Histopathologic Findings in Canine Pituitary Glands

Veterinary Pathology 1-9

@ The Author(s) 2018 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0300985818766211

journals.sagepub.com/home/vet

S

Margaret A. Miller1,2, David S. Bruyette3, J. Catharine Scott-Moncrieff4, Tina Jo Owen5, José A. Ramos-Vara1,2, Hsin-Yi Weng2, Andrea L. Vanderpool1,2, Annie V. Chen5, Linda G. Martin5, Deidre M. DuSold’, and Sina Jahan6

Abstract

To optimize the histologic evaluation of hypophysectomy specimens, sections of 207 canine pituitary glands (196 postmortem, I I hypophysectomy specimens) were reviewed. Adenohypophyseal proliferation was the most common (n = 79) lesion. Pro- liferative lesions were sparsely to densely granulated; the granules were usually basophilic to chromophobic and periodic acid- Schiff-positive. Adenohypophyseal proliferation was classified as hyperplasia (n = 40) if ≤2 mm diameter with intact reticulin network, as microadenoma (n = 22) for 1-5 mm homogeneous nodules with lost reticulin network, or as macroadenoma (n = 17) for larger tumors. Craniopharyngeal duct cysts were common incidental lesions and the only lesion in 15 dogs. Uncommon diagnoses included lymphoma (n = 4), hemorrhagic necrosis (n = 4), metastatic carcinoma (n = 3), hypophysitis (n = 3), ependymoma (n = 2), craniopharyngioma (n = 2), and I case each of metastatic melanoma, pituicytoma, gliomatosis, germ cell tumor, meningioma, and atrophy. The pituitary histologic diagnosis was associated with hyperadrenocorticism (HAC; P < .001) and adrenocortical histologic diagnosis (P = . 025). Both HAC and adrenocortical hyperplasia showed a positive trend with the degree of adenohypophyseal proliferation. The association of adrenocortical hyperplasia with HAC was not significant (P = . 077). Dogs with adenohypophyseal proliferations were older than dogs with normal pituitary glands (P <. 05). Brachycephalic breeds were overrepresented among dogs with pituitary macroadenoma or craniopharyngeal duct cysts, but the association was not statistically significant (P = . 076). Adenohypophyseal hyperplasia was more common than adenoma among postmortem specimens, but was unexpected in >80% of cases. Pituitary macroadenoma was the most common diagnosis in hypophysectomy specimens.

Keywords

adenohypophyseal hyperplasia, dogs, hyperadrenocorticism, pituitary adenoma, pituitary-dependent hypercortisolism, transsphe- noidal hypophysectomy

The pituitary gland is capable of diverse responses to injury; however, in most species, the most common lesion is prolif- eration.14,28 When functional, adenohypophyseal prolifera- tions can result in dysfunction of secondary endocrine glands. Indeed, corticotroph adenoma is considered the most common cause of spontaneous canine hypercortisolism.26 Until recently, the medical and surgical treatment of canine pituitary-dependent hypercortisolism (PDH) has been aimed mainly at the adrenal gland,3,11,15,23 and histologic evaluation of canine pituitary glands has been predominantly a postmor- tem endeavor. In contrast, pituitary adenectomy or hypophy- sectomy is a mainstay in the treatment of human PDH (Cushing’s disease) and other pituitary diseases, and surgical pathology is an important part of diagnosis and manage- ment. 2,4,5,22,24,27,29-31

With improved imaging, microsurgical techniques, and post- operative care, transsphenoidal hypophysectomy has become a viable treatment option for canine pituitary disease, yet surgical pathology has been a minor component of most

publications.9,16,17,21,25 The main goal of this study was to classify canine pituitary lesions in archived case materials as the basis for development of a standardized approach to histologic evaluation of transsphenoidal hypophysectomy specimens. In addition to

“Indiana Animal Disease Diagnostic Laboratory, Purdue University, West Lafayette, IN, USA

2Department of Comparative Pathobiology, Purdue University, West Lafay- ette, IN, USA

3 Anivive Life Sciences, Long Beach, CA, USA

4Department of Veterinary Clinical Sciences and Veterinary Teaching Hospital, Purdue University, West Lafayette, IN, USA

5Department of Veterinary Clinical Sciences and Veterinary Teaching Hospital, Washington State University, Pullman, WA, USA

6Department of Internal Medicine, Small Animal Hospital of Veterinary Faculty, University of Tehran, Tehran, Iran

Corresponding Author:

Margaret A. Miller, Animal Disease Diagnostic Laboratory, Purdue University, 406 S University St, West Lafayette, IN 47907, USA. Email: pegmiller@purdue.edu

Table I. Antibodies, Reagents, and Procedures for Immunohistochemistry.ª
Antibody (Code or Clone)DilutionIncubation (min.)bHIER Antigen RetrievalDetection SystemAntibody Source
Adrenocorticotrophic hormone (ab75683)1:16 00045Target Retrieval Solution,“ pH 6.0EnVision+ HRP-labeled polymer anti-rabbit“AbCam
Growth hormone (PA1-85518- T16000)1:20 00045Diva Decloaker, pH 6.2EnVision+ HRP-labeled polymer anti-rabbit“Thermo Fisher Scientific
Melanocyte stimulating hormone (MZ1111-0050)1:50060Reveal Decloaker, pH 6.0Rabbit-on-canine HRP polymerdEnzo Life Sciences
Prolactin (A0569-T300)1:30045NoneEnVision+ HRP-labeled polymer anti-rabbit“Dako Agilent Pathology Solutions

Abbreviations: HIER, heat-induced epitope retrieval; HRP, horseradish peroxidase.

