Glucocorticoid-dependent hypoadrenocorticism with thrombocytopenia and neutropenia mimicking sepsis in a Labrador retriever dog

Abstract

Glucocorticoid-deficient hypoadrenocorticism (GDH) with immune-mediated-neutropenia (IMN) and -thrombocytopenia (IMT) were diagnosed in a 3-year-old Labrador retriever dog. Glucocorticoid-deficient hypoadrenocorticism is rare and diagnostically challenging as clinical signs and laboratory abnormalities are often nonspecific. Immune-mediated cytopenias and other autoimmune disorders, as part of an autoimmune polyglandular syndrome have been reported with hypoadrenocorticism in humans. This is the first reported case of hypoadrenocorticism and bicytopenia in a dog.

A 3-year-old, 39.5-kg, castrated male, Labrador retriever dog was referred to the Small Animal Clinic at the Western College of Veterinary Medicine, University of Saskatchewan, for lethargy of 24 h duration, hypoglycemia, fever (40.0°C), and a single episode of vomiting and diarrhea. Abnormalities on initial blood work and urinalysis by the referring veterinarian included mild hypoglycemia, mild hypocholesterolemia, and isosthenuria [urine specific gravity (USG) 1.008]. Despite administration of intravenous (IV) fluids at a rate of 5 mL/kg body weight (BW)/h supplemented with 5% dextrose, and a single dose each of enrofloxacin (Bayer Healthcare; Toronto, Ontario), 2.5 mg/kg BW, IV, and ampicillin (Pfizer, Kirkland, Quebec), 20 mg/kg BW, IV, the dog’s clinical signs persisted and he was referred for further work-up with a suspected diagnosis of sepsis.

Case description

Physical examination upon presentation revealed a depressed, febrile (40.3°C) dog in good body condition (body condition score 3/5). The results of the remainder of the physical examination, including heart rate (92 beats per min) and respiratory rate (28 breaths per min), were unremarkable. The results of fundoscopic and complete neurological examinations were normal. Systemic arterial blood pressure was slightly elevated with average systolic and diastolic readings of 140 mmHg and 121 mmHg, respectively. Initial values for packed cell volume/total protein (PCV/TP), blood glucose, blood urea nitrogen (BUN) were normal. A venous blood gas analysis (Table 1; day 1) showed a compensated, mild mixed respiratory alkalosis and metabolic acidosis. Panleukopenia and mild non-specific biochemical changes including a mild elevation in creatinine were present (Tables 2 and 3; day 1). The USG was low (1.013), likely from prior IV fluid therapy. A coagulation panel, to look for evidence of early disseminated intravascular coagulation (DIC), showed an elevated partial thromboplastin time (PTT) with a normal prothrombin time (PT) (Table 2; day 1).

Table 1.

Serial blood values from a 3-year-old Labrador retriever dog with glucocorticoid-deficient hypoadrenocorticism, neutropenia, and thrombocytopenia

Parameter	Reference range for blood values	Day 1 value	Day 14 value
pH	7.31–7.42	7.365	7.287
PCO2 venous	35–45 mmHg	24.9	30.7
PO2 venous	35–45 mmHg	121.7	43.5
HCO3−	24 ± 4 mmol/L	13.5	13.8
Base excess	0 ± 3 mmol/L	−8.7	−10.3
Sodium	145–158 mmol/L	141.6	140.3
Potassium	3.8–5.6 mmol/L	4.19	4.21
Chloride	103–118 mmol/L	119	116
Calcium (ionized)	1.21–1.51 mmol/L	1.28	1.23
Lactate	1.25 +/− 0.46 mmol/L	0.80	1.67
Glucose	3.1–6.3 mmol/L	3.2	2.5
Rectal temperature (°C)	37.2–39.4	39.9	40.7

Table 2.

