SEARCH

SEARCH BY CITATION

Keywords:

  • horse;
  • geriatric;
  • endocrine;
  • equine Cushing's syndrome;
  • epidemiology

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

Reasons for performing study: The sensitivity and specificity of basal plasma α-melanocyte-stimulating hormone (α-MSH) and adrenocorticotrophic hormone (ACTH) for the diagnosis of pituitary pars intermedia dysfunction (PPID) has not been evaluated in a population-based study.

Objectives: To evaluate basal plasma α-MSH and ACTH concentrations for the diagnosis of PPID in a population of horses aged ≥15 years.

Methods: Owner-reported data were obtained using a postal questionnaire distributed to an equestrian group. A subgroup of surveyed owners was visited and veterinary examination performed on horses aged ≥15 years. Blood samples were analysed for plasma α-MSH and ACTH concentrations. Seasonally adjusted cut-off values for α-MSH and ACTH concentrations for the diagnosis of PPID were obtained using Youden index values against a clinical gold standard diagnosis (hirsutism plus 3 or more clinical signs of PPID).

Results: α-melanocyte-stimulating hormone and ACTH were highly correlated with each other and with clinical and historical indicators of PPID. The increase in both α-MSH and ACTH with increasing numbers of clinical signs in affected horses supports a spectrum of disease. Both variables were affected by season, with derived cut-off values being higher in autumn compared with other seasons. Sensitivity and specificity were moderate and good in nonautumn seasons (59 and 93%, respectively) for α-MSH using a cut-off of 52.0 pmol/l. Sensitivity and specificity were good in nonautumn seasons (80 and 83%, respectively) for ACTH using a cut-off of 29.7 pg/ml. For both α-MSH and ACTH, sensitivity and specificity were close to 100% for samples obtained during the autumn period.

Conclusions and potential relevance: Basal plasma α-MSH and ACTH had moderate-to-good sensitivity and specificity for the diagnosis of PPID, which improved substantially during the autumn period, suggesting this may be the ideal time to test. Further studies to develop seasonally adjusted reference intervals for different geographical locations are warranted.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

Equine pituitary pars intermedia dysfunction (PPID) is a neurodegenerative disorder of ageing, usually affecting horses aged 15 years and older [1]. Hirsutism or hypertrichosis, which can manifest as increased length and/or delayed shedding of all or part of the haircoat, is the most commonly observed clinical sign [2,3]. Other clinical signs include laminitis, weight redistribution, polyuria/polydipsia, hyperhidrosis, susceptibility to infections and lethargy or depression [2,3].

A number of endocrine tests have been used in the diagnosis of PPID, including basal and dynamic tests. Dynamic tests include the dexamethasone suppression test [4], the combined dexamethasone suppression thyrotrophin-releasing hormone (TRH) response test [5] and the α-melanocyte-stimulating hormone (α-MSH) or adrenocorticotrophic hormone (ACTH) response to TRH stimulation tests [6,7]. Basal endocrine tests have the advantage of simplicity and the ability to be used in epidemiological studies involving large numbers of horses. Basal plasma α-MSH [8] and ACTH [9,10] concentration have been reported to have good sensitivity and specificity for PPID in small, selected groups of horses.

Basal plasma α-MSH and ACTH have both been shown to be affected by season, with elevations and potential false-positive diagnoses in the autumn [11,12]. More recently, seasonally adjusted median reference values of ACTH have been determined from laboratory submissions [13]. To date, however, studies have been limited to small and selected groups of horses with PPID and control animals or laboratory submissions with incomplete clinical data. The determination of the sensitivity and specificity of both basal plasma α-MSH and basal plasma ACTH in a population of aged horses ≥15 years demonstrating a spectrum of clinical PPID has not been evaluated.

In order to evaluate the sensitivity and specificity of a test, a gold standard is required. There is a lack of an ante mortem laboratory ‘gold standard’ for PPID, and clinical signs have remained central to diagnosis of PPID. A commonly utilised ‘gold standard’ for demonstration of disease, post mortem histological examination of the pituitary, is not ideal to detect horses with PPID because the histopathological criteria for clinical disease (especially considering that the disease is progressive with ageing) have not been well elucidated [14]. Lesions in the pars intermedia indicative of PPID have been observed in over a third of normal horses [14]. Furthermore, in horses with clinical signs of PPID, there was no significant correlation between relative pituitary weight and plasma ACTH, cortisol or insulin concentrations [14]. Of further concern is a lack of agreement between veterinary pathologists when evaluating equine pituitary glands, again indicating that histopathological diagnosis may not be an appropriate gold standard for diagnosis [15].

