Changes in thyroid volume and thyroid function in acromegaly after surgery in Chinese population

Background : An increased prevalence of thyroid lesions was observed in acromegaly patients. However, the change of thyroid after remission of acromegaly was not clear in Chinese populations. The aims were to assess the thyroid structure and function changes before and after transsphenoidal pituitary adenoma resection in patients with acromegaly and to investigate the correlation between GH, IGF-1, disease duration and thyroid structure and function. Methods : We retrospectively studied 78 patients with acromegaly who underwent surgery between 2015 January and 2018 January at Peking Union Medical College Hospital. The pituitary hormone: random growth hormone (GH), nadir GH and insulin-like growth factor-1 (IGF-1); the thyroid hormone: thyroid stimulating hormone (TSH), thyroxine (T4), triiodothyronine (T3), free thyroxine (FT4) and free triiodothyronine (FT3); four parameters of thyroid metabolism: thyroid’s secretory capacity (SPINA-GT), the sum activity of peripheral deiodinases (SPINA-GD), standard TSH index (sTSHI) and thyrotrophic thyroid hormone sensitivity index (TTSI); and thyroid ultrasound were assessed at baseline and 1 year after surgery. Results : Thyroid volume was significantly positively related random GH, nadir GH, and disease duration. TSH, sTSHI and TTSI was negatively related with pituitary hormone while IGF-1 showed a significant positive association with FT4 and SPINA-GT. After transsphenoidal resection of pituitary adenoma and over 1 year follow-up, the thyroid volume decreased significantly (p=0.000). T3 (p=0.049) and FT3 (p=0.022) also decreased significantly though within normal ranges. However, no significant changes were found in nodule maximum diameter and sTSHI. Thyroid volume change was positively correlated with GH change and nadir GH change. T3 change as well as SPINA-GD change was positively associated with IGF-1 change. Though no significant difference were observed between controlled patients and those who did not achieved “control” level, control patients had a larger decline in thyroid volume along with a smaller decrease in TSH. Conclusion: Enlarged thyroid volume, prevalent thyroid nodules, suppressive pituitary thyrotrophic function and elevated peripheral thyroid hormones are characteristic in acromegaly. A decrease in GH could have favorable effect on thyroid status on thyroid volume and thyroid hormones, while established thyroid nodules and impaired pituitary thyrotrophic function seemed to change little after surgery. Our study is the largest study implemented in Chinese population which investigated the effect of transsphenoidal resection of GH-secreting pituitary adenoma on thyroid volume and thyroid function. We found enlarged thyroid volume, prevalent thyroid nodules, suppressive pituitary thyrotrophic function and elevated peripheral thyroid secretory capacity and peripheral deiodinases in acromegaly. After transsphenoidal resection of GH secretion pituitary adenoma, a partial reverse in thyroid volume and thyroid hormones were observed, while established thyroid nodules and impaired pituitary thyrotrophic function seemeed to change little after surgery. Therefore, our study indicated that early diagnosis and regular follow-up of thyroid ultrasound and functions are necessary in acromegaly patients even after surgery.


Introduction
Acromegaly is a chronic disease associated with a persistent hyper-secretion of growth hormone (GH) and subsequent elevation of insulin-like growth factor-1 (IGF-1) which is usually caused by pituitary adenoma (1). An increased prevalence of thyroid lesions was observed in acromegaly patients and raised researchers' interest (2,3).
Goiter, a kind of thyroid lesion, was demonstrated as a common co-occurrence of acromegaly though the mechanism underlying it was not completely understood. It seems that IGF-1 could act as a thyroid growth factor and thus could stimulate thyroid growth in acromegaly patients (4,5). Thyroid function in acromegaly was also widely studied.
Though GH was proved to modulate the activity of throxine deiodinase, which could affect thyroid hormone level in acromegaly patients, euthyroidism was seen in most acromegaly patients (6). Recently, Andreas Jostel and Johannes W. Dietrich et al proposed that traditional parameters in thyroid function test like thyroid stimulating hormone (TSH), thyroxine (T4), triiodothyronine (T3), free thyroxine (FT4) and free triiodothyronine (FT3) might influence each other and could not reflect more subtle thyroid dysfunctions. Thus, they provided four mathematical models investigating thyroid function: thyroid's secretory capacity (SPINA-GT), the sum activity of peripheral deiodinases (SPINA-GD) and, as 4 markers of the set point, Jostel's TSH index (JTI) or standard TSH index (sTSHI) for assessment of thyrotrophic pituitary function and the thyrotrophic thyroid hormone sensitivity index (TTSI). Therefore, these four calculated parameters might reflect thyroid function in acromegaly patients though no observations were reported (7,8).
Thyroid volume and nodules changes after treatment of acromegaly were also contradictory. Some researchers found no significant changes in thyroid volume and nodules after neurosurgery, radiotherapy, or medical treatment (6). However, Herrmann et al. observed thyroid volume decrease in both medical control and cured patients (9), and Seyfullah et al. confirmed decrease in thyroid volume and nodules volume after over 6 months somatostatin analog use (10). Data on thyroid function before and after surgery were sparse. No significant variation of thyroid stimulating hormone (TSH) was observed by Cannavo (6). Ferdinand found decreased triiodothyronine (T3), increased rT3, and unchanged thyroxine (T4) after treatment (11)(12)(13)(14).
Considering the close relationship between acromegaly and thyroid, we conducted this study to investigate the thyroid structure and function change in acromegaly patients and their relationship in a large Chinese pituitary center. The aims of our study were: 1) to assess the thyroid structure and function changes before and after transsphenoidal pituitary adenoma resection in patients with acromegaly, 2) to investigate the correlation between GH, IGF-1, disease duration and thyroid structure and function changes.

