The Journal of Laryngology and Otology April 1979. Vol. 93. pp. 325-340

Acinic cell carcinoma: A clinicopathologic study of thirty-five cases By JOHN G. BATSAKIS, M.D.* EDWIN K. CHINN, M.D.* THOMAS A. WEIMERT, M.D.f WALTER P. WORK, M.D.f and CHARLES J. KRAUSE, M.D.t WHEN one reviews the literature concerning acinic cell neoplasms over the past 25 years, one cannot help but be reminded of the confusion and controversy surrounding the mixed tumor of salivary glands that was prevalent in the 193O's and 1940's. The corollaries between the two lesions concerning their histogenesis, biologic course and appropriate forms of management are sometimes striking. Acinic cell neoplasms were not recognized as a specific class of salivary gland tumors until the early 1950's. Credit is given to Buxton et al. (1953) for being the first to ascribe a 'malignant ability' to these lesions. Since that time there has been an evolution by pathologists and surgeons to consider acinic cell neoplasms as carcinomas. This trend has not been without its detractors. Evans and Cruickshank (1970) hold firmly to the notion that there are only occasional 'facultative malignant variants' of acinic cell tumors. Others assume a purely academic and post-hoc stance and consider the neoplasms merely as tumors; dividing them into 'benign' or 'malignant' on the basis of their biologic course in the host (Godwin et al., 1954; Sharkey, 1977; Thackray and Sobin, 1972). This type of reasoning has culminated in the rather grim outlook that some patients will have unnecessary radical surgery and some patients will have insufficient surgery for these lesions (Sharkey, 1977). Disconcerting as that may sound, it is more disturbing that pathologists appear to be prisoners of their own nomenclature. The epitome of such semanticism appears to have been reached by Thackray and Sobin (1972): 'there seems to be no more necessity to change the name to carcinoma if a tumor happens to metastasize than there is justification for calling those that have not yet done so adenomas.' Would that the patients who have died of their acinic cell carcinomas be so reassured! Our purpose in this report is to present the clinical and pathologic findings in 35 cases of acinic cell carcinoma. In so doing we intend to affirm the malignant potential of this neoplasm, to present a histologic hypothesis of origin and to provide information for optimum treatment. * Department of Pathology, The University of Michigan Medical School, Ann Arbor, Michigan 48109. t Department of Otorhinolaryngology, The University of Michigan Medical School, Ann Arbor, Michigan 48109. 325

326 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE

Materials and methods During the course of an on-going histologic reclassification of salivary gland tumors from patients treated at the University of Michigan Hospital over the past 50 years, thirty-five cases fulfilling histologic criteria of acinic cell carcinomas were selected for analysis. While follow-up information was available on all patients, four were excluded from statistical analysis of biologic course because their surveillance period was less than six months. The remaining 31 patients had an average follow-up period of 7-5 years. One patient was followed for 42 years after his primary treatment. Eleven of the 35 patients had primary surgery for their lesions elsewhere before referral to University Hospital. An additional five patients had biopsy before referral. Since one of the purposes of this study was to evaluate the effect of treatment, regardless of the origin of that treatment, no further segregation was carried out. Hematoxylin and eosin stained sections from all cases were studied and where necessary, additional sections were prepared for study from blocks or additional wet tissue. All metastases were verified either by review of antemortem biopsy material or of necropsy specimens. The carcinomas were independently graded (without historical information) by one of the authors (J.G.B.) into histomorphologic high and low grade carcinomas. This subjective grading was then correlated with biologic course of the neoplasm and outcome for the patient. Criteria for the grading classification is presented below under histomorphologic features. Results Primary sites Only two of the 35 acinic cell carcinomas were not primary in the parotid gland. One tumor was in the palate; the other, in the submandibular gland. Histomorphologic features Table I presents a listing of the various pathologic diagnoses rendered for the 35 acinic cell carcinomas. Less than half of the cases were correctly diagnosed at the initial treatment. Twelve of the lesions carried diagnoses which imply a benign tumor. Fifteen of the 20 acinic cell carcinomas diagnosed after 1960 were correctly diagnosed; only one of 15 cases presenting before 1960 was classified correctly. The histologic growth pattern of the carcinomas fell into one or a combination of the following: (1) acinar-lobular, (2) microcystic, (3) follicular, (4) papillary cystic, (5) medullary, (6) ductulo-glandular, and (7) primitive tubular. A definable 'capsule' was observed in only four cases; the usual delimitation being only a thin condensation of stroma. Multinodularity was common and a prominent feature. This was more pronounced in recurrent

