Abstract
Objectives
To investigate the imaging and anatomic features of the anterior lobe (AL) of the superficial parotid gland (SPG).
Methods
Computed tomographic sialography examinations were undertaken for 142 parotid glands in 77 patients. Whole computer tomography (CT) data were analyzed using multi-planar reformation and maximum intensity projection to generate sialographic CT images. The tributary ducts of the SPG were analyzed to classify the parotid morphology. Three-dimensional analyses were used to investigate the AL and its relationship with adjacent anatomic landmarks.
Results
Four major types (I-IV) and 2 minor types (V-VI) of the AL and the superficial parotid gland were observed. Type I AL (83/142) was contiguous and not separated from the retromandibular parotid gland. Type II AL (16/142) was detached from the retromandibular parotid gland with 1-4 tributary ducts. Type III AL (12/142) showed a small isolated lobe above the Stensen duct around the anterior edge of the masseter. Type IV (28/142) showed the absence of the AL. Type V (3/142) shows the absence of the retromandibular parotid gland. Type VI (3/142) showed the presence of ectopic salivary gland beneath the Stensen duct anterior to the retromandibular parotid gland.
Conclusions
The AL gives rise to the morphological variations of the superficial parotid gland. AL also gives rise to the accessory parotid gland when it is detached from the retromandibular parotid gland.
Keywords: salivary glands, parotid gland, anatomy, sialography
Introduction
The parotid gland, the largest among the salivary glands, is commonly depicted as an organ with an inverted pyramid shape. It is anatomically divided into superficial and deep lobes, with the separation delineated by the facial nerve. Anatomy of the superficial parotid gland (SPG) is important for the clinical practice of parotid gland diseases. Surgeries of the parotid gland have been evolving towards more conservative protocols in recent years focused on the management of the parotid gland.1 Superficial parotidectomy is the traditional standard approach for surgical excision of the parotid tumors.2 Incomplete excision of the SPG is considered to give rise to postoperative recurrence. Less radical techniques including partial superficial parotidectomy and extracapsular dissection have been introduced in recent years to minimize injury to the facial nerve.3
The morphology of the SPG is highly variable depending on the anterior extending part.4 The “anterior extension” or “facial process” of SPG described by Frommer explained this anatomic variance.5 Three variations of the parotid gland have been described. The first is the parotid gland without the anterior extension. The second is the presence of an attached anterior extension or facial process. The third variation is the presence of a detached glandular mass or “accessory parotid gland” (APG). When an “anterior extension” or “facial process” is present, the gland extends much closer to the anterior margin of the masseter. There is no exclusive nomenclature for this part of the glandular tissue in the literature. And relevant anatomic study is also limited.
The purpose of this study is to investigate the morphological variations and ductal system of the SPG using the analysis of plain axial, sialographic, and 3-dimensional CT images. Based on these observations, we further delineate the anterior lobe (AL) of the parotid gland, thus furnishing vital anatomical insights imperative for procedures—both invasive and non-invasive—within this region.
Methods
This study was approved by the institutional review board of Peking University School and Hospital of Stomatology (PKUSSIRB-201523070). All patients were informed of the details of this study.
Patients
The inclusion criteria were patients who presented with salivary gland-related clinical symptoms at Peking University School and Hospital of Stomatology between 2016 and 2023 and underwent CT sialography examinations of bilateral parotid glands. The exclusion criteria included individuals with specific diseases or treatment histories such as tumors, those who had previously undergone salivary gland surgery or radiotherapy, and cases with poor image quality. Seventy-seven subjects were included in this study. The patients included 46 females and 31 males, aging from 22 to 73 years old with the mean age of 42.9. The CT data of 142 parotid glands were retrospectively analyzed.
CT sialographic examination
All procedures were performed by one experienced doctor using the same standard technique: 1-2 mL of contrast medium (iopamidol, 370 g I/100 mL, Bracco, Italy) was gently introduced into the orifice of the parotid duct. The maxillofacial region was scanned using a 16-row spiral computed tomography scanner (GE Optima, Boston, MA). The following scanning parameters were used: 200-380 mA (automatic exposure control); 120-140 kV; pitch: 1.625; field of view: 20 cm; reconstruction thickness: 1.25 mm. The reconstructed images had been stored in the picture archiving and communication system of our hospital.