ªThe chromogen for all markers was 3, 3’ diaminobenzidine (Betazoid DAB Chromogen Kit).d

bAll incubations were performed at room temperature.

“Dako Agilent Pathology Solutions.

Biocare Medical.

evaluating the prevalence and nature of adenohypophyseal proliferations and other pituitary lesions, the primary study objectives were to evaluate the association between (1) pitui- tary histologic diagnosis and a clinical diagnosis of hyperadre- nocorticism, (2) pituitary histologic diagnosis and histologic evidence of adrenocortical hyperplasia, and (3) adrenocortical hyperplasia and a clinical diagnosis of hyperadrenocorticism. The secondary objectives were to investigate whether age, breed (brachycephalic vs nonbrachycephalic), or sex of dogs was associated with pituitary histologic diagnosis, adrenocor- tical hyperplasia, or hyperadrenocorticism.

Materials and Methods

Records of the Indiana Animal Disease Diagnostic Laboratory from July 2007 through August 2015 were retrieved for all dogs in which the pituitary gland had been examined histologically. Cases were included if histologic sections or paraffin blocks of suitable quality (adequate sample size without severe autolysis or tissue artifact) were available. Signalment, history, clinical impression, and gross and histologic findings were tabulated from submission forms and pathology reports. All pituitary sections were reviewed by 2 authors (MAM, ALV) at a double-headed microscope and classified14,28 independently from the original pathology report into one or more of the following categories: histologically normal, adenohypophyseal proliferation (hyperplasia, microadenoma, or macroadenoma), craniopharyngeal duct cysts, or miscellaneous lesions (other pituitary neoplasms, secondary neoplasia, hypophysitis, atro- phy, etc). Proliferative adenohypophyseal cells were categor- ized as basophilic, chromophobic, or acidophilic based on tinctorial characteristics of the cytoplasmic granules in hema- toxylin and eosin-stained (HE) sections. Periodic acid-Schiff (PAS) and Gordon and Sweet’s reticulin histochemistry, and immunohistochemistry (IHC, Table 1) for adrenocortico- trophic hormone (ACTH), growth hormone (GH), melanocyte stimulating hormone (MSH), or prolactin were used for selected cases when authorized by the referring veterinarian

or as part of development of an IHC panel for evaluation of hypophysectomy specimens of pituitary adenoma. For trophic hormone IHC, neutral buffered formalin-fixed (1-2 days), paraffin-embedded sections of normal canine pituitary gland (with pars nervosa, pars intermedia, and pars distalis) were used to standardize each test. The final IHC protocol used the primary antibody dilution, antigen retrieval method, and detec- tion system that produced the best signal-to-noise ratio in the targeted cells. Normal canine pituitary gland was the positive control tissue. Adenohypophyseal proliferations in which the cells expressed ACTH, but not other trophic hormones, were classified as corticotroph hyperplasia or adenoma. Lesions in which the cells expressed MSH and ACTH were classified as melanotroph hyperplasia or adenoma. Lesions in which the cells expressed ACTH and GH were classified as plurihormo- nal. Adenohypophyseal nodular proliferations were classified as hyperplasia if ≤2 mm in diameter, with expanded but intact reticulin network; as microadenoma if homogeneous (one cell type) and 1-5 mm in diameter with loss of the reticulin net- work; and as macroadenoma if homogeneous and >5 mm in diameter with loss of the reticulin network (because adenomas >5 mm in diameter enlarge the canine pituitary gland). 14,19 However, the diameter of proliferative lesions could not be measured in histologic sections of the hypophysectomy speci- mens, so classification as macroadenoma in those cases was based on pituitary dimensions >1 cm in magnetic resonance (MR) images. For nodular proliferations >1 mm in diameter, maintenance of the reticulin network was the criterion used to distinguish between hyperplasia and microadenoma. Adenohy- pophyseal hyperplasia, microadenoma, and macroadenoma were mutually exclusive classifications. Thus, a pituitary gland with both hyperplasia and microadenoma was categorized as microadenoma, and a gland with microadenoma and macroa- denoma was categorized as macroadenoma. Mucin-filled structures lined by pseudostratified columnar epithelium with ciliated cells and goblet cells were classified as craniopharyngeal duct cysts if at least 1 mm in diameter. The categories of

Table 2. Descriptive Statistics of 207 Dogs With Pituitary Histologic Diagnoses.
DiagnosisªHistologically normalAdenohypophyseal hyperplasiaPituitary microadenomaPituitary macroadenomaMiscellaneousbCraniopharyngeal duct cysts
Number (%)92 (44)40 (19)22 (11)17 (8)21 (10)15 (7)
Median age (range)7.0 (0-15.0)10.0 (2.0-16.0)"11.5 (8.0-17.0)"10.0 (5.0-13.0)"8.0 (3.0-15.5)9.0 (2.0-14.0)
Brachycephalic breed, n (%)
Yes11 (44)3 (12)0 (0)4 (16)3 (12)4 (16)
No81 (45)37 (20)22 (12)13 (7)18 (10)11 (6)
Sex, n (%)
Male38 (41)19 (21)12 (13)9 (10)9 (10)5 (5)
Female54 (47)21 (18)10 (9)8 (7)12 (10)10 (9)

ªSee Materials and Methods for diagnostic criteria.

bMiscellaneous included neoplasms other than pituitary adenoma, inflammation, or degeneration.