One of the most specific indicators of PPID is the presence of hirsutism, which has been suggested to be pathognomonic for PPID [10]. Researchers have used numerous criteria for ante mortem diagnosis; most include hirsutism and then at least one other clinical sign [6,9,10,16]. Hirsutism alone as a diagnostic test against post mortem diagnosis has been shown to have a very high specificity (95%) but only a moderately good sensitivity (71%) in a small group of selected horses [5]. In that study, hirsutism alone had a greater accuracy than the combined dexamethasone suppression and TRH stimulation test or either of its components [5]. Detection of hirsutism on veterinary clinical examination can be difficult if the horse has been clipped, or if it is autumn or winter when hair coats are seasonally longer, or if the horse is only demonstrating delayed shedding and is examined once shedding has occurred. For this reason, owner-reported hirsutism may be more sensitive, because owners have an historical and seasonal perspective instead of a single examination. Therefore, owner-reported hirsutism (defined as observation of delayed shedding and/or failure to shed and/or a long haircoat), combined with other signs of PPID reported in the history or detected on veterinary clinical examination, was considered the most appropriate clinical ‘gold standard’ for diagnosis of PPID in a population-based study of client-owned horses.

The aim of this study was to evaluate basal plasma α-MSH and ACTH concentrations for the diagnosis of PPID in a population of horses aged 15 years and older in Southeast Queensland (SE QLD), Australia. Our objectives were to determine the relationship of basal plasma α-MSH and ACTH to clinical signs of PPID in aged horses, to determine the correlation between basal plasma α-MSH and ACTH and to develop seasonally adjusted cut-off values for basal plasma α-MSH and ACTH concentrations for the diagnosis of PPID by testing values obtained against a clinical gold standard diagnosis of hirsutism plus 3 or more clinical or historical signs of PPID.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

Ethical approval was obtained by the University of Queensland Ethical Review Committee and Animal Ethics Committee.

Study 1 (survey data) consisted of a questionnaire designed to obtain data on horses aged ≥15 years in SE QLD. Details of the selection criteria have been provided elsewhere [17]. In brief, the questionnaire was distributed via mail together with a newsletter of the Queensland Equestrian Federation of Australia (EQ; 2300 members). Members were asked to complete a questionnaire for each horse aged 15 years or older. Data were collected on horse details, history, owner-reported clinical signs, management practices and owners' welfare concerns of aged horses. A copy of the questionnaire is available as Supplementary Item 1.

A history of owner-reported hirsutism was determined on the basis of their responses to a question concerning hair coat changes detected in their horse and was considered to be present if the owner had indicated that their horse had shown signs of a long hair coat and/or delayed shedding and/or failure to shed over the past 12 months. A further question asked ‘Does your horse suffer from any definite disease or disorder that you know of?’, where owners answering ‘PPID, equine Cushing's syndrome’ or synonyms were counted as a previous diagnosis of PPID.

Study 2 (clinical data) consisted of a subsample (n = 340) of the population in Study 1 and was restricted to horses owned by members of Equestrian Queensland (EQ) who resided within defined geographical regions of SE QLD. These regions were Brisbane City, Gold Coast and hinterlands, Sunshine Coast and hinterlands, Ipswich and Lockyer Valley, Toowoomba and surrounds, and Warwick and surrounds. These areas ranged in latitude from -27° 29′ to -24° 54′, with the maximal change in day length between the summer and winter solstices being 3 h and 30 min. Average temperatures (minimum–maximum) in summer (December, January and February) in Brisbane are 20–28°C, autumn (March, April and May) 15–25°C, winter (June, July, August) 11–21°C and spring (September, October, November) 15–25°C.

Owners who indicated willingness to participate further in the study were contacted and a visit arranged. During each visit, all horses aged ≥15 years owned by the EQ member underwent a complete clinical examination and blood sample collection by a qualified veterinary surgeon. After obtaining owner consent, blood collection was performed, via jugular venipuncture, prior to moving the horse or commencing the clinical examination. This was done to minimise stress or excitement. Blood was collected into a 10 ml glass EDTA vacutainer tubea, then immediately centrifuged at 1000 g for 10 min using a portable centrifuge. The plasma was harvested via pipette into 1.5 ml Eppendorf™ tubes, and placed on ice packs before storage at -80°C (within 6 h) until being shipped frozen to the laboratory (1–24 months). Blood samples were analysed for plasma α-MSH (Euria alpha-MSH RIA Kit)b and ACTH (Immulite 1000 assay)c concentrations.

Data analysis

Plasma α-MSH and ACTH concentrations of all horses (n = 325) were plotted in numerical order to examine the population distribution of values. The plasma concentrations of α-MSH and ACTH were transformed, using natural logarithms for both, to create a normal distribution. Pearson's correlation coefficient and linear regression were used to assess relationships between loge plasma α-MSH and ACTH concentrations. Loge plasma α-MSH and ACTH concentrations were compared between horses and ponies using Student's unpaired t test.

Loge plasma α-MSH and ACTH concentrations of all horses were initially compared over season using a one-way ANOVA to determine whether values were significantly affected by season. Post hoc analysis using Bonferroni correction was performed.

