Head and neck squamous cell carcinoma (HNSCC) is the sixth most common malignancy worldwide, with over 40,000 new cases per year in the US. Of all primary HNSCCs, oropharyngeal carcinomas are the third most common, with the tonsil being the most common site of malignancy within the oropharynx.
Historically, the standard treatment of tonsil cancer has consisted of surgery with or without adjuvant radiotherapy. Given the important role the oropharynx plays in speech and swallowing, nonsurgical therapy with organ-preserving chemoradiation has gained a role in the treatment of tonsil carcinoma in attempts to avoid the morbidity of surgery. However, organ-preserving chemoradiation is not without its own morbidity. Additionally, minimally invasive transoral surgical techniques have become more widespread in the treatment of tonsillar carcinoma and may decrease the morbidity associated with surgical therapy. As such, the optimal treatment of tonsillar carcinoma is debated among head and neck oncologists.
A study by Liederbach et al indicated that the trend toward surgical management of oropharyngeal squamous cell carcinoma (SCC) has increased since the 2009 approval by the US Food and Drug Administration of transoral robotic surgery and that tonsillar origin of the cancer is one of the independent predictors of surgical treatment. According to the study, the rate of surgery for oropharyngeal SCC fell from 41.4% in 1998 to 30.4% in 2009 but then rose again, reaching 34.8% in 2012. The investigators reported that between 2009 and 2012, surgery was more likely in patients who not only had SCC of tonsillar origin but who were young, female, white or Hispanic, and of high socioeconomic status and who had stage I disease, had comorbidities, lived more than 75 miles from the hospital, had private insurance, used academic hospitals, and used hospitals in the West North Central region.[1]
This article will detail the frequency, etiology, and work-up of tonsillar SCC. It will then focus on the surgical treatment of the disease.
Historically, the mainstay of treatment for tonsillar SCC has consisted of open surgery, neck dissection, and tracheostomy, with or without adjuvant radiotherapy. Typical open approaches include a lip-splitting mandibulotomy to provide maximal access to the tonsillar fossa and to achieve wide margins and control of critical neurovascular structures. Frequently, this approach requires a reconstruction with either a regional rotational flap or a microvascular free flap. Although this approach is effective, it is not without morbidity. In an effort to decrease morbidity, surgeons have developed minimally invasive transoral techniques to treat tonsillar SCC. With the increased use of transoral techniques, surgical therapy remains a viable option in the treatment of tonsillar SCC.
Tonsillar SCC is a very treatable disease when diagnosed at an early stage. However, advance stage disease continues to have a poor prognosis, with stage IV disease carrying a 5-year survival approximately 50% or less. In addition to the poor prognosis associated with advanced stage disease, treatment may be associated with significant morbidity to the patient with respect to speech and swallowing. Both radical surgery and radical chemoradiotherapy alike can lead to significant impairment of swallowing, with reliance on a tracheostomy and gastrostomy tube. This can have a negative impact on patient quality of life.
Malignancy of the tonsils is an uncommon entity that accounts for little more than 0.5% of new malignancies in the United States every year. More than 8,000 oropharyngeal carcinomas are diagnosed in the United States each year.[2] The Armed Forces Institute of Pathology (AFIP) registry from 1945-1976 determined that more than 70% of malignancies in this region are squamous cell carcinoma. Squamous cell carcinomas are about 3-4 times more common in men than in women, and they are largely tumors that develop in the fifth decade of life or later.
Lymphomas of the tonsil are the second most frequent malignancy in this area. Other more uncommon malignancies include minor salivary gland tumors and metastatic lesions.
A study by Weatherspoon et al reported that approximately 75,468 incident oral cancer cases were diagnosed in the United States between 2000 and 2010, with the tonsils being the most frequent site of diagnosis (23.1%).[3]
According to the National Cancer Institute, accepted risk factors for squamous cell carcinoma include smoking and ethanol abuse. More recently, however, some indications show that viral etiology should also be considered. Although Epstein-Barr virus (EBV) is a major consideration in nasopharyngeal carcinoma, human papilloma virus (HPV) has been shown as more of a menace in this region.
Some studies have identified indications of HPV presence in approximately 60% of tonsillar carcinomas.
When the tonsils are included in studies of the entire oropharyngeal region, the risk factors include the following:
A diet deficient in fruits and vegetables
Consumption of the South American beverage mate
Chewing of betel quid
Infection with HPV
Tobacco smoking
Ethanol use
HPV is a double-stranded DNA virus that infects basal cells of the epithelium and can be found in up to 36% of squamous cell carcinomas of the oropharnyx.[4] Although more than 100 strains have been isolated, HPV type 16 and 18 are most commonly associated with cancers.[5] The viral genome codes for the oncoproteins E6 and E7, which have increased activity in highly oncogenic strains. The E6 oncoprotein causes degradation of the tumor suppressor p53, preventing programmed cell death. The E7 oncoprotein results in loss of the retinoblastoma (Rb) tumor suppressor. Loss of pRb leads to accumulation of p16, which would normally inhibit cell cycle progression through cyclin D1 and CDK4/CDK6 mediated events. However, E7 overrides this cell cycle checkpoint, and cell progress from G1 into S phase. Because of this accumulation, p16 can be used as a marker of HPV activity.