327

ACINIC CELL CARCINOMA TABLE I ORIGINAL DIAGNOSES RENDERED ON ACINIC CELL CARCINOMAS

Frequency Initial diagnosis

Total

Acinic (Serous) Cell Carcinoma Acinic (Serous) Cell Adenoma Mucoepidermoid tumor Mixed tumor Adenocarcinoma Adenoma Oncocytoma Papillary cystoadenoma 'Tumor'

16 4 4 3 3 2 1 1 1

Before 1950

1950-1960

After 1960

1 4

15 4

1 2

2 1 2

1 1 1

tumors. Remote, satellite micronodules were present in 10 cases. Three carcinomas manifested clear histological evidence of a multifocal origin (Fig. 1). One of these foci arose in a paraparotid lymph node. A conspicuous lymphoid infiltrate was present in association with the neoplasm in 13 cases. In five of these, the character and circumscription of the infiltrate were such as to strongly suggest origin within intra-parotid lymphoid aggregates or in small lymph nodes.

FIG. 1 Multifocal nodules of acinic cell carcinoma in a previously untreated parotid gland (H. & E. x8).

328 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE

Calcification was present in only four tumors. In three of these, it took the form of calcopherites and in the other, the calcification was of a dystrophic type in broad areas of dense collagenous tissue. The cells within the aforementioned growth pattern varied in their proportions within given tumors and consisted of cells designated as acinic, intercalated, vacuolated and atypical or primitive, undifferentiated types. As indicated above, the acinic cell carcinomas were histologically divided into high- and low-grade lesions based on several criteria. Features characterizing high-grade acinic cell carcinomas were (1) an aggressive local invasion, manifested by an infiltrative rather than blunt growth pattern; (2) extra-parenchymal invasion of bone, muscle, soft tissue, nerve or vessels (Fig. 2); (3) a medullary as opposed to a lobular growth architecture (Fig. 3); (4) a ductulo-glandular or tubular morphology in any part of the tumor (Figs. 4 and 5); and (5) prominence of undifferentiated cells in the composition of a tumor (Fig. 6). It is to be emphasized here that the ductular pattern and undifferentiated foci never dominated an acinic cell carcinoma but were present in areas of sections and readily identifiable. Low-grade carcinomas manifested acinar-lobular, cystic and papillary growth patterns and were composed of cells that had no ductulo-glandular or tubular elements. These tumors most closely approximated the normal lobular arrangement of acini (Fig. 7).

FIG. 2 Endolymphatic invasion by a high grade acinic cell carcinoma (H. & E. X480).

FIG. 3 High-grade acinic cell carcinoma, parotid gland demonstrating a medullary pattern (H. & E. X150).

FIG. 4 High-grade acinic cell carcinoma, parotid gland manifesting tubular and acinar differentiation (H. &E. x200).

FIG. 5 High-grade acinic cell carcinoma. In this field there is little resemblance to acini found elsewhere in ths neoplasm. These neoplastic ducts infiltrate parotid gland and by electron microscopic examination appear to be the neoplastic equivalent of intercalated ducts or terminal tubules (H. & E. x 160).

FIG. 6 Highly undifferentiated focus in an acinic cell carcinoma (H. & E. x 200).

ACINIC CELL CARCINOMA

FIG. 7 Low grade acinic cell carcinoma manifesting blunt invasive pattern (H. & E. x 100).

