Citalopram and sertraline exposure compromises embryonic bone development.

Selective serotonin reuptake inhibitors (SSRIs) are the most commonly prescribed treatments for depression and, as a class of drugs, are among the most used medications in the world. Concern regarding possible effects of SSRI treatment on fetal development has arisen recently as studies have suggested a link between maternal SSRI use and an increase in birth defects such as persistent pulmonary hypertension, seizures and craniosynostosis. Furthermore, SSRI exposure in adults is associated with decreased bone mineral density and increased fracture risk, and serotonin receptors are expressed in human osteoblasts and osteoclasts. To determine possible effects of SSRI exposure on developing bone, we treated both zebrafish, during embryonic development, and human mesenchymal stem cells (MSCs), during differentiation into osteoblasts, with the two most prescribed SSRIs, citalopram and sertraline. SSRI treatment in zebrafish decreased bone mineralization, visualized by alizarin red staining and decreased the expression of mature osteoblast-specific markers during embryogenesis. Furthermore, we showed that this inhibition was not associated with increased apoptosis. In differentiating human MSCs, we observed a decrease in osteoblast activity that was associated with a decrease in expression of the osteoblast-specific genes Runx2, Sparc and Spp1, measured with quantitative real-time PCR (qRT-PCR). Similar to the developing zebrafish, no increase in expression of the apoptotic marker Caspase 3 was observed. Therefore, we propose that SSRIs inhibit bone development by affecting osteoblast maturation during embryonic development and MSC differentiation. These results highlight the need to further investigate the risks of SSRI use during pregnancy in exposing unborn babies to potential skeletal abnormalities.

Thyroid function and pregnancy: before, during and beyond.

Thyroid disturbances are common in women during the reproductive years of their lives. Autoimmunity and altered iodine status together account for a high proportion of the abnormalities. Autoimmune thyroid disease is present in around 4% of young females, and up to 15% are at risk because they are thyroid antibody-positive. There is a strong relationship between thyroid immunity on the one hand and infertility, miscarriage, and thyroid disturbances in pregnancy and postpartum on the other hand. Suboptimal iodine status affects a large proportion of the world's population, and pregnancy further depletes iodine stores. There is controversy surrounding the degree to which iodine should be supplemented and the duration of supplementation. Recent studies have helped to clarify the relationship between maternal thyroid status and neuropsychological development of the child. The role of other environmental factors including smoking and selenium status is also now recognised. Universal screening for thyroid hormone abnormalities is not routinely recommended at present. However, measurement of thyroid function and autoantibodies should certainly be considered in those who are at high risk of thyroid disease and in those whose pregnancy is otherwise high risk. The practicing clinician needs to be aware of the thyroid changes which accompany pregnancy.

Q fever cases at a North Queensland centre during 1994-2006.

This study reviewed the epidemiological features, management and outcomes of patients with Q fever treated at a tertiary facility in North Queensland during the period from July 1994 to January 2006. Twenty-seven patients were identified. Our findings were consistent with the observations about Q fever that have been made in other regions of Australia. A diagnosis of Q fever should be considered in patients with a non-specific febrile illness.

Human chorionic gonadotropin may not be responsible for thyroid-stimulating activity in normal pregnancy serum.

Although hCG can activate thyroid cells in culture there is considerable doubt as to whether it is responsible for the changes in thyroid function which occur during normal pregnancy. Thyroid-stimulating activity (TSA), measured using iodide uptake into FRTL-5 cells, was demonstrated in 32/51 (63%) first and 15/29 (52%) third trimester sera. Free T3 was increased (P less than 0.001) and TSH decreased (P less than 0.01) in first trimester. In the early pregnancy group there was a positive correlation between hCG and TSA (r = 0.594, P less than 0.001) and a negative correlation between hCG and TSH (r = -0.329, P less than 0.02). In third trimester hCG concentration was often below that required to produce TSA in vitro and TSA could not be neutralized by antibodies to hCG. There was no correlation between hCG and TSH or thyroid hormones in the third trimester. In 26 women TSA decreased in parallel with serum hCG concentration after termination of pregnancy (P less than 0.001). Free T3 also decreased (P less than 0.01) and TSH increased (P less than 0.05) after termination. TSA persisted in many patients even after hCG was either very low or undetectable. Purified hCG stimulated iodide uptake in a concentration-dependent manner. Stimulation of iodide uptake by TSH was inhibited by the simultaneous presence of low concentrations of hCG while activity was restored with high concentrations. hCG may contribute to the thyroid changes in early pregnancy. However the poor correlation between TSA and thyroid tests suggests that other factors may be involved. The partial agonist activity of hCG may account for some of the inconsistencies observed but TSA in serum from late pregnancy or after termination of pregnancy is almost certainly due to another hormone or growth factor.

