Computed tomographic scan showing the inclusion criteria for all patients—the goiter and the carina tracheae are seen on the same computed tomographic section.
Sancho JJ, Kraimps JL, Sanchez-Blanco JM, Larrad A, Rodríguez JM, Gil P, Gibelin H, Pereira JA, Sitges-Serra A. Increased Mortality and Morbidity Associated With Thyroidectomy for Intrathoracic Goiters Reaching the Carina Tracheae. Arch Surg. 2006;141(1):82-85. doi:10.1001/archsurg.141.1.82
Complications associated with thyroidectomy for intrathoracic goiters have been underestimated because of the lack of a precise definition of high-risk patients.
Retrospective multicenter multinational review of medical records and radiographic images of patients who underwent thyroidectomy for intrathoracic goiters reaching the carina tracheae. Demographic, clinical, operative, anatomical, and pathological data were recorded.
There were 35 patients (mean ± SE age, 63 ± 11 years) included in the study. In 4 patients, the goiter was asymptomatic; 10 patients had dysphagia, 24 patients had dyspnea, and 3 patients had superior vena cava syndrome. A median sternotomy was required in 12 patients and a right-sided thoracotomy in 1 patient. The mean ± SE operative time was 145 ± 72 minutes (range, 50-360 minutes). Transient hypoparathyroidism developed in 13 patients. Four patients experienced transient hoarseness, and 1 patient had permanent vocal cord paralysis. There were no significant differences between the proportion of patients who underwent or did not undergo sternotomy or thoracotomy regarding vocal cord dysfunction (2 [15%] of 13 patients vs 3 [13%] of 22 patients) or hypoparathyroidism (5 [38%] of 13 vs 6 [28%] of 22 patients). The mean postoperative hospital stay was 10 days (range, 2-84 days). Four patients required reoperation. Two patients died. Nine of 14 patients with thyroid glands weighing at least 260 g required sternotomy vs 3 of 14 patients with thyroid glands weighing less than 260 g (P = .02). Overall, 18 [52%] of 35 patients were discharged without any complication.
Intrathoracic goiters reaching the carina tracheae carry a high unreported risk of sternotomy, postoperative complications, reoperation, and death.
Intrathoracic goiters (ITGs) are of clinical importance because of the severity of the compressive symptoms and the potential high rate of postoperative complications in the most severe cases. There is a common belief that most substernal goiters are easily removed via cervical incision and that the cervical or transternal approach can be used without significant mortality or morbidity.1,2
Not all ITGs, however, are equal. Several factors are indicators of a difficult approach requiring sternotomy or causing postoperative complications.3 Size, location, and depth of an ITG are risk factors that may require a thoracic approach and carry significant morbidity.
This study was designed to investigate whether deep ITGs are associated with major surgical risk. A multicenter series of patients with substernal goiters extending as far down as the carina tracheae was studied, and the risks of sternotomy and postoperative complications were compared with those published in the recent literature.
Six endocrine surgery units in Spain and France with comprehensive medical records databases participated in the study. Medical records and radiographic images of patients who underwent thyroidectomy for substernal goiter reaching the carina tracheae between January 1, 1987, and December 31, 2003, were evaluated. Operative and discharge notes were reviewed, and results were entered in a database. Demographics, preoperative clinical and laboratory data, intraoperative findings, indications for sternotomy, and postoperative complications were recorded.
The study inclusion criterion was at least 1 radiographic image, preferably a thoracic section of a contrast-enhanced computed tomographic (CT) image, clearly showing the carina tracheae and the ITG. The CT or magnetic resonance images were reassessed by 2 of us (J.J.S. and A.S.-S.), who made a final decision whether to include or exclude the case. In all cases, it was required that the lower pole of the ITG could be visualized in the same image section as the main bronchial bifurcation (Figure).
Data are reported as arithmetic means and as standard errors of the means, as most of the numerical values differed significantly from a normal distribution. Frequency distributions are reported as percentages when appropriate. Stepwise multiple logistic regression (StatView version 5.5; SAS Institute Inc, Cary, NC) was used to identify which variables and which cut points could help predict the need for sternotomy.