Tonsillar SCC may be confined to the tonsillar fossa, but extension to adjacent structures is common. Carcinoma commonly spreads along the glossotonsillar sulcus to involve the tongue base to a variable degree. In addition, spread is frequently superior to involve the soft palate or nasopharynx. The tonsillar fossa is bounded laterally by the superior constrictor muscle, which may contain the spread of carcinoma.
However, when the constrictor muscle is transgressed, the tumor gains access to the parapharyngeal space. It may involve the pterygoid musculature or the mandible. Superior extension in the parapharyngeal space can lead to skull base involvement, and inferior extension can lead to involvement of the lateral neck. Finally, extensive involvement within the parapharyngeal space may involve the carotid artery.
Metastasis to regional lymphatics is common. Neck metastases are present in approximately 65% of patients. In patients with a clinically negative neck, approximately 30% of these patients will have occult neck disease. Most lymph node metastases are to level II and to a lesser extent level III. Nodal metastases to level I or level IV occur in approximately 10%, and skip lesions in both these locations have been encountered.
Tonsillar SCC may also metastasize to retropharyngeal lymph nodes. These nodes are not the primary echelon nodes, but metastasis to this location may occur when the lymphatics are disrupted in the case of node positive disease in the jugulodigastric nodes or in the case of prior treatment with either surgery or radiation. Distant metastasis from tonsillar SCC occurs in approximately 15-30% of patients. The most common sites are lung, followed by liver, and then bone.[6]
Patients with tonsillar carcinomas may present with a neck mass. This is because carcinomas arise deep within the aforementioned crypts. These are deep epithelial invaginations of the surface epithelium.
A squamous carcinoma may originate at 1 or more sites within the deep nests or branches within the tonsil. In addition, the tonsil can enlarge considerably, bulging into empty oral space before it causes alarm to the individual.
Finally, the tonsils are lymphoid rich and contain abundant lymphatics that help the neoplasm access and metastasize to neck nodes.
All of these factors, and perhaps other unknown ones, explain why patients may present with a neck mass.
One of the unusual aspects of the neck node metastasis is the fact that a very large number of these are cystic. This has led to many being erroneously called branchial cleft carcinomas. In fact, the literature debates the existence of such an entity as a branchial cleft carcinoma.[7] Many pathologists feel that branchial cleft carcinoma is actually either a metastasis or a direct extension from a tonsil squamous cell carcinoma.
Regardless, cystic neck lymph node with an occult primary tumor must prompt an investigation of the tonsil. Occult primary squamous cell carcinomas that manifest as neck lymphadenopathy are a common problem faced by otolaryngologists.
Although the hypopharynx and the nasopharynx are often suspected as being the seed area, the tonsil and the tongue base are also very likely (perhaps more likely) sites and should also be promptly investigated.
In addition to a neck mass presentation, usually in the jugulodigastric region, other symptoms and signs may develop. These may be in conjunction with a neck mass or may be the only presentation.
Sore throat, ear pain, foreign body or mass sensation, and bleeding are all possible. Trismus is an ominous sign because it probably indicates involvement of the parapharyngeal space. Such tumors may be large enough to involve or encase the carotid sheath. In addition, the tumor may extend to the skull or mediastinum.
Even if the neck mass is not evident on casual inspection, careful palpation may reveal cervical lymphadenopathy.
If the tumor has involved the tongue base, contralateral nodes may be involved.
Primary tonsillar tumors may grow entirely beneath the surface. The clinician may therefore see nothing suspicious or may see only a slight increase in the size of the tonsil or the firmness of the area.
Alternatively, an exophytic fungating mass with central ulceration and heaped-up edges may be present. It may be deep red to white. Cutting into the lesion during biopsy may demonstrate a gritty texture (a function of the degree of keratinization), a firm resistance (a function of the degree of fibrosis), and cystification (a function of necrosis). Obviously, these findings vary depending on the specifics of the tumor according to the parameters parenthetically described.
The constitutional signs and symptoms of weight loss and fatigue are not uncommon with this neoplasm.
When considering surgery for tonsillar SCC, one must consider several factors prior to surgery, as follows:
Is the primary tumor resectable?
Is the neck disease resectable?
Is there distant metastatic disease?
What is the expected functional outcome following surgery?
Are there comorbid patient conditions that will affect surgical outcomes?
What is the patient’s preference for treatment?