Extra-parenchymal local invasion was found in 12 patients. This included invasion of bone (6 cases); nerve (5 cases); skin (2 cases); blood vascular (2 cases); and lymph node (2 cases). In none of the neoplasms was there an associated co-existent salivary tumor. Sex and age distribution Twenty-four of the patients were females, yielding a female dominance of nearly 2:1. The age at the time of diagnosis ranged from 9 years to 73 years (median 40: mean, 43 years). There were two observable peak age incidences; a major one in the fourth decade and a lesser one in the seventh decade of life. Presenting signs and symptoms Asymptomatic swelling was the most common presenting complaint (68 per cent). Pain in company with swelling occurred in 16 per cent of the patients. Table II associates clinical signs and symptoms with fatal outcome. The time between onset of clinical manifestations and diagnosis ranged to 18 years, with a mean duration of 2-9 years.

332 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE TABLE II PRESENTING SYMPTOMS A N D FREQUENCY OF DEATH FROM ACINIC CELL CARCINOMA

Presenting symptoms

Frequency

Died of tumor

Swelling alone Pain/tenderness Pain and swelling Facial paralysis Facial weakness Difficult jaw movement

21 1 5 2 1 1

0 2 2 1 1

3

Clinicopathologic correlation Table III summarizes the relationship of four clinicopathologic factors to patients' outcome. A preoperative biopsy does not appear to influence the eventual biologic course of the neoplasm. The relationship between type of surgical treatment and outcome is seriously hampered by a small sample size. The only strong relationship is between the rate of recurrence and local excision/enucleation of the neoplasm. The first recurrence in patients manifesting that event occurred between 2 months and 25 years after the initial surgical procedure. Only four of 29 patients with a follow-up of more than six months had recurrences in the first post-operative year, but 16 of 29 patients (55 per cent) eventually developed local recurrences during the course of their follow-up (Table IV). TABLE m RELATION OF HISTOLOGIC GRADE, LOCAL INVASION AND PRIMARY TREATMENT TO OUTCOME

Death due Number patients 12 19 31

7-6 7-4 7-5

7 9 16

(58) (47) (52)

7 2 9

(58) (11) (29)

7 1 8

(58)

12 19

4-7 9-2

8 8

(67) (42)

6 3

(50) (16)

6 2

(50)f

9 22

6-7 7-8

2 14

(22) (64)

4 5

(44) (23)

2 6

(22) (27)

13

121

11

(85)

4

(3D

4

(3D

5 10 3

4-3 41 4-3

2 1 2

(40) (10) (67)

0 3 2

(0)

(20) (67)

0 4 0

(40)

of

Histologic grade High grade Low grade All

Local invasion* Present Absent Preoperative biopsy Yes No

Type of surgical treatment Local excision (or enucleation) Subtotal parotidectomy (superficial) Total parotidectomy Extended parotidectomy

Local recurrence No. (%)

to

Average following (years)

Metastasis No. (%)

neoplasm No. (%)

(5)

(26) (16)

(0) (0)

* Local invasion denned as neoplastic invasion of nerve, bone, muscle, skin, vessels. f Includes three of four patients with clinical and pathological evidence of VII invasion.

ACINIC CELL CARCINOMA

333

Metastases developed in nine patients and occurred in 14 sites (Table V). Lymphogenous metastases to cervical lymph nodes occurred in four patients. Two other patients had involvement of their paraglandular lymph TABLE IV RECURRENCES OF ACINIC CELL CARCINOMAS: FREQUENCY RELATED TO POST-OPERATIVE TIME

Follow-up period (years)

Frequency of first local recurrences

Less than 1 1-4 5-9 10 and greater

4 10 2 0 TABLE V

METASTASES OF ACINIC CELL CARCINOMA: SITES AND TIME INTERVAL

Years after primary treatment

Cervical lymph nodes Lung Liver Knee (femur) Trapezius muscle Stomach 'Widespread'

