Parotid Gland Tumors Clinicopathologic
Study
Birendra K. Sinha, MD, FRCS (C), FACS,’ Winnipeg, Manitoba, Canada David W. Buntine, MBBS, FRCP (C), FRCPA, Winnipeg, Manitoba, Canada
Tumors of the parotid gland continue to be a surgical challenge due to their various degrees of malignancy and close proximity with the facial nerve. Although approximately 80 per cent of these tumors are considered benign [I-3], earlier reports indicated a high recurrence rate after limited excision [3-51. This has led to a more aggressive approach to the treatment of these lesions in recent years, in the form of either subtotal (superficial or deep lobectomy) or total conservative parotidectomy, with excellent results [1,6-81. At our Centre, however, the surgical treatment of benign tumors of the parotid gland has generally been less aggressive, being based on complete extirpation of the tumor with an adequate margin of normal glandular tissue and preservation of the facial nerve. Subtotal and total parotidectomy, although used in certain cases of benign tumors, were performed mainly for malignant lesions. This study was undertaken in an attempt to answer the following questions about the surgical treatment of parotid gland tumors: (1) whether subtotal or total parotidectomy should always be performed for control of benign tumors; (2) whether a more radical procedure is associated with a greater risk of facial nerve injury despite careful attempts to identify and dissect out the nerve; and (3) whether recurrence after a conservative procedure is controllable by further operation without undue risk to the patient. Material and Methods The clinical findings
and histologic features
were re-
viewed for 154 parotid gland tumors treated at the Health Sciences Centre, University of Manitoba, Winnipeg, from 1957 to 1970. The description of treatment From the Department of Surgery, Division of Oncology and Pathology, The University of Manitoba, and Health Sciences Centre, Winnipeg, Manitoba, Canada. ’ Present address and address for reprint requests: Elk Grove MedicalDental Park. 758 Arlington Heights Road, Elk Grove Village, Illinois60007.
Volume
129, June 1975
and follow-up data were obtained from the records of the Health Sciences Centre, the Manitoba Cancer Treatment and Research Foundation, and from respective attending surgeons. Pathology reports and original slides were available for review in all cases. The classification of the lesions was modified from that of Foote and Frazell [3]. The tumors were divided into two main groups, benign and malignant. There were one hundred twenty-two benign lesions (79 per cent of the entire series) and thirty-two malignant. Of the benign tumors, eighty-seven or 71 per cent were mixed tumors, including seventy-nine primary and eight recurrent mixed tumors. Warthin’s tumor was the second most common lesion, comprising twenty-six or 21 per cent of the total benign tumors, a relatively higher incidence than that in most reported series [1,3,6]. Other less frequently occurring benign tumors included five benign lymphoepithelial lesions (Godwin), one oncocytoma, two vascular tumors, and one parotid cyst. The malignant tumors included ten mucoepidermoid carcinomas, ten adenoid cystic carcinomas, three acinic cell carcinomas, five adenocarcinomas not otherwise specified, three anaplastic carcinomas in Eskimos, and one malignant mixed tumor. Benign Lesions. Primary mixed tumors: There were thirty males and forty-nine females in this group. The youngest patient was 13 and the oldest 81 years of age, with average age of 47.5 years. (Figure 1.) An asymptomatic parotid lump was the commonest presenting feature. Pain or tenderness was present in only six patients. None had facial palsy. The size of the lesion ranged from 1 to 8 cm, the majority being between 2 and 3 cm. Recent increase in size of the lesion was recorded in twenty-one patients. The shortest duration of the presence of lesion was two weeks and the longest thirty-five years, approximately 50 per cent being present for one to five years prior to diagnosis. Morphologically these tumors were pleomorphic with either epithelial or mesodermal elements predominating. (Figure 2.) It was not possible to correlate differences in clinical behavior with the histologic features of the benign mixed tumor. Recurrent mixed tumors: A total of eight patients had recurrent mixed tumors of the parotid gland, four males and four females. Two patients had a history of three
675
Sinha and Buntine
0 MIxed ,umor
n
Warthtn s
0 Molqnonl
M-30.F=49
M=l5.F=9 Tumors M=l5.F=17
DECADES
Figure 1. Age distribution of benlgn mixed tumor, Warthin’s tumor, and malignant tumors of the parotid gland.
of cyclic spaces lined with a double layer of pink epitheha1 cells (oncocytes) and varying amounts of lymphoid tissue in the surrounding stroma. Benign lymphoepithelial lesions (Godwin): There were five patients with this lesion, one male and four females. The youngest patient was 37 and the oldest 80 years of age. The size of the lesion ranged from 1 to 3 cm and duration from three to eight months. These patients presented with painless parotid lumps and underwent surgery mainly because the diagnosis of benign or malignant neoplasms could not be excluded. The microscopic picture consisted of loss of acinar tissue, ductal and myoepithelial cell proliferation with foci of lymphoid infiltration. Others: The two vascular lesions were a cavernous hemangioma and arteriovenous malformation. The other two were oncocytoma and benign epithelial cyst of parotid gland origin
previous operations over a period of forty years, including one who also had received radiation therapy after the initial local excision. One patient had a history of two previous excisions and the remaining five only one. The youngest patient was 17 years and the oldest 78 years of age, with an average age of 52.5 years. The size of the recurrent tumor was between 1 and 3 cm and the duration ranged from two months to eight years. Recent change in size was noted in two patients and pain was recorded in one. None had facial palsy. Warthin’s tumor: Twenty-four patients presented with Warthin’s tumor, two of whom had bilateral lesions. There were fifteen males and nine females. The youngest patient was 46 years and the oldest 79 years of age, with an average age of 60 years. (Figure 1.) The size of the lesion ranged from 1.5 to 6 cm and the duration from four weeks to twelve years. Recent increase in size was noted in nine patients. Two patients presented with pain and the rest were asymptomatic. The lesion generally was located in the tail of the parotid gland and was mobile and soft. Six of these were cystic. Microscopically this lesion had a characteristic appearance consisting
Malignant Tumors. Thirty-two patients with primary malignant tumors ranged in age from 17 to 83 years, with an average age of 48 years. (Figure 1.) There were fifteen males and seventeen females. The size of the lesion ranged from 1 to 8 cm and the duration from five months to six years. Fifteen patients presented with a painless lump in the parotid region. Pain and tenderness were present in another fifteen patients, fixation of the tumor to the underlying structures in three patients, and facial palsy in two patients. In contrast, of the onehundred twenty-two patients with benign lesions, onehundred fourteen presented with an asymptomatic lump, eight with pain and tenderness, and none with facial palsy. Mucoepidermoid carcinoma: There were seven males and three females in this group, their ages ranging from 17 to 76 years. The size of the lesion varied from 2 to 8 cm and the duration from five months to one year. Microscopically, these tumors showed various proportions
Figure 2. Photomicrograph (high power) of a benign mixed tumor showing markedly pieomorphic ceils in hyaline-like stroma.