Tumors of the tonsil are considered unresectable when there is invasion of the lateral pterygoid muscle, pterygoid plates, lateral nasopharyngeal wall, skull base, or carotid artery encasement.
Metastases to the neck are considered unresectable when the tumor circumferentially involves the carotid artery, if the tumor invades the deep muscles of the neck, vertebral column invasion, skull base invasion, Horner’s syndrome, phrenic nerve palsy, or brachial plexus invasion.
The anatomy of the tonsillar area is responsible for the fact that the vast majority of malignant tumors in this region present in advanced stages. In addition, the tonsils themselves have ill-defined boundaries that merge with other anatomic landmarks. Often, tumors involve these areas by the time a tonsillar primary tumor is palpable. A primary tonsillar malignancy that involves the base of tongue or palate is not unusual.
The anterior border of the tonsil is the anterior faucial pillar, which contains the palatoglossal muscle and is covered by squamous mucosa. The posterior border of the tonsil is the posterior faucial pillar, which contains the glossopharyngeal muscle and is covered by squamous epithelium. Occasionally, the ciliated columnar epithelium is also contained. Superiorly, these areas merge into the soft palate. Inferiorly, the pillars merge at the base of the tongue at the glossotonsillar fold. Large tumors of the tonsil often extend to the tongue base along this fold. No truly medial margin exists because this is an anatomic space at the junction of the oral cavity and pharynx. The lateral border is the superior constrictor muscle, which lines the tonsillar bed. Beyond the constrictor muscle is the parapharyngeal space.
When performing open surgery for tonsillar carcinoma, the anatomy is well known to head and neck surgeons. Furthermore, resection of the primary tumor usually follows a neck dissection, such that the carotid artery is under direct vision and can be protected. However, as minimally invasive transoral techniques become more prominent, the anatomy of the tonsil is approached from “inside-out,” in a manner that may be unfamiliar to many head and neck surgeons.
Holsinger et al described the transoral lateral oropharyngectomy.[7] This technique describes resecting the tonsil from lateral to medial, such that the transoral anatomy is approached differently than when performing a standard tonsillectomy.
The operation begins by making an incision through mucosa near the retromolar trigone and through the superior constrictor muscle. This allows the constrictor muscle to be reflected off the medial pterygoid muscle, and deeper dissection in this plane will encounter the parapharyngeal space fat. This collection of fat is traversed by the blood vessels supplying the tonsillar fossa.
As the dissection continues deeply, the styloglossus muscle, and then the stylopharyngeus muscle will be encountered. Between these muscles runs the glossopharyngeal nerve, which may be identified during a transoral resection. The internal carotid artery runs deep to these muscles, and they can be used as a landmark.
Liver function tests: Knowledge of hepatic function is necessary because (1) the patient's dietary and ethanol histories frequently lead to poor function, (2) hepatically metabolized chemotherapeutic agents or other medications (eg, pain medication) may be used, and (3) liver metastases are always possible.
Pulmonary function tests:
Any head and neck surgery carries additional risks of perioperative and postoperative respiratory complications.
Respiratory reserve is a necessary bit of knowledge before such surgery is performed.
Renal function tests: When certain chemotherapeutic agents are considered, renal function tests are necessary to ascertain whether the patient can eliminate agents that are handled by the kidneys.
Clotting and coagulation studies (including platelet count, typing, cross-matching)
The head and neck is one of the richest areas of vascularity in the human body.
Hemorrhage is one of the biggest problems in tonsillar surgery.
CT scanning of the neck, with and without contrast, is necessary to evaluate for metastases and to assess the extent of the tumor. In addition, if extended upward to include the bony areas, bone invasion is part of the new knowledge base. This is essential in staging tonsillar tumors.
MRI is also extremely useful for assessing tumor size and soft tissue invasion.
CT scan of the chest is the single most sensitive imaging study used to reveal lung metastasis and, therefore, should be the modality of choice, at least in high-risk patients (stage 4 disease, T4 tumor, N2 or N3 nodal disease, tumors that arise from the oropharynx, larynx, hypopharynx, or supraglottis).[8]
Biopsy is the only tool for obtaining diagnostic tissue.
Tonsillar malignancies may be lymphoma; therefore, the pathologist and team should be immediately ready to handle the tissue properly.
Special fixatives must be prepared. Some tissue may be needed for fresh studies, which are time dependent and require immediate handling. Some tissue should be frozen in liquid nitrogen. Given the nature of frozen sections and the type of unexpected events in a pathologist's day, alerting the pathologist 24 hours in advance of a possible lymphoma biopsy is wise.
Another very important consideration is the fact that squamous cell carcinomas commonly arise deep in the crypts. This necessitates the surgeon taking a deep biopsy so that the true neoplasm is not missed. Given the propensity for these lesions to bleed, this is a tricky procedure, and the surgeon should be ready for the unexpected.
