Introduction

Hyperthyroidism is characterized by an accumulation of thyroid hormones in the tissues, which is a pathological syndrome. Genetic and environmental factors lead to Graves’ hyperthyroidism, which is an organ-specific autoimmune disease characterized by abnormally high thyroid hormone levels1,2. Treatment for Graves’ disease involves three types of therapy that have been used for decades: antithyroid drugs, 131I therapy, and thyroidectomy3,4. Different countries have different preferences for hyperthyroid treatment. Radioiodine is preferred over antithyroid drugs in the US, whereas it is only favored in Europe and the UK5. There are many reasons for this difference, such as social, racial, and cultural factors, and different attitudes towards the complications of treatment. ATA guidelines recommend providing sufficient radioiodine to induce hypothyroidism for therapy6. In China, it is generally used for relapsed or persistent hyperthyroidism as a second-line therapy. China’s national guidelines for treating Graves’ disease recommend that doctors adhere to the principle of individualized treatment when selecting a regimen7. Therefore, the predictive factors influencing hypothyroidism following radioactive iodine treatment of GD are of concern to Chinese clinicians.

Hypothyroidism is the most significant complication of 131I therapy. Some studies claimed early hypothyroidism have higher incidence than late hypothyroidism, and others believe early hypothyroidism after 131I therapy finally develop permanent hypothyroidism8,9. Some risk factors, such as female sex, high 24-hour 131I uptake rate10, thyroid weight11, antithyroid antibodies12,13, etiology of hyperthyroidism14, use of antithyroid drugs15,16,17, and goiter size18,19 have been reported to influence the risk of hypothyroidism. Many Chinese clinicians believe hypothyroidism may be aggravated if timely substitution treatment is not administered. To maintain euthyroidism and improve quality of life, most hypothyroid patients should receive seasonable thyroxine replacement, frequent follow-up, and dose adjustments after 131I therapy20. Identifying predictive factors for the incidence of hypothyroidism and finding a way to decrease the incidence of hypothyroidism are important for Chinese patients and clinicians. In the present study, we explored the clinical factors that may affect hypothyroidism following 131I treatment in the treatment of hyperthyroidism in Graves’ disease in our center. Aim of this study was to identify any predictive factors for hypothyroidism and find the appropriate time for timely replacement therapy for hypothyroidism so that patients’ quality of life can be improved. Considering these factors and hypothyroidism, we further explored iodine treatment options for hyperthyroid patients.

Patients and methods

Patients

A total of 1,264 GD patients were enrolled with 131I in our center from 2010 to 2016. Out of 1,264 patients, 876 were female and 388 were male, with average ages of 42.31 ± 11.89 years and 42.54 ± 12.74 years, respectively.

Data

Before 131I therapy, the patients underwent routine examinations, including symptoms of GD, effective 131I half-life in the thyroid gland (Teff), triiodothyronine (T3), thyroxine (T4), free triiodothyronine (FT3), thyroid-stimulating hormone (TSH), free thyroxine (FT4), iodine uptake tests at 24-h (T24) and thyroglobulin antibody (TGAb), thyroid microsome antibody (TMAb), thyrotropin receptor antibody (TRAb). The therapeutic 131I dose was calculated according to the Marinelli–Quimby formula21

$$\:\text{A}=\frac{\text{R}\text{a}\text{d}/\text{g}\times\:\text{G}\times\:8}{{\text{T}}_{\text{e}\text{f}\text{f}}\times\:{\text{T}}_{24}\times\:120}$$

(A, 131I dose; G,thyroid weight; Rad/g, iodine dose per gram of thyroid tissue; Teff, effective 131I half-life in the thyroid gland; T24,24-h iodine uptake; and W, thyroid weight). The thyroid gland’s weight was estimated using palpation, thyroid scanning, and thyroid B-mode ultrasound by three or more physicians. In our center, based on the calculated dose and the condition of the patient, 131I was given once. After 131I treatment, follow-ups were conducted at 3 months, 6 months, and 1 year. Data were recorded, including age (named X1), duration of Graves’ hyperthyroidism (X2), thyroid weight (X3), Teff (X4), 24-h iodine uptake (X5), total dose of iodine(X6 ), T3(X7), T4 (X8), FT3(X9), FT4(X10), TPO(X11), TGAb(X12), TMAb(X13), TSH (X14), the usage time of ATD treatment before iodine therapy (X15), the number of weeks of antithyroid drugs should be withdrawn before 131I therapy (X16), iodine dose absorbed per gram of thyroid tissue (X17), administration of iodine dose per gram of thyroid tissue (X18), T4/T3 (X19), FT4/FT3 (X20), sex(X21), family history of thyroid disease(X22), iodine intake peak forward(X23), the usage of ATD(X24), the number of times of taking iodine(X25), nodules(X26), thyroid texture(X27), hyperthyroidism heart(X28), periodic paralysis(X29), abnormal liver function(X30), hematological abnormalities(X31), and prior ATD use before 131I therapy for hyperthyroidism (X32). The non-hypothyroidism and hypothyroidism groups were named as Y. Y1 = 1 implies non-hypothyroidism, whereas Y1 = 2 implies hypothyroidism.