4 4 2 1 1 1 1

5-9

10-14

1

2 1

15 and longer

4 1 1

1 1 1 1

nodes. This was regarded as involvement by contiguity rather than metastasis. Metastasis to cervical lymph nodes occurred relatively early; within 5 years after the primary treatment. Distant metastases, with one exception, occurred after 5 years (Fig. 8). In four of the patients with metastases to distant sites, local control of the disease appeared to have been achieved. Even though histologic evidence of local invasiveness is one of the criteria for a high grade carcinoma, that feature is separated for statistical analysis of comparison. Table 6 presents a Z-test comparison of local invasion and other histologic features characterizing a high-grade carcinoma against low grade carcinomas without invasion (American Joint Committee, 1977). As can be seen by the Table, all Z values are statistically significant. Table 7 presents the adjusted survival rates as calculated by actuarial (life table) method for the 31 patients with follow-up. Discussion Acinic cell carcinomas account for between 2- 5 per cent and 4 per cent of all tumors of the parotid gland (Abrams et al, 1965; Chong et ah, 1974;

334 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE TABLE VI COMPARISON OF 5-YEAR CUMULATIVE SURVIVAL: HISTOLOGIC GRADE A N D LOCAL INVAS1VENESS

Number of patients

Cumulative 5-year survival rate

High grade

12

0-626

Low grade Local invasion

19 12

1000 0-576

No local invasion

19

1-000

Value of Z statistic

Probability

2-49

Less than 0-05

2-51

Less than 0-05

FIG. 8 Metastatic high grade acinic cell carcinoma from parotid gland in lung. (H. & E. x 100).

Eneroth et al., 1966). Their incidence in salivary tissues other than the parotid gland is of a low order and is dominated by an origin from minor salivary glands of the oral cavity (Levin et al., 1975). The carcinoma ranks only behind Warthin's tumor in its frequency of bilateral parotid gland involvement; approximately 3 per cent (Levin et al., 1975). Characteristically presenting in patients who are in their fifth decade of life, the acinic cell carcinoma also figures prominently in the malignancies of salivary

ACINIC CELL CARCINOMA

335

TABLE VII ADJUSTED SURVIVAL: ACTUARIAL METHOD OF ALL PATIENTS WITH FOLLOW-UP*

Last year after definitive treatment

Patients alive at beginning of period

Patients exposed to risk of dying

Patients dying of tumor

Proportion surviving

Cumulative proportion surviving

1 2 3 4 5 6 7 8 9 10 15

31 26 23 21 16 14 11 10 7 4 4

290

3-5

1 2 0 0 1 0 0 1 1 0 1

0-966 0-920 1000 1000 0-935 1000 1000 0-889 0-883 1000 0-714

0-966 0-889 0-889 0-889 0-832 0-832 0-832 0-740 0-617 0-617 0-441

20

1

25-5 220

18-5 15-5 12-5 10-5 90 60

.

40

Greater than 16

2

* Standard error of 5 and 10 year survival rate are 0-079 and 0-145, respectively.

glands in childhood. Only the mucoepidermoid carcinomas are more common (Krolls et ai, 1972). Demographics aside, the clinicopathologic importance of acinic cell carcinomas lies in their pathogenesis and rationale for optimum therapy based on their biologic course and prognosis. Pathogensis Acinic cell carcinomas, because of their unique position at the distal end of the salivary unit, are particularly well suited to pathogenetic hypotheses. Our theory of the pathogenesis of salivary gland tumors rests on the preeminent role of duct stem cells or reserve cells in replenishing the epithelium of the salivary gland unit (Batsakis et ah, 1977; Regezi and Batsakis, 1977). It is in concert with Pierce's (1974) concept of the development of general neoplasia; a 'pathology' of the cells involved in tissue renewal. The development of a neoplasm then becomes an epigenetic event superimposed upon the process of cell renewal by stem or reserve cells. Malignant stem cells are derived from normal stem cells by a process equivalent to postembryonic differentiation. Malignant stem cells, however, have a capacity for proliferation and differentiation that operate on a different level of central than the normal phenotypic expression. Studies using partial extirpation and chemical injury clearly substantiate that salivary gland regeneration following injury is analogous to embryonic development (Cutler and Chaudhry, 1974; Hanks and Chaudhry, 1971; Redman and Sreebny, 1970). After such injury, the recovery phase is characterized by a proliferation and differentiation of the most terminal portion of the salivary duct system. It is from these cells also that acinic cell carcinoma is most likely derived.

