The treatment of glottic cancer is greatly influenced by the secondary goal of voice preservation. Mobility of the vocal cords is a critical factor in selecting treatment. For small cancers (T1, T2) with mobile vocal cords, radiation therapy alone for cure achieves excellent local control rates (T1, 85% to 95%; T2, 65% to 75%) and overall survival rates similar to those for surgical resection. Voice quality, although often impaired by radiation, is generally better than that following surgical resection. Local control rates are 10% to 15% better with primary surgery, but local recurrences after definitive radiation can often be salvaged by subsequent surgery, and this combined approach results in overall survival figures comparable to those with primary surgery.

Tumor involvement of the anterior commissure or arytenoids has been associated with higher local recurrence rates for radiation alone, but this may have been related to understaging. As with supraglottic cancers, careful clinical tumor staging is necessary since underestimation of tumor extent is common. The “irradiate-and-watch” treatment strategy is predicated on close follow-up in order to detect recurrences when they are still salvageable by surgery. Delay in the diagnosis of recurrent glottic cancers after radiation is more frequent than with supraglottic cancers and often requires total laryngectomy for cure. Thus, unreliable patients or patients who are difficult to examine may be more suitable for primary surgical treatment.

Survival figures in radiotherapy series are comparable to local control figures for surgery, reflecting the effectiveness of surgical salvage and the fact that few patients with early-stage glottic cancer die of their disease.

The 5-year survival rates for T1 lesions range from 80% to 95% with either primary surgery or radiation (

Table 90-16). Rates for T2 lesions are generally in the range of 70% to 80%, but these rates are decreased 10% to 15% (local control rates drop 20% to 25%) when the mobility of the vocal cords is impaired or transglottic spread is present. Lesions with impairment due to invasion of muscle behave more like T3 cancers and have a poorer response with radiation alone. Transglottic cancers and those with subglottic extension have higher rates of regional metastases and often require total laryngectomy for cure. In selected patients with these more advanced lesions or impaired vocal cord mobility, extended hemilaryngectomy, or subtotal laryngectomy with resection of a major portion of the cricoid cartilage, can achieve excellent cure rates. Voice quality is poor with these extensive procedures, and chronic aspiration or permanent tracheostomy may result. Additionally, these procedures are technically challenging and experience dependent. However, with proper patient selection, these procedures can be well tolerated.

Management of advanced T3 glottic cancers has historically consisted of total laryngectomy with or without postoperative radiation therapy. While older series show suboptimal control rates (20% to 35%) and survival rates (10% to 50%) for unselected sets of T3 and T4 tumors treated with radiation alone, it is now recognized that with proper selection, radiotherapy control rates for T3 lesions can approach 80%. In patients without regional metastases, local tumor control rates with surgery alone are excellent. Significant increases in local control with the addition of radiation therapy have not been clearly demonstrated. However, in patients with regional metastases, overall prognosis is poor and recurrence in the neck is a major problem when surgery alone is used. Better regional tumor control rates are achieved with the addition of adjuvant radiation therapy. Because rates of occult regional metastases approach 30% in patients with advanced glottic (T3, T4) cancers, elective modified or selective node dissections for staging purposes are recommended when surgery is performed for primary disease. Demonstration of histologically positive nodal metastases has been used as an indication for postoperative radiation. Surgery alone is curative in 50% to 80% of patients without nodal metastases, and this decreases to less than 40% if metastases are present.

Considerable controversy surrounds the use of definitive radiation with surgical salvage in patients with advanced (T3N0, T4N0) but localized glottic cancers. The radiation-alone concept presumes equal overall survivorship as compared with primary laryngectomy, with associated low complication rates. Overall survival rates range from 50% to 55%, with larynx preservation in 60% to 70% of these patients. High complication rates, however, have been reported with late surgical salvage of radiation failures. The overall patterns are confusing and based entirely on retrospective series. The resolution of this controversy will require carefully designed prospective studies that include assessments not only of survival, but also of voice and quality-of-life issues and complication rates.

A subset of laryngeal cancers that warrant special consideration are those that involve both the glottic and supraglottic regions (transglottic). These cancers are usually advanced and are associated with a high incidence (30% to 50%) of regional metastases, extralaryngeal spread, and vocal cord fixation. Although clinical understaging is common, occasionally these cancers are quite superficial and amenable to conservation surgical techniques. Most patients, however, require total laryngectomy. In a careful review of 152 cases of transglottic carcinomas, Mittal reported a 55% cure rate with combined therapy as compared with a 5-year survival of 8% with radiation alone.

Primary subglottic carcinomas account for less than 5% of laryngeal cancers. Limited data may support the use of primary radiotherapy for early-stage (T1, T2) lesions. However, these lesions are usually advanced at diagnosis and require surgery (laryngectomy) and bilateral peritracheal lymph node dissections since regional metastases occur in at least 20% of these patients. Many reported series contain glottic primaries with subglottic extension and confuse these analyses. Surgical treatment generally requires total laryngectomy combined with resection of adjacent soft tissues (thyroid gland, strap muscles, peritracheal lymph nodes). Involvement of the cricoid precludes any conservation laryngeal surgery. Five-year survival rates of 36% for radiation therapy and 42% for surgery have been reported. Also, cure rates as high as 70% have been reported in a small number of patients treated with combined therapy. The addition of adjuvant radiation offers the advantage of improved regional control rates and treatment of peritracheal and upper mediastinal lymph nodes. Histologically positive lymph nodes can be found in 65% of cases. The risk of stomal recurrence increases substantially with cancers that involve the subglottic larynx, particularly if prior tracheostomy was necessary for impending airway obstruction. Early aggressive treatment (often within 24 h) has been recommended for patients requiring tracheostomy for subglottic extension of laryngeal cancers.

Patterns of failure for glottic carcinoma differ somewhat from other laryngeal sites. Local failures are uncommon with primary surgical therapy and account for fewer than 10% of recurrences. However, after primary radiation therapy for local glottic primaries, recurrences account for 10% to 50% of failures. Regional nodal recurrences are seen in 15% to 30% of patients with advanced disease who are treated with surgery alone. This contrasts to supraglottic cancers, in which regional recurrences are a major site of failure.

It was previously thought that distant metastases from laryngeal cancers were uncommon, accounting for less than 10% of failures. However, with improved locoregional control, the recognized incidence of distant metastasis is increasing. Distant spread is approximately four times more common with supraglottic than with glottic cancers. Rates of distant metastases associated with glottic cancer have increased, however, with the use of combined therapy and have been reported in approximately 20% of patients with advanced disease. Rates appear to be directly related to the extent of nodal disease, with reported rates as high as 40% to 50% of failures attributed to distant metastases in patients with N2 or N3 disease.