The treatment of cancer of the larynx has undergone considerable evolution over the past 20 years. Prior to the 1990s, patients with advanced cancer of the larynx were usually treated with total laryngectomy (TL) and postoperative radiation therapy. Currently most patients with advanced cancer of the larynx are treated with chemotherapy and radiotherapy, reserving surgery for salvage. In this situation, with a hostile local tissue environment from the prior chemoradiation, partial laryngectomy is often impractical and salvage TL is common. Salvage TL, however, is not without risk, and there are often perioperative complications, including fistula formation and its sequelae, among others. In order to minimize this risk, the routine use of flap reconstruction is recommended. Even with flap reconstruction, however, wound complications can occur, incurring additional morbidity to the patient.

Recently, transoral techniques, either robotic or microsurgical, have been recognized as appropriate methods to surgically approach select tumors of the pharynx and larynx. Transoral laryngeal surgery, albeit effective, has been limited to partial laryngectomy techniques due to limitations in reconstructive options, primarily the need for healing by second intention. These procedures, since they are transoral, have a very low risk of fistula formation, either as a primary approach or for salvage. They would be particularly applicable for salvage if adequate margins can be obtained, as the complication profile should be less, in both incidence and severity. The use of the da Vinci surgical robot (Intuitive Surgical Inc., Sunnyvale, CA) has furthered this approach by facilitating transoral suturing, allowing direct closure of the mucosa and the ability to inset a flap transorally. Therefore, procedures such as TL, which would require mucosal closure, may be performed transorally. The consideration as to whether or not a transoral approach is appropriate is the most critical step in the process. At this time, transoral TL is most appropriate for salvage laryngectomy in patients who do not have nodal metastases.

A meticulous physical examination is necessary to accurately assess the extent and location of the tumor. This includes a requisite endoscopy and biopsy, during which the cancer must be carefully mapped and adjacent structures evaluated. This should be undertaken with the knowledge of the location and extent of the primary
cancer, to avoid overlooking potential residual disease outside the area of the obvious persistence or recurrence. A thorough understanding of the local patterns of spread of laryngeal cancer is essential in anticipating potential areas of submucosal involvement. This is one of the keys to assure resection margins free of tumor. As an example, vocal fold fixation may represent paraglottic extension of disease, raising the possibility of direct extralaryngeal extension through the cricothyroid membrane. The need for a complete assessment of the tumor and adjacent areas cannot be overstated, and such an evaluation is not possible during a routine office examination. Careful assessment of the character of the surrounding tissues (fibrosis, stenosis, telangiectasia) is important to help to identify any impediments to a transoral approach.

A main concern for transoral TL is adequate transoral access. This must be evaluated at endoscopy. Critical patient features to assess include the following: the width of the posterior mandibular arch, dentition, scarring of the oropharyngeal inlet, anterior positioning of the larynx, inferior positioning of the larynx, fibrosis of the laryngopharynx and cervical structures, and the presence of a tracheostomy tube. A narrow posterior mandibular arch will, in most cases, prevent adequate retraction of the tongue base, subsequently limiting the exposure of the vallecula and supraglottis. Most of the retractors available will not fit into such a confined space. Without anterior and superior retraction of the tongue base, transoral TL is not possible. Intact dentition, particularly the maxillary incisors, will severely limit transoral access to the larynx. Central maxillary dentition will inhibit anterior rotation of the surgical endoscope and further crowd an already limited space given the spatial needs of the operating arms, retractors, and suction instruments. Anterior positioning of the larynx has a similar effect, and inferior positioning, if one is unable to mobilize the larynx in a cephalad direction, will result in incomplete access to the hypopharynx and cricoid regions due to the geometry of the area and working depth of the current instrumentation. Fibrosis of the laryngopharyngeal or cervical tissue may similarly limit cephalic mobilization of the larynx and access to the caudal portions of the dissection. Finally, the presence of a tracheostomy, while not a contraindication, has a significant effect on the ease of the operation. The resultant fibrosis will increase the difficulty in dissection the cricotracheal region, a critical step in the procedure. In addition, if the tracheostomy has been performed by another surgeon, it is critical to know which tracheal rings were entered. It is fairly common, particularly if the tracheostomy was urgent, that the tracheal rings entered are more inferior than that anticipated by the surgeon. At this time, the retractors available for transoral access are limited, and one size does not fit all. Frequently multiple retractors, or blades, may be required to complete the operation.

Patients with cancer of the pyriform sinus, unless they are limited to the medial wall, are also not currently candidates for this procedure as instrumentation design limits access to the pyriform apex. That is likely to change as advances in instrument and robot design are made. Primary TL, which is indicated for a large T4a cancer with extralaryngeal spread, would not be appropriate for a transoral approach at this time either. Such a cancer would require resection of the strap muscles, neck dissection, and often the need for flap reconstruction, all of which would negate the current benefits of transoral TL. These indications, however, are likely to change and evolve over time.

Adequate imaging is critical to preoperative planning in the salvage setting. At a minimum, this should include a contrast computed tomography scan with thin cuts through the larynx. This will help to define extralaryngeal extension, which may preclude a transoral approach. Posterior submucosal extension of the tumor is also important to define, as is the course of the carotid artery and its branches. Careful assessment of the cervical lymph nodes is essential, as the need for a lateral neck dissection may also eliminate the benefits of a transoral resection. Metabolic imaging, such as a positron emission tomography scan, will also provide important information about both the local extent of disease and cervical lymph node metastasis and is recommended where available. The application of transoral surgery should never compromise the ability to completely remove the cancer, and anatomic imaging is an important component of tumor assessment and surgical planning.