Serial hematological and coagulation values from a 3-year-old Labrador retriever dog with glucocorticoid-deficient hypoadrenocorticism, neutropenia, and thrombocytopenia

CBC	Reference range	Day 1	Day 2	Day 3	Day 4	Day 14	Day 21
WBC	4.8–13.9 × 109/L	1.3	17	21.5	10.3	1.2	12.3
RBC	5.2–8.2 × 109/L	6.61	5.74	5.57	6.37	5.27	5.48
HGB	128–196 g/L	145	126	121		112	117
HCT	0.365–0.573 L/L	0.42	0.37	0.36	0.40	0.325	0.34
MCV	65.2–73.6 fL	63.9	64.3	64	63.2	61.7	62
MCH	22.5–25.5 pg	21.9	22	21.8	21.8	21.3	21
MCHC	335–357 g/L	343	342	341	345	345	347
Reticulocytes	80–120 × 109/L	0	0	0	0	0.4	0.2
Segmented neutrophils	3.0–10.0 × 109/L	0.143	13.7	19.1	7.5	0.6	8.8
Band neutrophils	0.0–0.1 × 109/L	0.208	1.19	0.21	0	0	0
Toxic change		2+	2+	1+	ND	2+	ND
Eosinophils	0.0–1.1 × 109/L	0.026	0	0	0.1	0	0
Lymphocytes	1.2–5.0 × 109/L	0.845	1.36	0.86	1.5	0.875	1.2
Monocytes	0.08–1.0 × 109/L	0.078	0.68	1.2	1.13	0.26	2.2
Platelets	200–900 × 109/L	79	91	72	119	80.8	204
		Clumping present
Coagulation tests
Prothrombin time (s)	7.5–9.9	10.6	9.3	12.5	ND	9.3	ND
Partial prothrombin time (s)	9.6–13.8	19.0	19.1	12.5	ND	20.7	ND
Fibrin degradation products	< 10	< 10	< 10	< 10	ND	< 10	ND

CBC — complete blood (cell) count; WBC — white blood (cell) count; RBC — red blood (cell) count; HGB — hemoglobin; HCT — hematocrit; MCV — mean corpuscular volume; MCH — mean corpuscular hemoglobin; MCHC — mean corpuscular hemoglobin concentration; s — seconds; ND = not done.

Table 3.

Serial serum biochemical values from a 3-year-old Labrador retriever dog with glucocorticoid-deficient hypoadrenocorticism, neutropenia, and thrombocytopenia

Serum content	Reference range	Day 1	Day 2	Day 14	Day 193
Sodium (Na)	145–158 mmol/L	145	152	144	152
Potassium (K)	3.8–5.6 mmol/L	4.2	3.4	4.2	5
Chloride (Cl)	103–118 mmol/L	119	122	121	109
Bicarbonate	15–25 mmol/L	12	16	11	19
Anion Gap	16–30 mmol/L	18	17	16	29
Calcium (Ca)	1.91–3.03 mmol/L	2.36	2.4	2.34	2.68
Phosphorus (P)	0.63–2.41 mmol/L	1.0	1.17	1.14	0.88
Urea	3.5–11.4 mmol/L	3.6	1.7	3.5	4.1
Creatinine	41–121 μmol/L	126	98	113	72
Glucose	3.1–6.3 mmol/L	6.0	4.6	3.5	4.7
Cholesterol	2.70–5.94 mmol/L	1.93	2.86	2.18	6.33
Total bilirubin	1–4 μmol/L	3	13	25	3
ALKPHOS	18–128 U/L	134	143	140	345
ALT	19–59 U/L	104	119	96	115
CK	51–318 U/L	4138	893	538	137
GGT	0–8 U/L	4	2	3	10
Total protein	55–71 g/L	47	47	48	68
Albumin	28–38 g/L	22	21	24	37
Globulin	20–34 g/L	25	26	24	31

ALKPHOS — alkaline phosphatase, ALT — alanine aminotransferase, CK — creatine kinase, GGT — gamma glutamyltransferase.

Cervical and thoracic radiographs, an abdominal ultrasound, an echocardiogram, blood and urine cultures, and cytological analysis of lymph node and joint fluid aspirates were done to elucidate the cause of the dog’s fever and hypoglycemia; the findings were all unremarkable. Fecal cultures were negative. The cause of the presumed sepsis remained unexplained despite exhaustive testing. Nonspecific supportive treatments were initiated; they included administration of IV fluids (LRS supplemented with 20 mEq/L KCl (Hospira Health Care, Montreal, Quebec) at a rate of 3.7 mL/kg BW/h, and ranitidine (Sandoz, Boucherville, Quebec), 1.5 mg/kg BW, IV q12h as a non-specific gastrointestinal protectant. Towards the end of day 1, asymptomatic hypoglycemia [1.8 mmol/L; reference range (RR): 3.1 to 6.3 mmol/L] was documented on a glucometer, prompting addition of 2.5% dextrose solution to the IV fluids. Antibiotic treatment [enrofloxacin (Bayer Healthcare), 10 mg/kg BW, IV, q24h and ampicillin (Novapharm, Toronto, Ontario), 22 mg/kg BW, IV, q8h] was reinstated after samples for cultures of the urine, blood, and feces had been collected.

