Hyponatremia (serum sodium concentration <135 mEq/L) is one of the most commonly encountered electrolyte abnormalities in daily practice. Its clinical manifestations may range from asymptomatic, lethargy, convulsion, and coma, depending on the severity and rapidity of the occurrence of hyponatremia. Hyponatremia itself is not a disease but a manifestation of variable kinds of diseases.1,2 Most of the underlying causes of hyponatremia are associated with low plasma osmolarity, and evaluation of the extracellular fluid status and the effective circulating volume are the first steps in the differential diagnosis.3
Hyponatremia is frequently observed in cancer patients, with an incidence ranging from 3.7% to 47%.4,5 The etiology include, but are not restricted to: (1) tumor-secreting factors such as antidiuretic hormone (ADH) and atrial natriuretic peptide (ANP), (2) tumor-related anorexia and emotional stress, and (3) treatment-induced side effects such as vomiting, diarrhea, and renal toxicity. Although the severity of hyponatremia is usually mild to moderate and its development is protracted in most cancer patients, an accurate diagnosis is important because inappropriate management may lead to further reduction of the serum sodium concentration and adverse clinical outcome.
Recently, we observed a lung cancer patient who presented to the emergency department with cardiac tamponade. He also had severe hyponatremia but with no significant neurologic deficit. A percutaneous indwelling catheter was inserted in order to evacuate the massive pericardial effusion for both diagnosis and symptom relief. Interestingly, the patient's severe hyponatremia recovered spontaneously and rapidly without any other treatment. We reviewed the literature and found only four isolated case reports.6–9 Because cardiac tamponade was not previously considered a cause of hyponatremia,1–3 it is not clear whether these observations were simply coincidental or important hints that were heretofore ignored. In this retrospective study, to determine the relationship between hyponatremia and cardiac tamponade, we systemically analyzed the frequency of hyponatremia and the effect of pericardial drainage on serum sodium concentrations in patients with cardiac tamponade.
Because a cytology test is part of the standard group of examinations for undetermined pericardial effusion, we used our cytology database to retrieve cases that had significant and/or undiagnosed pericardial effusion between January 2007 and December 2009. We then performed an electronic chart review to find patients who had cardiac tamponade that warranted percutaneous indwelling catheter drainage or pericardial window surgery for symptom relief. Patients who had serum sodium concentrations checked within 48 h before the evacuation of pericardial effusion were considered eligible for analysis. Those who had paired serum sodium concentrations measured after evacuation (<48 h) and did not receive any specific medication to correct hyponatremia were enrolled for further analysis of the effect of pericardial drainage on hyponatremia. This study was carried out in a tertiary referral medical center and conducted in accordance with the Declaration of Helsinki.
2.2. Statistical analysis
Patients with pretreatment serum sodium concentrations were included and Pearson's chi-square test was used to analyze the differential frequency of hyponatremia in patients with malignancy-related versus. nonmalignancy-related cardiac tamponade. Wilcoxon signed-rank test was used to analyze paired serum sodium concentrations before and after pericardial effusion drainage. A p value less than 0.05 was considered statistically significant. SPSS Statistics 17.0 (SPSS Inc., Chicago, IL, USA) was used to carry out the analysis and Microsoft Excel 2007 (Microsoft Corp., Redmond, WA, USA) was employed to construct the figures.
Fifty-seven patients were identified in a 3-year period and nine patients were excluded due to lack of serum sodium level measurement within 48 h before pericardial drainage. As a result, 48 patients were enrolled and patient characteristics are summarized in Table 1. Most of the patients were male (60%) and the median age was 61 years (range = 23–81). Thirty-seven patients had active malignancy at the time of presentation of cardiac tamponade, and 33 of them were lung cancers. Six patients had benign pericardial effusion, caused by tuberculosis, systemic lupus erythematosus, hypothyroidism, uremia, iatrogenic (caused by percutaneous transluminal coronary angioplasty), and idiopathic (survived without recurrence of effusion and was free from any malignancy for more than 2 years), respectively. A definite diagnosis of pericardial effusion could not be established in five patients. Four of them died soon thereafter, in spite of the effective management of cardiac tamponade, and one was discharged without specific diagnosis and lost to follow-up. The cytology and microbiology examinations were all negative in these five patients.
