Levothyroxine (LT4) suppression treatment for benign thyroid nodules alters coagulation

A machine learning-based approach to predicting the malignant and metastasis of thyroid cancer

Background

Thyroid Cancer (TC) is the most common malignant disease of endocrine system, and its incidence rate is increasing year by year. Early diagnosis, management of malignant nodules and scientific treatment are crucial for TC prognosis. The first aim is the construction of a classification model for TC based on risk factors. The second aim is the construction of a prediction model for metastasis based on risk factors.

Methods

We retrospectively collected approximately 70 preoperative demographic and laboratory test indices from 1735 TC patients. Machine learning pipelines including linear regression model ridge, Logistic Regression (LR) and eXtreme Gradient Boosting (XGBoost) were used to select the best model for predicting deterioration and metastasis of TC. A comprehensive comparative analysis with the prediction model using only thyroid imaging reporting and data system (TI-RADS).

Results

The XGBoost model achieved the best performance in the final thyroid nodule diagnosis (AUC: 0.84) and metastasis (AUC: 0.72-0.77) predictions. Its AUCs for predicting Grade 4 TC deterioration and metastasis reached 0.84 and 0.97, respectively, while none of the AUCs for Only TI-RADS reached 0.70. Based on multivariate analysis and feature selection, age, obesity, prothrombin time, fibrinogen, and HBeAb were common significant risk factors for tumor progression and metastasis. Monocyte, D-dimer, T3, FT3, and albumin were common protective factors. Tumor size (11.14 ± 7.14 mm) is the most important indicator of metastasis formation. In addition, GGT, glucose, platelet volume distribution width, and neutrophil percentage also contributed to the development of metastases. The abnormal levels of blood lipid and uric acid were closely related to the deterioration of tumor. The dual role of mean erythrocytic hemoglobin concentration in TC needs to be verified in a larger patient cohort. We have established a free online tool (http://www.cancer-thyroid.com/) that is available to all clinicians for the prognosis of patients at high risk of TC.

Conclusion

It is feasible to use XGBoost algorithm, combined with preoperative laboratory test indexes and demographic characteristics to predict tumor progression and metastasis in patients with TC, and its performance is better than that of Only using TI-RADS. The web tools we developed can help physicians with less clinical experience to choose the appropriate clinical decision or secondary confirmation of diagnosis results.

Differences reported in the lifespan and aging of male Wistar rats maintained on diets containing fat with different fatty acid profiles (virgin olive, sunflower or fish oils) are not reflected by histopathological lesions found at death in central nervous and endocrine systems

The present study was designed to examine if dietary fat sources that have shown differences in lifespan and if some aging-related aspects can modulate the range of histopathologic changes in central nervous and endocrine systems that occur during the lifespan of Wistar rats. Moreover, it was attempted to gain insight into the relationship between longevity and the development of the different pathological changes, as well as possible interaction with diet. In order to achieve this, male Wistar rats were randomly assigned to three experimental groups fed semisynthetic and isoenergetic diets from weaning until death with different dietary fat sources, namely virgin olive, sunflower, or fish oil. An individual follow-up until death of each animal was performed. Incidence, severity, and burden of specific or group (i.e., neoplastic or non-neoplastic proliferative and non-proliferative) of lesions was calculated along with individual's disease and individual organ lesion burden. Most of the histopathological lesions found have been described in previous studies. Neoplasms, and in particular pituitary adenomas followed by brain tumors, were the most prevalent lesions found in the rats and the main cause of death involving both systems. Incidence of brain lesions was associated with age-at-death. Assayed dietary fats did not present differential effects on pathological changes occurring in endocrine and central nervous systems throughout rat lifespan.

Platelet function in cats with hyperthyroidism

OBJECTIVES

Cats with hyperthyroidism have been reported to develop thromboembolism, with and without echocardiographic abnormalities consistent with hyperthyroidism. The objective of this study was to compare platelet function in cats with hyperthyroidism with euthyroid age-matched cats. We hypothesized that cats with hyperthyroidism have shortened collagen and adenosine diphosphate (C-ADP) closure times as measured with the platelet function analyzer (PFA-100) in comparison with healthy, age-matched controls.

METHODS

Sixteen hyperthyroid and nine euthyroid healthy cats >7 years of age were recruited from the hospital population. Platelet function, measured using the C-ADP closure times by the PFA-100, and platelet count were measured in healthy euthyroid cats and cats with hyperthyroidism.