336 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE

Although the reserve or stem cell theory is attractive, the identification of these cells is elusive. From a composite of morphological, histochemical, and ultrastructural studies (Bloom and Henriksson, 1977; Cutler and Chaudhry, 1974; Erlandson and Tandler, 1972; Redman and Sreebny, 1970) the cells making up the distal end of developing salivary glands may be characterized as follows: (1) the intercalated duct cell; (2) the terminal tubule cell; (3) the proacinar cell, and (4) the acinar cell. Transferring our knowledge from subhuman forms, it would appear most likely that as it is in post-natal development, acini and intercalated ducts evolve from the cells of the terminal tubule. This link between the functioning duct septa and the acini, is the source of the reserve or stem cells. The neoplastic equivalents of the terminal tubules are seen in those acinic cell carcinomas considered by us a high grade and represented by the undifferentiated tubules and solid epithelial masses (Fig. 9). The latter have a striking light microscopic similarity to embryonic terminal epithelial clusters in which the differentiation of terminal tubules occurs. In addition, our earlier ultrastructural findings on one of the neoplasms (Batsakis et al., 1977) and those of other investigators (Bloom and Henriksson, 1977;

I FIG. 9 A solid epithelial mass of undifferentiated cells in a high-grade acinic cell carcinoma. Elsewhere differentiation to ducts and acini was evident. Masses such as this appear nearly identical to the terminal epithelial clusters of the embryo which give rise by differentiation into terminal tubules, intercalated ducts and acini (H. & E. X480).

337

ACINIC CELL CARCINOMA

Erlandson and Tandler, 1972; Regezi and Batsakis, 1977) nearly exclude all but the terminal duct system as the source for these structures. Since de-differentiation of a fully developed acinar cell is not probable, the varying patterns of differentiation seen in acinic cell carcinomas are attributed to the attempts of the terminal tubule to approximate the neoplastic equivalent of the normal phenotypic expression of acinic lobules. The fullest neoplastic expression of this differentiation is the low-grade acinic cell carcinoma. Prediction of biologic behavior, therapy and prognosis The findings of this study and of others (Table VIII) indicate that a patient with an acinic cell neoplasm is ill-served by considering the tumor TABLE VIII RECURRENCE A N D METASTASIS: ACINIC CELL CARCINOMA

Authors Abrams et al. (1965) Beahrs et al (1950) Buxton et al. (1953) Eneroth et al. (1966) Foote and Frazell (1953) Fox et al. (1963) Godwin et al. (1954) Gorlin and Chaudhry (1957) Grage et al. (1961) The University of Michigan

Number of patients

Local recurrence

18 12 21 63 21 46 27 10 8 31

10 1 8 12 4 22 16 2 5 14

Metastasis indeterminate 4 — 10 —

—. 4 5 4 11

Death due to neoplasm 3 4 7 7 7 7 3 — 1 8

as benign. They should also dispel further ambiguity over the malignant potential of these neoplasms. Clinical experience clearly refutes statements such as: 'The number of cases that ultimately become malignant and metastasize is so small that these tumors are not entitled to be grouped all together under the generic term acinic cell carcinomas. Neither are they to be regarded as semi-malignant growths' (Evans and Cruickshank, 1970). A lack of definable histologic criteria of malignancy in many of these tumors cannot be justification for a conservative stance towards their management. The subjective and objective evidence of high- and low-grade carcinomas used in our study require more than frozen-section evaluation and as such cannot serve as a complete intra-operative guide to the surgeon. Furthermore, the fallibility of even such a post-hoc grading system is evident by the occasional patient with a low grade carcinoma who follows the same post-operative course as a patient with a high grade carcinoma (Chong et al, 1974; Eneroth et al, 1966). Whether or not pre-operative DNA content assessment is practical is conjectural. From the work of Eneroth and Zetterberg (1976), it is sug-