Panendoscopy
Operative endoscopy allows the surgeon to assess the full extent of the tumor. This can be very helpful when choosing between open and endoscopic surgical approaches. It also allows for a biopsy if it cannot be performed in the office.
Bronchoscopy and esophagoscopy are utilized to assess for second primary tumors that may be present at the time of diagnosis.
HPV testing
NCCN guidelines recommend HPV testing for prognostic factors.
Quantitative reverse transcriptase PCR (QRT-PCR) allows calculation of relative amounts of mRNA present in the sample.
Able to calculate copy number
Susceptible to false positives
Type-specific HPV DNA in situ hybridization
HPV-16 is most commonly used to examine oropharyngeal carcinomas.
It is both sensitive and specific.
P16 can be tested as a biomarker for HPV E7 activity.
Most palatine tonsil squamous cell carcinomas are moderately to poorly differentiated.
The following variants, although essentially squamous cell carcinomas, in this area have been described with some frequency:
Basosquamous carcinoma Nonkeratinizing carcinoma (transitional cell or sinonasal type)
Undifferentiated or lymphoepithelioma type
Lymphomas
Lymphoma type determination is crucial and can be achieved only with the help of special studies obtained by the pathologist. The cell and tissue markers used to type lymphomas are quite sensitive. These require fresh frozen tissue and unusual fixatives, in addition to immunohistochemical stains.
All of these studies help in the crucial determination of lymphoma type. Many require fresh or frozen tissue for immunohistochemical studies.
Most tonsillar carcinomas are diffuse non-Hodgkin large B-cell lymphomas.
Mucosa-associated lymphoid tissue (MALT) low-grade B-cell lymphomas composed of small cells are uncommon in the tonsil. This is surprising because the tonsil consists of a very intimate intermingled arrangement of epithelium and lymphocytes, which, in theory, would make an ideal environment for the development of MALT lymphomas. In reality, they are so uncommon in this region that they are case reportable.
Minor salivary gland malignancies
Minor salivary gland malignancies are the third most common lesion of the tonsil. These lesions include mucoepidermoid carcinoma, adenoid cystic carcinoma, acinic cell carcinoma, and adenocarcinoma.
Metastatic lesions to the tonsil
Although the palatine tonsils are a rich source of lymphatics and lymphoid tissue, metastases to the palatine tonsils are rare. Case reports have described an extraordinarily wide spectrum of malignancies metastatic to this area. Breast, various lung primaries, renal carcinomas, and pancreatic and colorectal malignancies have been reported. Documented cases of Wilms tumor and choriocarcinoma metastasizing to this distant site also exist.
Staging of tonsil carcinoma is according to the 6th edition of the AJCC Cancer Staging Manual. Clinical information is taken from all sources, including physical examination and any available imaging studies.
AJCC tumor staging of tonsil carcinoma is as follows:
Tx: Primary tumor cannot be assessed
T0: No evidence of primary tumor
Tis: Carcinoma in situ
T1: Tumor ≤ 2 cm in greatest dimension
T2: Tumor >2 cm but < 4 cm in greatest dimension
T3: Tumor >4 cm in greatest dimension
T4a: Tumor invades the larynx, deep or extrinsic muscles of the tongue, medial pterygoid muscle, hard palate, or mandible
T4b: Tumor invades the lateral pterygoid muscle, pterygoid plates, lateral nasopharynx, skull base, or encases carotid artery
AJCC nodal categories (except thyroid and nasopharyngeal carcinoma):
Nx: Regional lymph nodes that cannot be assessed
N0: No regional node metastasis
N1: Metastasis in a single ipsilateral lymph node, 3 cm or smaller
N2: Metastasis in a single ipsilateral lymph node, larger than 3 cm but not larger than 6 cm in greatest dimension is found; multiple ipsilateral lymph nodes, none larger than 6 cm; bilateral or contralateral lymph nodes, none larger than 6 cm
N2a: Metastasis in a single ipsilateral lymph node larger than 3 cm but not larger than 6 cm
N2b: Metastasis in multiple ipsilateral lymph nodes, none larger than 6 cm
N2c: Metastasis in bilateral or contralateral lymph nodes, none larger than 6 cm
N3: Metastasis in a lymph node larger than 6 cm
Distant metastasis:
Mx: Distant metastasis cannot be assessed
M0: No distant metastasis
M1: Distant metastasis
The combination of the primary tumor, nodal status, and presence or absence of distant metastasis is used as a part of the overall staging of the patient’s disease according to AJCC guidelines:
Non-surgical therapy of tonsil carcinoma consists of radiation therapy to the primary site and neck for early stage T1-2N0 tumors. For advanced stage tumors T3-4N+, non-surgical therapy consists of organ-preservation concurrent chemoradiation. This article focuses on surgical treatment, and therefore these therapies will not be discussed in detail.