One-way ANOVA and post hoc analysis using Bonferroni correction were used to compare mean loge plasma α-MSH and ACTH concentrations for horses with owner-reported hirsutism with zero to 12 clinical signs or historical factors to determine at what point the means of ‘affected’ and ‘unaffected’ horses were significantly different in order to set the clinical gold standard. For both loge plasma α-MSH and loge plasma ACTH, this occurred when owner-reported hirsutism was combined with 3 or more clinical signs (P<0.001).

A linear regression model was used to compare the number of clinical and historical signs of PPID demonstrated in an individual horse and the loge value of α-MSH and ACTH concentration measured in that horse. Student's unpaired t tests were used to compare the means of each test against commonly occurring clinical signs, history and subject details.

The clinical gold standard was therefore defined as owner-reported history of hirsutism (where hirsutism includes excessive long hair growth and/or delayed shedding and/or failure to shed) plus at least 3 other clinical or historical signs of PPID (hirsutism detected on clinical examination, increased supraorbital fat, wasted top line, pot belly, laminitis, divergent rings on hooves, seedy toe/separation of the white line, dropped soles, polyuria/polydipsia, depression/lethargy, owner-reported previous diagnosis of PPID). Loge plasma α-MSH and ACTH values were then evaluated against the clinical gold standard. Horses with hirsutism plus zero, one or 2 clinical signs were removed from analysis (n = 21). Sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios and Youden index values were derived using specific diagnostic test statistical software (MedCalc version 11.5.1.0)d. The value derived using the Youden index was used as the cut-off value [18]. These values were determined during the autumn months (March, April and May) and all other months excluding autumn separately to give autumn and nonautumn or seasonally adjusted reference intervals.

Unless specifically indicated, the term ‘horses’ is used in this article to indicate collectively all horses and ponies.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

Study 1 (survey data)

Completed questionnaires were received from a total of 536 QLD horse-owners who owned a total 2873 horses. Of these, 1089 (37.9%) horses were ≥15 years of age, and useable questionnaires were obtained for 89.4% (974) horses.

Study 2 (clinical data)

Veterinary clinical examination was performed on 340 of the 974 aged horses from the survey group (35%). Blood samples were obtained from 339 horses. Plasma α-MSH was measured on all 339 horses; however, plasma ACTH was not available for 14 horses, leaving 325 horses with both analyses completed. The frequency of owner-reported clinical signs and veterinary clinical examination findings typically associated with PPID are presented in Tables 1 and 2. None of the horses for which there was a previous diagnosis of PPID was receiving medical treatment for PPID.

Table 1.  Frequency of owner-reported history of clinical signs that may be associated with pituitary pars intermedia dysfunction in 974 horses aged ≥15 years
Study 1
HistoryPercentage95% confidence interval
Weight loss16.8 (164 of 974)14.5–19.2
Hirsutism16.7 (163 of 974)14.4–19.1
Depression7.6 (74 of 973)5.9–9.3
Laminitis4.8 (47 of 974)3.5–6.2
Polyuria3.7 (36 of 974)2.5–4.9
Polydypsia3.4 (33 of 974)2.3–4.5
Table 2.  Frequency of veterinary clinical examination findings that may be associated with pituitary pars intermedia dysfunction in 340 horses aged ≥15 years
Study 2
Clinical examination findingsPercentage95% confidence interval
Wasted top line32.1 (109 of 340)27.1–37.0
Pot belly18.8 (64 of 340)14.7–23.0
Seedy toe (separation of the white line)14.2 (46 of 323)10.4–18.1
Hirsutism13.8 (47 of 340)10.2–17.5
Supraorbital fat10.0 (34 of 340)6.8–13.2
Dropped sole5.6 (19 of 340)3.1–8.0
Divergent rings5.0 (17 of 340)2.7–7.3
Lame in all 4 legs2.1 (7 of 340)0.6–3.6

Graphs of the plasma concentrations of α-MSH and ACTH in the study population revealed an exponential distribution in α-MSH and ACTH concentrations. There was a good correlation between the plasma concentrations of logeα-MSH and loge ACTH for all horses that had both tests (n = 325), with a Pearson correlation coefficient of 0.79 (P<0.001; Fig 1). There were no significant differences between ponies (63 of 339) and horses (276 of 339) for logeα-MSH (P = 0.07). Likewise, there were no differences between ponies (57 of 325) and horses (268 of 325) for loge ACTH (P = 0.22). However, ponies were more likely to show more clinical signs of PPID (median 3, interquartile range 1–3.5) than horses (median 1, interquartile range 0–3; P<0.05).

image

Figure 1. Correlation between logeα-melanocyte-stimulating hormone (α-MSH) vs. loge adrenocorticotrophic hormone (ACTH) for 325 horses aged ≥15 years in Southeast Queensland. Pearson's r = 0.79 (P<0.001).

