Definition, Classification, and Pathogenesis of RG
From an etymological point of view, RG must at least partially be located in the mediastinum, but a myriad of factors affects this seemingly simple definition: a clinical versus a radiological evaluation, the extension of the head at the time of examination, the proportion of thyroid mass that is required to extend into the thoracic cavity, just to name a few. As of today, we still haven’t found a commonly accepted definition of RG which suffers from an “identity crisis”2, and a historical overview of the over ten definitions available has just been published5. Notably, Rios et al. analyzed the various criteria to define RGs and they found none to be clinically meaningful6. The most recent (and somehow broadest) definition comes from the 2020 American Association of Endocrine Surgeons (AAES) guidelines, where “if a mediastinal extension is present, meaning that the gland extends caudally past the sternal notch on physical examination, computer tomography (CT) imaging, or at the time of surgery” it can be considered as a true RG7. Since an authoritative international consensus has not been reached, the most recent literature is still heavily affected by these heterogeneous definitions (seeTable 1 ), and this represents an enormous obstacle for comparing surgical outcomes.
RGs can be variously defined according to their extension and relationships with anatomical structures of the mediastinum: the most recent classification divides them into a Type I RG (extending anteriorly to the trachea and RLN), a Type II RG (where at least a part of the mass extends posteriorly to the mediastinal vessels, the trachea, and the RLN), and a Type III (purely mediastinal) RG8. Unfortunately, a formal correlation of these classifications with surgical outcomes has not been performed. We must remember that over 99% of RGs are secondary (type I and II), i.e. they represent an inferior extension of the normal thyroid gland, and they are vascularized by branches of superior and inferior thyroid arteries. Instead, primary (type III) RGs are clearly separated from the orthotopic cervical thyroid, they receive arterial branches from intrathoracic arteries, and they are usually incidental findings9. In these cases, a missed gland can be diagnosed only when postoperative TSH levels remain unchanged or a (functional) imaging of the chest is performed10–13. From an embryological point of view, type III RG is explained by an erroneous migration of the two median and lateral anlagen from the pharyngeal pouch and the ultimobranchial body, as it was recently reviewed by Rico and Lung14. Migration of the median anlage may also lead to a presternal goiter, and in 2019 the seventh case worldwide was reported15. Sometimes, after subtotal thyroidectomy, patients may present with only a residual central mediastinal mass shifting the trachea laterally (“pseudo-primary RG”)16–18. This may also happen decades after the initial surgery and it may mimic other mediastinal or pulmonary pathologies, such as paraganglioma or adenocarcinoma of the lung19,20.
Regarding the pathogenesis of RG, no recent study has ever explored if there are different molecular mutations in RG compared to non-intrathoracic goiters. The issue remains unsolved and anatomical and functional factors (no limitations downwards encountered by the expanding gland, traction forces associated with swallowing, a negative intrathoracic pressure during inspiration), and gravity itself has been called into question21. A 1939’s study by Crile proposed that patients with a short neck and hypertrophic cervical muscles might theoretically favor the retrosternal expansion of the goiter22, but no formal anthropometric study exists to corroborate this assumption. In the end, and besides the simple time during which a goiter has been allowed to grow undisturbed, risk factors for the development of RGs remain still unexplored.