Introduction
Hypopharyngeal cancer is relatively uncommon, with a higher prevalence in South Asia.Reference Franceschi, Bidoli, Herrero and Munoz1 For organ preservation, primary radiotherapy or chemoradiotherapy is the preferred choice of treatment in the early stages. However, surgery followed by adjuvant radiotherapy or chemoradiotherapy is required for more advanced tumours, and for salvage after radiotherapy.
Most hypopharyngeal cancers are squamous cell carcinomas, and the pyriform sinus is the most common hypopharynx subsite involved.Reference Barnes and Johnson2, Reference Mendenhall, Werning, Pfister, DeVita, Lawrence and Rosenberg3 Surgery for hypopharyngeal cancer often requires laryngopharyngectomy, or laryngectomy with partial pharyngectomy. Tumour from the hypopharynx can spread to adjacent structures, including the thyroid or cricoid cartilages, the thyroid gland, and the oesophagus. The thyroid gland is therefore routinely removed during laryngectomy to achieve adequate tumour-free margins. The thyroid gland adjacent to the involved side is routinely removed; however, some surgeons prefer to perform a total thyroidectomy.
The incidence of thyroid gland involvement varies from 1 to 50 per cent.Reference Ogura4–Reference Ceylan, Koybasioglu and Yilmaz8 Most of the relevant literature concerns laryngeal cancer; few studies have reported the incidence of thyroid gland involvement in hypopharyngeal cancer specifically. Hypopharyngeal cancer has a different pattern of spread to laryngeal cancer due to the slight difference in anatomical location. Hence, the incidence and the factors predicting thyroid gland involvement may vary for laryngeal and hypopharyngeal cancer.
Resection of the thyroid gland can lead to hypothyroidism and hypoparathyroidism. When both resection and radiotherapy are carried out, the risk of hypothyroidism increases to 70 per cent.Reference Fried9–Reference Ballenger and Snow11 Hypothyroidism in itself can result in fatigue, weight gain, irritability and depression. Hence, thyroid resection should only be carried out if it is absolutely necessary.Reference Croce, Moretti, Bianchedi, Boccia and de Vincentiis12, Reference Gilbert, Cullen, van Nostrand, Bryce and Harwood13
This retrospective study aimed to determine the incidence of thyroid gland involvement in carcinoma of the hypopharynx and to evaluate the various factors predicting thyroid gland involvement.
Materials and methods
A retrospective analysis was conducted of all patients with hypopharyngeal cancer who underwent total laryngectomy with partial or total pharyngectomy at Tata Memorial Hospital, Mumbai between 2004 and 2010. Cases were selected based on the following eligibility criteria: biopsy-confirmed squamous cell carcinoma of the hypopharynx; surgery carried out with curative intent at Tata Memorial Hospital as a primary or salvage treatment; and total laryngectomy with partial or total pharyngectomy performed.
Between 2004 and 2010, 1012 patients underwent laryngectomy for carcinoma of the larynx or hypopharynx. After the eligibility criteria were applied, there were 358 cases of hypopharyngeal cancer for evaluation. Relevant data were retrieved from the hospital's electronic medical records. Those patients treated exclusively for laryngeal cancer or those who underwent partial laryngectomy were not included in this study.
All patients underwent a complete pre-operative evaluation, which included a direct laryngoscopy, contrast-enhanced computed tomography and barium study. The lesions of these patients predominantly involved the hypopharynx. Treatment therefore required partial or complete excision of the pharyngeal walls, achieved via partial or total laryngopharyngectomy.
Eighty-five per cent of patients (304 out of 358) underwent surgery as the first modality of treatment, and the remaining 15 per cent of patients (54 of 358) received radiotherapy or chemoradiotherapy prior to surgery. Cartilage erosion was the most common indication for surgery (in 55 per cent of cases), followed by exolaryngeal spread, or bulky and extensive soft tissue disease.
Statistical analysis was carried out using the Statistical Package for the Social Sciences software version 20.0 (SPSS; IBM, Armonk, New York, USA). P values were calculated using chi-square or t-tests. A p value of less than 0.05 was considered significant. Univariate and multivariate analyses were conducted using the chi-square test and the binary logistic regression test respectively.
Results
The mean age of the patient population was 61 years, with a male to female ratio of 12:1. The pyriform sinus was the most common site of disease (Table I). The adjacent laryngeal framework was involved in the majority of cases (70 per cent, 250 of 358). Seventy-one per cent of patients (253 of 358) had neck nodal metastases; 81 per cent of these patients (205 of 253) had extracapsular spread.
Table I Demographic factors
Patients were aged 24–88 years (mean 61 years, median 62 years). *Total n = 358. †Refers to treatment naïve patients (n = 304)
The common indications for thyroidectomy as a part of total laryngectomy included: cartilage erosion, in 53 per cent of cases (191 out of 358); extralaryngeal disease, in 30 per cent of cases (107 of 358); and disease extending to the subglottis, in 25 per cent of cases (90 of 358). Only 4 per cent of patients (13 of 358) underwent total thyroidectomy; hemi-thyroidectomy was carried out in the majority of patients (96 per cent, 345 of 358). Of those who underwent total thyroidectomy, four had bilateral and three had unilateral thyroid lobe involvement. Two patients had gross extralaryngeal disease without thyroid gland involvement. One patient had micropapillary thyroid carcinoma, and one had nodal metastasis of papillary carcinoma with a normal thyroid gland. Two patients had benign nodules in the ipsilateral thyroid lobe.
