Presumptive Postoperative Nontraumatic Adrenal Haemorrhage After Orthopaedic Surgery in a Dog

ABSTRACT

Background

Nontraumatic adrenal haemorrhage (NTAH) is rarely reported in dogs and may be associated with clinically important adrenal dysfunction when both adrenal glands are affected. Surgery has been recognised as a potential precipitating stressor in humans, but postoperative NTAH has been rarely described in veterinary medicine.

Case Presentation

A 7‐year‐old, 2.8‐kg spayed female Chihuahua developed abrupt decompensation on postoperative day 1 after bilateral distal femoral osteotomy, characterised by hypothermia, hypotension, and recurrent hypoglycaemia. Clinicopathologic abnormalities included anaemia, thrombocytopenia, hypoalbuminaemia, increased C‐reactive protein concentration, and increased D‐dimer concentration. Abdominal ultrasonography identified bilateral adrenal enlargement with hypoechoic parenchyma and marked periadrenal hyperechogenicity, supporting a presumptive diagnosis of postoperative NTAH.

Treatment and Outcome

Prednisolone and supportive care resulted in rapid clinical stabilisation. Basal cortisol, initially within the reference interval, subsequently declined below the assay detection limit (<1.0 µg/dL), accompanied by a decrease in the sodium‐to‐potassium ratio. Serial ultrasonography documented progressive reductions in adrenal thickness, with the left adrenal thickness decreasing below commonly cited ultrasonographic cutoffs used to support hypoadrenocorticism. Cortisol and electrolyte indices normalised after discontinuation of glucocorticoids.

Conclusion

This case highlights presumptive postoperative NTAH as an important differential diagnosis in dogs with abrupt postoperative decompensation and supports the value of serial ultrasonography combined with endocrine and electrolyte monitoring.

A small‐breed dog developed abrupt hypothermia, hypotension, and recurrent hypoglycaemia on postoperative day 1 after orthopaedic surgery. Serial ultrasonography showed bilateral adrenal thickening with periadrenal hyperechogenicity, followed by progressive reduction in adrenal thickness. Concurrent transient cortisol and sodium‐to‐potassium ratio changes supported presumptive postoperative nontraumatic adrenal haemorrhage with transient adrenal dysfunction

.

1. Introduction

Nontraumatic adrenal haemorrhage (NTAH) has been reported rarely in critically ill human patients and may lead to clinically important adrenal insufficiency when both adrenal glands are affected (Ali et al. 2018; Di Serafino et al. 2017). The exact pathogenesis is incompletely understood, but stress‐related physiologic changes – including increased adrenal vascularisation, oedema, venous stasis, and subsequent haemorrhage – have been proposed (Di Serafino et al. 2017; Elhassan et al. 2023; Kawashima et al. 1999). Reported predisposing factors include sepsis, hypotension, recent surgery, and coagulopathy. Because clinical signs and routine clinicopathologic abnormalities are often nonspecific, NTAH may mimic other severe postoperative or critical illness syndromes (Ali et al. 2018; Di Serafino et al. 2017).

Because early manifestations can be ambiguous, diagnostic imaging is important for early recognition (Elhassan et al. 2023; Kawashima et al. 1999). Abdominal ultrasound can identify early adrenal changes and is practical for serial monitoring; computed tomography, when available, can further characterise adrenal haemorrhage and evaluate for an underlying adrenal lesion (Ali et al. 2018; Kawashima et al. 1999). In veterinary medicine, NTAH with clinically relevant adrenal dysfunction has been reported rarely (Lee et al. 2022). To our knowledge, postoperative cases – particularly those characterised by serial ultrasonographic adrenal changes coupled with a transient course of adrenal dysfunction – are limited. We report a small‐breed dog that developed abrupt postoperative decompensation after orthopaedic surgery, with ultrasonographic adrenal changes consistent with a presumptive diagnosis of postoperative NTAH and a transient course of adrenal dysfunction that improved during serial monitoring.