RESULTS

Mean ± SD closure times were not significantly different between control (66.3 ± 9.6 s) and hyperthyroid cats (65.9 ± 11.5 s; P = 0.75). The mean ± SD closure times of hyperthyroid cats that either were untreated or received methimazole for ⩽3 weeks (n = 6; mean 68.5 ± 15.4 s) was not different than that of cats treated for >3 weeks (n = 10; mean 64.3 ± 8.9 s; P = 0.57). The mean automated platelet count was higher in the hyperthyroid group than in the control group (P = 0.023).

CONCLUSIONS AND RELEVANCE

Platelet function, as measured by closure time under high shear conditions using C-ADP as an agonist, was not affected by hyperthyroidism in this group of cats. Further research is needed to determine if a hypercoagulable state exists in hyperthyroid cats and the potential roles platelets and von Willebrand factor may have.

The influence of thyroid function on the coagulation system and its clinical consequences

Several studies indicate that low plasma levels of thyroid hormone shift the hemostatic system towards a hypocoagulable and hyperfibrinolytic state, whereas high levels of thyroid hormone lead to more coagulation and less fibrinolysis. Low levels of thyroid hormone thereby seem to lead to an increased bleeding risk, whereas high levels, by contrast, increase the risk of venous thromboembolism. Hypothyroidism leads to a higher incidence of acquired von Willebrand's syndrome and with increasing levels of free thyroxine, levels of fibrinogen, factor VIII and von Willebrand factor, amongst others, increase gradually, to the extent that they may lead to symptomatic venous thromboembolism in patients with hyperthyroidism. Here, we discuss the literature on the effect of thyroid hormone on the hemostatic system and the associated risk of bleeding and venous thromboembolism. Patients with hypothyroidism are at increased risk of developing bleeding complications, which could be relevant in patients undergoing invasive procedures. Furthermore, physicians should be aware of the possibility of hyperthyroidism as an underlying risk factor for venous thromboembolism, especially in unexplained cases. Clinical studies are needed to further investigate the significance for general practice of these findings. Besides the effects of hyperthyroidism on venous thromboembolism, its effects on embolism secondary to atrial fibrillation are described.

The effect of hyperthyroidism on procoagulant, anticoagulant and fibrinolytic factors

Several coagulation and fibrinolytic parameters appear to be affected by thyroid hormone excess;however, the net effect on the haemostatic system remains unclear. We aimed to update our previous review and systematically summarise and meta-analyse the data by assessing the effects of thyrotoxicosis on the coagulation and fibrinolytic system in vivo. Data sources included MEDLINE (2006–2012), EMBASE (2006–2012), and reference lists. The sources were combined with our previous search containing studies from 1980–2006. Eligible studies were all observational or experimental studies. Two investigators independently extracted data and rated study quality. Weighted mean proportion and 95% confidence intervals were calculated and pooled using a fixed and a random-effects model. A total of 29 articles consisting of 51 studies were included, as in several articles more than one study was described. We included four intervention (before and after treatment in hyperthyroid patients), five cross-sectional (hyperthyroid subjects and euthyroid controls), and four experimental (before and after use of thyroid hormone in euthyroid subjects) medium/high quality studies for meta-analysis. We found that thyrotoxicosis shifts the haemostatic balance towards a hypercoagulable and hypofibrinolytic state with a rise in factors VIII and IX, fibrinogen, von Willebrand factor, and plasminogen activator inhibitor-1. This was observed in endogenous and exogenous thyrotoxicosis, and in subclinical as well as overt hyperthyroidism. We conclude that both subclinical and overt hyperthyroidism induce a prothrombotic state, which is therefore likely to be a risk factor for venous thrombosis.

Prothrombotic changes due to an increase in thyroid hormone levels

OBJECTIVE

With increasing free thyroxine levels, a gradually rising risk of venous thromboembolism has been described in case-control studies. However, reports on the influence of thyroid hormones on haemostasis, while suggesting a hypercoagulable state in thyrotoxicosis, have often been inconclusive. This study evaluates multiple markers of haemostasis and fibrinolysis in a paired design, making it more sensitive to changes in thyroid hormone levels.

DESIGN

We analysed multiple variables in patients who shifted from severe hypothyroidism to mild hyperthyroidism during thyroid cancer treatment. Those with possible residual disease were excluded.

METHODS

Ninety patients following total thyroidectomy were tested on two occasions: i) before radioiodine remnant ablation and ii) 6 weeks later, on levothyroxine (lT4) suppression treatment, and the results were compared using the Wilcoxon's test for paired data.