On day 2, leukopenia, neutropenia, and a degenerative left shift were replaced by a mild leukocytosis, mild neutrophilia, and a moderate left shift; however, 2+ toxic change persisted (Table 2). In addition, a mild thrombocytopenia with evidence of red blood cell fragmentation (keratocytes and shizocytes) had developed. Results of a serum biochemistry panel and clotting panel were similar to those for day 1 except for a dramatic decrease in creatinine kinase (CK) and development of mild hyperbilirubinemia (Tables 2 and 3; day 2). Due to the concern about DIC, the dog was premedicated with diphenhydramine (Sandoz), 1.2 mg/kg BW, IM, and a fresh-frozen plasma (FFP) transfusion was administered. Mini-doses of heparin (Leo Pharma, Thornhill, Ontario), 76 IU/kg BW, SQ, q8h were administered. On day 3 the leukogram was improved; neutrophilia accompanied by a reduced left shift and milder toxic change (1+) were present. However, worsening thrombocytopenia, persistent keratocytes, and a new finding of a mild, non-regenerative anemia supported a suspicion of ongoing systemic inflammation predisposing to DIC (Table 2; day 3).

A second FFP transfusion was initiated. A mild transfusion reaction, manifested by swollen eyelids and facial rubbing, occurred and responded well to a single dose of dexamethasone (Vétoquinol, Lavaltrie, Quebec), 0.1 mg/kg BW, IV. Coagulation parameters were normal on day 4 and the inflammatory leukogram had resolved with improvement in platelet numbers (Table 2; day 4). The dog was discharged on day 4 with a presumptive diagnosis of sepsis from an unknown cause that had responded to antibiotics. Enrofloxacin (Bayer Healthcare), 11 mg/kg BW, PO, q24h, and amoxicillin (Pfizer), 25 mg/kg BW, PO, q8h were prescribed for a further 10 d and a follow-up complete blood (cell) count (CBC) in 7 d was requested.

Despite an initial good response to symptomatic treatment, the dog was returned 9 d after discharge (day 14) for recurrence of fever and lethargy. At this time the owner raised the possibility of longer term disease as she believed the dog had “not been himself” since being boarded at a kennel 5 mo earlier and, over the last 7 mo he had lost almost 6 kg in BW. His BW indicated he had lost 1.1 kg since his last visit. The dog was depressed and febrile (40.6°C), mildly tachycardic (130 beats per min), and tachypnic (40 breaths per min). No other abnormalities were identified. Neutropenia with 2+ toxic change, thrombocytopenia, and a mild, microcytic, normochromic, and non-regenerative anemia were present on the CBC (Table 2; day 14). Abnormalities on the serum biochemistry panel continued to be mild and nonspecific but included a persistent hypocholesterolemia, mild hypoalbuminemia, and mild hyperbilirubinemia (Table 3; day 14). Repeat thoracic radiographs and abdominal ultrasound were normal, and blood, fecal, and urine cultures were negative. Later that day hypoglycemia was again detected along with a metabolic acidosis (Table 1; day 14), so 2.5% dextrose was added to the IV fluids.

A coagulation panel revealed prolonged PTT (Table 2; day 14). Cytological analysis of a bone marrow aspirate and biopsy showed appropriate cellularity with evidence of granulocytic and megakaryocytic hyperplasia, with no evidence of any infectious agents or neoplastic cells. This indicated a normal response to both the neutropenia and thrombocytopenia; the cause of the anemia was unclear. The dog was premedicated with diphenhydramine, 1.3 mg/kg, BW, IM, and another 200 mL FFP transfusion for possible DIC was administered. Titers for tick-borne diseases were not done as the dog had no travel history and no tick-borne diseases are endemic to the area.