CT interpretation and analysis
All CT images were interpreted by 2 experienced radiologists together using GE workstation (AW4.6, Boston, MA) and Mimics 21.0 software (Materialise Technologies, Leuven, Belgium). The relationship between the SPG and the masseter was determined on axial CT image first and then on the sialographic CT (Figure 1). The tributary ductal system of the SPG was reconstructed to observe the 3-dimensional morphology and analyze its relationship with the masseter.
Figure 1.
Line 1 is tangent to the posterior curves of the mandibular ramus and the masseter. Line 2 is tangent to the lateral side of the masseter and parallel to the sagittal plane. Tangent points A and B define a turning curve of the masseter from the lateral to the posterior side. Line 3 represents the posterior border of the masseter. The intersection of Line 3 and the Stensen’s duct (point C) is used as landmark for classification.
On axial CT image (Figure 1A), line 1 (tangent to the posterior border of the ramus and the masseter) and line 2 (tangent to the lateral side of the masseter and parallel to the middle sagittal plane) were drawn. The curved turning area between 2 tangent points A and B was used as landmark to determine whether there was abundant gland tissue lateral or superficial to the masseter. If there is gland tissue anterior to this turning curve, we name it as AL on axial image. On sialographic CT image (Figure 1B), line 3 representing the posterior border of the masseter was used as landmark (Figure 1B). If there was a tributary duct originating anterior to the intersection of line 3 and the Stensen duct, we named it as AL on sialographic image.
The tributary ducts numbers were counted and the anterior-posterior length and height of the AL were quantitatively measured on sialographic CT images. Multi-planar reformatted maximum density projections were generated for measurements.
Statistical analysis
Statistical analysis was performed using SPSS version 26.0 software (SPSS, Inc., Chicago, IL). The chi-square test was used to compare the difference in tributary ducts incidence between groups. The 1-way ANOVA test was used to compare the quantitative measurements of the glands. The significance level was set at P < .05.
Results
Based on plain axial and sialographic CT images, the parotid gland can be divided into 2 parts: the AL that is lateral or superficial to the masseter and the retromandibular parotid gland (RPG) that is posterior to the masseter. The AL gives rise to the different types of the SPG both on plain axial and sialographic CT images. The APG is a variant form of AL in Type II and III SPG.
Classification on plain axial CT images
Based on plain axial CT images, 5 types of SPG were observed (Figure 2). Axial Type A (85/142) presented with abundant gland tissue that existed superficial to the masseter (AL) and was closely attached to the RPG (Figure 2A). Axial Type B (16/142) presented with abundant gland tissue that was superficial to the masseter (AL) and obviously separated from the RPG (Figure 2B). Axial Type C (8/142) appeared with a small lobe gland tissue that was superficial to the masseter (AL) and close to the anterior border of the masseter (Figure 2C). Axial Type D (30/142) showed that there was no gland tissue superficial to the masseter and only the RPG could be observed (Figure 2D). Axial Type E (3/142) appeared that most of the RPG was absent (Figure 2E).
Figure 2.
Classifications of the SPG on plain axial images. (A) Type A shows that abundant gland tissue (arrow) exists lateral to the masseter and is not separated from the RPG. (B) Type B appears with abundant gland tissue (arrow) lateral to the masseter and separated from the RPG. An obvious gap is observed anterior to the RPG (arrow head). (C) Type C appears with a small lobe gland tissue (arrow) lateral to the masseter. (D) Type D shows that no gland tissue lateral to the masseter is observed. (E) Type E shows that most of the RPG is absent.
Classification on sialographic CT images
Four major types (I-IV) and 2 minor types (V-VI) of AL and SPG could be observed on the sialographic CT and 3-dimensional images. The sialographic Type I SPG (83/142, 58.5%) showed the existence of abundant gland tissue along the Stensen duct with numerous tributary ducts (Figure 3). Type I AL located superficial to the masseter and was not separated from the RPG.
Figure 3.
Type I AL and SPG on sialographic and 3-dimensional CT images. (A-C) Type Ia (A), Type Ib (B), and Type Ic (C) SPG shows abundant ALs (arrows) with numerous tributary ducts. (D) Three-dimensional image shows a large AL (arrow) is combined to the RPG.