Comparison of left anterior oblique, anterior and geometric mean methods for determining gastric emptying times.

The accurate determination of gastric emptying time requires correction or compensation for tissue attenuation. The gold standard for tissue attenuation correction for gastric emptying is the geometric mean of the gastric counts from the anterior and posterior views. For reasons of efficiency, many community hospitals acquire only the anterior projection. This study addressed the hypothesis that, using the left anterior oblique view alone, one can minimize the effect of variation in attenuation as the meal moves from the fundus to the stomach to the more anterior antrum to a degree equal to that of the geometric mean technique. We studied 42 consecutive patients using a standardized 300-g meal labeled with 650 muCi of 99mTc-sulfur colloid. The patients were imaged in the anterior (ANT), posterior (POST) and left anterior oblique (LAO) views every 15 min for 90 min. Linear regressions were obtained using the ANT, LAO and GM data. Cross-correlation of the T1/2 for 35 cases showed an R value for the GM versus LAO of 0.95 and GM versus ANT of 0.84. The p value greater than 0.49, for the paired two-tailed t-test of the LAO and GM methods. The p value for the ANT and GM methods is 0.0058 indicating a significant difference between these methods. The cross-correlation, F-test p and t-test p values support the hypothesis that there is no significant difference between the geometric mean and left anterior oblique gastric emptying times. It is therefore reasonable to substitute the left anterior oblique for routine GET when using a solid meal in patients with normal gastric anatomy, albeit altered physiology.

Release of interleukin-6 by human thyroid epithelial cells immortalized by simian virus 40 DNA transfection.

Interleukin-6 (IL-6) has actions on a variety of endocrine tissues. The cytokine is secreted by cells of the anterior pituitary and endocrine pancreas and has recently been shown to be produced by cultures of thyroid epithelial cells. In this study we have examined some of the factors which regulate IL-6 release from an immortalized human thyroid line (HTori3). IL-6 release over 24 h was stimulated by TSH (5000 microU/ml), by forskolin (0.01 mmol/l), by fetal calf serum (1-20%) and by epidermal growth factor (20 ng/ml). Stimulation was also apparent with gamma-interferon and with tumour necrosis factor at concentrations known to enhance class II major histocompatibility antigen expression by thyroid epithelium. The most potent factor tested was interleukin-1 (IL-1), which controls IL-6 release from other cell types. Threefold stimulation was found with 1 U/ml rising to 350-fold with 1000 U/ml. The effect of IL-1 took 2 h to develop and was blocked by cycloheximide (100 mumol/l). Stimulation was not markedly inhibited by pertussis toxin. Many of the actions of IL-1 are mediated by prostaglandin E2 (PGE2). At concentrations as low as 30 nmol/l, PGE2 stimulated IL-6 release but the maximum stimulation obtained with PGE2 was only threefold. The effect of IL-1 was not inhibited by indomethacin. These data provide further evidence that IL-6 is produced by human thyrocytes. The effect of IL-1 has not been demonstrated previously. Stimulation of IL-6 release by IL-1 did not appear to be mediated by prostaglandin.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of glucocorticoids and oestrogen on interleukin-6 production by human thyrocytes from patients with Graves' disease and toxic multinodular goitre and from HTori3 cells.

Interleukin-6 (IL-6) is a cytokine released by thyrocytes and is involved in disease processes such as autoimmune thyroid disease. The secretion of IL-6 can be stimulated by interleukin-1 (IL-1), tumour necrosis factor-alpha (TNF), serum, TSH and agents which increase intracellular cyclic AMP levels. Antithyroid drugs such as methimazole inhibit IL-6 production by thyrocytes but the effects of glucocorticoids and oestrogen have not been investigated. The effects of dexamethasone and 17 beta-oestradiol on IL-1-, TNF-, TSH-, forskolin- and phorbol 12-myristate 13-acetate (PMA)-stimulated IL-6 release in serum-free conditions were studied in human thyrocytes derived from patients with Graves' disease and toxic multinodular goitres, and in the immortalised human thyrocyte cell line, HTori3. Dexamethasone inhibited IL-6 production under stimulated conditions. In serum-free conditions, no basal release of IL-6 was assayable. In all but one of the primary thyroid cultures, TSH did not stimulate IL-6 release above the lower detectable limit of the assay. In Graves' and multinodular goitre thyrocytes, inhibition of IL-1 (100 U/ml)-stimulated IL-6 release by dexamethasone (100 nmol/l) was 62.51% +/- 10.43 (S.E.M.), and in HTori3 cells it was 78.35% +/- 3.9. The degree of IL-1 stimulation of IL-6 release and inhibition by dexamethasone was not significantly different in thyrocytes derived from either Graves' or multinodular glands. 17 beta-Oestradiol had no effect on IL-1-stimulated IL-6 release in either primary thyroid cell culture or in HTori3 cells.

