Anterior commissure involvement

The optimal treatment for early glottic cancer is currently under debate.Reference Yoo, Lacchetti, Hammond and Gilbert^6, Reference Hartl, Ferlito, Brasnu, Langendijk, Rinaldo and Silver⁷ Anterior commissure involvement has been reported in between 17.6 and 45.4 per cent of cases. However, although this is of the utmost clinical importance, it currently has no impact on laryngeal cancer staging.Reference Aydil, Akmansu, Kizil, Yazici, Ustün and Karaloğlu^8–Reference Reddy, Mohideen, Marra and Marks¹²

Anterior commissure involvement is a challenging issue for head and neck oncologists and surgeons for several reasons. Anatomically, the thyroid cartilage is in close proximity, there are no anatomical barriers to prevent tumour spread to the supraglottic portion and cricothyroid membrane, and the perichondrium is absent at this location. These factors can all lead to understaging of the tumour extent. Ossification occurs earlier at this point, which may facilitate tumour invasion, and overstaging is possible if sclerosis encountered in the radiological evaluation is assumed to be cartilage invasion.Reference Hartl, Landry, Bidault, Hans, Julieron and Mamelle¹³ The small size of this region makes radiological evaluation difficult. Tumours may show occult progression without any change in vocal fold mobility, and laryngological examination and endoscopic surgical exposure are suboptimal.

Treatment options for early glottic cancers include radiotherapy (RT), endolaryngeal surgery with or without laser, and open partial laryngectomy. The most important tumour-related factors for deciding between RT and endolaryngeal surgery are probably tumour location and extent. The anterior commissure is a glottic subsite with a high risk of local recurrence, and this should always be considered during treatment planning.

The impact of anterior commissure involvement on local control is considered differently from surgical and RT standpoints. For surgeons, anterior commissure involvement is important for surgical modality selection. Open surgical procedures are suggested to provide the best oncological clearance in cases of anterior commissure involvement.Reference Agrawal and Ha¹⁴ According to Rifai and Khattab, tumours with anterior commissure involvement have a higher tendency toward cartilage invasion, and the anterior portion of the thyroid cartilage should be excised.Reference Rifai and Khattab¹⁵ Many surgeons perform open partial laryngectomy in the presence of anterior commissure involvement, and more extended procedures such as supracricoid laryngectomy are often considered.Reference Spector, Sessions, Chao, Haughey, Hanson and Simpson¹⁶ Anterior commissure involvement is believed to have a large impact on local control; numerous histological and clinical studies have published on this issue.Reference Rifai and Khattab^15, Reference Bradley, Rinaldo, Suárez, Shaha, Leemans and Langendijk^17–Reference Nozaki, Furuta, Murakami, Izawa, Iwasaki and Takahashi²² A consensus statement paper from the UK stated that ‘tumours involving the anterior commissure are associated with a higher local residual tumour and recurrence rate than those without anterior commissure involvement’.Reference Bradley, Mackenzie, Wight, Pracy and Paleri²³ In 2007, a new cordectomy type (type VI), encompassing the anterior commissure and the anterior part of both vocal folds, was proposed by European Laryngological Society Working Committee on Nomenclature.Reference Remacle, Van Haverbeke, Eckel, Bradley, Chevalier and Djukic²⁴

Sachse et al. reported that anterior commissure involvement is associated with an increased recurrence rate (23 per cent) compared with an absence of anterior commissure involvement (5 per cent), and that four out of five recurrences are located at the anterior commissure.Reference Sachse, Stoll and Rudack²⁵ Laccourreye et al. reported that the local failure rate was 23 per cent if there is anterior commissure involvement; this rate is higher than for any other glottic subsite involvement in early glottic cancers.Reference Laccourreye, Weinstein, Brasnu, Trotoux and Laccourreye²⁶

There is also a higher tumour recurrence rate after endolaryngeal surgery for lesions with anterior commissure involvement (Table I). In a large series comprising 263 patients treated by laser microsurgery, the 5-year local control rates for patients with and without anterior commissure involvement were 84 per cent and 90 per cent, respectively, for pathologically staged T_1a tumours and 73 per cent and 92 per cent, respectively, for pathologically staged T_1b tumours.Reference Steiner, Ambrosch, Rödel and Kron²¹ In this series, 62 per cent of patients with local failure initially had anterior commissure involvement, whereas the anterior commissure was tumour free in only 38 per cent of patients with local failure.

