Hyponatremia portends a poor prognosis in patients with heart failure (HF). The aim of this study was to evaluate prognostic implication of hyponatremia on adverse events in patients with HF receiving cardiac resynchronization therapy (CRT). Additionally, the impact of improvement of hyponatremia after CRT device implantation was also evaluated. In this retrospective analysis, we included patients in whom a CRT device was implanted between April 2004 and April 2010 at our institution and had a baseline sodium level obtained within 72 hours of implantation. The patients were followed up for 3 years after implantation for subsequent primary composite end points, that is, hospitalization for HF, left ventricular assist device or heart transplant, and all-cause death. Sodium levels were followed up at 3 to 6 months after device implantation. Hyponatremia was defined as a serum sodium level of <135 mmol/L. A total of 402 patients were included (age 68.7 ± 12.3 years, women 20.9%). One hundred seventy-nine adverse events were noted in this period. In a Cox proportional hazards univariate model, hyponatremia (hazard ratio [HR] 1.54, 95% confidence interval [CI] 1.113 to 2.131, p = 0.009), creatinine (HR 1.267, 95% CI 1.156 to 1.389, p <0.001), and diuretics (HR 2.652, 95% CI 1.401 to 5.019, p = 0.003) were associated with occurrence of the composite end point. A total of 57.9% of patients with hyponatremia at baseline had the composite end point compared with 40.7% of those with normal sodium concentration (p = 0.004). Kaplan-Meier curve showed that hyponatremic patients fared worse. Also, patients in whom hyponatremia resolved after CRT device implantation had lower incidence of the composite end point compared with patients who had normal pre-CRT sodium levels but developed hyponatremia later. In conclusion, baseline hyponatremia is associated with poor prognosis in patients with HF. CRT can resolve hyponatremia in some patients after device implantation. Patients with postimplantation hyponatremia (either newly developed or persistent from baseline) have a poor clinical outcome. Post-CRT improvement of hyponatremia is associated with improved clinical outcomes.
Hyponatremia is a common electrolyte abnormality, seen in approximately 20% of patients with heart failure (HF), and it has been particularly demonstrated to be of much prognostic significance in the HF population. However, there is a paucity of data looking at how serum sodium levels impact patients who receive cardiac resynchronization therapy (CRT). This study aims to examine (1) the effect of hyponatremia on patients with HF receiving CRT and (2) the influence of CRT on this electrolyte abnormality and consequently clinical outcomes.
Methods
This was a single-center retrospective study of database comprising 569 consecutive patients who underwent CRT device implantation from April 2004 to April 2010. Patients were followed up postimplantation in the multidisciplinary clinic by integrated care visits with HF, echocardiography, and electrophysiology specialists. The physicians belonging to this group were responsible for defining New York Heart Association class, defining cause of HF, calculating left ventricular ejection fraction (LVEF), and implanting CRT device according to criteria. The database comprised patients’ demographic characteristics, medical history, disease course after device implantation, diagnostic testing before and after device implantation, and drug therapies. Laboratory values for basic metabolic panel inclusive of serum sodium, potassium, chloride, blood urea nitrogen, creatinine, glucose, and bicarbonate were collected at baseline, which was defined as 72 hours before device implantation. Sodium level was obtained at follow-up at 3- to 6-month visit. Hyponatremia was defined as sodium level <135 mEq/L based on the previous description in the OPTIME-CHF (Outcomes of a Prospective Trial of Intravenous Milrinone for Exacerbations of Chronic Heart Failure) trial. Patients underwent a transthoracic echocardiography before implantation and a follow-up echocardiography within a year of the baseline echocardiography (Philips iE33, SONOS 5500/7500; Philips Healthcare, Andover, Massachusetts and Vivid 7; General Electric, Milwaukee, Wisconsin). LVEF was calculated using the biplane method of discs from the apical 4- and 2-chamber views.