2. Case Description

A 7‐year‐old spayed female Chihuahua (2.8 kg) was referred for evaluation of acute deterioration 1 day after orthopaedic surgery. The dog had no history of underlying disease, including endocrine disease, and was not receiving prescription medications, including glucocorticoids, nonsteroidal anti‐inflammatory drugs, or anticoagulants, before surgery.

Preoperative laboratory testing was performed on the day of surgery, and selected clinicopathologic variables for comparison with the referral presentation are summarised in Table 1.

TABLE 1.

Selected clinicopathologic variables before surgery and at presentation to the referral centre (postoperative day 1; case day 0) in a dog with presumptive postoperative nontraumatic adrenal haemorrhage.

Variable	Day of surgery	Referral presentation (POD1; day 0)
Glucose (mg/dL)	144	47 a
PCV (%)	58.6	30.8
Platelets (×103/µL)	176	91
Total protein (g/dL)	6.8	4.4
Albumin (g/dL)	3.7	2.0
CRP (mg/L)	10	164

Abbreviations: Alb, albumin; CRP, C‐reactive protein; PCV, packed cell volume; PLT, platelet count; TP, total protein.

^a Point‐of‐care glucose.

The dog underwent distal femoral osteotomy (DFO) for bilateral grade 4 medial patellar luxation. Premedication consisted of butorphanol (0.1 mg/kg IV) and midazolam (0.2 mg/kg IV). Anaesthesia was induced with propofol IV, administered incrementally to effect; up to 6 mg/kg was prepared for induction. Anaesthesia was maintained with isoflurane in oxygen. Perioperative antimicrobial prophylaxis consisted of cefazolin (22 mg/kg IV), administered within 30 min before skin incision. Surgical time was approximately 3 h. No intraoperative complications or immediate anaesthetic recovery abnormalities were reported.

The dog recovered uneventfully from anaesthesia and remained clinically stable through the morning of postoperative day 1 (POD1), when it ate normally. At approximately 3 PM on POD1, the dog developed acute lethargy with hypothermia (37.2°C), hypotension (Doppler systolic blood pressure, 60 mm Hg), and hypoglycaemia (point‐of‐care blood glucose, 35 mg/dL). Oral 50% dextrose and an IV bolus of 10% dextrose resulted in transient improvement, but hypoglycaemia recurred and referral was initiated. The dog arrived at our hospital at approximately 8 PM on POD1, and hypothermia, hypotension, and hypoglycaemia persisted.

On presentation to our hospital (POD 1; day 0), point‐of‐care blood glucose was 47 mg/dL (RI, 75–128). Compared with preoperative values, clinicopathologic abnormalities included anaemia, thrombocytopenia, hypoproteinaemia, hypoalbuminaemia, and a marked increase in CRP concentration (Table 1). Serum electrolyte concentrations included mild hyponatraemia (136 mmol/L; RI, 139–150) with potassium of 3.8 mmol/L (RI, 3.4–4.9), yielding a sodium‐to‐potassium (Na:K) ratio of 35.78. Basal serum cortisol concentration, measured before prednisolone administration, was 1.3 µg/dL (RI, 1.0–6.0). D‐dimer concentration was increased (1028 ng/mL; RI, 0–250). A complete leukocyte differential was not performed; therefore, eosinophilia, lymphocytosis, or absence of a stress leukogram could not be assessed. Other serum biochemical variables, including cholesterol and renal and hepatobiliary parameters, were within their respective reference intervals.

Abdominal ultrasonography on day 0 showed adrenal thickness of 5.9 mm (left) and 6.3 mm (right). The adrenal parenchyma was hypoechoic, and periadrenal fat was markedly hyperechoic (periadrenal hyperechogenicity) (Figure 1A, E). No discrete adrenal mass was identified. Based on the abrupt POD1 clinical deterioration (hypothermia and hypotension), recurrent hypoglycaemia, basal cortisol, and the ultrasonographic findings, postoperative NTAH was suspected.

FIGURE 1.

