Inf. J Rudiaion Oncology Bid. Phys., Vd Printed in the U.S.A. All rights reserved.
I 8, pp. 4 13-4 I7
0360-3016/90 $3.00 + .OO copyright 0 1990 Pergamon Press plc
??Brief Communication
SALIVARY
GLAND
TUMORS
OF THE ORAL CAVITY
Luu TRAN, M.D.,’ JONAS SIDRYS, M.D.,2 AHMAD SADEGHI, M.D., ’ NANCY ELLERBROEK, M.D.,3 DAVID HANSON, M.D.4 AND ROBERT G. PARKER, M.D.2 ‘Radiation Therapy Service, 4ENT Service, VA Medical Center, West Los Angeles, CA; ‘Department of Radiation Oncology, University of California, Los Angeles, CA; and ‘Division of Radiotherapy, MD Anderson Hospital and Tumor Institute, Houston, TX Between 1961 and 1985, 62 patients with malignant salivary gland tumors of the oral cavity underwent surgery with curative intent at the University of California, Los Angeles (UCLA) Medical center. All patients had a minimum follow-up of 2 years. Fifty of 62 (81%) patients presented with Tl-2 primary tumors. The tumors arose from the palate in #41/62 (66%) patients. The most common histologic type was adenoid cystic carcinoma comprising 34 of 62 (55%) cases. Radical resection was performed in 46 cases and wide local excision in 16 patients. Postoperative radiation therapy was used in 24 cases due to advanced stage and/or positive surgical margins. Results of treatment were analyzed by stage of disease, modes of treatment, histology, and surgical extent. Local control of small lesions reached 100% at IO years with either radical resection alone or local excision. With residual tumor at the surgical margins, the incidence of local recurrence was 4/14 (29%) for those who received adjuvant radiation therapy and 5/10 (50%) for thlose who did not. The vast majority of mucoepidermoid carcinomas were early stage and low grade. These lesions had an excellent prognosis with a control rate of 100%. In contrast, there was a 29% (10/34) failure rate for adenoid cystic carcinoma. The 5-, lo-, and 15-year actuarial survivals for the whole group were 94%, 84%, and 73%, respectively. Our results indicate that for early stage disease, wide local excision may offer patients the chance to avoid the cosmetically and functionally debilitating effects of radical surgery without compromising treatment outcome. Adjuvant radiation therapy appears to reduce the local recurrence for those with residual tumor at lthe surgical margins. Oral cavity, Adenoid cystic carcinoma, Radical surgery.
California, Los Angeles (UCLA) Medical Center are presented.
INTRODUCTION Minor salivary glands are located throughout the oral cavity: palate, lip, buccal mucosa, retromolar trigone, tongue. They contribute most of the saliva secreted between mealtimes, which moistens and protects the mucous membranes, facilitates taste and swallowing, and protects against dental caries (7). Tumors can arise from any of these glands. However, they represent only 2-4% of all head and neck cancers. Because of the superficial location of these tumors, patients usually present with limited disease. Consequently, a high success rate has been reported in the treatment of oral salivary gland tumors compared with tumors in other locations (4, 17, 19). Surgery, radical more often than conservative, has been the mainstay of therapy of these tumors. Postoperative radiation therapy is commonly added for advanced and poorly differentiated tumors with favorable results. In this series, the results of 62 cases of malignant salivary gland tumors of the oral cavity treated with curative intent at the University of
METHODS
AND MATERIALS
Between 1961 and 1985, 75 patients with malignant salivary gland tumors of the oral cavity were registered in the UCLA Tumor Registry. Thirteen patients were eliminated from the study because of consultation and treatment for recurrent disease ( lo), lack of follow-up (2), and death in the postoperative period due to myocardial infarction (1). The remaining 62 patients treated primarily with surgery with and without adjuvant radiation therapy are the subject of this report. The mean age of these patients was 50 years with a range of 15 to 77 years. Male and female patients were equally distributed. An asymptomatic mass occurring in over 40 (65%) patients was the most common presenting symptom. The distribution of histologies within each primary site of involvement is listed in Table 1. Tumors in
Accepted for publication 19 July 1989.
Reprint requests to: Luu ‘Tran,M.D. Radiation Therapy Service, W214, Wadsworth VA Medical Center, Los Angeles, CA 90073. 413
414
I. J. Radiation Oncology 0 Biology 0 Physics Table 1. Distribution
of patients
February 1990, Volume 18, Number 2 according
to histology and site Histology
Site
Adenoid cystic ca
Palate Buccal mucosa Floor of mouth Retromolar trigone Lip Tongue
25 4 2 2
Total
34
Adenoca 2
1
ca
Malignant mixed tumor
3
Table 2. Distribution of patients according and modes of treatment
to stage
Tl-2
T3
RS + RT WLE + RT RS only WLE only Histology Adenoidcystic ca Adenocarcinoma Mucoepidermoid ca Malignant mixed tumor
8 7 26 9
9 0 3 0
25 2 22 1
9 1 2
Total
50
12
RS = Radical surgery; WLE = Wide local excision; RT = Radiation therapy.
