Pulsus alternans in a critically ill dog hospitalized for xylitol toxicity

Abstract

A 2-year-old spayed female Great Pyrenees cross dog was presented following the consumption of pure xylitol sweetener. Blood tests revealed hepatocellular leakage and cholestasis, hyperlactatemia, thrombocytopenia, and prolonged prothrombin and activated partial thromboplastin times. Thoracic radiographs on day 2 of hospitalization were consistent with pulmonary hemorrhage. Prior to death, the dog developed pulsus alternans suggestive of myocardial dysfunction secondary to severe systemic inflammation. This is the first report of pulsus alternans in a critically ill dog prior to clinical deterioration and death. This is also the first documentation of pulsus alternans with a high-definition oscillometric device.

Key clinical message:

Increased recognition of pulsus alternans and its potential implications in veterinary medicine may contribute to the identification of cardiovascular complications associated with systemic disease.

First described in 1872, pulsus alternans (PA) is a rarely reported phenomenon defined as the alternation of high and low pulse amplitude in the face of a regular cardiac rhythm (1). Often an indicator of severe ventricular dysfunction and myocardial depression, PA has been associated with a poor prognosis in the human literature (2–4). Despite multiple experimental studies investigating the underlying pathophysiological mechanisms of PA in dogs (5–8), this finding has only rarely been reported in the clinical veterinary setting (9–11). Furthermore, PA has yet to be reported in the absence of primary cardiac disease (10), general anesthesia (9), or anticholinergic therapy (11) in veterinary medicine.

In the human critical care setting, it has been suggested that PA may serve as an important clinical finding that warrants further investigation; however, its significance may often be overlooked (4,12). This report describes the clinical course of a dog with acute hepatic failure and coagulopathy secondary to xylitol intoxication that developed PA prior to severe clinical deterioration and death.

Case description

A 2-year-old spayed female Great Pyrenees cross dog (body weight 31.9 kg) was referred to the Veterinary Medical Centre (VMC) following ingestion of an estimated 190 to 380 g [5.96 to 11.91 g/kg body weight (BW)] of pure xylitol sweetener alongside one other dog in the household. Both dogs were suspected to have contributed to the consumption; consequently, the true consumed dose was unknown. The dog had been unsupervised for 7 to 8 h before owner identification of the ingestion upon returning home. No clinical signs were noted at this time. The dog was previously healthy with no medical concerns.

The dog was immediately brought to the primary care veterinarian. On presentation, the dog was bright, alert, and responsive. Mucous membranes were tacky. The dog had a complete blood (cell) count (CBC) performed that identified a normal white blood cell (WBC) count [6.23 × 10⁹ cells/L, reference interval (RI): 6.00 to 17.00 × 10⁹ cells/L], lymphopenia (0.30 × 10⁹ cells/L, RI: 0.83 to 4.91 × 10⁹ cells/L), and an elevation in hematocrit (HCT: 0.591 L/L, RI: 0.33 to 0.56 L/L). The automated platelet count revealed thrombocytopenia (69 × 10⁹ cells/L, RI: 117 to 490 × 10⁹ cells/L). No manual platelet evaluation was performed at this time. The remainder of the CBC was unremarkable. One episode of vomiting occurred at the primary care veterinary clinic prior to referral to the VMC. No medical or fluid therapy was initiated.

On presentation to the VMC (day 1), the dog was ambulatory and had appropriate mentation. Mucous membranes were dark pink and tacky with a capillary refill time of 3 s. The dog had a normal heart rate (132 beats/min) and rhythm and had strong, synchronous femoral pulses. No heart murmur was detected on cardiac auscultation. The dog was tachypneic (40 breaths/min), and on abdominal palpation was tense and sensitive to palpation of the cranial abdomen. The remainder of the physical examination was within normal limits.