Serial ultrasonographic images of the adrenal glands in a dog with presumptive postoperative nontraumatic adrenal haemorrhage. (A–D) Left adrenal gland; (E–H) right adrenal gland. Images were obtained on day 0 (A, E; presentation), day 5 (B, F), day 12 (C, G), and day 27 (D, H). On day 0, both adrenal glands were surrounded by markedly hyperechoic periadrenal fat (white arrows), and adrenal thickness subsequently decreased on follow‐up examinations with resolution of the periadrenal changes.

On day 0, supportive care was continued, including dextrose supplementation and IV fluid therapy with lactated Ringer's solution supplemented with 5% dextrose (D5‐LRS). Ampicillin–sulbactam (20 mg/kg IV q12h) was administered for 5 days. Prednisolone was administered at 0.5 mg/kg PO q12h for 2 days and then decreased to 0.25 mg/kg PO q12h for a total treatment duration of 5 days, after which it was discontinued. Hypothermia, hypotension, and hypoglycaemia resolved the following day.

Basal cortisol and serum electrolyte concentrations were monitored. Basal cortisol concentrations were 2.1 µg/dL on day 1, 2.2 µg/dL on day 3, 1.1 µg/dL on day 5, and <1.0 µg/dL on day 12 (Figure 2A). The Na:K ratio was 28.60 on day 5 (Na 143 mmol/L; K 5.0 mmol/L) and 29.38 on day 12 (Na 144 mmol/L; K 4.9 mmol/L) (Figure 2B). Despite the low basal cortisol concentration on day 12, prednisolone was not restarted because the dog remained clinically stable without recurrent hypoglycaemia, hypotension, or clinically relevant electrolyte abnormalities, and close monitoring was continued. Thereafter, basal cortisol increased to 2.3 µg/dL on day 19 and 3.9 µg/dL on day 27, and the Na:K ratio was 34.19 on day 19 and 35.71 on day 27 (Figure 2A, B).

FIGURE 2.

Serial changes in basal serum cortisol concentration (A) and sodium‐to‐potassium (Na:K) ratio (B) in the same dog. The shaded box indicates the period of prednisolone treatment from day 0 to day 5. In panel A, the day 12 cortisol concentration was below the assay detection limit (<1.0 µg/dL) and is shown as an open circle. In panel B, the dashed horizontal line indicates a Na:K ratio of 27.

Serial abdominal ultrasonography was performed (Figure 1). Adrenal thickness on day 5 was 3.3 mm (left) and 4.5 mm (right). On day 12, adrenal thickness was 2.7 mm (left) and 3.5 mm (right). On day 27, adrenal thickness was 2.6 mm (left) and 3.5 mm (right), and measurements remained similar on subsequent rechecks. Thus, the left adrenal thickness decreased to below commonly cited ultrasonographic cutoffs used to support hypoadrenocorticism, whereas the right adrenal thickness remained above this range.

Serial haematologic and biochemical reassessments were performed. D‐dimer concentration was within the reference interval on day 2 (214 ng/mL; RI, 0–250). The platelet count was 161 ×10³/µL (RI, 117–490) on day 4. Total protein and albumin concentrations on day 12 were 6.4 g/dL (RI, 5.0–7.2) and 3.9 g/dL (RI, 2.6–4.0), respectively. The PCV on day 27 was 42.4% (RI, 32.5–58). After prednisolone discontinuation, hypoglycaemia, hypotension, or electrolyte abnormalities did not recur, and the dog remained clinically stable without ongoing therapy through approximately 6 weeks after surgery.

3. Discussion

Nontraumatic adrenal haemorrhage (NTAH) is an uncommon but potentially life‐threatening condition that may be clinically overlooked because early signs and routine clinicopathologic abnormalities are often nonspecific, particularly in postoperative or critically ill patients (Ali et al. 2018; Di Serafino et al. 2017; Kawashima et al. 1999; Lee et al. 2022). Early recognition is clinically important because adrenal haemorrhage can be associated with clinically relevant adrenal insufficiency when both adrenal glands are involved, and delayed corticosteroid supplementation in adrenal crisis may adversely affect outcome (Ali et al. 2018; Elhassan et al. 2023).