Total
13 1 1 5 4
1
41 5 3 8 4 1
24
1
62
1
the palate accounted for 41 of 62 (66%) cases. The remaining sites were involved with decreasing frequency in the retromolar trigone (8), buccal mucosa (5), lip (4) floor of mouth (3) and oral tongue (1). Adenoid cystic carcinoma was the predominant histology, comprising 55% (34/62) of the group. The second most frequent histology was mucoepidermoid carcinoma which accounted for 39% (24/62) of the group. All but one of the mucoepidermoid carcinomas were of low or intermediate grade. All patients were retrospectively staged according to the AJC (TNM) staging system used for squamous cell carcinoma of the oral cavity. Overall, 8 1% (50/62) of patients presented with early (T l-2) disease. Adenoid cystic carcinoma accounted for 9 of the 12 patients with T3 disease (Table 2). Two patients with adenoid cystic carcinoma had clinically positive neck nodes and six had perineural invasion. All 62 patients were treated primarily with surgery. Forty-six patients underwent radical resection; 17 of these had postoperative radiation therapy. Sixteen patients underwent wide local excision; seven of these had postoperative radiation therapy. Overall, 24 patients were treated with adjuvant radiation therapy because of advanced disease (9 pt), and/or positive surgical margins ( 14 pt). Eighteen of these 24 patients had adenoid cystic carcinoma, five of whom had perineural invasion. Radical resection consisted of palatectomy with or
Mode of treatment
Mucoepidermoid
without maxillectomy for palatal tumors, and composite resection for tumors of the retromolar trigone, tongue or floor of the mouth. Wide local excision was used primarily for 16 patients with small lesions located in the lip, buccal mucosa, or palate. Nine of these patients had low grade mucoepidermoid carcinoma. Radical neck dissection was performed in four patients; two of these had clinically positive neck adenopathy. Radiation therapy was given with megavoltage equipment, either 60Cobalt or 6 Mev linear accelerator photons. The average dose was 60 Gy with a range of 50 Gy-76 Gy given 5 days a week with daily doses of 180 cGy-225 cGy. The technique was generally the same as for squamous cell carcinoma of the oral cavity except that the neck was not electively treated. Statistical software (3) was used for statistical analysis and the actuarial survival curve was calculated according to the Kaplan-Meier life table method (11). Curves were compared by two tailed log-rank test. The number of patients at risk at a given time is shown in parenthesis in the figures. Patients were followed a minimum of 24 months to a maximum of 24 years. Seventy-nine percent of patients had follow-up of 5 years and 20% had follow-up of more than 20 years. RESULTS Patterns of failure: Overall, 13 patients failed initial treatment. Local failure (recurrence within the primary site with or without other components) occurred in 10 patients, five of whom also had distant metastases. Three patients developed distant metastases; two of these had neck failures despite control of the primary site. Adenoid cystic carcinoma accounted for 10 ofthe 13 failures (77%) (Table 3). The frequency of local control was related to the initial stage of disease and the mode of treatment (Table 4). All seven Tl-2 patients treated with wide local excision and adjuvant radiation therapy achieved local control. On the other hand, 4 of 9 Tl-2 tumors treated with wide local excision alone recurred locally. All four local failures, however, were subsequently salvaged by further surgery with or without adjuvant radiation therapy. Seventeen patients were treated with radical resection followed by postoperative radiation; five developed local
415
Salivary gland tumors of the oral cavity 0 L. TRAN et al. Table 3. Results of treatment No. of patients
Histology
by histology Distant mets
Local failure
NED
Local + distant failure
Adenoid cystic ca Adenoca Muco-epidermoid ca M. mixed tumor
34 3 24 1
24 2 22 1
2 1 2
3
5
Total
62
49
5*
3
5
* All were salvaged by further resection
+/-
irradiation.
Nine patients in this group presented with advanced (T3) disease. Radical resection alone was used in 29 patients, 26 of whom had T l-2 disease. Initial local control in this group was 27/29 (88%). Early stage lesions (Tl-2) were controlled in 44 of 50 patients (88%) after initial treatment and in 49 of 50 patients (98%) overall after salvage treatment compared to 7/ 12 (58%) and 8/12 (67%), respectively, for T3 lesions. The local control was also dependent on residual disease at the surgical margins. Twenty-four patients had positive surgical margins after resection. Tumor control was greater in those who had adjuvant radiation therapy (lo/ 14 or 7 1%) than in those who did not (5/10 or 50%). In contrast, of the 38 patients with negative margins, local control was achieved in 9/10 and 27/28 patients in the combined modality and surgery alone group, respectively. The actuarial NED survival curves by histology are shown in Figure 1. Adenoid cystic carcinomas continued to relapse after 2 years reflecting the long natural course of these cancers. Mucoelpidermoid carcinomas were controlled locally in 22 of 24 patients. The interval between initial treatment and clinical evidence of recurrence varied from 5 months to 10 years with an average relapse interval of 36 months (range 7-120 months) for adenoid cystic carcinoma and 32 months (range 5-60 months) for mucoepidermoid carcin0m.a. The location of the lesion did not appear to affect the possibility of local control. Distant metastasis. A.11 eight patients who developed distant metastases had adenoid cystic carcinomas. Of these, five had concurrent local recurrence and two had neck failure despite 1oca.lcontrol. The most common site of distant metastases wa.s the lung (7), followed by bone (2) and brain ( 1). Three patients are still alive 5 years after the appearance of distant metastases. The other five patients died at an average interval of 4 years after developing metastases.
recurrences.
Table 4. Local recurrence
according
No. of patients RS+RT WLE + RT RS only WLE only
to modes of treatment
Local recurrence
17 7 29 9
RS = Radical surgery; WLE = Wide local excision.