Materials And Methods
We retrospectively analyzed data from patients who were diagnosed with acromegaly and underwent transsphenoidal pituitary adenoma resection between January 2015 and January 2018 in the Department of Neurosurgery, Peking Union Medical College Hospital (PUMCH). The inclusion criteria were as follows: (1) presented with symptoms for acromegaly; (2) patients who satisfied the diagnostic endocrine standard (fasting GH > 1 5 ng/ml, nadir GH > 0.4 ng/ml after oral administration of 75 g of glucose, and a fasting IGF-1 level higher than the age related reference range)(15); (3) a pituitary adenoma identified by contrast-enhanced magnetic resonance imaging (MRI); (4) underwent transsphenoidal pituitary adenoma neurosurgery in PUMCH; (5) having undergone examinations of thyroid ultrasound and thyroid hormone functional test before surgery.
Exclusion criteria were: having undergone thyroidectomy, radiotherapy, or medical treatment before surgery; having pregnancy within 1 year before surgery; having ever experienced nervous or psychological disease, such as Parkinsonism and Schizophrenia. Figure 1 showed the flow chart of participants in the present study. Seventy-eight patients were included in our study at baseline. 3 patients were diagnosed with thyroid cancer and underwent thyroidectomy or radiotherapy after neurosurgery. All of them were papillary thyroid carcinomas (PTC) confirmed by pathology, and were followed at clinic and showed no recurrence during observable period. 6 patients lost follow-up. The rest completed follow-up in our hospital 1 year after surgery including examinations of pituitary hormone test, thyroid hormone functional test and thyroid ultrasound examinations. Twenty-eight patients achieved "control" level while the rest 41 did not according to the 2014 guideline (15). Of the patients who did not meet the "control" criteria, 2 patients had residual tumors after surgery and second surgery was recommended for them, while the GH level of 3 patients remained higher than 5ng/ml (13.1, 16.6 and 33.4) and medical therapy were recommended for them. Observation were recommended for the rest 36 patients who did not achieve "control" but reached a GH level lower than 5ng/ml at 1 year after surgery.

Statistical analysis
Categorical variables were presented as a number (percentage). Quantitative data were presented as the mean (±standard deviation) or median (±standard deviation). Normality was tested using the Kolmogorov-Smirnov and ShapiroWilk W test. Comparisons between categorical variables were performed using the chi-square test. Comparisons between numerical variables were performed using the independent sample t-test and Mann-Whitney test. The paired-samples t-test or Wilcoxon's signed-rank test was used to compare the differences between two measurements (beginning and end). Correlation between variables were conducted using Pearson or Spearman test. Analyses were  At baseline, we found positive relationship between thyroid volume and random GH (r = 0.277, P = 0.032), nadir GH (r = 0.383, P = 0.003), and disease duration (r = 0.283, P = 0.027). With regard to thyroid nodules, no significant relationship was found between pituitary hormone and nodule diameters. Considering relationship between pituitary hormone and thyroid function, we found negative relationship between TSH and random GH (r = -0.230, P = 0.049) and nadir GH (r = -0.307, P = 0.008). sTSHI was also negatively associated with nadir GH (r = -0.252, P = 0.031), while TTSI was negatively However, no significant changes were found in total nodule maximum diameter, single nodule maximum diameter, or multiple nodule maximum diameters. Nodule maximum diameter and single nodule maximum diameter even increased though not reach significance. In addition, at 1 year follow-up, T3 decreased significantly from 1.13±0.27 at baseline to 1.06±0.19 at follow-up (p = 0.049), and FT3 declined significantly from 3.12±0.57 to 2.95±0.38 (p = 0.022). The change of TSH, FT4, T4, SPINA-GT, SPINA-GD and TTSI showed a declined trend but were not significant while sTSHI was almost unchanged (p = 0.772) ( Table 2). Comparing controlled patients with those who did not achieved "control", controlled patients had a larger decline in thyroid volume along with a smaller decrease in TSH (Table 3).

