Predictive Factors for Thyroid Complications After Radiation Therapy

Data From a Cohort of Cancer Patients Closely Followed Since They Were Irradiated

Vitoria Duarte; Joana Maciel; Daniela Cavaco; Sara Donato; Inês Damásio; Sara Pinheiro; Ana Figueiredo; Ana Ferreira; Joana S. Pereira


Clin Endocrinol. 2022;96(5):728-733. 

In This Article


The general characteristics of patients are summarised in Table 1. HD was the most common type of cancer, followed by CNS tumours. The mean age at diagnosis of cancer was 18.5 ± 15.1 years, and the median follow-up period in our department was 13.5 ± 5.1 years. All patients received radiation therapy, with a mean dose of 35 ± 19 Gy.

Thyroid Dysfunction

Primary hypothyroidism was observed in 160 patients (56.7%). Of these, seven (4.4%) presented positive antibodies. Decreased thyroid function was observed in 19.4% 59.7%, 62.5%, and 65.3% of leukaemia group, HD, HNT, and CNS tumours, respectively. The mean time interval between irradiation and hypothyroidism diagnosis was 6.8 ± 5.9 years. A shorter period of time was found in the HNT group with an average of 5.6 years, whereas the Leukaemia group had a delay of 10 years before the occurrence of hypothyroidism. The mean irradiation dose for those who were diagnosed with hypothyroidism was 40 ± 14.7 Gy. Thyrotoxicosis was not documented in our cohort.

Risk Factors for Hypothyroidism

In the univariate analysis, we found that the radiation field (p < .001) and radiation dose are higher than 35 Gy (p < .01) had a marked impact on the development of hypothyroidism.

The multivariate analysis confirmed that the irradiation field was independently associated with thyroid dysfunction. Compared to TBI, those who received neck and CSI had higher odds of developing hypothyroidism (adjusted odds ratio = 3.5, 95% confidence interval: 1.5–7.9).

Thyroid Nodules and Cancer

A total of 75 patients (26.6%) were found to have thyroid nodules on ultrasonography. Of those, 54 patients underwent cytological examination: 24 had papillary thyroid carcinoma (PTC), 4 had follicular lesions of undetermined significance, and 26 presented a benign nodule. The overall incidence of thyroid malignancy was 8.5%. In leukaemia/NHL group, 20% of patients were found to have PTC, while in HD, HNT, and CNS thyroid cancer incidence rates were 9%, 2%, and 5.5%, respectively (Table 2). The median age at diagnosis of thyroid cancer was 29.8 (IQR = 10.75; 17–62) years. The mean time interval between irradiation and thyroid cancer diagnosis was 20.5 ± 5.5, 16 ± 4.7, 17 ± 0, and 21 ± 2.2 years for leukaemia, HD, HNT, and CNS groups, respectively.

Patients with thyroid cancer were treated with total thyroidectomy, with additional radioactive iodine treatment in 18 cases. All but one patient achieved disease-free status.

In the univariate analysis, three variables had an impact on the incidence of thyroid cancer: gender (p = .027), age at irradiation (p = .006), and irradiation dose (p = .001).

After adjusting for other variables, the prognostic value of all three remained significant (Table 3). The female gender, younger patients, and those who received irradiation doses <35 Gy were three to five times more likely to develop thyroid cancer than their counterparts. No PTC was diagnosed in patients who received >70 Gy.