The sialographic Type II SPG (16/142, 11.3%) had a large AL with one or more tributary ducts (Figure 4A and C). The AL located superficial to the masseter. Type II AL showed significantly larger size and more identifiable tributary ducts in comparison with those of Type III (P < .05).
Figure 4.
Type II (A, C) and Type III (B, D) AL and SPG on sialographic and 3-dimensional CT images. (A, B) Type II and III SPG show that the ALs (arrows) are detached from the RPG. C-D: Three-dimensional views of Type II (C) and Type III (D) show that the ALs (arrows) are separated from the RPG.
The sialographic Type III SPG (12/142, 8.5%) showed a small and isolated AL above the Stensen duct and lateral to the anterior border of the masseter (Figure 4B and D). No identifiable tributary duct could be observed. Except for the AL, no other gland tissue could be observed along the Stensen duct anterior to the RPG in Type III. Occurrence of Type III AL in Type I or II SPG was not observed in any individual parotid gland.
The sialographic Type IV SPG (28/142, 19.7%) showed no AL or other gland tissue anterior to the RPG (Figure 5A and B). Type IV SPG located posterior to the masseter and no gland tissue could be observed superficial to the masseter.
Figure 5.
Type IV AL and SPG on sialographic and 3-dimensional CT images. (A) Sialographic view shows that no tributary duct originates anterior to the posterior border of the masseter. (B) Three-dimensional view shows the absence of the AL.
The sialographic Type V SPG (3/142, 2.1%) showed that the RPG was absent and only the AL could be observed (Figure 6A). The sialographic Type VI SPG (3/142, 2.1%) showed that a small lobe of gland tissue was observed beneath the Stensen duct and anterior to the RPG (Figure 6B). Type VI SPG was observed in 1 Type IV and 2 Type II parotid glands.
Figure 6.
Type V and Type VI SPG on sialographic images. Type V SPG shows the absence of most of the gland tissue posterior to the masseter (RPG). Type VI SPG shows a small lobe of gland tissue (arrow) beneath the Stensen’s duct.
The classification on plain axial images (A-E) correlated with that on the sialographic images (I-V) positively (R = 0.913, P = .000).
Quantitative measurements of AL
The means and standard deviations of anterior-posterior lengths of the AL in sialographic Type I, II, and III groups were 23.5 ± 9.9 (mm), 24.4 ± 4.73 (mm), and 13.9 ± 8.6 (mm), respectively. The means and standard deviations of vertical heights of the AL in sialographic Type I, II, and III groups were 21.0 ± 6.2 (mm), 20.7 ± 6.2 (mm), and 12.0 ± 5.5 (mm), respectively. No significant statistical difference was observed between Type I and II in the anterior-posterior lengths (P = .728) and heights (P = .846). The AL of Type III showed significantly reduced lengths (P = .002) and heights (P = .000) in comparison with the AL of Type I and II.
The sialographic Type I ALs were composed of 1 (23/83), 2 (47/83), 3 (12/83) or 4 (1/83) tributary ducts (Table 1). The first tributary duct of Type I AL originated from the anterior third (Type Ia, 33/83), middle third (Type Ib, 34/83), and posterior third (Type Ic, 16/83) of the Stensen ducts (Figure 3A and C). The lengths and heights of the AL decreased as the first originating sites backwardly moved along the Stensen duct (P = .000). There was an obvious trend that the first originating site from the Stensen duct moved backwardly as the tributary ducts number decreased.
Table 1.
Tributary duct numbers and quantitative measurements of the anterior lobe.
| Tributary duct number |
||||||
|---|---|---|---|---|---|---|
| One | Two | Three | Four | Length (mm) | Height (mm) | |
| Type Ia (n = 33) | 6 | 18 | 8 | 1 | 31.4 ± 8.8 | 22.8 ± 6.2 |
| Type Ib (n = 34) | 5 | 27 | 2 | 0 | 19.5 ± 4.9 | 19.5 ± 4.9 |
| Type Ic (n = 16) | 12 | 2 | 2 | 0 | 12.3 ± 3.5 | 18.3 ± 6.4 |
| Type II (n = 16) | 14 | 2 | 0 | 0 | 25.4 ± 5.3 | 21.1 ± 6.1 |
| Type III (n = 12) | 0 | 0 | 0 | 0 | 19.7 ± 5.1 | 16.1 ± 7.9 |
| Total (n = 111) | 37 | 49 | 12 | 1 | 22.9 ± 9.0 | 20.2 ± 6.3 |
The sialographic Type II ALs were composed of 1 (14/16) or 2 (2/16) tributary ducts (Table 1). The first tributary duct of Type II originated from the anterior third (10/16) or middle third (6/16) of the Stensen ducts.