4 0 2 4
Survival. The 5-, lo-, 15-year actuarial survivals for the whole group were 94%, 84%, and 73% (SE values: .032, .079 and .12), respectively. Actuarial survival curves by histology are shown in Figure 2. The 5-, lo-, 15-year survival rates were 89%, 69%, and 50%, respectively (S.E values: .05X, .144 and .210) for adenoid cystic carcinoma and 100% for mucoepidermoid carcinoma. DISCUSSION
The distribution of the tumors in the present series corresponds to that reported in the literature ( 1,5,6, 17) and confirms that the most common primary site of minor salivary gland tumors of the oral cavity is the palate 41/ 62 (66%) and the most common histologic type is adenoid cystic carcinoma 34/62 (55%). Because of the superficial location of the tumors, they are readily detected by the patient. Most lesions are visible by direct inspection and are easy to biopsy. In as much as these cancers are detected when relatively small, a better prognosis can be expected compared to those with tumors in other sites such as the paranasal sinus or ororopharynx. Weisberger et al. (20) compared the treatment results of 37 tumors arising in the palate and 14 tumors arising in the paranasal sinus. Twenty-three of 37 palate tumors were controlled in comparison to 6 of 14 sinus tumors. Spiro and co-workers ( 17) reported the results of 3 11 malignant minor salivary gland tumors, mainly treated by surgery alone. The local control of 198 cases of oral cavity and oropharynx tumors was 68% versus 28% of 58 cases of sinus and nasal cavity
2 3
70-
Y
60-
!
50-
%
40-
a
30-
: E 8
zoIOI 0
1 2
mIIIIIII 4 6
6
10
TIME
Fig. 1. Actuarial
disease-free
12
14
16
16
(years)
survival according
to histology.
416
1. J. Radiation Oncology 0 Biology0 Physics (31
2
60-
j
50-
;
40-
6
30-
8
20-
12,
MUCOEPlDERMOlD CPiRClNOMPl
&DENO,D CYSTK CARClNOMA
IO-
0
I 2
I 4
I 6
I1 6 10
I I 12 14
TIME
I1 I 11 11 16 16 20 22 24 26 26
1 30
(years)
Fig. 2. Actuarial survival curve according to histology.
tumors. Ellis et al. (4) reported the results of 52 patients with minor salivary gland tumors treated with either radiation therapy alone or in combination with surgery. The 5- and 1O-year actuarial survival was 80% for tumors of the oral cavity compared to 45% and 35%, respectively, for tumors of the paranasal sinuses. Previous reports (12, 19) on salivary gland tumors from UCLA include 14 base of tongue tumors and 35 paranasal sinus tumors. Of these, only 4 of 14 and 11 of 35 patients, respectively, were free
February 1990, Volume 18, Number 2
of relapse. In the present series, 49 of 62 (87%) oral cavity lesions have had no evidence of disease at last follow-up. This significantly higher success rate can be attributed to the earlier stage of the lesions at presentation as well as aggressive surgical therapy. Surgery remains the primary treatment of these tumors. However, the surgical extent should be individualized based on anatomical site, size, and histology. There have been no definite conclusions in the literature regarding the superiority of one surgical approach over another. Spiro et al. ( 17) in a series of 292 malignant salivary gland tumors of the oral cavity state that there is no proven benefit to radical surgery compared to local excision alone for small and accessible lesions. In contrast, Olsen et al. (14) in a report of 54 cases of mucoepidermoid carcinoma (most of them small and low grade) of the oral cavity note that wide surgical excision (radical resection) is the treatment of choice for these tumors due to a high rate of local failure associated with local excision. In our series, ultimate local control approached 100% for early lesions regardless of whether they were treated with radical resection or wide local excision. It appears that surgery alone with adequate margins is an effective modality for the treatment of small lesions. Small and low grade lesions can be treated by wide local excision and radical resection should be
Table 5. Summary of reported series Survival (year) Authors
No. of patients
Histology
Treatment
Control rate*
5
10
Leaftedt et al. + ( 13)
29
Adenoid cystic ca
41%
30
All
41% ultimate control 79% 100% 100%
76%
Kadish et al. ( 10)
292
All
Mainly S 14% S + RT 19 s 5S+RT 3 RT 93% s 2%S+RT 5% RT 93% s 6% RT 93% s 4%S+RT 3% RT 16 S 3 RT 30 s 5S+RT 2RT 52 S 2 RT 25 S 13S+RT Mainly S + RT and RT alone 38 S 24S+RT
Spiro et al.* (17)
Spiro et al. * (15)
125
Spiro et al.* (16)
77
Chung et al. (2)
Adenoid cystic ca Adenoca
All (Pa?Ze) All
Weisberger et al. (20) (Pa;la:e) Olsen et al.* (14)
54
Tran et al. (18) Ellis et al. (4) Present series
Muco-epidermoid All
(Pa?ie) 20 62
All All
* Oral cavity site only. + Includes patients initially treated for recurrent diseases. S = Surgery, RT = radiation therapy.
70%
42% cure rate 23% 10 year cure rate 48% 10 year cure rate 63% 100% 70% relapse-free
58% initial 0% control 88% 85%
90% M-E ca 5 1% adenoidcystic with 10 YrDFS
65%
57%
88%
67%
91%
84%
93%
80%
76%
76%
94%
84%
Salivary gland tumors of the oral cavity 0 L. TRAN et al.
reserved for treatment failure. This approach offers patients the chance to avoid the cosmetically and functionally debilitating effects of radical surgery without compromising the treatment outcome. The treatment results of extensive, advanced lesions (T3) in our series is disalppointing with only 58% free of disease. Recently, many reports have established the efficacy of postoperative irradiation in improving local control of salivary gland tumors. Kadish et af. (10) reported the results of 27 salivary gland tumors of the oral cavity treated with radiation therapy alone, combined modality, and surgery alone. The control rates were 3/3, 5/5, and 13/ 19, respectively. Ellis et al. (4) in a report of 52 patients treated either with radiation therapy alone or in combination with surgery, state that the combined approach yielded a superior local control rate as compared with radiation therapy alone (lo/13 vs 2/l 3). Table 5 summarizes the treatment results of oral cavity salivary gland tumors with respect to modes of therapy, control rate, and survival. In the present series, the difference in local control rate between the surgery alone group and the adjuvant radiation therapy group is not statistically significant (Table 4). However, the combined therapy group, as in most retrospective studies, was biased towards patients with poor prognostic factors. Despite this fact, postoperative irradiation did appear to improve the local control in patients with1 positive surgical margins (7 1% vs 50%).