In this series, 26 patients (54.2%) had hyponatremia. The condition was mild (130–134 mEq/L) in eight; moderate (120–129 mEq/L) in 16; and severe (<120 mEq/L) in two (Table 2). Among the 37 patients exhibiting malignancy-related cardiac tamponade, 23 (62.2%) had hyponatremia and the condition was mild in 6 and moderate to severe in 17. Among those with benign or undetermined pericardial effusion, only three (27.3%) had hyponatremia, with the condition being mild in two and moderate in the other. The presence of hyponatremia was not associated with age or sex; however, it was significantly associated with the underlying disease (malignancy vs. non-malignancy, 62.2% vs. 27.3%, p = 0.041). In patients who had hyponatremia at presentation, 19 had paired serum sodium levels measured within 48 h after evacuation of pericardial effusion. None of them had received specific treatment to correct hyponatremia. The postevacuation serum sodium levels were significantly higher than the pre-evacuation levels (p < 0.001). In fact, all patients had higher post-treatment serum levels with a median increase of 8 mEq/L (range, 1–14 mEq/L). This is shown in Fig. 1. Twelve patients (63.2%) showed an increase by at least 5% and seven of them showed an absolute increase of 10 mEq/L or more. Despite the dramatic changes in serum sodium concentrations, none of the patients developed major neurologic deficit during or after evacuation of pericardial effusion.
Hyponatremia is a common clinical manifestation of several kinds of diseases. The list of differential diagnosis is lengthy;1–3 however, cardiac tamponade is not a part of this list. We reviewed the literature and found only four isolated case reports.6–9 To the best of our knowledge, this study is the first to systemically investigate serum sodium concentrations in patients with cardiac tamponade. We showed that hyponatremia was indeed frequently encountered in patients with cardiac tamponade, at least in those who had active malignancy. In addition, by comparing the serum sodium concentration before and after pericardiocentesis, we provided strong evidence suggesting a causative role of cardiac tamponade in the development of hyponatremia. However, it is noteworthy that only one-half of the patients' serum sodium levels returned to normal range after pericardial drainage. Therefore, we considered cardiac tamponade to be a critical but not the sole contributing factor to hyponatremia in these patients.
The mechanism as to how cardiac tamponade causes hyponatremia is not clear. Previous animal study had shown that circulating ADH, which is stimulated by increased atrial pressure,10 was elevated in dogs with experimental cardiac tamponade.11 Indeed, elevated serum ADH levels were observed in clinical patients.6,12 Then it was suggested that the “inappropriate” secretion of ADH was the principle factor for the development of hyponatremia. However, this proposal was contradicted by the clinical finding that patients' serum sodium levels were paradoxically decreased when they were treated by water restriction.7 Atrial natriuretic factor (ANF) had also been implied to be responsible for hyponatremia in cardiac tamponade. However, it is now clear that atrial distension, but not atrial pressure elevation, causes ANF release from atrial myocardiocyte. In cardiac tamponade, although atrial pressure is elevated, transmural pressure is maintained and then ANF is not secreted.11,13 Reduced effective volume, and then decreased glomerulus filtration rate in the presence of cardiac tamponade may play a key role in the development of hyponatremia. In fact, simple fluid hydration is adequate to correct hyponatremia. However, the reasons why hyponatremia occurs in some patients with cardiac tamponade but not others are still unknown.
In this study, almost all patients experienced spontaneous and rapid recovery of hyponatremia after evacuation of pericardial effusion. We consider this result multifactorial. The relief of cardiac tamponade may (1) reduce atrial pressure and then inhibit further ADH secretion, (2) increase the atrial transmural pressure resulting in atrial distension and ANF release, and (3) improve the effective circulating volume and then increase glomerulus filtration rate. All of these effects will result in diuresis, which is commonly seen in clinical patients undergoing pericardiocentesis, and correction of hyponatremia. In fact, the consequence of pericardial drainage on serum sodium levels has even been demonstrated in an animal model of pericardial effusion.14
Interestingly, our study showed a significantly higher incidence of hyponatremia in patients with malignancy-related than those with nonmalignancy-related cardiac tamponade. In addition, most patients with malignancy-related cardiac tamponade had moderate to severe hyponatremia, but for those without active malignancy history, only one had a serum sodium level less than 130 mEq/L. Although the mechanism is yet to be determined, it is plausible that cancer cells may not only result in large amount of pericardial effusion but also induce significant pericardial fibrosis, which may have an influence on atrial transmural pressure and ANF secretion. A detailed hemodynamic-pathological correlation study is needed to clarify this issue.