An immune-mediated cause for the neutropenia was suspected, based on lack of support for either a neoplastic or infectious cause, so blood was collected and submitted for determination of the antinuclear antibody (ANA) titer and for anti-neutrophil antibodies using flow cytometry. Dexamethasone was administered, 0.2 mg/kg BW, IV, prior to these results being available. The dog continued to receive enrofloxacin (Bayer Healthcare), 10 mg/kg BW, IV q24h; ampicillin (Pfizer), 22 mg/kg BW, IV q8h, ranitidine (Sandoz), 1 mg/kg BW, IV q24h; and heparin (Leo Pharma), 75 IU/kg BW, SQ q24h. The ANA titer and direct and indirect anti-neutrophil titers were negative. An adrenocorticotrophic hormone (ACTH) stimulation test was done to rule out hypoadrenocorticism (HoAC) as the cause of the dog’s hypoglycemia, hypocholesterolemia, microcytic anemia, and vague gastrointestinal signs. Results of the ACTH stimulation test using synthetic ACTH (0.25 mg, IV) (Amphastar Pharmaceuticals, Rancho Cucamongo, California, USA) were consistent with HoAC (resting cortisol < 20 nmol/L, RR: 20 to 270 nmol/L; 1 h after ACTH cortisol < 20 nmol/L, RR: 230 to 570 nmol/L). Replacement therapy with oral prednisone (Apotex, Toronto, Ontario), 0.6 mg/kg BW, PO, q12h for 20 d was initiated for suspected glucocorticoid-deficient hypoadrenocorticism (GDH). The owner was instructed to decrease the dose to 0.5 mg/kg BW, q24h for 6 wk and then to 0.3 mg/kg BW, PO, q24h thereafter.

By the next day the dog was feeling better; his fever had resolved, there had been no recurrence of hypoglycemia, and his appetite and energy level were substantially improved. Intravenous fluids were discontinued. The leukon and thrombon on day 21 had normalized and the anemia was less severe (Table 2). Follow-up CBC’s on days 29, 60, 77, 109, and 180 after were normal, as were follow-up serum biochemistry panels on days 109 and 180 (data not included). The dog did well with treatment for his GDH for the next 6 mo; during this time his BW steadily increased to 44 kg. However, on day 193 he was presented to the referring veterinarian in hypothermic and hypovolemic shock; he died despite aggressive IV fluid therapy. Electrolytes were normal on a biochemistry panel done prior to death (Table 3; day 193). Necropsy revealed a small intestinal obstruction by a foreign body (plastic dog toy), widespread congestion and hemorrhage in the gastrointestinal tract, liver, thymus, and lungs, and fibrin thrombi in renal glomerular capillaries consistent with DIC. Both adrenal glands were atrophied. Microscopic examination confirmed fibrosis and selective loss of the zona fasciculata and zona reticularis with preservation of the zona glomerulosa and adrenal medulla. No other significant histological abnormalities were seen in other organs, but the pituitary was not examined histologically.

Discussion

This is the first report to describe a dog with concurrent peripheral blood cytopenias and HoAC: it is a rare disease in dogs (1–5). Most cases are the result of adrenal gland failure (primary HoAC); less commonly it is caused by hypothalamic or pituitary failure (spontaneous secondary HoAC), or it can be iatrogenic (acquired secondary HoAC) (1,3–6). Primary HoAC most commonly results from immune-mediated destruction of all layers of the adrenal cortex causing life-threatening miner-alocorticoid and glucocorticoid-deficient hypoadrenocorticism (MGDH) (1,3–5). When primary adrenocortical destruction targets only the zona fasciculata, or if the HoAC is secondary, a glucocorticoid only deficiency state results (GDH) (1,3–5). As in our dog, patients with GDH lack the classic electrolyte abnormalities (hyponatremia and hyperkalemia) seen with MGDH (1,3–5), and this contributes to the difficulty in diagnosing GDH (4,7–11). Most dogs with primary GDH go on to develop MGDH over time (4,8–9,11), but progression from GDH to MGDH was not documented over a 6-month follow-up period in this dog, and there was no evidence of destruction of the adrenocortical zona glomerulosa on the postmortem examination.

