Estradiol Immunoreactivity and Plasma Estradiol Concentrations in Domestic Ferrets (Mustela putorius furo) with Adrenal Gland Hyperplasia and Neoplasia

Laura A. Adams, Ashley L. Ayoob, Cheryl B. Greenacre, Raymond P. Campagnoli, and Kenneth S. Latimer

Honors Biology Program, Biological Sciences (Adams, Ayoob) and Department of Pathology, College of Veterinary Medicine (Campagnoli, Latimer), The University of Georgia, Athens, GA 30602 (USA)

Abstract: A pilot study was performed to evaluate estradiol immunoreactivity in tissue sections of adrenal glands from 41 domestic ferrets (Mustela putorius furo). The adrenal glands were diagnosed histologically as normal (n=7), nodular adrenocortical hyperplasia (n=10), adrenocortical adenoma (n=4), or adrenocortical carcinoma (n=20). Plasma estradiol concentrations were available for 21 ferrets with adrenal-associated endocrinopathy. Normal adrenal glands infrequently exhibited estradiol immunoreactivity in tissue section. Generally, immunoreactivity in nodular hyperplasia was less than that of adrenal neoplasms. Slightly greater estradiol expression was observed in adrenocortical adenomas than in adrenocortical carcinomas. Plasma estradiol concentrations showed some correlation with immunohistochemical findings; however, hormonal concentrations could not distinguish nodular cortical hyperplasia from neoplasia. Complete characterization of adrenal-associated endocrinopathy in the ferret is best accomplished using a combination of H&E-stained tissue sections, estradiol immunohistochemistry, and determination of plasma estradiol concentrations simultaneously.

Key Words: Ferret, Mustela putorius furo, Adrenal gland, Endocrinopathy, Hyperplasia, Adenoma, Carcinoma, Estradiol, Immunohistochemistry, Plasma

Introduction

Endocrine disease is observed commonly in middle age to older domestic ferrets (Mustela putorius furo) and is often associated with the adrenal gland. ¹ Histologic lesions of the adrenal gland of ferrets include nodular cortical hyperplasia, adrenocortical adenoma, and adrenocortical carcinoma. Any of these diseases of the adrenal cortex can cause excessive production of steroid hormones such as estradiol, progesterone, and testosterone. ^1,5 Unilateral or bilateral adrenalectomy is often required to treat adrenal-associated endocrinopathy of ferrets. ⁸

Adrenal-associated endocrinopathy in the ferret is suspected with the presence of bilateral (occasionally pruritic) alopecia, vulvar swelling in spayed female ferrets, dysuria in neutered male ferrets, myelosuppression, abnormal bleeding, and behavioral changes. ^4,6,8 Behavioral changes are related to reproductive habits wherein spayed female ferrets become interested in males as would during normal estrus. ⁸ Suspicion of adrenal neoplasia may be based upon abdominal palpation or visualization of masses near the pole of the kidney by ultrasonography or radiography. Determination of plasma hormone concentrations is necessary to document functional endocrinopathy, which may occur with adrenocortical hyperplasia or neoplasia. ^1-7 However, none of these methods can reliably distinguish between cortical , adenoma, or carcinoma. Although histology has been used as the "gold standard" to diagnose adrenal lesions in humans and domesticated animals, correct diagnosis of adrenal lesions in ferrets is often difficult.

The purpose of this study was to develop a reliable and consistent method to diagnose nodular cortical hyperplasia, adrenocortical adenoma, and adrenocortical carcinoma in the ferret using H&E-stained tissue sections in conjunction with estradiol immunohistochemistry. In addition, we wished to determine whether estradiol immunoreactivity in tissue sections was predictive of estradiol concentrations in the plasma.

Materials and Methods

Case selection: All adrenal gland biopsies from ferrets submitted to the Department of Pathology and the Athens Diagnostic Laboratory between 1996 and 1999 were reviewed. The submission sheets were reviewed for signalment and significant medical history. Plasma estradiol concentrations, if determined, were recorded for future reference. The initial H&E-stained tissue sections were retrieved from the archival files and reviewed for accuracy of histologic diagnosis and the presence of an intact adrenal gland that was free of crush artifact, distortion, and autolysis or necrosis. Subsequently, the paraffin embedded tissue blocks were procured, if available, and examined to determine that enough tissue remained for replicate sectioning.

Immunohistochemical staining: Formalin fixed, paraffin embedded tissue blocks were sectioned at 3 to 4 m m and placed on ProbeOn Plus ^a glass slides. Tissue sections were deparaffinized in three changes of limonene (HemoDe)^a and rehydrated through graded ethanol solutions to deionized water. Endogenous peroxidase activity was quenched by immersing the tissue sections in 3% hydrogen peroxide for 15 minutes. Subsequently, the sections were rinsed again in deionized water, transferred to Autobuffer, ^b and inserted face-to-face in a MicroProbe^a slide holder.

To reduce background staining, the tissues were blocked before exposure to the primary antibody. The blocking solution consisted of 1 drop of equine serum from a universal polyclonal kit ^c added to 10mL PBS with 0.3% triton X-100 detergent. The gaps between slide pairs were filled by capillary action with this blocking solution for 5 minutes. The blocking solution was removed by blotting.