The diagnostic relevance of this case lies in the temporal convergence of abrupt POD1 shock‐like deterioration, evolving adrenal ultrasonographic abnormalities, and a transient endocrine/electrolyte pattern compatible with adrenal dysfunction. This combined pattern helped distinguish presumptive postoperative NTAH from nonspecific postoperative complications and supported serial ultrasonography with endocrine/electrolyte monitoring as a practical diagnostic approach (Kawashima et al. 1999; Lee et al. 2022; Wenger et al. 2010). Although definitive confirmation was not possible without computed tomography, histopathology, or ACTH stimulation testing, the parallel clinical, ultrasonographic, and endocrine/electrolyte evolution supported a presumptive diagnosis of postoperative NTAH with transient adrenal dysfunction.

Postoperative adrenal haemorrhage has been reported rarely in humans, supporting surgery as a plausible precipitating stressor even when the procedure is unrelated to the adrenal glands (Andres et al. 2023; Belmore et al. 1995; Dahan et al. 2016; McNicol et al. 2014). Although the exact pathogenesis remains incompletely defined, a ‘stress‐adrenal’ framework has been proposed in which stress‐related physiologic changes – including increased adrenal blood flow and vascularisation, oedema, venous stasis, and microthrombotic or haemostatic perturbations – predispose the adrenal microvasculature to haemorrhage (Ali et al. 2018; Andres et al. 2023; Di Serafino et al. 2017; Kawashima et al. 1999). Adrenal haemorrhage may be under‐recognised because its clinical manifestations overlap with other postoperative syndromes, including pain, hypovolaemia, anaesthetic effects, gastrointestinal complications, and systemic inflammatory responses (Ali et al. 2018; Andres et al. 2023; McNicol et al. 2014). The present case extends the veterinary literature by describing presumptive NTAH occurring after orthopaedic surgery, characterised by serial ultrasonographic adrenal changes paired with a transient endocrine/electrolyte phenotype that normalised without ongoing glucocorticoid supplementation.

In dogs, adrenal haemorrhage and clinically relevant adrenal dysfunction appear to be rare, and published veterinary reports most commonly involve severe systemic infection (e.g. Waterhouse–Friderichsen–like syndromes) or iatrogenic/diagnostic interventions, as well as diverse adrenal pathologies (Cole et al. 2021; Lee et al. 2022; Wyatt et al. 2024). In the limited veterinary case literature, as well as in the broader human literature, imaging has been emphasised as a means to raise suspicion early and to track lesion evolution (Kawashima et al. 1999; Lee et al. 2022; Wenger et al. 2010). The serial ultrasonographic pattern in the present dog – initially increased adrenal thickness bilaterally with prominent periadrenal hyperechogenicity, followed by progressive decreases in adrenal thickness, including left adrenal thickness below commonly cited hypoadrenocorticism‐supportive cutoffs on later examinations – was consistent with an acute adrenal insult with subsequent remodelling. Although adrenal ultrasonography cannot definitively distinguish haemorrhage from other acute adrenal insults such as inflammation or necrosis, ultrasonographic characterisation of adrenal gland lesions can aid diagnostic assessment in dogs when interpreted together with clinical and endocrine findings (Pagani et al. 2016; Reagan et al. 2022). A similar trajectory (enlargement/haemorrhage followed by later atrophy) has been described in a dog with presumed spontaneous bilateral NTAH preceding hypoadrenocorticism (Lee et al. 2022). Although computed tomography can further characterise adrenal haemorrhage and evaluate for an underlying lesion in humans (Ali et al. 2018; Andres et al. 2023; Kawashima et al. 1999), abdominal ultrasound is practical for bedside assessment and serial monitoring in small animal patients and can provide a clinically useful time‐course when paired with endocrine and electrolyte indices (Lee et al. 2022; Reagan et al. 2022; Wenger et al. 2010).