Inverse correlation between activated T cells and TSH receptor antibodies in Graves' disease.

TSH receptor antibodies and peripheral blood lymphocyte subsets have been measured in fourteen patients with untreated Graves' thyrotoxicosis. CD8 (suppressor) cells were reduced (p less than 0.01) and helper/suppressor cell ratio was increased in Graves' patients. Increased levels of 4F2 positive (activated) T cells were found in the patients compared to controls (p less than 0.001) and there was a negative correlation between 4F2 positive cells and TSH receptor antibodies (TBII). It may be possible, with multiple immunological markers, to identify different stages in the pathogenesis of autoimmune thyroid disease.

The role of hCG in regulation of the thyroid gland in normal and abnormal pregnancy.

There is strong evidence that at least some forms of hCG can interact with and stimulate the thyroid both in vitro and in vivo. Changes in thyroid tests are sufficiently common in normal pregnancy for us to regard them as physiologic. The evidence that hCG is the agent responsible for these changes remains largely circumstantial but is now supported by an increasing body of evidence from laboratory studies. Trophoblastic tumors secrete variant forms of hCG that can stimulate the thyroid, but we do not know if they have any role in the extreme examples of thyroid stimulation encountered in normal pregnancy. Preparations of hCG from pregnancy urine bind to thyroid membranes from a wide variety of species, but they do not activate adenylate cyclase in all assay systems. The enzyme in human thyroid cells or membranes is, at best, only weakly stimulated by hCG. There are ample in vitro data that hCG can stimulate the thyroid, but studies using human thyroid cells have yielded conflicting results. The most direct evidence comes from the study of thyroid tests in normal pregnancy. In early pregnancy, when hCG concentrations are highest, free thyroid hormones are increased and serum TSH concentration is decreased. An inverse correlation exists between serum hCG and TSH concentrations, but hCG generally correlates poorly with individual thyroid tests. An activity in pregnancy serum related to hCG is able to stimulate FRTL-5 cells and may account for the changes in thyroid function observed in pregnancy. Structural considerations, along with data from biologic assays and sensitive thyroid function tests, suggest that hCG has significant thyroid-stimulating activity. This information suggests that the thyroid may be under dual control from both hCG and TSH in early pregnancy.

Application of the fuzzy ARTMAP neural network model to medical pattern classification tasks.

This paper presents research into the application of the fuzzy ARTMAP neural network model to medical pattern classification tasks. A number of domains, both diagnostic and prognostic, are considered. Each such domain highlights a particularly useful aspect of the model. The first coronary care patient prognosis, demonstrates the ARTMAP voting strategy involving 'pooled' decision-making using a number of networks, each of which has learned a slightly different mapping of input features to pattern classes. The second domain, breast cancer diagnosis, demonstrates the model's symbolic rule extraction capabilities which support the validation and explanation of a network's predictions. The final domain, diagnosis of acute myocardial infarction, demonstrates a novel category pruning technique allowing the performance of a trained network to be altered so as to favour predictions of one class over another (e.g. trading sensitivity for specificity or vice versa). It also introduces a 'cascaded' variant of the voting strategy intended to allow identification of a subset of cases which the network has a very high certainty of classifying correctly.

Application of autonomous neural network systems to medical pattern classification tasks.

This paper presents a study of the application of autonomously learning multiple neural network systems to medical pattern classification tasks. In our earlier work, a hybrid neural network architecture has been developed for on-line learning and probability estimation tasks. The network has been shown to be capable of asymptotically achieving the Bayes optimal classification rates, on-line, in a number of benchmark classification experiments. In the context of pattern classification, however, the concept of multiple classifier systems has been proposed to improve the performance of a single classifier. Thus, three decision combination algorithms have been implemented to produce a multiple neural network classifier system. Here the applicability of the system is assessed using patient records in two medical domains. The first task is the prognosis of patients admitted to coronary care units; whereas the second is the prediction of survival in trauma patients. The results are compared with those from logistic regression models, and implications of the system as a useful clinical diagnostic tool are discussed.

Autoantibodies in feline hyperthyroidism.

