Brain damage and postoperative hyponatremia

Over the past decade, a number of reports have described the phenomenon of postoperative hyponatremia resulting in death or permanent brain damage in generally healthy individuals who have recently had elective surgery. ^[1-9] Although most reported patients who have died or suffered permanent brain damage as a complication of postoperative hyponatremia have been women, only recently have statistical data demonstrated the propensity of menstruant women to suffer this disorder. ^[10] However, epidemiologic data are difficult to acquire for a sporadic condition that affects hospitalized patients, and as a result, the frequency of permanent brain damage from hyponatremia following elective surgery is not generally known.

Postoperative hyponatremia is a common clinical problem in the United States and western Europe, with an occurrence of 1% to 5%. ^[10-12] Death from postoperative hyponatremia was initially reported in the mid-1930s, ^[13] and there have been multiple subsequent reports of such patients dying or suffering permanent brain damage. ^[14-17] There has been an increasing awareness that postoperative hyponatremia can lead to death or permanent brain damage following elective surgery in generally healthy adults. *RF 1-4,6-10,18 19* However, hyponatremia may be unsuspected as the cause of death or brain damage. Patients typically have a prodrome of headache, nausea, emesis, and weakness, and because such symptoms are somewhat nonspecific, they are often thought to be associated with the postoperative state. Respiratory insufficiency, manifested by either respiratory arrest or pulmonary edema, is often the initial manifestation of hyponatremic encephalopathy. ^[20] Respiratory insufficiency is often associated with grand mal seizures or sudden obtundation. ^[7]

Role of intravenous fluids in the pathogenesis of postoperative hyponatremia.

As early as 1953, water retention out of proportion to sodium retention was reported in surgical patients, accompanied by decreases in urine output and plasma sodium and an increase in urine salt excretion. ^[15] In postsurgical patients, arginine vasopressin (AVP, or antidiuretic hormone) levels are universally increased when compared with preoperative values. ^[5,11,21] Despite abnormalities in AVP metabolism or renal function, hyponatremia will not develop unless excess free water is administered. Only a small minority of surgical patients, approximately 1%, develop hyponatremia, and symptomatic hyponatremia occurs in approximately 20% of these patients. ^[10,22] It is the failure to recognize the compromised ability of the patient to maintain water balance that most often leads to hyponatremia in the postoperative patient. Although at least 15 liters of fluid can usually be excreted daily in a normal individual, ^[23] studies from our laboratory have demonstrated that the intravenous administration of as little as 3 to 4 liters of hypotonic fluid over 2 days can result in fatal hyponatremic encephalopathy in women postoperatively. ^[7,10,19] Thus, the most important step to take in the prevention of postoperative hyponatremia is a careful consideration of the choice of intravenous fluids administered. Hypotonic solutions (280 mM glucose, 77 mM NaCl, 280 mM glucose/38 mM NaCl) are seldom, if ever, appropriate in patients postoperatively. Unless there is a definite reason to suspect the existence of a deficit of free water, isotonic sodium chloride (154 mM NaCl) is virtually always preferable.

Prevalence of postoperative hyponatremia and brain damage in women.

Although postoperative hyponatremic encephalopathy can occur in anyone, the occurrence of brain damage or death demonstrates a clear predisposition for women. ^[10] A review of the literature since about 1935 reveals multiple cases of symptomatic postoperative hyponatremia often accompanied by coma and seizures. ^[22] Recent studies demonstrate that the age and gender of the patient are major determinants of brain damage due to hyponatremia. ^[10] In general, most adult patients with symptomatic hyponatremia (serum sodium level of less than equals 128 mmol/L) who did not suffer permanent neurologic damage were older women or men of any age, whereas those who died or developed permanent brain injury were younger women. ^[10] Many recent textbooks and journal articles state that brain damage due to hyponatremia is uncommon unless the serum sodium is below 120 mmol/L. ^[24-27] Recent information documents that in postoperative women, hyponatremic brain damage can occur with a serum sodium level of 128 mmol/L or less. ^[10,19,20] Is the increased susceptibility of women to brain damage from postoperative hyponatremia a new finding? Figure 1 shows a literature review of 159 cases of postoperative hyponatremia from 1935 to 1990 in which enough information was available to determine the gender distribution of patients suffering morbidity. The gender distribution of patients with postoperative hyponatremia did not differ from the period 1935-90 to now. However, morbidity and mortality occur almost exclusively in women, both currently and in 1935-90 (see Figure 1). Thus, postoperative hyponatremia occurs equally in both genders, but morbidity occurs primarily in women (see Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. The morbidity, mortality, and gender distribution in 159 patients with postoperative hyponatremia reported in the literature between 1935 and 1990 is compared to recent data (1992) on 739 patients with postoperative hyponatremia. ^[10] The upper panel shows the gender distribution (men vs. women). For both periods, there was no significant difference in the percentage of patients with postoperative hyponatremia (p more than 0.1). The lower panel shows the mortality (men vs. women); for both periods, the mortality in women is significantly higher than that in men (p less than 0.001). (Data extracted from the following references: 7,13-15,17,19,21,32,58-71).