417
Besides using conventional radiation therapy as an adjuvant treatment, fast neutron therapy should be investigated for the management of advanced, inoperable tumors due to the high relative biological effectiveness (RBE) for adenoid cystic carcinoma. Since mid- 1986, the UCLA medical center has been participating in the RTOG protocol for treatment of advanced salivary gland tumors with fast neutrons. The latest data from RTOG (8) indicate an 85% local control rate with neutrons as compared to 33% with photons. Given the success of neutron therapy for inoperable lesions and the failure rate of adjuvant photon irradiation in the positive margin patients (29%) neutron irradiation in conjunction with surgery might be an option for those patients. However, it should be approached with caution since Griffin et al. (9) have reported a high complication rate for patients who undergo surgery following neutron irradiation alone. In summary, complete resection is an effective treatment of salivary gland tumors of the oral cavity. Wide local excision with adequate margins can be used for tumors with favorable prognosis including small lesions and low grade histologies. Radiation therapy should be added in case of positive surgical margins. Advanced lesions should be treated with radical surgery as well as adjuvant radiation therapy. After treatment, patients require careful follow-up because of the tendency for late relapse (especially with adenoid cystic carcinoma) and the high success rate with salvage therapy.
REFERENCES 1. Chaudhry, A. P.; Lab.ay, G. R.; Yamane, G. M.; Jacobs, M. S. Clinical pathologic and histogenetic study of 189 intraoral minor salivary gland tumors. J. Oral Med. 39(2):5877; 1984. 2. Chung, C. K.; Rahman, S. M.; Constable, W. C. Malignant salivary gland tumors of the palate. Arch. Otolaryngol. 104: 501-504; 1978. 3. Dixon, W. J., ed. BMDP statistical software. Los Angeles: University of California Press; 1983:47-62. 4. Ellis, E. R.; Million, R. R.; Mendenhall, W.; Cassini, N. J. The use of radiation therapy in the management of minor salivary gland tumors. Int. J. Radiat. Oncol. Biol. Phys. 15: 613-617; 1988. 5. Everson, J. W.; Lawson, R. A. Tumor of the minor (oropharyngeal) salivary glands: a demographic study of 336 cases. J. Oral Pathol. 14:500-509; 1985. 6. Fu, K.; Leibel, S.; Phillips, T. Carcinoma of the major and minor salivary glands. Cancer 40:2882-2890; 1977. 7. Gates, G. Malignant neoplasm of the minor salivary glands. N. Engl. J. Med. 306:718-721; 1982. 8. Griffin, T. W.; Pajak, ‘T. F.; Hendrickson, F. R.; Maor M. Neutron vs photon irradiation of inoperable salivary gland tumors: results of a RTOG-MRC Cooperative Randomized study. Int. J. Radiat. Oncol. Biol. Phys. 15: 1085-1090; 1988. 9. Griffin, T. W.; Weisberger, E. C.; Laramore, G. E.; Tong, D.; Biasko, J. C. Complications of combined surgery and neutron radiation therapy in patients with advanced carcinoma of the head and neck. Radiology 132: 177-l 78; 1979. 10. Kadish, S. P.; Goodman, M. L.; Wang, C. C. Treatment of
11. 12.
13.
14. 15.
16.
17.
18.
19.
20.
minor salivary gland malignancies of upper food and air passage epithelium. Cancer 29: 102 1- 1026; 1972. Kaplan, E. L.; Meier, P. Nonparametric estimations from incomplete observations. Am. Stat. J. 53:457-480; 1985. Kessler, D. J.; Mickel, R. A.; Calcaterra, T. C. Malignant salivary gland tumors of the base of the tongue. Arch. Otolaryngol. 111:664-666; 1985. Leafstedt, S. W.; Gaeta, J. F.; Sako, K.; Shedd, D. P. Adenoid cystic carcinoma of major and minor salivary glands. Am. J. Surg. 122:756-762; 1971. Olsen, K.; Devine, K. D.; Weiland. Mucoepidermoid carcinoma of the oral cavity. Otolaryngol. 89:783-79 1; 198 1. Spiro, R. H.; Huvos, A. G.; Strong, E. W. Adenoid cystic carcinoma of salivary gland origin: a clinical study of 242 cases. Am. J. Surg. 128:572-580; 1974. Spiro, R. H.; Huvos, A. G.; Strong, E. W. Adenocarcinoma of salivary gland origin: a clinicopathologic study of 204 patients. Am. J. Surg. 144:423-431; 1982. Spiro, R. H.; Koss, L.; Hajdu, S. I.; Strong, E. W. Tumor of minor salivary origin: a clinicopathologic study of 492 cases. Cancer 3 1: 117- 129; 1973. Tran, L.; Sadeghi, A.; Calcaterra, T.; Parker, R. G. Salivary gland tumors of the palate: the UCLA experience. Laryngoscope 97(11):1343-1345; 1987. Tran, L.; Sidrys, J.; Parker, R. G. Salivary gland tumors of the paranasal sinuses and nasal cavity. Am. J. Clin. Oncol. 12(5):387-392; 1989. Weisberger, E.; Luna, M. A.; Guillamondegui, 0. M. Salivary gland cancers of the palate. Am. J. Surg. 138:584-587; 1979.
