Introduction

Hypophysitis refers to a group of rare inflammatory disorders of the pituitary gland, lymphocytic hypophysitis (LHP) being the most common histologic form identified in people.^1,2 This disease is defined as an immune-mediated lymphocytic infiltration of the pituitary gland, by a predominance of T cells, leading to destruction of the pituitary tissue and subsequent hormonal dysfunction.¹ Corticotrophs are predominantly affected, followed by gonadotrophs, thyrotrophs, somatotrophs, and ultimately antidiuretic hormone, causing diabetes insipidus (DI) when the neurohypophysis and/or the hypothalamus become infiltrated.¹ Headaches and visual impairment are commonly described in people, along with signs related to partial or total hypopituitarism.²

Despite its low prevalence (approximately one person in nine million individuals affected), LHP is a well-described condition in the human population. In dogs, it represents ∼0.4% (1/207) of canine pituitary lesions on postmortem examination,³ and veterinary literature regarding this condition is scarce. Few cases of presumed or confirmed hypophysitis are described in dogs, resulting in death in almost all cases.^4–8 This report describes clinical, biochemical, and advanced imaging features of a dog with histologically diagnosed lymphocytic panhypophysitis and panhypopituitarism, with a satisfying but transient response to medical management.

Case Report

A 3 yr old intact male Doberman pinscher dog was referred with a 3 mo history of diarrhea, along with a 2 wk history of lethargy, disorientation, increased appetite, and severe polyuria-polydipsia (PUPD) of acute onset. Most metabolic and infectious causes of diarrhea were already ruled out by the dog’s primary care veterinarian, with unremarkable hematology and biochemistry panels and negative coproscopy including Parvovirus, Giardia, Campylobacter, and Cryptosporidium polymerase chain reaction tests on feces. Various therapeutic trials with antibiotics, broad-spectrum deworming treatment, probiotics, and symptomatic treatment of diarrhea were already performed before referral without significant improvement. At referral, physical examination revealed moderate lethargy, aggressiveness, bilateral testicular atrophy, and hypothermia (96.8°F/36°C). Neurological examination was normal. Serum electrolytes, ammonia, and bile acids concentrations did not yield any abnormality. A serum hormonal panel was performed to explore the cause of PUPD and polyphagia, revealing decreased thyroxin levels (<6 nmol/L, reference interval [RI] 13–51) and basal cortisolemia (6.5 nmol/L, RI 50–200) with undetectable thyroid-stimulating hormone (<0.25 ng/mL, RI 0–0.7) and endogenous adrenocorticotropic hormone (<5 pg/L, RI 10–100) concentrations. Urinalysis showed hyposthenuria (urine specific gravity [USG] = 1.005, RI >1.025). Urinary tract infection was ruled out on cytologic and bacteriologic examinations of the urine sediment. Abdominal ultrasound findings included mild bilateral adrenal gland atrophy, marked prostate, and bilateral testicular atrophy.

As biochemical results were suggestive of central hypothyroidism, central hypoadrenocorticism, and DI, a 1.5-T brain MRI^a was performed and revealed an ill-defined lesion, involving both the pituitary gland and hypothalamus, that was hyperintense to the cortical gray matter on T2-weighted and fluid-attenuated inversion recovery sequences and hypointense on T1-weighted images. It showed marked heterogeneous contrast enhancement after IV gadoteric acid^b injection and was associated with a mass effect on adjacent tissue. A multifocal marked and homogeneous pachymeningeal and leptomeningeal thickening was observed (Figure 1). Cerebrospinal fluid (CSF) analysis from a lumbar puncture showed a mild lymphocytic pleocytosis (total nucleated cell count of 11/μL, RI <5/µL, with 94% small lymphocytes, RI <70%), with a mildly increased protein concentration (45 mg/dL, RI <40 mg/dL). A qualitative serologic leishmaniasis test (Speed Leish K) and a neurological polymerase chain reaction panel on CSF, submitted to detect the presence of canine distemper virus, Toxoplasma gondii, Neospora caninum, Leishmania infantum, and Ehrlichia sp., were negative. Based on these results, a presumed diagnosis of primary, immune-mediated lymphocytic hypophysitis, meningitis, and hypothalamitis causing panhypopituitarism was made. A round cell neoplastic process, such as histiocytic sarcoma or lymphoma, could not be ruled out considering the MRI findings.