When evaluating a patient with tonsil carcinoma for surgery, one must determine the optimal surgical approach. For most early stage tumors and select late-stage tumors, a transoral approach may be appropriate. Transoral approaches include using a standard mouth gag and headlight as performing a standard tonsillectomy, transoral laser microsurgery (TLM), or a new technique, transoral robotic surgery (TORS).[9]
However, for most advanced stage tumors, the standard open approaches are typically appropriate. Open approaches may include a lip-splitting mandibulotomy or a lateral pharyngotomy to achieve access to the tumor. Typically in these open approaches, reconstruction with either a local, regional, or free tissue flap is needed to close the surgical defect. When deciding upon an approach, the surgeon must carefully assess the extent of the tumor, and when considering a transoral approach, the surgeon must determine if transoral access is possible.
Factors preventing transoral access include trismus, large teeth, small transverse mandibular dimensions, mandibular tori, large tongue, poor atlanto-occipital extension, and prior radiation that may obscure tissue planes or determination of resection margin. Furthermore, if the tumor has significant lateral extension, transoral resection may put the carotid artery at risk, making this approach unsafe.
In addition, if transoral resection may leave a positive margin (such as with skull base extension), then an open approach should be chosen. Finally, the experience of the surgeon must be considered. Transoral resection tonsil carcinoma approaches the anatomy from “inside-out” in a way that may not be familiar to many surgeons. This can make transoral surgery difficult and compromise the resection margin.
When the tumor has significant involvement of adjacent sites such as the soft palate, tongue base, or nasopharynx, a transoral resection may not be appropriate due to the need for reconstruction. When more than half the soft palate or tongue base is resected, these patients may benefit from reconstruction with a flap, and an open surgical approach may be more appropriate. Most of these factors can be accurately assessed prior to taking the patient to the operating room for definitive treatment. Physical examination in the office or during operative endoscopy can accurately map out the extent of the tumor. Careful examination of preoperative imaging can determine the proximity of the carotid artery to the tumor.
A retrospective study by Spellman et al indicated that in patients with palatine tonsil SCC, staging of the malignancy can determine candidates for primary TORS. The investigators found that in patients with early stage (T1/2) tonsillar SCC, definitive treatment with TORS in those who had pathologically confirmed N0/N1 necks without extracapsular extension resulted in clear margins in all patients. No significant intraoperative or postoperative complications occurred in these cases, and no recurrences of the cancer were found in mean and median 28-month follow-up.[10]
If the patient is deemed a candidate for transoral resection, the surgeon must then decide how to resect the tumor. Options include the using standard mouthgags and a headlight, TLM, or TORS. Holsinger et al described their approach to transoral lateral oropharyngectomy for removal of tonsil carcinomas using standard mouthgags and a headlight.[7] This approach involves incising the superior constrictor muscle at the pterygomandibular raphe, and then reflecting the constrictor muscle from lateral to medial. This approach takes the superior constrictor muscle as the lateral margin of resection.
Transoral laser microsurgery techniques have been described in detail by Steiner and Ambrosch.[11] The standard approach is to gain exposure via standard mouthgags or distending oropharyngoscopes. The operating microscope and a CO2 laser is then used to remove the tumor piecemeal. Using the enhanced visualization of the microscope and the differential cutting of the laser through normal tissue versus tumor allows the surgeon to follow the tumor and preserve the maximal amount of normal tissue. Because the tumor is removed piecemeal, it is imperative that the surgeon communicate effectively with the pathologist to ensure a true negative margin.[12]
Transoral Robotic Surgery is new technique pioneered by Weinstein et al.[9] TORS achieves exposure of the tumor using standard mouthgags. The robotic endoscope and operating instruments are then inserted into the mouth and used to resect the tumor in a modified fashion described by Holsinger et al. Advantages of TORS include enhanced 3-D visualization and use of wristed instruments and angled endoscopes that allow the surgeon to achieve access to the tumor that are otherwise difficult with the standard transoral approach or TLM.[11]
In all cases of transoral resection, the surgeon must be able to control bleeding from branches of the carotid arterial system. This is achieved either by electrocautery or through the use of surgical hemoclips. Typically, the surgical bed is left to heal by secondary intention. This makes protecting the carotid artery critical, to prevent erosion of the vessel wall by exposure to saliva.
In cases in which an open approach is needed, exposure of the tumor may be through a lip-splitting mandibulotomy, a lateral pharyngotomy, or a combination of transoral exposure and lateral pharyngotomy. This leaves the patient with an open communication between the neck and the pharynx, necessitating a reconstruction with a soft tissue flap. Common options for reconstruction include a pectoralis major myocutaneous flap, radial forearm fasciocutaneous free flap, or anterolateral thigh fasciocutaneous or myocutaneous free flap. Advantages of open approaches include direct access to the tumor while providing maximal control of the great vessels. However, patients are subjected to longer operative times, the need for reconstruction, and longer time for recovery of swallowing.