“Significantly different from “Histologically normal” (P < . 05).

craniopharyngeal duct cysts and miscellaneous lesions were not considered mutually exclusive.

For all dogs that had been patients of Purdue University Veterinary Teaching Hospital (PUVTH), medical records were reviewed for evidence of hyperadrenocorticism (HAC), defined as the presence of ≥3 of the following clinical signs: polyuria, polydipsia, polyphagia, abdominal distension, alope- cia, or hypertension. A confirmed diagnosis of HAC was made in dogs with both clinical signs of HAC and a positive endo- crine test, for example, failure to suppress cortisol concentra- tions to <1.9 µg/dL at 8 hours on the low-dose dexamethasone suppression (LDDS), or a post-ACTH serum cortisol >17 µg/dL. The dogs were categorized as HAC not suspected, HAC suspected but not confirmed, or HAC confirmed.6,26 For cases with adenohypophyseal proliferation, available histolo- gic sections of adrenal gland were reviewed for adrenocortical proliferation, which was classified as within normal limits, hyperplasia, adenoma, or carcinoma.14,28 For data analysis, the adrenocortical histologic variable was dichotomized into hyperplasia versus other classifications. Eleven cases were non-PUVTH transsphenoidal hypophysectomy specimens (sur- gery by author TJO); all 11 dogs had been evaluated for PDH by one or more of the following tests: urinary cortisol to crea- tinine ratio, LDDS, ACTH-stimulation test, basal ACTH con- centration, or high-dose dexamethasone suppression (HDDS).6,8,26

Fisher’s exact tests were used to evaluate the bivariate cor- relation between pituitary histologic diagnosis, a clinical diag- nosis of hyperadrenocorticism, and a histologic diagnosis of adrenocortical hyperplasia. The age distribution was compared across the classifications of pituitary histologic diagnosis and clinical diagnosis of HAC using Kruskal-Wallis tests, followed by the pairwise comparisons with a significant result. Mann- Whitney U test was used to compare age distribution between a histologic diagnosis of adrenocortical hyperplasia and other adrenocortical diagnoses. The association of breed (brachyce- phalic vs nonbrachycephalic) and sex with pituitary histologic diagnosis, clinical diagnosis of HAC, or histologic diagnosis of adrenocortical hyperplasia was evaluated by Fisher’s exact test. Significance was set at P < . 05 for all statistical analyses.

Results

Of the 234 archived canine pituitary specimens, 207 met the inclusion criteria. Of these pituitary glands, 196 were collected at autopsy and 11 were transsphenoidal hypophysectomy speci- mens. The histologic diagnoses are compared with signalment of the dogs in Table 2. In 92 (44%) dogs, the pituitary gland was histologically normal. Adenohypophyseal proliferation (hyperplasia, microadenoma, or macroadenoma) was the most common pituitary lesion, found in 79 (38%) dogs. Among dogs with adenohypophyseal proliferation, 36 (46%) had been con- sidered histologically unremarkable on initial postmortem examination.

Adenohypophyseal proliferative lesions were classified as hyperplasia in 40 dogs (all examined postmortem), including 27 (60%) for which pituitary lesions were not recorded in the original histopathology report. Hyperplastic nodules (Fig. 1) typically appeared as multiple microscopic aggregates of hypertrophied cells with ample densely granulated basophilic cytoplasm; the cytoplasmic granules were PAS-positive and immunolabelled for ACTH. Most hyperplastic nodules con- tained a few scattered acidophils. The largest hyperplastic nodule was 2 mm in diameter (ie, larger than the smaller micro- adenomas), but was classified as hyperplasia rather than neo- plasia because the reticulin network was expanded rather than disrupted, and because rare acidophils were scattered among the basophils.

Adenohypophyseal proliferations were classified as micro- adenoma in 22 pituitary glands (all examined postmortem), including 9 (41%) that had been considered unremarkable on the original examination. Microadenomas (Fig. 2) appeared as 1 or more homogeneous nodules of basophilic to chromopho- bic cells with exclusion of acidophils. The hypertrophied cells had densely granulated cytoplasm, and the granules were PAS- and ACTH-positive. Ten microadenomas were accom- panied by 1 or more hyperplastic nodules. Mitotic figures were noted in only 1 microadenoma (1 per ten 400x fields).

