Central diabetes insipidus in a cat with central nervous system B cell lymphoma.

A 6-year-old male neutered cat presented with blindness, lethargy, polydipsia, hyposthenuria and severe hypernatraemia. Central diabetes insipidus was demonstrated by means of a low measured anti-diuretic hormone (ADH) concentration in the face of hypernatraemia, and clinical response to supplementation with desmopressin. Magnetic resonance imaging of the brain showed a discrete mass in the region of the hypothalamus. The cat was euthanased and post-mortem histological examination demonstrated B cell lymphoma involving the brain, optic nerves, urinary bladder wall and diaphragm. To the authors' knowledge, this case report is the first to describe central diabetes insipidus caused by central nervous system lymphoma in the cat.

A 6-year-old male neutered cat was presented to the referring veterinary surgeon for blindness. Fundoscopic examination by a registered veterinary ophthalmology specialist identified patchy chorioretinitis in both retinas. These changes were considered consistent with protozoal chorioretinitis, but it was felt they were not severe enough to cause blindness. Toxoplasma serology was run, and IgG was positive at 1:1024, and IgM was positive at 1:128. Treatment was initiated with prednisolone (Delta-Cortef; Upjohn) 0.8 mg/kg q 12 h and clindamycin (Antirobe; Pfizer) 25 mg/kg q 12 h. Within 3 days, the cat’s vision returned, and the cat remained clinically normal for approximately 1 month. The prednisolone dosage was decreased to q 24 h after 2 weeks, and q 48 h after 4 weeks. Within 2 weeks of discontinuing the prednisolone, the cat was represented to the referring veterinary surgeon with blindness and severe lethargy. At that time, the cat was referred to Melbourne University Veterinary Hospital.

On presentation, the cat was quiet but responsive and ambulatory. The mucous membrane colour, capillary refill time, pulse quality and skin turgour were all normal. Mean systolic blood pressure (Doppler, Parks, Oregon, USA) was 145 mmHg. The cat was clinically blind. The pupils were fixed and dilated, and pupillary light responses, menace and dazzle reflexes were absent bilaterally. Fundus examination demonstrated numerous grey ‘blebs’ and ‘trails’ on the tapetal fundus. Other than the blindness, neurological examination was unremarkable. When a water bowl was offered to the cat it drank compulsively and unremittingly until the bowl was removed. This behaviour was observed repeatedly.

The results of pathology tests are attached (Table 1 ). The significant abnormalities were marked hypernatraemia, hyperchloraemia, hypoglycaemia, increased total serum calcium, increased phosphate, increased alanine aminotransferase (ALT), markedly increased creatine kinase (CK), increased total protein, hyperalbuminaemia, neutropenia, lymphopenia and marked hyposthenuria. A seropositive toxoplasma status was confirmed, but other infectious disease testing was negative.

Table 1

Results of pathology tests.

Haematology	Result	Reference range	Biochemistry	Result	Reference interval
Haemoglobin (g/dl)	13.0	10.0–15.0	Sodium	178	150–165
Packed cell volume (PCV) (l/l)	0.41	0.30–0.45	Potassium	4.7	3.5–5.8
Red cell count (× 1012/l)	7.7	5.0–10.0	Chloride	141	112–129
Mean cell volume (MCV) (PCV/RCC) (fl)	58	37–49	Calcium	3.14	1.75–2.50
Mean corpuscular haemoglobin (MCH) (Hb/RCC) (pg)	17	13–17	Phosphate	2.4	1.3–2.3
White cell count (× 109/l)	2.50	5.5–19.5	Urea	13.6	5.4–10.7
Bands (× 109/l)	0.15	0–0.13	Creatinine	0.28	0.07–0.16
Neutrophils (× 109/l)	2.23	2.5–12.5	Glucose	2.8	3.9–7.5
Lymphocytes (× 109/l)	0.10	1.5–7.0	Cholesterol	4.7	1.9–3.9
Monocytes (× 109/l)	0.02	0–0.9	Total bilirubin	2	0–15
Eosinophils (× 109/l)		0–1.5	ALT	115	5–80
Basophils (× 109/l)		Rare	ALP	14	10–120
Platelets (× 109/l)	Adequate	300–700	CK	2603	50–400
nRBC (/100 WBC)		Rare	Total protein	83	56–80
Reticulocytes (/100 RBC)		0–0.4	Albumin	39	22–35

ACTH = adrenocorticotropic hormone, ALT = alanine aminotransferase, ALP = alkaline phosphatase, CK = creatine kinase.