The primary antibody was a prediluted, polyclonal, rabbit anti-estradiol antibody. ^d The capillary gaps were filled with primary antibody solution and incubated for 30 minutes and 30° C. The antibody solution subsequently was removed by blotting and the tissue sections were rinsed multiple times with Autobuffer. A biotinylated secondary antibody from a commercial universal immunostaining kit^c was mixed with 1.5% horse serum in PBS. The capillary gaps between slides were filled with the secondary antibody solution and incubated for 30 minutes at 30° C. The antibody was then removed by blotting, and the slides were rinsed multiple times with Autobuffer.

Sites of primary antibody binding were visualized by the avidin-bioin immunoperoxidase technique. The capillary gaps were filled with avidin-biotin complex and incubated for 30 minutes at 30° C. The slides were rinsed multiple times with Autobuffer, and the capillary gaps were filled with a chromagen solution composed of 20 m L 3% hydrogen peroxide and 2.5mL diaminobenzidine (DAB, 1mg/mL). The slides were incubated for 2 minutes or until the tissues were adequately stained. The chromagen solution was then removed by blotting. Following brief counterstaining with Gill’s II Hematoxylin, the tissue sections were dehydrated through graded ethanol solutions to xylene and coverslipped.

Image acquisition and analysis: Each slide was examined microscopically for the presence of estradiol immunoreactivity. Digital images containing cortical cells were captured from each slide at 40X magnification for subsequent analysis. Sites of image capture were chosen randomly in areas of uniform tissue thickness that were free of tissue distortion. All images were captured under identical illumination settings; images were not altered or enhanced prior to analysis. The images were saved by computer into Image-Pro Plus Version 3.0 for Windows. ^e The settings for image analysis were adjusted to detect and quantify the total area (from which the percent area was calculated) and the percentage of area exhibiting estradiol immunostaining. The resultant values for estradiol immunoreactivity in tissue sections were compared with the histologic diagnosis for each tissue and the plasma estradiol concentration, if known.

Statistical analyses: The percentage area of estradiol immunoreactivity for adrenal glands with a diagnosis of normal gland, nodular cortical hyperplasia, adrenocortical adenoma, and adrenocortical carcinoma were tested for significant differences by One-way analysis of variance. Significant differences between group means were localized using Fisher’s protected least significant difference. The level of significance tested was P <0.05. Linear regression analysis was used to determine the correlation, if any, between plasma estradiol concentration and the percentage of estradiol immunoreactivity within the tissue sections.

Results

Adrenal tissues from 41 ferrets were selected for immunohistochemical staining. The histologic diagnoses of these tissues, based upon examination of H&E-stained tissue sections was as follows: normal adrenal gland (n=7), nodular adrenocortical hyperplasia (n=10), adrenocortical adenoma (n=4), and adrenocortical carcinoma (n=20). Histologic diagnosis, sex, plasma estradiol concentrations, and tissue section immunoreactivities for estradiol are presented in Table 1. Immunostaining revealed variable intensities of estradiol staining in the adrenocortical cells (Figs. 1 through 4). Plasma estradiol concentrations were determined for 21 ferrets with adrenal-associated endocrinopathy; however, plasma hormone concentrations were not determined for the ferrets with normal adrenal glands.

The variation within the four diagnostic groups was high and significant differences were not apparent between group means. Two of seven normal adrenal gland tissue sections revealed slight immunoreactivity for estradiol (0.2% and 0.9% of the total area evaluated). Tissue sections diagnosed as nodular adrenocortical hyperplasia had estradiol immunoreactivity that ranged from 0.1% to 7.7% of the total tissue area; however, mean estradiol activity was intermediate between that of normal glandular tissue and adrenal neoplasms. The highest immunoreactivity for estradiol generally was observed in adrenocortical adenomas (range = 0.0% to 22.6% of area) and carcinomas (0.0% to 31.0% of total area). Considerable overlap was observed between the estradiol expression in adenomas and carcinomas; however, mean immunoreactivity was slightly greater in the adenoma group (Table 2).

Plasma estradiol concentrations (pMol/L) were measured in 21 ferrets; however, none of the ferrets sampled had histologically normal adrenal glands. Estradiol concentrations varied widely in nodular hyperplasia (181 to 276 pmol/L), adrenocortical adenoma (22 to 229 pmol/L), and adrenocortical carcinomas (135 to 398 p mol/L). The correlation between the measured plasma estradiol concentrations and the percent area of estradiol immunoreactivity in tissue section was marginally significant (p=0.052). The correlation coefficient (R² ) was relatively low with a value of only 0.183.

Fig. 1. Normal adrenal gland, ferret, estradiol immunostaining with hematoxylin counterstain, 40X magnification. Adrenocortical cells lack estradiol reactivity.	Fig. 2. Nodular adrenocortical hyperplasia, ferret, estradiol immunostaining with hematoxylin counterstain, 40X magnification. Foci of adrenocortical cells exhibit mild but variable estradiol reactivity.