I 8, pp. 4 13-4 I7
0360-3016/90 $3.00 + .OO copyright 0 1990 Pergamon Press plc
??Brief Communication
SALIVARY
GLAND
TUMORS
OF THE ORAL CAVITY
Luu TRAN, M.D.,’ JONAS SIDRYS, M.D.,2 AHMAD SADEGHI, M.D., ’ NANCY ELLERBROEK, M.D.,3 DAVID HANSON, M.D.4 AND ROBERT G. PARKER, M.D.2 ‘Radiation Therapy Service, 4ENT Service, VA Medical Center, West Los Angeles, CA; ‘Department of Radiation Oncology, University of California, Los Angeles, CA; and ‘Division of Radiotherapy, MD Anderson Hospital and Tumor Institute, Houston, TX Between 1961 and 1985, 62 patients with malignant salivary gland tumors of the oral cavity underwent surgery with curative intent at the University of California, Los Angeles (UCLA) Medical center. All patients had a minimum follow-up of 2 years. Fifty of 62 (81%) patients presented with Tl-2 primary tumors. The tumors arose from the palate in #41/62 (66%) patients. The most common histologic type was adenoid cystic carcinoma comprising 34 of 62 (55%) cases. Radical resection was performed in 46 cases and wide local excision in 16 patients. Postoperative radiation therapy was used in 24 cases due to advanced stage and/or positive surgical margins. Results of treatment were analyzed by stage of disease, modes of treatment, histology, and surgical extent. Local control of small lesions reached 100% at IO years with either radical resection alone or local excision. With residual tumor at the surgical margins, the incidence of local recurrence was 4/14 (29%) for those who received adjuvant radiation therapy and 5/10 (50%) for thlose who did not. The vast majority of mucoepidermoid carcinomas were early stage and low grade. These lesions had an excellent prognosis with a control rate of 100%. In contrast, there was a 29% (10/34) failure rate for adenoid cystic carcinoma. The 5-, lo-, and 15-year actuarial survivals for the whole group were 94%, 84%, and 73%, respectively. Our results indicate that for early stage disease, wide local excision may offer patients the chance to avoid the cosmetically and functionally debilitating effects of radical surgery without compromising treatment outcome. Adjuvant radiation therapy appears to reduce the local recurrence for those with residual tumor at lthe surgical margins. Oral cavity, Adenoid cystic carcinoma, Radical surgery.
California, Los Angeles (UCLA) Medical Center are presented.
INTRODUCTION Minor salivary glands are located throughout the oral cavity: palate, lip, buccal mucosa, retromolar trigone, tongue. They contribute most of the saliva secreted between mealtimes, which moistens and protects the mucous membranes, facilitates taste and swallowing, and protects against dental caries (7). Tumors can arise from any of these glands. However, they represent only 2-4% of all head and neck cancers. Because of the superficial location of these tumors, patients usually present with limited disease. Consequently, a high success rate has been reported in the treatment of oral salivary gland tumors compared with tumors in other locations (4, 17, 19). Surgery, radical more often than conservative, has been the mainstay of therapy of these tumors. Postoperative radiation therapy is commonly added for advanced and poorly differentiated tumors with favorable results. In this series, the results of 62 cases of malignant salivary gland tumors of the oral cavity treated with curative intent at the University of
METHODS
AND MATERIALS
Between 1961 and 1985, 75 patients with malignant salivary gland tumors of the oral cavity were registered in the UCLA Tumor Registry. Thirteen patients were eliminated from the study because of consultation and treatment for recurrent disease ( lo), lack of follow-up (2), and death in the postoperative period due to myocardial infarction (1). The remaining 62 patients treated primarily with surgery with and without adjuvant radiation therapy are the subject of this report. The mean age of these patients was 50 years with a range of 15 to 77 years. Male and female patients were equally distributed. An asymptomatic mass occurring in over 40 (65%) patients was the most common presenting symptom. The distribution of histologies within each primary site of involvement is listed in Table 1. Tumors in
Accepted for publication 19 July 1989.
Reprint requests to: Luu ‘Tran,M.D. Radiation Therapy Service, W214, Wadsworth VA Medical Center, Los Angeles, CA 90073. 413
414
I. J. Radiation Oncology 0 Biology 0 Physics Table 1. Distribution
of patients
February 1990, Volume 18, Number 2 according
to histology and site Histology
Site
Adenoid cystic ca
Palate Buccal mucosa Floor of mouth Retromolar trigone Lip Tongue
25 4 2 2
Total
34
Adenoca 2
1
ca
Malignant mixed tumor
3
Table 2. Distribution of patients according and modes of treatment
to stage
Tl-2
T3
RS + RT WLE + RT RS only WLE only Histology Adenoidcystic ca Adenocarcinoma Mucoepidermoid ca Malignant mixed tumor
8 7 26 9
9 0 3 0
25 2 22 1
9 1 2
Total
50
12
RS = Radical surgery; WLE = Wide local excision; RT = Radiation therapy.