Solid phase chemiluminescent competitive immunoassay on Immulite, Siemens Healthcare, Gwynedd, UK.

ACTH DA Kit, MP Biomedicals, Solon, OH.

Solid phase chemiluminescent, competitive immunoassay on Immulite, Siemens Healthcare, Gwynedd, UK.

Nichols method of equilibrium dialysis, Idexx laboratories, Brisbane, QLD, using radioimmunoassay, Antech, Irvine, CA.

Chemiluminescent competitive immunoassay on Immulite, Siemens Healthcare, Gwynedd, UK.

Double antibody radioimmunoassay, Buhlmann, Schonenbuch, Switzerland.

SensPERT FeLV Ag/FIV Ab test kit, VetAll, Kyunggi-Do, Korea.

Immunofluorescent antibody test, Mt Pleasant laboratory, Launceston, TAS, using Anti-Feline fluorescent conjugate, Kirk Garden Perry, Gaithersburg, MD.

Feline infectious peritonitis virus immunofluorescent antibody kit, Fuller, Fullerton, CA.

Abdominal ultrasound and thoracic radiographs were unremarkable.

Possible causes for hypernatraemia in general consist of pure water loss, hypotonic fluid loss, or increased sodium intake. In the normovolaemic patient, pure water loss is the most likely cause. Clinical examination indicated normovolaemia. Underlying differentials for pure water loss include primary hypodipsia/adipsia, lack of access to water, fever, burns, panting, secondary nephrogenic diabetes insipidus (NDI), central diabetes insipidus (CDI), or primary NDI (rare). Of these differentials, only diabetes insipidus would account for hypernatraemia and hyposthenuria.

Intravenous fluid therapy was commenced at 5 ml/kg/h. The fluid type initially chosen was lactated Ringer’s solution (LRS) with 2.5% dextrose supplemented with additional NaCl to produce a sodium concentration of 160 mmol/l. This was undertaken with the aim of decreasing the serum sodium concentration at a rate no faster than 0.5 mEq/l/h. Electrolytes were rechecked repeatedly during this infusion (see Table 1). After the first 2 h, the sodium concentration increased to 182 mmol/l, and oscillated between 182 and 185 mmol/l throughout the first 24 h of treatment. For this reason the sodium concentration of the intravenous fluid was serially decreased. The cat remained clinically the same throughout this phase of the treatment.

Blood was submitted for anti-diuretic hormone (ADH) assay. The serum sodium of the same sample was 185 mmol/l. The ADH concentration of this sample was 1.0 pmol/l. Although no published reference interval is available for the cat, one study demonstrated a mean ADH value in water restricted cats of 78 pmol/l. Accordingly, the result in this case was inappropriate in the face of hypernatraemia and supported a diagnosis of CDI.

Exogenous desmopressin was administered (Minirin solution; Ferring AB, Sweden) one drop (1.5–4 mg) in the conjunctival sac. Two hours later sodium was 171 mmol/l and the urine specific gravity (USG) of a free-catch urine sample was 1.010. Although the rate of decrease in sodium concentration was undesirably rapid, there was no clinical deterioration immediately or over the subsequent 24 h. The sodium concentration 6 h after vasopressin administration was 169 mmol/l, with a USG of 1.012, and 12 h after administration the sodium was 166 mmol/l. The same dose of vasopressin was subsequently administered twice daily, and the sodium concentration remained between 157 and 165 mmol/l over the subsequent 24 h.

