Management of Malignant Tumors of the Salivary Glands

ABSTRACT: Results of treatment for patients with salivary gland carcinoma have improved in recent years, most likely due to earlier diagnosis and the use of more effective locoregional therapy. Salivary gland tumors are treated surgically, often in conjunction with postoperative radiation therapy when the tumor is malignant. Good results rest strongly on the performance of an adequate, en bloc initial resection. Radical neck dissection is indicated in patients with obvious cervical metastasis, and limited neck dissection may be appropriate in patients with clinically negative nodes in whom occult nodal involvement is likely. Postoperative radiation therapy should be administered when the tumor is high stage or high grade, the adequacy of the resection is in question, or the tumor has ominous pathologic features. Neutron beam therapy shows promise in controlling locoregional disease but requires further study. No single chemotherapeutic agent or combination regimen has produced consistent results. At present, chemotherapy is clearly indicated only for palliation in symptomatic patients with recurrent and/or unresectable cancers. Patients with salivary gland carcinomas must be followed for long periods, as recurrence may occur a decade or more following therapy. Distant metastasis appears to occur in approximately 20% of patients.[ONCOLOGY 12 (5): 671-683, 1998]

Salivary gland tumors pose a special challenge to clinicians because of their infrequency and remarkable variation in presentation and behavior. The current incidence of malignant salivary tumors in the United States is less than 10 cases per million people. This means that approximately 2,500 new cases are diagnosed each year and that salivary gland neoplasms account for about 7% of all epithelial cancers arising in the upper aerodigestive tract.

In our hospital, 1% of all admissions and 6% to 7% of patients treated on the head and neck service have salivary gland tumors, 70% of which arise in the parotid gland. The submandibular gland is the site of origin in 8% of patients, and the sublingual gland, the most uncommon of the three paired, or major, salivary tumor sites, accounts for only 0.05% of salivary tumors. About 22% of salivary gland tumors originate in the so-called minor salivary glands, the tiny, predominantly mucus-secreting glands that are found everywhere beneath the mucous membranes of the upper aerodigestive tract but are most densely clustered in the palate.[1]

The probability of a malignant diagnosis is less than 25% in patients with parotid gland tumors, about 50% in those with submandibular gland primaries, more than 80% in those with minor salivary lesions, and virtually 100% in those few who present with sublingual gland lesions. It is important to remember that statistics on the distribution of salivary gland tumors and the proportion that are malignant usually derive from the tumor registries of large tertiary care centers, where there is obvious referral bias. In the community hospital setting, virtually all of the salivary gland tumors encountered originate in the parotid and the incidence of malignant tumors is usually lower.

Clinical Presentation

Salivary gland tumors can occur at any age, and incidence does not differ significantly by gender. In our experience, patients with benign lesions tend to be younger than those with carcinomas (median age, 46 vs 54 years). Moreover, low-grade, less aggressive malignant tumors are the rule for younger patients, whereas older patients more often have high-grade or anaplastic tumor types.

Whether these tumors arise in the major salivary glands or the minor glands in the oral cavity or oropharynx, the typical presenting feature is asymptomatic swelling. The fact that swelling may have been present for several years is no assurance of a benign diagnosis. Conversely, pain or rapid growth often, but not invariably, indicates that the tumor is malignant. Patients with a tumor arising in the nasal cavity, paranasal sinuses, hypopharynx, or larynx develop symptoms identical to those described for squamous cell carcinomas occurring in the same sites.

Small, asymptomatic parotid or submandibular gland tumors are clinically indistinguishable from their benign counterparts. Conversely, facial nerve palsy, cervical node enlargement, and skin adherence, in the absence of prior treatment, are virtually certain indicators that a tumor is malignant and usually advanced.

About 10% of parotid tumors arise below the plane of the facial nerve in the so-called deep “lobe.” This is never obvious clinically unless swelling of the palate or tonsil indicates the presence of a retromandibular component. Another unusual source of “parotid” tumors (1%) involves the nubbin of accessory tissue adjacent to Stensen’s duct at the anterior margin of the gland. In our experience, the incidence of malignant salivary tumors at either of these sites is similar to that noted in the rest of the gland.[2,3]

Minor salivary gland tumors typically present as a nodule or mass beneath an intact mucous membrane. As with major salivary gland tumors, small benign minor gland tumors have a similar appearance as their malignant counterparts. When ulceration is present, either related to biopsy, denture irritation, or other trauma, these lesions may be confused with squamous cell carcinomas.

