Hyperosmotic solutions such as 50% dextrose, mannitol, and hypertonic saline may be administered as a continuous infusion or as a small-volume bolus. Significant hemodynamic changes are often not observed with the administration of such boluses. Animal studies have reported acute decreases in systemic arterial blood pressure1–8 and bradycardia2,4–6 with the administration of hyperosmotic solutions. Furthermore, transient hypotension has been reported in humans after the administration of 25% mannitol.8 The mechanism of hyperosmolar-induced hypotension is believed to involve a neural reflex with cardiopulmonary afferents and vasomotor efferents.4 An animal study4 has demonstrated that, for a given solution, the rate of infusion and the degree of osmolarity factor into not only the presence of the hypotensive response, but also its magnitude, if present. This is most likely because both the rate of infusion and the degree of osmolarity influence the osmotic load at the proposed osmoreceptor at a given point in time. To our knowledge, this is the first case detailing acute hypotensive responses to 50% dextrose in a human subject. The patient’s power of attorney reviewed the case report and gave written consent for publication as a result of the patient’s inability to provide consent.
A 60-year-old woman who had undergone an aortic valve replacement and repair of an ascending aortic aneurysm in 2009 presented with chest pressure. A computed tomography scan revealed a mediastinal hematoma resulting from graft anastomosis dehiscence (Fig. 1). The etiology of the dehiscence was endocarditis and subsequent graft infection. She had her aortic valve, root, and arch replaced. She remained on a ventilator secondary to complications of an embolic stroke and required an epinephrine infusion at 2 μg/min to maintain adequate cardiac output and arterial blood pressure. She was instrumented with a left-sided internal jugular triple-lumen catheter and a right-sided pulmonary artery catheter. The location of the tip of the internal jugular catheter and of the pulmonary artery catheter was confirmed in the distal superior vena cava and in the main pulmonary artery, respectively. Her arterial blood pressure was measured via a left-sided radial arterial line.
A few days postoperatively, the nurse observed that administration of 50% dextrose for hypoglycemia, as governed by hospital protocol, was followed by an acute decrease in systemic arterial blood pressure on 2 occasions. Per nurse report, she administered 50 mL of 50% dextrose over approximately 30 to 60 seconds through the internal jugular catheter for a serum glucose of 59 mg/dL. This was followed by a decrease in systemic arterial blood pressure from 108/46 to 52/18 mm Hg (Table 1). A few hours later, 40 mL of 50% dextrose was administered through the same catheter for hypoglycemia (serum glucose 64 mg/dL). The nurse reported that the patient’s systemic arterial blood pressure decreased acutely from 106/48 to 58/24 mm Hg (Table 1).
Once again, the patient experienced hypoglycemia (serum glucose 52 mg/dL). We realized that we needed to devise a safe way to administer dextrose to treat recurrent hypoglycemia. Speculating that the previous episodes of hypotension occurred secondary to a disruption in the epinephrine infusion simultaneous to the dextrose injection, we moved the epinephrine infusion from the internal jugular catheter to the pulmonary artery catheter and attempted to administer 50 mL of 50% dextrose through the internal jugular catheter. After her systemic arterial blood pressure decreased from 114/56 to 78/40 mm Hg with administration of 20 mL of 50% dextrose, administration of the remaining 30 mL was aborted.
Hours later, the patient’s serum glucose was 60 mg/dL. After 3 observations of decreases in systemic arterial pressure with 50% dextrose administration through the internal jugular triple-lumen catheter and hypothesizing that administration through this catheter somehow contributed to the hemodynamic alteration, we attempted to administer 50 mL of 50% dextrose through the pulmonary artery catheter. After injection of 20 mL, her systemic arterial blood pressure decreased from 110/56 to 74/38 mm Hg. Administration of the remaining 30 mL was aborted. Administration of 50 mL normal saline had no effect on her arterial blood pressure.