Statistical analysis

Statistical analysis was performed using SPSS for Windows (version 23.0; SPSS Inc., Chicago, Illinois, USA). An independent-sample t-test was used to investigate the influence of the measurement data (X1 –X20). The χ2 test was used to investigate the influence of the count data (X21–X32). Moreover, logistic regression analysis was used to evaluate the impact of particular parameters on the occurrence of hypothyroidism after 131I treatment (non-hypothyroidism group and hypothyroidism group). Logistic regression analysis considered those parameters that were statistically significant for the outcome (P < 0.05). All P values were two-tailed.

Result

In this study, 1,264 Graves’ hyperthyroidism patients, 388 males and 876 females, were treated with 131I therapy.

Analysis of influential factors of hypothyroidism after treated by 131I therapy 3 months later

GD patients are described in this study (Table 1), including clinical and physiological characteristics, and the results of 131I treatment (non-hypothyroidism, hypothyroidism group). An independent-samples t-test compared the measurement data (X1–X20) between the euthyroidism and persistent hyperthyroidism group (group 1) and the hypothyroidism group (group 2). The χ2-test compared count data (X21–X32) between the two groups. Table 1 shows 131I therapy outcomes were influenced by X2-13, X15, X18, X20, X21, X23, X24, X25, X27, X28, X32(P < 0.05). The results (Table 2) show that X3(g), X12 (%), X15(months), and X32(%) were more likely associated with hypothyroidism. The regression equation is Y=-0.014 × 3 + 0.021 × 12-0.006 × 15 + 0.587 × 32. This equation was tested using the likelihood ratio χ2 = 123.87, P < 0.05. The first three months after 131I therapy, hypothyroidism risk was higher among patients with lighter thyroid weight, higher levels of TGAb, and shorter durations of ATD treatment before 131I therapy (P < 0.05).

Table 1 Characteristics of patients under study with respect to clinical and physical parameters and 131I therapy outcomes: non-hypothyroidism, hypothyroidism within 3 months.
Full size table
Table 2 Variables and constants of the regression equation: the chance of hypothyroidism within 3 months after 131I therapy.
Full size table

Analysis of influential factors of hypothyroidism after treated by 131I therapy 6 months later

GD patients are described in this study (Table 3), including clinical and physiological characteristics, and the results of 131I treatment (non-hypothyroidism, hypothyroidism group). An independent-samples t-test compared the measurement data (X1–X20) between the euthyroidism and persistent hyperthyroidism group (group 1) and the hypothyroidism group (group 2). The χ2-test compared count data (X21–X32) between the two groups. Table 3 shows 131I therapy outcomes were influenced by X2-13, X15, X18, X20, X21, X23, X24, X25, X27, X28, X32(P < 0.05). The results (Table 3) show that 131I therapy outcomes were influenced by X2, X3, X5, X6, X15, X17, X18, X23, X25, X32(P < 0.05).

Table 3 Characteristics of patients under study with respect to clinical and physical parameters and131I therapy outcomes: non-hypothyroidism, hypothyroidism within 6 months.
Full size table

The results (Table 4) showed that X3 (g), X15(months), and X25(%) were more possible associated with hypothyroidism. The regression equation was Y=-0.014 × 3-0.006 × 15-0.775 × 25. This equation was tested using the likelihood ratio: χ2 = 40.35, P < 0.05.

Table 4 Variables and constants of the regression equation: the chance of hypothyroidism within 6 months after 131I therapy.
Full size table

This equation shows that there was an increased risk of hypothyroidism in patients with lighter thyroid, shorter duration of ATD treatment before 131I therapy, and increased iodine intake after 6 months.

Analysis of influential factors of hypothyroidism after treated by 131I therapy 1 year later

GD patients are described in this study (Table 5), including clinical and physiological characteristics, and the results of 131I treatment (non-hypothyroidism, hypothyroidism group). An independent-samples t-test compared the measurement data (X1–X20) between the euthyroidism and persistent hyperthyroidism group (group 1) and the hypothyroidism group (group 2). The χ2-test compared count data (X21–X32) between the two groups. Table 5 shows that 131I therapy outcomes were influenced by X3, X5, X8, X10 (P < 0.05).

Table 5 Characteristics of patients under study with respect to clinical and physical parameters and 131I therapy outcomes: non-hypothyroidism, hypothyroidism within 1 year.
Full size table

The results (Table 6) show that X5(%) were more possible associated with the occurrence of hypothyroidism. The regression equation was Y=-0.027 × 5. This equation was tested using the likelihood ratio χ2 = 11.69, P < 0.05. Thus, there was an increased risk of hypothyroidism in patients after one year with lower 24-h iodine uptake.