Classification

The histologic classification system now used by most centers is basically a modification of that proposed more than 40 years ago in the classic paper by Foote and Frazell (Table 1).[4] In 1978, Batsakis and Regezi proposed a more detailed classification of epithelial salivary gland tumors that incorporated newer concepts of histogenesis, with emphasis on the role of the myoepithelial cell (Table 2).[5] The second edition of the World Health Organization’s histologic classification of salivary gland tumors, published in 1992, proposed an even more complex histologic breakdown (Table 3).[6]

Although unique subtypes of malignant tumors are better defined in this newest classification, overall it is rather unwieldy. What emerges from all this refinement is the reality that detailed subclassification may be exciting to pathologists but is confusing to clinicians.

In our experience, more than 80% of benign tumors are pleomorphic adenomas. This is the histology most often encountered in the submandibular gland, as well as the parotid gland. In patients with malignant neoplasms, mucoepidermoid carcinoma is the most frequent diagnosis (34%), followed by adenoid cystic carcinoma (22%), adenocarcinoma (18%), and malignant mixed tumor (13%). Less frequently diagnosed malignant salivary neoplasms include acinic cell carcinoma (7%), epidermoid carcinoma (4%), and other anaplastic variants (3%).

There is an interesting correlation between the histologic diagnosis and the site of origin. About 40% of minor salivary gland tumors in our patients involve the palate, by far the most common site; this is also where almost all of the relatively few benign minor salivary tumors in our patients originate. Aside from a few other patients with benign tumors in the lips or nasal cavity, minor salivary tumors arising in other anatomic sites are almost invariably malignant.[7]

With respect to the distribution of malignant salivary tumors, we find that mucoepidermoid carcinoma is the most common diagnosis in the parotid gland, whereas adenoid cystic carcinoma is the malignant tumor most often encountered in submandibular or minor salivary sites.[8] Most adenocarcinomas histologically resemble breast carcinomas of ductal origin, but some have unusual features, characterized as papillary, mucus-secreting, colonic type, clear cell, and so on. More recently, it has been appreciated that some adenocarcinomas arising in the nasal cavity or paranasal sinuses have morphologic features that distinguish them as being of mucus membrane, rather than minor salivary, origin.

Grading

The concept of tumor grading can be traced back to a 1945 report by Stewart et al, in which the term “mucoepidermoid” was first used for a neoplasm that previously had been poorly characterized under a variety of names. For many years thereafter, pathologists argued over whether some of these mucoepidermoid tumors were benign, despite the fact that Stewart et al had divided their patients into “relatively favorable” and “highly unfavorable” groups, with the caveat that the term “benign” was “. . . scarcely ever applicable in the absolute sense.”[9]

The importance of grading salivary gland tumors has become widely accepted. Most centers now categorize mucoepidermoid carcinomas as low, intermediate, or high grade. We have been able to make similar distinctions for many adenocarcinomas despite the bewildering variety of subtypes mentioned above.[10] When we graded acinic cell carcinomas, we designated the uncommon papillocystic carcinoma as a high-grade variant of this tumor type.[11]

Significant differences in tumor behavior are apparent when certain salivary tumor types are subdivided according to histologic grade. Low-grade mucoepidermoid carcinomas, for example, almost never metastasize and typically behave in a relatively benign fashion. Similar, less aggressive growth patterns are evident in patients who have low-grade acinic cell carcinomas or adenocarcinomas.

Experts currently disagree about the value of grading in patients with adenoid cystic carcinoma, an indolent but highly aggressive tumor. Tumors that display a cribriform, rather than a solid, pattern are considered to be less aggressive and more favorable in some centers,[12,13] but our experience indicates that differences in survival based on tumor appearance alone disappear when patients are followed for more than 10 years.[14,15]

Clinicians need to appreciate that not all malignant salivary can be graded. Moreover, different pathologists may disagree about the grade of a given tumor, even when they are using similar criteria. It is important to remember that the classification of salivary gland tumors is an evolving art. In our experience, diagnoses are frequently changed when histologic material is reviewed retrospectively, which confirms that the identification of these tumors can be a formidable challenge even to experienced pathologists.

Clinical Staging

In 1975, we reported our experience with 288 previously untreated patients who underwent definitive treatment for carcinoma of the parotid gland. In addition to the obvious indicators of advanced disease (facial nerve palsy and cervical node involvement), it was apparent that the patients could be categorized by tumor size and local extension. From these data, we proposed a simple staging system that accurately predicted the probability of survival.[16] With modification, the system was subsequently adopted by the American Joint Committee on Cancer (AJCC) in 1978 (Table 4).[17] Subsequent revisions in 1983 and 1988 may have rendered the current AJCC major salivary gland staging system more complex, but few will question that staging is valuable for treatment planning and absolutely essential for a meaningful comparison of end results.

