Investigation of the pathophysiological mechanism for altered calcium homeostasis in hyperthyroid cats

Authors


  • Results from this study were presented at ACVIM Forum 2013, Seattle, USA.

Abstract

Objective

To investigate possible pathophysiological mechanisms (reduced plasma calcitriol concentrations and/or presence of concurrent or masked chronic kidney disease) for hypocalcaemiain hyperthyroid cats.

Methods

Prospective cohort study. Routine plasma biochemical parameters, plasma parathyroid hormone and calcitriol concentrations, ionized calcium concentrations, and venous pH, were measured at diagnosis and following treatment of hyperthyroidism. Linear regression analysis was used to determine predictors of ionized calcium concentration.

Results

Hyperthyroid cats (n=45) had lower ionized calcium concentrations than healthy geriatric cats (n=52), however, ionized calcium concentrations were higher in hyperthyroid cats with concurrent or masked chronic kidney disease than non-azotaemic hyperthyroid cats. Plasma calcitriol concentrations were higher in hyperthyroid cats than control cats. Plasma total thyroxine concentration and venous pH were independent predictors of ionized calcium concentration. Plasma total thyroxine concentration was also a predictor of ionized calcium concentration after adjustment for plasma parathyroid hormone and calcitriol concentrations.

Clinical Significance

Hypocalcaemia in hyperthyroid cats is not associated with the presence of concurrent or masked chronic kidney disease or reduced plasma calcitriol concentrations. Increased thyroid hormone concentrations might influence ionized calcium concentrations through a mechanism, yet to be determined, that is independent of control by parathyroid hormone and calcitriol.

Abnormalities in calcium and phosphate homeostasis are common in hyperthyroid cats (Archer & Taylor 1996, Barber & Elliott 1996, Williams et al. 2012). Hyperthyroid cats have elevated plasma concentrations of phosphate and parathyroid hormone (PTH), and reduced plasma concentrations of ionized calcium and fibroblast growth factor-23 (FGF-23) (Archer & Taylor 1996, Barber & Elliott 1996, Williams et al. 2012). In contrast, human patients with Graves’ disease have elevated ionized calcium, phosphate and FGF-23 concentrations in conjunction with decreased PTH concentrations (Mosekilde et al. 1990, Iqbal et al. 2003, Yamashita et al. 2005). Many of the pathophysiological changes in calcium and phosphate homeostasis that are observed in hyperthyroid cats mimic the changes seen in cats with chronic kidney disease (CKD) and secondary renal hyperparathyroidism (Barber & Elliott 1996). However, not all hyperthyroid cats with derangements in calcium and phosphate homeostasis go on to develop azotaemic CKD after treatment (Williams et al. 2012).

The differences in calcium and phosphate homeostasis between hyperthyroid cats and human patients with Graves’ disease could be explained by the disparity in ionized calcium concentrations between hyperthyroid cats (hypocalcaemic) and Graves’ disease patients (hypercalcaemic). The pathophysiological mechanism for hypocalcaemia in hyperthyroid cats is currently unknown. Ionized calcium concentrations might be decreased because of the presence of concurrent hyperphosphataemia in hyperthyroidism, as phosphate will complex with calcium to decrease the ionized calcium concentration (Barber & Elliott 1996). Hypocalcaemia might also occur as a result of decreased intestinal absorption of calcium, decreased renal reabsorption of calcium or decreased bone turnover. Renal reabsorption of calcium is increased by PTH (Guyton & Hall 2000), which is elevated in 55 to 77% of hyperthyroid cats (Barber & Elliott 1996, Williams et al. 2012), therefore renal reabsorption of calcium should be increased in hyperthyroid cats. Bone turnover would also appear to be increased, rather than decreased, in hyperthyroid cats (Archer & Taylor 1996).