An arterial blood gas and packed cell volume (PCV), total protein, blood glucose, and blood urea nitrogen (BUN) (Azostix Reagent Strips; Siemens Canada, Oakville, Ontario) were determined immediately on presentation. Arterial blood gas analysis revealed a metabolic acidosis (pH: 7.327, RI: 7.350 to 7.450) with respiratory compensation (PaCO₂: 26.0 mmHg, RI: 32 to 43 mmHg; PaO₂: 102.9 mmHg, RI: 80 to 105 mmHg), hypobicarbonatemia (13.1 mmol/L, RI: 18 to 26 mmol/L), a low base excess (−9.7 mmol/L, RI: −5.0 to 1.0 mmol/L), hyponatremia (143.7 mmol/L, RI: 145 to 151 mmol/L), hyperglycemia (12.2 mmol/L, RI: 3.6 to 6.2 mmol/L), and hyperlactatemia (7.1 mmol/L, RI: 0.5 to 2.0 mmol/L). The emergency panel revealed an elevated PCV (58%, RI: 37% to 55%). The remaining parameters were within normal limits. Coagulation parameters were also evaluated. Prothrombin time (PT) and activated partial thromboplastin time (aPTT) were both prolonged (PT: 28 s, RI: 11 to 17 s; aPTT: 132 s, RI: 72 to 102 s) and a manual platelet count confirmed the presence of thrombocytopenia.

The dog had an IV catheter placed and was started on Normosol R (Abbott Laboratories, Mississauga, Ontario), 11.5 mL/kg BW per hour, supplemented with 20 mEq/L potassium chloride (KCl Concentrate 2 mEq/mL; Hospira, Saint-Laurent, Quebec) and 2.5% dextrose (Baxter, Mississauga, Ontario). This rate was reduced to 3.7 mL/kg BW per hour after 4 h. A loading dose of acetylcysteine (Acetylcysteine Inj. 200 mg/mL; Alveda Pharmaceuticals, Toronto, Ontario), 140 mg/kg BW, IV was administered once followed by 70 mg/kg BW, IV, q6h. A combination of s-adenosylmethionine, silybin, and phospholipids (Zentonil Advanced, 400 mg; Vétoquinol, Lavaltrie, Quebec), was administered at 12.5 mg/kg BW, PO, q24h. The dog was monitored hourly for mucosal bleeding and every 2 h for blood glucose concentration. Despite 1 episode of vomiting and 1 episode of hypoglycemia that resolved without intervention (2.9 mmol/L, RI 3.6 to 6.2 mmol/L), the dog remained stable in the ICU overnight.

The next morning (day 2), the dog was noted to be hypothermic (36.9°C) and a venous blood gas and emergency panel were repeated. Venous blood gas revealed hypobicarbonatemia (15.8 mmol/L, RI 18 to 26 mmol/L), a low base excess (−6.7 mmol/L, RI: −5 to 1 mmol/L), hypokalemia (3.68 mmol/L, RI: 3.9 to 5.1 mmol/L), and ionized hypercalcemia (1.41 mmol/L, RI: 1.16 to 1.40 mmol/L), in addition to hyperlactatemia (3.5 mmol/L, RI: 0.5 to 2.0 mmol/L). Emergency panel revealed hyperproteinemia (75 g/L, RI: 51 to 72 g/L) and icteric serum. Vitamin K1 (Phytonadione Vitamin K1 10 mg/mL; WCVM Pharmacy Compounding, Saskatoon, Saskatchewan) supplementation was started at 2 mg/kg BW, SC, q24h and maropitant citrate (Cerenia 10 mg/mL; Zoetis, Kirkland, Quebec) was started at 0.5 mg/kg BW, IV, q24h. The dog ate a small amount of food after the morning assessment; however, she proceeded to regurgitate following consumption. Hypoglycemia (2.9 mmol/L, RI: 3.6 to 6.2 mmol/L) was transiently noted; however, it resolved 2 h later (5.3 mmol/L, RI: 3.6 to 6.2 mmol/L) without additional dextrose supplementation.