More recently, we reviewed our 45-year experience with 378 previously untreated minor salivary gland cancers in order to verify a long-standing presumption. Using criteria that were identical for squamous cell carcinomas in similar sites, sufficient data were available to retrospectively stage all but 25 of these tumors. Multivariate analysis confirmed that these criteria were just as useful for predicting results in patients who received treatment for minor salivary gland carcinomas in comparable sites.[18]

Patient Evaluation

Experienced clinicians are well aware that any lump near the ear is best considered a parotid gland neoplasm until proven otherwise. Open biopsy of a parotid or submandibular mass is mentioned only to be condemned. Such ill-advised intervention is unnecessary and risks tumor seeding with multifocal recurrence, as well as an unsightly scar.

In patients with minor salivary gland tumors, on the other hand, biopsy is important because the differential diagnosis includes sarcoma, lymphoma, and melanoma, and treatment will vary accordingly. When minor salivary tumors are small, biopsy is preferable to excision unless the latter provides a specimen with margins that will be adequate if the lesion proves to be malignant. Otherwise, the patient will need a reexcision of the biopsy wound, which leads to a larger surgical defect and the potential for unnecessary morbidity.

Role of Fine-Needle Aspiration Biopsy

Through the efforts of Dr. Hayes Martin decades ago, our hospital pioneered the use of needle aspiration for the diagnosis of solid tumors. Despite our early enthusiasm for this technique, we do not routinely perform needle aspiration of parotid masses because we find that a tissue diagnosis is not essential for treatment planning in the patient who presents with a small, mobile mass that is obviously within the substance of the gland. Clearly, fine-needle aspiration biopsy is appropriate when parotid gland origin is uncertain, or when the size and location of the tumor suggest that a facial nerve dissection may prove tedious.

The indication for fine-needle aspiration biopsy of a mass in the submandibular triangle is much more compelling, as relatively few such “lumps” prove to be primary neoplasms of the submandibular gland.[19] Many patients with these masses will have inflammation or neoplasm involving adjacent lymph nodes, and knowledge of the histology is likely to influence the treatment plan.

Although some centers claim great accuracy for fine-needle aspiration biopsy in the diagnosis of salivary tumors, we and others have noted a significant error rate even when frozen-section technique is used on adequate tissue fragments or surgical specimens. Fine-needle aspiration biopsy certainly has a role in selected patients, but we continue to place great reliance on careful clinical assessment.

Role of Imaging Studies

Another controversial area is the role of computed tomography (CT) and magnetic resonance imaging (MRI) in the assessment of patients with salivary gland tumors. Few will question the importance of imaging in assessing a patient with a deep lobe parotid tumor, a high-stage tumor that has invaded adjacent structures, or a high-grade tumor that may be associated with occult cervical metastases. Conversely, CT should not be routine for all patients with parotid masses because it is not likely to change the management of small- to moderate-sized, freely movable tumors that are obviously confined to the gland.

Imaging assumes much more importance in the evaluation of patients with minor salivary gland tumors, particularly those arising in the palate or sinuses. Without it, there is real risk of underestimating tumor extent, as, for example, mistaking the inferior palatal extension of a large antral carcinoma for a palatal primary. A good-quality CT scan may indicate unsuspected skull base involvement by a malignant sinus tumor and raise questions about resectability.

Clearly, the availability of high-quality imaging has greatly enhanced treatment planning in certain patients with malignant salivary gland tumors. However, clinicians must always exercise caution when radiographic findings are inconsistent with the clinical presentation. Even the best radiographs can occasionally be misleading, and good clinical judgment should remain the overriding factor in the treatment decision.

Given current concerns about cost-effectiveness, there is little published evidence that justifies ordering a much more expensive MRI initially when a good-quality CT will often suffice. The well-recognized advantages of MRI include multiplanar imaging capability, the elimination of dental artifact, and the ability to distinguish between tumor and obstructed secretions in patients with sinus lesions.

Treatment

The treatment of salivary gland neoplasms is surgical, often in conjunction with postoperative radiation therapy when the tumor is malignant.

Surgical Treatment of the Primary Tumor

Parotid Gland Tumors—Size and location usually determine the extent of resection of a malignant parotid gland tumor. Most T1 and T2 lesions lateral to the facial nerve are suitable for conventional, superficial parotidectomy with nerve preservation, regardless of the precise histologic diagnosis.