Intestinal absorption of calcium is regulated by calcitriol (DeLuca 1988), therefore reduced plasma calcitriol concentrations could cause decreased intestinal absorption of calcium and hypocalcaemia. One in vitro study reported that thyroid hormones could suppress calcitriol synthesis in a perfused rat kidney model (Kano & Jones 1984), and human patients with Graves’ disease have decreased serum calcitriol concentrations (Bouillon et al. 1980, Jastrup et al. 1982, Czernobilsky et al. 1988). Therefore, if hyperthyroidism was associated with decreased plasma calcitriol concentrations in cats, this could be the pathophysiological mechanism for hypocalcaemia. Plasma calcitriol concentrations are also reduced, relative to the prevailing PTH concentration, in cats with CKD (Barber & Elliott 1998), and therefore hyperthyroid cats with concurrent CKD might also have calcitriol deficiency and thus hypocalcaemia. As calcitriol also inhibits PTH synthesis by the parathyroid gland (Silver et al. 1986), it is possible that reduced plasma calcitriol concentrations could contribute to the elevated plasma PTH concentrations observed in hyperthyroid cats (Barber & Elliott 1996, Williams et al. 2012). One previous cross-sectional study reported that plasma calcitriol concentrations were not significantly different between hyperthyroid cats and a group of healthy geriatric cats (Barber & Elliott 1996), however, this study only included eight hyperthyroid cats, and so was statistically underpowered to detect any significant differences.

The aim of this study was to assess ionized calcium and plasma calcitriol concentrations in hyperthyroid cats with and without concurrent CKD, in order to investigate the association between calcitriol, ionized hypocalcaemia and elevated PTH concentrations.

METHODS

Patient selection

Newly diagnosed hyperthyroid cats from two London-based first opinion practices (Peoples Dispensary for Sick Animals, Bow and Beaumont Sainsbury Animal Hospital, Camden) were prospectively recruited into the study between January 1, 2011 and February 8, 2012. Diagnosis of hyperthyroidism was based on a plasma total thyroxine (T4) concentration by chemiluminescence (Immulite) greater than the laboratory reference range (>55 nmol/L), or a plasma total T4 greater than 40 nmol/L in conjunction with a serum free T4 concentration by equilibrium dialysis greater than 40 pmol/L and compatible clinical signs.

The following data were collected at the time of diagnosis; plasma total T4 concentration, routine plasma biochemical variables, whole blood ionized calcium concentration and venous pH. Cats included in the study were all treated for hyperthyroidism with anti-thyroid medication (carbimazole or methimazole) at standard dosages, except for those cats which had adverse reactions to anti-thyroid medication which were treated by thyroidectomy.

Cats were reassessed and blood sampled every 3 to 4 weeks after commencement of treatment until hyperthyroidism was controlled (total T4 <40 nmol/L). At the time of first documentation of controlled hyperthyroidism, plasma biochemistry, ionized calcium and venous pH measurement were repeated. Following this visit, cats were re-evaluated every 8 weeks and plasma biochemistry (and total T4 if indicated) was repeated every 4 months.

Renal azotaemia was defined as a plasma creatinine concentration greater than 177 µmol/L in conjunction with inadequate urine concentrating ability (USG <1·035), or persistent azotaemia on two or more consecutive occasions without evidence of a pre-renal cause. Cats that were azotaemic at the time of diagnosis of hyperthyroidism and initially non-azotaemic hyperthyroid cats which developed azotaemia within approximately 4 months of establishment of euthyroidism were combined and included in the azotaemic-CKD group. All other cats were included in the non-azotaemic group.

A control group of non-azotaemic euthyroid (total T4<40 nmol/L) geriatric cats (>9 years old) were also recruited into the study for comparison of ionized calcium concentrations. Cats with evidence of significant systemic disease (e.g. diabetes mellitus or neoplasia) were excluded from this control group. In addition, plasma calcitriol and total 25-hydroxyvitamin D (total 25OHD, defined as the sum of 25(OH)D2 and 25(OH)D3) concentrations were measured in a further group of 10 euthyroid (total T4 <40 nmol/L) geriatric cats that had serial plasma creatinine concentrations less than 140 µmol/L for a one-year period.