The dog was closely monitored in the ICU and that afternoon high-definition oscillometry (HDO) (VET HDO Monitor; S + B metVET GmbH, Babenhausen, Hesse, Germany) was initiated for non-invasive blood pressure monitoring. Continuous ECG inspection revealed sinus tachycardia at 184 beats/min; however, the HDO monitor was simultaneously recording a pulse of 92 beats/min with a reliable waveform. A subjectively slow pulse rate was detected on manual palpation of peripheral arterial pulses. Comparison of the ECG waveform to the HDO monitor showed that every alternate QRS complex was registering as a significantly smaller pulse wave amplitude on the HDO blood pressure monitor. Pulsus alternans was identified as alternating strong and weak pressures on the HDO waveform resulting in the inaccurate pulse rate measurement (Figure 1). Blood pressure monitoring with the HDO system was discontinued at this time.

Figure 1.

Non-invasive high definition oscillometric blood pressure waveform of (A) a dog hospitalized for xylitol toxicity with pulsus alternans and simultaneous sinus tachycardia on visual ECG inspection and of (B) a healthy adult dog. Pulsus alternans (A) is defined by the presence of alternating strong and weak pulses, represented by the alternation of tall and short vertical pulse amplitudes, in the face of a regular cardiac rhythm. Comparatively, vertical pulse amplitudes of the healthy adult dog (B) consistently reach the horizontal trend line of the oscillometric waveform with no evidence of regular alternation.

Serum biochemistry, PT and aPTT time were evaluated within the hour. Abnormalities are summarized in Table 1. Serum biochemistry showed evidence of hepatocellular leakage and cholestasis in addition to elevations in creatinine and creatine kinase. The dog had phosphorus and blood glucose concentrations within normal limits. Both PT and aPTT were significantly prolonged beyond the upper limit of detection.

Table 1.

Abnormalities on serum biochemistry and coagulation panel (PT, aPTT) on day 2 of hospitalization in a dog with xylitol toxicity that went on to develop pulsus alternans prior to cardiac arrest.

Parameter	Value (reference interval)
Na:K ratio	39 (28 to 38)
Anion gap	27 mmol/L (12 to 26 mmol/L)
Creatinine	123 μmol/L (41 to 121 μmol/L)
Amylase	3.52 U/L (5.83 to 23.38 U/L)
Cholesterol	8.92 mmol/L (2.70 to 5.94 mmol/L)
Total bilirubin	88.5 μmol/L (1 to 4 μmol/L)
Direct bilirubin	43.1 μmol/L (0 to 2 μmol/L)
Indirect bilirubin	45.4 μmol/L (0 to 2.5 μmol/L)
ALP	3.43 U/L (0.15 to 1.5 U/L)
GGT	0.18 U/L (0 to 0.13 U/L)
ALT	383.47 U/L (0.32 to 0.98 U/L)
GLDH	23.52 U/L (0 to 0.12 U/L)
CK	22.53 U/L (0.85 to 6.97 U/L)
PT	> 60 s (7.5 to 9.9 s)
aPTT	> 60 s (9.6 to 13.8 s)

ALP — alkaline phosphatase; GGT — gamma-glutamyl transferase; ALT — alanine aminotransferase; GLDH — glutamate dehydrogenase; CK — creatine kinase; PT — prothrombin time; aPTT — activated partial thromboplastin time.

Approximately 3 h later the dog’s mentation began to severely deteriorate. The dog developed tachypnea (52 breaths/min) and her heart rate had increased to 240 beats/min. Body temperature was 38.2°C and blood glucose concentration was 5.3 mmol/L. Thoracic focused assessment with sonography was negative for both pleural and pericardial effusion. Inspection of the continuous ECG revealed ventricular tachycardia and 2 doses of propranolol (Propranolol Hydrochloride Inj. 1 mg/mL; Sandoz Canada, Boucherville, Quebec), 0.02 mg/kg BW, IV were administered 5 min apart. A normal sinus rhythm was obtained, although the dog remained tachycardic (223 beats/min). At this time, the dog began to cough and increased bronchovesicular sounds were detected on thoracic auscultation. Pulse oximetry revealed a peripheral capillary oxygen saturation of 85% on room air. Oxygen supplementation (100%) was initiated with bilateral nasal prongs; subsequently, peripheral capillary oxygen saturation increased to 94%.