Discussion
Anatomy and imaging of the parotid gland are important bases for clinical practice, especially surgery.4,6,7 The anatomy knowledge of the parotid gland derives from cadaver, surgical and radiologic studies.1,5,8,9 Cadaver studies provide the fundamental knowledge of anatomy.5,10 But due to the limitation in the number of specimens, it is difficult to realize a cadaveric study with large sample size. Surgery also plays an important role in revealing more knowledge of anatomy. However, because the surgery of parotid gland is evolving towards more and more conservative protocols of partial parotidectomy and extracapsular dissection, chances become less for dissection of the whole parotid gland during surgery. Imaging studies have the advantages in obtaining large samples and discovering holistic anatomical characteristics. Sialography is far more than an image.11 Sialographic study has the advantage in revealing the ductal system accurately both on axial images and 3-dimensional projections.12,13 CT or MRI sialography also creates a non-distorted images with the high spatial resolution.14
This study improves the understandings of the morphological variations of the SPG and its appearances on the cross-sectional imaging of CT. The anterior edge of the SPG may diminish around the posterior border of the masseter or extend to the anterior edge of the masseter, which gives rise to the highly variable shape of the anterior extending part of the SPG. Reportedly, about half of the SPG border outlines are roughly triangular, while about 30% of the SPG is more or less even throughout the anterior-posterior width if the gland extends forwards to the anterior border of the masseter.15 This anatomic variation is due to the AL of the SPG. The AL can be combined with the posterior RPG to form a whole entity or be an isolated lobe separated from the RPG. When the AL is florid and combined to the RPG, the SPG could extend to the anterior border of the masseter (Figure 3A).
There are 2 pathways for the degradation of the AL, as shown in Figure 7. The posterior pathway is that the AL becomes gradually smaller and smaller (Type Ib to Type Ic, Figure 7A-C and F), but keeps combined with the RPG until it disappears. The anterior pathway is that the AL becomes detached from the RPG first (Type II, Figure 7D) and then degrades to small size (Type III, Figure 7E) before it disappears. Type II and Type III AL obviously correspond to the well-known APG. This degradation hypothesis of AL and SPG also explains why the APG manifested many variations.
Figure 7.
Illustration of the evolution of the AL. The AL (arrows) could gradually reduce its size from Type Ia (A) to Type Ib (B), Type Ic (C), and then Type IV (F). The AL also could detach from the RPG first (Type II, D) and gradually reduce its size (Type III, E) to give rise to Type IV (F).
The concept of AL is supported by the terms “anterior process” or “facial lobe” of the parotid gland, which have been described in literatures.16 The anterior process of parotid gland refers to lobes protruding to the front edge of the masseter muscle according to Toh.16 We named it “AL” because it is observed that the AL derived from one or more inherent tributary ducts from the Stensen duct observed on sialography. The AL is defined according to the following characteristics. First, the AL refers to the gland tissue mainly lying superficial to the masseter that may extend from the anterior border of the masseter to the posterior edge. Secondly, the AL is named in contrast with the RPG that locates posterior to the masseter and ramus. Thirdly, from the sialographic views the AL has its own ductal system from the Stensen duct. Finally, 4 major types of AL give more substantial explanation for the anatomic variation of the SPG and APG. The AL varies from florid to none. The AL can be non-separable or visually detached from the RPG.
