Hypersomatotropism in diabetic cats in Australia

Abstract

Objectives

Hypersomatotropisim is an excessive production of growth hormone from the anterior pituitary gland, typically secondary to a pituitary tumour, which causes insulin-resistant diabetes and the clinical syndrome of acromegaly. Studies have shown the prevalence of hypersomatotropism among diabetic cats in the UK, Switzerland and the Netherlands to be in the range of 17.8–26%. The prevalence of hypersomatotropism in Australia is not known. Our objective was to determine the prevalence of hypersomatotropism in diabetic cats in Australia.

Methods

Residual serum samples from cats with increased fructosamine or increased blood glucose and a clinical history of diabetes were submitted for the measurement of insulin-like growth factor 1 (IGF-1). Hypersomatotropism was defined as an IGF-1 of 1000 ng/ml or more. The prevalence and associated confidence interval were calculated (Jeffrey’s method). Clinicopathological features between diabetic cats with and without hypersomatotropism were compared.

Results

Serum samples from 87 cats were included in the final analysis. IGF-1 was above 1000 ng/ml in 14 cats. The absolute prevalence of IGF-1 was 16%; therefore, the prevalence of hypersomatotropism (IGF-1 levels >1000 ng/ml) in an Australian population is estimated to be in the range of 9.5–24.9%. No significant difference was detected between breed (pedigree vs domestic), sex, age nor location (metropolitan vs regional) in cats with and without hypersomatotropism. Glucose and fructosamine concentrations did not differ between cats with and without hypersomatropism (P = 0.9 and P = 0.57, respectively).

Conclusions and relevance

Hypersomatotropism is an increasingly recognised condition in the feline population as a major contributor to uncontrolled diabetes mellitus. The prevalence of hypersomatotropism in Australian diabetic cats is 16%, which is similar to results from other countries. Clinical features cannot be used to distinguish diabetic cats with and without hypersomatotropism, so screening using a validated IGF-1 assay is necessary.

Introduction

Hypersomatotropism is defined as an excessive production of growth hormone from the anterior pituitary gland (adenohypophysis). Hypersomatropism, and the clinical syndrome that it causes (acromegaly), is almost always due to a growth hormone-producing pituitary tumour. ¹

Acromegaly usually results in characteristic clinical abnormalities, such as broad facial features, clubbed feet, stridor and insulin-resistant diabetes. ¹ If not diagnosed and treated, cats with hypersomatotropism often develop poorly controlled diabetes with less common manifestations, including progressive cardiomyopathy, pancreatopathy and neurological signs.^2,3

Most cats with hypersomatotropism do not display typical phenotypical acromegaly signs until the later stages of disease; therefore, increased clinician awareness and screening of diabetic cats for hypersomatotropism are important for early detection and treatment. ⁴

The pathophysiology and progression of acromegaly in cats is similar to that in humans, which means naturally occurring disease in cats provides an interesting model for research to improve both feline and human health. ²

Insulin-like growth factor 1 (IGF-1) is a small, single-chain polypeptide that is structurally like proinsulin. IGF-1 is produced by hepatocytes and regulated by growth hormone. It is increased in both cats and humans with hypersomatotropism. IGF-1 is also relatively stable, unlike growth hormones that are secreted in pulses, making single measurements difficult to interpret. ⁵ IGF-1 levels can be used to evaluate treated diabetic cats for the presence of hypersomatotropism. An IGF-1 concentration above 1000 ng/ml on a validated assay is strongly suggestive of hypersomatotropism. ⁶ Advanced imaging is recommended to confirm the diagnosis. The ideal treatment for humans is transphenoidal hypophysectomy. Similarly, the evidence suggests that hypophysectomy is the ideal treatment for cats, with two recent case series from the Royal Veterinary College and Utrecht University demonstrating high rates of endocrine cure and low mortality.^{7 –9}

An initial study in the UK estimated a prevalence of hypersomatotropism among diabetic cats in the range of 27–32%. ¹⁰ A subsequent, larger study (1221 cats) confirmed these preliminary results, with 26% of diabetic cats having IGF-1 results consistent with hypersomatotropism. ⁴ A similar study in Switzerland and the Netherlands found a slightly lower prevalence of hypersomatotropism in diabetic cats (17.8%). ¹¹ All of the above studies evaluated all diabetics, regardless of degree of diabetic control.

The prevalence of feline hypersomatotropism in Australia is not known. Although hypersomatotropism is more common in older male cats in studies in other countries, it is not known if similar trends are present in Australian cohorts. The diagnostic test of choice (validated IGF-1) is not available locally and is reasonably expensive. Additional diagnostic tests for hypersomatotropism, including CT and MRI, have limited availability, substantial cost and require general anaesthesia. Ideally, advanced imaging would only be performed in patients after a positive IGF-1 test. For these reasons, knowledge of the prevalence and features of diabetic cats with hypersomatotropism is important for the appropriate management of feline diabetes.