Epidemiology and incidence of hyponatremic brain damage.

How frequent is the syndrome of permanent brain damage as a consequence of postoperative hyponatremia? Until recently, the frequency has been unclear because in many patients suffering morbidity from postoperative hyponatremia, the cause of the brain damage was attributed to other causes, ranging from postpartum vasomotor collapse, air embolism, herpes encephalitis, and arteriovenous malformation to narcotic overdose. *RF 1,3,7,9,16 19* In the United States, the projected mortality and morbidity rate (death and permanent brain damage) among postoperative patients with hyponatremia is over 10,000 cases per year. ^[10] This Figure isderived from 25 million surgeries in the United States per year ^[28,29] with an incidence of hyponatremia of 1%, or 250,000 cases per year. ^[10] In a recent study of patients with postoperative hyponatremia, the frequency of hyponatremic encephalopathy was 8% (66/740) ^[10] or a projected 20,000 cases per year. Among patients with hyponatremic encephalopathy, the morbidity rate was 52% (34/66), or a projected mortality and morbidity rate (death and permanent brain damage) of over 10,000 cases per year in the United States Figure 2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. The mortality associated with postoperative hyponatremia in the United States. Among 25 million inpatient surgical procedures per year in the United States (including cesarean section but excluding uncomplicated inhospital childbirth), the incidence of hyponatremia is about 1%. Among patients with postoperative hyponatremia, approximately 8% develop hyponatremic encephalopathy, approximately 52%, or about 10,000, of whom either do not survive or suffer permanent brain damage. (Data from Ayus JC, Wheeler JM, Arieff AI. Postoperative hyponatremic encephalopathy in menstruant women. Ann Intern Med 1992;117:891-897.)

A recent publication from the Mayo Clinic stated that in their institution, from 1976 to 1992, there were no cases of postoperative hyponatremia and death in women. ^[30] Cardiorespiratory arrest was searched for by a retrospective computer analysis of 290,815 surgical cases in women. Among these, 1,498 female patients had a cardiorespiratory arrest, a frequency of 0.5%, whereas the frequency of hyponatremia was below 0.005%. Several aspects of this data raise serious concerns. The frequency of cardiorespiratory arrest was 0.5% among women undergoing surgery, or about 1 death per 200 operations, an extraordinarily high postoperative death rate. By comparison, at the University of California at San Francisco, a recent study of postoperative complications among 474 elderly male patients in the highest cardiovascular risk category demonstrated no instances of postoperative cardiorespiratory arrest. ^[31] Although no cases of women suffering fatal postoperative hyponatremia were discovered, ^[30] two earlier studies from the Mayo Clinic described postoperative hyponatremia with neurologic symptoms and death in women. ^[14,32] It appears likely that the type of surveillance used was unable to retrieve such cases from the computer records. Furthermore, many individuals probably received sodium bicarbonate after their arrest, an almost universal therapeutic maneuver known to increase the serum sodium substantially. ^[33,34] Thus, the alleged failure of the authors to find such cases at the Mayo Clinic ^[30] cannot be interpreted as showing that postoperative hyponatremia resulting in permanent brain damage in women is a rare condition.

Cerebral effects of hyponatremia in men versus women.