Treatment of tonsillar carcinoma requires management of the regional lymphatics. Selective neck dissection to include levels I-IV is the standard operation. In open approaches, neck dissection is performed with the resection of the primary tumor. In transoral approaches, neck dissection can be performed concomitantly, or it may be staged to avoid the risk of a salivary fistula. In addition, the retropharyngeal nodes may be at risk. These can be removed at the time of primary tumor resection.
Patients undergoing surgery for tonsillar carcinoma must be monitored carefully in the postoperative period. Issues that are critical to consider include airway management, potential for bleeding, and diet. Airway management is dependent on the approach used and the extent of resection. When a transoral approach is utilized, patients may remain intubated following surgery, depending on the extent of resection, the potential risk for bleeding, or the preference of the surgeon. In most cases, patients undergoing transoral resection will not need a tracheostomy, as swelling is generally less than in open resections.
In the case of open resections, edema is usually significant, especially with flap reconstruction, and most patients will require a tracheostomy. The tracheostomy is temporary in most cases. Bleeding following resection of tonsillar carcinoma can be significant and life-threatening. Most cases of transoral resection allow the wound to heal by secondary intention. As a result, branches of the external carotid artery that have been ligated during surgery are at risk for bleeding. Bleeding can be brisk, and due to the proximity to the airway, aspiration of blood can be a significant problem. Rich et al reported a 3.6% bleeding rate in a large series of oropharyngeal carcinomas treated with TLM.[12]
Resumption of an oral diet is also an important consideration following surgical treatment of tonsillar carcinoma. Nearly all patients will have some level of dysphagia that can interfere with resumption of a normal diet. Transoral resections typically have less dysphagia, although many patients will require a temporary feeding tube. Timing of feeding tube removal can be dictated by a clinical examination, with or without a modified barium swallow. Long-term percutaneous gastrostomy (PEG) tube requirement is rare with transoral resections. Weinstein et al reported a 3.7% PEG tube rate in their initial study of TORS radical tonsillectomy.[9] This is similar to the rate (4%) reported by Moore et al in their study of transoral resection of tonsil carcinomas.[13]
Adjuvant treatment with radiation therapy or chemoradiation is frequently indicated following surgery, and is dictated by the final pathology. Indications for postoperative radiotherapy include perineural or lymphovascular invasion, multiple positive nodes, close margins, and T4 disease. Indications for postoperative chemoradiation include positive margins and extracapsular spread in the lymph nodes.[14]
Routine follow-up care of patients with tonsil cancer is important, particularly because the risk of developing a second primary tumor is highest in this group. Patients with head and neck cancers have a 20% overall risk of developing a second primary tumor, while patients with tonsil cancer have as high as a 30% risk.
The prognosis as determined by 5-year survival rate of treated squamous cell carcinoma of the tonsillar region is as follows:
Stage I - 80%
Stage II - 70%
Stage III - 40%
Stage IV - 30%
The survival from tonsillar carcinoma has historically been considered poor, especially for late stage (III and IV) disease. However, more recent literature has shown promising results with surgical therapy for tonsillar carcinoma, even for advanced stage disease. A study by Rahmati et al of patients who had undergone a combination of surgery and postoperative radiotherapy for tonsillar SCC found the 5-year rates of overall survival (OS), disease-specific survival (DSS), and recurrence-free survival (RFS) to be 66%, 82%, and 80%, respectively. The study involved 88 patients, 48% of whom had T3-T4 tumors and 75% of whom had neck lymph node metastases. The study results indicated that neck metastases were not predictive of survival, with DSS being 80% for patients without such metastases and 82% for those with metastases, but that lymphovascular invasion was predictive of OS, DSS, and RFS.[15]
Rich et al reported 2- and 5-year disease-specific survival rates of 96% and 92% for stage III and IV oropharyngeal carcinoma treated with transoral laser microsurgery (TLM) and adjuvant therapy.[12] Likewise, Moore et al reported 94% disease-specific survival in stage III and IV tonsillar carcinomas treated with transoral resection and adjuvant therapy. The results of these studies indicate that select patients, when treated appropriately, can have excellent survival outcomes despite historically poor outcomes.
The safety and feasibility of transoral robotic surgery (TORS) as a minimally invasive treatment alternative for malignancies of the head and neck was established by investigators at the University of Pennsylvania, University of Alabama-Birmingham, and the Mayo Clinic. This resulted in TORS being approved by the US Food and Drug Administration (FDA) in 2009 for transoral otolaryngology surgical procedures in adults restricted to benign and malignant tumors classified as T1 and T2. Advanced T-stage tumors were not approved since these studies only had a small number of advanced-stage tumors. While these initial studies reported their oncologic outcomes, the follow-up was short due to the relative infancy of the procedure. However, the oncologic data for TORS are beginning to mature and longer-term follow-up data are being reported.