RESULTS

During lT4 treatment, significant increases (all P<0.001) in fibrinogen (from median 3.4 to 3.8 g/l), von Willebrand factor (from 85 to 127%), factor VIII (from 111 to 148%) and plasminogen activator inhibitor 1 (from 6.5 to 13.9 μg/l) were observed. In addition, the activation times of platelet adhesion and aggregation stimulated with collagen and epinephrine (EPI)/ADP, i.e. closure times in platelet function analyser (PFA-100), were significantly shortened (P<0.001): for EPI from median 148 to 117 s and for ADP from 95 to 80 s. Changes in other tests were less prominent or insignificant.

CONCLUSIONS

An increase in thyroid hormone levels shifts the haemostatic balance towards a hypercoagulable, hypofibrinolytic state. This may contribute to the increased cardiovascular morbidity and mortality observed even in mild thyrotoxicosis.

Alterations in coagulation and fibrinolysis after levothyroxine exposure in healthy volunteers: a controlled randomized crossover study

BACKGROUND

Several hemostatic abnormalities have been reported in hyperthyroidism, but the overall effect of thyroid hormone excess on coagulation and fibrinolysis is unclear.

OBJECTIVE

Our aim was to assess whether the use of supraphysiological doses of levothyroxine leads to coagulation activation and inhibition of fibrinolysis.

PATIENTS AND METHODS

Healthy volunteers were randomized to receive levothyroxine or no medication for 14 days with a washout period of at least 28 days in a crossover design. To study the effects of different degrees of thyroid hormone excess, 16 participants received levothyroxine in a dose of 0.3 mg per day, and 12 received levothyroxine 0.45 or 0.6 mg per day depending on body weight. Several variables of coagulation and fibrinolysis were measured.

RESULTS

Levels of von Willebrand factor activity (VWF:RiCo) and antigen (VWF:Ag), factor (F) VIII, plasminogen activator inhibitor-1 (PAI-1) and clot-lysis time were slightly higher after levothyroxine 0.3 mg per day than after the control situation, but only levels of VWF showed a significant increase from baseline. After levothyroxine 0.45 or 0.6 mg per day, levels of fibrinogen increased by 17%, VWF activity by 24%, VWF antigen by 26%, FVIII by 19%, FIX by 14%, FX by 7%, PAI-1 by 116% and clot-lysis time by 14%, and activated partial thromboplastin time decreased by 3%; all were significant changes compared with the control situation. We did not observe clear evidence of coagulation activation.

CONCLUSIONS

Our data suggest that thyroid hormone excess increases coagulation factor levels and inhibits fibrinolysis in a dose-dependent fashion. This implies an increased risk of venous thrombosis during hyperthyroidism.

Acute myocardial infarction as the first presentation of thyrotoxicosis in a 31-year old woman - case report

A 31-year old woman, previously fit & well was admitted with pressing retrosternal chest pain and palpitations of sudden onset. Her body weight was normal (BMI 20.5 kg/m²) and there was no significant family history of cardiac disease. She smoked, however, about 15 cigarettes a day and she had been taking combined oral contraceptive pill (Cilest^®) for about three years. On admission she appeared sweaty and in pain, blood pressure 130/70 mmHg, heart rate about 110/min, mild lid-lag sign. Heart sounds were normal and chest was clear. ECG revealed 2-3 mm ST segment elevations in II, III, aVF as well as V2 to V5. Troponin I was raised and she was qualified to an emergency coronary angiography. This revealed a massive spasm of left anterior descending (LAD) coronary artery that responded to intracoronary glyceryl trinitrite administration, however, with the presence of critical narrowing of the LAD apical segment with possible superimposed thrombus. Cardiac ultrasound revealed akinesis of 1/2 of apical area consistent with myocardial infarction

Treatment and progress

She was started on Aspirin, Simvastatin, and Diltiazem, but continued to have persistent tachycardia and tremor. Thyroid function tests were ordered and showed thyrotoxicosis [free T4-46.9 pmol/l (ref. range 9-25), free T3-11.9 pmol/l (2-5), TSH - 0.02 mIU/l (0.27-4.2)]. She was referred for an endocrine opinion and started on Thiamazole. Other investigations revealed elevated anti-TPO and anti-TSH receptor antibodies consistent with Graves' disease. Thrombophilia screen was negative. She had remained euthyroid on a "block & replace" regimen (Thiamazole plus L-Thyroxine) that was discontinued after 18 months. She denies any anginal symptoms, but continues to smoke against medical advice.

Conclusions

Our case highlights the possibility of development of an acute myocardial infarction in a young subject with thyrotoxicosis. We speculate that patient's smoking habit combined with subtle thyrotoxicosis-induced prothrombotic state and/or coronary-artery spasm had lead to the above-mentioned acute coronary event.