Thoracic radiographs showed a diffuse interstitial to coalescing alveolar pattern most severe in the hilar and cranioventral regions of the lungs (Figures 2A, B). A pleural fissure line was present between the right middle and caudal lung lobes suggestive of pleural thickening, although pleural effusion could not be ruled out. Gastric dilation was consistent with functional ileus. The cardiac silhouette and mediastinum were within normal limits and there was no evidence of lymphadenopathy. Pulmonary hemorrhage was identified as the top differential due to the patient’s history and concurrent coagulopathy. Other differential diagnoses for acute respiratory distress included cardiogenic pulmonary edema, pulmonary thromboembolism, anaphylaxis, severe infection, septicemia, and neurogenic disease.

Figure 2.

Radiographs of a dog hospitalized for xylitol toxicity that had exhibited pulsus alternans approximately 4 hours prior to acquisition. A — Left lateral thoracic radiograph. B — Ventrodorsal thoracic radiograph. Final interpretation consistent with a diffuse interstitial to coalescing alveolar pattern, most severe in the hilar and cranioventral regions of the lungs. In the periphery of the pulmonary parenchyma there is soft tissue infiltration surrounding the pulmonary vasculature. A pleural fissure line is seen between the right middle and caudal lung lobes. The cardiac silhouette and mediastinum are within normal limits. There is no evidence of lymphadenopathy.

The dog was blood-typed as DEA 1.1 positive and a fresh frozen plasma transfusion was prepared. Prior to starting the transfusion, the dog went into cardiac arrest and CPR was initiated. The dog received 1 dose of epinephrine (Epiclor; Rafter 8 Products, Calgary, Alberta), 0.01 mg/kg BW, IV and 1 dose of atropine sulphate (Atro-SA 0.5 mg/mL, Rafter 8 Products), 0.04 mg/kg BW, IV. Cardiopulmonary resuscitation was discontinued after 5 min as per the owners’ request. The owners did not consent to a post-mortem evaluation.

Discussion

To the authors’ knowledge, this is the first report of PA in a critically ill dog prior to clinical deterioration and death. Furthermore, this also serves as the first documentation of pulsus alternans with an HDO device. Defined as the alternation of high and low pulse amplitude in the face of a regular cardiac rhythm (1), PA can be detected using various techniques. While it has been noted that alternating pulses in human medicine may be undetectable by palpation in extreme cases, peripheral arterial palpation generally permits recognition of PA when the difference between strong and weak systolic beats is ≥ 20 mmHg (13). The threshold for detecting PA in small animal patients is unknown. Direct arterial monitoring (9,11) and echocardiography (10) facilitate reliable detection of PA in dogs; however, their clinical applications are limited by facilities and expertise. Plethysmographic waveforms of a pulse oximeter can serve as an alternative diagnostic tool for the detection of PA (4), although poor quality tracings are a common limitation in practice. As evidenced by this report, HDO monitoring allows rapid and easy visual examination of the pulse wave and serves as a reasonable non-invasive technique for clinical application.

In human medicine, myocardial injury has been identified as a complication of critically ill patients that is often unrecognized (14). More recently, this phenomenon has been highlighted in veterinary medicine with regard to critically ill dogs (15). Acute ventricular dysfunction and myocardial depression have been recognized as consequences of severe systemic inflammation, independent of primary cardiac disease (14,15). This relationship is thought to be mediated through the generation of various cytokines serving as negative inotropes [e.g., tumor necrosis factor (TNF), interleukin-1 (IL-1)] (16,17). Central to the development and progression of the systemic inflammatory response syndrome (SIRS), these cytokines are thought to depress cardiac contractility through the activation of inducible nitric oxide synthase (15–18). Furthermore, cardiac function may be further impaired by local thrombosis, myocardial hypoxia, and the generation of oxygen free-radicals in states of systemic inflammation (15,16).

It has been suggested that pulsus alternans may serve as one of the first recognizable clinical indicators of underlying myocardial dysfunction in the critical care setting (4,12). While the mechanism of the regular alternation between strong and weak pulses remains a focus of pathophysiologic investigations, dysfunctional calcium cycling of the sarcoplasmic reticulum in addition to hemodynamic alterations have been identified as central contributors to the development of PA (3,10,12,19–21).