The University of Pennsylvania reported the results of 47 consecutive patients with advanced-stage oropharyngeal squamous cell carcinoma (OPSCC) and a minimum of 18 months follow-up who were treated with primary TORS, staged neck dissection, and adjuvant radiation or chemoradiation as indicated.[16] Seventy-seven percent of patients had T1 and T2 tumors, while 51% of patients had N1 disease and 49% of patients had N2 disease.
Negative margins were achieved in 98% of patients. Five patients avoided radiotherapy altogether, while 13 patients received radiotherapy only, 2 received chemotherapy only, and 27 received concurrent chemoradiation. Local control and regional control were 98% and 96%, respectively. Actuarial 1- and 2-year overall survival was 96% and 82%, respectively. Actuarial disease-specific survival at 1 year and 2 years was 98% and 90%, respectively. Disease-free survival was 96% at 1 year and 79% at 2 years. When compared with similar organ-preservation chemoradiation trials, oncologic control was similar.[17, 18]
With the changing landscape of OPSCC and more tumors caused by human papillomavirus (HPV) infection, the same authors analyzed the oncologic outcomes of a cohort of TORS patients with respect to HPV status.[19] In 50 patients treated with TORS, staged neck dissection, and adjuvant radiation or chemoradiation as indicated, 37 patients (74%) had HPV-positive tumors and 13 patients (26%) had HPV-negative tumors. There was no statistical difference between the 2 groups with respect to the margin status, presence of cervical metastases, recurrence, or survival curves. Many HPV-induced tumors are presenting with smaller tumors at the primary site but with advanced-stage neck disease.
To determine the rates of regional recurrence in their patients, the authors examined 31 patients, all with negative-margin TORS who underwent selective neck dissection and adjuvant therapy, and they found only one regional recurrence. Examination of the pathological specimens in the neck showed that 33% and 43% of the clinical N0 and N1 patients, respectively, were pathologically upstaged, while 4 of the 14 clinical N1 patients had negative pathological necks. Pathological staging of the necks allowed the authors to selectively administer adjuvant therapy and to deintensify therapy in some cases.[20]
The University of Alabama-Birmingham (UAB) and the Mayo Clinic pooled their data to report their 2-year survival analysis in a cohort of 89 patients with carcinoma of the oral cavity, oropharynx, and supraglottic larynx.[21] Seventy-nine percent of patients were stage T1 and T2, and negative margins were achieved in all patients. Seventy-six percent of patients underwent staged or concomitant neck dissection. Of the patients who underwent TORS as primary treatment, 63% received adjuvant radiation therapy and 48% had chemotherapy either before or after surgical treatment. Two-year recurrence-free survival in patients who underwent TORS as primary treatment was 89.3%.
Genden et al reported 18-month survival data in 30 patients with head and neck squamous cell carcinoma, the majority of who had OPSCC.[22] All patients underwent TORS, concomitant neck dissection, and adjuvant therapy as indicated. Because all patients were pathologically staged, adjuvant therapy was deintensified in 4 patients, while it was escalated in 5 patients. Eighteen-month locoregional control, distant control, disease-free survival, and overall survival rates were 91%, 93%, 78%, and 90%, respectively. Comparison with a matched group of patients undergoing primary chemoradiation showed no statistically significant survival differences.
In addition to excellent oncologic outcomes, patients treated with TORS have shown excellent functional outcomes with longer follow-up. The University of Pennsylvania reported a 2.4% (one patient) gastrostomy tube dependency rate at a minimum of 1-year follow-up, while the rate of feeding tube dependency in the combined UAB and Mayo Clinic, as well as the Mt. Sinai study, was 0%.[16, 21] In addition to gastrostomy tube dependency rate, quality-of-life data are now available in patients undergoing TORS.
Leonhardt et al found that in 38 patients with OPSCC undergoing TORS, staged neck dissection, and adjuvant therapy, declines in the eating and diet domains on the Performance Status Scale that were seen at 6 months returned to baseline at 12 months.[23] Declines in the speech domain remained significantly decreased at 6 and 12 months. Patients receiving adjuvant chemoradiation had significantly lower diet domain scores at 6 and 12 months compared with those who underwent surgery only.