Table I Impact of ACI on local control in patients with early glottic cancer treated by endolaryngeal laser surgery

ACI = anterior commissure involvement; w/o = without

Although the local control rate is almost always worse in the presence of anterior commissure involvement after RT, it has not always reached statistical significance in multivariate analyses (Table II). Published reports indicate that anterior commissure involvement is an important tumour parameter for deciding on glottic tumour treatment and an important prognostic factor for local control. In our opinion, the effect of anterior commissure involvement on glottic cancer is comparable with the effect of pre-epiglottic space invasion on supraglottic cancer in terms of occult invasion, and may lead to understaging and an unfavourable outcome if not managed properly.

Table II Impact of ACI on local control in patients with early glottic cancer treated by radiotherapy

ACI = anterior commissure involvement; w/o = without; UVA = univariate analysis; MVA = multivariate analysis; N/A = not available

Subglottic location or extension

Only 1–2 per cent of all laryngeal cancers are derived from the subglottic region: direct invasion of glottic cancers or supraglottic cancer spread via the paraglottic space are more common.Reference Dahm, Sessions, Paniello and Harvey^46–Reference Smee, Williams and Bridger⁴⁸ The rate of subglottic extension is reported to be roughly 10 per cent in early glottic cancers.Reference Khan, Koyfman, Hunter, Reddy and Saxton^41, Reference Murakami, Nishimura, Baba, Furusawa, Ogata and Yumoto⁴⁴ On the other hand, 13 per cent of patients who underwent total laryngectomy for advanced local disease had subglottic extension, and the subglottic extension rate was 31 per cent for those with a transglottic tumour.Reference Dadas, Uslu, Cakir, Ozdoğan, Caliş and Turgut⁴⁹

Since the subglottic mucosa is thin, involvement of the underlying cartilage framework and spread to the neck through the cartilage or cricothyroid membrane can easily occur. In many cases, subglottic extension is considered a contraindication for conservative laryngeal surgery. The prognosis of primary subglottic cancer is generally believed to be poor.Reference Dahm, Sessions, Paniello and Harvey^46, Reference Garas and McGuirt^47, Reference Paisley, Warde, O'Sullivan, Waldron, Gullane and Payne^50, Reference Hata, Taguchi, Koike, Nishimura, Takahashi and Komatsu⁵¹ According to Chen et al., the subglottic extension of glottic cancer predicts a poorer outcome and the 5-year local control rates for patients with and without subglottic extension are 9 per cent and 77 per cent, respectively.Reference Dagan, Morris, Bennett, Mancuso, Amdur and Hinerman⁵² In two separate studies, subglottic extension was the only tumour-related factor to affect RT outcomes in T₂ glottic cancers in multivariate analyses.Reference Garden, Forster, Wong, Morrison, Schechter and Ang^53, Reference Chen, Chang, Tsang, Liao and Chen⁵⁴ Mortuaire et al. reported that subglottic invasion has a significant impact on the local control rate.Reference Mortuaire, Francois, Wiel and Chevalier⁵⁵ Subglottic extension is also an independent risk factor for thyroid gland involvement and level VI lymph node metastasis.Reference Mendelson, Al-Khatib, Julien, Payne, Black and Hier^56, Reference Medina, Ferlito, Robbins, Silver, Rodrigo and de Bree⁵⁷

Stomal recurrence is the most devastating event and has the highest mortality rate (2.6–5 per cent) after total laryngectomy.Reference Rubin, Johnson and Myers^58–Reference Yuen, Wei, Ho and Hui⁶² Almost all patients with stomal recurrence die within 12 months despite aggressive therapy.Reference Ampil, Ghali, Caldito and Baluna^61–Reference Reddy, Narayana, Melian, Kathuria, Leman and Emami⁶³ The most important risk factor for stomal recurrence is believed to be subglottic extension. Rubin et al. reported that 80 per cent of patients with stomal recurrence had subglottic extension.Reference Rubin, Johnson and Myers⁵⁸ Reddy et al. reported that stomal recurrence occurred in 100 per cent and 6 per cent, respectively, of patients with and without subglottic extension.Reference Reddy, Narayana, Melian, Kathuria, Leman and Emami⁶³

Currently, impaired vocal fold mobility and subglottic extension have the same impact on tumour staging for glottic cancer. Since subglottic extension is a significant negative prognostic indicator, it should be categorised as a higher tumour stage and primary subglottic cancer staging should start at a higher tumour stage (Table III).