Recognition of cardiac dysfunction in critically ill patients may have the potential to improve patient management in the ICU and reduce morbidity and mortality (14,15). The human literature has recommended that the identification of PA in a critical patient warrants further investigation and intensive monitoring (4,12). As cardiac arrest was the final outcome in this critically ill dog, this report suggests that identification of PA may help to alert the clinician to clinical deterioration. Future studies are warranted to investigate the incidence and potential clinical significance of PA in veterinary medicine.

Although a direct cardiotoxic effect of xylitol cannot be ruled out, the authors speculate that this case of PA occurred following severe hepatic necrosis, the development of SIRS, and subsequent myocardial depression and ventricular dysfunction. However, it is important to note that while PA is generally indicative of myocardial dysfunction, it has less frequently been recognized in association with severe tachycardia and normal myocardial function (7). It is suspected that sinus tachycardia at the time of PA identification in this dog was only a minor contributor to the development of PA in the face of severe systemic inflammation. Although primary cardiac disease cannot be completely ruled out in this dog, it was not suspected based on the dog’s history and thoracic radiographs.

As a result of ongoing tissue injury, severe hepatic necrosis has the potential to result in a dysregulated inflammatory response, macrophage activation, cytokine production, and the subsequent development of systemic inflammatory response system (SIRS) (16). Defined as the clinical manifestation of the systemic response to injury or microbial invasion, the classic criteria for SIRS in dogs consist of a heart rate > 120 beats/min, respiratory rate > 40 breaths/min or PaCO₂ < 30 mmHg, body temperature < 38°C or > 40°C, and > 18.0 or < 5.0 WBC × 10⁹ cells/L (16). Although the dog in this report did not initially present with these abnormalities, she went on to develop transient hypothermia, tachypnea, and tachycardia on day 2 of hospitalization. The authors suspect that ongoing hepatic injury evidenced by elevations in liver enzymes and a progressive coagulopathy contributed to a disproportionate response between proinflammatory and anti-inflammatory mediators and the subsequent development of SIRS (16). The presence of thrombocytopenia in this case provides further support for the presence of disseminated intravascular coagulation (DIC), which is often triggered by disease states with severe systemic inflammation.

Canine xylitol toxicosis generally carries a good to excellent prognosis with effective management and only mild liver enzyme elevations (22,23). While hypoglycemia appears to be the more common syndrome associated with toxicity (23), xylitol hepatopathy has also been reported without hypoglycemia (22). Furthermore, it has been suggested that liver involvement may be idiosyncratic in nature, as the ingestion of high doses of xylitol does not always result in hepatic failure (23–25). The development of hepatic necrosis not only has the potential to result in a coagulopathy from the decreased production of clotting factors, but may also contribute to the development of DIC and SIRS (16,24,25). Consequently, evidence of liver failure secondary to xylitol toxicity has been associated with a more guarded to poor prognosis in dogs (25).

This case report is not without limitations. Primarily, no echocardiogram was performed to evaluate cardiac function and to exclude primary cardiac disease. Consequently, there is no confirmation that systolic dysfunction was present in this dog. In addition, alternative ways to document PA such as direct arterial blood pressure and pulse oximetry were not used to confirm the HDO results. While the authors suspect that the PA reported in this case was secondary to cardiac dysfunction, it is also important to recognize that PA has also been reported in cases of acute ischemia and pulmonary embolism despite normal systolic function (26,27). As the dog in this report had data supporting DIC, dyspnea secondary to pulmonary thromboembolism cannot be ruled out. Finally, it could be argued that the dog may have suffered negative consequences from the use of propranolol for the treatment of ventricular tachycardia in the face of suspected cardiac dysfunction. However, many experimental studies have failed to demonstrate deleterious effects following the administration of propranolol in dogs with severe cardiac changes (28–31).

This is the first report of a critically ill dog developing PA prior to clinical deterioration and cardiac arrest following xylitol ingestion. Increased recognition of PA and its potential implications in veterinary medicine may contribute to the identification of cardiovascular complications associated with systemic disease. Furthermore, identification of this clinical finding broadens the spectrum of information provided by non-invasive monitoring techniques in the ICU. CVJ