Characteristics and clinical course of thyroid abnormalities arisen in long term survivors of childhood cancer

Medical records were retrospectively reviewed to obtain the demographic and medical characteristics of CCSs. We reviewed the diagnoses and treatment modalities including chemotherapy, radiotherapy, and hematopoietic stem cell transplantation (SCT). The patients who visited National Cancer Center between January 2013 and August 2021 were included in the study. They were diagnosed with cancer at younger than 20 year of age. CCSs who were follow-up for at least 5 years after completion of cancer therapy were enrolled. Patients who received any cancer therapy including chemotherapy, radiotherapy, or targeted therapy such as TKIs and immune check-point inhibitors within 5 years prior to the last follow-up were excluded. The patients with residual tumor after completion of cancer therapy were also ruled out other than 3 patients with inoperable but stationary CNS tumors (2 optic nerve glioma, 1 pilocytic astrocytoma).

The subjects were categorized into 6 subgroups including leukemia or myelodysplastic syndrome (MDS), Hodgkin lymphoma (HD), non-Hodgkin lymphoma (NHL), brain tumor, nasopharyngeal cancer, and other tumors based on the diagnosis. Other tumors were sarcomas, Wilms tumor, neuroblastoma, Langerhans cell histiocytosis, and ovarian tumors.

The subjects were further classified into 3 subgroups according to the type of cancer therapy, which included chemotherapy without irradiation (chemotherapy only), irradiation group 1 and irradiation group 2. The irradiation fields in the irradiation group 1 were direct to thyroid gland or near hypothalamus-pituitary-thyroid (HPT) axis, which were as follows: head and neck, craniospinal irradiation (CSI), total body irradiation (TBI), cervical spine, nasopharynx, and mediastinum. The remaining irradiation fields (abdomen, pelvis, and limbs) were regarded as irradiation group 2. The two irradiation groups received combinations of chemotherapy.

To analyze the additional risk of SCT, the patients were categorized into 4 groups, which were SCT accompanied irradiation near HPT axis or TBI (SCT & irradiation group 1), SCT without irradiation, irradiation group 1 without SCT (no SCT & irradiation group 1), and the remained CCSs (chemo only or irradiation group 2).

Free thyroxine (fT4) and thyroid-stimulating hormone (TSH) were measured at the time of diagnosis, at the end of treatment, and usually, annually thereafter. Neck palpation was also performed annually. Thyroid ultrasound was conducted when abnormal finding was identified in annual surveillance of thyroid or in case of radiotherapy directly or near HPT axis. FT4 and TSH were measured using a chemiluminescent immunoassay (ADVIA Centaur XP, Siemens, Erlangen, Germany). The reference values were 80–200 ng/dL for fT4, and 0.3–4.5 mU/L for TSH. CCSs with results outside the reference value limits were defined as having thyroid dysfunction and followed-up with thyroid hormone testing. Thyroid dysfunction was categorized into subgroups of primary overt hypothyroidism, subclinical hypothyroidism, and central hypothyroidism. Low fT4 and increased TSH were defined as primary overt hypothyroidism, which was divided into two subgroups according to a TSH cutoff level of 10 mU/L. Low fT4 and normal TSHs level were defined as central hypothyroidism. Normal fT4 and elevated TSH were defined as subclinical hypothyroidism, which was also divided into two subgroups based on a TSH level of 10 mU/L.

Thyroxine medication was usually prescribed for fT4 levels below 80 ng/dL or TSH levels exceeding ≥10 mU/L with consideration of the clinical symptoms and patient’s age. Thyroxine medication or abnormal thyroid hormone levels (out of the reference range) at the last follow-up were designated as persistent thyroid dysfunction. If thyroid hormone levels normalized without medication at the last follow-up, the condition was regarded as transient dysfunction. Autoimmune thyroiditis was defined when thyroid dysfunction was present with positive antithyroid peroxidase antibodies (TPOAbs) or anti-thyroglobulin antibodies (TgAbs), which were measured by a radioimmunoassay method (DIAsource Immuno Assay S.A., Louvain-la-Neuve, Belgium). TPOAbs and TgAbs over 60 U/mL were regarded as positive. Laboratory tests of TPOAbs and TgAbs were conducted when abnormal thyroid function tests were detected during follow-up period. The thyroid cancer group included CCSs who underwent total or partial thyroidectomies due to malignant pathologic findings after fine-needle aspiration biopsy regardless of the thyroid hormone levels. Thyroid dysfunction, autoimmune thyroiditis, and thyroid cancer were considered thyroid abnormalities. Thyroid nodules were not included as thyroid abnormalities in this study because ultrasound was performed in only half CCSs.

Their baseline characteristics of the CCSs are described in Table 1. The total number of subjects was 257 (male: 147, female: 110), and the mean age at diagnosis was 10.7 ± 5.3 years. The median age was 11.8 years (0.1–19.8). The median duration from initial diagnosis of cancer to the end of treatment was 0.8 years (0.3–15.1). The median follow-up period after completion of therapy was 9.6 years (5.0–19.5), and the median follow-up time from diagnosis was 11.2 years (5.4–29.5).