View Figure

• Download as Powerpoint
• Download Figure

The dog was hospitalized as he developed moderate hypernatremia (163 mmol/L, RI 138–151), that resolved when desmopressin^c (3 µg/kg sublingual q 12 hr) was initiated. Thyroid hormonal supplementation (levothyroxine^d at 20 µg/kg orally q 24 hr) and immunosuppressive steroid therapy (dexamethasone^e 0.3 mg/kg, IV q 24 hr then prednisolone^f 1.5 mg/kg orally q 24 hr) were also initiated. The dog was discharged after 3 days of hospitalization.

At 15-day follow-up, PUPD had dramatically improved. Relapse was observed only when the administration of desmopressin was delayed by 2 hr. Although the lethargy and diarrhea had resolved, the dog persistently showed aggressiveness toward members of the family and other dogs. Polypnea was also reported and assumed to be steroid-induced. Six hours after medication administration, USG was 1.030, blood thyroxine levels were within the lower third of the RI (19 nmol/L), and serum electrolytes were normal. Therefore, the levothyroxine dosage was doubled (20 µg/kg orally q 12 hr), and prednisolone was decreased by 25% (1.1 mg/kg orally q 24 hr).

At 1 mo follow-up, polypnea had resolved. Persistent PUPD was noted in the evening. At 6 hr after medication administration, USG was 1.007 and blood thyroxine levels were within the lower half of the RI (25 nmol/L). Levothyroxine dosage was not changed and desmopressin dosage was increased by 50% (3 µg/kg sublingual q 8 hr).

At 2 mo follow-up, a relapse of PUPD and polyphagia was reported, attributed to corticosteroid adverse events, as urinalysis showed an almost adequate USG (1.021) without other abnormalities. Prednisolone dosage was decreased (0.85 mg/kg orally q 24 hr).

At 3 mo follow-up, PUPD, polyphagia, and aggressiveness were persistent. Repeat abdominal ultrasound showed persistent adrenal, prostate, and testicle atrophy. Brain MRI was repeated, showing complete resolution of signs of pituitary inflammation and meningeal thickening, along with atrophy of the pituitary gland. On repeat CSF analysis, lymphocytic pleocytosis had resolved (0 total nucleated cell per μl; protein concentration 25 mg/dL). Corticosteroid dosage was tapered over 6 wk to reach physiological supplementation. Persistent aggressiveness was deemed of behavioral origin or related to persistent ongoing hypothalamitis. As no improvement was reached over the following weeks, the dog becoming dangerous for people and other dogs at home, euthanasia was elected by the owners.

Necropsy was performed immediately after death with the owners’ consent. Brain tissue was submitted to the Institute for Animal Pathology at Ludwig Maximilian University (Munich, Germany). Histologic examination revealed chronic mild multifocal lymphocytic hypophysitis and mild focal lymphohistiocytic peri-infundibular meningoencephalitis. Multiple pituitary cysts were noted. Although the neurohypophysis was affected to a far lesser extent, there was some extension of the inflammation surrounding the infundibular recess, resulting in infundibuloneurohypophysitis. An attenuated stage of lymphocytic panhypophysitis was therefore identified (Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

Kidneys, liver, full-thickness digestive biopsies, and spleen were submitted to the Veterinarian Pathology Laboratory VetDiagnostics (Charbonnières-les-Bains, France) to rule out any lymphomatous infiltration. Adrenal glands were not sampled during the necropsy. None of the examined pieces showed significant abnormality. There was no evidence of neoplastic infiltration in any organ submitted for histological analysis.

Discussion

Only a few cases of confirmed or presumed primary hypophysitis are described in the veterinary literature, with an overall poor prognosis.^4–8 Among them, only one dog with presumed LHP (based on MRI and CSF findings) successfully responded to a corticosteroid treatment and remained free of signs 11 mo after the medication was discontinued.⁷ The other dogs suffered sudden death or were euthanized as a result of severe clinical deterioration. In people, the mortality rate is up to 6–7%, mostly owing to adrenal insufficiency.² Once primary hypophysitis is diagnosed and managed appropriately, the overall prognosis seems good.¹ The poor outcome in dogs may be explained by the low prevalence of the disease and associated difficulty to establish a diagnosis.