Blood and urine sampling and processing

Blood and urine samples were collected as part of a geriatric screening and health care programme at the time of diagnosis with the consent of the owner. Jugular venous blood samples were collected and ionized calcium concentration and venous pH were determined immediately following blood sampling using whole blood (ISTAT; Woodley Equipment). Residual blood was placed in heparinized and ethylenediaminetetraacetic acid (EDTA) tubes, and urine was collected by cystocentesis. Samples were kept at 4°C before sample processing which occurred within 6 h of collection. Blood samples were centrifuged at 2016 × g for 10 min to enable separation of plasma from cellular components. Heparinized plasma was submitted to a single external laboratory (IDEXX Laboratories) for biochemical analysis including total T4. EDTA plasma was stored at −80°C until batch analysis of PTH, FGF-23, calcitriol and total 25OHD. Plasma PTH concentrations were measured using a recently validated human intact PTH immunoradioimmetric assay (total intact PTH immunoradiometric assay; Scantibodies) (Pineda et al. 2012, Williams et al. 2012). Plasma FGF-23 concentrations were measured in EDTA plasma using a commercially available human intact FGF-23 ELISA that has been recently validated for use in feline plasma samples (human intact FGF-23 ELISA; Kainos Laboratories) (Geddes et al. 2013). Vitamin D metabolites were measured as described in detail previously (Mawer et al. 1985, Mawer et al. 1990, Berry et al. 2000). Briefly, samples were extracted using acetonitrile and applied to C18 Silica Sep-paks. Separation of metabolites was by straight phase high-performance liquid chromatography (HPLC) (Waters Associates) using a Hewlett-Packard Zorbax-Sil Column (Hichrom) eluted with hexane: propan2ol: methanol (92:4:4). The aliquot containing 25(OH)D2 plus 25(OH)D3 (i.e. total 25OHD) was collected and applied to a second Zorbax-Sil Column eluted with hexane: propan2ol (98:2), quantified by UV absorbance at 265 nm and corrected for recovery. Following separation by HPLC, the aliquot containing 1,25(OH)2D (both D2 and D3) was quantified by radioimmunoassay as described in detail elsewhere (Mawer et al. 1990). The inter-assay coefficient of variation (CV) between two assay runs for calcitriol and total 25OHD was calculated.

Statistical analysis

Statistical analyses were performed using computerized statistical software (PASW Version 18.0). A reference range for ionized calcium concentration was derived from mean ±1·96*standard deviation (sd) of ionized calcium concentration in control cats. Results are reported as median (25th, 75th percentile) and statistical significance was defined as P≤0·05. The Mann–Whitney U test was used to compare baseline variables between hyperthyroid cats and control cats and between the azotaemic CKD and non-azotaemic groups. The Wilcoxon-signed rank test was used to compare selected variables before and after treatment of hyperthyroidism.

Correlations were assessed with Spearman's correlation coefficient. Linear regression analysis was performed to identify clinicopathological variables of interest that were significantly associated with ionized calcium concentration in hyperthyroid cats (P<0·1). Variables that were significantly associated with ionized calcium concentration at the 10% level were carried forward to a multi-variable linear regression model to identify independent predictors of ionized calcium concentration. In addition, a second multi-variable linear regression model which included plasma concentrations of total T4, PTH and calcitriol was constructed, in order to determine if plasma total T4 was associated with ionized calcium concentration independent of changes in plasma PTH and calcitriol concentrations.

RESULTS

Ionized calcium

Fifty-two control cats were included in the study and from these cats a reference range for ionized calcium concentration was derived as 1·18 to 1·34 mmol/L. Forty-five hyperthyroid cats were recruited into the study and post-treatment data was available in 25 cats. Three hyperthyroid cats were azotaemic at baseline, and nine hyperthyroid cats developed azotaemia within 4 months of establishment of euthyroidism, thus there were 12 cats in the azotaemic-CKD group. The remaining 33 hyperthyroid cats constituted the non-azotaemic group. Baseline clinicopathological data are summarized in Table 1.