Figure 3. Photomicrograph (high power) of an infiitrating ductai carcinoma with dense fibrous stroma.
676
The American
Journal
of Surgery
Parotid
of both squamous and mucus-secreting cells. The histologic gradings appeared to correlate well with the biologic behavior of the tumor. Of the seven tumors regarded as low grade, none showed evidence of metastasis or recurrence. Of the three regarded as high grade, one metastasized t.o the regional lymph node and another involved the facial nerve causing facial paralysis. Adenoid cystic carcinoma: There were five males and five females from 23 to 83 years of age. These tumors were generally slow growing. The size of the lesion varied from 1 to 4.5 cm and the duration from one to eight years. Grossly, the tumor appeared well circumscribed but nonencapsulated, and microscopically it was corn.posed of epithelial and myoepithelial cells surrounding cores of hyaline of various sizes giving rise to the distinc tive Swiss cheese appearance. The characteristic ability of this tumor to invade the perineural space was seen in one patient. However, no lesion had metastasized to the regionai lymph nodes. Acinic cell carcinoma: This was diagnosed in three patients, two males and one female, between 25 and 48 years of age. The size of the lesion varied from 2 to 3 cm and the duration from five months to five years. These tumors were poorly circumscribed and microscopicall> composed of acinar cells showing basophilic granular cy., toplasm. Adenocarcinoma (not otherwise specified): This unusual group of lesions presented a microscopic appearance of ductal carcinoma, often with a dense fibrous stroma (Figure 3), and showed a great propensity to metastasize to distant organs. This was diagnosed in five patients, four males and one female, their ages ranging from 28 to 76 years. The lesions were 2 to 3 cm in size and generally of short duration, less than a year. Anaplastic carcinoma in Eskimos: All three patients were female, 17, 31, and 45 years of age. These tumors varied from 3 to 7 cm in size and were irregular and fixed to the underlying structures. One patient had pain and tenderness as well as involvement of the facial nerve. Grossly, the lesion was locally invasive and in two instances had metastasized to regional lymph nodes. Microscopically, the tumor was composed of groups of anaplastic epithelial cells in a stroma heavily infiltrated with lymphoid cells. (Figure 4.) This lesion seems to be a malignant variant of the benign lymphoepithelial lesion of the parotid gland and appears peculiar to the Alaskan Indians and the Canadian Eskimos as reported previously [9,10]. Malignant mixed tumors: Only one patient, a 45 year old male, had a malignant mixed tumor of the parotid gland according to the criteria used at this Centre. This particular lesion was 8 cm in diameter and had been present for ten years withL a recent increase in size. As mentioned earlier, the degree of pleomorphism and cellularity did not correlate with the biologic behavior of mixed tumors. The only reliable criteria for diagnosis of malignant mixed tumor in our experience is vascular invasion, although a high mitotic index must be considered suspicious.
Volume 129, June 1975
Gland Tumors
Figure 4. Photomicrograph (high power) of an Eskimo tumor showing anaplastic epithelial cells in a stroma heavily inffftrated with lymphocytes.
Diagnosis, Treatment, Complications, and Results Diagnosis was based mainly on the clinical evaluation of the patient. Plain x-ray st,udy and sialography were performed in only a few cases, which aided in the recognition of non-neoplastic disorders but did not contribute significantly to the distinction between benign and malignant neoplasms. In most instances the diagnosis was confirmed after wide local excision. When malignancy was suspected, an incisional biopsy was performed on operative exposure of the parotid gland prior to the definitive procedure. More recently, the disposable Vim-Silverman needle has been used to obtain an adequate core of tissue using local anesthesia for histologic diagnosis. The procedure is relatively simple, reliable, and in our experience, free of the commonly feared risk of tumor spillage along the needle track. Tables I and II present the treatment of benign and malignant tumors of the parotid gland. Patients undergoing enucleation, limited local exci-
TABLE ____
I -..
Type of Tumor ___ Mixed primary Mixed recurrent Warthin’s Godwin’s Others Total _~~ ~_ -
Treatment Gland
of Benign Tumors
of the Parotid
Local Excision
Subtotal Parotidectomy
Total Parotidectomy
Postoperative Radiation
79
61
18
0
2
8 26 5 4 122
7 15 2 2 87
1 10 3 2 34
0 1 0 0 1
0 0 1 0 3
Number
677
Sinha
TABLE
and Buntine
II
Treatment
of Malignant
Type of Tumor Mucoepidermoid Adenoid cystic Acinic cell Adenocarcinoma specified) Anaplastic Mixed Total
(not
Tumors of the Parotid Gland
Number
Local Excision
10 10 3
1 6 2
5 3 1 32
2 0 0 11
(Eskimo)
III
Postoperative Complications of Parotid Gland Surgery for Benign and Malignant Tumors Complication
Benign*
Malignant?
8 11 9 1 2
... 1 3 ... 1
1 1
... 1
Facial paralysis Upper trunk Lower trunk Total paralysis Salivary fistula Auriculotemporal syndrome Wound complications Seroma Hematoma and infection * Facial nerve was sacrificed benign tumor. t Facial nerve was sacrificed of malignant tumors.