Total
13 1 1 5 4
1
41 5 3 8 4 1
24
1
62
1
the palate accounted for 41 of 62 (66%) cases. The remaining sites were involved with decreasing frequency in the retromolar trigone (8), buccal mucosa (5), lip (4) floor of mouth (3) and oral tongue (1). Adenoid cystic carcinoma was the predominant histology, comprising 55% (34/62) of the group. The second most frequent histology was mucoepidermoid carcinoma which accounted for 39% (24/62) of the group. All but one of the mucoepidermoid carcinomas were of low or intermediate grade. All patients were retrospectively staged according to the AJC (TNM) staging system used for squamous cell carcinoma of the oral cavity. Overall, 8 1% (50/62) of patients presented with early (T l-2) disease. Adenoid cystic carcinoma accounted for 9 of the 12 patients with T3 disease (Table 2). Two patients with adenoid cystic carcinoma had clinically positive neck nodes and six had perineural invasion. All 62 patients were treated primarily with surgery. Forty-six patients underwent radical resection; 17 of these had postoperative radiation therapy. Sixteen patients underwent wide local excision; seven of these had postoperative radiation therapy. Overall, 24 patients were treated with adjuvant radiation therapy because of advanced disease (9 pt), and/or positive surgical margins ( 14 pt). Eighteen of these 24 patients had adenoid cystic carcinoma, five of whom had perineural invasion. Radical resection consisted of palatectomy with or
Mode of treatment
Mucoepidermoid
without maxillectomy for palatal tumors, and composite resection for tumors of the retromolar trigone, tongue or floor of the mouth. Wide local excision was used primarily for 16 patients with small lesions located in the lip, buccal mucosa, or palate. Nine of these patients had low grade mucoepidermoid carcinoma. Radical neck dissection was performed in four patients; two of these had clinically positive neck adenopathy. Radiation therapy was given with megavoltage equipment, either 60Cobalt or 6 Mev linear accelerator photons. The average dose was 60 Gy with a range of 50 Gy-76 Gy given 5 days a week with daily doses of 180 cGy-225 cGy. The technique was generally the same as for squamous cell carcinoma of the oral cavity except that the neck was not electively treated. Statistical software (3) was used for statistical analysis and the actuarial survival curve was calculated according to the Kaplan-Meier life table method (11). Curves were compared by two tailed log-rank test. The number of patients at risk at a given time is shown in parenthesis in the figures. Patients were followed a minimum of 24 months to a maximum of 24 years. Seventy-nine percent of patients had follow-up of 5 years and 20% had follow-up of more than 20 years. RESULTS Patterns of failure: Overall, 13 patients failed initial treatment. Local failure (recurrence within the primary site with or without other components) occurred in 10 patients, five of whom also had distant metastases. Three patients developed distant metastases; two of these had neck failures despite control of the primary site. Adenoid cystic carcinoma accounted for 10 ofthe 13 failures (77%) (Table 3). The frequency of local control was related to the initial stage of disease and the mode of treatment (Table 4). All seven Tl-2 patients treated with wide local excision and adjuvant radiation therapy achieved local control. On the other hand, 4 of 9 Tl-2 tumors treated with wide local excision alone recurred locally. All four local failures, however, were subsequently salvaged by further surgery with or without adjuvant radiation therapy. Seventeen patients were treated with radical resection followed by postoperative radiation; five developed local
415
Salivary gland tumors of the oral cavity 0 L. TRAN et al. Table 3. Results of treatment No. of patients
Histology
by histology Distant mets
Local failure
NED
Local + distant failure
Adenoid cystic ca Adenoca Muco-epidermoid ca M. mixed tumor
34 3 24 1
24 2 22 1
2 1 2
3
5
Total
62
49
5*
3
5
* All were salvaged by further resection
+/-
irradiation.
Nine patients in this group presented with advanced (T3) disease. Radical resection alone was used in 29 patients, 26 of whom had T l-2 disease. Initial local control in this group was 27/29 (88%). Early stage lesions (Tl-2) were controlled in 44 of 50 patients (88%) after initial treatment and in 49 of 50 patients (98%) overall after salvage treatment compared to 7/ 12 (58%) and 8/12 (67%), respectively, for T3 lesions. The local control was also dependent on residual disease at the surgical margins. Twenty-four patients had positive surgical margins after resection. Tumor control was greater in those who had adjuvant radiation therapy (lo/ 14 or 7 1%) than in those who did not (5/10 or 50%). In contrast, of the 38 patients with negative margins, local control was achieved in 9/10 and 27/28 patients in the combined modality and surgery alone group, respectively. The actuarial NED survival curves by histology are shown in Figure 1. Adenoid cystic carcinomas continued to relapse after 2 years reflecting the long natural course of these cancers. Mucoelpidermoid carcinomas were controlled locally in 22 of 24 patients. The interval between initial treatment and clinical evidence of recurrence varied from 5 months to 10 years with an average relapse interval of 36 months (range 7-120 months) for adenoid cystic carcinoma and 32 months (range 5-60 months) for mucoepidermoid carcin0m.a. The location of the lesion did not appear to affect the possibility of local control. Distant metastasis. A.11 eight patients who developed distant metastases had adenoid cystic carcinomas. Of these, five had concurrent local recurrence and two had neck failure despite 1oca.lcontrol. The most common site of distant metastases wa.s the lung (7), followed by bone (2) and brain ( 1). Three patients are still alive 5 years after the appearance of distant metastases. The other five patients died at an average interval of 4 years after developing metastases.
recurrences.
Table 4. Local recurrence
according
No. of patients RS+RT WLE + RT RS only WLE only
to modes of treatment
Local recurrence
17 7 29 9
RS = Radical surgery; WLE = Wide local excision.