Adenohypophyseal proliferations were classified as macro- adenoma in 17 dogs. Of these, 10 were in hypophysectomy specimens, 4 were postmortem cases from PUVTH, and 3 were

Figure 1. Adenohypophyseal hyperplasia. a. The hyperplastic nodules (asterisks) are up to 1.3 mm in diameter and are composed of basophilic cells. Hematoxylin and eosin (HE). b. In the hyperplastic nodule (arrows), hypertrophied cells enlarge the acini, but the reticulin fiber network is intact. Nonproliferative adenohypophysis is in the upper right. Gordon and Sweet's reticulin stain. c. A few acidophils (arrows) are scattered among the basophils in the center of a hyperplastic nodule. HE. d. The hyperplastic corticotrophs have PAS-positive granules. An acidophil (arrow) in this hyperplastic nodule is PAS-negative. Periodic acid-Schiff (PAS). Fig. 2. Microadenoma, pituitary gland, dog. a. The microadenoma (asterisk) is 1.1 mm in diameter and consists of coalescing nests of basophilic cells. HE. b. In the microadenoma (arrows), the reticulin fiber network is lost. Nonproliferative adenohypophysis is in the upper right. Gordon and Sweet's reticulin. c. The adenoma consists of a homo- geneous population of densely granulated cells. HE. d. The cytoplasmic granules are PAS-positive. PAS. Fig. 3. Macroadenoma, pituitary gland, dog. a. The adenoma consists of a sheet of basophilic cells. HE. b. The reticulin fiber network is lost in the adenoma (arrows). Nonproliferative adenohypophysis is compressed to the right. Gordon and Sweet's reticulin. c. The adenoma consists of a homogeneous population of densely granulated cells. The cytoplasmic granules are PAS-positive. PAS.

1c

1a

1b

1d

2a

2b

2d

3a

3b

3c

postmortem specimens from other clinics. One macroadenoma, a hypophysectomy specimen, was composed of pale acidophi- lic and sparsely granulated cells that immunolabeled for ACTH and for GH, and negative for MSH and prolactin; that dog had no evidence of hyperadrenocorticism or hypersomatotropism. One dog with PDH and a macroadenoma diagnosed by MR imaging alone was treated with the somatostatin analog pasir- eotide instead of hypophysectomy. Autopsy of this dog revealed severe pan-adenohypophyseal atrophy with no evi- dence of neoplastic tissue in histologic sections of pituitary gland. The remaining macroadenomas were composed of chro- mophobic to pale basophilic cells with sparsely to densely granulated cytoplasm (Fig. 3). Minimal to moderate nuclear atypia was noted in 4 cases, but mitotic figures were rare (0-1 per ten 400x fields) with the exception of 1 case with a mitotic index of 7. Necrosis with cholesterol granulomas and neutrophilic inflammation was observed in or adjacent to some macroadenomas.

Miscellaneous lesions included neoplasms other than pituitary adenoma, hypophysitis, and degeneration. Other pituitary, sellar, or secondary neoplasms were diagnosed in 16 dogs. One dog had a 2-mm-diameter pituicytoma in the pars nervosa. A suprasellar germ cell tumor (Fig. 4) was diagnosed in a 4-year-old dog. Craniopharyngioma was diagnosed in 2 dogs: 1 postmortem case (Fig. 5) and 1 hypophysectomy case. Secondary neoplasms included 3 metastatic carcinomas (urothelial carcinoma; Fig. 6), nasal carcinoma, and carcinoma of unknown origin), 4 lympho- mas (each part of multicentric lymphoma; Fig. 7), 1 metastatic oral melanoma, local extension of 2 ependymomas, and 1 case each of oligodendroglial gliomatosis and meningioma.

Inflammatory or degenerative pituitary lesions were found in 8 dogs. Hypophysitis, diagnosed in 3 dogs, included 1 case each of systemic blastomycosis (Fig. 8), idiopathic granulomatous meningoencephalitis (GME), and idiopathic lymphocytic hypo- physitis (Fig. 9). Hemorrhage and necrosis of unknown cause was the sole diagnosis in 4 dogs. Adenohypophyseal atrophy was diagnosed in 1 dog with bilateral adrenocortical carcinoma.

Craniopharyngeal duct cysts were diagnosed in 32 dogs, and were the sole pituitary lesion in 15. These cysts typically appeared in the pars intermedia as uni- to multilocular mucin-filled cysts, 1-5 mm in diameter, and lined by pseudos- tratified columnar epithelium with ciliated cells and goblet cells (Fig. 10). The cysts were not associated with clinical disease in any dog.

The frequency count (joint distribution) of pituitary histolo- gic diagnosis by HAC clinical diagnosis in 129 dogs is sum- marized in Table 3. Medical records were reviewed for 118 PUVTH patients and for all 11 hypophysectomy cases. The pituitary histologic diagnosis was significantly (P < .001) associated with HAC. Overall, HAC was not suspected in 100, suspected but not confirmed in 18, and confirmed in 11 dogs. There was a positive trend between adenohypophyseal proliferation and HAC with increasing percentage of dogs with confirmed HAC as the degree of adenohypophyseal prolifera- tion increased. Only 1 dog with a histologically normal pitui- tary gland had a confirmed diagnosis of HAC; this dog had

adrenocortical nodular hyperplasia of unknown cause. Of the 3 PUVTH macroadenoma cases, HAC had not been suspected in 1, was suspected but not confirmed in 1, and was confirmed in 1 dog. Of the 11 hypophysectomy specimens (10 of which had pituitary macroadenoma), PDH had been confirmed in 6. Hyperadrenocorticism was not suspected in any of the PUVTH dogs with other pituitary neoplasms or nonproliferative pitui- tary lesions.