TABLE
IV
intentionally intentionally
in one
case
in all four
676
of
cases
Incidence of Postoperative Facial Paralysis Related to Extent of Surgery for Benign Tumors with and without Exposure of the Nerve Incidence of Facial Paralysis
Local excision Subtotal parotidectomy Total parotidectomy
6 3 1
1 0 0
2 0 0
7 0 0
2 1 0
3 2 1 16
0 1 0 2
0 2 0 4
4 3 0 14
0 0 0 3
Total
Postoperative Radiation
Biopsy and Radiation
otherwise
Son, or wide local excision as indicated in the operative report are all grouped as having had local excision. Those undergoing excision of either the entire superficial lobe or the deep lobe (in only one instance) with or without excision of the isthmus are classified as having had subtotal parotidectomy, and removal of the entire gland is synonymous with total parotidectomy. For benign lesions, TABLE
Parotidectomy
Radical Neck Dissection
Subtotal Parotidectomy
Total Number
Cases of Facial Paralysis
Nerve Exposed
Nerve Not Exposed
87
20
7150
13137
34
7
7134
1
1
l/l
eighty-seven local excisions, thirty-four subtotal parotidectomies, and one total parotidectomy were carried out. For malignant lesions, eleven local excisions (six of these for adenoid cystic carcinoma) and sixteen subtotal parotidectomies were performed. Two patients, one with mucoepidermoid carcinoma and the other with anaplastic carcinoma (an Eskimo), underwent total parotidectomy. In four patients, two with mucoepidermoid and two Eskimos with anaplastic carcinoma, either subtotal or total parotidectomy was combined with radical neck dissection. Fourteen patients received postoperative irradiation, including seven patients with mucoepidermoid carcinoma, four with adenocarcinoma (not otherwise specified), and three with anaplastic carcinoma. Two patients with mucoepidermoid carcinoma and one with adenoid cystic carcinoma had only palliative radiotherapy after biopsy. Table III presents the incidence of postoperative complications in benign and malignant neoplasms. The most frequent complication was facial nerve injury, either partial or total, which occurred in twenty-eight patients with benign lesions and four patients with malignant lesions. In the latter cases it occurred as a result of intentional sacrifice of the nerve located close to the tumor. However, in the case of benign tumors sacrifice of the facial nerve was intentional in one case only; the remaining instances were due to inadvertent injury of the facial nerve during operation. The incidence of postoperative facial paralysis in these cases was further analyzed relative to the extent of surgery and whether or not an attempt had been made to preserve the facial nerve. (Table IV.) The findings suggest an apparent correlation between the extent of surgery and postoperative facial paralysis even when an effort was made to preserve the nerve. In the group with local excision, however, the incidence appeared to be higher when blind
The American Journal of Surgery
Parotid
excision was carried out than when the facial nerve was identified and exposed. All patients except three with benign tumors were followed up three to twelve years. (Table V.) Of the seventy-nine primary mixed tumors, there were ten recurrences in seven patients, three patients having two recurrences each three months to nine years after the initial procedure. All patients with recurrence had undergone local excision. In this group histologic slides and pathology reports were reviewed with regard to the completeness of excision. Of sixty-one local excisions, twenty-nine were considered histologically complete and among these there were no recurrences. Thirty-two either were considered incomplete or the tumor extended to the line of excision and all ten recurrences were noted in this group, correlating well with the histologic findings. However, all recurrences were controlled by further local excision except in one patient who required total parotidectomy because of multifocal recurrence only three months after the initial operation. There were two recurrences in one patient in the group with recurrent mixed tumor that were controlled by repeated local excision. In the remaining patients with benign lesions, there was only one recurrence in a patient with Warthin’s tumor, controlled by local excision. There were no deaths attributable to benign tumors. Table VI presents the survival data of patients with malignant tumors during a follow-up period of three to twelve years. A total of fifteen patients or approximately 50 per cent are alive and free of disease, two are alive with disease, and fifteen are dead, including eight who died from the disease. Of ten paCents with mucoepidermoid carcinoma? four are alive free of disease, one is alive with disease, three are dead from disease, and two dead from other causes. Of ten patients with adenoid cystic carcinoma, six are alive with no evidence of disease, one is dead from disease, and three are dead from other causes. Of three patients with acinit cell carcinoma, one is alive free of disease, one is alive with disease, and one is dead from other causes. Of five patients with adenocarcinoma (not otherwise specified), two are alive free of disease and three are dead, including two from disease. One Eskimo patient with anaplastic carcinoma is alive free of disease, and the remaining two are dead from disease. Two patients who presented with involvement of the facial nerve by malignant tumor died within three years of diagnosis. There was no difference in the results of surgery and postoperative radiation as compared with surgery alone. Volume
129, June 1975
Tumors
Results of Treatment of 122 Benign Parotid Gland Tumors in 120 Patients during a Follow-Up Period of 3 to 12 Years
TABLE V
Type of Tumor
Gland
Dead from Other Causes+
Lost to FollowUp
68
9
2
4
3
1
19 3 4
5 2 0
0 0 0
Total Number of Number of Pabents Recurrences ,“ir * ~ ..~ _____ _~.
Mixed primary Mixed recurrent Warthin’s Godwin’s Others
10 in 7 pts
79
2 in 1 pt
8 24 5 4
1 0 0
Study
_~~
* NED = no evidence of disease. t None died from disease.
This series illustrates the wide variation in the natural history and degree of malignancy of parotid tumors, confirming previous reports [2,6-81. A benign tumor may be very slow growing or may remain unchanged for many years. On the other end of the spectrum, anaplastic carcinoma (Figure 5) may grow rapidly and metastasize rapidly causing death within a short period of time. There is unanimity of opinion that benign tumors must be treated surgically since they can attain an enormous size and cause serious disability (Figure 6) or local complications, including ulceration, bleeding and infection (Figure 7), and occasionally death, as reported previously [6,8]. Previous authors have reported the great propensity of benign mixed tumors to recur locally after surgical removal 13~51.
TABLE
VI
Type of Tumor
Results of Treatment of 32 Malignant Parotid Gland Tumors during a Follow-Up Period of 3 to 12 Years
Total Number of Patients
Alive NED*
Alive with Disease
Dead from Disease
Dead from Other Causes
Mucoepidermoid Adenoid cystic Acinic cell Adenocarcinoma (not otherwise specified) Anaplastic (Eskimo) Mixed Total
10 10 3
4 6 1
1 0 1
3 1 0
2t 3 1
5
2
0
2
1
3 1 32
1 1 15
0 0 2
2 0 8
0 0 7
* NED = no evidence of disease. t One patient had residual disease
at the time
of death.