Download figure to PowerPoint

The total number of clinical and historical signs of PPID displayed in horses with owner-reported hirsutism (48 of 339 for α-MSH and 44 of 325 for ACTH) was significantly correlated to the loge value of α-MSH (r = 0.63, P<0.001) and ACTH (r = 0.69, P<0.001; Fig 2a, b). Historical signs of PPID that were associated with a significantly (P<0.05) higher logeα-MSH and/or ACTH concentration included hirsutism, a veterinary diagnosis of PPID (or equine Cushing's syndrome), laminitis, depression and polyuria and polydipsia (Fig 3a, b). Signs of PPID detected on veterinary clinical examination that were associated with a significantly (P<0.05) higher logeα-MSH and/or ACTH included hirsutism, wasted top line, pot belly, divergent rings on the hoof wall, presence of supraorbital fat and seedy toe (separation of the white line; Fig 4a, b). When mean logeα-MSH and loge ACTH concentrations from horses with clinical or historical signs of PPID were compared with those of horses without the clinical or historical signs of PPID, there was very little overlap between 95% confidence intervals. Confidence intervals were narrow in horses with no clinical or historical signs (Figs 3, 4).

image

Figure 2. a) Simple scatter plot of the linear regression model of the natural log of α-melanocytes-stimulating hormone (α-MSH) and the total number of clinical and historical signs of pituitary pars intermedia dysfunction (PPID) displayed by 48 of 339 horses with regression line of best fit (r = 0.63, P<0.001). Owner-reported hirsutism was the inclusion criterion (if no further signs were displayed, total signs = 1). b) Simple scatter plot of the linear regression model of the natural log of adrenocorticotrophic hormone (ACTH) and the total number of clinical and historical signs of PPID displayed by 44 of 325 horses with regression line of best fit (r = 0.69, P<0.001). Owner-reported hirsutism was the inclusion criterion (if no further signs were displayed, total signs = 1).

Download figure to PowerPoint

image

Figure 3. a) Mean of the loge of the concentrations of α-melanocytes-stimulating hormone (α-MSH) (error bars represent the 95% confidence intervals) for horses with and without the owner-reported history of the clinical signs or having been previously diagnosed with pituitary pars intermedia dysfunction (PPID). The numbers of horses with and without a history were as follows: clinical signs of depression (30 and 309; P<0.01), polydypsia (10 and 330; P<0.001), polyuria (12 and 328; P<0.001), laminitis (21 and 319; P<0.001), hirsutism (48 and 292; P<0.001) and a history of previous diagnosis of equine Cushing's syndrome (ECS; 10 and 330; P<0.001; in this instance, ECS and PPID are synonymous). b) Mean of the loge of the concentrations of adrenocorticotrophic hormone (ACTH) (error bars represent the 95% confidence intervals) for horses with and without the owner-reported history of the clinical signs or having been previously diagnosed with PPID (P<0.001 for each sign). The numbers of horses with and without a history of clinical signs were as follows: depression (29 and 295), polydypsia (8 and 317), polyuria (10 and 315), laminitis (17 and 308), hirsutism (44 and 218) and a history of previous diagnosis of ECS (8 and 317; in this instance, ECS and PPID are synonymous).

Download figure to PowerPoint

image

Figure 4. a) Mean of the loge of the concentrations of α-melanocytes-stimulating hormone (α-MSH) (error bars represent the 95% confidence intervals) for horses with and without the clinical signs found on clinical examination. The numbers of horses with and without a clinical signs were as follows: hirsutism (48 and 292; P<0.001), divergent rings (17 and 322; P = 0.001), seedy toe (56 and 283; P<0.01), supraorbital fat (34 and 305; P<0.002), pot belly (64 and 275; P<0.01) and wasted top line (109 and 230; P<0.02). b) Mean of the loge of the concentrations of adrenocorticotrophic hormone (ACTH) (error bars represent the 95% confidence intervals) for horses with and without the clinical signs found on clinical examination. The numbers of horses with and without a clinical signs were as follows: hirsutism (43 and 282; P<0.01), divergent rings (14 and 310; P<0.01), seedy toe (54 and 270; P<0.001), supraorbital fat (29 and 295; P<0.01), pot belly (59 and 265; P<0.02) and wasted top line (103 and 221; P<0.001).

Download figure to PowerPoint

Seventy-five horses were tested in the autumn, 80 in the winter, 113 in the spring and 71 in the summer. Loge plasma α-MSH and ACTH concentrations were significantly greater in the months March, April and May compared with other months (P<0.05) and therefore these 3 months were analysed together collectively as ‘autumn’. Loge plasma α-MSH and ACTH concentrations were both significantly greater in the autumn period compared with spring, summer and winter (P<0.01; Fig 5).

image

Figure 5. Natural log basal plasma α-melanocytes-stimulating hormone (α-MSH) a) and adrenocorticotrophic hormone (ACTH) concentrations b) in spring, summer, autumn and winter for 27 horses with owner-reported hirsutism and 3 or more signs of pituitary pars intermedia dysfunction (PPID) and for 291 age-matched horses without hirsutism within the same subpopulation.

