Leptin and ghrelin concentration in hyperthyroid cats before and after radioactive iodine therapy compared to euthyroid control cats

Summary Objective: Leptin and ghrelin, two peptide hormones with antagonistic effects on satiety and energy balance, could be involved in the pathogenesis of weight loss and polyphagia in cats with hyperthyroidism. Leptin generally decreases appetite and increases energy expenditure, while ghrelin exerts the opposite effects. Materials and methods: Leptin and ghrelin were measured in 42 client owned hyperthyroid cats with a body condition score (BCS) 5/9 before (T0) and 4 weeks after radioactive iodine treatment (RAIT) (T1). Dependent on the serum total thyroxine concentration concentration at T1, cats were sub-classified as still hyperthyroid (ht-ht) (n = 4), euthyroid (ht-eu) (n = 10) or hypothyroid (ht-hypo) (n = 28). Results were compared to those of 22 healthy, euthyroid control cats with a comparable BCS ( 5/9) and age ( 8 years) to hyperthyroid cats. Results: At T0, there were no significant differences between hyperthyroid and control cats for leptin (p = 0.06) or ghrelin concentrations (p = 0.27). At T1, leptin significantly decreased in ht-hypo cats compared to T0 (p = 0.0008) despite a significantly increased body weight in this group (p = 0.0001). Serum ghrelin concentrations did not differ between hyperthyroid cats with a history of polyphagia compared to non-polyphagic cats (p = 0.42). After RAIT, ghrelin concentration significantly increased in all hyperthyroid cats (p < 0.0001), as well as in the subgroups ht-eu (p = 0.014) and ht-hypo (p < 0.0001) compared to their respective T0 baseline concentrations. Conclusion: Leptin and ghrelin fluctuations may be indicative of changes in metabolic functions in cats with thyroid dysfunction. Leptin fluctuations occurred independently of body weight in different states of thyroid dysfunction; increasing ghrelin concentrations after RAIT suggest a ghrelin-independent mechanism for polyphagia in hyperthyroid cats.


Background
Food intake and energy expenditure are processes tightly regulated by multiple peripheral and central acting signaling molecules. Studies in human and rodent models show leptin and ghrelin to be two major regulating hormones with antagonistic effects on satiety and energy balance (16,23,33). Acylated ghrelin is the active form of ghrelin predominantly secreted by the stomach in response to fasting (10). Ghrelin stimulates short-term food intake centrally via secretion of neuropeptide Y (NPY) and agouti-related peptide (AGRP) within the arcuate nucleus in the hypothalamus and peripherally via the afferent activity of the vagus nerve (23). After food intake, circulating ghrelin concentration rapidly decreases (33). Ghrelin also acts as a secretagogue of growth hormone and has effects on the long-term metabolic balance by the reduction of energy expenditure, promotion of weight gain and adipogenesis (7,17,32).
Leptin is predominantly secreted by adipocytes, with concentrations directly proportional to the amount of white adipose tissue as demonstrated in rodents, humans and cats (1,19). Hereby, leptin acts as a main peripheral cue signaling the body's energy stores to the brain and inhibits appetite and food intake centrally (6,25). Leptin and ghrelin appear to act as antagonists in a continuum of complex and partly unelucidated signaling pathways controlling feeding behavior and energy balance.
Hyperthyroidism is considered the most common endocrine disease in elderly cats. Thyroid hormone excess results in a hypermetabolic state associated with increased energy expenditure causing weight loss despite increased food intake. Given this apparent divergence, dysfunctions of the signaling circuits controlling feeding behavior as well as energy expenditure with implications on circulating leptin and ghrelin concentrations could be expected. To our knowledge, there is only one study investigating serum leptin concentration in hyperthyroid cats both before and after treatment with methimazole. It concluded that leptin concentration increases in cats controlled with this antithyroid medication (15).
As would be expected in a polyphagic condition such as hyperthyroidism, reduced periods of fasting result in suppressed ghrelin production and ghrelin production would be expected to increase following successful treatment of hyperthyroidism. This has been the finding in studies of hyperthyroidism in both rodents and humans (14,28,30). There has been a paucity of data on ghrelin concentration in cats with hyperthyroidism.
The aims of the present study were to determine serum leptin and plasma acylated ghrelin concentration in hyperthyroid cats before and after radioactive iodine treatment (RAIT) and to compare results to euthyroid control cats. Both leptin and ghrelin, have been reported to be affected by body composition (1,2,9). Hyperthyroi dism usually occurs in middle-aged to older cats (26,27). Therefore, cats with a similar age and body condition served as a control group. Moreover, we investigated possible similarities or discordances with the reported effects of hyperthyroidism, hypothyroidism and anti-thyroid therapy in other species or other treatment types.