Table 6 Variables and constants of the regression equation: the chance of hypothyroidism within 1 year after 131I therapy.
Full size table

Discussion

In our study, we used logistic regression and found in the first three months, patients with lower thyroid weight were more likely to develop early hypothyroidism, and logistic regression analysis showed that a higher TGAb level was a contributing factor towards the development of hypothyroidism (odds ratio = 1.022, 95% confidence interval:1.014–1.029, P < 0.05). The thyroid weight is more likely to be overestimated when using the individual dose method, resulting in a dose larger than necessary and hypothyroidism. However, an different analysis by Wang found a correlation between hyperthyroidism and the course of GD, the thyroid uptake ratio of 131I, the effective half-life time, and the level of TMAb20. Wilson and James reported that early changes observed in serum T3, T4, TSH, thyroid microsomal, and thyroglobulin antibody levels were not predictive of hypothyroidism risk22. Moreover, in patients with higher TGAb levels, the cause of hyperthyroidism was more likely to be hyperthyroidism combined with Hashimoto’s thyroiditis. Strigari and Sciuto claimed that higher TGAb levels could lead to a higher risk of hypothyroidism23. Meanwhile, we found that a shorter time to take antithyroid drugs before trying iodine increased the risk of early hypothyroidism (odds ratio = 1.798, 95% confidence interval: 1.313–2.461, P < 0.05). Shi GM claimed that pre-treatment with PTU, or sequential PTU and MMI pret-reatment, reduced the rate of hypothyroidism24. This difference may be because the pre-treatment of ATD was different, resulting in varied outcomes.

Compared with 3 months after 131I therapy, we found that a lighter thyroid and a shorter duration of ATD treatment before 131I therapy could increase the risk of hypothyroidism 6 months after 131I therapy. This may be because ATD increases the iodine radioactive tolerance of the thyroid. Walter M, Metso S and Jaatinen P claim that ATD preceding RAI(radioactive iodine, RAI) therapy decreased the risk of hypothyroidism in patients with Graves’ disease16,25. In contrast, Abd and El-Sayed found that there was no correlation between the development of hypothyroidism and factors such as age, sex, and previous antithyroid medications17. In contrast, Hu and Liu found that Male gender, smaller thyroid weight and lower 6-h RAIU are the main risk factors for early hypothyroidism26. Moreover, we found that more treatments decreased the risk of hypothyroidism. In China, hypothyroidism is very unacceptable in Chinese patients, therefore, they always ask doctors to provide fractional treatment to reduce the probability of hypothyroidism. In our clinical practice, fractional 131I treatment has been proven to effectively reduce the incidence of hypothyroidism in the first six months.

When the time point was 1 year after iodine treatment, there was only one factor that could affect the risk of hypothyroidism, which was 24-h iodine uptake. In our study, patients with a lower 24-h iodine uptake had a higher risk of early hypothyroidism. Possibly, the final occurrence of hypothyroidism was related to the initial body indices before 131I treatment. Sankar et al. claimed that hypothyroidism eventually occurs in nearly all patients treated with radioiodine27. Metso and Jaatinen also claimed that hypothyroidism develops in 82% of the patients 25 years after treated16. John E and Karounwi O believe that hypothyroidism is a predictable sequela of RAI therapy28. As a result of the lower 24-hour iodine uptake, our study found a higher risk of early hypothyroidism. This might be because the calculated dose increases with a decrease in 24-hour iodine uptake, and the absorption rate varies during the therapy29,30. This means that the development of hypothyroidism is only due to patients’ health status before treatment with 131I, such as patient radiosensitivity or genetic specificity. In our previous studies, whether hypothyroidism occurred after iodine treatment was related to two genes, Bc1-2 and Egr-1, which regulate individual radiosensitivity31,32. Therefore, research on individual radio sensitivities may be a point for future studies. Which may finally do some help on the treatment methods for Graves’ hyperthyroidism.

Our study had some limitations, including the possibility of selection bias in retrospective, single-institutional studies, and the short follow-up period of some patients. Further studies with longer follow-up periods are required to validate these findings.

Conclusion

Our study indicated that with time, whether hypothyroidism occurs or not only depends on the patient’s physical state, such as radiosensitivity before 131I therapy. However, in the first 3 or 6 months, we could still do something to effectively decrease the incidence of hypothyroidism to improve the quality of life of patients. Even after calculating the Marinelli–Quimby formula, different radiation doses should be provided to patients with different conditions. In patients with lighter thyroid, higher TGAb titer, shorter duration of ATD treatment before 131I therapy, and hyperthyroidism without using ATD before 131I therapy, we could decrease the radiation dose or provide fractional 131I treatments. As a result, multi-factor analysis allows for an ideal and individualized approach to treatment.