Transthoracic echocardiogram conducted during the injection of dextrose through the internal jugular catheter demonstrated grossly unchanged left ventricular function during and after the decreases in systemic arterial blood pressure. Performing the Valsalva maneuver showed the lack of a functioning patent foramen ovale. The filling pressures in both ventricles appeared unchanged. There were no noticeable changes in inferior vena cava collapsibility. A chest radiograph demonstrated unchanged and proper placement of both catheters (Fig. 2). Ultrasound examination did not show evidence of hematoma or injury at the catheter sites.
It has been reported that intravenous (IV) infusions of glucose, mannitol, and hypertonic saline may induce transient, although substantial, decreases in systemic arterial blood pressure and heart rate in animals.1–8 The incidence of clinically significant hypotension and bradycardia in human subjects receiving boluses of hyperosmotic solutions is unclear. To our knowledge, this is the first case detailing acute hypotensive responses to 50% dextrose in a human subject. In our case, we had the benefit of real-time arterial blood pressure, electrocardiogram, and pulmonary artery pressure monitoring during and after the administration of 50% dextrose. Because of the lack of literature citing hyperosmolar-induced hypotension to 50% dextrose in human subjects, one should ask whether the phenomenon occurs commonly but goes unnoticed as a result of the lack of adequate monitoring and/or whether it may be more likely to occur in patients with significant cardiovascular compromise, like in this patient.
It is important to note that in animal studies that document hemodynamic alterations with the rapid administration of hyperosmotic solutions, simultaneous hypotensive and bradycardic responses are cited.2,4–6 The patient described here maintained a stable heart rate (Table 1). The most likely explanation for the patient’s unchanged heart rate is that she was receiving an epinephrine infusion, potentially masking bradycardia. Because hypotension occurred without bradycardia, it may be concluded that bradycardia is more likely in response to hypotension rather than vice versa. For instance, systemic hypotension may lead to decreased coronary perfusion, resulting in bradycardia.
Because hyperosmolar-induced hypotension has been described with various hyperosmotic solutions,1–8 it can be concluded that the response depends on osmotic load rather than on chemical composition of the solution. The mechanism of hyperosmolar-induced hypotension remains unclear. In our case, examination of the catheter sites excludes the potential for mass effect and baroreceptor activation as the impetus for the acute decrease in systemic arterial blood pressure. Cardiac dysfunction is an unlikely mechanism for the hypotensive response because left ventricular function was grossly unchanged on transthoracic echocardiogram during and after the decreases in systemic arterial blood pressure. Furthermore, increases in cardiac output, coronary perfusion, and ventricular contractility occurred simultaneously to the decrease in systemic arterial blood pressure with the rapid administration of hypertonic saline in dogs.3
The most plausible explanation for the decrease in arterial blood pressure proposed in animal studies is systemic vasodilation.2,4 Raizner et al.2 report that infusion of 50% glucose over 30 seconds into the superior vena cava of dogs resulted in a mean decrease in systemic arterial pressure of 34.3% at 40 seconds with a return to baseline within 2 minutes. Infusion at half the rate resulted in a statistically significant decrease in brachial artery plasma osmolality; an attenuation of vasodilation in muscle, kidney, and fat vascular beds; and an attenuation of the decrease in systemic arterial blood pressure.
An animal study by Zhang et al.4 reports vasodilation via a neural reflex with cardiopulmonary afferents and vasomotor efferents as the mechanism for the decrease in systemic arterial blood pressure and bradycardia after administration of hypertonic saline. A threshold concentration of 1.5% to 2.0% sodium chloride was required to elicit hypotensive and bradycardic responses, and the magnitude of the hypotensive and bradycardic responses increased in a concentration-dependent manner. IV hypertonic saline evoked the hypotensive response with a larger magnitude and with a shorter latency than if injected directly into the aorta. The hypotensive response was not affected by bilateral cervical vagotomy. A plausible explanation for the larger magnitude and shorter latency of the hypotensive response to IV hypertonic saline and for the lack of effect of vagotomy is that there is an osmoreceptor associated with cardiopulmonary afferents between the vena cava and the left ventricle, with this pathway independent of vagus nerve.4,9 Such a peripheral osmoreceptor afferent pathway not via the vagus nerve but through the spinal cord has been described in animals in an unrelated study.10
In conclusion, rapid infusions of even small volumes of hyperosmotic solutions may result in a dramatic decrease in systemic arterial blood pressure (with or without bradycardia). Although the decrease in systemic arterial blood pressure (with or without bradycardia) is transient, it may be dangerous if uninterrupted blood flow is critical such as in patients with an acute ischemic stroke for which strict blood pressure parameters are warranted or with underlying cardiac dysfunction who are incapable of compensating from reduced coronary perfusion.