679
Sinha and Buntine
Figure 6. A benign mixed tumor grown to an enormous size (incidental finding in a patient admitted with myocardiai infarction, not included in this series). Figure 5. Anapiastic carcinoma of the parotid gland in an Eskimo patient.
Figure 7. Warthin’s tumor showing ulceration and hemorrhage.
These reports also indicate that local recurrence results from inadequate surgical excision, which led to the more aggressive approach of subtotal or conservative total parotidectomy, with no recurrence or very low recurrence rates [1,6-81. This approach is supported by the findings of the present study, which show that all recurrences of primary
680
benign mixed tumors followed local excisions and none followed subtotal or total parotidectomy. However, in this study all local recurrences were noted in the group with incomplete local excision and none in those with complete excision as verified histologically. Furthermore, the risk of facial nerve injury increased with the extent of the surgical procedure, in contrast to previous reports [11,12]. The incidence of facial paralysis was.highest with total parotidectomy and lowest with local excision, although in the latter group facial nerve injury generally occurred when exposure and dissection of the nerve were not carried out. Thus, wide local excision with an adequate margin of normal glandular tissue verified by frozen section is the logical treatment for most benign tumors of the parotid gland except when subtotal or conservative total parotidectomy is required because of the size or location of the tumor. The facial nerve should be identified and preserved. This approach would prevent recurrence and minimize facial nerve injury as indicated by this study. In benign lesions, intentional sacrifice of the facial nerve is rarely justifiable since actual invasion or fixation of the nerve by benign tumors generally is absent, as indicated by the fact that none of the patients with benign tumor presented with facial paralysis. In our experience, careful dissection can permit removal of benign tumor lying close to the nerve without neural injury. If the facial nerve is
The American Journal of Surgery
Parotid Gland Tumors
divided during surgery it should be repaired immediately by end to end anastomosis or interposition of nerve graft. Analyzing the recurrences, we find them generally controllable by further surgery without facial nerve injury and without risk to life, contrary to other reports [1,6,12]. Total removal of the gland, based on the multicentricity of these tumors and possibility of malignant, transformation as reported previously [3,13], is not supported by the findings of this study, since in only one instance was the recurrence multifocal as confirmed by histologic study and required total parotidectomy. None of the recurrences underwent malignant changes. For malignant tumors, a more radical procedure in the form of subtotal or total parotidectomy is advisable. In general agreement with previous reports [1,6-81, radical neck dissection was performed only in presence of clinically involved regional lymph nodes except in cases of anaplastic carcinoma in Eskimo patients. Intentional sacrifice of part or all of the facial nerve may be necessary for adequate treatment of malignant lesions on a selective basis depending on the type and anatomic location of the tumor. Regional lymph node metastases and involvement of the facial nerve usually suggest a poor prognosis, confirming other reports [ 1,2]. The role of radiation therapy in the management of these tumors is uncertain, except for palliation in cases of advanced tumors. This modality was used in too few cases to evaluate its role in improving the survival data. A program of combined operation and either preoperative or postoperative irradiation for anaplastic carcinoma’ in Eskimo patients is currently underway and will be reported separately. Summary and Conclusions The clinical findings and histologic features of gland tumors treated at the Health Sciences Centre, Winnipeg, Canada, from 1957 to 1970 were reviewed. The study illustrates a wide variation in the natural history and degree of malignancy of parotid tumors. Approximately 80 per cent were benign; among these, mixed tumors were by far the commonest. For most benign tumors, 154 parotid
Volume
129,
June
1975
the recommended surgical treatment is wide local excision with an adequate margin of normal glandular tissue, except when subtotal or conservative total parotidectomy is required because of the size or location of the tumor. The facial nerve should be visualized and preserved. This approach prevents recurrences and minimizes facial nerve injury, since risk of neural injury increases with the extent of the surgical procedure. Total removal of the gland on the basis of the multicentricit,y or malignant transformation of these tumors is not supported by the findings of this study. For malignant tumors a more radical procedure, that is, subtotal or total parotidectomy with or without node dissection, with sacrifice of the facial nerve if necessary is advisable, depending on the type and anat.omic location of the tumor. Acknowledgment: We wish to thank Dr R. 1,. Cooke, Department of Surgery, Division of Oncology, Health Sciences Centre, for his guidance in this study; Dr Christopher Man for his help; and MS Mildred Marsh for typing the manuscript. References 1. Beahrs OH, Woolner LB, Carveth SW, Devine KD: Surgical management of parotid lesions. Review of 760 cases. Arch Surg 80: 890, 1960. 2. Berdal P, Gronas HE, Mylius E: Parotid tumors: clinical and histological aspects. Acta Otolaryngol (Suppl) 263: 160, 1970. 3. Foote FW, Frazell EL: Tumors of the major salivary glands. Cancer 6: 1065, 1953. 4. Kirklin JW, McDonald JR, Harrington SW, New GB: Parotid tumors: histopathology. clinical behaviour and end reports. Surg Gynecol Obstet 92: 721, 1951. 5. McFarland J: 300 mixed tumors of the parotid of which 69 recurred. Surg Gynecol Obstet 63: 457, 1939. 6. Toraya AA, Berens J, Hale HW, Wagner J: Parotid gland tumors. Am J Surg 120: 629, 1970. 7. Hugo NE, McKinney P, Griffith BH: Management of tumors of the parotid gland. Surg C/in Notih Am 53: 105, 1973. 8. Frazell EL: Observations on the management of salivary gland tumors. Cancer 18: 235, 1968. 9. Wallace AC, MacDougall JT, Hildes JA, Lederman JM: Salivary gland tumors in Canadian Eskimo. Cancer 16: 1338, 1963. 10. Arthand JB: Anaplastic parotid carcinoma (malignant lymphoepithelial lesion) in seven Alaskan natives. Am J C/in Pathol57: 275, 1971. 11. Gore DO, Annamunthodo H, Harland A: Tumors of salivary gland origin. Surg Gynecol Obstet 119: 1290, 1964. 12. Potdar GG, Paymaster JC: Salivary gland tumors. Am J Surg 118: 440. 1969. 13. Molnar L, Ronway P, Dobrossy L: Mixed tumors of the parotid gland. Oncology 25: 143. 1971.