Why are women so much more likely than men to suffer such complications? The brain is the major target organ of hyponatremia, and the brain's early defenses against hyponatremia-induced brain edema include an initial fall in cerebral blood flow ^[35] and loss of CSF via bulk flow. Following these initial adjustments, extrusion of sodium is the brain's most important early defense against hyponatremia, ^[36] and the most important pathway is via the Na^plus-Kplus adenosine triphosphatase (ATPase) system. ^[37,38] Estrogens inhibit Na^plus-Kplus ATPase activity in several tissues, including the brain, *RF 37 39* whereas androgens may enhance such adaptation. ^[40,41]

An hypoxic event, such as respiratory insufficiency, is a major factor militating against survival without permanent brain damage in patients with symptomatic hyponatremia. ^[7] Hypoxia impairs the ability of the brain to adapt to hyponatremia, which serves to decrease the effectiveness of the brain in the prevention of permanent damage. *RF 10,38 42*

Postoperative elevation of vasopressin in postoperative patients is a universal finding. *RF 11,21,43 44* The effects of vasopressin on the cerebral circulation are primarily mediated via V₁ vasopressin receptors. ^[45] Chronic AVP-induced hyponatremia is characterized by decreased cerebral perfusion and brain oxygen utilization, but only in female rats. ^[45] Studies of peripheral blood vessels suggest a possible difference in vascular reactivity to vasopressin between females and males, ^[46-49] and in several different encephalopathies (stroke, subarachnoid hemorrhage, cold lesion), brain edema and cerebral vasospasm are mediated by intracerebral vasopressin. ^[50-52] A recent study supports this contention. ^[53] When hyponatremia was induced in rats with 140 mM glucose and doses of vasopressin were deliberately kept too low to stimulate vasopressin V₁ receptors, there was no gender difference in mortality. Thus, in female laboratory animals, cerebral effects of the hormones estrogen and vasopressin appear to both impair adaptation of the brain to experimental hyponatremia and lead to cerebral hypoxia; both of these effects result in an increased propensity of females to develop permanent brain damage compared with males. ^[41]

Mechanisms of brain damage in hyponatremic patients.

There are at least two distinct mechanisms that can induce brain injury in hyponatremic patients. The first entity, hyponatremic encephalopathy, appears to result from a combination of brain swelling and increased intracranial pressure, leading to brain damage. Hyponatremic encephalopathy is not related to therapy for hyponatremia, since most of such brain damage occurs in untreated patients. *RF 10,19 54* The second entity is brain damage associated with therapy for hyponatremia. Over a period of 19 years in our laboratory we described 117 patients with postoperative hyponatremia who suffered death or permanent brain damage, and 4% were attributed to improper therapy for hyponatremia ^[22] Figure 3. A recent survey of 4,100 nephrologists yielded 56 hyponatremic patients who had been treated, 10 of whom had suffered permanent brain damage. ^[55] The mean absolute correction in serum sodium in these 10 patients over the initial 48 hours was 32 plus minus 7 (plus minus SD) mmol/L. These data support our prospective study of patients with symptomatic hyponatremia that identified an absolute change in serum sodium of more than 25 mmol/L over the initial 48 hours as an independent risk factor for hyponatremic brain damage. ^[56]

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Of 847 hospitalized patients with postoperative hyponatremia, from nine published series from our laboratory, 19% (158/847) developed hyponatremic encephalopathy and 117 of those patients developed permanent brain damage or died. The major risk factors associated with permanent brain damage in these 117 patients with hyponatremic encephalopathy are shown. Most patients (96%) suffered an hypoxic episode because active therapy was not initiated in a timely manner. In 4% of patients suffering permanent brain damage, improper therapy for hyponatremia was implicated. (Data extracted from references: 7,8,10,19,20,42,54,56,72.)

Avoidance of postoperative hyponatremic brain damage.

When hyponatremic encephalopathy occurs, the initial symptoms may be dramatic, including such severe manifestations as seizures and respiratory arrest, which may indicate a far advanced process. ^[7] In fact, radiologic evaluation of the brain of such patients often demonstrates evidence (by MRI or CT) of tentorial herniation. ^[54] Thus, awareness of the situations in which unrecognized symptomatic hyponatremia may be present is the first and most important step in management. Sometimes, the diagnosis of postoperative hyponatremic encephalopathy is unsuspected. ^[7] Even more important, there is a failure to recognize that a serum sodium level in the range of 120 to 128 mmol/L can lead to permanent brain damage. ^[10,19] Even when the serum sodium level is known, the neurologic manifestations of hyponatremic encephalopathy may be attributed to other neurologic disorders. In many cases, affected patients are young and otherwise healthy, and preliminary diagnoses such as subarachnoid hemorrhage, herpes encephalitis, drug reaction, and arteriovenous malformation are commonly entertained. In a study of 15 female patients with hyponatremic encephalopathy, despite the serum sodium level being known, many patients were subjected to multiple invasive procedures, including open brain biopsy for suspected herpes encephalitis. ^[7] Thus, every postoperative patient should be considered at risk for the development of hyponatremia, particularly if they are receiving hypotonic fluids. Because both improper treatment and untreated hyponatremic encephalopathy can produce brain damage, the single most important factor in the prevention of this complication is the avoidance of routine use of intravenous hypotonic fluids in postoperative patients. Commonly used hypotonic fluids include 280 mM dextrose in water (5% dextrose in water) and 77 mM NaCl (half-normal saline). Because headache, nausea, and vomiting are the most common clinical manifestations of postoperative hyponatremic encephalopathy, ^[7,10] electrolytes should be monitored in a timely fashion. If hyponatremia is diagnosed, appropriate therapy should be initiated before an hypoxic episode takes place. ^[20]