Of 257 subjects, transient or persistent thyroid abnormalities were identified in 107 (41.6%). Sixty-five out of 257 (25.3%) had subclinical hypothyroidism, and 16 (6.2%) developed central hypothyroidism. Five CCSs (1.9%) had primary overt hypothyroidism. 5 (1.9%) and 6 (2.3%) CCSs were diagnosed with autoimmune thyroiditis and thyroid cancer, respectively. Of 5 CCSs with autoimmune thyroiditis, 2 patients received SCT with TBI as a conditioning regimen. Of 6 CCSs with thyroid cancer, 4 were exposed to radiation therapy (either nasopharynx, lung, mediastinum or TBI). The initial thyroid hormone test results were not available in 10 CCSs with brain tumors because they were brought to the National Cancer Center with thyroxine medication from the diagnosis at another hospital. Nevertheless, the 10 CCSs were regarded as having persistent thyroid abnormalities due to necessity of thyroxine medication.

Among the different diagnostic groups, all nasopharyngeal cancer (n = 3) patients had thyroid abnormalities. Two with nasopharyngeal cancer had subclinical hypothyroidism, and the other patient had primary overt hypothyroidism, with TSH levels over 10 mU/L in all 3 patients. Thyroid abnormalities were found in 66.7% (6 out of 9) and 57.6% (49 out of 85) of the HD and brain tumor patients, respectively (Table 1). The chi-square test indicated a significantly different thyroid abnormality rate between different diagnoses (Table 1). The irradiation group 1 had more CCSs (58.4%) with thyroid abnormalities compared to the remaining CCSs (28.5%) (P < 0.0001) (Table 1). CCSs with SCT had an increased prevalence of thyroid abnormalities (60.5%) compared to the other CCSs (37.9%) (P = 0.0069) (Table 1).

The time to identify thyroid abnormalities from diagnosis of primary cancer was a median of 3.3 (0–14.8) years (Table 2). The cumulative event rate (thyroid abnormalities) was 21.2, 30.3, and 43.6% at 3 years, 5 years, and 10 years, respectively, after diagnosis of primary cancer. At the end of cancer therapy, 34 of 257 CCSs had developed thyroid abnormalities. Seventy-three CCSs exhibited thyroid abnormality during follow-up period after cancer treatment. The time to detect thyroid abnormalities from the end of treatment of cancer was a median of 1.3 (0–14.4) years in 107 subjects with thyroid abnormalities. Forty-five (42%) of 107 subjects with thyroid abnormalities had transient thyroid abnormalities. Forty (88.9%) of 45 transient thyroid abnormality were subclinical hypothyroidism. At the last follow-up, 62 CCSs of 257 CCSs (24.1%) had thyroid abnormalities. Forty-six of 58 (79.3%) non-brain tumor CCSs had subclinical hypothyroidism, showing transient thyroid abnormalities were more common in non-brain tumor CCSs compared to CCSs with brain tumors. The median follow-up time after completing cancer therapy was 9.6 years (5–16.7) and 9 years (5–16.6) in transient and persistent thyroid abnormalities, respectively. Thirty-five (70%) out of 50 CCSs with subclinical hypothyroidism (4.5 mU/L ≤ TSH < 10 mU/L) showed normalized thyroid hormone tests at the last follow-up. Ten (66.7%) out of 15 with subclinical hypothyroidism (TSH ≥ 10 mU/L) had persistent subclinical hypothyroidism (Table 2). Twenty-one CCSs with primary overt hypothyroidism or central hypothyroidism were categorized into persistent thyroid abnormalities because all 21 have been on thyroxine medication at the last follow-up.

The irradiation group 1 had the highest event rate compared to the chemo only or irradiation group 2 (P < 0.0001) (Fig. 1). CCSs who received TBI and neck/nasopharynx had thyroid abnormalities in 7 of 9 subjects and 8 of 12 subjects, respectively. However, among different radiation fields such as CSI, head, TBI, and neck/cervical/nasopharynx/spinal areas, there was no significant thyroid abnormalities rate (P = 0.0592). Among 65 CCSs with subclinical hypothyroidism, only 2 subjects had TSHs level above 10 mU/L in 24 CCSs who did not received radiotherapy, while 13 of 41 irradiated CCSs had TSH levels over 10 mU/L (P = 0.0309). Twenty-two (53.7%) of 41 irradiated subjects had normalized TSH levels, whereas 18 (75%) of the remained 24 non-irradiated subjects had normalized TSH levels at the last follow-up (Table 2). Of 107 subjects with thyroid abnormalities, transient cases were more prevalent in chemo only group than the other enrolled CCSs (P = 0.0025), and persistent cases were more common in irradiation group 1 than the rest CCSs (P = 0.0004). In the analysis to observe the effect of SCT, there was a statistically significant difference among the 4 different groups (Fig. 2) (P < 0.0001). The highest event rate was found in CCSs treated with both SCT and radiation therapy near HPT axis or TBI. Multiple comparisons using Bonferroni method were shown in suppl table. Two patients with sarcoma had been on tyrosine kinase inhibitor, and 1 patient with brain tumor used to take bevacizumab. Among those 3 subjects, 2 had transient thyroid dysfunction and 1 had persistent thyroid dysfunction.