Download figure to PowerPoint

Twenty-seven horses had an owner-reported history of hirsutism and at least 3 other clinical or historical signs of PPID. A further 21 horses had a history of owner-reported hirsutism with zero to 2 additional signs and were excluded from the analysis, leaving 291 horses with no clinical evidence of PPID and 318 horses remaining in the analysis to determine the Youden index for α-MSH and 304 horses for ACTH. Based on the overall prevalence of 14% of owner-reported hirsutism, assuming some inaccuracy, a prevalence of 13% was used as an estimate to determine positive and negative predictive values (Table 3).

Table 3.  Youden index derived seasonal cut-off values for plasma α-melanocyte-stimulating hormone (α-MSH) and adrenocorticotropin (ACTH) concentrations for the diagnosis of pituitary pars intermedia dysfunction (PPID)
VariableLoge α-MSHLoge α-MSHLoge ACTHLoge ACTH
  1. Representing the total sample size (n = 318 for logeα-MSH and n = 304 for loge ACTH) divided by cases (positive group = hirsutism and 3 or more signs) and controls (negative group = no owner-reported hirsutism) and by season (autumn only vs. other seasons). The derived values include calculated area under the receiver operating characteristic (ROC) curve with standard error, 95% confidence interval, z statistic and P value. The criterion point (cut-off) is shown for loge of both ACTH and α-MSH for horses tested during the autumn and for horses tested in seasons other than autumn. The criterion point corresponds to the maximum of the Youden index (J= max [SEi+ SPi− 1], where SEi and SPi are the sensitivity and specificity over all possible threshold values and correspond to the point on the ROC curve furthest from the diagonal line). Sensitivity (95% confidence interval) and specificity (95% confidence interval), as well as positive and negative likelihood ratios and predictive values, are shown only for the criterion point (cut-off). Estimated prevalence of PPID is 13%.

SeasonSpring, summer, winterAutumnSpring, summer, winterAutumn
Sample size2496923767
Positive group225204
Negative group2276421763
Area under the ROC curve (95% confidence interval)0.79 (0.74–0.84)0.98 (0.92–1.00)0.85 (0.80–0.89)0.98 (0.91–1.00)
Standard error0.0630.0140.060.017
z statistic4.6635.226.3628.73
Significance level P (area = 0.5)<0.0001<0.0001<0.0001<0.0001
Criterion (cut-off) at maximum Youden index loge (inverse loge pmol/l and pg/ml for α-MSH and ACTH respectively) >3.95 (52.0 ) >5.11 (165.4) >3.39 (29.7) >4.35 (77.4)
Sensitivity (95% confidence interval)59.09 (36.4–79.3)100 (47.8–100.0)80 (56.3–94.3)100 (39.8–100.0)
Specificity (95% confidence interval)92.95 (88.8–95.9)96.87 (89.2–99.6)82.49 (76.8–87.3)95.24 (86.7–99.0)
Positive likelihood ratio8.38324.5721
Negative likelihood ratio0.4400.240
Positive predictive value55.682.740.675.8
Negative predictive value93.810096.5100

Youden index derived cut-off values for nonautumn seasons were 52.0 pmol/l for α-MSH and 29.7 pg/ml for ACTH. When autumn was evaluated separately, these values rose to 165.4 pmol/l for α-MSH and 77.4 pg/ml for ACTH. Sensitivity, specificity, likelihood ratios and positive and negative predictive values for a prevalence estimate of 13% are shown in Table 3.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

This study is the first to examine a combination of basal plasma α-MSH and ACTH concentrations, owner-reported historical signs and veterinary clinical examination findings in a population-based group of aged horses. Such a group is more likely to represent the clinical spectrum of PPID in an age-matched group, compared with previous studies, which have relied on small subgroups of advanced PPID cases [5,7,19]. It has been suggested that PPID is a progressive syndrome associated with ageing and that affected horses may fall across a spectrum of disease [1]. The implication is that there is likely to be a gradual increase in both α-MSH and ACTH concentrations in PPID-affected horses, corresponding to an increase in clinical signs as the disease progresses. This was supported by the positive correlation between both logeα-MSH and loge ACTH and the number of historical or clinical signs demonstrated by affected horses in this study. That is, an increase in the number of clinical or historical signs was associated with increasing values for basal plasma α-MSH and ACTH.