4 0 2 4
Survival. The 5-, lo-, 15-year actuarial survivals for the whole group were 94%, 84%, and 73% (SE values: .032, .079 and .12), respectively. Actuarial survival curves by histology are shown in Figure 2. The 5-, lo-, 15-year survival rates were 89%, 69%, and 50%, respectively (S.E values: .05X, .144 and .210) for adenoid cystic carcinoma and 100% for mucoepidermoid carcinoma. DISCUSSION
The distribution of the tumors in the present series corresponds to that reported in the literature ( 1,5,6, 17) and confirms that the most common primary site of minor salivary gland tumors of the oral cavity is the palate 41/ 62 (66%) and the most common histologic type is adenoid cystic carcinoma 34/62 (55%). Because of the superficial location of the tumors, they are readily detected by the patient. Most lesions are visible by direct inspection and are easy to biopsy. In as much as these cancers are detected when relatively small, a better prognosis can be expected compared to those with tumors in other sites such as the paranasal sinus or ororopharynx. Weisberger et al. (20) compared the treatment results of 37 tumors arising in the palate and 14 tumors arising in the paranasal sinus. Twenty-three of 37 palate tumors were controlled in comparison to 6 of 14 sinus tumors. Spiro and co-workers ( 17) reported the results of 3 11 malignant minor salivary gland tumors, mainly treated by surgery alone. The local control of 198 cases of oral cavity and oropharynx tumors was 68% versus 28% of 58 cases of sinus and nasal cavity
2 3
70-
Y
60-
!
50-
%
40-
a
30-
: E 8
zoIOI 0
1 2
mIIIIIII 4 6
6
10
TIME
Fig. 1. Actuarial
disease-free
12
14
16
16
(years)
survival according
to histology.
416
1. J. Radiation Oncology 0 Biology0 Physics (31
2
60-
j
50-
;
40-
6
30-
8
20-
12,
MUCOEPlDERMOlD CPiRClNOMPl
&DENO,D CYSTK CARClNOMA
IO-
0
I 2
I 4
I 6
I1 6 10
I I 12 14
TIME
I1 I 11 11 16 16 20 22 24 26 26
1 30
(years)
Fig. 2. Actuarial survival curve according to histology.
tumors. Ellis et al. (4) reported the results of 52 patients with minor salivary gland tumors treated with either radiation therapy alone or in combination with surgery. The 5- and 1O-year actuarial survival was 80% for tumors of the oral cavity compared to 45% and 35%, respectively, for tumors of the paranasal sinuses. Previous reports (12, 19) on salivary gland tumors from UCLA include 14 base of tongue tumors and 35 paranasal sinus tumors. Of these, only 4 of 14 and 11 of 35 patients, respectively, were free
February 1990, Volume 18, Number 2
of relapse. In the present series, 49 of 62 (87%) oral cavity lesions have had no evidence of disease at last follow-up. This significantly higher success rate can be attributed to the earlier stage of the lesions at presentation as well as aggressive surgical therapy. Surgery remains the primary treatment of these tumors. However, the surgical extent should be individualized based on anatomical site, size, and histology. There have been no definite conclusions in the literature regarding the superiority of one surgical approach over another. Spiro et al. ( 17) in a series of 292 malignant salivary gland tumors of the oral cavity state that there is no proven benefit to radical surgery compared to local excision alone for small and accessible lesions. In contrast, Olsen et al. (14) in a report of 54 cases of mucoepidermoid carcinoma (most of them small and low grade) of the oral cavity note that wide surgical excision (radical resection) is the treatment of choice for these tumors due to a high rate of local failure associated with local excision. In our series, ultimate local control approached 100% for early lesions regardless of whether they were treated with radical resection or wide local excision. It appears that surgery alone with adequate margins is an effective modality for the treatment of small lesions. Small and low grade lesions can be treated by wide local excision and radical resection should be
Table 5. Summary of reported series Survival (year) Authors
No. of patients
Histology
Treatment
Control rate*
5
10
Leaftedt et al. + ( 13)
29
Adenoid cystic ca
41%
30
All
41% ultimate control 79% 100% 100%
76%
Kadish et al. ( 10)
292
All
Mainly S 14% S + RT 19 s 5S+RT 3 RT 93% s 2%S+RT 5% RT 93% s 6% RT 93% s 4%S+RT 3% RT 16 S 3 RT 30 s 5S+RT 2RT 52 S 2 RT 25 S 13S+RT Mainly S + RT and RT alone 38 S 24S+RT
Spiro et al.* (17)
Spiro et al. * (15)
125
Spiro et al.* (16)
77
Chung et al. (2)
Adenoid cystic ca Adenoca
All (Pa?Ze) All
Weisberger et al. (20) (Pa;la:e) Olsen et al.* (14)
54
Tran et al. (18) Ellis et al. (4) Present series
Muco-epidermoid All
(Pa?ie) 20 62
All All
* Oral cavity site only. + Includes patients initially treated for recurrent diseases. S = Surgery, RT = radiation therapy.
70%
42% cure rate 23% 10 year cure rate 48% 10 year cure rate 63% 100% 70% relapse-free
58% initial 0% control 88% 85%
90% M-E ca 5 1% adenoidcystic with 10 YrDFS
65%
57%
88%
67%
91%
84%
93%
80%
76%
76%
94%
84%
Salivary gland tumors of the oral cavity 0 L. TRAN et al.
reserved for treatment failure. This approach offers patients the chance to avoid the cosmetically and functionally debilitating effects of radical surgery without compromising the treatment outcome. The treatment results of extensive, advanced lesions (T3) in our series is disalppointing with only 58% free of disease. Recently, many reports have established the efficacy of postoperative irradiation in improving local control of salivary gland tumors. Kadish et af. (10) reported the results of 27 salivary gland tumors of the oral cavity treated with radiation therapy alone, combined modality, and surgery alone. The control rates were 3/3, 5/5, and 13/ 19, respectively. Ellis et al. (4) in a report of 52 patients treated either with radiation therapy alone or in combination with surgery, state that the combined approach yielded a superior local control rate as compared with radiation therapy alone (lo/13 vs 2/l 3). Table 5 summarizes the treatment results of oral cavity salivary gland tumors with respect to modes of therapy, control rate, and survival. In the present series, the difference in local control rate between the surgery alone group and the adjuvant radiation therapy group is not statistically significant (Table 4). However, the combined therapy group, as in most retrospective studies, was biased towards patients with poor prognostic factors. Despite this fact, postoperative irradiation did appear to improve the local control in patients with1 positive surgical margins (7 1% vs 50%).