Interpretation of basal cortisol in this case warrants careful contextualisation. Basal cortisol is commonly used as a screening (rule‐out) test for naturally occurring hypoadrenocorticism in dogs, with higher concentrations providing strong negative predictive value, whereas low or low‐normal values are nonspecific and require clinical correlation and, when feasible, ACTH stimulation testing (Bovens et al. 2014; Lennon et al. 2007). ACTH stimulation testing was not performed during the acute phase because immediate stabilisation was prioritised after collection of the initial basal cortisol sample, and dynamic endocrine testing could not be arranged before glucocorticoid supplementation. This precluded definitive assessment of adrenal functional reserve and represents an important limitation of this report. Importantly, the clinical question in this dog was not whether a single resting cortisol value met a numerical threshold, but whether the adrenal response was appropriate for the degree of physiologic stress and whether the endocrine trajectory aligned with the clinical course and imaging evolution (Martin 2011). In a dog experiencing abrupt postoperative shock‐like decompensation with recurrent hypoglycaemia and a concurrent decline in sodium‐to‐potassium ratio, basal cortisol values that remain within reference limits yet potentially inappropriately low given the severity of physiologic stress, followed by a transient result below the assay detection limit (<1.0 µg/dL), support an episode of adrenal dysfunction (absolute or relative) rather than an appropriately mounted stress response (Martin 2011). Subsequent recovery of basal cortisol and electrolyte indices without ongoing glucocorticoid supplementation further supports a transient course of adrenal dysfunction rather than persistent primary hypoadrenocorticism. Prednisolone was not resumed after the low day 12 basal cortisol result because the dog remained clinically stable without recurrent hypoglycaemia or hypotension, and the electrolyte changes were not progressive. Therefore, continued monitoring was considered more appropriate than automatic continuation of glucocorticoid therapy based on a single basal cortisol value.

Several alternative or concurrent possibilities were also considered as differential diagnoses. First, postoperative inflammation can account for some clinicopathologic abnormalities. In dogs undergoing surgical procedures, CRP commonly rises rapidly and then declines in an uncomplicated postoperative course; therefore, a marked early increase does not, in isolation, confirm infection (Christensen et al. 2015; Kanno et al. 2019). Accordingly, the CRP elevation in this dog was compatible with a postoperative acute‐phase response, and CRP was interpreted in conjunction with the overall time‐course (clinical status, haemodynamics, imaging, serial trends) rather than as a standalone discriminator between sterile inflammation and infection. Second, occult infection or sepsis could not be excluded definitively and remains an important differential, particularly because sepsis is a recognised predisposing factor for adrenal haemorrhage in humans and veterinary patients (Ali et al. 2018; Cole et al. 2021). However, the temporal pattern – initially stable postoperative period followed by abrupt deterioration on POD1 – and the subsequent serial adrenal ultrasonographic evolution were most consistent with an acute adrenal event rather than progressive infectious decompensation alone. Third, perioperative haemostatic activation with transient consumption related to acute tissue injury and/or haemorrhage was considered, given the thrombocytopenia and increased D‐dimer concentration, and could have contributed to adrenal vascular injury (Ali et al. 2018; Andres et al. 2023). However, because coagulation testing was incomplete, these findings were interpreted conservatively as supportive but not diagnostic of a defined coagulopathy (e.g. DIC).

Fourth, critical illness–related corticosteroid insufficiency (CIRCI) was considered because it can present with inadequate corticosteroid activity in severely ill patients (Martin 2011). However, CIRCI is a functional construct and would not be expected to account for the conspicuous serial structural adrenal changes observed here, including initially increased adrenal thickness with prominent periadrenal hyperechogenicity followed by progressive decreases in adrenal thickness, with the left adrenal thickness falling below commonly cited hypoadrenocorticism‐supportive cutoffs on later examinations. These findings favoured a structural adrenal insult.

Fifth, perioperative haemorrhage and haemodilution could contribute to anaemia and hypoalbuminaemia, but would not readily explain the combined evolution of adrenal imaging abnormalities and the endocrine/electrolyte trajectory.

Treatment and outcome in this case underscore two practical points. First, rapid clinical stabilisation after glucocorticoid supplementation and supportive care is consistent with clinically meaningful adrenal dysfunction in the acute phase, but therapeutic response alone should not be construed as diagnostic (Martin 2011). Second, serial monitoring was pivotal. The dog's endocrine and electrolyte indices normalised without ongoing glucocorticoid administration, suggesting recovery of clinically relevant adrenal function.