Animal selection
Twenty-two healthy cats presented to the Small Animal Clinic, Justus-Liebig University (Giessen, Germany), for a geriatric check-up between March 2011 and October 2012 were included and served as a control group. Control cats were included if classified as healthy based on their medical history (including absence of recent weight loss and inappetence), physical examination, complete blood count (CBC), biochemistry profile (BC) and serum total thyroxine concentration (tT4). In order to include cats with a similar age and body condition compared to hyperthyroid cats of our study, control cats had a minimum age of ≥ 8 years and a BCS ≤ 5/9 (8). Cats were considered euthyroid if their tT4 was within the laboratory's 1 reference interval (RI) (1.0-4.0 µg/dl; 13-51 nmol/l).
During the same time 42 client owned hyperthyroid cats admitted to the Royal Veterinary College, London, UK for RAIT were included. Cats were previously diagnosed as hyperthyroid at the primary veterinarian based on an increased serum tT4 concentration. All cats underwent a physical examination, followed by a CBC, BC, urinalysis, and systolic blood pressure measurement. Hyperthyroid cats were included if they exhibited a similar and common lean phenotype with a normal to decreased BCS ≤ 5/9. Cats with a serum creatinine concentration above the laboratory's 2 RI (74.5-185.30 µmol/l) while being hyperthyroid (based on preRAIT tT4) were excluded as were cats with significant concurrent disease (e. g. diabetes mellitus) or significant abnormalities on general physical examination (e. g. abdominal mass). Thirty-six out of 42 cats were previously treated with antithyroid medication, which was discontinued for a minimum of 10 days prior to RAIT. Diagnosis of hyperthyroidism was confirmed on admission to the Royal Veterinary College based on a tT4 above the laboratory's 2 RI (1.5-5.0 µg/dl; 19-65 nmol/l) 1-2 days prior RAIT.
Each cat received a dose of 111, 148 or 185 MBq of radioactive iodine (I 131 ), depending on severity of clinical signs and tT4 (4,26,27). Measurement of tT4 was repeated 4 weeks after RAIT. Cats were classified into three subgroups according to their tT4 after RAIT. Cats with a tT4 within, below or above the RI were classified as euthyroid (ht-eu), hypothyroid (ht-hypo) or hyperthyroid (ht-ht), respectively.

Blood sampling and measurement of leptin and ghrelin
All blood samples were collected after the cats were fasted for a minimum of 8 hours. Residual blood samples for analysis of 1 Biocontrol, Ingelheim am Rhein, Germany 2 Diagnostic Laboratories, Department of Pathology and Pathogen Biology, Royal Veterinary College, London UK serum leptin and plasma acylated ghrelin were handled in a standardized fashion and obtained at baseline and in hyperthyroid cats 4 weeks after RAIT at the same time blood samples for CBC, BC and tT4 were obtained. Whole blood was placed in plain tubes (1.5 ml SC Micro Tube, Sarstedt, Nümbrecht, Germany) for measurement of serum leptin or EDTA + aprotinin tubes (Micro tube 1.3 ml K3E, Sarstedt, Nümbrecht, Germany; aprotinin from bovine lung, Sig ma-Aldrich®, Darmstadt, Germany, 0.005-0.01 trypsin inhibitor units/ml blood) for measurement of plasma acylated ghrelin, respectively. Blood samples were processed within 30 minutes, spun down in a cooled centrifuge (2 °C) and stored at -80 °C before analysis. At the time of analysis, samples were shipped as a batch to the analyzing laboratory 3 via overnight express on dry ice.
Plasma ghrelin concentrations were measured with a human active ghrelin RIA kit from Linco Research (Cat.# GHRA-88HK). Due to a high structural similarity between feline and human ghrelin, human kits have been shown to be appropriate for the measurement of feline ghrelin (12). The cross reactivity of the RIA kit is 100% with human ghrelin, rat ghrelin, canine ghrelin and feline ghrelin. The kit was previously validated for parallelism responses to increasing dilutions of plasma from five cats. Intraand interassay coefficients of variation (CV) were both 6.7% for a mean concentration of 140 pg/ml (20,21).