Patients with obvious, high-stage cancers (facial palsy, skin involvement) usually require a total parotidectomy with resection of part or all of the facial nerve. On occasion, extraparotid tumor extension may necessitate skin excision, mandibulectomy, or resection of portions of the maxilla or temporal bone. When a malignant diagnosis is not obvious, however, the facial nerve can usually be spared unless it is involved by, or adherent to the tumor. In our experience, facial nerve dysfunction was apparent initially in 14% of patients with malignant parotid tumors, and partial or complete nerve sacrifice was required in about 30% of parotidectomies performed for carcinoma.

Retromandibular parotid gland tumors usually require facial nerve exposure, or even superficial parotidectomy, before the lesion is resected through a transcervical approach. A few may need a mandibulectomy for access. We prefer a paramedian, rather than a lateral, osteotomy, because the latter will be directly in the portal should postoperative radiation therapy be required. Rare lesions arising in accessory parotid tissue anterior to the gland are best approached through an extended, conventional parotidectomy incision rather than directly through the overlying skin of the cheek. The extent of the tumor will determine how much parotid gland and which nerve branches must be resected.

Should sacrifice of the main trunk or major branches of the facial nerve become necessary, immediate cable grafting is indicated when proximal and distal branches can be identified. Branches of the cervical plexus or the sural nerve can be used. Implantation of a gold weight in the upper eyelid is preferable to tarsorrhaphy for protection of the eye in this setting.

Submandibular Gland Tumors—Gland excision alone is adequate only for those few patients whose tumors are small and confined to the parenchyma. When a carcinoma involves, or extends beyond, the capsule, the addition of a lymphadenectomy may seem to be appropriate but is not oncologically sufficient. An adequate resection should also remove the tumor bed (ie, any adjacent structures in contact with the tumor), which may include the mylohyoid, digastric, and hyoglossus muscles, as well as the hypoglossal and lingual nerves and the marginal branch of the facial nerve.

For lesions that involve the mandibular periosteum or bone, a composite resection may be required in order to secure adequate margins.[19] Whether limited resection and aggressive postoperative radiation therapy can achieve comparable locoregional control with less morbidity remains to be seen.

Minor Salivary Gland Tumors—When lesions arise in the palate, sinuses, pharynx, or larynx, imaging is essential for treatment planning. Conservative local excision usually suffices for patients with small, low-grade malignant tumors arising in the palate. At times, resection of the underlying bone may be necessary to secure an adequate deep margin, even when radiographs fail to show osseous invasion. Larger lesions require either a peroral palatectomy or a conventional subtotal maxillectomy through a cheek flap approach.

In general, resections for cancers in all sites are similar to those required for their malignant squamous counterparts, although adenoid cystic primaries require more generous margins because of their propensity for insidious perineural spread. The microscopic extent of adenoid cystic tumors is always greater than can be appreciated either clinically or radiographically.

Neck Dissection

In our patients with major salivary gland carcinomas, we have found that the overall incidence of cervical node involvement is 26% and that neck failure is uncommon whether or not lymphadenectomy is performed.[8,20] Conventional radical neck dissection is indicated in any patient who has obvious cervical metastasis.

Limited selective neck dissection may be appropriate in patients who are N0 when there is a significant chance of occult nodal involvement, as is the case in some patients with high-grade mucoepidermoid or squamous cell carcinomas (which are associated with 44% and 58% rates of occult nodal disease, respectively, in previously untreated patients). In addition, elective neck dissection may facilitate the resection of a bulky parotid or submandibular gland primary that extends into the upper neck.

Postoperative Radiation Therapy

Adjunctive postoperative irradiation therapy is indicated when a tumor is high stage or high grade and carries a significant risk of locoregional recurrence, when the surgeon is concerned about the adequacy of the resection, or when the pathology report describes ominous features, such as margin involvement or an unusually aggressive histologic appearance. These criteria will automatically include almost all patients with retromandibular, deep lobe tumors because they are seldom resectable with adequate margins.

Most radiation oncologists prefer to start treatment within 6 weeks of the resection and to include at least the upper half of the ipsilateral neck within the treatment field. This may be appropriate when there is no clinical or radiographic evidence of adenopathy, but few will question the need to include the entire neck when metastases are present. Irradiation to the contralateral neck seems pointless in patients with parotid or submandibular gland primaries, considering the negligible incidence of contralateral nodal involvement. Regardless of technique, clinicians need to realize that postoperative irradiation is unlikely to salvage patients who have had inadequate surgery that does not remove all gross tumor.

Several reports in the literature claim that postoperative radiation therapy significantly improves locoregional control.[21-23] As a rule, this conclusion is based on retrospective analyses of relatively small patient groups, often with limited follow-up. Unfortunately, this question is difficult to answer prospectively, considering the infrequent occurrence of malignant salivary gland tumors and the desirability of at least 10 years of posttreatment follow-up.