There were 35 patients included in the study; 9 (26%) of them were men. The mean age was 63 ± 11 years (age range, 36-83 years). Six patients were asymptomatic, 4 of whom were referred because the goiter had been discovered on a chest x-ray film and 2 because of the cosmetic defect. The remaining 29 patients were symptomatic, among whom 10 (34%) had dysphagia, 24 (69%) had some degree of dyspnea (generally worse in the recumbent position), and 3 (9%) had superior vena cava syndrome. Two patients in whom the ITG was a recurrence had undergone subtotal thyroidectomies 8 and 22 years previously. If known to be preexisting, the goiter had been neglected a mean of 9.3 ± 2.6 years (range, 0.1-42.0 years) since the first recorded symptom.
Some degree of hyperthyroidism was detected in 13 of the patients, with 6 taking antithyroid drugs at the time of diagnosis. Direct laryngoscopy was selectively performed to assess vocal cord function in patients with hoarseness. Two patients had unilateral preoperative vocal cord palsy, due to probable stretching or compression of the inferior laryngeal nerve in one patient and following a previous subtotal thyroidectomy in the other patient.
Computed tomographic images were available for 32 patients (91%), magnetic resonance images for 2 patients (6%), and thyroid images for 18 patients (52%). Two patients had only plain chest x-ray films, in which the goiter could be clearly identified as reaching the carina tracheae. In 2 cases, 1 of which was a recurrence, the goiter was completely intrathoracic, without a palpable cervical component. Fine-needle aspiration cytology, aiming at a dominant nodule within the enlarged thyroid gland, was performed in 13 patients, revealing a papillary cancer in 1.
The intrathoracic extensions of the goiters were predominantly on the right side in 20 patients (57%). These occupied the anterior, lateral, posterolateral, and posterior spaces in an almost even distribution.
Total thyroidectomy was performed in 32 patients (91%), and bilateral subtotal lobectomy was performed in the remaining 3 patients (9%). All cases were performed by senior endocrine surgeons. The thyroid glands removed had a mean maximum diameter of 12.2 ± 3.0 cm (range, 6-19 cm) and a mean weight of 257 ± 148 g (range, 60-650 g) (weight was available in 28 patients). The mean operative time was 145 ± 72 minutes (range, 50-360 minutes).
Parathyroid gland identification was rarely exhaustive (Table 1), and the inferior parathyroid glands were identified and salvaged in 17 patients. Four patients had undergone parathyroid gland autotransplantation.
A thoracic approach to thyroidectomy was undertaken in 13 patients, with a median sternotomy required in 12 patients and a right-sided thoracotomy in 1 patient. Median sternotomy was preplanned in 6 cases, it was intraoperatively decided on in 3 cases, and this information was unavailable in the remaining 3 cases. Morcellation was performed in 2 cases using a cervical approach. Tracheostomy was performed in 3 patients, during thyroidectomy in 2 and during reoperation in 1. Tracheostomy was decided on during the initial operation as a safety measure for suspected tracheomalacia in 1 patient and as a precaution after stretching the recurrent laryngeal nerve in the patient with contralateral vocal cord palsy undergoing reoperation. Nine of 14 patients with thyroid glands weighing at least 260 g required sternotomy vs 3 of 14 patients with thyroid glands weighing less than 260 g (P = .02).
Transient hypoparathyroidism developed in 13 patients. Parathyroid function was assessed by the need for calcium or cholecalciferol treatment during the postoperative period and, in some cases, by the parathyroid hormone levels. The external branch of the superior laryngeal nerve was identified in 6 patients, none bilaterally. The recurrent laryngeal nerve was not identified in 4 patients on the left side and in 1 patient on the right side. Five patients had unilateral vocal cord dysfunction, which was permanent in 1 patient; all patients had both recurrent laryngeal nerves identified at surgery. Postoperative vocal cord examination was not conducted in all cases. There were no significant differences between the percentages of patients who underwent or did not undergo sternotomy or thoracotomy regarding vocal cord dysfunction (3 [15%] of 13 patients vs 3 [13%] of 22 patients, P = .30) or hypoparathyroidism (5 [38%] of 13 patients vs 6 [28%] of 22, P = .40).