Table 1. Baseline clinicopathological data for hyperthyroid cats with and without concurrent azotaemic chronic kidney disease
VariableNon-azotaemic group Median (25th, 75th percentile)nAzotaemic-CKD group Median (25th, 75th percentile)nSig.
  1. Sig. significance

  2. Non-azotaemic group were non-azotaemic hyperthyroid cats which remained non-azotaemic for at least 4 months after establishment of euthyroidism. Azotaemic-CKD group were hyperthyroid cats which were azotaemic at the time of diagnosis of hyperthyroidism and hyperthyroid cats which developed azotaemia within 4 months of establishment of euthyroidism. Comparisons between groups were made by Mann–Whitney U test

  3. Variables that were significant at the 5% level (P≤0·05) are in bold

Age (years)15·0 [11·5, 16·5]3216·6 [13·7, 18·0]120·054
Plasma total thyroxine concentration (nmol/L)154·0 [75·0, 201·0]3375·4 [62·8, 86·5]120·019
Plasma creatinine concentration (µmol/L)95·4 [83·0, 114·3]33144·3 [106·8, 176·1]120·001
Plasma phosphate concentration (mmol/L)1·75 [1·55, 1·96]331·63 [1·23, 2·04]120·355
Plasma total calcium concentration (mmol/L)2·37 [2·28, 2·47]332·50 [2·40, 2·55]120·007
Plasma albumin concentration (g/L)31·1 [29·3, 32·8]3331·8 [29·6, 32·9]120·563
Plasma globulin concentration (g/L)40·6 [37·7, 45·8]3341·1 [38·7, 47·9]120·626
Plasma parathyroid hormone concentration (pg/mL)35·2 [22·7, 57·0]1134·1 [6·1, 61·8]90·704
Plasma fibroblast growth factor-23 concentration (pg/mL)169·6 [145·1, 295·2]11317·5 [214·4, 800·0]90·062
Venous pH7·361 [7·329, 7·383]327·346 [7·319, 7·365]120·406

Hyperthyroid cats had significantly lower ionized calcium concentrations than control cats [hyperthyroid group 1·24 (1·21, 1·27) mmol/L, control group 1·26 (1·23, 1·29) mmol/L; P=0·003]. Five hyperthyroid cats (11%) were hypocalcaemic (ionized calcium concentration <1·18 mmol/L) and no hyperthyroid cats were hypercalcaemic (ionized calcium concentration >1·34 mmol/L). Ionized calcium concentration did not change significantly following treatment of hyperthyroid cats [pre-treatment 1·24 (1·22, 1·28) mmol/L to 1·25 (1·23, 1·28), n=25; P=0·085, Fig 1].

Figure 1.

Ionized calcium concentrations in 45 hyperthyroid cats. Post-treatment ionized calcium concentrations were available in 25 cats. The dotted lines represent the upper and lower limits of a reference range for ionised calcium (1·18-1·34 mmol/L) which was derived from 52 healthy geriatric cats

Hyperthyroid cats in the azotaemic-CKD group had significantly higher baseline ionized calcium concentrations than cats in the non-azotaemic group [1·27 (1·24, 1·28) mmol/L, n=12 versus non-azotaemic group 1·23 (1·20, 1·26) mmol/L, n=33; P=0·008]. There was a weak positive correlation between ionized calcium concentration and plasma total calcium concentration (rs=0·403; P=0·006) and plasma creatinine concentrations (rs=0·297; P=0·047). Ionized calcium concentrations also showed a moderate negative correlation with plasma total T4 (rs=−0·581; P<0·001). There were no significant correlations between ionized calcium concentration and plasma concentrations of PTH (P=0·335), FGF-23 (P=0·927), calcitriol (P=0·697) or total 25OHD (P=0·182).

Calcitriol and 25-hydroxyvitamin D

Mean (±sd) inter-assay CV for plasma calcitriol and total 25OHD were 14·7 ±9·3% (n=4) and 4·1 ±2·1% (n=5), respectively. In 10 healthy cats, plasma calcitriol concentrations ranged between 35 and 174 pmol/L and plasma total 25OHD concentrations ranged between 65·9 and 154·4 nmol/L. Calcitriol and total 25OHD were only measured in 20 hyperthyroid cats before and after treatment due to the high costs of analysis. Hyperthyroid cats had significantly higher plasma calcitriol concentrations than control cats [hyperthyroid group 115 (99, 144) pmol/L, n=20, control group 89 (61, 108) pmol/L, n=10; P=0·037], however, only two hyperthyroid cats (10%) had a plasma calcitriol concentration above the highest value recorded for a control cat (>174 pmol/L). There was no significant difference in plasma total 25OHD concentrations between hyperthyroid and control cats [hyperthyroid group 122·9 (94·4, 146·1) nmol/L, n=20, control group 100·0 (79·6, 115·9) nmol/L, n=10; P=0·135]. Plasma calcitriol concentration did not change significantly following treatment of hyperthyroidism [pre-treatment 115 (99, 144) pmol/L, post-treatment 131 (109, 145) pmol/L, n=20; P=0·287, Fig 2]. Similarly, there was no significant change in plasma total 25OHD concentration after treatment of hyperthyroidism [pre-treatment 122·9 (94·4, 146·1) nmol/L, post-treatment 120·4 (103·0, 162·0), n=20; P=0·167, Fig 3].