Plasma α-MSH and ACTH concentrations were highly correlated and, despite previous reports of higher values in horses than in ponies [9], there were no differences between ponies and horses for either variable in this study. As plasma concentrations of α-MSH are less affected by extraneous factors, such as stress or diurnal variation [11], this might be expected to be a better screening endocrine test when compared with plasma ACTH concentration. However, the present results do not support its superiority over plasma ACTH concentration for screening for PPID. There were some discrepancies between positive horses on one and not the other test. However, these involved values close to the cut-off in almost all cases, highlighting a probable ‘grey zone’. This grey zone reinforces the contention that a combination of historical report, veterinary clinical examination findings and endocrinological tests provides the best method for diagnosing PPID.

Pituitary pars intermedia dysfunction manifests in different ways in individual affected animals [2,20,21]. Individual historical and clinical examination findings typically associated with PPID were associated with a significantly higher mean logeα-MSH and/or ACTH concentration. However, there was some overlap in confidence intervals for some individual signs. This is likely to be due to the nonspecific nature of many of the clinical signs. Furthermore, many of the signs are related to each other, e.g. hoof changes consistent with laminitis and a history of laminitis. A combination of owner-reported hirsutism and 3 or more signs was identified using ANOVA as representing significantly different logeα-MSH and/or ACTH concentrations between affected and unaffected groups and supports the use of such a clinical gold standard with which to evaluate basal plasma α-MSH and ACTH concentrations.

The Youden index (J), a summary measure of the receiver operating characteristic, has become more frequently used over the past decade in the biomedical field to establish a cut-off point, where patients who have a test value on one side are considered to be diseased and patients with a test value on the other side are considered to be disease free. The Youden index provides a cut-off point that maximises the test's ability to differentiate positive and negative patients, with equal importance placed on sensitivity and specificity. This cut-off point does not take into consideration prevalence of the disease or misclassification costs and as such also does not take a ‘grey zone’ into consideration [18,22,23]. By evaluation against a clinical gold standard, we were able to establish sensitivity, specificity and cut-off values for both plasma α-MSH and plasma ACTH in this sample of horses.

In evaluation of a biomarker to determine whether a subject is diseased or not, there also needs to be a diagnosis by a separate criterion, unrelated to the biomarker under evaluation. Despite this being a population-based study, and the spectrum of disease represented, it was necessary to create an ante mortem clinical gold standard, which represented more definite disease, to determine cut-off values for plasma α-MSH and ACTH concentrations for diagnosis of PPID. By using the clinical gold standard of owner-reported hirsutism plus 3 or more clinical or historical signs, we have selected for more advanced disease and, as such, the cut-off values represented may under-represent earlier cases of PPID. However, this is a situation common to all studies of PPID and is unlikely to be improved upon until an early biomarker is validated.

Plasma ACTH concentration has previously been shown to be a sensitive [9,19] and specific [10] indicator of PPID. Sensitivity and specificity of ACTH have been reported as 84 and 78%, respectively, against hirsutism only [10] using a cut-off of >35 pg/ml; 90.9–81.8% and 100%, respectively, against hirsutism and one other sign [9] using a cut-off of >27 pg/ml for ponies and >50 pg/ml for horses; and 84 and 89.5%, respectively, against a dexamethasone suppression test [8] using a cut-off of >50 pg/ml. Basal plasma α-MSH sensitivity and specificity have been reported as 88 and 84%, respectively, against a dexamethasone suppression test using a cut-off of >59.6 pmol/l. The major drawback of both basal plasma α-MSH [11] and basal plasma ACTH [12] concentrations has been reported to be the marked seasonal variation in values obtained from horses sampled in the Northern hemisphere. Despite SE QLD, Australia, having more limited seasonal changes in photoperiod or temperature, it was apparent that there was an autumn (March, April and May) rise in both plasma α-MSH and plasma ACTH concentrations. In areas further from the equator in the Northern hemisphere, this peak can occur earlier than in regions closer to the equator, and in Finland the peak will come as early as late summer (August) [24]. No evidence of a summer or February peak was found in this population, which is consistent with a location much closer to the equator [24], albeit in the Southern hemisphere. Owing to the seasonal differences, it was necessary to establish different cut-off values for autumn and the other 3 seasons.

Other studies examining potential cut-off values used either different groups [9,13] or laboratory submissions lacking consistent clinical inclusion criteria [10,13,25]. Studies using laboratory submissions have the disadvantage that the main inclusion criterion is also the variable being studied. Despite this, the cut-off values found in this study compare well with previous reports. The cut-off for plasma ACTH concentration (29.7 pg/ml) for seasons excluding autumn, with a sensitivity of 80% and specificity of 83%, was identical to the value derived from laboratory submissions by Copas and Durham [13]. Likewise, the cut-off for α-MSH (52.0 pmol/l) for nonautumn seasons, with a sensitivity of 59% and specificity of 93%, was only slightly lower than that calculated by Horowitz et al. [8], although season had not been accounted for in their study. When translated into likelihood ratios, α-MSH performed better than ACTH for ruling in disease (likelihood ratio 8.4 vs. 4.6, ideal >10), while α-MSH performed worse than ACTH for ruling out disease (likelihood ratio 0.4 vs. 0.2, ideal <0.1) during the nonautumn months. Both tests were better at ruling out than ruling in disease based on the predictive values [26].