Thyroid autoantibodies have been demonstrated by indirect immunofluorescence in the sera of 10 of 29 (34 per cent) cats with hyperthyroidism. Antinuclear factor, rare in healthy cats, was found in a further four animals. Twenty-eight of the cats had a palpable goitre at first presentation. In 16 cases the goitre was unilateral, while in the others it was bilateral. Lymphocytic infiltration was present in nine of the 27 (33 per cent) thyroids examined histologically. Five of the sera gave a particularly strong reaction for thyroid antibodies. Four of these cases had bilateral goitres and lymphocytic infiltrations were found in four of the five thyroids (P less than 0.05). Twenty-one of the cats were followed up for a mean period of 11 months after operation, during which time three cats developed recurrent hyperthyroidism. Two had bilateral goitres with lymphocytic infiltration and the serum of both was strongly positive for thyroid microsomal antibodies. The third had unilateral goitre with lymphocytic infiltration and serum which was positive for antinuclear factor. In this case, the recurrence involved the lobe which had been previously operated on. Some cases of feline hyperthyroidism may be immunologically mediated and the condition is thus a potential model for some aspects of autoimmune thyrotoxicosis in man.

An artificial neural network system for diagnosis of acute myocardial infarction (AMI) in the accident and emergency department: evaluation and comparison with serum myoglobin measurements.

Recent studies have confirmed that artificial neural networks (ANNs) are adept at recognising patterns in sets of clinical data. The diagnosis of acute myocardial infarction (AMI) in patients presenting with chest pain remains one of the greatest challenges in emergency medicine. The aim of this study was to evaluate the performance of an ANN trained to analyse clinical data from chest pain patients. The ANN was compared with serum myoglobin measurements--cardiac damage is associated with increased circulating myoglobin levels, and this is widely used as an early marker for evolving AMI. We used 39 items of clinical and ECG data from the time of presentation to derive 53 binary inputs to a back propagation network. On test data (200 cases), overall accuracy, sensitivity, specificity and positive predictive value (PPV) of the ANN were 91.8, 91.2, 90.2 and 84.9% respectively. Corresponding figures using linear discriminant analysis were 81.0, 77.9, 82.6 and 69.7% (P < 0.01). Using a further test set from a different centre (91 cases), the accuracy, sensitivity, specificity and PPV for the admitting physicians were 65.1, 28.5, 76.9 and 28.6% respectively compared with 73.6, 52.4, 80.0 and 44.0% for the ANN. Although myoglobin at presentation was highly specific, it was only 38.0% sensitive, compared with 85.7% at 3 h. Simple strategies to combine clinical opinion, ANN output and myoglobin at presentation could greatly improve sensitivity and specificity of AMI diagnosis. The ideal support for emergency room physicians may come from a combination of computer-aided analysis of clinical factors and biochemical markers such as myoglobin. This study demonstrates that the two approaches could be usefully combined, the major benefit of the decision support system being in the first 3 h before biochemical markers have become abnormal.

Using patient-reportable clinical history factors to predict myocardial infarction.

Using a derivation data set of 1253 patients, we built several logistic regression and neural network models to estimate the likelihood of myocardial infarction based upon patient-reportable clinical history factors only. The best performing logistic regression model and neural network model had C-indices of 0.8444 and 0.8503, respectively, when validated on an independent data set of 500 patients. We conclude that both logistic regression and neural network models can be built that successfully predict the probability of myocardial infarction based on patient-reportable history factors alone. These models could have important utility in applications outside of a hospital setting when objective diagnostic test information is not yet be available.

Using classification tree and logistic regression methods to diagnose myocardial infarction.

Early and accurate diagnosis of myocardial infarction (MI) in patients who present to the Emergency Room (ER) complaining of chest pain is an important problem in emergency medicine. A number of decision aids have been developed to assist with this problem but have not achieved general use. Machine learning techniques, including classification tree and logistic regression (LR) methods, have the potential to create simple but accurate decision aids. Both a classification tree (FT Tree) and an LR model (FT LR) have been developed to predict the probability that a patient with chest pain is having an MI based solely upon data available at time of presentation to the ER. Training data came from a data set collected in Edinburgh, Scotland. Each model was then tested on a separate Edinburgh data set, as well as on a data set from a different hospital in Sheffield, England. Previously published models, the Goldman classification tree[1] and Kennedy LR equation[2], were evaluated on the same test data sets. On the Edinburgh test set, results showed that the FT Tree, FT LR, and Kennedy LR performed equally well, with ROC curve areas of 94.04%, 94.28%, and 94.30%, respectively, while the Goldman Tree's performance was significantly poorer, with an area of 84.03%. The difference in ROC areas between the first three models and the Goldman model is significant beyond the 0.0001 level. On the Sheffield test set, results showed that the FT Tree, FT LR, and Kennedy LR ROC areas were not significantly different (p > = 0.17), while the FT Tree again outperformed the Goldman Tree (p = 0.006). Unlike previous work[3], this study indicates that classification trees, which have certain advantages over LR models, may perform as well as LR models in the diagnosis of patients with MI.