Discussion
This study retrospectively investigated the thyroid structure and function changes in acromegaly patients in a large pituitary center in northern China. Our study showed that reduction in GH after surgical management for acromegaly patients could influence the size of enlarged thyroid and changed thyroid function.
Thyroid disease, especially goiter, is a frequent complication of acromegaly. Previous research showed that goiter developed in 20%-90% acromegaly patients, usually presenting as thyroid enlargement and thyroid nodularity (16,17). However, the correlation of thyroid volume with disease duration and pituitary hormone is unclear.
Gasperi et al. (16) observed thyroid volume had a positive correlation with disease duration, but not with GH and IGF-1. While Miyakawa et al. found an analogous association between higher GH, IGF-1 and thyroid volume (4). In this study, we reported thyroid abnormalities in 93.6% of acromegaly patients and found a positive correlation between thyroid volume and disease duration, serum random GH and nadir GH levels. Moreover, we found a significant decrease in thyroid volume after transsphenoidal resection of GH-secreting pituitary adenoma, and proved a positive correlation between thyroid volume reduction and random GH, nadir GH decrease, which was consistent with previous researches (9,18). The results of our study support the theory that GH and IGF-1 could stimulate DNA synthesis and activate antiapoptotic signaling pathway in thyroid cells (19)(20)(21). The decline of thyroid volume after surgery indicated that thyromegaly might be reversible with disease control. Our study excluded those having ever received treatment with octreotide during observable period, thus proved the effect of GH, IGF-1 decrease on thyroid volume without the interference that octreotide could itself inhibit cell proliferation through SSTR2 and 5 expressing on thyroid cells (22)(23)(24)(25). We found no correlation between thyroid volume change and IGF-1change, which may be explained by the slow descent rate of IGF-1 after surgery. We also found that the decrease in thyroid volume was more significant in patients with longer duration, higher GH, higher nadir GH, and higher IGF-1 before treatment, which provided a value in predicting the change of thyroid volume.
Thyroid nodule change after treatment in acromegaly patients was also controversial. To our knowledge, our study was the first to investigate parameters like SPINA-GT, SPINA-GD, sTSHI and TTSI in acromegaly patients. Traditional thyroid function tests were considered limited in evaluating thyroid function individually due to the interference of TSH and thyroid hormones (27). Thus, parameters like SPINA-GT, SPINA-GD, sTSHI and TTSI were proposed and considered stable and more reliable than thyroid function test parameters. sTSHI was demonstrated a better estimate of true pituitary thyrotrophic function which adjusting for the negative feedback inhibition of TSH by peripheral FT4 concentrations (7) and TTSI was demonstrated to be a valuable marker for estimating thyrotrophic function. Also, reliability of SPINA-GT and SPINA-GD was demonstrated higher than that of measured hormone concentrations (8). Therefore, our study was to investigate these parameters as well as traditional hormones in acromegaly patients.
Our study observed a negative relationship between nadir GH, IGF-1 levels and TSH. This negative correlation might be due to the inhibited leptin observed in active acromegaly considering the stimulating effect of leptin on TSH (28-30). The somatostatin secretion accompanied with excess GH might also suppress TSH secretion(31). In addition, the negative relationship between sTSHI, TTSI and pituitary hormones indicated an impaired pituitary thyrotrophic function in acromegaly patients more specifically. After transsphenoidal resection of pituitary adenoma, though TSH showed a declined trend, sTSHI was almost unchanged, indicating that the impaired pituitary thyrotrophic function might not recover from the decrease of GH. Thus, thyroid function should be followed-up in acromegaly patients even after remission of the disease to avoid hypothyroidism.
Though negative correlation was observed between GH or IGF-1 and TSH, an expected negative correlation between these pituitary hormones and thyroid hormones was not observed. What's more, higher FT4 and SPINA-GT were related with higher IGF-1 and IGF-1 change was significantly positively associated with T3 change and SPINA-GD change. After surgery, decreases in T3, FT3, T4, FT4 SPINA-GT and SPINA-GD were found after surgery though some were not significant. A kind of modulation of thyroid secretion independent of TSH was proposed to explain this discrepancy. Gotzsche et al. observed the stimulation of thyroxine deiodinase by GH, which could lead to T3 increase (32, 33). Yoshinari et al. Though no significant difference were observed between controlled patients and those who did not achieved "control" level, control patients had a larger decline in thyroid volume, with a smaller decrease in TSH. This indicated the effect of surgery on thyroid volume and TSH, and a larger sample might be implemented to compare the difference in patients with or without remission. This study has some limitations. First, it is a retrospective study and thus a standard follow-up was failed to be established. Second, the thyroid function was estimated by regular hormone test. However, we used SPINA-GT, SPINA-GD, sTSHI and TTSI to evaluate the thyroid function. More sensitive methods such as radionuclide imaging are needed to assess thyroid function in the future. 13

Conclusion
Our study is the largest study implemented in Chinese population which investigated the effect of transsphenoidal resection of GH-secreting pituitary adenoma on thyroid volume and thyroid function. We found enlarged thyroid volume, prevalent thyroid nodules, suppressive pituitary thyrotrophic function and elevated peripheral thyroid secretory capacity and peripheral deiodinases in acromegaly. After transsphenoidal resection of GH secretion pituitary adenoma, a partial reverse in thyroid volume and thyroid hormones were observed, while established thyroid nodules and impaired pituitary thyrotrophic function seemeed to change little after surgery. Therefore, our study indicated that early

Consent for publication
Informed consent was obtained from all individual participants included in the study.
Availability of data and materials 14 The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.