All patients were followed up over a period of 3 years for hard clinical end points, that is, all-cause mortality, HF hospitalizations, left ventricular (LV) assist device implantation, and cardiac transplant. HF hospitalization was defined as inpatient admission with signs and/or symptoms of HF, including dyspnea, peripheral edema, and/or congestion on the chest x-ray, and improvement of these signs and/or symptoms with medical therapy. The composite end point included a combination of (1) HF hospitalization, (2) implantation of LV assist device, (3) cardiac transplant, and/or (4) all-cause death. Notably, clinical outcomes were reconfirmed with review of the electronic medical record and comparison with the Social Security Death Index.
Statistical analyses were performed using SPSS, version 21.0 (SPSS Inc., Chicago, Illinois). Results are presented as mean ± SD for continuous variables. Categorical data are summarized as frequencies and percentages. For the assessment of differences chi-square test was applied, as appropriate. Continuous variables were compared by independent samples t test. Kaplan-Meier curves were constructed to compare event rates in hyponatremic and normonatremic groups with respect to the composite end point of HF hospitalization, LV assist device implantation, cardiac transplant, and all-cause death. Cox proportional hazards models were used for cumulative risk assessment of primary end point as described previously. The models were adjusted for clinical variables, which were shown as significant univariate predictors (age, creatinine, hyponatremia, LVEF, LVIDd (Left Ventricular Internal Diameter [diastolic]), LVIDs (Left Ventricular Internal Diameter [systolic]), diabetes mellitus, diuretics, and aldosterone antagonist). Hyponatremia was also assessed separately for hospitalization for HF exacerbation and mortality. Parameters with p <0.05 were selected for multivariate analysis. All statistical tests were considered significant at p <0.05. We also performed analysis based on follow-up sodium levels, and Kaplan-Meier curves were constructed with respect to the composite end point. Sample size calculation was not performed.
Results
Of 569 patients, a total of 402 patients with available serum sodium levels within the defined time period were included. Mean baseline characteristics of the cohort were as follows: age 68.7 ± 12.3 years, 84 (20.9%) were female, 164 had history of diabetes mellitus (40.8%), creatinine 1.51 ± 0.8 mg/dl, LVEF before device implantation 24.2 ± 7.1%, and QRS duration 160.2 ± 27.4 ms. In total, 300 patients (74.6%) had New York Heart Association class III or IV, and there was no significant difference between hyponatremic and normonatremic patients.
Table 1 lists baseline characteristics in detail for the entire cohort. Hyponatremic patients were significantly younger compared with normonatremic patients. They also were more likely to have history of diabetes mellitus and had significant difference in baseline serum creatinine. The baseline LVEF was significantly less in them; and they had significantly larger LV internal diameters, both diastolic and systolic. These patients were also receiving a higher dose of furosemide, and aldosterone antagonist usage was significantly higher in them. There was no significant difference in any other co-morbidities and medication usage. Postimplantation evaluation at 6 months did not show any significant difference in change in New York Heart Association classification and in absolute and percent change in LVEF in both the groups (in regard to baseline serum sodium levels).