The use of AL deepens the understanding of the APG, which frequently leads to the parotid diseases. Tumors of APG are considered in the differential diagnosis of a mid-cheek mass.17–19 Incomplete excision of the APG potentially gives rise to recurrence. APG has been generally described as a mall glandular tissue lying above the Stensen duct on the anterior portion of the masseter, that is distinctly separate from the main parotid gland.20 APG is also highly variable in anatomy.20 APGs are usually round, ellipsoidal, triangular, or indeterminate in shape. Some small APGs may be rice-like.16 An essential criterion for the definition of APG is its complete detachment from the main body of the gland. According to this definition, the incidence ranges from 21% to 69%.5,16
Our study indicates that the AL gives rise to the APG. First, the AL in Type II and Type III SPG (19.7%) conforms to the diagnostic definition of the APG. Type II AL is larger and closer to the RPG. Type III AL is smaller and closer to the anterior edge of the masseter. These findings account for the morphological variations of the APG. Secondly, the AL in Type I SPG is closely associated with the RPG, but its anatomic position is equivalent to that of the APG. We hypothesize that, from the biological and evolutionary perspective, Type I AL of SPG is actually the prototype of the APG (Figure 7), which is just not separated from the RPG. Finally, the APG is constantly absent when the AL disappears in Type IV SPG, which also reinforces our hypothesis.
There are some shortcomings in the present study. First, this study focused on the AL of the SPG and its ductal system. The whole ductal system of the RPG including the deep lobe and the tip lobe was not involved in the present study. Secondly, as CT radiological study could not show the peripheral branches of the facial nerves, we could not prove the relationship between the facial nerve and the AL. Thirdly, we are unable to prove the existence of connective tissue septa between AL and RPG in Type I parotid gland. We do not think these connective tissue septa are necessarily present, because when the AL is not separated from the RPG, they will all be enclosed by the dense superficial fascia of the parotid gland. This also means that the AL may not be easily dissected from RPG in Type I SPG. At last, the sample size is relatively restricted to reveal more detailed features of SPG.
Conclusions
There are 4 major types and 2 minor types of the AL in the SPG. The ALs lie superficial to the masseter and give rise to anatomic variations of the SPG. The APG could be considered as a variant form of the AL.
Contributor Information
Junru Zhao, Department of Oral and Maxillofacial Radiology, Peking University School and Hospital of Stomatology, Beijing 100081, China; National Engineering Laboratory for Digital and Material Technology of Stomatology & Beijing Key Laboratory of Digital Stomatology, Beijing 100081, China.
Chang Han, Department of Oral and Maxillofacial Radiology, Peking University School and Hospital of Stomatology, Beijing 100081, China; National Engineering Laboratory for Digital and Material Technology of Stomatology & Beijing Key Laboratory of Digital Stomatology, Beijing 100081, China.
Minghu Huang, Department of Oral and Maxillofacial Radiology, Peking University School and Hospital of Stomatology, Beijing 100081, China; National Engineering Laboratory for Digital and Material Technology of Stomatology & Beijing Key Laboratory of Digital Stomatology, Beijing 100081, China.
Lisha Sun, National Engineering Laboratory for Digital and Material Technology of Stomatology & Beijing Key Laboratory of Digital Stomatology, Beijing 100081, China; Key Laboratory of Oral Pathology, Peking University School and Hospital of Stomatology, Beijing 100081, China.
Gang Li, Department of Oral and Maxillofacial Radiology, Peking University School and Hospital of Stomatology, Beijing 100081, China; National Engineering Laboratory for Digital and Material Technology of Stomatology & Beijing Key Laboratory of Digital Stomatology, Beijing 100081, China.
Zhipeng Sun, Department of Oral and Maxillofacial Radiology, Peking University School and Hospital of Stomatology, Beijing 100081, China; National Engineering Laboratory for Digital and Material Technology of Stomatology & Beijing Key Laboratory of Digital Stomatology, Beijing 100081, China.
Author contributions
J. Zhao and C. Han contributed equally to this work.
Funding
Program for New Clinical Techniques and Therapies of Peking University School and Hospital of Stomatology (PKUSSNCT-23A13). National Clinical Key Discipline Construction Project (PKUSSNKP-202111).
Conflicts of interest
No potential sources of conflict of interest exist. There is no interest or relationship, financial or otherwise that might be perceived as influencing an author's objectivity. There is no patent or stock ownership, membership of a company board of directors, membership of an advisory board or committee for a company, and consultancy for or receipt of speaker’s fees from a company.