Our experience indicates that adjunctive radiotherapy is not cost-effective for every patient. We addressed this issue by retrospectively analyzing a small cohort who had received postoperative radiation therapy; these patients were carefully matched by site, gender, histology, grade, and stage with another cohort from prior years who received only surgery. There was no enhancement of locoregional control or survival when irradiation followed resection of stage I or II tumors, and the benefit was limited to those with more advanced disease (stage III or IV). The evidence suggests that high-tumor grade alone is not a compelling indication for postoperative irradiation.[24]

Neutron Therapy—Recently, it has been appreciated that neutrons may be more effective than photons when directed against malignant salivary tumor cells. Whether neutron therapy has been used for residual, recurrent, or unresectable salivary gland carcinoma, preliminary reports describe impressive locoregional control at the expense of significant radiation-related morbidity.[25-27] Neutron beam therapy is clearly a viable option in patients with unresectable tumor. Better definition of the role of this new modality in other settings will require a larger cohort of treated patients and longer follow-up.

Chemotherapy

Although anecdotal reports have described tumor responses to doxorubicin- and platinum-based chemotherapy, to date no single agent or drug combination has produced consistent results. For this reason, the use of chemotherapy as a preoperative adjunct or concurrently with radiation therapy cannot be justified unless the drugs are administered as part of a well-designed clinical trial. As of this writing, the only clear indication for chemotherapy is for palliation in symptomatic patients with recurrent and/or unresectable carcinoma.

Results and Prognostic Factors

Prolonged observation is necessary to appreciate the clinical course of some salivary gland carcinomas. Recurrence is possible a decade or more following the treatment of low-grade malignant tumors. In patients with adenoid cystic carcinoma, disease-related death occurs for as long as they are followed, and the survival curve never levels out.[14]

In our experience, distant metastasis occurs in about 20% of patients with salivary gland carcinomas. As with squamous cell carcinomas, dissemination is more common when lymph nodes are involved. The incidence of distant metastasis is more than 40% in patients with squamous cell or anaplastic carcinomas, about 38% in those with adenoid cystic carcinoma, and varies from 16% to 28% in those with malignant mixed tumors, adenocarcinomas, or intermediate/high-grade mucoepidermoid carcinomas.[8,18]

Overall, cumulative survival rates at 5, 10, and 15 years in our patients with previously untreated major salivary gland carcinomas were 82%, 67%, and 55%, respectively. Similarly, survival rates at 5, 10, and 15 years for those with previously untreated, malignant minor salivary gland tumors were 73%, 56%, and 46%, respectively.[8,18] Before comparing these figures with results achieved elsewhere, it is essential to realize that outcomes can vary widely depending on the stage distribution and the proportion of aggressive tumor types treated.

A comparison of survival by stage showed that results have significantly improved in patients with high-stage tumors (stage III or IV).[8] It seems clear that this improvement occurred after we began to employ adjunctive radiotherapy. Moreover, as noted above, the increase in survival was limited to patients with high-stage disease (Figure 1). Earlier impressions that “cure” rates were lower in patients with submandibular gland or minor salivary carcinomas have not been substantiated in more recent studies, when it became possible to eliminate bias by comparing tumors stage for stage.[8,18]

Significant survival differences became apparent when results were analyzed according to the histologic diagnosis. The lowest survival rate was noted in patients with squamous cell or anaplastic carcinomas, and the highest rate was found in those with low-grade mucoepidermoid or acinic cell carcinomas. As shown in Figure 2, no significant differences were noted when survival curves were plotted for the remaining tumor types (ie, intermediate/high-grade mucoepidermoid carcinoma, adenocarcinoma, adenoid cystic carcinoma and malignant mixed tumor).[8] When survival was compared according to stage or grade, the differences were striking, although it should be remembered that only certain tumor types are gradable (Figure 3).[1]

Determinants of Survival

Univariate analysis of patients with salivary gland carcinoma suggests that survivorship depends on histologic tumor type and grade, anatomic site of origin, and clinical stage. In those with parotid gland cancers, male gender and age > 50 years seem to be significant adverse factors.

Anatomic site can be an important consideration. For example, patients with minor salivary gland carcinomas of sinus origin have a lower survival than their counterparts with parotid gland cancers because a larger proportion of the former have high-stage disease, and because a maxillectomy can be a more formidable technical challenge in terms of tumor clearance than is a standard parotidectomy.