At the same time as desmopressin was commenced a single dose of dexamethasone sodium phosphate 0.1 mg/kg was given by intravenous injection. The cat’s vision returned, there was a partial return of the pupillary light responses, its demeanour improved, and it began to eat voluntarily.

Magnetic resonance imaging (MRI) of the brain was performed on the following day (Fig 1 ). This showed a rounded mass lesion, approximately 10.5 mm diameter, ventral to the thalamus, rostrodorsal to and apparently continuous with the pituitary gland. The mass was causing dorsal displacement of the thalamus; on the sagittal plane there was caudodorsal displacement of the interthalamic adhesion. Differentials considered were neoplasia, particularly meningioma or pituitary macroadenoma, or granuloma.

Fig 1

MRI scan of the head of a 6-year-old neutered male domestic shorthair cat with CDI: post-contrast transverse slice at the level of the diencephalon. A well-defined space-occupying lesion (10 mm high, 10 mm wide) is evident in the hypothalamic–pituitary region (arrows).

Based on the MRI findings, and despite the clinical improvement, the owners elected to have the cat euthanased and consented for necropsy. At necropsy, grossly there was a round mass bulging from the base of the brain at the level of the optic chiasm (Fig 2 ).

Fig 2

Necropsy specimen of the brain of a 6-year-old neutered male domestic shorthair cat with CDI: transverse slice at the level of the diencephalon. A well-defined space-occupying lesion (10 mm high, 10 mm wide) is evident in the hypothalamic-pituitary region (arrows).

Other pertinent gross findings were that the adrenal cortices were thinned with a cortical to medullary ratio of 1:4, and there were depressions in the surface of both kidneys, with loss of parenchyma and replacement with fibrous tissue.

The tissues were fixed in neutral buffered 10% formalin and embedded in paraffin, followed by routine sectioning and staining with haematoxylin and eosin (HE). Histologically, there was a fairly well demarcated, approximately 1 cm diameter round mass centred over the area of optic chiasm and extending caudally into the hypothalamus. The mass consisted of dense sheets of neoplastic round cells supported by minimal vascular stroma. There was a very high mitotic rate [76/10 high powered fields (HPF)], which included abnormal mitoses. The round cells had a small amount of amphophilic cytoplasm and distinct cell borders. The nuclei were round to indented with vesicular chromatin. The nuclei contained a large, prominent nucleolus. There was moderate anisokaryosis and anisocytosis, with cells ranging in size from approximately 15–25 μm in diameter. Scattered throughout the mass were moderately sized areas of necrosis and elsewhere, focal infiltrates of macrophages.

Immunohistochemistry was performed on deparaffinised tissue sections following citrate buffer microwave antigen retrieval. The sections were treated with fetal calf blocking serum and incubated overnight in primary antibody at 4°C. Universal labelled streptavidin biotin (LSAB) reagents were used and detection was by a Nova Red substrate kit. The primary antibodies used were polyclonal rabbit anti-human CD3 and purified mouse monoclonal anti-PAX5 at dilution factors of 1:5. Positive control tissue was from a canine lymph node. The neoplastic round cells had positive nuclear staining for PAX5 and negative cytoplasmic staining for CD3. The sections were counterstained with Gill’s haematoxylin (Fig 3 ).

Fig 3

Hypothalamus of a 6-year-old neutered male domestic shorthair cat with CDI: the infiltrative neoplasm is causing lateral deviation of the third ventricle. The small inset at the upper right shows a higher magnification of the neoplastic B cells (HE).

Histologically, the optic chiasm, tuber cinereum, infundibulum (pituitary stalk) and ventral hypothalamus [including supraoptic nuclei (SON) and paraventricular nuclei (PVN)] were effaced by the tumour. The pars distalis remained, although was poorly demarcated from the neoplasm. The dorsal third of the hypothalamus was predominantly intact, while containing moderate numbers of large, white matter vacuoles. The subfornical organ (SFO) remained histologically intact. There was marked Wallerian degeneration of the optic tracts. Neoplastic lymphocytes were also detected in the diaphragm, urinary bladder wall and around the optic nerves, just caudal to the eyes. There was no histological evidence of toxoplasmic chorioretinitis and histological changes within the eyes were minimal, showing only small areas of cataract formation.