Adrenal gland sections from 35 dogs that had adenohypo- physeal hyperplasia were classified histologically as within normal limits in 18 (51%) dogs, adrenocortical hyperplasia in 14 (40%), adrenocortical adenoma in 2 (accompanied by adre- nocortical hyperplasia in 1 case), and adrenocortical carcinoma in 1 (3%). Adrenal sections from 21 dogs with pituitary micro- adenoma were classified as within normal limits in 5 (24%) dogs, adrenocortical hyperplasia in 12 (57%), adrenocortical adenoma in 2 (10%), and adrenocortical carcinoma in 2 (10%). All 4 pituitary macroadenoma cases in which adrenal gland was evaluated histologically had adrenocortical hyper- plasia. The data suggested a significant (P = . 025) monotonic association between adenohypophyseal proliferation and adre- nocortical hyperplasia, in which 40% of dogs with adenohypo- physeal hyperplasia, 57% of dogs with pituitary microadenoma, and 100% of dogs with pituitary macroade- noma had adrenocortical hyperplasia (Fig. 11). Dogs with adre- nocortical hyperplasia or neoplasia were 5.7 times more likely to be suspected of having or confirmed to have HAC (P = .035), but the association between adrenocortical hyperplasia and HAC was not significant (P = . 077).

The age of dogs in the study ranged from 1 day to 17 years old with a median age of 9 years old. Age was significantly (P< .001) associated with pituitary histologic diagnosis (Table 2). Dogs with adenohypophyseal proliferation (hyperplasia or ade- noma) were older than dogs in which the pituitary gland was within normal limits (P < . 05). Age was also significantly associated with a diagnosis of HAC (P = . 026), but not with adrenocortical hyperplasia (P = . 715).

Brachycephalic breeds comprised 25 (12%) of the 207 dogs (Table 2). These were over-represented among dogs with pitui- tary macroadenoma or craniopharyngeal duct cysts, but the difference was not statistically significant (P = . 076). There was no apparent association of brachycephalic phenotype with HAC (P = . 279) or with adrenocortical hyperplasia (P = . 492).

There were 115 female (81 spayed) and 92 male (59 castrated) dogs in the study (Table 2). No association of sex with pituitary histologic diagnosis (P = . 747) or a clinical diagnosis of HAC (P = . 601) was apparent, but sex was asso- ciated with adrenal histopathology (P = . 038). The data sug- gested that male dogs were more likely to have a histologic diagnosis of adrenocortical hyperplasia.

Discussion

Greater than 50% of the archived canine pituitary glands had histologic changes. Adenohypophyseal proliferation was the most common lesion and occurred in older dogs without sex

Figures. 4-10. Suprasellar germ cell tumor, pituitary fossa, dog. The germ cell component is in the upper left; lipid-laden hepatoid cells in the middle, and compressed adenohypophysis in the bottom right. Inset: Detail of the seminoma-like germ cells. Hematoxylin and eosin (HE). Fig. 5. Craniopharyngioma, pituitary fossa, dog. The neoplastic tissue consists of coalescing nests of squamous epithelial cells in fibrous stroma. Anucleate, heavily keratinized "ghost" cells are in the center of the nests. HE. Fig. 6. Metastatic urothelial carcinoma, pituitary fossa, dog. The neoplastic urothelial cells form solid nests in the pars distalis. HE. Fig. 7. Multicentric lymphoma, pituitary gland, dog. The neoplastic lymphocytes are smaller than the basophils, acidophils, and chromophobes of the pars distalis. HE. Fig. 8. Systemic blastomycosis, pituitary gland, dog. Granulomatous inflammation with macrophages and neutrophils surrounded by lymphocytes and plasma cells. Inset: Detail of yeast of Blastomyces dermatitidis. HE. Fig. 9. Idiopathic hypophysitis, pituitary gland, dog. Lymphocytes and plasma cells infiltrate the pars distalis. HE. Fig. 10. Craniopharyngeal duct cyst, pituitary gland, dog. A multilocular cyst is lined by respiratory epithelium and filled with mucus. HE.