338 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE

gested that the DNA content of acinic cell carcinoma is equivalent to that of cells from a poorly differentiated adenocarcinoma. While this finding argues strongly against the existence of a totally benign variety of acinic cell tumor, there is further evidence that the DNA content correlates best with the invasiveness of the carcinoma. Ambiguity over the malignancy of acinic cell neoplasms has had its counterpart in the therapeutic philosophy applied to the management of these neoplasms. Eneroth et al. (1966) have recommended the most aggressive, and Godwin et al. (1954), the most conservative forms of treatment. The former advocate total parotidectomy and neck dissection. The latter suggest the 'best' treatment is an excision of the tumor with a margin of parotid gland. Considering the tumors as essentially benign, Abrams et al. (1965) recommend 'lobectomy for lesions in the superficial lobe or tail of the parotid gland or total parotidectomy for those in deeper portions'. The fate of the facial nerve in patients with these carcinomas has also been debated. Grage et al. (1961) advise routine VII nerve resection. Others sacrifice the nerve only when there is clinical or gross involvement (Chong et al, 1974; Spiro et al, 1975; Woods et al, 1977). The place of neck dissection is also controversial. Only Eneroth and his associates (1966) have recommended a routine radical neck dissection. It is of interest that Fox et al. (1963) advise against radical neck dissection even though that judgment was made after a study of a series in which the frequency of metastases to cervical lymph nodes (3/46 cases) was nearly equal to that occurring in Eneroth's (1966) series (4/47 cases). Based on our data and that provided by Chong et al. (1974) and Abrams et al. (1965), enucleation and local excision of these carcinomas is to be condemned. An 85 per cent recurrence rate in our series and a 67 per cent rate in Chong's (1974) series should underscore that opinion. Recurrences are markedly reduced by total parotidectomy, but as is evident in our study and that of Chong et al. (1974) the capriciousness of the neoplasm, clearly enhanced by failure of primary management, mitigates statements concerning the ultimate prognosis in these patients.- In the final analysis, it is the clinical stage of the carcinoma which determines prognosis (Spiro et al, 1975). It would seem quite clear that the best opportunity for cure lies in complete surgical removal of the neoplasm at the time of initial treatment. For this, a total parotidectomy is the procedure of choice. This is predicted on the unpredictable invasiveness of the carcinoma, high grade foci that occur in a significant number of the carcinomas and a potential for multifocal origin of some of the tumors. A relatively low incidence of metastases to the cervical lymph nodes (approximately 10 per cent) does not warrant elective neck dissection. The management of the facial nerve cannot be predicted on the basis of a retrospective analysis of the behavior of these carcinomas. That decision is best left to the surgeon who in each patient bases his decision on clinical and/or gross findings.

ACINIC CELL CARCINOMA

339

The full malignant potential of acinic cell carcinoma can only be realized in studies with long term follow-up. Given that, there is clear evidence that these neoplasms exhibit a higher degree of malignancy than is usually ascribed to them. It has also been demonstrated that both the recurrences and appearance of metastases may be very late (Chong et al., 191A; Grage et al, 1961; Spiro et al, 1975). Four of the eleven patients studied by Grage et al. (1961) died of their disease, but the average survival before death was 13 years. Eneroth et al. (1966) record a determinate survival at 5 years of 90 per cent; declining to 56 per cent at 20 years. Of the 69 patients with acinic cell carcinoma reported by Chong et al. (1974) 16 died of their disease. The influence of primary treatment on prognosis is quite clear in the latter study. All deaths occurred in patients who had local excision. Summary

A clinicopathologic study of 35 cases of acinic cell carcinomas is presented. Complete follow-up information was available on 31 patients (average period, 7-5 years). The malignant potential of these neoplasms is affirmed by observations that metastases (local and distant) occurred in nine patients and that eight patients died as a consequence of their carcinoma. Retrospective classification of the carcinomas into high and low grade lesions correlated well with ultimate biologic behavior but is unlikely to be successful in an intra-operative (frozen section) mode. This limitation is due to: (a) sampling limitations at the time of primary surgery and (b) the malignant behavior of the occasional low grade carcinoma. Histologic features characterizing high grade carcinomas are local aggressive infiltration and areas of the tumor that appear analogous to the embryonic and post-embryonic terminal tubules and intercalated ducts. The best opportunity for cure of these neoplasms lies in their complete surgical removal at the time of initial treatment. For this, a total parotidectomy is the procedure of choice. Enucleation and local excision is to be condemned. REFERENCES ABRAMS, A. M., CORNYN, J., SCOFIELD, H. H., and HANSEN, L. S. (1965) Cancer, 18,1145. AMERICAN JOINT COMMITTEE FOR CANCER STAGING AND END RESULTS REPORTING: Manual for