Only by using actuarial methodology and multivariate analysis has it become clear that clinical stage and histologic grade are the only significant determinants of survival. Although both are independent variables that influence treatment results, a recent, retrospective review of 378 patients with previously untreated minor salivary gland carcinomas confirmed that tumor extent (ie, stage) is more important. When patients with mucoepidermoid carcinoma, adenocarcinoma, or adenoid cystic carcinoma were divided into three groups—low grade, high grade but low stage, and high grade and high stage—the survival of patients with high-grade but low-stage tumors was the same as those with low-grade lesions. As would be expected, the outcome was poor when a tumor was both high grade and high stage (Figure 4).[18]

Conclusions

The improvement in treatment results for patients with salivary gland carcinoma noted in recent years can probably be attributed to earlier diagnosis and the more effective locoregional treatment that has been achieved with combination therapy. Hopefully, with better physician education and growing patient awareness, the proportion of early, favorable cases will continue to rise as the need for extended, radical surgery declines.

Conversely, it is important that surgeons avoid inadequate or piecemeal resection with the unrealistic expectation that postoperative radiation therapy will control any tumor that remains. Good results depend heavily on an adequate, en bloc initial resection. Reoperation to resect residual tumor or recurrence almost always causes more morbidity with much less likelihood of complete tumor extirpation.

For parotid tumors, however innocent the presentation, and regardless of a benign fine-needle aspiration biopsy, prudence dictates that the surgeon and patient must always be prepared for the possibility—small as it may be—that part or all of the facial nerve may have to be sacrificed.

A word of caution is appropriate for surgeons who are contemplating extended radical surgery for the unfortunate few patients who present with locally advanced, borderline-resectable salivary gland carcinomas. Whether it be a parotid gland tumor that is ulcerating skin and eroding the temporal bone, or a sinus primary with skull base destruction and extension into the anterior cranial fossa, these tumors are typically high grade and the chance of “cure” is minimal. The ability to repair virtually any surgical defect with free, revascularized tissue has encouraged some to undertake supraradical ablative procedures based on the assumption that palliation is likely even if the resection is incomplete.

Before heroic extirpations are undertaken, the surgeon has a weighty responsibility. As mentioned above, some data suggest that the palliative results of neutron beam irradiation may compare favorably with surgery in patients with borderline-resectable disease. Our group also has had encouraging preliminary results using aggressive, concomitant chemoradiation treatment for unresectable, head and neck squamous cell and non-squamous cell carcinomas.[28] Until this issue is addressed in a prospective, cooperative trial, it is incumbent on the surgeon to be certain that his or her patient is aware of these nonoperative alternatives, as well as the significant risks and very limited expectations following surgical removal of very extensive, borderline-resectable salivary gland carcinomas.

Finally, we need to be aware that distant metastasis all too often defeats us despite successful locoregional treatment. Resection of isolated pulmonary metastases from carcinomas arising in the head and neck may be reasonable in carefully selected patients, but the literature on pulmonary metastastectomy contains minimal information about its role in the treatment of malignant salivary tumors.[29,30]

Thoracotomy to excise a solitary lung metastasis may be worthwhile when the salivary histology is low grade and the disease-free interval is measured in years. In contrast, the outcome is not likely to justify the effort in patients with a high-grade tumor that has metastasized to lung parenchyma within months. In particular, metastatectomy for adenoid cystic carcinoma would seem to be highly questionable. Past experience indicates that solitary pulmonary metastasis is quite unusual in these patients, and some metastatic lesions can remain relatively stable for more than 10 years.[31] In order to establish that metastectomy has improved survival in patients with adenoid cystic carcinoma, longer follow-up is essential, as well as comparison with matched patients whose lung metastases have not been excised.

A more logical and efficient approach to this problem would obviously involve the use of chemotherapy, given preoperatively and/or postoperatively, to reduce the incidence of distant metastasis. Aside from the fact that consistently effective agents are not presently available, we need biomarkers that will allow us to better define the cohort at greatest risk of distant dissemination in order to make this approach cost-effective. Considering that the disappointing experience with chemotherapy in patients with salivary gland carcinomas is almost exclusively limited to those with gross, inoperable or recurrent tumors, the use of appropriate agents in an adjunctive setting is another issue that might be addressed in a prospective, cooperative study.

The Spiro Article Reviewed

Ernest A. Weymuller, Jr., MD, Professor and Chairman, Otolaryngology-Head and Neck Surgery, University of Washington, Seattle, Washington

Dr. Spiro has provided a great service to the oncologic community with this summary of the management of malignant salivary gland tumors. Of particular value is his thorough analysis of an institutional experience large enough to generate meaningful conclusions. Because the points made are so important, this commentary will attempt to highlight them for reemphasis.