Based on the necropsy findings, secondary HoAC resulting from inadequate ACTH secretion cannot be conclusively ruled out in this case. Selective atrophy of the zone fasciculata and the reticularis with preservation of the outer zona glomerulosa seen on necropsy is consistent with either primary GDH or spontaneous secondary HoAC, since ACTH plays only a minor role in regulating adrenocortical aldosterone secretion (1,2,4,5,12). Iatrogenic secondary HoAC can result when exogenous glucocorticoid medications given long-term are abruptly discontinued (1,2,4,5). Iatrogenic secondary HoAC was unlikely in this dog because exposure to prior exogenous glucocorticoids was limited to 2 low doses of dexamethasone given during the diagnostic workup. There were no central nervous signs, or signs of injury to support spontaneous secondary HoAC from damage to the hypothalamus or pituitary gland from neoplasia, trauma, or inflammation. Differentiating between primary and secondary HoAC requires measurement of endogenous plasma ACTH concentration (1,2,13). This was not performed in our dog because it would have necessitated an unacceptable delay in treatment for GDH. Prior treatment with any type or dose of glucocorticoid predictably lowers the plasma ACTH concentration into the low normal or normal range, mimicking the results expected with secondary HoAC (1,2,13). Relative adrenal insufficiency associated with critical illness has been reported in dogs and humans (14–16), but was unlikely as the cause of HoAC in this dog for the following reasons: 1) the degree of adrenal insufficiency in response to ACTH stimulation seen with critical illness is not as dramatic as in our case, 2) we were unable to confirm infection, inflammation, or neoplasia consistent with critical illness, and 3) the postmortem examination supported a diagnosis of GDH (15,16).

As in this case, clinical signs associated with HoAC are nonspecific and may include lethargy, vomiting, diarrhea, anorexia, and abdominal pain (1,4,5,12). The nonspecific nature of the clinical signs and the fact that the signs often wax and wane until adrenal insufficiency progresses to an acute medical crisis (1,4,5,12) are major reasons why HoAC can be such a diagnostic challenge in dogs and humans (1–5,17). An acute adrenal crisis, from either GDH or MGDH, manifests as acute collapse due to either hemorrhagic or hypovolemic shock (1,4–5,12). In this case the dog was presented on 2 occasions with symptoms mimicking sepsis with secondary DIC (fever, hypoglycemia, neutropenia, elevated PTT, thrombocytopenia, hyperbilirubinemia) for which no microbiological etiology could be established (18). Cultures of blood, urine, feces, and joint fluid were all negative and there were no other major signs compatible with sepsis (hypotension, cold extremities, bounding pulses). Systemic inflammatory response syndrome (SIRS) resulting from sterile inflammation was also considered; however, a thorough diagnostic work-up failed to detect a site of significant inflammation (18). Despite not being as severe as in many dogs with immune-mediated thrombocytopenia, based on exclusion of other causes, it was concluded that the thrombocytopenia in this dog was triggered by immune-mediated disease. Granulocytic anaplasmosis caused by Anaplasma phagocytophilum is another possible explanation for the thromobocytopenia in this dog (19), but at the time of presentation this disease had not been reported in Saskatchewan and there were no morulae within granulocytes in the dog’s blood film. An explanation for the prolonged PTT in this dog on both occasions when he was hospitalized could be the presence of circulating anticoagulants as is seen in some human patients with autoimmune disorders (20). Fever is not commonly reported in dogs with HoAC (21), but is listed as a common clinical finding of HoAC in humans (22). The underlying cause for this dog’s fever was most likely cytokine release associated with immune-mediated disease (23).

Neutropenia is common in humans (24) but is uncommon in dogs (3,6–8,21) with HoAC. In fact, neutrophil counts are usually normal to slightly elevated in dogs with HoAC (6–8,21). Mechanisms for neutropenia include increased demand due to acute severe inflammation (infectious or non-infectious) with a shift of neutrophils from the circulating to the marginal and or tissue pools, immune-mediated destruction of circulating neutrophils, and decreased bone marrow production (25). Primary immune-mediated neutropenia in dogs is rare but reported (23,26–28). In our dog there was no history or evidence of toxic or drug-induced bone-marrow arrest, and there was no convincing evidence of an infectious, inflammatory, or neoplastic disease to initiate secondary peripheral immune destruction of the neutrophils. By exclusion of other causes, and based on response to corticosteroid therapy, a diagnosis of primary immune-mediated neutropenia (IMN) was established (23,26–29). The dramatic neutropenia in our dog is consistent with that seen in other dogs with IMN (median neutrophil count 0.1 × 10⁹/L) (23,26–29). While definitive diagnosis of IMN requires demonstration of antineutrophil antibodies, confirmatory testing is rarely done due to limited availability and the low reliability of these tests in veterinary patients (23,26–30). In the case described, both indirect and direct measurements of anti-neutrophil antibodies using flow cytometry were negative. However, severe neutropenia as seen in this case increases the chance of a false negative result by limiting material needed for the test. For these reasons, most dogs, including the dog in this case, are diagnosed with IMN based on exclusion of other causes and resolution of the neutropenia with immunosuppressive therapy (23,26–29).

