Abstract
Background/Aim: To evaluate the incidence of radiation-induced hypothyroidism (RIHT) and its association with thyroid dose in patients undergoing volumetric modulated arc therapy (VMAT) for early glottic cancer.
Patients and Methods: We retrospectively reviewed 23 patients with early glottic cancer who received VMAT between 2018 and 2023. All included patients had normal baseline thyroid function tests. RIHT was defined as an increase in thyroid-stimulating hormone levels with or without a decrease in free-T4 or T3 levels. Dose-volume parameters (DVPs) of the thyroid gland, including mean dose and the relative thyroid volume receiving at least 10-60 Gy (V10Gy-V60Gy), were analyzed for correlation with RIHT.
Results: The median follow-up time was 36.4 months, during which all patients survived. The 1-year and 3-year local failure-free survival rates were 91.3% and 82.4%, respectively. The median mean dose was 25.2 Gy. RIHT developed in five patients (21.7%), with a median onset time of 16.7 months after VMAT. Among them, one patient received thyroid hormone therapy. Among the DVPs, V10Gy>70% was significantly associated with a higher risk of RIHT. The 2-year rates of RIHT were 6.2% in patients with V10Gy≤70% and 44.4% in patients with V10Gy>70% (p=0.006). Age and underlying diseases were not associated with RIHT.
Conclusion: A considerable proportion of patients developed RIHT after VMAT for early glottic cancer. The thyroid gland should be recognized as an important organ at risk. For VMAT planning, V10Gy may serve as a useful optimization constraint.
Introduction
The larynx is the most common site of malignancies in the head and neck, with an overall incidence of approximately 2.3 per 100,000 people. Among laryngeal malignancies, glottic cancer is the most prevalent subtype (1, 2). Glottic cancer which has T1-2N0 disease is referred to as early glottic cancer. Treatment options for early glottic cancer are transoral microsurgery and radiotherapy (RT), with ongoing debate regarding their effectiveness in terms of local control and voice preservation (3-5). RT is generally preferred in cases where post-therapy voice preservation is a priority or when the anterior commissure is involved (6).
Early glottic cancer has a low risk of nodal involvement, less than 5%. Therefore, the irradiated field is typically confined between the hyoid bone and the cricoid cartilage. Conventional RT techniques employ two opposed lateral beams with a field size of 5-6 cm. Conventional RT techniques have been widely used due to their simplicity, even in the absence of a computed tomography (CT) simulator (7). However, there are limitations in sparing adjacent organs at risk (OARs), such as the thyroid gland, esophagus, and carotid arteries.
Intensity-modulated RT (IMRT) offers improved OAR sparing while maintaining target coverage compared to conventional RT techniques. Volumetric modulated arc therapy (VMAT) is a novel IMRT technique that utilizes rotational arcs. VMAT allows shorter treatment times compared to other IMRT techniques that use multiple gantry angles (8). Given the high cure rate of early glottic cancer, the use of VMAT is increasing due to growing concerns about patients’ quality of life (9).
Radiation-induced hypothyroidism (RIHT) is one of the most common adverse effects following RT for head and neck cancers, occurring in more than half of the patients (10). In early glottic cancer, the upper portion of the thyroid gland falls within the RT field. However, the thyroid gland is not currently considered an important OAR in these patients. Moreover, there are no clear guidelines regarding thyroid dose constraints for VMAT planning. In this study, we aimed to evaluate the incidence of RIHT and its association with thyroid dose in patients who received VMAT for early glottic cancer.
Patients and Methods
This study was approved by the Institutional Review Board of Hallym University Sacred Heart Hospital, which waived the requirement for informed consent due to the retrospective nature of the study. Between March 2018 and December 2023, 35 patients underwent VMAT for early glottic cancer at the Department of Radiation Oncology of Hallym University Sacred Heart Hospital. In our institution, thyroid functional tests (TFTs) are not routinely recommended for patients with early glottic cancer who are expected to undergo laryngomicrosurgery followed by RT. Therefore, we excluded seven patients who did not have TFT results and three patients who had abnormal baseline TFT results. Additionally, two patients with other malignancies were excluded. Finally, 23 patients were included in this study, and we reviewed their medical records and treatment details.