Fig. 3. Adrenocortical adenoma, ferret, estradiol immunostaining with hematoxylin counterstain, 40X magnification. Diffuse, moderately intense estradiol immunoreactivity is apparent.	Fig. 4. Adrenocortical carcinoma, ferret, estradiol immunostaining with hematoxylin counterstain, 40X magnification. Diffuse, intense estradiol immunoreactivity is present.

Discussion

In general, tissue sections of normal adrenal glands from ferrets exhibited little to no immunoreactivity for estradiol. Only two histologically normal adrenal glands had slight estradiol immunoreactivity. This observation is expected because normal adrenal glands only produce or secrete extremely small amounts of estradiol. ⁷ Conversely, these two ferrets may have been in the early or subclinical stages of developing cortical hyperplasia or neoplasia. Adrenal glands with nodular cortical hyperplasia usually displayed estradiol immunoreactivity that was intermediate between that of normal adrenal glands and adrenal neoplasms (adenomas or carcinomas). The total areas of estradiol immunoreactivity ranged from 0.1% to 7.7%. Estradiol staining was similar in cortical adenomas and carcinomas; however, the degree of immunoreactivity varied widely. The adrenal carcinomas displayed the broadest range of immunoreactivity. The most likely explanations for the wide range of immunoreactivity could be extremely variable estradiol production or lack of recognition of the antigenic epitope by the monoclonal antibody. Alternatively, some of the lesions may have been misdiagnosed by the pathologist. Although the wide ranges of immunostaining observed in these neoplasms indicate that estradiol immunoreactivity alone cannot be used to differentiate adenomas from carcinomas, neoplasia is highly suspect when 8% or more of the tissue area exhibits immunostaining. However, a threshold of 8% is dependent upon the parameters of image analysis that have been specified within the software program.

Plasma estradiol concentrations cannot reliably differentiate nodular cortical hyperplasia, adrenocortical adenoma, and adrenocortical carcinoma. However, each of these conditions may produce adrenal-associated endocrinopathy in the ferret.

Lack of strong correlation between plasma estradiol concentration and tissue immunoreactivity suggests that some hyperplastic or neoplastic cortical cells may produce and quickly secrete large quantities of estradiol, resulting in an increased plasma hormone concentration with lack of tissue immunoreactivity. Alternatively, some cells may produce large quantities of estradiol with inhibited cellular secretion. This would result in low plasma estradiol concentrations with marked tissue immunoreactivity.

In summary, this study demonstrates that abnormal adrenal glands of ferrets often exhibit estradiol immunoreactivity. The intensity of immunostaining varies somewhat according to the type of adrenal gland tumor that is present. Furthermore, the degree of estradiol immunoreactivity in tissue sections roughly correlates with plasma concentrations of this hormone. Although estradiol immunoreactivity in tissue sections and plasma hormone concentrations can document hyperestrogenism in diseased ferrets, careful examination of H&E-stained tissue sections is still necessary to distinguish cortical nodular hyperplasia from neoplasia.

Sources and Manufacturers

a. Fisher Scientific, Pittsburgh, PA, USA.
b. Biomeda Corporation, Foster City, CA, USA.
c. Vector Laboratories, Burlingame, CA, USA.
d. BioGenex, San Ramon, CA, USA.
e. Media Cybernetics, Silver Spring, MD, USA.

References

1. Ackerman J, Carpenter JW, Godshalk CP, et al. Ultrasonographic detection of adrenal gland tumors in two ferrets. J Am Vet Med Assoc 1994; 205:1001-1003.

2. Caticha O, Odell WD, Wilson DE, et al. Estradiol stimulates cortisol production by adrenal cells in estrogen-dependent primary adrenocortical nodular dysplasia. J of Clinical Endocrinology and Metabolism 1993; 21:494-495.

3. Hillyer EV. Ferret Endocrinology. Diseases of Caged Birds and Exotic Pets. In: Kirk RW, Bonagura JD (eds): Current Veterinary Therapy XI. Philadelphia, W. B. Saunders Co. , 1992, pp. 1185-1189.

4. Lipman NS Marini RP, Murphy JC, et al. Estradiol-17b -secreting adrenocortical tumor in a ferret. J Am Vet Med Assoc 1993; 203:1552-1555.

5. Rosenthal KL, Peterson ME. Evaluation of plasma androgen and estrogen concentrations in ferrets with hyperadrenocorticism. J Am Vet Med Assoc 1996; 209:1097-1102.

6. Rosenthal KL, Peterson ME, Quesenberry KE, et al. Hyperadrenocorticism associated with adrenocortical tumor or nodular hyperplasia of the adrenal gland in ferrets: 50 cases (1987-1991). J Am Vet Med Assoc 1993; 203:271-275.

7. Wagner RA, Dorn DP. Evaluation of serum estradiol concentrations in alopecic ferrets with adrenal gland tumors. J Am Vet Med Assoc 1994;205:703-707.

8. Wheler CL, Kamieniecki CL. Ferret adrenal-associated endocrinopathy. Can Vet J 1998; 39:175-17

This Page Last Updated November 15, 1999

^  Top of page