Figure 2.

Plasma calcitriol concentrations in 20 hyperthyroid cats before and after treatment. The dotted lines represent the upper and lower limits of a reference interval for plasma calcitriol (35–174 pmol/L) which was derived from 10 healthy geriatric cats

Figure 3.

Plasma total 25-hydroxyvitamin D (total 25OHD) concentrations in 20 hyperthyroid cats before and after treatment. The dotted lines represent the upper and lower limits of a reference interval for plasma total 25OHD concentration (65·9-154·4 nmol/L) which was derived from 10 healthy geriatric cats

Plasma calcitriol concentrations were not significantly different between hyperthyroid cats in the azotaemic-CKD and non-azotaemic groups [azotaemic-CKD group 119·0 (74·0, 149·0) pmol/L, n=9, non-azotaemic group 113·0 (103·0, 144·0) pmol/L, n=11; P=1·000]. Plasma total 25OHD concentrations were also not significantly different between hyperthyroid cats in the azotaemic-CKD and non-azotaemic groups [azotaemic-CKD group 110·2 (97·5, 143·1) nmol/L, n=9, non-azotaemic group, 124·3 (66·6, 154·6) nmol/L, n=11; P=0·970].

Plasma calcitriol concentrations were not correlated with plasma concentrations of total T4 (rs=−0·175, n=20; P=0·462) at baseline. There was also no significant correlation between plasma calcitriol concentration and plasma concentrations of PTH (rs=−0·188, n=20; P=0·427), FGF-23 (rs=0·026, n=20; P=0·915), total 25OHD (rs=0·108, n=20; P=0·652) or ionized calcium (rs=−0·093, n=20; P=0·697) at baseline. Following treatment, plasma calcitriol concentration was moderately positively correlated with plasma PTH (rs=0·619, n=20; P=0·004) and FGF-23 (rs=0·531, n=20; P=0·016) concentrations, however, there was no significant correlation with ionized calcium concentration (rs=−0·104, n=20; P=0·664).

Variables associated with ionized calcium concentration

Linear regression analysis identified total T4 (P<0·001), venous pH (P=0·019), and plasma globulin (P=0·002) as being significantly associated with ionized calcium concentration at the 10% level. Plasma concentrations of creatinine (P=0·109), albumin (P=0·125), PTH (P=0·691) and calcitriol (P=0·646) were not significantly associated with ionized calcium concentration at the 10% level. In the first multi-variable linear regression model (Table 2), plasma total T4 and venous pH were independent predictors of ionized calcium concentration and plasma globulin concentration tended towards an independent association with ionized calcium concentration. In the second multi-variable model which included total T4, PTH and calcitriol, only total T4 was an independent predictor of ionized calcium concentration (Table 3).

Table 2. Multivariable linear regression model of predictors of ionized calcium concentration in hyperthyroid cats at diagnosis (adjusted r2=0·417, n=44; P<0·001)
VariableBseSig.95% CI for B
  1. se standard error, CI confidence interval, Sig. significance

  2. Plasma total thyroxine concentrations were divided by 10 to allow easier interpretation of B values

  3. Variables that were significant at the 5% level (P≤0·05) are in bold

Plasma total thyroxine concentration/10 (nmol/L)−0·0030·0010·003−0·005 to 0·001
Venous pH−0·2910·1260·026−0·545 to 0·036
Plasma globulin concentration (g/L)0·0020·0010·0840·000 to 0·004
Constant3·3380·9220·0011·476 to 5·201
Table 3. Multi-variable linear regression model of predictors of ionized calcium concentration in hyperthyroid cats at diagnosis (adjusted r2=0·338, n=20; P=0·022)
VariableBSESig.95% CI for B
  1. se standard error, CI confidence interval, Sig. significance