4

5

6

7

8

9

10

predilection. Brachycephalic dogs may have been more likely to have pituitary macroadenoma, which warrants further inves- tigation. Hyperplastic nodules were more common than

pituitary adenomas in the postmortem specimens and were composed of basophilic to chromophobic cells with acidophils lightly interspersed, whereas adenomas were a homogeneous

Table 3. Histologic Diagnosis in Canine Pituitary Glands with Clinical Diagnosis of Hyperadrenocorticism (n = 129).
Pituitary Histologic DiagnosisHAC Not Suspected®HAC SuspectedªHAC ConfirmedªTotal
Within normal limits43 (80%)10 (19%)1 (2%)54 (100%)
Adenohypophyseal hyperplasia19 (83%)3 (13%)1 (4%)23 (100%)
Microadenoma13 (68%)4 (21%)2 (11%)19 (100%)
Macroadenoma6 (43%)1 (7%)7 (50%)14 (100%)
Miscellaneousb10 (100%)0 (0%)0 (0%)10 (100%)
Craniopharyngeal duct cysts9 (100%)0 (0%)0 (0%)9 (100%)

Abbreviation: HAC, hyperadrenocorticism.

aSee Materials and Methods for HAC diagnosis in 118 Purdue Veterinary Teaching Hospital dogs and II dogs treated by hypophysectomy.

bMiscellaneous diagnoses included other neoplasms (6 dogs) and inflammatory or degenerative lesions (4 dogs).

“This diagnosis was made in cases in which cysts >I mm diameter were the sole pituitary lesion.

Figure 11. The association of adrenocortical hyperplasia with ade- nohypophyseal proliferation. The percentage of dogs with adrenocor- tical hyperplasia increased as the degree of adenohypophyseal proliferation increased from hyperplasia through microadenoma to macroadenoma.

% Adrenocortical hyperplasia

100%

100%

80%

60%

57%

40%

40%

20%

0%

Hyperplasia (n=35)

Microadenoma (n=21)

Macroadenoma (n=4)

11

Adenohypophyseal proliferation

proliferation of a single cell type. As in human pathology,2,4 reticulin histochemistry highlighted the expansion of the reti- culin network in adenohypophyseal hyperplasia versus its dis- ruption in adenoma, and was particularly useful in distinguishing hyperplastic nodules that approached or exceeded 1 mm in diameter from small microadenomas. This distinction of hyperplasia from adenoma was particularly use- ful in hypophysectomy fragments, in which the diameter of a lesion was not apparent. Microadenomas were usually com- posed of densely granulated basophilic cells, whereas macro- adenoma, the most common diagnosis in hypophysectomy specimens, tended to be less well granulated and more chromophobic.

In a 1967 review,10 61 of 67 previously reported canine pituitary neoplasms had been classified as basophil (23), chro- mophobe (20), acidophil (4), or unspecified, mixed or pars intermedia (14) adenomas. In the same article, 26 newly reported adenomas were classified with a broad histochemical panel as 20 chromophobe adenomas and 6 pars intermedia adenomas. The authors concluded that most canine pituitary adenomas were composed of chromophobic cells, and in con- trast to previous reports, they had not encountered a canine basophil adenoma in their collective experience.

Although tinctorial classification of pituitary adenomas has been succeeded by immunohistochemistry for trophic hormone expression,14,28,31 the term “chromophobe adenoma” has lin- gered in veterinary medicine as the expected appearance of corticotroph adenoma, perhaps because larger adenomas can be sparsely granulated and chromophobic. However, many nonneoplastic corticotrophs, those in hyperplastic nodules or microadenomas, and even many of those in macroadenomas in the current study were densely granulated and at least faintly basophilic with HE. Furthermore, neoplastic cells in hypophy- sectomy specimens of 46 canine pituitary adenomas, although described as chromophobic, often had fine basophilic granules.20

Human corticotroph microadenomas are typically com- posed of densely granulated cells; the granules are basophilic or amphophilic, PAS-positive, and strongly immunoreactive for ACTH.31 In contrast, macroadenomas can be sparsely granulated and chromophobic.2,4 In human chromophobic tumors, PAS-reactivity of the granules and immunoreactivity for ACTH are weaker, but still helpful in identifying the cells as corticotrophs.2,29 However, canine melanotroph adenomas are also sparsely to densely granulated with weak to strong PAS and ACTH reactivity, so IHC for MSH is needed in a canine pituitary panel to distinguish melanotroph from corticotroph adenoma.

As in other studies,10,14,28 adenohypophyseal hyperplasia and adenoma were the most common canine pituitary diag- noses, and other pituitary or sellar neoplasms (germ cell tumor, craniopharyngioma, secondary neoplasms, granular cell tumor, and pituicytoma) were rare. Corticotroph carcinoma with sys- temic metastasis has been reported in a dog,13 but pituitary carcinoma of any type is exceedingly rare in any species,4,14,29 and was not observed in the current study.

Only 3 cases of hypophysitis were recorded in the current study, and in 2 cases, inflammation was part of meningoence- phalitis or systemic inflammation, in keeping with the litera- ture.28 Lymphocytic hypophysitis, seemingly primary, was observed in only 1 dog. Lymphocytic hypophysitis was diag- nosed in a hypophysectomy specimen from a dog with diabetes insipidus and a pituitary mass.18 Two postmortem cases of lymphocytic hypophysitis were accompanied by adrenalitis in 1 dog1 and adrenocortical atrophy in the other.32 Craniophar- yngeal duct cysts were relatively common but seemingly inci- dental lesions of the pars intermedia in this and in other canine14,20,28 or human2,4 case series.

