); therefore, although results can be concluded that limited parotid surgery was favourable, further research was required to compare between the now mainstay limited parotid surgical approaches of ECD and PSP.

Both approaches involve excision of the tumour with a cuff of healthy parotid tissue however PSP includes the addition of pre-identifying and dissecting the facial nerve trunk. An Australian study by O’Brien [2003], in which the author analysed the outcomes of his own partial superficial parotidectomies from 1988–2002, demonstrated lower rates of complications and recurrence compared to reported outcomes from the “Cleveland Clinic and a large European Center” (17). Laskaris evaluated 266 patients comparing these two approaches and concluded both groups had similar post-operative rates with no statistically significant difference in recurrence rates and facial nerve outcomes (18).

Interestingly, the study presented here revealed that patients undergoing PSP had superior outcomes in respect to facial nerve palsy in comparison to those who underwent extracapsular dissection (P<0.01), thereby suggesting pre-identification and dissection of the facial nerve leads to better facial nerve outcomes. A randomised control trial comparing limited surgery (PSP) with conventional surgery (CSP) found that PSP reduced duration of surgery and operative morbidity (19).

Further, the patients presented here that underwent ECD had greater odds of recurrence than those with PSP (OR =74.00, 95% CI: 5.83–938.80, P<0.01). Rather, this result is supported by the findings in a Viennese study on 894 patients, in which the rates of positive resection margins (29.4% vs. 10.2%, P<0.001), disease recurrence (7.2% vs. 2.2%, P=0.0003) and facial nerve palsy (2.2% vs. 0.6%, P=0.0396) were found to be more common after ECD compared to SP (20).

Regarding capsule breach, this is a well-known risk factor for developing recurrence. Likewise, in the sample of this paper, patients whose surgeries were documented to have capsule breach due to intra-operative spill had greater odds of recurrence (OR =6.42, 95% CI: 2.03–20.34, P<0.01). Notably, there was no association found between tumour size and recurrence or facial nerve dysfunction

Location of the tumour also had an impact on post-operative facial nerve dysfunction with tumours crossing the nerve and involving both lobes of the parotid gland or the accessory lobe having greater odds of facial paralysis compared to those with tumours affecting the superficial lobe only (OR =28.71, 95% CI: 3.52–234.45, P<0.01 and OR =47.86, 95% CI: 9.49–241.25, P<0.001, respectively). This is supported by one of the largest retrospective series performed on incidence of facial nerve weakness in parotid surgery similarly demonstrating tumours affecting both lobes to have increased facial nerve palsy (21).

Limitations

The authors acknowledge there may be limitations within this study; firstly, due to the retrospective nature of the study, the accuracy of the data lies in the detail of documentation including the operation reports collected. Further, the nomenclature of parotid surgery is exceptionally diverse (e.g., subtotal parotidectomy which can be synonymous with PSP). To overcome this when there was ambiguity in the operation reports (in 9 cases), the surgeons were contacted directly to clarify the definition of nomenclature recorded to ensure accuracy in classifying the surgical approach. In cases where there was insufficient information or incomplete medical records (41 cases), these were entirely excluded from the study. Further, the degree of facial nerve dysfunction, as graded by the House Brackmann score or details on individual branches, could not be analysed as this was not consistently recorded within patient documentation. Length of surgery was not recorded, which may be a possible confounding factor. Future research including this as a variable may be considered to assess its potential impact on outcomes.

It should also be noted that pre-identification and dissection of the facial nerve is the gold standard teaching within this teaching tertiary hospital—therefore, the majority of parotid surgeries were CSP or PSP with comparatively smaller numbers for the ECD which should be taken into account when interpreting results. This retrospective analysis was conducted at a single tertiary centre which provides a central service for a large population. As it is a training hospital, surgeries were performed by operators with varying levels of experience. However, all procedures were directly supervised by a RACS-accredited consultant. This supervision ensured that surgical techniques and decision-making adhered to established standards, minimising the potential impact of operator variability on patient outcomes. Additionally, this model reflects the real-world practice in Australian healthcare settings, where supervised training is integral to surgical education and the public healthcare system. Patient selection in this study was based on predefined clinical criteria, i.e., all patients with cytological or histopathological diagnosis of PA, thereby ensuring a homogenous cohort and minimising patient selection bias. While it is recognised that some inherent selection bias may remain due to the metropolitan location, this hospital serves a large catchment inclusive of regional and rural patients, thereby reducing bias and ensuring a diverse patient population is represented.

Overall, the authors believe this local data remains valuable and would be applicable to similar settings (particularly at RACS credentialled surgical training hospitals) where surgical practices and standards of care are generally comparable. However, there may be value in expanding future research across multiple centres to capture surgeon and patient diversity to further strengthen and validate these findings.