For TFTs, we collected test dates and serum levels of thyroid-stimulating hormone (TSH), T3, and free-T4 (fT4). RIHT was defined as an increase in TSH levels with or without a decrease in fT4 and T3 levels following the completion of VMAT. The upper normal limit for TSH was 5.0 μIU/ml, while the lower normal limits were 0.93 ng/dL for fT4 and 83.2 ng/dl for T3. The time interval between the last day of VMAT and the date of abnormal TFT results indicating RIHT was calculated.
For CT planning, slices with 3 mm thickness were acquired using either a Brilliance CT Big Bore (Philips Healthcare, Cleveland, OH, USA) or a Somatom Confidence (Siemens Healthiness, Erlangen, Germany). All patients underwent CT simulation in the supine position with a Type S thermoplastic mask (Civco, Kalona, IA, USA). Target volumes were delineated by a single radiation oncologist (TK). Clinical target volumes (CTVs) were defined from the upper border of the thyroid cartilage to the inferior border of the cricoid cartilage. Planning target volumes (PTVs) were generated by expanding the CTVs with a 3-5 mm margin in all directions. To account for the build-up region, PTVs were cropped 2-3 mm below the skin surface. Normal organs, including the thyroid gland, esophagus, and carotid arteries, were contoured.
For VMAT planning, RayStation (RaySearch Laboratories, Stockholm, Sweden) versions 4.7, 8A, 8B, and 9B were used, and the Collapsed Cone algorithm was applied for a final dose calculation. Two arcs with beam angles ranging from 200° to 240° were used with 6 MV photon beams. An optimization constraint was set to limit the mean dose (Dmean) of the thyroid gland to less than 25 Gy. Figure 1 presents the axial, sagittal, and coronal views of the VMAT plan for early glottic cancer. Regarding dose-volume parameters (DVPs), we calculated the Dmean and the relative volume of the thyroid gland receiving at least 10 Gy (V10Gy), 20 Gy (V20Gy), 30 Gy (V30Gy), 40 Gy (V40Gy), 50 Gy (V50Gy), and 60 Gy (V60Gy) of radiation. Since these DVPs are continuous variables, we used the Cox proportional hazards model to identify candidate DVPs associated with RIHT. The maximally selected rank method was applied to categorize the candidate DVPs into two groups. The categorized DVPs were then compared using the log-rank test to assess their association with RIHT.
The axial, sagittal, and coronal views of the volumetric modulated arc therapy planning.
Survival rates were calculated from the end of RT to the date of an event or the last follow-up. The Kaplan-Meier method was used to estimate survival rates and generate survival curves. Statistical significance was defined as a p-value <0.05. All statistical analyses were performed using R 4.3.1 (R Development Core Team, Vienna, Austria).
Results
Patient characteristics and RT outcomes. Details of patient characteristics and RT are described In Table I. The median age of all patients was 62 years (range=41-76 years). All patients received a curative dose of RT following transoral microsurgery for tumor biopsy. The total radiation dose was 65.25 Gy in 29 fractions, with a daily dose of 2.25 Gy. Elective neck irradiation was not performed for any patient, as all had T2 or lower and N0 disease.
Patient and tumor characteristics.
The median follow-up time was 36.4 months (range=7.7-66.1 months) for all patients. No deaths were observed during the follow-up period. Local failure was observed in two patients, occurring at the glottis at 6.6 and 30.0 months after RT. One patient experienced regional failure at the ipsilateral Level III neck at 6.5 months, followed by transglottic local failure at 21.9 months after RT. Another patient had transglottic recurrence along with metastases to the lower neck nodes and mediastinal nodes at 5.8 months after RT.