Four patients underwent reoperation. The indications were hemorrhage less than 4 hours after thyroidectomy in 2 patients, tracheomalacia with asphyxia in 1 patient, and sternotomy wound dehiscence in 1 patient.
Two patients died. A 76-year-old woman with mild aortic valve stenosis underwent reoperation for sternotomy wound dehiscence. After admission to the intensive care unit, she developed a Staphylococcus aureus catheter-related bloodstream infection and died of endocarditis. An 83-year-old woman died of bilateral pneumonia resulting from a tracheobronchial fistula following tracheomalacia that caused an acute upper airway obstruction, which necessitated an emergency tracheostomy.
On stepwise multiple logistic regression analysis, only the weight of the thyroid gland predicted the need for sternotomy. Overall, 18 (52%) of 35 patients were discharged without any complication (Table 2). The mean postoperative hospital stay was 10 days (range, 2-84 days).
When an enlarged thyroid gland grows downward to the thorax, surgeons consider it substernal, intrathoracic, or mediastinal. This categorization is open to several interpretations, as the plane of the superior thoracic inlet is not horizontal but rather is frontally tiled. Extension of the neck pulls the thyroid gland out of the thorax, with resultant confusion in considering the clavicles instead of the first rib as the inferior limit of the neck. Some authors define ITGs as goiters with more than 50% of their mass below the plane of the first ribs.4
Thyroidectomy for ITG usually includes goiters with minimal thoracic involvement and those with massive mediastinal occupation. Using this definition of ITG, the rate of sternotomy is less than 5%, the complication rates are acceptable, and the mortality is less than 1%.5,6 These data do not reflect the seriousness of ITGs, as they do not emphasize the difficult ITGs, which comprise most of the surgical and postoperative problems.
To define a potential high-risk group of patients with ITGs, we evaluated a subset of patients with ITGs with the deepest thoracic extension and selected the carina tracheae as an unambiguous anatomical landmark that is easily identifiable on CT or magnetic resonance imaging. Our objective was to delineate the more easily removable goiters with limited intrathoracic extension from those dangerously protruding past the aortic arch to the tracheal division. This landmark includes a subgroup of ITGs that challenge the surgeon’s skills.
The present series is the largest study of the thoracic approach to thyroidectomy published to date. On multiple logistic regression analysis, the weight of the thyroid gland predicted to a certain extent the need for sternotomy. Nine of 14 patients with thyroid glands weighing at least 260 g required sternotomy. Although this is a postoperative finding and may seem inappropriate to guide the surgical approach, preoperative thyroid gland volume estimation by CT may be used to predict which patients are likely to require a sternotomy.
The rates of postoperative complications and mortality in the present series are higher than previously reported. Transient hypoparathyroidism developed in 13 of the patients. The number of parathyroid glands identified during the thyroidectomies was less than that expected in a series of 32 (91%) total thyroidectomies (Table 1), and the inferior parathyroid glands were identified and salvaged in only 18 of 35 patients. There was a high prevalence of vocal cord dysfunction, compared with other series of ITGs.7- 10 The likely explanation was our inclusion criterion of the deepest reaching glands, which selected patients at higher risk of complications. It appears that deeper glands are associated with greater risk of stretching the recurrent laryngeal nerve.
There were no significant differences between the percentages of patients who underwent or did not undergo sternotomy or thoracotomy regarding the complications of vocal cord dysfunction or hypoparathyroidism. This finding and the retrospective nature of the study preclude any recommendation to perform sternotomy to avoid vocal cord paralysis or transient hypoparathyroidism. Tracheostomy was performed in 3 patients, for different reasons. Two patients died of postoperative complications (sternotomy wound dehiscence and tracheobronchial fistula). In summary, ITGs reaching the carina tracheae, especially those weighing at least 260 g, are difficult to approach, often require a thoracic procedure, and carry significant surgical morbidity and mortality.
Correspondence: Antonio Sitges-Serra, MD, Unitat de Cirurgia Endocrina, Hospital del Mar, Passeig Marítim 25-29, Barcelona, E-08003 Spain (firstname.lastname@example.org).
Accepted for Publication: April 19, 2005.
Previous Presentation: This study was presented at the First International Congress of the European Society of Endocrine Surgeons; May 13, 2004; Pisa, Italy.