Though results of this study may not reflect the prevalence of pituitary disease in the general canine population, 35% of postmortem cases in which the pituitary gland was examined had adenohypophyseal proliferation, and 15% had pituitary adenoma. The prevalence of pituitary adenoma has been esti- mated at 14% in human postmortem studies.12 Lactotroph ade- noma is the most common (and often incidental) human pituitary adenoma encountered postmortem; however, lacto- troph adenomas are usually treated medically and are rare among surgical specimens.4,5,29,31 In most surgical case series, somatotroph adenomas and corticotroph adenomas each account for up to 15% of human pituitary adenomas. 4,29,31 Melanotroph adenomas are generally absent in human case series, but relatively common in dogs.14,28

In another canine study,10 HAC was diagnosed in 14/20 dogs with chromophobe adenoma of the pars distalis and in 2/6 dogs with pars intermedia adenoma.

In the current study, HAC was confirmed in only a small proportion of the cases of adenohypophyseal proliferation, but was increasingly common with increasing degree of adenohy- pophyseal proliferation from no proliferation through hyper- plasia, microadenoma, and macroadenoma. In elderly humans, corticotroph hyperplasia is also seldom associated with HAC, but in a few cases was the only adenohypophyseal lesion in patients with Cushing’s disease.30

At autopsy of 318 laboratory beagles 6-17 years of age,7 10% had pituitary adenoma, and 17% had adrenocortical neo- plasia. Approximately half those dogs with adenohypophyseal or adrenocortical neoplasia had pathologic or clinical features of HAC. In the current study, all 6 of the postmortem cases with confirmed HAC had adrenocortical hyperplasia or neoplasia.

Hyperplastic nodules or small microadenomas can be missed in the plane of section on histologic examination of pituitary specimens.2,30 This could explain the absence of pitui- tary lesions in the current study in a dog with adrenocortical nodular hyperplasia and confirmed HAC. It is also possible that adrenocortical nodular hyperplasia, as opposed to diffuse adre- nocortical hyperplasia, could be difficult to distinguish from a small adrenocortical adenoma.8 Nevertheless, in the current study, the prevalence of adrenocortical hyperplasia increased with the increase in degree of adenohypophyseal proliferation from hyperplasia through microadenoma to macroadenoma.

In summary, adenohypophyseal proliferation was the most common canine pituitary lesion in this retrospective study. Hyperplastic nodules and adenomas were typically composed of basophilic to chromophobic cells with PAS-positive gran- ules, consistent with origin from corticotrophs or melano- trophs. Reticulin histochemistry was used to accentuate the loss of the reticulin network in adenoma versus its preservation in hyperplasia, and this facilitated recognition of adenoma in fragmented hypophysectomy specimens. In postmortem speci- mens, a diagnosis of hyperplasia or microadenoma was more common, but pituitary macroadenoma was the most common diagnosis in biopsy specimens. Dogs with a histologic classi- fication of adenohypophyseal proliferation were older and were more likely to have a clinical diagnosis of HAC or PDH. The

association between adenohypophyseal proliferation and hyperadrenocorticism underscores the importance of distin- guishing among hyperplasia, microadenoma, and macroadenoma.

Acknowledgements

We thank the histology technicians, pathologists, and pathology res- idents at the Indiana Animal Disease Diagnostic Laboratory and the Washington Animal Disease Diagnostic Laboratory; Connie Fraser, Washington State University; and Jan Shivers, University of Minnesota.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, author- ship, and/or publication of this article.

References

1. Adissu HA, Hamel-Jolette A, Foster RA. Lymphocytic adenohypophysitis and adrenalitis in a dog with adrenal and thyroid atrophy. Vet Pathol. 2010;47(6): 1082-1085.

2. Al-Brahim NYY, Asa SL. My approach to pathology of the pituitary gland. J Clin Pathol. 2006;59(12):1245-1253.

3. Alenza DP, Arenas C, Lopez ML, et al. Long-term efficacy of trilostane admi- nistered twice daily in dogs with pituitary-dependent hyperadrenocorticism. J Am Anim Hosp Assoc. 2006;42(4):269-276.

4. Asa SL. Pituitary histopathology in man: normal and abnormal. In: De Groot LJ, Beck-Peccoz F, Chrousos G, Dungan K, Grossman A, Hershman JM, Koch C, McLachlan R, New M, Rebar R, Singer F, Vink A, Weickert MO, eds. Endotext. South Dartmouth, MA: MDText.com, Inc; 2007.

5. Asa SL. Practical pituitary pathology: what does the pathologist need to know? Arch Pathol Lab Med. 2008;132(8):1231-1240.

6. Behrend EN, Kooistra HS, Nelson R, et al. Diagnosis of spontaneous canine hyperadrenocorticism: 2012 ACVIM consensus statement (small animal). J Vet Intern Med. 2013;27(6):1292-1304.

7. Berry M, Bruyette DS. A retrospective study of pituitary and adrenal neoplasia and the incidence of hyperadrenocorticism in a geriatric beagle colony (abstract). J Vet Intern Med. 1994;8:163.

8. Bruyette D. Canine pituitary dependent hyperadrenocorticism series. Part 2: diagnostic approach. Today’s Vet Pract. 2016;6(1):36-42.