| Characteristic (Frequency or Mean) | Whole Cohort (n = 402) | Hyponatremia | p Value | |
|---|---|---|---|---|
| Yes (n = 88) | No (n = 314) | |||
| Age (year) | 68.7 ± 12.3 | 65.4 ± 13.6 | 69.7 ± 11.6 | <0.003 |
| Female | 84 (22%) | 14 (17%) | 70 (22%) | 0.193 |
| Creatinine (mg/dL) | 1.51 ± 0.80 | 1.69 ± 1.20 | 1.47 ± 0.64 | 0.012 |
| NYHA class: III/IV | 300 (75%) | 67 (76%) | 233 (74%) | 0.713 |
| LVEF (%) | 24.2 ± 7.1 | 22.7 ± 6.8 | 24.6 ± 7.2 | 0.026 |
| LVIDd (mm) | 62.09 ± 9.07 | 64.55 ± 10.05 | 61.38 ± 8.69 | 0.006 |
| LVIDs (mm) | 54.22 ± 9.74 | 56.72 ± 11.13 | 53.5 ± 9.22 | 0.022 |
| QRS duration (msec) | 160.24 ± 27.46 | 160.71 ± 32.87 | 160.88 ± 25.5 | 0.837 |
| Coronary bypass | 168 (42%) | 42 (48%) | 126 (40%) | 0.201 |
| Chronic atrial fibrillation | 113 (28%) | 20 (23%) | 93 (30%) | 0.204 |
| Paroxysmal atrial fibrillation | 124 (31%) | 33 (38%) | 91 (29%) | 0.126 |
| Diabetes mellitus | 164 (41%) | 48 (55%) | 116 (37%) | 0.003 |
| Hypertension | 302 (75%) | 66 (75%) | 236 (75%) | 0.976 |
| Coronary artery disease | 272 (68%) | 55 (63%) | 217 (69%) | 0.241 |
| Ischemic cardiomyopathy | 239 (60%) | 49 (56%) | 190 (61%) | 0.396 |
| Percutaneous coronary intervention | 107 (27%) | 23 (26%) | 84 (27%) | 0.908 |
| Valve surgery | 66 (16%) | 15 (17%) | 51 (16%) | 0.857 |
| Medications | ||||
| Angiotensin converting enzyme inhibitor | 242 (60%) | 56 (64%) | 186 (59%) | 0.456 |
| Angiotensin receptor blocker | 79 (20%) | 13 (15%) | 66 (21%) | 0.192 |
| Aldosterone antagonist | 131 (33%) | 42 (48%) | 89 (28%) | 0.001 |
| Beta-blockers | 353 (88%) | 77 (88%) | 276 (88%) | 0.921 |
| Digoxin | 161 (40%) | 41 (47%) | 120 (38%) | 0.157 |
| Diuretics | 355 (88%) | 83 (94%) | 272 (87%) | 0.047 |
| Total daily dose of Lasix (mg) | 68.8 ± 58.3 | 84.8 ± 75.3 | 63.9 ± 51.3 | 0.007 |
In univariate Cox proportional hazards model for predictors of the primary end point, hyponatremia, creatinine, and diuretics were associated with clinical outcome ( Table 2 ). In a multivariate model, creatinine (hazard ratio 1.40, 95% confidence interval 1.238 to 1.595, p <0.001) and diuretics (hazard ratio 3.916, 95% confidence interval 1.791 to 8.694, p = 0.001) continued to remain significantly associated with the composite end point ( Table 2 ). Kaplan-Meier curve for event-free survival showed that patients with hyponatremia had significantly greater incidence of the composite end point compared with normonatremic patients ( Figure 1 ). Chi-square comparison between the 2 groups for the primary composite end point showed that 57.9% of hyponatremic patients had adverse events compared with 40.7% of normonatremic patients (p = 0.004). Notably, in comparison with patients with a normal serum sodium level, the hyponatremic patients were more likely to be hospitalized with HF exacerbation (p = 0.033), but there was not a significant difference between mortality (p = 0.159).
| Characteristic | Univariate | Multivariate | ||
|---|---|---|---|---|
| HR (95% CI) | p Value | HR (95% CI) | p Value | |
| Age | 1.011 (0.998–1.025) | 0.087 | — | — |
| Creatinine | 1.267 (1.156–1.389) | <0.001 | 1.405 (1.238–1.595) | <0.001 |
| Hyponatremia | 1.54 (1.113–2.131) | 0.009 | 1.254 (0.879–1.789) | 0.212 |
| LVEF | 0.995 (0.974–1.016) | 0.612 | — | — |
| LVIDd | 1.007 (0.992–1.023) | 0.33 | — | — |
| LVIDs | 1.008 (0.993–1.023) | 0.33 | — | — |
| Diabetes mellitus | 1.336 (0.998–1.789) | 0.052 | — | — |
| Diuretics | 2.652 (1.401–5.019) | 0.003 | 3.916 (1.791–8.694) | 0.001 |
| Aldosterone antagonist | 0.913 (0.668–1.248) | 0.567 | — | — |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