From this study, there are at least three important clinical implications. First, cardiac tamponade should be in the differential diagnosis of hyponatremia, especially in patients with active malignant disease. Secondly, it is better to defer treatment until a diagnosis is made. Inappropriate secretion of ADH is the leading cause of hyponatremia in cancer patients,15 and water restriction is the treatment of choice. However, in patients with cardiac tamponade-related hyponatremia, water restriction may actually worsen the hyponatremia. In fact, it could have a detrimental effect on hemodynamic status in these patients. Thirdly, the presence of moderate to severe hyponatremia in patients with cardiac tamponade may serve as a clue for the diagnosis of pericardial effusion, since it was relatively rare in benign pericardial disease in our series.
This study had some limitations. It was a retrospective clinical observation; therefore, the patient population might include some bias. The diagnosis of cardiac tamponade and the decision to undergo pericardiocentesis were made by an individual in-charge physician. We considered that both the echocardiographic findings and the hemodynamic changes in patients with cardiac tamponade were important in the severity evaluation of tamponade, which, in turn, would help to elucidate the underlying pathogenesis of hyponatremia in these patients. However, these data were not informative in a retrospective study. We believe a prospectively designed clinical trial in which echocardiography and hemodynamic evaluation are performed in a predefined methodology is warranted. In addition, we retrieved patients from the cytology database; therefore, we might fail to enroll patients who already been diagnosed before pericardiocentesis, where the resulting pericardial effusion was not sent for cytology examination. This may partially explain the unusual ratio of malignancy-related to non-malignancy-related cardiac tamponade in our cohort.
In conclusion, cardiac tamponade should be considered in the differential diagnosis of hyponatremia, especially in patients with active malignant disease. It is of significant clinical importance because prompt management of cardiac tamponade per se can be sufficient to result in rapid recovery of hyponatremia.
We would like to thank Ms. Tracy Wang for English editing.
1. Adrogué HJ, Madias NE. Hyponatremia. N Engl J Med
2. Lien YH, Shapiro JI. Hyponatremia: clinical diagnosis and management. Am J Med
3. Singer GG, Brenner BM. Fluid and electrolyte disturbances. Fauci AS, Braunwald E, Isselbacher KJ, et al, editors. Harrison's principles of internal medicine. 14th ed. New York: The McGraw-Hill Companies; 1998: pp. 265-277.
4. Berghmans T, Paesmans M, Body JJ. A prospective study on hyponatraemia in medical cancer patients: epidemiology, aetiology and differential diagnosis. Support Care Cancer
5. Doshi SM, Shah P, Lei X, Lahoti A, Salahudeen AK. Hyponatremia in hospitalized cancer patients and its impact on clinical outcomes. Am J Kidney Dis
6. Groves PH, Shah AM, Hutchison SJ. Hyponatraemia secondary to an inappropriately high release of antidiuretic hormone in cardiac tamponade. Br Heart J
7. Mouallem M, Wolf I, Mindlin G, Farfel Z. Pericardial tamponade-associated hyponatremia. Am J Med Sci
8. Shafique R, Sarwar S, Wall BM, Cooke CR. Reversible hyponatremia related to pericardial tamponade. Am J Kidney Dis
9. Weekes MP, Reddi BAJ, Wharton S, Gazis A. Hyponatraemia associated with cardiac tamponade and chronic fluid excess. BMJ Case Reports
Published online 12 July 2010.
10. Ledsome JR, Wilson N. The relationship between plasma vasopressin concentration and urinary excretion during left atrial distension in anaesthetized dogs. Pflugers Arch
11. Mancini GB, McGillem MJ, Bates ER, Weder AB, DeBoe SF, Grekin RJ. Hormonal responses to cardiac tamponade: inhibition of release of atrial natriuretic factor despite elevation of atrial pressures. Circulation
12. Douglas JE, Loon N. Pericardial tamponade and excessive secretion of antidiuretic hormone. South Med J
13. Edwards BS, Zimmerman RS, Schwab TR, Heublein DM, Burnett JC Jr. Atrial stretch, not pressure, is the principal determinant controlling the acute release of atrial natriuretic factor. Circ Res
14. de Laforcade AM, Freeman LM, Rozanski EA, Rush JE. Biochemical analysis of pericardial fluid and whole blood in dogs with pericardial effusion
. J Vet Intern Med
15. Raftopoulos H. Diagnosis and management of hyponatremia in cancer patients. Support Care Cancer