Statistical analysis
Results were tested for normality by D' Agostino & Pearson omnibus normality test. For the purpose of statistical analysis any tT4 concentration below the laboratory's limit of detection (< 5.1 nmol/l) was assigned a value of 5.1 nmol/l. Age, weight and BCS were compared between control and hyperthyroid cats using the unpaired t-test and the chi-square test for evaluation of the gender ratio. The Mann-Whitney U-test was used for comparisons of leptin, ghrelin and tT4 between control and hyperthyroid cats before RAIT. The treatment effect was evaluated using the Wilcoxon signed-rank test between the three subgroups. Correlations between leptin, ghrelin and tT4 and body weight were examined using the Spearman's coefficient of correlation. For all tests, level of statistical significance was set at p < 0.05. Normally distributed data are reported as mean ± standard deviation (SD). Data not normally distributed are presented as median [25th, 75th percentile]. Data analysis was conducted using computerized statistical software (Prism6, GraphPad, San Diego, California).

Leptin
At baseline (T0) residual serum samples for measurement of leptin were available for all control cats (n = 22), 36/42 hyperthyroid cats before and 35/42 cats after treatment. Before treatment (T0), leptin was not significantly different between control and hyperthyroid cats (p = 0.06) (▶ Table 2, ▶ Fig. 1). After RAIT, leptin concentration significantly decreased in hyperthyroid cats compared to their baseline concentration (p = 0.002) even though their body weight significantly increased (p ≤ 0.0001). Leptin concentration in RAI treated cats was found to be significantly lower compared with euthyroid control cats (p = 0.0004).
Upon analysing different subgroups, the decline in leptin concentration at T1 compared to T0 was driven by the group of hypothyroid cats (p = 0.0008) while there was no difference in cats considered euthyroid after RAIT (p = 0.74). Body weight significantly increased in both groups (p < 0.0001). 3 Department of Biology, Pathology and Food Science, Oniris, National College of Veterinary Medicine, Food Science and Engineering, Nantes-Atlantic, France  Results of the four ht-ht cats were not analyzed statistically, due to the small sample size. Leptin was not correlated with either body weight, tT4 or ghrelin in control or hyperthyroid cats or in the subgroups ht-eu and ht-hypo after RAIT.

Ghrelin
At baseline (T0) residual serum samples for ghrelin meassurement were available for all control cats (n = 22), 42/42 hyperthyroid cats before and 40/42 cats after treatment. Similar to leptin, ghrelin was not significantly different between control and hyperthyroid cats at T0 (p = 0.27) (▶ Table 2 Table 2, ▶ Fig. 2). Comparing different subgroups, there remained an increase in ghrelin concentration at T1 compared to T0 (ht-eu p = 0.014, ht-hypo p < 0.0001). Again, results of ht-ht cats were not evaluated statistically, given the small sample size. Ghrelin was not correlated with body weight, tT4 or leptin in control or hyperthyroid cats or in the subgroups ht-eu and ht-hypo after RAIT.

Discussion
Results of our study showed a significant decrease of leptin after RAIT in hyperthyroid cats despite a significant increase in body weight. Comparing different subgroups of cats, this decline was solely driven by a significant decrease of leptin in ht-hypo cats. In ht-eu cats, leptin concentration did not change following RAIT. The increase in body weight, however, remained significant in both subgroups. Leptin has previously been shown to be correlated with body weight or body fat in cats, humans and rodent models and to increase in cats treated for hyperthyroidism (1,3,13,15,19).
However, results of several studies suggest that the control of leptin by adipocytes may be overridden by the thyroid axis in hyper-or hypothyroid states (11,13,35). In human patients thyroid hormones play a critical role in leptin metabolism independent of body mass index or body fat (11). In hypothyroid people, hypoleptinemia has previously been reported and leptin increased with levothyroxine therapy (13). In addition, preadipocytes from mice in vitro show leptin mRNA expression and leptin secretion to be dependent on physiological levels of serum triiodthyronine (35). These findings led to the hypothesis that thyroid hormones may act as secretagogues for leptin from adipocytes. Thus, in hypothyroid states, leptin might be negatively influenced, which could explain the decreased leptin concentration in the ht-hypo subgroup of this study. To the authors' knowledge, there is only one study in cats comparing leptin in hyperthyroid cats before and after treatment (15
p value < 0.0001 0.003 0.0001 Table 2 Leptin and ghrelin concentration in control cats and hyperthyroid cats, before and after treatment with radioactive iodine therapy.