In people, patients are presented with signs related to pituitary deficiencies and often show headache (50%) and visual defect (10–30%).^1,2 In dogs, these clinical signs are likely to remain unnoticed by the owners, leading to a late diagnosis of the disease. Moreover, even when a pituitary function impairment is identified, diagnosing hypophysitis remains challenging as it requires a brain MRI combined with CSF analysis, and MRI findings of the pituitary gland with LHP are not pathognomonic of the disease.⁹ The main differential diagnoses for such pituitary lesions remain pituitary neoplasms, with macroadenomas being the most common.² Interestingly, most cases of canine LHP described in veterinary medicine were first mistaken as pituitary adenomas. Gutenberg et al. developed a clinical and MRI scoring system in people to distinguish LH from macroadenomas with a reported sensitivity of 92% and a specificity of 99%.¹⁰ Some of the features suggesting LHP such as young age, symmetric sellar and pituitary stalk enlargement, and intact sellar floor might be of use in dogs, but further studies are warranted.^7–9 A recent case series established an association between visceral leishmaniasis and the presence of lymphocytic pituitary infiltration in 21 dogs, but none of them showed neurological or hormonal-deficiency related signs.¹¹ In our case, leishmaniasis was deemed very unlikely based on a negative qualitative Leishmania infantum serology and the absence of amastigote organisms or pyogranulomatous inflammation in any organ on postmortem examination.

Our case showed atypical MRI findings. Although the expansile nature of the lesion with T2-weighted hyperintensity and T1-weighted hypointensity should make one consider a primitive suprasellar neoplasia, the ill-defined aspect is uncommon for such condition.¹⁰ Moreover, a severe thickening of the meningeal layers raised concerns for atypical nonspecific meningitis or a diffuse neoplasm such as histiocytic sarcoma or lymphoma.^12,13 Both are observed in aging dogs with a median age around 7–8 yr, and CSF analysis often shows neoplastic cells in lymphoma (68% of cases) or marked pleocytosis (mean count of 1509 cells/µL) with small and large mononuclear cells in histiocytic sarcoma.^12,13 As our dog was relatively young and CSF cytology showed mild pleocytosis without atypical cells, an inflammatory process was deemed more likely despite atypical MRI features. Unfortunately, no imaging or biochemical feature allows a certain distinction between LHP and pituitary neoplasms. Transsphenoidal biopsies have been described as a diagnostic tool in people, but they are rarely performed in both species because of their high complication rate.² Ultimately, repeat brain MRI at 3 mo showed a resolution of meningeal lesions. These findings further support the inflammatory nature of our dog’s condition, as both brain lymphoma (median survival time of less than 2 mo) and histiocytic sarcoma (median survival time of 3 days) are not expected to respond to steroid therapy for such a long time. A pituitary atrophy was also noted. This feature is often observed in people in late disease stages where chronic inflammation leads to fibrosis and subsequent shrinkage of the gland.¹

Definitive diagnosis was provided by the histologic examination, revealing an attenuated stage of lymphocytic panhypophysitis with a majority of CD3-positive T cells. Such a pattern is described in people² and in other dogs^7,11 with LHP, with a predominance of CD17 T-helper and CD3 T cells, suggesting an autoimmune process. This hypothesis is further supported in humans by a common association of LHP with concurrent or familial history of autoimmune diseases. Our dog showed multiple pituitary encapsulated cystic lesions that are likely caused by a loss of the tissue during phases of active inflammation. However, cystic lesions are not commonly observed in people, where the histologic endpoint of LHP is mainly pituitary fibrosis.¹ The latter usually inversely correlates with the degree of lymphocyte infiltration. Therefore, the attenuated aspect of the lymphocytic inflammation in our case may be linked to the chronicity of the disease or to a positive response to the ongoing immunosuppressive therapy.