This report highlights presumptive postoperative NTAH occurring after orthopaedic surgery as an important differential diagnosis in dogs with abrupt postoperative decompensation, especially when shock‐like signs (hypotension, hypothermia), recurrent hypoglycaemia, and evolving adrenal ultrasonographic abnormalities are present. Serial abdominal ultrasound combined with serial endocrine/electrolyte monitoring may provide a practical framework for early recognition and follow‐up in similar cases. Limitations include the lack of preoperative baseline adrenal ultrasonography, absence of a complete leukocyte differential, incomplete infectious and coagulation evaluation, absence of cross‐sectional imaging, absence of specific mineralocorticoid function testing, and potential confounding effects of glucocorticoid administration on subsequent endocrine testing; thus, definitive confirmation of adrenal haemorrhage and determination of the precise mechanism underlying the transient adrenal dysfunction were not possible. Because a complete leukocyte differential was not performed, leukogram features that may support glucocorticoid‐deficient hypoadrenocorticism, such as eosinophilia or absence of a stress leukogram, could not be evaluated (Reagan et al. 2022). The initial basal cortisol concentration was measured before prednisolone administration and was therefore not affected by prednisolone cross‐reactivity. However, subsequent cortisol measurements obtained during or shortly after prednisolone treatment require cautious interpretation for two reasons. First, short‐term prednisolone administration could have contributed to hypothalamic–pituitary–adrenal axis suppression, as glucocorticoid‐induced adrenocortical suppression has been documented in dogs, and recovery after intermediate‐acting glucocorticoid treatment can be variable (Chastain and Graham 1979; Del Baldo et al. 2024). Second, the degree of cross‐reactivity between prednisolone and the cortisol assay used was not determined; if cross‐reactivity occurred, measured cortisol concentrations during treatment may have overestimated endogenous cortisol, and the true cortisol concentrations may have been lower than reported. Therefore, serial cortisol values were interpreted together with the clinical course, sodium‐to‐potassium ratio, and adrenal ultrasonographic changes rather than as definitive endocrine confirmation alone. The subsequent decline in the Na:K ratio was considered supportive because crystalloid fluid therapy and the weak mineralocorticoid activity of prednisolone would be expected, if anything, to improve rather than worsen this ratio. However, the Na:K ratio did not decrease below 27:1, and specific assessment of mineralocorticoid function, such as aldosterone concentration or plasma renin activity, was not performed because immediate stabilisation was prioritised and the electrolyte abnormalities subsequently resolved. Therefore, mineralocorticoid dysfunction could not be definitively confirmed.

4. Conclusion

This case supports presumptive postoperative NTAH as an important differential diagnosis in dogs with abrupt postoperative decompensation. A small‐breed dog developed hypothermia, hypotension, and recurrent hypoglycaemia on postoperative day 1, with serial ultrasonographic findings consistent with an acute adrenal insult and subsequent remodelling. In parallel, basal cortisol declined transiently to below the assay detection limit and the Na:K ratio decreased, followed by normalisation during follow‐up. Serial abdominal ultrasonography combined with endocrine and electrolyte monitoring may facilitate early recognition, guide timely supportive care (including glucocorticoid supplementation when clinically indicated), and help document recovery in similar cases.

Author Contributions

Byung‐Jun Kim: conceptualisation, investigation, data curation, formal analysis, writing – original draft, writing – review and editing. Hyoung‐Sun Cho: investigation, writing – review and editing. Kun‐Ho Song: supervision, conceptualisation, writing – review and editing.

Funding

This research was funded by the Rural Development Administration, Republic of Korea (grant number: RS‐2023‐00231965).

Ethics Statement

The authors confirm that the ethical policies of Veterinary Medicine and Science have been adhered to. Ethical approval was not required for this clinical case report involving client‐owned animals.

Patient Consent Statement

Written informed owner consent was obtained for publication of this case report and associated images.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: vms371028‐sup‐0001‐SuppMat.mp4

Supporting File 1: vms371028‐sup‐0002‐SuppMat.mp4

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.