Hyponatremia in the perioperative period: When and how to correct

Watch a video presentation of this article

Watch the interview with the author

Abbreviations

ADH

antidiuretic hormone

CNS

central nervous system

HE

hepatic encephalopathy

IV

intravenous

MELD

Model for End‐Stage Liver Disease

MELDNa

MELD sodium

MRI

magnetic resonance imaging

ODS

osmotic demyelination syndrome

Hyponatremia in cirrhosis is a common and challenging issue. Approximately 50% of patients have serum sodium concentrations less than 135 mEq/L, whereas only 1% have serum sodium values less than 120 mEq/L.1 The accepted definition of hyponatremia in cirrhosis is a serum sodium value less than 130 mEq/L with a prevalence rate of 21.6%.1, 2 Data on the prevalence of hyponatremia that is commonly cited is now more than 10 years old, and with increasing Model for End‐Stage Liver Disease (MELD) scores at the time of transplant and the introduction of the MELD sodium (MELDNa) score, the present‐day prevalence of hyponatremia may in fact be significantly higher.

A comprehensive discussion of the pathophysiology of hyponatremia in cirrhosis is beyond the scope of this concise review. In brief, systemic vasodilation occurs leading to decreased systemic blood pressure and renal perfusion. Carotid artery and renal baroreceptors, through neurohormonal‐mediated mechanisms, activate vasoconstrictor pathways including renin‐angiotensin system, sympathetic nervous system, and antidiuretic hormone (ADH). This results in renal sodium and water retention leading to ascites and hyponatremia, respectively.3

The impact of hyponatremia on patient outcomes is significant. Hyponatremia is associated with bacterial infections, ascites, spontaneous bacterial peritonitis, renal failure, hepatic encephalopathy, and reduced quality of life.4 Furthermore, decreasing values of serum sodium between 140 and 125 mEq/L are associated with increased risk for death after adjustment for the MELD score.5, 6 Therefore, in January 2016, the MELDNa score was implemented in the United States as the national organ allocation model.

Because of this newly implemented MELDNa model, the approach to management in the wait‐listed cirrhotic patient with hyponatremia is evolving. Because of the fact that patients with lower sodium levels derive higher MELDNa scores, in turn increasing their priority for LT, the impetus for correction of hyponatremia is likely less than in prior eras, although this is conjecture. Most hepatologists will initiate a management strategy for hyponatremia when the serum sodium concentration reaches 120 mEq/L or less, and it is recommended that any patient, regardless of serum sodium level, with symptomatic hyponatremia, undergo treatment. Patients with serum sodium concentrations greater than 125 mEq/L are usually asymptomatic; however, at lower serum sodium concentration, symptoms may develop including headache, disorientation, nausea, vomiting, and lethargy. These symptoms are relatively nonspecific, adding to the management challenges. The majority of patients experience development of hyponatremia over a period of weeks, allowing time for cerebral adaptation through efflux of intracellular osmoles. In the event of rapid and severe hyponatremia, significant manifestations can occur including coma, respiratory arrest, seizures, and brain herniation.

The critical issue in the correction of hyponatremia is the risk for osmotic demyelination syndrome (ODS) when hyponatremia is corrected too quickly. The classic presentation of this syndrome is biphasic in nature with hepatic encephalopathy (HE) or seizures during hyponatremic phase and then clinical improvement as the sodium level is restored, followed by development of dysarthria, dysphagia, oculomotor dysfunction, and/or quadriparesis over a period of days.7 The incidence rate of ODS is only 0.5% to 0.88%, but the outcomes are devastating, with a rate of death or disability of 77% or more.7, 8 Clinical suspicion as well as brain magnetic resonance imaging (MRI) are used to diagnose this condition. The MRI findings may lag several weeks behind the clinical presentation, and thus does not exclude ODS if performed early in the course. A classic MRI appearance of ODS would be a well‐defined low T₁ intensity lesion in the pons (or extrapontine central nervous system [CNS]) and high T₂ intensity trident‐shaped pontine lesion.9

The initial management of hyponatremia includes evaluation of the volume status. Those with hypovolemic hyponatremia require careful volume resuscitation and suspension of diuretics. Reasons for volume loss must be addressed, such as excessive diarrhea from lactulose, poor oral intake, or excessive diuresis. It is important to correct hypokalemia in all patients because this can result in improvement in serum sodium levels.

The majority of patients will have hypervolumic hyponatremia.10 The initial management approach is also withdrawal of diuretics followed by a free‐water restriction.3, 11 There is debate about the appropriate degree of free‐water restriction and whether this includes all oral intake and intravenous (IV) sources or only free water. The free‐water restriction can begin at 1.5 L/day in patients with less severe hyponatremia. In those with more significant hyponatremia, a 1 L/day free‐water restriction is required. The impact of free‐water restriction is usually seen within the first 2 to 3 days, and the free water restriction can be adjusted appropriately based on the effect seen. Free‐water restrictions of 500 mL/day or less are not well tolerated and generally avoided. In patients who do not achieve the desired correction of hyponatremia with diuretic suspension and free‐water restriction, IV 25% albumin can be used despite limited data. Twenty‐five percent albumin in small studies has led to improvement in serum sodium in hypervolemic patients.12, 13 However, this is not a durable solution for chronic hyponatremia.