  2. Plasma total thyroxine, parathyroid hormone and calcitriol concentrations were divided by 10 to allow easier interpretation of B values

  3. Variables that were significant at the 5% level (P≤0·05) are in bold

Plasma total thyroxine concentration/10 (nmol/L)−0·0050·0010·003−0·008 to 0·002
Plasma parathyroid hormone concentration/10 (pg/mL)0·0010·0010·442−0·002 to 0·004
Plasma calcitriol concentration/10 (pmol/L)−0·0020·0030·492−0·007 to 0·004
Constant1·3170·038<0·0011·237 to 1·397

DISCUSSION

The results of this study demonstrate that hyperthyroid cats have a lower ionized calcium concentrations than healthy geriatric cats, as reported previously (Barber & Elliott 1996), however, the vast majority of hyperthyroid cats had ionized calcium concentrations within the reference interval.

Ionized calcium concentrations were higher in hyperthyroid cats with concurrent or masked CKD than non-azotaemic hyperthyroid cats in this study. These data suggest that the presence of concurrent or masked CKD is not the cause of ionized hypocalcaemia in hyperthyroid cats. The reason why hyperthyroid cats with concurrent or masked CKD have increased ionized calcium concentrations compared to non-azotaemic cats is unknown. Hyperthyroid cats which develop azotaemia after treatment have lower plasma total T4 than non-azotaemic hyperthyroid cats (Williams et al. 2010), and plasma total T4 was negatively correlated with ionized calcium concentration (independent of other factors) in this study. Therefore, it could be speculated that the higher ionized calcium concentration observed in hyperthyroid cats with concurrent or masked CKD in this study reflects the lower plasma total T4 observed in this group. However, the relationship between plasma total T4 and ionized calcium concentration in hyperthyroid cats with concurrent or masked CKD is likely to be confounded by the suppression of plasma total T4 in cats with non-thyroidal illnesses such as CKD (sick euthyroid effect) (Mooney et al. 1996). It is possible that there is an indirect relationship between plasma total T4 and ionized calcium concentration in hyperthyroid cats with concurrent or masked CKD; plasma total T4 may be suppressed secondary to the sick euthyroid effect, and ionized calcium concentration may be increased by another mechanism.

Calcitriol is an important hormone for calcium homeostasis as calcitriol can stimulate increased calcium absorption in the intestine and potentiate increased bone turnover (DeLuca 1988). Serum calcitriol concentrations are suppressed in human patients with Graves’ disease (Bouillon et al. 1980, Jastrup et al. 1982, Czernobilsky et al. 1988), however, serum calcitriol concentrations are increased in human patients with toxic multinodular goitre, despite concurrent suppression of PTH (Czernobilsky et al. 1988). In vitro studies have also reported that thyroid hormone can suppress calcitriol production (Kano & Jones 1984), therefore, if feline hyperthyroidism was associated with reduced plasma calcitriol concentrations, this could be a pathophysiological mechanism for the development of hypocalcaemia in hyperthyroidism. In this study, hyperthyroid cats had higher plasma calcitriol concentrations than healthy geriatric cats, and plasma total T4 was not correlated with plasma calcitriol concentrations. Therefore, hyperthyroidism appears to be associated with increased rather than decreased calcitriol production, and so reduced plasma calcitriol concentrations do not appear to be the pathophysiological mechanism for ionized hypocalcaemia or elevated PTH concentrations in hyperthyroid cats. Elevated PTH concentrations (>17·3 pg/mL) were present in the majority of hyperthyroid cats in the present study (75%, data not shown), as has been reported previously (Barber & Elliott 1996, Williams et al. 2012). PTH stimulates calcitriol production, therefore increased plasma calcitriol concentrations in hyperthyroid cats might occur secondary to increased plasma PTH concentrations. In addition, FGF-23 can inhibit 1α-hydroxylase activity and suppress calcitriol production, and plasma FGF-23 concentrations are reported to be decreased in hyperthyroid cats (Williams et al. 2012). Therefore, the increased plasma calcitriol concentrations in hyperthyroid cats might also be secondary to reduced plasma FGF-23 concentrations. In this study, plasma calcitriol concentrations were not significantly correlated with either plasma PTH or FGF-23 concentrations, however, these variables were only measured in a relatively small number of cats and so the study may have been statistically underpowered to detect significant correlations.