For basal plasma ACTH, the cut-off increased during the autumn season to over twice the nonautumn value, but so did both the sensitivity and specificity, which increased to 100 and 95%, respectively. Likewise, for α-MSH, the cut-off value increased by more than threefold for autumn only, but also corresponding to a sensitivity of 100% and a specificity of 97%. Furthermore, the likelihood ratios in the study show that for autumn months both the positive and negative likelihood ratios are good for ruling in and out disease for both α-MSH and ACTH [26]. These results support the findings of Copas and Durham [13], where seasonally adjusted reference intervals for ACTH were proposed, but the results of the present study go further to support that diagnostic testing in the autumn period may be superior to that in other seasons. It may be that the autumn provides a natural stimulation test to the hypothalamic-pituitary axis similar to that which results from TRH stimulation [6,7]. Further research using larger populations from different latitudes is warranted to verify this finding.

In conclusion, plasma α-MSH and ACTH were highly correlated with each other and with clinical and historical indicators of PPID. The increase in both plasma α-MSH and plasma ACTH with increasing numbers of clinical signs in affected horses supports a spectrum of disease, which is reflected in the spectrum of plasma α-MSH and ACTH values. When the population sample was evaluated for diagnostic cut-off values, moderate and good sensitivity for plasma α-MSH and good specificity for plasma ACTH were found in seasons other than autumn. However, despite higher cut-off values, sensitivity and specificity were higher during the autumn period, suggesting this may be the ideal time to test. Further studies to develop seasonally adjusted cut-off values for different geographical locations are warranted.

Source of funding

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

The authors gratefully acknowledge the support of the University of Queensland's internal grant scheme in funding this project.

Manufacturers' addresses

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information

a Vacutainer Systems, Becton, Dickinson & Co., Rutherford, New Jersey, USA.

b Euro diagnostica, Malmö, Sweden.

c VETPATH Laboratory Services, Ascot, Western Australia, Australia.