681
Study
Birendra K. Sinha, MD, FRCS (C), FACS,’ Winnipeg, Manitoba, Canada David W. Buntine, MBBS, FRCP (C), FRCPA, Winnipeg, Manitoba, Canada
Tumors of the parotid gland continue to be a surgical challenge due to their various degrees of malignancy and close proximity with the facial nerve. Although approximately 80 per cent of these tumors are considered benign [I-3], earlier reports indicated a high recurrence rate after limited excision [3-51. This has led to a more aggressive approach to the treatment of these lesions in recent years, in the form of either subtotal (superficial or deep lobectomy) or total conservative parotidectomy, with excellent results [1,6-81. At our Centre, however, the surgical treatment of benign tumors of the parotid gland has generally been less aggressive, being based on complete extirpation of the tumor with an adequate margin of normal glandular tissue and preservation of the facial nerve. Subtotal and total parotidectomy, although used in certain cases of benign tumors, were performed mainly for malignant lesions. This study was undertaken in an attempt to answer the following questions about the surgical treatment of parotid gland tumors: (1) whether subtotal or total parotidectomy should always be performed for control of benign tumors; (2) whether a more radical procedure is associated with a greater risk of facial nerve injury despite careful attempts to identify and dissect out the nerve; and (3) whether recurrence after a conservative procedure is controllable by further operation without undue risk to the patient. Material and Methods The clinical findings
and histologic features
were re-
viewed for 154 parotid gland tumors treated at the Health Sciences Centre, University of Manitoba, Winnipeg, from 1957 to 1970. The description of treatment From the Department of Surgery, Division of Oncology and Pathology, The University of Manitoba, and Health Sciences Centre, Winnipeg, Manitoba, Canada. ’ Present address and address for reprint requests: Elk Grove MedicalDental Park. 758 Arlington Heights Road, Elk Grove Village, Illinois60007.
Volume
129, June 1975
and follow-up data were obtained from the records of the Health Sciences Centre, the Manitoba Cancer Treatment and Research Foundation, and from respective attending surgeons. Pathology reports and original slides were available for review in all cases. The classification of the lesions was modified from that of Foote and Frazell [3]. The tumors were divided into two main groups, benign and malignant. There were one hundred twenty-two benign lesions (79 per cent of the entire series) and thirty-two malignant. Of the benign tumors, eighty-seven or 71 per cent were mixed tumors, including seventy-nine primary and eight recurrent mixed tumors. Warthin’s tumor was the second most common lesion, comprising twenty-six or 21 per cent of the total benign tumors, a relatively higher incidence than that in most reported series [1,3,6]. Other less frequently occurring benign tumors included five benign lymphoepithelial lesions (Godwin), one oncocytoma, two vascular tumors, and one parotid cyst. The malignant tumors included ten mucoepidermoid carcinomas, ten adenoid cystic carcinomas, three acinic cell carcinomas, five adenocarcinomas not otherwise specified, three anaplastic carcinomas in Eskimos, and one malignant mixed tumor. Benign Lesions. Primary mixed tumors: There were thirty males and forty-nine females in this group. The youngest patient was 13 and the oldest 81 years of age, with average age of 47.5 years. (Figure 1.) An asymptomatic parotid lump was the commonest presenting feature. Pain or tenderness was present in only six patients. None had facial palsy. The size of the lesion ranged from 1 to 8 cm, the majority being between 2 and 3 cm. Recent increase in size of the lesion was recorded in twenty-one patients. The shortest duration of the presence of lesion was two weeks and the longest thirty-five years, approximately 50 per cent being present for one to five years prior to diagnosis. Morphologically these tumors were pleomorphic with either epithelial or mesodermal elements predominating. (Figure 2.) It was not possible to correlate differences in clinical behavior with the histologic features of the benign mixed tumor. Recurrent mixed tumors: A total of eight patients had recurrent mixed tumors of the parotid gland, four males and four females. Two patients had a history of three
675
Sinha and Buntine
0 MIxed ,umor
n
Warthtn s
0 Molqnonl
M-30.F=49
M=l5.F=9 Tumors M=l5.F=17
DECADES
Figure 1. Age distribution of benlgn mixed tumor, Warthin’s tumor, and malignant tumors of the parotid gland.
of cyclic spaces lined with a double layer of pink epitheha1 cells (oncocytes) and varying amounts of lymphoid tissue in the surrounding stroma. Benign lymphoepithelial lesions (Godwin): There were five patients with this lesion, one male and four females. The youngest patient was 37 and the oldest 80 years of age. The size of the lesion ranged from 1 to 3 cm and duration from three to eight months. These patients presented with painless parotid lumps and underwent surgery mainly because the diagnosis of benign or malignant neoplasms could not be excluded. The microscopic picture consisted of loss of acinar tissue, ductal and myoepithelial cell proliferation with foci of lymphoid infiltration. Others: The two vascular lesions were a cavernous hemangioma and arteriovenous malformation. The other two were oncocytoma and benign epithelial cyst of parotid gland origin
previous operations over a period of forty years, including one who also had received radiation therapy after the initial local excision. One patient had a history of two previous excisions and the remaining five only one. The youngest patient was 17 years and the oldest 78 years of age, with an average age of 52.5 years. The size of the recurrent tumor was between 1 and 3 cm and the duration ranged from two months to eight years. Recent change in size was noted in two patients and pain was recorded in one. None had facial palsy. Warthin’s tumor: Twenty-four patients presented with Warthin’s tumor, two of whom had bilateral lesions. There were fifteen males and nine females. The youngest patient was 46 years and the oldest 79 years of age, with an average age of 60 years. (Figure 1.) The size of the lesion ranged from 1.5 to 6 cm and the duration from four weeks to twelve years. Recent increase in size was noted in nine patients. Two patients presented with pain and the rest were asymptomatic. The lesion generally was located in the tail of the parotid gland and was mobile and soft. Six of these were cystic. Microscopically this lesion had a characteristic appearance consisting
Malignant Tumors. Thirty-two patients with primary malignant tumors ranged in age from 17 to 83 years, with an average age of 48 years. (Figure 1.) There were fifteen males and seventeen females. The size of the lesion ranged from 1 to 8 cm and the duration from five months to six years. Fifteen patients presented with a painless lump in the parotid region. Pain and tenderness were present in another fifteen patients, fixation of the tumor to the underlying structures in three patients, and facial palsy in two patients. In contrast, of the onehundred twenty-two patients with benign lesions, onehundred fourteen presented with an asymptomatic lump, eight with pain and tenderness, and none with facial palsy. Mucoepidermoid carcinoma: There were seven males and three females in this group, their ages ranging from 17 to 76 years. The size of the lesion varied from 2 to 8 cm and the duration from five months to one year. Microscopically, these tumors showed various proportions
Figure 2. Photomicrograph (high power) of a benign mixed tumor showing markedly pieomorphic ceils in hyaline-like stroma.