Our dog showed a marked improvement of clinical signs once treated with hormonal supplementation and corticosteroids at an immunosuppressive dose. Historically, the treatment of primary hypophysitis in people only relied on hormonal supplementation, as spontaneous remission occurs quite frequently: 25–46% of cases exhibit radiological improvement and 15–33% had hormonal recovery.^1,14 However, recent studies revealed higher improvement and pituitary recovery rates with high-dose glucocorticoid therapy, leading to a change of practices in managing primary hypophysitis.^1,14,15 Earlier IV initiation of corticosteroid has also been associated with better pituitary function outcomes. Yet ∼70% of patients still require long-term replacement with at least one hormone, and DI rarely resolves, as documented in our dog. Interestingly, the presence of DI at the time of presentation, along with hypogonadism and specific MRI findings (thicker pituitary stalk, smaller pituitary volume, and evidence of posterior pituitary involvement) has been associated with a better clinical outcome in people.¹⁵ Controversially, DI is also reported as a negative factor in another study.¹⁶ Azathioprine, methotrexate, mycophenolate mofetil, and rituximab have also been described in managing refractory hypophysitis.²

Persistence of histological lymphocytic inflammation of the pituitary gland despite resolution of hormonal deficits should raise awareness of the need for long-term high-dose corticotherapy and/or other immunosuppressive drugs in these patients. The cause of worsening aggressiveness in our dog remained unclear. As an attenuated stage of lymphocytic panhypophysitis was identified at histological examination, such aggressiveness could be a clinical feature of persistent active lymphocytic inflammation. However, behavioral changes secondary to LHP are not described in people and were not yet reported in dog cases.^1,4–8 Brain areas associated with anxiety and aggressive behaviors in dogs, such as the limbic system and caudolateral part of the hypothalamus, showed no lesion on histologic examination in our dog. Aggressiveness was therefore more likely attributed to a corticosteroid side effect or a behavioral origin. An imbalance in thyroid function may be involved, but hypothyroidism was deemed well controlled in our dog, and evidence of a link between aggressive behavior and hypothyroidism or levothyroxine therapy remains unproven in dogs.¹⁷ On the other hand, psychiatric adverse events among humans on glucocorticoid therapy are widely reported, with patients with pre-existing psychiatric disorders being more at risk of developing behavioral side effects under such treatment.¹⁸ Moreover, recent studies show that dogs receiving corticosteroids tend to be less playful and more nervous and prone to react aggressively.¹⁹ Glucocorticoid therapy might have been a contributing factor to the worsening of a pre-existing aggressiveness in our dog, as such behavior had been reported in previous veterinary consultations. Furthermore, the owners reported the arrival of a newborn at home. Lifestyle changes and the presence of children have been reported as potential triggers for aggression in dogs²⁰ and could be another contributing factor in our case. Unfortunately, tapering the prednisolone dosage did not lead to any improvement of the aggressive behavior. Psychotropic drugs such as selective serotonin reuptake inhibitors or tricyclic antidepressants have shown some benefit in treating aggression in dogs²⁰ and could have been considered in our case.

Conclusion

Canine hypophysitis might carry a slightly better prognosis than what was previously described in dogs, as a partial clinical and histological remission was observed in our case. A few features such as young age, symmetric sellar and pituitary stalk enlargement, homogeneous gadolinium enhancement, and intact sellar floor on MRI investigation might help diagnosing hypophysitis in dogs, but further studies are warranted. Hypophysitis should be included in the differential diagnosis of pituitary lesions or deficiencies, even given its low prevalence, as it may offer a better prognosis than some pituitary masses when managed accordingly with immunosuppressive and hormone replacement therapies. Bimodal immunosuppressive treatment might be indicated as a first-line medical approach in those cases. Adjustment of corticotherapy remains challenging as some treatment side effects can also be attributed to poor control of LHP (including PUPD, behavioral change) and active pituitary inflammation can persist despite resolution of MRI lesions. Therefore, long-term immunosuppressive therapy and MRI follow-up are advocated in these patients.

The authors would like to thank Pr. Kaspar Matiasek for the histopathological examination and interpretation of this case.