d MedCalc, Mariakerke, Belgium.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information
  • 1
    McFarlane, D. (2007) Advantages and limitations of the equine disease, pituitary pars intermedia dysfunction as a model of spontaneous dopaminergic neurodegenerative disease. Ageing Res. Rev. 6, 54-63.
  • 2
    Schott, H.C. (2002) Pituitary pars intermedia dysfunction: equine Cushing's disease. Vet. Clin. N. Am.: Equine Pract. 18, 237-270.
  • 3
    Schott, H.C. Fun facts about pituitary pars intermedia dysfunction in horses. Proceedings in: Equine endocrinology summit, Charleston SC, USA. 2011.
  • 4
    Dybdal, N.O., Hargreaves, K.M., Madigan, J.E., Gribble, D.H., Kennedy, P.C. and Stabenfeldt, G.H. (1994) Diagnostic testing for pituitary pars-intermedia dysfunction in horses. J. Am. Vet. Med. Ass. 204, 627-632.
  • 5
    Frank, N., Andrews, F.M., Sommardahl, C.S., Eiler, H., Rohrbach, B.W. and Donnell, R.L. (2006) Evaluation of the combined dexamethasone suppression/thyrotropin-releasing hormone stimulation test for detection of pars intermedia pituitary adenomas in horses. J. Vet. Intern. Med. 20, 987-993.
  • 6
    Beech, J., Boston, R., Lindborg, S. and Russell, G.E. (2007) Adrenocorticotropin concentration following administration of thyrotropin-releasing hormone in healthy horses and those with pituitary pars intermedia dysfunction and pituitary gland hyperplasia. J. Am. Vet. Med. Ass. 231, 417-426.
  • 7
    McFarlane, D., Beech, J. and Cribb, A. (2006) Alpha-melanocyte stimulating hormone release in response to thyrotropin releasing hormone in healthy horses, horses with pituitary pars intermedia dysfunction and equine pars intermedia explants. Domest. Anim. Endocrinol. 30, 276-288.
  • 8
    Horowitz, M.L., Neal, L. and Watson, J.L. (2003) Characteristics of plasma adrenocorticotropin, β-endorphin and α-melanocyte stimulating hormone as diagnostic tests for pituitary pars intermedia dysfunction in the horse. J. Vet. Intern. Med. 17, 386.
  • 9
    Couetil, L., Paradis, M.R. and Knoll, J. (1996) Plasma adrenocorticotropin concentration in healthy horses and in horses with clinical signs of hyperadrenocorticism. J. Vet. Intern. Med. 10, 1-6.
  • 10
    Perkins, G.A., Lamb, S., Erb, H.N., Schanbacher, B., Nydam, D.V. and Divers, T.J. (2002) Plasma adrenocorticotropin (ACTH) concentrations and clinical response in horses treated for equine Cushing's disease with cyproheptadine or pergolide. Equine Vet. J. 34, 679-685.
  • 11
    McFarlane, D., Donaldson, M.T., McDonnell, S.M. and Cribb, A.E. (2004) Effects of season and sample handling on measurement of plasma alpha-melanocyte-stimulating hormone concentrations in horses and ponies. Am. J. Vet. Res. 65, 1463-1468.
  • 12
    Donaldson, M.T., McDonnell, S.M., Schanbacher, B.J., Lamb, S.V., Mcfarlane, D. and Beech, J. (2005) Variation in plasma adrenocorticotropic hormone concentration and dexamethasone suppression test results with season, age, and sex in healthy ponies and horses. J. Vet. Intern. Med. 19, 217-222.
  • 13
    Copas, V.E. and Durham, A.E. (2011) Circannual variation in plasma adrenocorticotropic hormone concentrations in the UK in normal horses and ponies, and those with pituitary pars intermedia dysfunction. Equine Vet. J. Epub ahead of print; doi: 10.1111/j.2042-3306.2011.00444.x.
  • 14
    van der Kolk, J.H., Heinrichs, M., van Amerongen, J.D., Stooker, R.C., in de Wal, L.J. and van den Ingh, T.S. (2004) Evaluation of pituitary gland anatomy and histopathologic findings in clinically normal horses and horses and ponies with pituitary pars intermedia adenoma. Am. J. Vet. Res. 65, 1701-1707.
  • 15
    McFarlane, D., Miller, L.M., Craig, L.E., Dybdal, N.O., Habecker, P.L., Miller, M.A., Patterson, J.S. and Cribb, A.E. (2005) Agreement in histologic assessments of the pituitary pars intermedia in aged horses. Am. J. Vet. Res. 66, 2055-2059.
  • 16
    Miller, M.A., Pardo, I.D., Jackson, L.P., Moore, G.E. and Sojka, J.E. (2008) Correlation of pituitary histomorphometry with adrenocorticotrophic hormone response to domperidone administration in the diagnosis of equine pituitary pars intermedia dysfunction. Vet. Pathol. 45, 26-38.
  • 17
    McGowan, T.W., Pinchbeck, G.L., Phillips, C., Perkins, N., Hodgson, D.R. and McGowan, C.M. (2010) A survey of aged horses in Queensland, Australia. Part 1: management and preventive health care. Aust. Vet. J. 88, 420-427.
  • 18
    Ruopp, M.D., Perkins, N.J., Whitcomb, B.W. and Schisterman, E.F. (2008) Youden index and optimal cut-point estimated from observations affected by a lower limit of detection. Biom. J. 50, 419-430.
  • 19
    Van der Kolk, J.H., Wensing, T., Kalsbeek, H.C. and Breukink, H.J. (1995) Laboratory diagnosis of equine pituitary pars intermedia adenoma. Domest. Anim. Endocrinol. 12, 35-39.
  • 20
    Hillyer, M.H., Taylor, F.G., Mair, T.S., Murphy, D., Watson, T.D.G. and Love, S. (1992) Diagnosis of hyperadrenocorticism in the horse. Equine Vet. Educ. 4, 131-134.
  • 21
    McGowan, C. (2003) Diagnostic and management protocols for equine Cushing's syndrome. In Pract. 25, 586-592.
  • 22
    Fluss, R., Faraggi, D. and Reiser, B. (2005) Estimation of the Youden Index and its associated cutoff point. Biom. J. 47, 458-472.
  • 23
    Greiner, M., Pfeiffer, D. and Smith, R.D. (2000) Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests. Prev. Vet. Med. 45, 23-41.
  • 24
    McFarlane, D., Paradis, M.R., Zimmel, D., Sykes, B., Brorsen, B.W., Sanchez, A. and Vainio, K. (2011) The effect of geographic location, breed, and pituitary dysfunction on seasonal adrenocorticotropin and α-melanocyte-stimulating hormone plasma concentrations in horses. J. Vet. Intern. Med. 25, 872-881.
  • 25
    Lee, Z.Y., Zylstra, R. and Haritou, S.J. (2010) The use of adrenocorticotrophic hormone as a potential biomarker of pituitary pars intermedia dysfunction in horses. Vet. J. 185, 58-61.
  • 26
    Petrie, A. and Watson, P. (2006) Additional techniques. In: Statistics for Veterinary and Animal Science, 2nd edn., Wiley-Blackwell, Oxford. pp 191-195.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Authors' declaration of interests
  8. Source of funding
  9. Manufacturers' addresses
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
evj575_sm_Item1.pdf81KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.