Most patients with differentiated thyroid carcinoma (DTC) are treated by complete surgical excision (thyroidectomy or hemithyroidectomy) and, depending on the risk of recurrence, may receive postoperative radioiodine(¹³¹I) combined with TSH suppressive therapy both of which have been reported to reduce locoregional recurrence and improve survival in higher risk patients [1].

Locally advanced DTC with clinical extrathyroid extension is an uncommon clinical scenario, although more frequent in older patients, especially men. It can be a significant challenge if it recurs, as uncontrolled cancer in the neck can be devastating with laryngeal, tracheal, and esophageal obstruction, skin ulceration and necrosis, and even carotid artery rupture. Following extensive surgery, the disrupted blood flow to the thyroid bed may result in poor distribution of ¹³¹I to residual tumor foci. Further, patients in this group often have less well-differentiated tumors that may not take up iodine as avidly as those that are more well differentiated, and therefore, surgery and postoperative ¹³¹I may not adequately control persistent cervical disease. Consequently, additional treatment with external beam radiotherapy (EBRT) may be beneficial. However, there is a lack of randomized evidence to guide its use and confirm its benefit. The only randomized trial comparing observation with EBRT was in patients with pT3/4 pN0/1 with DTC, and it closed early due to poor accrual [2]. The study randomized patients without distant metastasis following an R0 (no residual disease) or R1 (microscopic residual disease) resection and TSH suppression therapy. The dose of EBRT was 59.4 Gy if an R0 resection had been performed and 66.6 Gy if R1 resection had been performed. The data were subsequently reported as a prospective cohort study which showed EBRT was well tolerated and there were no recurrences; however, there was only a 3 % recurrence rate in the observation group. Many patients did not meet the American Thyroid Association (ATA) guideline criteria [3] on the use of EBRT and had relatively low risk disease; therefore, a benefit from EBRT may not have been expected [4].

With the lack of randomized data, institutional series are used to inform management. In a typical series of high-risk, (pT2-T4, N0/1) DTC patients who received EBRT (median dose 62Gy), the 4-year locoregional control was 72 %. EBRT however can cause significant side effects, and in this series 5 % remained dependent on a feeding tube [5]. The extent of disease following surgery remains an important prognostic factor despite treatment with EBRT. In another report of high-risk patients who received EBRT, the 5-year locoregional control rate for patients with clear or microscopic margins was 89 %, compared to a 69 % rate of control in those with macroscopic residual or unresectable disease [6]. Our own data have shown that EBRT in older patients (age >60) with extrathyroidal extension (ETE) has significantly improved 5-year local relapse-free rate. In addition, in patients with postoperative gross residual disease treated with radiotherapy, the 10-year cause-specific survival and local recurrence-free rate were 48 and 90 %, respectively. This demonstrates clearly that even though local disease can be controlled with EBRT, patients continue to die from metastatic disease [7].

The ATA guidelines recommend consideration of EBRT in patients older than 45 years with gross extrathyroidal extension and high likelihood of microscopic residual disease [3]. EBRT should also be considered in patients with gross residual disease. It is our policy to recommend EBRT in patients over age 50 with gross residual disease after surgery or with extrathyroid extension that invades posteriorly into the tracheal-esophageal groove that is unlikely to be controlled by ¹³¹I, and in whom salvage surgery may require an ablative surgical procedure such as laryngectomy (T4a or T4b). Patients with minimal ETE (T3) with positive margins or ETE that invades anteriorly into the strap muscles can usually be resected with clear margins and do not require EBRT. Similarly as ¹³¹I is more likely to be effective, we rarely recommend EBRT in younger patients unless they have T4b disease.

Minimizing toxicity from EBRT is a critical consideration in DTC. Toxicity is related to both dose and the volume of tissue that is irradiated. The optimal target volume for EBRT is controversial. Larger elective target volumes have the potential to reduce recurrence, but are associated with increased toxicity. The late effects following EBRT to the head and neck region are also dependent on the volumes radiated as well as the dose to normal structures. Structures of particular concern include the parotid glands [8] radiation to which potentiates the risk of xerostomia, which may already be of concern if high activities of ¹³¹I have already been administered; the pharyngeal constrictors [9], which can result in feeding tube dependence; and, although more of a risk in other head and neck cancers, the mandible [10], which can be associated with a risk of osteo-radionecrosis. Our usual radiation volume includes the surgical thyroidectomy bed and nodal levels III, IV, VI, and part of level V, extending from the hyoid bone superiorly to the aortic arch inferiorly. Other centers use larger volumes. Two reports have recently concluded that larger volumes result in fewer recurrences and that larger volumes should be used. In one study, limited field EBRT was compared to extended radiotherapy volumes. They had no out of field recurrences in the larger volume patients in contrast to those with smaller volume irradiation. There was also an improved relapse-free rate of 89 % compared to a perhaps surprisingly low 40 % with the smaller volume [11]. The second report recommended extending the treatment volumes to the level of the carina to avoid upper mediastinal failures [6]. If radiotherapy is extended into the mediastinum, the simulation CT scan should include the whole lungs so the dose to the lungs can be ascertained and the radiation planning technique chosen to ensure a safe radiation dose is given and that risk of pneumonitis is minimized. As we reported 90 % local relapse-free rate in our series [7], we do not think that volumes need to be extended.

Intensity-modulated radiotherapy (IMRT) which allows modulation of the intensity of the radiotherapy beam ensures better distribution of radiation dose to the intended treatment volumes and is especially suitable for treating complex volumes such as the thyroid bed. It results in a more homogenous dose distribution and sparing of normal tissues than other radiation techniques. In randomized studies in squamous cell carcinomas of the head and neck, IMRT reduced side effects and improve quality of life [12]. The MD Anderson group, in a study of high-risk patients given EBRT, also reported that IMRT was associated with less frequent late toxicity in DTC [13]. The local relapse-free rate was high at 79 % at 4 years.

It is our institutional policy to deliver 60Gy in 30 fractions to the thyroid bed and areas of surgical dissection if there is concern for microscopic residual disease and a lower dose of 54Gy in 30 fractions to undissected areas at risk of microscopic disease. In the case of gross residual disease, a dose of 66Gy in 33 fractions is given to the residual disease plus a margin for uncertainty with 56Gy in 33 fractions to the areas at risk of microscopic disease. A preoperative CT scan with contrast is a great aid in planning any postoperative radiation therapy, as well as being helpful to the surgeon in planning the operation. In our institution, surgeons routinely perform CT scans with contrast in any patient with a potential T4 tumor (a large mass, pain, or hoarseness). Previously, concern that the use of iodinated contrast may interfere with the effectiveness of ¹³¹I by reducing uptake has been expressed. However, with modern water-soluble contrast media, only a 1- or 2-month delay is required for the urinary iodine levels to fall, which is not an undue delay. Other centers have a preference for cervical MRI, but identification of lymph nodes and laryngeal cartilage involvement may be inferior when compared to a contrast-enhanced CT scan. Although in theory EBRT could reduce the effectiveness of ¹³¹I, there is no good evidence to support this; however, our preference is usually to give ¹³¹I, then perform postoperative CT scans after the post ¹³¹I therapy scan, and reassess the extent of disease after surgery as seen on all imaging modalities. A PET scan if available may provide additional information. If however there is concern about the extent of local disease that may cause an oncological emergency without control of that disease, such as gross residual disease after spinal cord decompression, then we will give EBRT before ¹³¹I.