Table III Suggested changes to laryngeal, upper/alveolar ridge and hard palate cancer staging

Underlying bone invasion in hard palate and maxillary alveolar ridge cancers

The hard palate mucosa is very close to the underlying bone; tumours derived from the hard palate can easily invade the underlying bone at an early stage of development. The same is true for cancers of the upper alveolar ridge. Most studies report that the incidence of T₄ tumours at these locations may be abnormally high: some have reported that 58–71 per cent of upper alveolar ridge and the hard palate tumours are T₄ stage.Reference Aydil, Kızıl, Bakkal, Köybaşıoğlu and Uslu^64–Reference Brown, Bekiroglu, Shaw, Woolgar and Rogers⁶⁷ According to a detailed radiological study, maxillary floor perforation was detected in 57 per cent of cases of maxillary alveolar ridge cancer.Reference Ogura, Kurabayashi, Sasaki, Amagasa, Okada and Kaneda⁶⁸ However, it is doubtful that bone invasion is an important poor prognostic factor for cancer of these locations. The local failure rate was only 12 per cent in a series reported by Mourouzis et al., despite almost three-quarters of tumours being initially staged as T₄.Reference Mourouzis, Pratt and Brennan⁶⁵ Brown et al. reported local recurrence in 8 out of 43 patients despite 25 tumours being classified as T₄ stage.Reference Brown, Bekiroglu, Shaw, Woolgar and Rogers⁶⁷ In two large series comprising 139 and 97 patients, nearly half of the tumours were T₄ stage.Reference Morris, Patel, Shah and Ganly^69, Reference Eskander, Givi, Gullane, Irish, Brown and Gilbert⁷⁰ In one of these studies, the five-year disease-specific survival rate was 78 per cent; this is surprisingly high considering the large proportion of T₄ tumours.Reference Eskander, Givi, Gullane, Irish, Brown and Gilbert⁷⁰ The other study also reported a high five-year survival rate of 81 per cent in the absence of regional recurrence.Reference Morris, Patel, Shah and Ganly⁶⁹

In our opinion, invasion of the upper and lower jaw should not be considered in the same way. Unlike the mandible, the hard palate is not a mobile part of the maxillofacial skeleton and prosthetic rehabilitation is much easier to achieve and more effective. Bone invasion may not result in serious morbidity related to surgical treatment and may not have much impact on the survival rate. In contrast, maxillary cancers causing bone erosion or destruction including extension into the hard palate and/or middle nasal meatus are categorised as T₂. In conclusion, the effect of bone invasion should be decreased for tumour staging of malignant hard palate and upper alveolar ridge neoplasms in the next edition of The AJCC Cancer Staging Manual, similar to maxillary cancer staging (Table III).

Tumour thickness in early oral cancer

The histopathological characteristics of oral cancer specimens (including surgical margin clearance, tumour thickness and perineural and lymphovascular invasion) are probably more important than their gross clinical and radiological features. Tumour thickness was shown to be the most important predictor of survival for early oral cancer in a large retrospective study, and integrating tumour thickness into routine staging was suggested.Reference Hubert Low, Gao, Elliott and Clark⁷¹ In their prospective studies, Po Wing Yuen et al. and Yuen et al. reported that tumour thickness was the only factor to significantly influence nodal metastasis, local recurrence rate and survival.Reference Po Wing Yuen, Lam, Lam, Ho, Wong and Chow^72, Reference Yuen, Lam, Wei, Lam, Ho and Chow⁷³ Liao et al. also identified tumour thickness as an independent risk factor for neck control and survival in early oral cancer in a large series.Reference Liao, Lin, Fan, Wang, Ng and Lee⁷⁴ According to Gonzalez-Moles et al., tumour thickness has the greatest influence on survival among many factors including tumour and nodal stages in tongue cancer.Reference Gonzalez-Moles, Esteban, Rodriguez-Archilla, Ruiz-Avila and Gonzalez-Moles⁷⁵ Numerous other studies have investigated the relationship between tumour thickness and neck metastasis, and tumour thickness was consistently identified as a strong predictor of cervical metastasis in a meta-analysis.Reference Huang, Hwang, Lockwood, Goldstein and O'Sullivan⁷⁶ In conclusion, there is strong evidence that tumour thickness has a significant influence on disease control and survival compared with many other factors influencing oral cancer. Therefore, tumour thickness should be considered in oral cancer staging.