9. Bruyette D. Canine pituitary dependent hyperadrenocorticism series. Part 3: current and investigative options for therapy. Today’s Vet Pract. 2016;6(2): 30-38.

10. Capen CC, Martin SL, Koestner A. Neoplasms in the adenohypophysis of dogs: a clinical and pathologic study. Path Vet. 1967;4:301-325.

11. den Hertog E, Braakman JC, Teske E, et al. Results of non-selective adreno- corticolysis by o, p’-DDD in 129 dogs with pituitary-dependent hyperadreno- corticism. Vet Rec. 1999;144(1):12-17.

12. Ezzat S, Asa SL, Couldwell WT, et al. The prevalence of pituitary adenomas: a systematic review. Cancer. 2004; 101(3):613-619.

13. Gestier S, Cook RW, Agnew W, et al. Silent pituitary corticotroph carcinoma in a young dog. J Comp Pathol. 2012;146(4):327-331.

14. Kiupel M, Capen C, Miller M, et al. Tumors of the hypophysis (pituitary gland) and sellar region. In: Histological Classification of Tumors of the Endocrine System of Domestic Animals. 2nd series, vol. 12. Washington, DC: Armed Forces Institute of Pathology; 2008:17-25.

15. Lien YH, Huang HP. Use of ketoconazole to treat dogs with pituitary-dependent hyperadrenocorticism: 48 cases (1994-2007). J Am Vet Med Assoc. 2008; 233(12):1896-1901.

16. Mamelak AN, Owen TJ, Bruyette D. Transsphenoidal surgery using a high definition video telescope for pituitary adenomas in dogs with pituitary dependent hypercortisolism: methods and results. Vet Surg. 2014;43(4): 369-379.

17. Martin LG, Owen TJ, Chen-Allen AV, et al. Clinical characteristics and out- come in dogs treated with transsphenoidal hypophysectomy for non-functional sellar masses. J Vet Intern Med. 2016;30(6):1941.

18. Meij BP, Voorhout G, Gerritsen RJ, et al. Lymphocytic hypophysitis in a dog with diabetes insipidus. J Comp Pathol. 2012;147(4):503-507.

19. Meij BP, Voorhout G, Rijnberk A. Progress in transsphenoidal hypophysectomy for treatment of pituitary-dependent hyperadrenocorticism in dogs and cats. Mol Cell Endocrinol. 2002;197(1-2):89-96.

20. Meij BP, Voorhout G, Van Den Ingh TSGAM, et al. Results of transsphenoidal hypophysectomy in 52 dogs with pituitary-dependent hyperadrenocorticism. Vet Surg. 1998;27(3):246-261.

21. Meij BP, Voorhout G, Van Den Ingh TSGAM, et al. Transsphenoidal hypophy- sectomy in beagle dogs: evaluation of a microsurgical technique. Vet Surg. 1997;26(4):295-309.

22. Mete O, Ezzat S, Asa SL. Biomarkers of aggressive pituitary adenomas. J Molec Endocrinol. 2012;49(2):R69-R78.

23. Neiger R, Ramsey I, O’Connor J, et al. Trilostane treatment of 78 dogs with pituitary-dependent hyperadrenocorticism. Vet Rec. 2002;150(26):799-804.

24. Osamura RY, Kajiya H, Takei M, et al. Pathology of the human pituitary adenomas. Histochem Cell Biol. 2008;130:495-507.

25. Owen TJ, Chen AV, Martin LG. Why, when, how and post-operative care of dogs undergoing transsphenoidal hypophysectomy for large sellar masses. J Vet Intern Med. 2016;30(6):1939.

26. Pérez-Alenza MD, Melián C. Hyperadrenocorticism in Dogs. In: Ettinger SJ, Feldman EC, Côté E, eds. Textbook of Veterinary Internal Medicine: Diseases of the Dog and the Cat. 8th ed., vol. 2. St. Louis, MO: Elsevier; 2017: 1796-1811.

27. Raverot G, Jouanneau E, Trouillas J. Clinicopathological classification and molecular markers of pituitary tumours for personalized therapeutic strategies. Eur J Endocrinol. 2014;170(4):R121-R132.

28. Rosol TJ, Gröne A. Endocrine Glands. In: Maxie MG, ed. Jubb, Kennedy, and Palmer’s Pathology of Domestic Animals. 6th ed., vol. 3. St. Louis, MO: Else- vier; 2016:269-356.

29. Saeger W, Lüdecke DK, Buchfelder M, et al. Pathohistological classification of pituitary tumors: 10 years of experience with the German pituitary tumor reg- istry. Eur J Endocrinol. 2007;156(2):203-216.

30. Syro LV, Rotondo F, Cusimano MD, et al. Current status on histological clas- sification in Cushing’s disease. Pituitary. 2015;18(2):217-224.

31. Syro LV, Rotondo F, Ramirez A, et al. Progress in the diagnosis and classifi- cation of pituitary adenomas. Front Endocrinol. 2015;6(97):1-8.

32. Wolfesberger B, Fuchs-Baumgartinger A, Schwendenwein I, et al. Sudden death in a dog with lymphoplasmacytic hypophysitis. J Comp Pathol. 2011; 145(2-3):231-234.