Hypersomatotropism is a common cause of insulin resistance, poor diabetic control and subsequent reduction in quality of life for diabetic cats in other countries. ³ However, Australian and New Zealand cats are genetically and environmentally isolated from European cats, and the prevalence of other diseases, particularly infectious diseases, has been very different in Australian and New Zealand cats than studies in other countries would predict. ¹² Knowledge of regional disease prevalence, the pretest probability, is important in informing the predictive value of a positive IGF-1 result for clinicians.

Our hypothesis was that the prevalence of hypersomatotropism in diabetic cats in Australia would be similar to that reported elsewhere, with an estimated prevalence of 20%. Our secondary hypothesis was that signalment and select pathological abnormalities would not differ significantly between diabetic cats with and without hypersomatotropism.

Materials and methods

Samples

As a prospective study using residual serum samples, cats did not undergo additional venipuncture, and specific ethics approval was not required.

Residual serum samples from cats with increased fructosamine or blood glucose submitted to IDEXX Veterinary Laboratory (Sydney) over 12 months were collected and de-identified, apart from date of birth, breed, sex, postcode and laboratory results. The samples were stored and submitted to the Nationwide Veterinary Services, Cambridge for IGF-1 analysis in a single batch. Samples were stored at –20°C because of the unavailability of a –80°C freezer and were transported frozen in a single batch. The stability of IGF-1 is considered good, with feline samples stored for more than 9 years at –80°C showing stability. ¹³

Inclusion criteria

Serum samples were only included from cats with any of the following: (1) fructosamine above 286 mmol/l (IDEXX reference interval, updated 2019 [https://www.idexx.fi/fi/veterinary/reference-laboratories/reference-laboratory-support/updates-canine-and-feline-fructosamine-reference-intervals]); (2) fructosamine below 286 mmol/and a clinical history of diabetes mellitus on laboratory submission; or (3) blood glucose above 15 mmol/l and a clinical history of diabetes mellitus on laboratory submission.

IGF-1 measurement was performed on an assay validated for cats as previously described. ⁶

Statistical analysis

Hypersomatotropism was defined as an IGF-1 of 1000 ng/ml or above. The prevalence and associated confidence interval were calculated (Jeffrey’s method). Continuous data were assessed for normality (Shapiro–Wilk test) and non-parametric statistics used where indicated.

Clinicopathological features between diabetic cats with and without hypersomatotropism were compared using Mann–Whitney U-tests (quantitative data) and Fisher’s exact test (categorical data). Correlations between IGF-1 and fructosamine were made (Spearman’s correlation).

Power analysis

Using an estimated prevalence of 20%, a test sensitivity of 100% and test specificity of 88%, ⁶ 104 samples would be required to estimate the prevalence with a confidence of 95% and a precision of 0.1 (https://epitools.ausvet.com.au/prevalencess).

Results

A total of 140 samples were available for analysis. For cats with the same name, date of birth and postcode, records were compared to identify multiple submissions from the same cat; in such cases, only the highest recorded IGF-1 result was included and duplicate results were excluded (n = 18). Samples with insufficient material (n = 4) or incomplete history (n = 40) were excluded, leaving a total of 87 samples from individual cats for the final analysis.

Of the samples, 75 were from cats with increased fructosamine concentrations. One sample was from a cat with normal fructosamine but a clinical history of diabetes mellitus. In addition, there were 11 samples from cats with blood glucose above 15 mmol/l and a clinical history of diabetes mellitus.

Information regarding sex was recorded in 86 cats. There were 33 spayed females, one entire male and 53 castrated males. Age was estimated from date of birth and date of submission in 84/87 cats. The median age was 12 years (interquartile range [IQR] 9.5–14). Breed information was available for 64 cats. Most cats were listed as domestic (n = 54) with 10 pedigrees (five Burmese, two Devon Rex and one Bombay, Maine Coone and Russian Blue each). The submitting postcode was available for all cats. Most cats (n = 66) were from a postcode within the greater Sydney metropolitan area, with 21 cats from postcodes within New South Wales outside this area.

IGF-1 levels were above 1000 ng/ml in 14 cats. Three cats had an IGF-1 level above 2000 ng/ml, which was recorded as 2000 ng/ml for statistical analysis. A further 11 cats had IGF-1 levels above 700 ng/ml but below 1000 ng/ml. The remaining 62 cats had IGF-1 levels below 700 ng/ml. The absolute prevalence of IGF-1 above 1000 ng/ml was 16%; therefore, the prevalence of hypersomatotropism (IGF-1 levels >1000 ng/ml) in an Australian population is estimated to be in the range of 9.5–24.9% (95% confidence interval).

Cats with hypersomatotropsim had a similar age (median 11 years, IQR 7–12) to cats with diabetes mellitus alone (median 12 years, IQR 10–14; P = 0.16). The proportion of pedigree and domestic and male and female cats did not differ between cats with and without hypersomatotropism (P = 0.19 and P = 0.38, respectively). Similarly, no significant difference was detected between location (metropolitan vs regional) in cats with and without hypersomatotropism (P = 1). Pathological features, specifically glucose (median 21.4 and 21.9 mmol/l) and fructosamine (median 443 and 452 µmol/l) concentrations did not differ between cats with and without hypersomatropism (P = 0.9 and p = 0.57, respectively).