The mechanisms that regulate thirst and ADH regulation in the mammalian brain have been reported in detail.4, 5 Increased serum osmolarity triggers both thirst and ADH release in the normal animal. Sodium is the main determinant of osmolarity, and so hypernatraemia results in an increased cerebrospinal fluid osmolarity, and this is detected by a series of so-called circumventricular organs. These organs, located in the anterior wall of the third ventricle, are the organum vasculosum of the lamina terminalis (OVLT) and the SFO. The median preoptic nucleus (MnPO) is also involved in thirst detection, and is located inside the blood–brain barrier, but is not considered part of the circumventricular organ system. In rodent models, experiments have demonstrated connections between these organs and thirst centres elsewhere in the brain.

ADH is produced by the SON and PVN in the hypothalamus, from where it is transmitted to, and subsequently released from, the posterior pituitary. The SON and PVN are osmoreceptive, and hyperosmolarity directly stimulates ADH production. Additionally, the SON and PVN receive input from the circumventricular organs described above.

In summary, hypernatraemia results in two distinct but interrelated processes: the stimulation of thirst, and the release of ADH. In the normal animal, these responses occur simultaneously, but it is possible to experimentally decouple these processes, and cause a selective incapacity of either thirst or ADH production (Fig 4 ).

Fig 4

MRI of the brain of 6-year-old neutered male domestic shorthair cat with CDI, sagittal reconstruction at the level of the midline, with diagram showing the anatomic distribution of the circumventricular osmoreceptor organs in the rat for comparison. (Reproduced with kind permission of M McKinley.)

In this cat, severe hypernatraemia resulted in compulsive thirst, but not increased serum ADH concentrations. A lesion that destroyed the organs of ADH production, but preserved the mechanism of thirst, would be consistent with these observations. The identified tumour had obliterated the region containing the PVN and SON, as well as the ventral circumventricular region, but had preserved the more dorsal circumventricular region. It was hypothesised that part or all of the SFO or MnPO may have been spared, which would be sufficient to preserve the thirst response, a hypothesis supported by the finding of an histologically intact SFO. The MnPO could not be confidently identified histologically, which is unsurprising, as it is normally not a histologically distinctive structure.

A previous published report of CNS lymphoma in a cat described severe hypernatraemia with a concurrent lack of thirst. That case is probably best described as hypodipsic hyponatraemia (HH), which may or may not have had concurrent CDI as well. ADH was not tested, nor any supplementary ADH administered in that case. In order for thirst to be lacking in that case, we hypothesise that the OVLT, MnPO and SFO were damaged by that tumour.

To the authors’ knowledge this is the first reported case of CDI caused by intracranial neoplasia in a cat. CDI has been associated with B cell lymphoma in a dog, and has also been described in several case series and reports in cats.8, 9, 10, 11, 12, 13, 14, 15 The underlying cause of CDI in the feline cases has been reported as congenital,10, 14 traumatic14, 15 or idiopathic.

The position of the lesion raised the possibility that other aspects of hypothalamopituitary function could be impaired. The results of an adrenocorticotropic hormone (ACTH) stimulation test, endogenous ACTH, thyroxine, and thyroid stimulating hormone (TSH) assessments were not supportive of this possibility. The testing of other hormones, including growth hormone and IGF-1 could also be considered in such a case but were not taken in this instance due to time, financial and sample handling constraints.

The blindness in this cat was likely due to direct injury to the optic chiasm and adjacent optic tracts. The retinal changes observed initially were consistent with toxoplasmal chorioretinitis, but that is not likely to have been the main cause of blindness in this patient. This is supported by the lack of a definitive improvement from medical management of toxoplasmosis with clindamycin, and a lack of supportive histological changes.