A few factors should be considered in preventing the hypotension (and possible bradycardia) associated with the bolus of 50% dextrose and other hyperosmotic solutions. Three factors, namely the degree of osmolarity of the solution, the rate of infusion, and the location of infusion, impact the osmotic load at the proposed osmoreceptor involved in hyperosmolar-induced hypotension. First and foremost, the osmolarity of the solution must be considered. All else equal, a solution of greater osmolarity will result in a greater osmotic load at the proposed osmoreceptor at a given point in time. Second, the rate of infusion must be considered because solutions infused slowly result in less of an osmotic load at the osmoreceptor at a given point in time. Perhaps, patients may benefit from the establishment of a maximal rate of infusion of boluses of 50% dextrose and other hyperosmotic solutions. Third, the site of infusion must be considered. Infusions through catheters terminating in or near the vena cava or heart, the speculated location of the osmoreceptor, will allow for no to minimal admixture and therefore no to minimal hemodilution of the osmotic load, whereas infusions through catheters of peripheral IV lines further from the central circulation will allow for admixture and therefore hemodilution of the osmotic load. Finally, patients with significant cardiovascular disease may have an exaggerated response to hyperosmolar-induced hypotension because they may be unable to compensate from the acute and profound systemic vasodilation. Special care should be taken when administering boluses of hyperosmotic solutions to patients who are in a state of cardiovascular compromise.
1. Muirhead EE, Lackey RW, Bunde CA, Hill JM. Transient hypotension following rapid intravenous injections of hypertonic solutions. Am J Physiol 1947;151:51624.
2. Raizner AE, Costin JC, Croke RP, Bishop JB, Inglesby TV, Skinner NS Jr. Reflex, systemic, and local hemodynamic alterations with experimental hyperosmolality. Am J Physiol 1973;224:132733.
3. Kien ND, Kramer GC, White DA. Acute hypotension caused by rapid hypertonic saline infusion in anesthetized dogs. Anesth Analg 1991;73:597602.
4. Zhang D, Sato T, Gong D, Fu L, Dai S, Xu H, Wu Q, Wang D, Peng Y, Sun Y. Neural reflex hypotension induced by very small dose of hypertonic NaCl solution in rats. Chin J Physiol 2009;52:815.
5. Pisarri TE, Jonzon A, Coleridge HM, Coleridge JC. Intravenous injection of hypertonic NaCl solution stimulates pulmonary C-fibers in dogs. Am J Physiol 1991;260:H152230.
6. Pinsky MR, Smith PL, Bleecker ER, Bromberger-Barnea B. Effects of antihistamines and indomethacin on hyperosmolar-induced vasodilation. Am J Physiol 1982;242:H4505.
7. Stiff JL, Munch DF, Bromberger-Barnea B. Hypotension and respiratory distress caused by rapid infusion of mannitol or hypertonic saline. Anesth Analg 1979;58:428.
8. Coté CJ, Greenhow DE, Marshall BE. The hypotensive response to rapid intravenous administration of hypertonic solutions in man and in the rabbit. Anesthesiology 1979;50:305.
9. Wang WZ, Gao L, Pan YX, Zucker IH, Wang W. Differential effects of cardiac sympathetic afferent stimulation on neurons in the nucleus tractus solitarius. Neurosci Lett 2006;409:14650.
10. Vallet PG, Baertschi AJ. Spinal afferents for peripheral osmoreceptors in the rat. Brain Res 1982;239:2714.