There was a weak, negative correlation (Spearman’s rho –0.21) between fructosamine and IGF-1 concentrations (P = 0.068).

Discussion

Hypersomatotropism is an important cause of diabetes mellitus in the Australian cat population, with a prevalence of 16% among a broad population of diabetic cats managed in all clinical settings. The prevalence of 16% is lower than reports in the UK, but similar to the estimated prevalence of 17.8% in Switzerland and the Netherlands, and marginally higher than the estimated prevalence of 14.9% in a referral population in Argentina.^4,11,14 The genetic pool of Australian cats is potentially different to European cats given the geographical distance. The findings of this study suggest that hypersomatotropism should be considered a cause of diabetes mellitus in cats worldwide. The estimated prevalence indicates a relatively significant incidence of hypersomatotropism as a cause of diabetes in Australian cats. Given that this disease can lead to more extensive complications, including reversible but life-limiting cardiomyopathy, which may occur before overt insulin-resistance diabetes, and that definitive treatment via hypophysectomy is available in Australia, it would be ideal to screen all diabetic cats for hypersomatotropism. ¹⁵ Early diagnosis allows owners to make informed decisions about treatment options and provides the potential to improve long-term outcomes through timely definitive treatment. The main disadvantage of screening for hypersomatotropism is the cost of IGF-1 measurement, as samples must be shipped overseas; however, this limitation is at the discretion of the owner rather than a medical disadvantage of screening.

Given these results, ideally all diabetic cats in Australia should be screened for hypersomatotropism, with IGF-1 measurement performed 6–8 weeks after the initiation of insulin therapy.

Although the present study included more male than female diabetic cats, no significant difference was found in the prevalence of hypersomatotropism between the sexes. This observation may be attributable to the limited sample size; it is possible that a larger cohort of cats with hypersomatotropism could reveal a predisposition associated with male sex, as has been suggested in previous studies.^4,14

One limitation of this study is that the treatment status (insulin or newly diagnosed) of most cats at the time of sample submission was unknown. This could have resulted in under-detection of hypersomatotropism in this population. It is possible that some of these cats, especially those with IGF-1 above 700 ng/ml but below 1000 ng/ml, may have hypersomatotropism, as IGF-1 is reduced in untreated diabetic cats. Retesting after initiation of insulin therapy is recommended for new diabetic cats with IGF-1 above 700 ng/ml.^5,6 The highest recorded IGF-1 result was included for cats with multiple samples; however, it remains possible that the prevalence of hypersomatotropism in Australian diabetic cats could be higher than that reported in this study.

Unfortunately, after excluding samples from animals with inadequate confirmation of a diagnosis of diabetes mellitus, the study’s sample size was smaller than required to give the preferred precision, resulting in a wider confidence interval, which is the statistic that should be applied when extrapolating findings to the larger population of cats. Ideally, this study would be performed with a larger number of samples. Recruitment of samples, however, was not biased towards cats suspected of having hypersomatotropism, nor a referral population that would be typically biased towards poorly controlled diabetics, as all viable residual samples from a reference laboratory were analysed. Therefore, this estimated prevalence is drawn from a reasonably representative sample of the Australian diabetic cat population.

A final limitation of this study is that hypersomatotropism was not confirmed with advanced imaging, and it is possible the prevalence may be overestimated when using a cutoff of 1000 ng/ml. ¹⁶ Our samples were tested at the same laboratory where results were previously validated, with a cutoff of above 1000 ng/ml showing a 95% positive predictive value in cats with follow-up imaging. ⁴

There was no significant correlation between IGF-1 and fructosamine, similar to previous findings.^4,5,11

The pathogenesis of hypersomatotropism in cats is hypothesised to be similar to humans, with a recent study looking at environmental toxin exposure in acromegalic cats as a model for human disease. ¹⁷ This study found acromegalic cats had detectable organohalogenated compounds in their plasma and hypothesised that cats in households in highly built-up areas may be at higher risk of hypersomatotropism. ¹⁷ Most cats in our study were from the Sydney metropolitan area, without a significant difference in location (metropolitan vs regional) between cats with and without hypersomatotropism. This epidemiological observation does not strongly support this theory but, as there were smaller numbers of cats from regional areas overall, the study was not likely to be adequately powered to refute this hypothesis. The aryl hydrocarbon receptor-interacting protein (AIP) mutation has also been identified as a cause of hypersomatotropism in humans and a potential cause in cats, with the mutation identified in 4/16 cats with hypersomatotropism in a small initial sequencing study. ¹⁸ Further larger studies are required to determine if this mutation is a significant cause of hypersomatotropism in cats as well as its hypothesised relationship to toxin exposure due to the AIP connection to the aryl hydrocarbon receptor pathway, which is involved in environmental toxin response. ¹⁹

Conclusions

Hypersomatotropism is an increasingly recognised condition in the feline population and a major contributor to poorly controlled diabetes mellitus. In this study, the prevalence of hypersomatotropism among Australian diabetic cats was 16%. No significant differences in signalment or fructosamine levels were identified between diabetic cats with or without hypersomatotropism.