Conclusions

In conclusion, the findings from this extensive 10-year retrospective analysis revealed that PSP were superior to ECD in both facial nerve morbidity and recurrence rates (P<0.01). Notably, capsule breach increased the risk for recurrence (P<0.01). Interestingly, there was no relationship found between tumour size with facial nerve injury or recurrence. Tumours involving both superficial and deep lobes also increased odds of facial nerve palsy (P<0.01). Ultimately, this study suggests PSP should be considered as the main surgical approach for parotid PA if clinically appropriate, in order to minimise surgical complications and improve patient outcomes.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/ajo-24-42/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-24-42/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-24-42/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-24-42/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and approved by the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (Ref No. LNR/2020/QRBW/64972). Because of the retrospective nature of the research, the requirement for informed consent was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Carlson ER, Ord RA. Salivary gland pathology: diagnosis and management. Second edition. Hoboken, NJ: John Wiley & Sons, Inc./Wiley Blackwell; 2016.
2. Zbären P, Vander Poorten V, Witt RL, et al. Pleomorphic adenoma of the parotid: formal parotidectomy or limited surgery? Am J Surg 2013;205:109-18. [Crossref] [PubMed]
3. Foresta E, Torroni A, Di Nardo F, et al. Pleomorphic adenoma and benign parotid tumors: extracapsular dissection vs superficial parotidectomy--review of literature and meta-analysis. Oral Surg Oral Med Oral Pathol Oral Radiol 2014;117:663-76. [Crossref] [PubMed]
4. Licitra L, Locati LD. Salivary gland cancer: from diagnosis to tailored treatment. Cham, Switzerland: Springer; 2019.
5. Seifert G. Histopathology of malignant salivary gland tumours. Eur J Cancer B Oral Oncol 1992;28B:49-56. [Crossref] [PubMed]
6. Witt RL, Eisele DW, Morton RP, et al. Etiology and management of recurrent parotid pleomorphic adenoma. Laryngoscope 2015;125:888-93. [Crossref] [PubMed]
7. Speight PM, Barrett AW. Salivary gland tumours. Oral Dis 2002;8:229-40. [Crossref] [PubMed]
8. Tweedie DJ, Jacob A. Surgery of the parotid gland: evolution of techniques, nomenclature and a revised classification system. Clin Otolaryngol 2009;34:303-8. [Crossref] [PubMed]
9. Dulguerov P, Todic J, Pusztaszeri M, et al. Why Do Parotid Pleomorphic Adenomas Recur? A Systematic Review of Pathological and Surgical Variables. Front Surg 2017;4:26. [Crossref] [PubMed]
10. Grasso M, Fusconi M, Cialente F, et al. Rupture of the Pleomorphic Adenoma of the Parotid Gland: What to Know before, during and after Surgery. J Clin Med 2021;10:5368. [Crossref] [PubMed]
11. Rooker SA, Van Abel KM, Yin LX, et al. Risk factors for subsequent recurrence after surgical treatment of recurrent pleomorphic adenoma of the parotid gland. Head Neck 2021;43:1088-96. [Crossref] [PubMed]
12. Papadogeorgakis N. Partial superficial parotidectomy as the method of choice for treating pleomorphic adenomas of the parotid gland. Br J Oral Maxillofac Surg 2011;49:447-50. [Crossref] [PubMed]
13. Li C, Xu Y, Zhang C, et al. Modified partial superficial parotidectomy versus conventional superficial parotidectomy improves treatment of pleomorphic adenoma of the parotid gland. Am J Surg 2014;208:112-8. [Crossref] [PubMed]
14. Loke WL, Rahimi S, Brennan PA. An update on extracapsular dissection for the management of parotid gland pleomorphic adenoma. J Oral Pathol Med 2022;51:219-22. [Crossref] [PubMed]
15. Cristofaro MG, Allegra E, Giudice A, et al. Pleomorphic adenoma of the parotid: extracapsular dissection compared with superficial parotidectomy--a 10-year retrospective cohort study. ScientificWorldJournal 2014;2014:564053. [Crossref] [PubMed]
16. Xie S, Wang K, Xu H, et al. PRISMA-Extracapsular Dissection Versus Superficial Parotidectomy in Treatment of Benign Parotid Tumors: Evidence From 3194 Patients. Medicine (Baltimore) 2015;94:e1237. [Crossref] [PubMed]
17. O'Brien CJ. Current management of benign parotid tumors--the role of limited superficial parotidectomy. Head Neck 2003;25:946-52. [Crossref] [PubMed]
18. Laskaris S, Chrysikos D, Koutrafouris I, et al. Partial Superficial Parotidectomy Versus Extracapsular Anatomical Dissection for the Treatment of Benign Parotid Tumors. Acta Med Acad 2022;51:85-91. [Crossref] [PubMed]
19. Roh JL, Kim HS, Park CI. Randomized clinical trial comparing partial parotidectomy versus superficial or total parotidectomy. Br J Surg 2007;94:1081-7. [Crossref] [PubMed]
20. Kadletz L, Grasl S, Grasl MC, et al. Extracapsular dissection versus superficial parotidectomy in benign parotid gland tumors: The Vienna Medical School experience. Head Neck 2017;39:356-60. [Crossref] [PubMed]
21. Jin H, Kim BY, Kim H, et al. Incidence of postoperative facial weakness in parotid tumor surgery: a tumor subsite analysis of 794 parotidectomies. BMC Surg 2019;19:199. [Crossref] [PubMed]