Carotid artery involvement

Carotid artery involvement usually has catastrophic consequences: surgical treatment is difficult, complication and mortality rates are high, quality of life is low, and survival is dismal even after complete resection of the tumour.Reference Freeman, Hamaker, Borrowdale and Huntley^77–Reference Biller, Urken, Lawson and Haimov⁸³ Freeman et al. reported a 1-year survival rate of only 40 per cent, and only 2 out of 58 patients were alive at the third year despite aggressive surgical intervention.Reference Freeman, Hamaker, Borrowdale and Huntley⁷⁷ According to Roh et al., the 2-year survival rate is 13 per cent.Reference Roh, Kim, Choi, Lee, Cho and Nam⁷⁸ Studies with the most optimistic results reported 2- and 5-year survival rates of 21–32 per cent and 12–28 per cent, respectively.Reference Zhengang, Colbert, Brennan, Xue, Yongfa and Pingzhang^79, Reference Ozer, Agrawal, Ozer and Schuller⁸⁰ Unfortunately more than half of patients had locoregional recurrence.Reference Freeman, Hamaker, Borrowdale and Huntley^77, Reference Manzoor, Russell, Bricker, Koyfman, Scharpf and Burkey⁸¹ It was reported that 16 per cent to 33 per cent of patients had distant metastasis within a relatively short follow-up period.Reference Freeman, Hamaker, Borrowdale and Huntley^77, Reference Ozer, Agrawal, Ozer and Schuller^80–Reference Biller, Urken, Lawson and Haimov⁸³

In the case of direct invasion of the carotid artery by a primary tumour, the tumour stage is classified as T_4b. However, carotid artery involvement almost always results from extracapsular spread of a metastatic neck node; direct invasion of the carotid artery by a primary tumour is rare.Reference Freeman, Hamaker, Borrowdale and Huntley^77, Reference Ozer, Agrawal, Ozer and Schuller^80, Reference Manzoor, Russell, Bricker, Koyfman, Scharpf and Burkey^81, Reference Freeman^84, Reference Pons, Ukkola-Pons, Clément, Gauthier and Conessa⁸⁵ Although the main route of carotid artery invasion is via neck nodes, carotid artery involvement is not taken into account in neck node staging. It is clear that carotid artery involvement is an important prognostic indicator that should be independently incorporated into nodal staging rather than into primary tumour assessment. Since it is a very poor prognostic indicator, carotid artery involvement should be considered separately; our suggested stage for carotid artery involvement is N_3b. In our opinion, carotid artery involvement should be taken into account in both clinical (radiological) and pathological node staging processes. However, there is evidence that extracapsular spread may not affect prognosis in the same manner in HPV-positive patients. This will be discussed in the next section. Similarly, nodal staging of nasopharyngeal cancer differs from that of other mucosal head and neck cancers, and the adverse prognostic effect of carotid artery involvement in nasopharyngeal cancer is currently unclear. For these reasons, we do not suggest making changes to nodal staging for nasopharyngeal cancer and HPV-positive oropharyngeal cancer.