The 1-year and 3-year local failure-free survival rates were 91.3% and 82.4%, respectively. The 1-year and 3-year locoregional failure-free survival rates were 87.0% and 82.4%, respectively. The 1-year and 3-year progression-free survival rates were 87.0% and 82.4%, respectively. A Grade 3 adverse effect was reported in one patient, who underwent a tracheostomy due to vocal cord palsy 5.0 months after RT.
TFT results and thyroid DVPs. Five patients (21.7%) had abnormal TFT results at 2.6, 3.6, 16.7, 27.6, and 40.4 months after RT. The 1-year, 2-year, and 3-year rates of RIHT were 9.6%, 15.6%, and 23.3%, respectively (Figure 2). One patient received synthetic thyroid hormone therapy. Hypertension (p=0.998), diabetes (p=0.127), and other underlying diseases (p=0.942) were not significantly associated with RIHT. Age also showed no significant correlation with RIHT (p=0.995).
The probability of radiation-induced hypothyroidism in all patients. The 1-year, 2-year, and 3-year rates of radiation-induced hypothyroidism were 9.6%, 15.6%, and 23.3%, respectively.
Analysis of thyroid DVPs revealed that the median Dmean of the thyroid was 25.2 Gy (range=13.2-36.5 Gy) for all patients. Among the continuous DVPs, V10Gy showed a trend toward an association with RIHT (p=0.080; Table II). Then patients were categorized into two groups: V10Gy>70% (N=6) and V10Gy≤70% (N=17). A significant difference in the rate of RIHT was observed between the two groups (p=0.006; Figure 3). The 2-year rates of RIHT were 6.2% in the lower V10Gy group and 44.4% in the higher V10Gy group.
Univariate analysis for thyroid dose-volume parameters.
The probability of radiation-induced hypothyroidism according to V10Gy. The 2-year rates of radiation-induced hypothyroidism were 6.2% in the lower V10Gy group and 44.4% in the higher V10Gy group.
Discussion
IMRT, including VMAT, is widely used for head and neck cancers to spare normal organs. The thyroid gland should be considered an important OAR since it is located among the neck nodes, and elective neck irradiation is a routine practice in head and neck cancer treatment. The incidence of RIHT in patients undergoing IMRT for head and neck cancers is estimated to be as high as 40% (11, 12). This rate remains higher than expected, considering that the incidence of RIHT was reported to be 20-50% before the introduction of IMRT (10, 13).
Although IMRT is an advanced technique, the thyroid gland can only be effectively spared when clear dose constraints are applied. Without appropriate constraints, scattered low doses or peak doses may lead to thyroid dysfunction. Currently, robust dose-volume constraints for the thyroid gland have not been established. Therefore, dose thresholds must be determined based on findings from previous studies on RT planning for head and neck cancers.
In terms of DVPs, previous studies have suggested V45Gy<50%, V50Gy<50%, and Dmean<40 Gy as indicators of RIHT (11, 14). Patients included in these studies received conventional techniques as well as IMRT. Nevertheless, more recent IMRT studies have reported similar results regarding DVPs related to RIHT. In nasopharyngeal carcinoma patients receiving IMRT, V40Gy≤85%, V45Gy<50%, and V50Gy<24-35% were recommended as dose constraints (15-18). Similarly, for patients with oropharyngeal cancer treated with IMRT, Dmean<42Gy and a spared thyroid volume of ≥3 cc at 45 Gy were suggested as optimization objectives (19, 20).
Early glottic cancer is a unique type of laryngeal malignancy, as RT is typically confined to the laryngeal space. The thyroid dose should be lower compared to other head and neck cancers, as only the upper portion of the thyroid gland falls within the RT field. Therefore, adopting thyroid dose constraints based on studies of other head and neck cancers is challenging. However, no previous study has reported DVPs associated with RIHT in patients with early glottic cancer treated with IMRT.