The incidence of pathological thyroid gland involvement in this series was 13 per cent (47 out of 358 patients). Associations between thyroid gland involvement and other various factors were evaluated, including disease site, cartilage erosion, extralaryngeal spread, subglottic extension, level VI lymph node metastasis, tumour stage, lateral neck nodes, extracapsular spread, prior tracheostomy, perineural invasion and lymphovascular emboli, and transglottic involvement.
Univariate analysis revealed statistically significant associations (p < 0.05) between thyroid gland involvement and the following factors: anatomical location of the tumour in the hypopharynx, perineural invasion, subglottic involvement, extralaryngeal spread, prior tracheostomy and cartilage erosion. Multivariate analysis revealed significant associations (p < 0.05) between thyroid gland involvement and prior tracheostomy, subglottic extension of the tumour, thyroid cartilage erosion, extralaryngeal spread and post-cricoid involvement (Table II).
Table II Factors associated with thyroid gland involvement*
* Statistically significant factors on multivariate analysis. Pts = patients
Discussion
Most of the previously published clinical studies analysed hypopharyngeal and laryngeal cancer cases together. This is partly because hypopharyngeal cancer is uncommon, and because hypopharyngeal and laryngeal tumours are located at anatomically nearby sites. Because of this anatomical proximity, hypopharyngeal cancer surgery often entails laryngectomy also. To our knowledge, the incidence of thyroid gland involvement in hypopharyngeal cancer has not been previously studied. In this study, as in many other studies, hemi or total thyroidectomy was carried out on all patients who underwent total laryngectomy.Reference Al-Khatib, Mendelson, Kost, Zeitouni, Black and Payne14, Reference Bahadur, Iyer and Kacker15 However, in our series, the disease involved the thyroid gland in only 13 per cent of cases (Table III).
Table III Incidence of thyroid gland involvement
Pts = patients
In 1955, Ogura recommended total laryngectomy with ipsilateral thyroidectomy for all cases of laryngeal cancer.Reference Ogura4 In 1973, Harrison proposed total thyroidectomy for post-cricoid and subglottic tumours, as they are associated with a high incidence of contralateral thyroid gland involvement.Reference Harrison5 A study by Kim et al. reported an incidence rate of 14 per cent for thyroid gland involvement in hypopharyngeal cancer; this incidence was associated with subglottic tumours.Reference Kim, Han, Byun, Lee, Cho and Kim17 A meta-analysis of laryngeal cancer cases showed that the incidence of thyroid gland involvement was 8 per cent.Reference Mendelson, Al-Khatib, Julien, Payne, Black and Hier18
In the meta-analysis by Mendelson et al. subglottic extension of the tumour (>10 mm), a transglottic tumour and a subglottic subsite significantly increased the risk of thyroid gland invasion by laryngeal carcinoma.Reference Mendelson, Al-Khatib, Julien, Payne, Black and Hier18 In the current study, subglottic extension was also an important factor for thyroid gland involvement. However, transglottic involvement was not a significant factor (p = 0.093).
Many previous studies indicate that prior tracheostomy is a risk factor for thyroid gland involvement. This could be attributed to subglottic extension of the disease, which requires tracheostomy. In fact, in our series, 49 per cent of patients with subglottic disease had had a prior tracheostomy.
• Thyroid glands are routinely removed en bloc during laryngectomy
• There are no objective criteria for deciding the extent of thyroid gland resection
• In this study, the thyroid gland was involved in 13 per cent of hypopharyngeal cancer cases
• The thyroid gland was removed unnecessarily in most patients
• Hemi-thyroidectomy should be performed when the post-cricoid area or subglottis are involved, or if tumour extends to neck soft tissues
In a study by Hilly et al. the incidence of thyroid gland involvement in laryngeal cancer was 21 per cent; thyroid gland involvement was associated with salvage procedures and paratracheal metastases.Reference Hilly, Raz, Vaisbuch, Strenov, Segal and Koren19 In the present study, neither of these factors were associated with thyroid gland involvement.
In a study by Brennen et al. the incidence of thyroid gland involvement was 8 per cent. All lesions were found to have transglottic growth and laryngeal cartilage invasion.Reference Brennan, Meyers and Jafek16 In the present study too, cartilage invasion was significantly associated with thyroid gland involvement (unlike transglottic involvement, which was not a significant factor).
The location of the tumour within the hypopharynx, involving the post-cricoid area for instance, is often associated with a higher incidence of thyroid gland involvement.Reference Harrison5, Reference Bahadur, Iyer and Kacker15 Likewise, in our series, we observed thyroid gland involvement in 47 per cent of the cases in which the post-cricoid region was implicated (p < 0.05).
In the study by Hilly et al. there was cartilage invasion in 10 of 11 cases with thyroid gland involvement, and all tumours were in proximity to (or in contact with) areas of calcification and ossification.Reference Hilly, Raz, Vaisbuch, Strenov, Segal and Koren19 Similarly, in the current study, multivariate analysis revealed a significant association between cartilage erosion and thyroid gland involvement. However, it must be noted that cartilaginous invasion is often associated with subglottic extension or extralaryngeal disease in hypopharyngeal cancer.
Conclusion
The thyroid gland was involved in 13 per cent of hypopharyngeal cancer cases (47 out of 358) and was thus removed unnecessarily in the majority of patients. Cases that involved the post-cricoid area, subglottic extension, extralaryngeal spread or prior tracheostomy were associated with a higher risk of thyroid gland involvement. It is suggested that hemi-thyroidectomy is sufficient in most such cases to achieve adequate tumour-free margins.