Extracapsular spread of human papillomavirus negative tumours

It is clear that the prognosis of patients with extracapsular spread in neck nodes is very poor.Reference Cooper, Pajak, Forastiere, Jacobs, Campbell and Saxman^86–Reference de Juan, García, López, Orús, Esteller and Quer⁹⁰ The presence of extracapsular spread increases the regional failure rate and the incidence of distant metastasis.Reference Myers, Greenberg, Mo and Roberts^91–Reference Greenberg, Fowler, Gomez, Mo, Roberts and El Naggar⁹³ The incidence of extracapsular spread within neck dissection specimens is reported to be between 17 per cent and 36 per cent.Reference Myers, Greenberg, Mo and Roberts^91–Reference Jose, Coatesworth, Johnston and MacLennan⁹⁵ Alvi and Johnson reported that extracapsular spread is even present in 17 per cent of N₀ necks and predicts a poor outcome in patients with occult cervical metastasis.Reference Alvi and Johnson⁹⁶ Jose et al. prospectively investigated the importance of microscopic extracapsular spread and found no difference in actual or recurrence-free survival between microscopic and macroscopic extracapsular spread.Reference Jose, Coatesworth, Johnston and MacLennan⁹⁵ According to these authors, both forms of extracapsular spread are of prognostic importance in survival. Greenberg et al. also reported that survival is not significantly different in patients with extracapsular spread of either less than or more than 2 mm.Reference Greenberg, Fowler, Gomez, Mo, Roberts and El Naggar⁹³

It should be kept in mind that extracapsular spread does not indicate a large, fixed neck mass. Ghadjar et al. retrospectively analysed the dimensions of nodes with extracapsular spread and found that 60 per cent of patients with at least one extracapsular spread positive lymph node have at least one extracapsular spread positive node smaller than 10 mm.Reference Ghadjar, Simcock, Schreiber-Facklam, Zimmer, Gräter and Evers⁹⁷ The median diameter of the small extracapsular spread positive lymph nodes was 7 mm, and presence of an extracapsular spread positive lymph node smaller than 7 mm was a significant prognostic factor for decreased regional relapse-free survival, distant metastasis-free survival and overall survival. According to these authors, small extracapsular spread positive metastatic nodes may represent a more invasive tumour. Ghadjar et al. also reported the mean and median extent of extracapsular spread as 2 mm and 1 mm, respectively.Reference Ghadjar, Simcock, Schreiber-Facklam, Zimmer, Gräter and Evers⁹⁷ These authors reported that more than 90 per cent of extracapsular spread positive lymph nodes are smaller than 2 cm and the extent of extracapsular spread is less than 5 mm in 97 per cent of nodes. These results are almost the same as those of a previous study by Apisarnthanarax et al., which showed that extracapsular spread positive nodes are usually smaller than the 3 cm cut-off value between N₁ and N_2a.Reference Apisarnthanarax, Elliott, El-Naggar, Asper, Blanco and Ang⁹⁸

Extracapsular spread was previously considered one of the worst prognostic features of head and neck cancers, and patients with extracapsular spread were reported to benefit from adjuvant chemoradiotherapy in two landmark clinical trials.Reference Cooper, Pajak, Forastiere, Jacobs, Campbell and Saxman^86, Reference Bernier, Domenge, Ozsahin, Matuszewska, Lefèbvre and Greiner⁸⁷ Today, regardless of metastatic node size and number, extracapsular spread is considered one of the two main adjuvant chemoradiotherapy indications (together with a positive surgical margin). As the most important tumour-related adverse feature to predict survival, extracapsular spread should be considered in nodal staging, as previously advocated by de Juan et al. and Shaw et al.Reference de Juan, García, López, Orús, Esteller and Quer^90, Reference Shaw, Lowe, Woolgar, Brown, Vaughan and Evans⁹⁴ In our opinion, extracapsular spread should have more impact on nodal staging. As it is a very important independent poor prognostic indicator and the depth of extracapsular spread is not related to the oncological outcome, the pathological nodal stage should be N₃ in the presence of extracapsular spread, regardless of node size or depth of extracapsular spread. However, the effect of extracapsular spread on prognosis is not the same in HPV-positive patients. Sinha et al. reported that extracapsular spread is not associated with poorer survival in HPV-positive oropharyngeal cancer patients.Reference Sinha, Lewis, Piccirillo, Kallogjeri and Haughey⁹⁹ More recently, Maxwell et al. also showed that extracapsular spread was not associated with worse disease-specific survival in HPV-positive oropharyngeal cancer patients.Reference Maxwell, Ferris, Gooding, Cunningham, Mehta and Kim¹⁰⁰