This is the first study to evaluate the relationship between RIHT and thyroid dose in patients with early glottic cancer receiving VMAT. In this study, 21.7% of patients developed RIHT, which is a considerable proportion and comparable to rates observed in other head and neck cancers (10, 13). Therefore, routine TFT follow-up may be necessary for patients with early glottic cancer, similar to patients with head and neck cancer undergoing elective neck irradiation.
Regarding thyroid dose constraints, our findings suggest that V10Gy may be a valuable parameter for VMAT planning. Patients with V10Gy>70% had a significantly higher incidence of RIHT than patients with V10Gy≤70%. Two years after VMAT, 44.4% of patients in the higher V10Gy group were predicted to develop RIHT, compared to only 6.2% in the lower V10Gy group. V10Gy may appear to be a relatively low threshold, as traditionally accepted dose thresholds for RIHT have been around V20Gy (21). However, because the thyroid gland generally receives a lower radiation dose in early glottic cancer compared to other head and neck cancers, more refined dose constraints are needed to improve RIHT prediction. Given the sensitivity of the thyroid gland to even low dose of irradiation, further studies are needed to establish optimal constraints.
A recent study, although not specific to early glottic cancer, has also suggested V10Gy as an appropriate threshold for predicting RIHT. Park et al. (22) analyzed DVPs to predict the risk of RIHT in patients with breast cancer undergoing whole breast RT. Similar to early glottic cancer, only very low radiation doses are delivered to the thyroid gland in patients with breast cancer. The thyroid Dmean was 0.34 Gy in patients without regional nodal irradiation and 4.48 Gy in patients with regional nodal irradiation. Among 1,063 patients, RIHT was observed in 43 (4.0%), with most cases being subclinical (65.1%) rather than clinical (34.9%). The study identified V10Gy as a significant prognostic factor for RIHT, with a cut-off point of 26%. A predictive model incorporating age, systemic therapy, and V10Gy was developed. The 3-year cumulative RIHT rates were 6.4% and 3.1% for the high- and moderate-risk groups, respectively (p=0.008).
Study limitations. First, the sample size was relatively small. As mentioned above, routine TFT follow-up is not mandatory for patients with early glottic cancer. Additionally, we included only patients who received VMAT, as the thyroid gland was not always contoured, and CT simulation was not mandatory for RT planning in conventional techniques. Second, while we identified V10Gy as a potential dose threshold for the thyroid gland, the cut-off value needs validation in a larger cohort. Lastly, concerns regarding locoregional control may arise, given that VMAT can limit the dose to the carotid arteries and create sharp dose fall-off regions (23). In the present study, regional failures were reported in two patients, and the 1-year and 3-year local failure-free survival rates were 91.3% and 82.4%, respectively. These results are promising, considering that the study included only patients with pre- and post-VMAT TFT results. Future studies should focus on validating the optimal V10Gy threshold and comparing survival outcomes between RT techniques.
Conclusion
We analyzed the relationship between thyroid dose and RIHT in patients with early glottic cancer treated with VMAT. RIHT developed in 21.7% of patients, with most cases presenting as subclinical hypothyroidism. Additionally, 4.3% of patients required synthetic thyroid hormone replacement. Regarding thyroid dose constraints, our findings suggest that V10Gy is a valuable parameter for VMAT planning. Patients with V10Gy>70% experienced a significantly higher incidence of RIHT compared to patients with V10Gy≤70%. Therefore, the thyroid gland should be considered an important OAR, with V10Gy serving as an optimization constraint in VMAT planning to minimize the risk of RIHT.
Footnotes
Authors’ Contributions
TK conceptualized and designed this study. HJP and TK collected and analyzed data. MYL, KHC, HKK, and YP participated in the interpretation of results. HJP and TK drafted the article, and TK revised the article. All Authors read and approved the article.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received March 28, 2025.
- Revision received April 15, 2025.
- Accepted April 16, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