Cyclic neutropenia was reported in a 17-year-old female human with MGDH, which was confirmed only after she had presented on a number of occasions to the emergency department in a hypoadrenal crisis, with each crisis being preceded by a drop in her blood neutrophil count (31). Sepsis was suspected on each occasion until the patient developed hyperkalemia and hyponatremia months after her initial presentation. Bone marrow cytology showed a maturation arrest in the neutrophil series in this patient (31). Cyclic neutropenia was considered in our patient, but follow-up CBCs after initiating treatment for GDH and suspected immune-mediated cytopenias did not show evidence of recurrent neutropenia, and the bone marrow examination when the dog was neutropenic did not show evidence of a maturation arrest in the neutrophil line. In the dog reported on here, the neutropenia definitely preceded and may have even triggered a hypoadrenocortical crisis on 2 occasions on which his symptoms and blood work abnormalities (hypoglycemia, neutropenia, thrombocytopenia, and prolonged PTT) mimicked sepsis. Hypoadrenocorticism has also been reported in humans mimicking sepsis, with the diagnosis being delayed based on apparent initial response to antibiotic therapy (32–34). After the diagnosis of HoAC was made in this dog, prednisone therapy at a low dose (1.2 mg/kg dose divided BID), resolved all the cytopenias. This dose is considered anti-inflammatory and not immunosuppressive, and is lower than doses traditionally used to treat dogs with presumed primary autoimmune cytopenias (1 to 2 mg/kg BID or > 30 mg/m² for a dog weighing > 30 kg) (28–29,35). However, there are no studies on the effect of prednisone dosages on short- and long-term outcomes in dogs with primary autoimmune cytopenias, and doses of 1 mg/kg BW/d are also commonly used to treat adult humans with autoimmune cytopenias (36). In addition, as in humans, there is considerable individual variability in dogs in their response to a similar dose of corticosteroids, so it is presumed that in this dog the dose was immunosuppressive.

Other hematologic abnormalities seen with HoAC include the absence of a stress leukogram and a mild to moderate normocytic, normochromic anemia (1,4–6,12). Absence of a stress leukogram, manifested as an eosinophilia and/or a lymphocytosis, however, is considered an insensitive clue for HoAC since this finding is reported in < 20% of dogs with HoAC in retrospective studies (1,3–4,9). The dog in this case had an appropriate eosinopenia and lymphopenia on days 1 and 14 when he was most acutely ill and when his neutropenia was most severe. Anemia has been reported in up to 26% of dogs with HoAC (3). The microcytic anemia is best explained by ongoing subclinical blood loss, but chronic inflammation may also have played a role. Although spherocytes were never reported in this dog, the unexplained mild bilirubinemia that occurred at the same time as the steady drop in the dog’s hematocrit, along with the detected schizocytes, keratocytes, and acanthocytes, also provide support for mild hemolysis contributing to the anemia (25). A Coombs test may have helped rule in autoimmune hemolysis. Concurrent immune-mediated hemolytic anemia has been reported in 1 dog with GDH (21), and concurrent immune-mediated anemia and thrombocytopenia have both been reported in people with HoAC (13,20). In addition, HoAC in humans commonly occurs in association with other endocrine and nonendocrine autoimmune diseases (20,37–39). The clinical response in the dog reported on suggests that immune-mediated cytopenias may have been concurrent with HoAC.

Biochemical abnormalities seen with HoAC include pre-renal azotemia, and a mild to moderate normal anion gap metabolic acidosis (1,2,4,6,14), hypercalcemia, hypoalbuminemia, hypoglycemia, and hypocholesterolemia (1–4,6). Hypocholesterolemia and hypoglycemia were marked in our dog and ultimately provided an important clue to the possibility of HoAC.

This case illustrates the diagnostic challenge of a rare but life-threatening disease such as GDH. Without a clinical index of suspicion for HoAC, patients with this disease are likely to be subjected to various and repeated testing, and specific treatment may be delayed (40–42). In addition to hyponatremia and hyperkalemia, which should prompt suspicion of MDGH, unexplained hypoglycemia, hypocholesterolemia, hypercalcemia, and/or signs that mimic sepsis despite failure to identify a source of infection should prompt suspicion of HoAC. Concurrent immune-mediated disorders, including cytopenias, should be considered in patients with HoAC.

To the authors’ knowledge, this is the first report of GDH with suspected concurrent immune-mediated neutropenia and thrombocytopenia in a dog.