Staging of Cancer (1977), p. 25. BATSAKIS, J. G., WOZNIAK, K. D., and REGEZI, J. A. (1977) Journal of Oral Surgery, 35, 904. BEAHRS, O. H., WOOLNER, L. B., CARVETH, S. W., and DEVINE, K. D. (1960) Archives ofSurgery,

90, 890. BLOOM, G. D., and HENRIKSSON, R. (1977) Journal of Laryngology and Otology, 91, 947. BUXTON, R. W., MAXWELL, J. H., and FRENCH, A. J. (1953) Surgery, Gynecology and Obstetrics,

97,401. CHONG, G. C , BEAHRS, O. H., and WOOLNER, L. B. (1974) Surgery, Gynecology and Obstetrics,

138, 65. CUTLER, L. S., and CHAUDHRY, A. P. (1974) Developmental Biology, 41, 31. ENEROTH, C.-M., HAMBERGER, C. A., and JACKOBSSON, P. A. (1966) Annals of Otology, 75,780. ENEROTH, C.-M., JAKOBSSON, P. A., and BLANCK, C. (1966) Cancer, 19,1761.

ENEROTH, C.-M., and ZETTERBERG, A. (1976) Ada Otolaryngologica, 81, 489.

340 J. G. BATSAKIS, E. K. CHINN, T. A. WEIMERT, W. P. WORK AND C. J. KRAUSE ERLANDSON, R. A., and TANDLER, B, (1972) Archives of Pathology, 93, 130. EVANS, R. W., and CRUICKSHANK, A. H. (1970) Epithelial Tumors of the Salivary Glands. W. B. Saunders Co., Philadelphia. FOOTE, F. W., JR., and FRAZELL, E. L. (1953) Cancer 6,1065. Fox, N. M., REMINE, W. H., and WOOLNER, L. B. (1963) American Journal ofSurgery, 106,860. GODWIN, J. T., FOOTE, F. W., and FRAZELL, E. L. (1954) American Journal of Pathology, 30,

465. GORLIN, R. J., and CHAUDHRY, A. (1957) Journal of Oral Surgery, 15, 304. GRAGE, T. B., LOBER, P. H., and ARHELGER, S. W. (1961) American Journal of Surgery, 102,765. HANKS, C. T., and CHAUDHRY, A. P. (1971) American Journal of Anatomy, 130,195. KROLLS, S. O., TRODAHL, J. N., and BOYERS, R. C. (1972) Cancer, 30,459.

LEVIN, J. M., ROBINSON, D. W., and LIN, F. (1975) Archives of Surgery, 110, 64. PIERCE, G. B. (1974) American Journal of Pathology, 77, 103. REDMAN, R. S., and SREEBNY, L. M. (1970) Anatomical Record, 168, 127. REGEZI, J. A., and BATSAKB, J. G. (1977) Otolaryngologic Clinics of North America, 10, 297. SHARKEY, F. E. (1977) American Journal of Clinical Pathology, 67, 272. Spmo, R. H., Huvos, A. G., and STRONG, E. W. (1975) American Journal of Clinical Pathology, 130, 452. THACKRAY, A. C , and SOBIN, L. H. (1972) W.H.O. Geneva. WOODS, J. E., WEILAND, L. H., CHONG, G. C , and IRONS, G. B. (1977) Surgical Clinics ofNorth

America, 57, 565. Requests for reprints to: John G. Batsakis, M.D., Department of Pathology, The University of Michigan Medical School, Ann Abor, Michigan 48109.

Acinic cell carcinoma: a clinicopathologic study of thirty-five cases.

The Journal of Laryngology and Otology April 1979. Vol. 93. pp. 325-340 Acinic cell carcinoma: A clinicopathologic study of thirty-five cases By JOHN...
2MB Sizes 0 Downloads 0 Views