Treatment of symptomatic postoperative hyponatremia.

Currently, the decision of whether to treat a patient with postoperative hyponatremia in a passive manner (water restriction being the cornerstone of such therapy) or with an active therapeutic intervention (infusion of hypertonic NaCl) is primarily based on the presence or absence of neurologic symptomatology. The presence of symptoms referable to the CNS has been shown to be associated with brain edema by neuroradiologic criteria, as confirmed by pathologic findings at autopsy. ^{[10,19,20,42]} Thus, active therapy is necessary in such patients to prevent brain herniation and respiratory insufficiency. The distinction between ``acute'' and ``chronic'' hyponatremia is often difficult to ascertain in the clinical setting, and has not proved useful in determining the appropriate therapy. If the patient is asymptomatic, water restriction (less than 1 liter/day) is usually appropriate. Water restriction of about 800 ml/day will increase the serum sodium level by about 1 to 2 mmol/L per day. Symptomatic postoperative hyponatremia is a medical emergency, where the overall morbidity is in excess of 20%, ^[10] and active therapy is indicated. Thus, after the airway has been secured, patients with symptomatic hyponatremia should be moved to a facility where constant monitoring can be provided, such as an intensive care unit. Such patients should be treated with intravenous hypertonic sodium chloride (514 mM), using an infusion pump, with the infusion designed to raise plasma sodium at a rate of about 0.5 to 1 mmol/L per hour. The rate of correction is not an important factor in such patients. Therapy with hypertonic NaCl should be discontinued when either of two suggested end points have been accomplished: (1) the patient becomes asymptomatic, or (2) the patient's plasma sodium has increased by 25 mmol/L within the initial 48 hours of therapy. During the interval that active correction of symptomatic hyponatremia is being accomplished, monitoring of plasma electrolytes should be carried out every 2 hours, until the patient has become neurologically stable. ``Neurologically stable'' implies that the patient is communicative and able to breathe without mechanical assistance. In addition to hypertonic sodium chloride, therapy may include, when required, assisted mechanical ventilation, endotracheal intubation, or administration of a loop diuretic (furosemide). This regimen may require modification in patients with severe renal or cardiac disease. Because of the possible dangers of brain damage, the serum sodium level should never be acutely elevated to hyper- or normonatremic levels (within 48 hr). In patients who have symptomatic hyponatremia associated with inappropriately elevated plasma vasopressin levels, plasma sodium can be raised at the same rate using 514 mM NaCl, in combination with furosemide, if vascular congestion is a likely complication. ^[57] The rate of correction of hyponatremia is not an important factor in the eventual outcome, but the absolute change during the first 48 hours of therapy must be monitored closely. ^[56]

Addendum.

After this manuscript was accepted for publication, preliminary data was presented that confirmed our observations that fatal post operative hyponatremia often occurs in menstruant women following elective surgery (Steele A, Gowrishankar M, Abrahmson S, Mazer D, Halperin ML. Post-operative hyponatremia: A phenomenon of `desalination.' Presented at the 28th annual meeting, American Society of Nephrology; 1995 Nov 7; San Diego. J Am Soc Nephrol 1995;6:444). The study described five women (4 from Toronto, Canada; 1 from Rochester, NY), ranging in age from 24 to 49 years, who from 1992-1995 had elective inpatient surgery Postoperatively, all developed nausea and headache, and all suffered respiratory arrest. At the time of respiratory arrest, the serum sodium (plus minus SD) was 123 plus minus 3 mmol/L (range 118 to 127 mmol/L). All five patients died from the hyponatremia.