Figure 3. Photomicrograph (high power) of an infiitrating ductai carcinoma with dense fibrous stroma.
676
The American
Journal
of Surgery
Parotid
of both squamous and mucus-secreting cells. The histologic gradings appeared to correlate well with the biologic behavior of the tumor. Of the seven tumors regarded as low grade, none showed evidence of metastasis or recurrence. Of the three regarded as high grade, one metastasized t.o the regional lymph node and another involved the facial nerve causing facial paralysis. Adenoid cystic carcinoma: There were five males and five females from 23 to 83 years of age. These tumors were generally slow growing. The size of the lesion varied from 1 to 4.5 cm and the duration from one to eight years. Grossly, the tumor appeared well circumscribed but nonencapsulated, and microscopically it was corn.posed of epithelial and myoepithelial cells surrounding cores of hyaline of various sizes giving rise to the distinc tive Swiss cheese appearance. The characteristic ability of this tumor to invade the perineural space was seen in one patient. However, no lesion had metastasized to the regionai lymph nodes. Acinic cell carcinoma: This was diagnosed in three patients, two males and one female, between 25 and 48 years of age. The size of the lesion varied from 2 to 3 cm and the duration from five months to five years. These tumors were poorly circumscribed and microscopicall> composed of acinar cells showing basophilic granular cy., toplasm. Adenocarcinoma (not otherwise specified): This unusual group of lesions presented a microscopic appearance of ductal carcinoma, often with a dense fibrous stroma (Figure 3), and showed a great propensity to metastasize to distant organs. This was diagnosed in five patients, four males and one female, their ages ranging from 28 to 76 years. The lesions were 2 to 3 cm in size and generally of short duration, less than a year. Anaplastic carcinoma in Eskimos: All three patients were female, 17, 31, and 45 years of age. These tumors varied from 3 to 7 cm in size and were irregular and fixed to the underlying structures. One patient had pain and tenderness as well as involvement of the facial nerve. Grossly, the lesion was locally invasive and in two instances had metastasized to regional lymph nodes. Microscopically, the tumor was composed of groups of anaplastic epithelial cells in a stroma heavily infiltrated with lymphoid cells. (Figure 4.) This lesion seems to be a malignant variant of the benign lymphoepithelial lesion of the parotid gland and appears peculiar to the Alaskan Indians and the Canadian Eskimos as reported previously [9,10]. Malignant mixed tumors: Only one patient, a 45 year old male, had a malignant mixed tumor of the parotid gland according to the criteria used at this Centre. This particular lesion was 8 cm in diameter and had been present for ten years withL a recent increase in size. As mentioned earlier, the degree of pleomorphism and cellularity did not correlate with the biologic behavior of mixed tumors. The only reliable criteria for diagnosis of malignant mixed tumor in our experience is vascular invasion, although a high mitotic index must be considered suspicious.
Volume 129, June 1975
Gland Tumors
Figure 4. Photomicrograph (high power) of an Eskimo tumor showing anaplastic epithelial cells in a stroma heavily inffftrated with lymphocytes.
Diagnosis, Treatment, Complications, and Results Diagnosis was based mainly on the clinical evaluation of the patient. Plain x-ray st,udy and sialography were performed in only a few cases, which aided in the recognition of non-neoplastic disorders but did not contribute significantly to the distinction between benign and malignant neoplasms. In most instances the diagnosis was confirmed after wide local excision. When malignancy was suspected, an incisional biopsy was performed on operative exposure of the parotid gland prior to the definitive procedure. More recently, the disposable Vim-Silverman needle has been used to obtain an adequate core of tissue using local anesthesia for histologic diagnosis. The procedure is relatively simple, reliable, and in our experience, free of the commonly feared risk of tumor spillage along the needle track. Tables I and II present the treatment of benign and malignant tumors of the parotid gland. Patients undergoing enucleation, limited local exci-
TABLE ____
I -..
Type of Tumor ___ Mixed primary Mixed recurrent Warthin’s Godwin’s Others Total _~~ ~_ -
Treatment Gland
of Benign Tumors
of the Parotid
Local Excision
Subtotal Parotidectomy
Total Parotidectomy
Postoperative Radiation
79
61
18
0
2
8 26 5 4 122
7 15 2 2 87
1 10 3 2 34
0 1 0 0 1
0 0 1 0 3
Number
677
Sinha
TABLE
and Buntine
II
Treatment
of Malignant
Type of Tumor Mucoepidermoid Adenoid cystic Acinic cell Adenocarcinoma specified) Anaplastic Mixed Total
(not
Tumors of the Parotid Gland
Number
Local Excision
10 10 3
1 6 2
5 3 1 32
2 0 0 11
(Eskimo)
III
Postoperative Complications of Parotid Gland Surgery for Benign and Malignant Tumors Complication
Benign*
Malignant?
8 11 9 1 2
... 1 3 ... 1
1 1
... 1
Facial paralysis Upper trunk Lower trunk Total paralysis Salivary fistula Auriculotemporal syndrome Wound complications Seroma Hematoma and infection * Facial nerve was sacrificed benign tumor. t Facial nerve was sacrificed of malignant tumors.