References
- 1. Mehta V, Nathan CA. Extracapsular dissection versus superficial parotidectomy for benign parotid tumors. Laryngoscope. 2015;125(5):1039-1040. [DOI] [PubMed] [Google Scholar]
- 2. Stathopoulos P, Igoumenakis D, Smith WP. Partial superficial, superficial, and total parotidectomy in the management of benign parotid gland tumors: a 10-year prospective study of 205 patients. J Oral Maxillofac Surg. 2018;76(2):455-459. [DOI] [PubMed] [Google Scholar]
- 3. 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(34):e1237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Park HJ, Hong SO, Kim HM, Oh W, Kim HJ. Positional deformation of the parotid gland: Application to minimally invasive procedures. Clin Anat. 2022;35(8):1147-1151. [DOI] [PubMed] [Google Scholar]
- 5. Frommer J. The human accessory parotid gland: its incidence, nature, and significance. Oral Surg Oral Med Oral Pathol. 1977;43(5):671-676. [DOI] [PubMed] [Google Scholar]
- 6. Yousem DM, Kraut MA, Chalian AA. Major salivary gland imaging. Radiology. 2000;216(1):19-29. [DOI] [PubMed] [Google Scholar]
- 7. Mcgeary R, Rhyner PA, Desai AB, Bhatt AA. Navigating the parotid glands: anatomy, imaging work-up and next steps. Clin Neuroradiol. 2022;32(3):615-623. [DOI] [PubMed] [Google Scholar]
- 8. Zhu W, Hu F, Liu X, Guo S, Tao Q. Role of the accessory parotid gland in the etiology of parotitis: statistical analysis of sialographic features. PLoS One. 2016;11(2):e0150212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Uzmansel D, Elvan O, Aktekin M. Precise localization of parotid duct: a cadaveric study. Anat Sci Int. 2022;97(1):59-64. [DOI] [PubMed] [Google Scholar]
- 10. Kochhar A, Larian B, Azizzadeh B. Facial nerve and parotid gland anatomy. Otolaryngol Clin North Am. 2016;49(2):273-284. [DOI] [PubMed] [Google Scholar]
- 11. Huston M, Collard B, Blanco-Guzman M. Sialography—more than an image. Clin Otolaryngol. 2015;40(6):740-741. [DOI] [PubMed] [Google Scholar]
- 12. Tanaka T, Ono K, Habu M, et al. Functional evaluations of the parotid and submandibular glands using dynamic magnetic resonance sialography. Dentomaxillofac Radiol. 2007;36(4):218-223. [DOI] [PubMed] [Google Scholar]
- 13. Canzi P, Capaccio P, Marconi S, et al. Feasibility of 3D printed salivary duct models for sialendoscopic skills training: preliminary report. Eur Arch Otorhinolaryngol. 2020;277(3):909-915. [DOI] [PubMed] [Google Scholar]
- 14. Smith JR, King WW, Tang WY, Metreweli C. Differentiating tumours of the deep and superficial lobes of the parotid gland by computed tomographic sialography. Clin Radiol. 1987;38(4):345-349. [DOI] [PubMed] [Google Scholar]
- 15. Standring S. Gray's Anatomy. The Anatomical Basis of Clinical Practice. Churchill Livingstone Elsevier; 2008:495-496. [Google Scholar]
- 16. Toh H, Kodama J, Fukuda J, Rittman B, Mackenzie I. Incidence and histology of human accessory parotid glands. Anat Rec. 1993;236(3):586-590. [DOI] [PubMed] [Google Scholar]
- 17. Choi HJ, Lee YM, Kim JH, Tark MS, Lee JH. Wide excision of accessory parotid gland with anterior approach. J Craniofac Surg. 2012;23(1):165-168. [DOI] [PubMed] [Google Scholar]
- 18. Klotz DA, Coniglio JU. Prudent management of the mid-cheek mass: revisiting the accessory parotid gland tumor. Laryngoscope. 2000;110(10 Pt 1):1627-1632. [DOI] [PubMed] [Google Scholar]
- 19. Ma H, Jin S, Du Z, Wang L, Zhang Z, Wang Y. Pathology and management of masses in the accessory parotid gland region: 24-year experience at a single institution. J Craniomaxillofac Surg. 2018;46(2):183-189. [DOI] [PubMed] [Google Scholar]
- 20. Batsakis JG. Accessory parotid gland. Ann Otol Rhinol Laryngol. 1988;97(4 Pt 1):434-435. [DOI] [PubMed] [Google Scholar]