The dietary vitamin D intake was not reported as the cats included in this study were client-owned cats which were fed a range of different diets that are likely to contain different amounts of vitamin D. However, as cats cannot readily synthesize vitamin D in the skin, the measured total 25OHD concentration should correlate with the amount of dietary vitamin D that is absorbed in the intestine, and thus indicate the amount of circulating vitamin D substrate that is available for the production of calcitriol.

In the multi-variable linear regression analysis, plasma total T4 remained an independent predictor of ionized calcium concentration after adjustment for venous pH and plasma globulin concentrations. In addition, plasma total T4 remained an independent predictor of ionized calcium concentration after adjustment for plasma PTH and calcitriol concentrations. This could suggest that increased thyroid hormone concentrations can cause alterations to ionized calcium concentration either directly, or through another currently undetermined mechanism, such as altered renal or intestinal handling of calcium, which is independent of control by PTH or calcitriol. Thyroid hormones stimulate increased calcium uptake into rat liver cells (Hummerich & Soboll 1989), and increase the expression of the cardiac sarcoplasmic reticulum calcium ATPase (Periasamy et al. 2008, Limas 1978), thus increasing the amount of calcium stored in the sarcoplasmic reticulum. However, it is possible that chronic exposure to thyroid hormone (as would be the case in a hyperthyroid cat) would result in a saturation of the sarcoplasmic reticulum stores and hence this mechanism could only account for a temporary decrease in extracellular ionized calcium concentrations which would be normalized by homeostatic mechanisms (such as increased intestinal calcium absorption and increased renal calcium reabsorption) that are mediated by PTH and calcitriol. Thyroid hormones also act directly on the proximal tubular cells of the kidney to cause increased phosphate reabsorption and hyperphosphataemia (Yusufi et al. 1985, Alcalde et al. 1999). It is possible that thyroid hormones might also have a direct effect on the kidney to decrease renal calcium reabsorption, which could lead to hypocalcaemia. Human patients with Graves’ disease have elevated urinary calcium excretion, whereas human patients with toxic nodular goitre have decreased urinary calcium excretion (Czernobilsky et al. 1988), however, the patients included in the aforementioned study were not on a calcium and phosphorus controlled diet therefore the results should be interpreted with caution. Furthermore, previous studies have demonstrated increased activity of the Na+Ca2+ exchanger in the intestine and kidney of hyperthyroid rats (Kumar & Prasad 2002, 2003), however, it is possible that the effect of hyperthyroidism on the Na+Ca2+ exchanger in cats is different to rats (and perhaps humans). Such differences in calcium handling by the intestine and/or kidney could perhaps account for the differences in calcium homeostasis between hyperthyroid cats and human patients with Graves’ disease. Further studies to investigate intestinal and renal calcium handling in hyperthyroid cats are warranted to determine if altered renal or intestinal handling of calcium is part of the pathophysiological explanation for hypocalcaemia in hyperthyroid cats.

In summary, hyperthyroidism is associated with increased plasma calcitriol concentrations and decreased ionized calcium concentrations. Hyperthyroid cats with concurrent or masked CKD have higher ionized calcium concentrations compared to non-azotaemic hyperthyroid cats, thus hypocalcaemia in hyperthyroidism does not appear to be associated with calcitriol deficiency or the presence of concurrent or masked CKD. Plasma total T4 was independently associated with ionized calcium concentration after adjustment for other factors including plasma PTH and calcitriol concentration. This suggests that thyroid hormones can influence ionized calcium concentration either directly or through a mechanism that is yet to be determined but which is independent of control by PTH and calcitriol.

Acknowledgements

The authors would like to acknowledge BSAVA Petsavers who funded this study.

Conflict of interest

None of the authors of this article has a financial or personal relationship with other people or organizations that could inappropriately influence or bias the content of the paper.

Ancillary