TABLE
IV
intentionally intentionally
in one
case
in all four
676
of
cases
Incidence of Postoperative Facial Paralysis Related to Extent of Surgery for Benign Tumors with and without Exposure of the Nerve Incidence of Facial Paralysis
Local excision Subtotal parotidectomy Total parotidectomy
6 3 1
1 0 0
2 0 0
7 0 0
2 1 0
3 2 1 16
0 1 0 2
0 2 0 4
4 3 0 14
0 0 0 3
Total
Postoperative Radiation
Biopsy and Radiation
otherwise
Son, or wide local excision as indicated in the operative report are all grouped as having had local excision. Those undergoing excision of either the entire superficial lobe or the deep lobe (in only one instance) with or without excision of the isthmus are classified as having had subtotal parotidectomy, and removal of the entire gland is synonymous with total parotidectomy. For benign lesions, TABLE
Parotidectomy
Radical Neck Dissection
Subtotal Parotidectomy
Total Number
Cases of Facial Paralysis
Nerve Exposed
Nerve Not Exposed
87
20
7150
13137
34
7
7134
1
1
l/l
eighty-seven local excisions, thirty-four subtotal parotidectomies, and one total parotidectomy were carried out. For malignant lesions, eleven local excisions (six of these for adenoid cystic carcinoma) and sixteen subtotal parotidectomies were performed. Two patients, one with mucoepidermoid carcinoma and the other with anaplastic carcinoma (an Eskimo), underwent total parotidectomy. In four patients, two with mucoepidermoid and two Eskimos with anaplastic carcinoma, either subtotal or total parotidectomy was combined with radical neck dissection. Fourteen patients received postoperative irradiation, including seven patients with mucoepidermoid carcinoma, four with adenocarcinoma (not otherwise specified), and three with anaplastic carcinoma. Two patients with mucoepidermoid carcinoma and one with adenoid cystic carcinoma had only palliative radiotherapy after biopsy. Table III presents the incidence of postoperative complications in benign and malignant neoplasms. The most frequent complication was facial nerve injury, either partial or total, which occurred in twenty-eight patients with benign lesions and four patients with malignant lesions. In the latter cases it occurred as a result of intentional sacrifice of the nerve located close to the tumor. However, in the case of benign tumors sacrifice of the facial nerve was intentional in one case only; the remaining instances were due to inadvertent injury of the facial nerve during operation. The incidence of postoperative facial paralysis in these cases was further analyzed relative to the extent of surgery and whether or not an attempt had been made to preserve the facial nerve. (Table IV.) The findings suggest an apparent correlation between the extent of surgery and postoperative facial paralysis even when an effort was made to preserve the nerve. In the group with local excision, however, the incidence appeared to be higher when blind
The American Journal of Surgery
Parotid
excision was carried out than when the facial nerve was identified and exposed. All patients except three with benign tumors were followed up three to twelve years. (Table V.) Of the seventy-nine primary mixed tumors, there were ten recurrences in seven patients, three patients having two recurrences each three months to nine years after the initial procedure. All patients with recurrence had undergone local excision. In this group histologic slides and pathology reports were reviewed with regard to the completeness of excision. Of sixty-one local excisions, twenty-nine were considered histologically complete and among these there were no recurrences. Thirty-two either were considered incomplete or the tumor extended to the line of excision and all ten recurrences were noted in this group, correlating well with the histologic findings. However, all recurrences were controlled by further local excision except in one patient who required total parotidectomy because of multifocal recurrence only three months after the initial operation. There were two recurrences in one patient in the group with recurrent mixed tumor that were controlled by repeated local excision. In the remaining patients with benign lesions, there was only one recurrence in a patient with Warthin’s tumor, controlled by local excision. There were no deaths attributable to benign tumors. Table VI presents the survival data of patients with malignant tumors during a follow-up period of three to twelve years. A total of fifteen patients or approximately 50 per cent are alive and free of disease, two are alive with disease, and fifteen are dead, including eight who died from the disease. Of ten paCents with mucoepidermoid carcinoma? four are alive free of disease, one is alive with disease, three are dead from disease, and two dead from other causes. Of ten patients with adenoid cystic carcinoma, six are alive with no evidence of disease, one is dead from disease, and three are dead from other causes. Of three patients with acinit cell carcinoma, one is alive free of disease, one is alive with disease, and one is dead from other causes. Of five patients with adenocarcinoma (not otherwise specified), two are alive free of disease and three are dead, including two from disease. One Eskimo patient with anaplastic carcinoma is alive free of disease, and the remaining two are dead from disease. Two patients who presented with involvement of the facial nerve by malignant tumor died within three years of diagnosis. There was no difference in the results of surgery and postoperative radiation as compared with surgery alone. Volume
129, June 1975
Tumors
Results of Treatment of 122 Benign Parotid Gland Tumors in 120 Patients during a Follow-Up Period of 3 to 12 Years
TABLE V
Type of Tumor
Gland
Dead from Other Causes+
Lost to FollowUp
68
9
2
4
3
1
19 3 4
5 2 0
0 0 0
Total Number of Number of Pabents Recurrences ,“ir * ~ ..~ _____ _~.
Mixed primary Mixed recurrent Warthin’s Godwin’s Others
10 in 7 pts
79
2 in 1 pt
8 24 5 4
1 0 0
Study
_~~
* NED = no evidence of disease. t None died from disease.
This series illustrates the wide variation in the natural history and degree of malignancy of parotid tumors, confirming previous reports [2,6-81. A benign tumor may be very slow growing or may remain unchanged for many years. On the other end of the spectrum, anaplastic carcinoma (Figure 5) may grow rapidly and metastasize rapidly causing death within a short period of time. There is unanimity of opinion that benign tumors must be treated surgically since they can attain an enormous size and cause serious disability (Figure 6) or local complications, including ulceration, bleeding and infection (Figure 7), and occasionally death, as reported previously [6,8]. Previous authors have reported the great propensity of benign mixed tumors to recur locally after surgical removal 13~51.
TABLE
VI
Type of Tumor
Results of Treatment of 32 Malignant Parotid Gland Tumors during a Follow-Up Period of 3 to 12 Years
Total Number of Patients
Alive NED*
Alive with Disease
Dead from Disease
Dead from Other Causes
Mucoepidermoid Adenoid cystic Acinic cell Adenocarcinoma (not otherwise specified) Anaplastic (Eskimo) Mixed Total
10 10 3
4 6 1
1 0 1
3 1 0
2t 3 1
5
2
0
2
1
3 1 32
1 1 15
0 0 2
2 0 8
0 0 7
* NED = no evidence of disease. t One patient had residual disease
at the time
of death.