Importance of Surgical Judgment and Planning

Most importantly, Dr. Spiro indicates that the impact of surgical judgment and planning cannot be overemphasized. There is no doubt that the first attempt at excision must be the best, and that underoperating on a malignant tumor sets the stage for locoregional recurrence, regardless of the quality of subsequent radiation therapy.

Although I agree with Dr. Spiro’s comments about fine-needle aspiration (FNA) and computed tomographic (CT) scans, surgeons must be aware of their own limitations. The highly experienced senior surgeon may be able to plan and execute salivary gland surgery effectively without these imaging modalities. However, more junior surgeons and those who perform the surgery with less frequency should use every available modality before beginning a procedure. In particular, CT can help predict deep lobe extension in tumors that have a subtle external presentation. Likewise, FNA can identify the occasional lymphoma or Warthin’s tumor that might be managed nonoperatively.

Another reason for using FNA and preoperative imaging is the ever-increasing need to accurately predict operating room time requirements. The difference between a superficial parotidectomy and a complex facial nerve dissection is substantial, and a relatively inexpensive set of tests (FNA and CT) will go a long way toward calculating the amount of time needed in a particular facility.

Another important point relative to preoperative imaging is the adjunctive benefits of magnetic resonance imaging (MRI) scans in assessing such lesions as adenoid cystic carcinoma, especially adenoid cystic carcinoma of the oral cavity and sinonasal regions; perineural spread to the skull base can make a major difference in planning treatment for such lesions and may be unidentifiable without MRI. I have seen more than one case that appeared to be suitable for some sort of surgical resection, which, on preoperative MRI, demonstrated unsuspected disease in the cavernous sinus or trigeminal nerve foramina.

The article also provides information emphasizing the prognostic importance of stage and grade. This is a very important concept for surgeons managing parotid disease.

Management of Aggressive Tumors

Dr. Spiro’s comments regarding the boundaries of the anterior and middle craniofossa as they relate to aggressive salivary malignancy are very appropriate. This is especially true for high-grade malignancies, which are usually rapid in their systemic progression.

We tend to take a less aggressive approach to high-grade tumors involving bone and neural structures of the skull base. We use concomitant chemoradiation in the squamous varieties and neutron irradiation in the nonsquamous varieties. For slower-growing tumors with more definable, discrete margins involving the anterior and middle cranial fossa, we tend to include surgical management and often add radioactive implants or intraoperative radiation therapy to augment locoregional control. Proof of the benefit of these modalities will be a long time coming, as noted by Dr. Spiro.

We generally proceed with immediate reconstruction after facial nerve sacrifice, as described by Dr. Spiro. We have found the temporalis sling to be an effective adjunct to facial nerve grafting and gold weights.

New Insights Into Histopathology

The histopathology of salivary carcinoma is, indeed, “an evolving art,” and is somewhat confusing to the surgical practitioner. This is particularly true of the rarer forms of salivary carcinoma. It will be important in the future to glean new predictive information through the application of molecular biology techniques. Hopefully, a more simple codification of salivary tumors will emerge over the next few years.

Neutron Beam Therapy

Our experience with neutron beam therapy is summarized in the three articles cited by Dr. Spiro. Having used this modality in the treatment of a number of patients with salivary gland tumors, I would offer the following comments from a surgeon’s standpoint:

There is no doubt that the best results occur when a tumor can be resected down to microscopic or clear margins and then supplemented with neutron radiation. This approach offers approximately a 10% to 15% increase in locoregional control over the utilization of neutron beam therapy alone.

Surgery after neutron radiation is challenging, infrequently curative, and associated with all of the side effects of intensive postoperative radiation.

Summary

In closing, I enjoyed and learned from this summary by Dr. Spiro. It is an important contribution to the literature on salivary gland carcinoma.

References:

1. Spiro RH: Salivary neoplasms: Overview of a 35-year experience with 2,807 patients. Head Neck Surg 8:177-184, 1986.

2. Nigro MF, Spiro RH: Deep lobe parotid tumors. Am J Surg 134:523-527, 1977.

3. Johnson FE, Spiro RH: Tumors arising in accessory parotid tissue. Am J Surg 138:576-578, 1979.

4. Foote FW Jr, Frazell EL: Tumors of the major salivary glands. Cancer 6:1065-1153, 1953.

5. Batsakis JG, Regezi JA: The pathology of head and neck tumors: Salivary glands, part 1. Head Neck Surg 1:59-68, 1978.

6. Seifert G, Sobin LH: The World Health Organization’s histologic classification of salivary gland tumors. Cancer 70:379-385, 1992.

7. Spiro RH, Koss LG, Hajdu SI, et al: Tumors of minor salivary gland origin: A clinicopathologic study of 492 cases. Cancer 31:117-129, 1973.