417
Besides using conventional radiation therapy as an adjuvant treatment, fast neutron therapy should be investigated for the management of advanced, inoperable tumors due to the high relative biological effectiveness (RBE) for adenoid cystic carcinoma. Since mid- 1986, the UCLA medical center has been participating in the RTOG protocol for treatment of advanced salivary gland tumors with fast neutrons. The latest data from RTOG (8) indicate an 85% local control rate with neutrons as compared to 33% with photons. Given the success of neutron therapy for inoperable lesions and the failure rate of adjuvant photon irradiation in the positive margin patients (29%) neutron irradiation in conjunction with surgery might be an option for those patients. However, it should be approached with caution since Griffin et al. (9) have reported a high complication rate for patients who undergo surgery following neutron irradiation alone. In summary, complete resection is an effective treatment of salivary gland tumors of the oral cavity. Wide local excision with adequate margins can be used for tumors with favorable prognosis including small lesions and low grade histologies. Radiation therapy should be added in case of positive surgical margins. Advanced lesions should be treated with radical surgery as well as adjuvant radiation therapy. After treatment, patients require careful follow-up because of the tendency for late relapse (especially with adenoid cystic carcinoma) and the high success rate with salvage therapy.
REFERENCES 1. Chaudhry, A. P.; Lab.ay, G. R.; Yamane, G. M.; Jacobs, M. S. Clinical pathologic and histogenetic study of 189 intraoral minor salivary gland tumors. J. Oral Med. 39(2):5877; 1984. 2. Chung, C. K.; Rahman, S. M.; Constable, W. C. Malignant salivary gland tumors of the palate. Arch. Otolaryngol. 104: 501-504; 1978. 3. Dixon, W. J., ed. BMDP statistical software. Los Angeles: University of California Press; 1983:47-62. 4. Ellis, E. R.; Million, R. R.; Mendenhall, W.; Cassini, N. J. The use of radiation therapy in the management of minor salivary gland tumors. Int. J. Radiat. Oncol. Biol. Phys. 15: 613-617; 1988. 5. Everson, J. W.; Lawson, R. A. Tumor of the minor (oropharyngeal) salivary glands: a demographic study of 336 cases. J. Oral Pathol. 14:500-509; 1985. 6. Fu, K.; Leibel, S.; Phillips, T. Carcinoma of the major and minor salivary glands. Cancer 40:2882-2890; 1977. 7. Gates, G. Malignant neoplasm of the minor salivary glands. N. Engl. J. Med. 306:718-721; 1982. 8. Griffin, T. W.; Pajak, ‘T. F.; Hendrickson, F. R.; Maor M. Neutron vs photon irradiation of inoperable salivary gland tumors: results of a RTOG-MRC Cooperative Randomized study. Int. J. Radiat. Oncol. Biol. Phys. 15: 1085-1090; 1988. 9. Griffin, T. W.; Weisberger, E. C.; Laramore, G. E.; Tong, D.; Biasko, J. C. Complications of combined surgery and neutron radiation therapy in patients with advanced carcinoma of the head and neck. Radiology 132: 177-l 78; 1979. 10. Kadish, S. P.; Goodman, M. L.; Wang, C. C. Treatment of
11. 12.
13.
14. 15.
16.
17.
18.
19.
20.
minor salivary gland malignancies of upper food and air passage epithelium. Cancer 29: 102 1- 1026; 1972. Kaplan, E. L.; Meier, P. Nonparametric estimations from incomplete observations. Am. Stat. J. 53:457-480; 1985. Kessler, D. J.; Mickel, R. A.; Calcaterra, T. C. Malignant salivary gland tumors of the base of the tongue. Arch. Otolaryngol. 111:664-666; 1985. Leafstedt, S. W.; Gaeta, J. F.; Sako, K.; Shedd, D. P. Adenoid cystic carcinoma of major and minor salivary glands. Am. J. Surg. 122:756-762; 1971. Olsen, K.; Devine, K. D.; Weiland. Mucoepidermoid carcinoma of the oral cavity. Otolaryngol. 89:783-79 1; 198 1. Spiro, R. H.; Huvos, A. G.; Strong, E. W. Adenoid cystic carcinoma of salivary gland origin: a clinical study of 242 cases. Am. J. Surg. 128:572-580; 1974. Spiro, R. H.; Huvos, A. G.; Strong, E. W. Adenocarcinoma of salivary gland origin: a clinicopathologic study of 204 patients. Am. J. Surg. 144:423-431; 1982. Spiro, R. H.; Koss, L.; Hajdu, S. I.; Strong, E. W. Tumor of minor salivary origin: a clinicopathologic study of 492 cases. Cancer 3 1: 117- 129; 1973. Tran, L.; Sadeghi, A.; Calcaterra, T.; Parker, R. G. Salivary gland tumors of the palate: the UCLA experience. Laryngoscope 97(11):1343-1345; 1987. Tran, L.; Sidrys, J.; Parker, R. G. Salivary gland tumors of the paranasal sinuses and nasal cavity. Am. J. Clin. Oncol. 12(5):387-392; 1989. Weisberger, E.; Luna, M. A.; Guillamondegui, 0. M. Salivary gland cancers of the palate. Am. J. Surg. 138:584-587; 1979.