679
Sinha and Buntine
Figure 6. A benign mixed tumor grown to an enormous size (incidental finding in a patient admitted with myocardiai infarction, not included in this series). Figure 5. Anapiastic carcinoma of the parotid gland in an Eskimo patient.
Figure 7. Warthin’s tumor showing ulceration and hemorrhage.
These reports also indicate that local recurrence results from inadequate surgical excision, which led to the more aggressive approach of subtotal or conservative total parotidectomy, with no recurrence or very low recurrence rates [1,6-81. This approach is supported by the findings of the present study, which show that all recurrences of primary
680
benign mixed tumors followed local excisions and none followed subtotal or total parotidectomy. However, in this study all local recurrences were noted in the group with incomplete local excision and none in those with complete excision as verified histologically. Furthermore, the risk of facial nerve injury increased with the extent of the surgical procedure, in contrast to previous reports [11,12]. The incidence of facial paralysis was.highest with total parotidectomy and lowest with local excision, although in the latter group facial nerve injury generally occurred when exposure and dissection of the nerve were not carried out. Thus, wide local excision with an adequate margin of normal glandular tissue verified by frozen section is the logical treatment for most benign tumors of the parotid gland except when subtotal or conservative total parotidectomy is required because of the size or location of the tumor. The facial nerve should be identified and preserved. This approach would prevent recurrence and minimize facial nerve injury as indicated by this study. In benign lesions, intentional sacrifice of the facial nerve is rarely justifiable since actual invasion or fixation of the nerve by benign tumors generally is absent, as indicated by the fact that none of the patients with benign tumor presented with facial paralysis. In our experience, careful dissection can permit removal of benign tumor lying close to the nerve without neural injury. If the facial nerve is
The American Journal of Surgery
Parotid Gland Tumors
divided during surgery it should be repaired immediately by end to end anastomosis or interposition of nerve graft. Analyzing the recurrences, we find them generally controllable by further surgery without facial nerve injury and without risk to life, contrary to other reports [1,6,12]. Total removal of the gland, based on the multicentricity of these tumors and possibility of malignant, transformation as reported previously [3,13], is not supported by the findings of this study, since in only one instance was the recurrence multifocal as confirmed by histologic study and required total parotidectomy. None of the recurrences underwent malignant changes. For malignant tumors, a more radical procedure in the form of subtotal or total parotidectomy is advisable. In general agreement with previous reports [1,6-81, radical neck dissection was performed only in presence of clinically involved regional lymph nodes except in cases of anaplastic carcinoma in Eskimo patients. Intentional sacrifice of part or all of the facial nerve may be necessary for adequate treatment of malignant lesions on a selective basis depending on the type and anatomic location of the tumor. Regional lymph node metastases and involvement of the facial nerve usually suggest a poor prognosis, confirming other reports [ 1,2]. The role of radiation therapy in the management of these tumors is uncertain, except for palliation in cases of advanced tumors. This modality was used in too few cases to evaluate its role in improving the survival data. A program of combined operation and either preoperative or postoperative irradiation for anaplastic carcinoma’ in Eskimo patients is currently underway and will be reported separately. Summary and Conclusions The clinical findings and histologic features of gland tumors treated at the Health Sciences Centre, Winnipeg, Canada, from 1957 to 1970 were reviewed. The study illustrates a wide variation in the natural history and degree of malignancy of parotid tumors. Approximately 80 per cent were benign; among these, mixed tumors were by far the commonest. For most benign tumors, 154 parotid
Volume
129,
June
1975
the recommended surgical treatment is wide local excision with an adequate margin of normal glandular tissue, except when subtotal or conservative total parotidectomy is required because of the size or location of the tumor. The facial nerve should be visualized and preserved. This approach prevents recurrences and minimizes facial nerve injury, since risk of neural injury increases with the extent of the surgical procedure. Total removal of the gland on the basis of the multicentricit,y or malignant transformation of these tumors is not supported by the findings of this study. For malignant tumors a more radical procedure, that is, subtotal or total parotidectomy with or without node dissection, with sacrifice of the facial nerve if necessary is advisable, depending on the type and anat.omic location of the tumor. Acknowledgment: We wish to thank Dr R. 1,. Cooke, Department of Surgery, Division of Oncology, Health Sciences Centre, for his guidance in this study; Dr Christopher Man for his help; and MS Mildred Marsh for typing the manuscript. References 1. Beahrs OH, Woolner LB, Carveth SW, Devine KD: Surgical management of parotid lesions. Review of 760 cases. Arch Surg 80: 890, 1960. 2. Berdal P, Gronas HE, Mylius E: Parotid tumors: clinical and histological aspects. Acta Otolaryngol (Suppl) 263: 160, 1970. 3. Foote FW, Frazell EL: Tumors of the major salivary glands. Cancer 6: 1065, 1953. 4. Kirklin JW, McDonald JR, Harrington SW, New GB: Parotid tumors: histopathology. clinical behaviour and end reports. Surg Gynecol Obstet 92: 721, 1951. 5. McFarland J: 300 mixed tumors of the parotid of which 69 recurred. Surg Gynecol Obstet 63: 457, 1939. 6. Toraya AA, Berens J, Hale HW, Wagner J: Parotid gland tumors. Am J Surg 120: 629, 1970. 7. Hugo NE, McKinney P, Griffith BH: Management of tumors of the parotid gland. Surg C/in Notih Am 53: 105, 1973. 8. Frazell EL: Observations on the management of salivary gland tumors. Cancer 18: 235, 1968. 9. Wallace AC, MacDougall JT, Hildes JA, Lederman JM: Salivary gland tumors in Canadian Eskimo. Cancer 16: 1338, 1963. 10. Arthand JB: Anaplastic parotid carcinoma (malignant lymphoepithelial lesion) in seven Alaskan natives. Am J C/in Pathol57: 275, 1971. 11. Gore DO, Annamunthodo H, Harland A: Tumors of salivary gland origin. Surg Gynecol Obstet 119: 1290, 1964. 12. Potdar GG, Paymaster JC: Salivary gland tumors. Am J Surg 118: 440. 1969. 13. Molnar L, Ronway P, Dobrossy L: Mixed tumors of the parotid gland. Oncology 25: 143. 1971.
681