8. Spiro RH, Armstrong J, Harrison L, et al: Carcinoma of major salivary glands: Recent trends. Arch Otolaryngol Head Neck Surg 115:316-321, 1989.

9. Stewart FW, Foote FW Jr, Becker WF: Muco-epidermoid tumors of salivary glands. Am Surg 122:820-844, 1945.

10. Spiro RH, Huvos AG, Strong EW: Adenocarcinoma of salivary origin: Clinicopathologic study of 204 patients. Am J Surg 144:423-431, 1982.

11. Spiro RH, Huvos AG, Strong EW: Acinic cell carcinoma of salivary origin: A clinicopathologic study of 67 cases. Cancer 41:924-935, 1978.

12. Perzin KH, Gullane P, Clairmont AC: Adenoidcystic carcinomas arising in salivary glands: A correlation of histologic features and clinical course. Cancer 42:265-282, 1978.

13. Nascimento AG, Amaral ALP, Prado LAF, et al: Adenoid cystic carcinoma of salivary glands: A study of 61 cases with clinicopathologic correlation. Cancer 57:312-319, 1986.

14. Spiro RH, Huvos AG, Strong EW: Adenoid cystic carcinoma: Factors influencing survival. Am J Surg 138:579-583, 1979.

15. Spiro RH, Huvos AG: Stage means more than grade in adenoid cystic carcinoma. Am J Surg 164:623-628, 1992.

16. Spiro RH, Huvos AG, Strong EW: Cancer of the parotid gland: A clinicopathologic study of 288 primary cases. Am J Surg 130: 452-459, 1975.

17. American Joint Committee for Cancer Staging and End Results Reporting: Manual for Staging of Cancer. Chicago, American Joint Committee, 1978.

18. Spiro RH, Thaler HT, Hicks WF, et al: The importance of clinical staging of minor salivary gland carcinoma. Am J Surg 162:330-336, 1991.

19. Spiro RH, Hajdu SI, Strong EW: Tumors of the submaxillary gland. Am J Surg 132:463-468, 1976.

20. Kelley DJ, Spiro RH: Management of the neck in parotid gland carcinoma (abstract). Proceedings of the International Conference on Head and Neck Cancer, p 147, 1996.

21. McNaney D, McNeese MD, Guillamondegui OM, et al: Post-operative irradiation in malignant epithelial tumors of the parotid. Int J Radiat Oncol Biol Phys 9:1289-1295, 1983.

22. Simpson JR, Matsuba HM, Thawley SE, et al: Improved treatment of salivary gland adenocarcinomas: Planned combination surgery and irradiation. Laryngoscope 96: 904-907, 1986.

23. Fitzpatrick PJ, Theriault C: Malignant salivary gland tumors. J Radiat Oncol Biol Phys 12:1743-1747, 1986.

24. Armstrong J, Harrison LB, Spiro RH, et al: Malignant tumors of major salivary gland origin: A matched-pair analysis of the role of combined surgery and postoperative radiotherapy. Arch Otolaryngol Head Neck Surg 116:290-293, 1990.

25. Laramore GE, Krall JM, Griffen TW, et al: Neutron versus photon irradiation for unresectable salivary gland tumors: Final report of an RTOG-MRC randomized clinical trial. Int J Radiol Oncol Biol Phys 27:235-240, 1993.

26. Bucholz TA, Laramore GE, Griffin BR, et al: The role of fast neutron radiation therapy in the management of advanced salivary gland neoplasms. Cancer 69:2779-2787, 1992.

27. Bucholz TA, Shimotakahara SG, Weymuller EA, et al: Neutron radiotherapy for adenoid cystic carcinoma of the head and neck. Arch Otolaryngol Head Neck Surg 119:747-552, 1993.

28. Harrison LB, Raben A, Pfister DG, et al: A prospective phase II trial of concomitant chemotherapy and radiotherapy with delayed accelerated fractionation in unresectable tumors of the head and neck (abstract). Proceedings of the International Conference on Head and Neck Cancer, p 102, 1996.

29. Mountain C, McMurtrey M, Hermes K: Surgery for pulmonary metastasis: A 20-year experience. Ann Thorac Surgery 38:323-330, 1984.

30. Pastorino U, Buyse M, Friedel G, et al: Long-term results of lung metastasectomy: Prognostic analyses based on 5,206 cases. J Thorac Cardiovasc Surg 113:37-49, 1997.

31. Spiro RH: Distant metastasis in adenoid cystic carcinoma of salivary origin. Am J Surgery 174:495-498, 1997.