Tab. 2 Leptin-und Ghrelinkonzentration bei Kontrollkatzen und hyper thyreoten Katzen vor und nach Radiojodtherapie
leptin at 3 months after initiation of medical therapy with methimazole. However, these results refer to cats with a normal or even high tT4 concentration on medical therapy. The present study reports decreasing leptin only in cats with low tT4 levels 4 weeks after RAIT. Again, this might point to the thyroid axis outweighing body fat mass in the ability to control serum leptin concentration. Plasma ghrelin concentration did not differ between control and hyperthyroid cats before treatment. After RAIT, ghrelin increased significantly in hyperthyroid cats and all subgroups analyzed compared to the concentration prior treatment and compared to control cats. Similar results were found in hyperthyroid rodent models and in people after various treatments for hyperthyroidism (5,9,28,30). It was suspected to reflect a possible response of this prometabolic peptide hormone to the abruptly declining metabolic rate after treatment for hyperthyroidism (28). It also led to the conclusion that ghrelin might not cause polyphagia frequently seen in hyperthyroidism in these species. In the current study, besides weight loss, polyphagia was among the most frequent clinical sign in hyperthyroid cats. However, ghrelin in polyphagic hyperthyroid cats was not different from those in nonpolyphagic cats. Moreover, ghrelin markedly increased after treatment of hyperthyroid cats compared to concentration at baseline and compared to control cats. Thus, a similar conclusion of ghrelin-independent polyphagia in hyperthyroid cats can be drawn from findings of this study.
There are several limitations associated with this study. Control and hyperthyroid cats were recruited from different locations and CBC, BC and tT4 assays were run in different laboratories. To address this potential error source, only standardized laboratories with established reference intervals for all parameters evaluated were used.
Also, the sampling time point with regards to the time after RAIT and time of the day may have had an impact on the results. Hyperthyroid cats often display temporarily low tT4 concentration  shortly after RAIT and thyrotroph recovery after suppression by thyroid hormones may take up to 3 months (34).
Blood sampling was performed 4 weeks after RAIT, which was reflected by the high rate of hypothyroid cats. It remains unclear whether these cats were transiently or permanently hypothyroid, since follow up samples were not available. Thyroid recovery ultimately is dependent on different factors like dosing regime for I 131 or pretreatment and may vary between individual cats and institutions. In addition, we cannot exclude the possibility that different time points of blood sampling during the day influenced our results, since circadian rhythmicity has been reported for leptin and ghrelin in humans and rodent models (19,29,31).
Cats were also not fed a standardized diet before or during the study period and food intake was not controlled, which might have affected our study results. However, ghrelin concentration appears to be independent of dietary macronutrients in humans and cats (22,24). Although, we intended to match euthyroid control cats and hyperthyroid cats with respect to their body condition, control cats were significantly heavier and had a significantly higher BCS -a result not entirely unexpected. The majority of healthy geriatric pet cats are expected to be even overweight, due to a typical modern lifestyle (e. g. indoor, ad libitum feeding etc.). To recruit healthy geriatric cats with an ideal or even lean BCS proved to be a challenge and thus the significant higher weight and BCS are not surprising. However, by carefully taking the history with special emphasis on the absence of recent weight loss, inappetence, or other signs of systemic disease in addition to a thorough physical examination and laboratory testing including tT4, we tried to assure that only truly healthy cats entered the study. Most cats were reported as being lean for their whole life or becoming more lean when having outdoor access (e. g. during summertime). The fact that besides BCS, body weight was significantly lower, reflects the correlation between body weight and body condition. As a final limitation of the study, we did not measure parameters that may influence leptin and ghrelin concentration, such as body fat, glucose or insulin, thus explanations for changes in leptin and ghrelin with regards to these factors remains elusive.