Hurtuketal et al used the Head and Neck Cancer Inventory to show that speech, aesthetics, attitude, and overall quality of life remained in the high domain at 12 months, while the eating domain dropped to the intermediate level at 12 months in 18 of 64 patients who underwent TORS and adjuvant therapy for head and neck squamous cell carcinoma at all sites.[24]
Genden et al compared TORS patients with a similar cohort of patients treated with primary chemoradiation and found that all TORS patients returned to baseline in the eating, speech, and diet domains on the Performance Status Scale for Head and Neck, as well as the Functional Oral Intake Score. In contrast, patients treated with chemoradiation had a lower-than-baseline diet score and lower Functional Oral Intake Score.[22]
Since its inception, TORS has proven to be a safe and efficacious, minimally invasive means of achieving an en bloc resection of head and neck malignancies. Functional outcomes have been excellent, and as the data mature, the oncologic outcomes appear to be equivalent to those with achieved with standard open surgery or organ-preservation chemoradiation protocols. By surgically staging patients with TORS and neck dissection, adjuvant therapy can be tailored to the individual patient and can be deintensified in some cases of low-risk disease. TORS offers the potential to maximize oncologic control while maximizing functional outcomes. This approach is in stark contrast to chemoradiation protocols in which all patients receive identical therapy. More studies are necessary to validate long-term outcomes.
It is becoming apparent that survival for tonsil cancer correlates with HPV status. There has been recent evidence from both the surgical and nonsurgical literature that patients with HPV-positive tumors have improved survival.[25] Several retrospective case series have shown that patients with HPV-positive oropharyngeal tumors have a better prognosis than patients with HPV-negative tumors.[26] Similar findings were reported in a prospective analysis of data from a small clinical trial.[27]
More recently, Ang et al analyzed the HPV status in a large number of patients randomized to receive either standard fractionated chemoradiation or accelerated fractionation chemoradiation for stage III-IV OPSCC.[28] Patients with HPV-positive tumors had improved 3-year overall survival (82.4% vs 57.1%, P< .001) and 58% risk-of-death reduction compared with patients with HPV-negative tumors.
Furthermore, these authors found that patients could be risk stratified according to HPV status and smoking history. Low-risk patients had HPV-positive tumors and less than 10 pack years smoking history or had HPV-positive tumors, more than 10 pack years smoking history, but had N0-N2a neck disease. Intermediate-risk patients had HPV-positive tumors with more than 10 pack years smoking history and advanced neck disease or they had HPV-negative tumors with less than 10 pack years smoking history and T2-T3 tumors. The high-risk group had HPV-negative tumors with less than 10 pack years smoking history and T4 tumors or they had HPV-negative tumors and more than 10 pack years smoking history. Three-year overall survival in the low-, intermediate-, and high-risk groups was 93%, 70.8%, and 46.2%, respectively.
In the surgical literature, HPV tumor status is proving to affect prognosis as well. Rich et al reported improved overall survival (HR = 0.04) and disease-specific survival (HR = 0.10) in patients with p16-positive tumors compared with p16-negative tumors.[12] While p16 serves as a biomarker for HPV positivity, the authors did not, however, find that HPV positivity affected survival. This may be related to discrepancies between the different detection techniques.
Results of the above studies, and others, indicate that HPV-positive tumors have improved outcomes compared with HPV-negative tumors. Knowledge of the HPV status of the tumor is important for prognosis and risk stratification of patients and should be included in the pathologic analysis of tumor specimens.
The largest controversy regarding the treatment of tonsillar carcinoma surrounds whether this cancer should be treated primarily with surgery or with organ-preservation chemoradiation. Each treatment is not without its own risks, and the decision regarding treatment should be made in conjunction with the recommendations of the multidisciplinary team and the preferences of the patient.
Recent surgical literature has been extremely promising in the treatment of tonsillar carcinoma. Minimally invasive techniques such as transoral robotic surgery and transoral laser microsurgery offer patients an excellent option for treatment both from a functional and an oncologic standpoint.
Niels Kokot, MD, Associate Professor, Residency Program Director, USC Tina and Rick Caruso Department of Otolaryngology-Head and Neck Surgery, Keck School of Medicine of the University of Southern California
Disclosure: Nothing to disclose.
Specialty Editors
Francisco Talavera, PharmD, PhD, Adjunct Assistant Professor, University of Nebraska Medical Center College of Pharmacy; Editor-in-Chief, Medscape Drug Reference
Disclosure: Received salary from Medscape for employment. for: Medscape.
Karen H Calhoun, MD, FACS, FAAOA, Professor, Department of Otolaryngology-Head and Neck Surgery, Ohio State University College of Medicine
Disclosure: Nothing to disclose.
Chief Editor
Arlen D Meyers, MD, MBA, Professor of Otolaryngology, Dentistry, and Engineering, University of Colorado School of Medicine
Disclosure: Serve(d) as a director, officer, partner, employee, advisor, consultant or trustee for: Cerescan;RxRevu;Cliexa;The Physicians Edge;Sync-n-Scale;mCharts<br/>Received income in an amount equal to or greater than $250 from: The Physicians Edge, Cliexa<br/> Received stock from RxRevu; Received ownership interest from Cerescan for consulting; .
Additional Contributors
Terance (Terry) Ted Tsue, MD, Vice-Chairman for Administrative Affairs, Professor, Residency Program Director, Department of Otolaryngology-Head and Neck Surgery, University of Kasnas School of Medicine
){kind=link}