The vaptans are ADH receptor antagonists that inhibit the principal cells of the collecting ducts from reabsorbing water, thereby resulting in aquaresis.14 Tolvaptan (oral) and conivaptan (IV) are the only vaptans that are approved in the United States. Conivaptan is a V2 and V1a receptor antagonist.15 Because of the IV formulation that complicates ease of use, potential for decreased blood pressure, variceal hemorrhage, and potential for renal compromise, this drug is rarely used. Tolvaptan was shown to be effective in the SALT‐1 and SALT‐2, which included 63 patients with cirrhosis, Child‐Pugh score less than 10, and serum sodium concentration less than 120mEq/L.16 The day 4 and day 30 serum sodium concentrations were higher (statistically significant) in cirrhotic patients receiving tolvaptan. However, based on a polycystic kidney disease study (TEMPO3:4), a higher frequency of liver enzyme elevations was noted in the tolvaptan group (≈5%) compared with placebo (≈1%).17 Therefore, a US Food and Drug Administration warning label was placed on this drug for use in cirrhotics. A long‐term follow‐up study (mean 1.9 years) for patients, including in the SALT‐1 and ‐2 trials, did not demonstrate an excess frequency of elevated liver tests.18 It is important to note that the dose of tolvaptan was as high as 90 mg in the morning and 30 mg in the evening in the TEMPO 3:4 study compared with a total daily dosage of 60 mg in the SALT‐1 and ‐2 trials. Nonetheless, the role of tolvaptan and management of hyponatremia in cirrhotic patients is still not clear, and at this point it is not used frequently. Demeclocycline, another ADH antagonist, should not be used because of the risk for kidney dysfunction.11

In general, serum sodium should not be corrected by more than 8 mEq/L in a 24‐hour period. Overcorrection of serum sodium appears to be more important than the initial sodium value. In a systematic review (n = 59), the majority of LT recipients experiencing ODS had sodium values between 121 and 135 mEq/L (63%), whereas 22% of those with serum sodium concentration greater than 135 mEq/L and 4% with concentration less than 120 mEq/L experienced ODS.7 However, caution should be exercised in interpreting these results because many LT programs will not perform LT in those with serum sodium concentration less than 120 mEq/L. The important point derived from this study is that ODS can occur in cirrhotic patients despite normal serum sodiums if overcorrection occurs. Although limited data are available, LT is generally not advisable in patients with serum sodium concentration less than 120 mEq/L, especially in those with significant malnutrition or a history of alcoholism.

Hypertonic saline is rarely used because of the concern for increasing the serum sodium level too quickly and precipitating osmotic demyelination. We reserve the use of hypertonic saline to those patients who have a serum sodium concentration less than 120 mEq/L and severe symptoms (such as seizures) attributable to abrupt hyponatremia. No randomized controlled trials have investigated the use of hypertonic saline in cirrhotic patients. During transplantation, the administration of fluids as well as a correction of hepatic function may lead to more abrupt changes in serum sodium, which lead to a greater risk for osmotic demyelinization; therefore, hypertonic saline should not be used in this setting.

Operative Management of Hyponatremia

During transplantation the serum sodium concentration typically increases, with greater changes in the more hyponatremic patients.19, 20 Larger intraoperative increases have been associated with longer posttransplant intubation and ICU stays; the relationship to other posttransplant outcomes is discussed later.8, 19 Larger increases are associated with higher volumes of intraoperative fluid administration, higher transfusion of blood and blood products requirements, and the use of sodium bicarbonate.19, 20 Efforts to ameliorate the increase in sodium intraoperatively include limiting the volume of isotonic fluids given and restricting transfusion; however, this is not always possible. The use of tromethamine instead of sodium bicarbonate for management of acidosis may be appropriate. Continuous renal replacement therapy has been used intraoperatively to limit Na increase. When used, the replacement fluid must be adjusted appropriately; otherwise, it can contribute to the sodium elevation.19, 21

The effect of hyponatremia on posttransplant outcomes has been controversial. Most studies demonstrate increased morbidity such as renal failure, septic complications, and neurological complications, and several studies indicate there is worse posttransplant survival.22, 23, 24 However, in the largest study on this topic, we demonstrated no association between the hyponatremia and survival relative to patients with normonatremia.4 In another study by Hackworth et al.25 no difference was noted in 6‐month posttransplant survival between patients with pretransplant hyponatremia versus resolved pretransplant hyponatremia versus those who had never had hyponatremia. A small perspective study from Region 11 in the United States involving 62 patients demonstrated no difference in 6‐month‐survival compared with normonatremic counterparts.26 In contrast, hypernatremia was associated with diminished posttransplant survival compared with patients with hyponatremia and normonatremia, suggesting that the correction of hypernatremia before transplant is of significant importance.4

In summary, hyponatremia remains a common and challenging clinical problem. Because of the implementation of the MELDNa system, the impetus to treat hyponatremia is likely evolving, but current criteria for treatment include sodium less than 120 mEq/L or symptoms of hyponatremia. For hypervolemic hyponatremia, free‐water restriction, cessation of diuretics, and correction of hypokalemia remain the mainstay of management. There are limited data to support 25% albumin infusions, and the effect is short‐lived. Unfortunately, there are safety concerns with the use of vaptans. However, in patients with chronic and severe hyponatremia who will not receive LT imminently, low‐dose tolvaptan can be considered. The role of tolvaptan is not completely clear at this time, and the risk of liver test elevations merits caution and careful laboratory monitoring in the use of this medication. Lastly, hypertonic saline can be considered in patients with serum sodium concentration less than 120 mEq with severe symptoms (i.e., seizure), but outside of severe symptoms, generally it is not required immediately before or during liver transplantation because of the natural rise in sodium during the transplant. Great attention needs to be paid to limiting the increase in serum sodium with these strategies to no more than 8 mEq/L in a 24‐hour period, because osmotic demyelination can result when this threshold is exceeded, which is often a devastating outcome.