Abstract
Background
Serum chloride is an important homeostatic biomarker in left heart failure, with significant prognostic implications. The impact of serum chloride in the long-term survival of patients with pulmonary arterial hypertension (PAH) is unknown. We tested whether serum chloride levels are associated with long-term survival in patients with PAH.
Methods
We included patients with idiopathic or heritable PAH who had a basic metabolic panel performed at the time of their diagnostic right heart catheterization. Laboratory results were recorded both at diagnosis and 6-month follow-up.
Results
We included 277 patients, mean age 51 ± 18 years and 73% women, of whom 254 had a follow-up electrolyte determination at 6 months. Serum chloride was 102.9 ± 3.9 mM/L at diagnosis. A serum chloride ≤ 100 mM/L was noted in 65 (24%) and 53 (21%) patients at diagnosis and 6 months, respectively. Patients with serum chloride ≤ 100 mM/L at 6 months tracked with increase mortality when adjusted by age, sex, pulmonary vascular resistance, diuretics or prostacyclin analogs usage, and serum creatinine and sodium at 6 months (hazard ratio, 1.83; 95% CI, 1.11-3.00). This group of patients was older, with decreased functional capacity, had worse renal function, took more diuretics, had higher pulmonary artery wedge pressure but lower mean pulmonary artery pressure, transpulmonary gradient, and pulmonary vascular resistance.
Conclusions
Serum chloride at 6 months from the PAH diagnosis is a strong and independent predictor of mortality in patients with idiopathic or heritable PAH, even after adjusting serum sodium, renal function, diuretic, and prostacyclin analog usage.
Key Words: electrolyte abnormalities, hypochloremia, outcomes, pulmonary arterial hypertension, serum chloride
Abbreviations: AUC, area under the curve; BNP, brain natriuretic peptide; HR, hazard ratio; IQR, interquartile range; mPAP, mean pulmonary artery pressure; NT-pro-BNP, N-terminal pro BNP; PAH, pulmonary arterial hypertension; PAWP, pulmonary artery wedge pressure; PH, pulmonary hypertension; PVR, pulmonary vascular resistance; RHC, right heart catheterization; SD, standard deviation
Pulmonary arterial hypertension (PAH) is a condition characterized by progressive narrowing of the pulmonary artery that can lead to right heart failure and death.1 PAH can be associated with a variety of conditions, including idiopathic and heritable forms.2 There is a growing need to identify biomarkers that can predict outcomes in PAH. At this point, the only blood biomarker routinely used in clinical practice and recommended by current guidelines is the brain natriuretic peptide (BNP).2, 3 There is mounting evidence to suggest that high creatinine4 and low sodium5 in serum are associated with increased mortality in PAH. In addition, patients with PAH with hyponatremia have worse functional class, more pronounced hemodynamic and ECG severity, and higher hospitalization rates.5, 6
Chloride is the principal extracellular anion accounting for two-thirds of negative charges and one-third of plasma tonicity.7 Serum chloride has been appreciated as an important homeostatic biomarker in left heart failure, with significant prognostic implications.8 In patients with left heart failure, evidence suggests that hypochloremia is associated with increased mortality and that serum chloride may be better than serum sodium in predicting outcomes,9, 10, 11 possibly by reflecting a broader homeostatic imbalance.10 The impact of serum chloride in right heart failure resulting from PAH has remained largely ignored. We hypothesize that in patients with idiopathic or heritable PAH, serum chloride levels are associated with long-term survival, particularly when adjusted by serum sodium levels, renal function, and prostacyclin analog use.
Methods
Patient Selection
This retrospective study was approved by the Cleveland Clinic institutional review board (study number 16-452). Written informed consent was waived. Patients with either idiopathic or heritable PAH were identified from the Cleveland Clinic Pulmonary Hypertension Registry. We included patients who had a basic metabolic panel done at the time of their diagnostic right heart catheterization (RHC) between January 2000 and January 2017. Data from patients seen by the pulmonary vascular group before May 2005 were entered retrospectively, whereas subjects seen after that date were entered prospectively. All patients had precapillary pulmonary hypertension (PH) characterized by a mean pulmonary artery pressure (mPAP) ≥ 25 mm Hg, pulmonary artery wedge pressure (PAWP) ≤ 15 mm Hg, and pulmonary vascular resistance (PVR) > 3 Wood units.2, 12 Two PH specialists (R. A. D. and A. R. T.) agreed on the PAH etiology based on the review of the patient’s medical records and current guidelines.2, 13
Data Collection
At the time of the diagnostic RHC, we collected the serum measurements of sodium, potassium, chloride, bicarbonate, BUN, and creatinine. All blood tests were performed at Cleveland Clinic using standard methodology. At the time of the serum determinations, we recorded whether patients were on diuretics (including type and dose). We also collected data on demographics, BMI, prior diagnosis of hypertension, diabetes, dyslipidemia and OSA, functional status, N-terminal pro BNP (NT-pro-BNP) in serum, distance walked in the 6-min walk test, left atrial area, and right ventricular function (based on visual estimation and tricuspid annular plane systolic excursion) by echocardiography as close as possible to the diagnostic RHC. Hemodynamic measurements included right atrial pressure, mean pulmonary arterial pressure (mPAP), PAWP, transpulmonary gradient (mPAP-PAWP), cardiac index, and PVR. At 6 months from the diagnostic RHC, we recorded the same serum measurements as baseline and documented whether patients were receiving PAH-specific therapies (particularly parenteral prostacyclin analogs) or any diuretics, including type and dose. We estimated the glomerular filtration rate, both at baseline and 6 months, using the Modification of Diet in Renal Disease study equation.14
Statistical Analysis
Continuous variables are summarized using mean ± SD or median and interquartile range (IQR) when appropriate. Categorical variables are presented as n (%). Continuous and categorical variables were tested using t and χ2 tests, respectively. Associations between variables were tested with Pearson correlation coefficient and/or linear regression. We examined the impact of different variables on survival using Kaplan-Meier analysis and Cox proportional hazards model. Cox models were adjusted by potential confounders that were prespecified and included age, sex, use of diuretics or parenteral PAH therapies, and PVR. The starting point for the survival analyses was the date of the electrolyte determination either at the time of the diagnostic RHC (when examining baseline serum chloride) or at 6 months (when testing the 6-month chloride determination or the difference in serum chloride between baseline and 6 months). Patients were censored at the time of lung transplantation and followed until death or end of the study in January 2017. Interactions among variables were checked. Results are given as hazard ratio (HR) with the 95% CI.
Receiver operating characteristic curves with Youden index were used to identify the best cutoff to predict mortality at 2 years. Results are reported as area under the curve (AUC) with 95% CI. All P values were reported as two tailed; a value < .05 was considered statistically significant. Statistical analyses were performed using the statistical packages SPSS, version 17 (IBM), and MedCalc, version 14.12.2 (MedCalc Software bvba).
Results
Baseline Characteristics
We identified 292 consecutive patients with idiopathic or heritable PAH who underwent a diagnostic RHC at Cleveland Clinic. We excluded 15 patients who did not have electrolyte or renal function measurements at the time of the initial RHC. A total of 22 patients did not have available electrolyte or renal function measurements 6 months after the initial RHC and one subject underwent lung transplantation before the 6-month interval; therefore, we included 277 patients in this study with a mean age of 51 ± 18 years (73% women). Idiopathic PAH was present in 243 (88%) patients, whereas the rest had heritable PAH. Functional, echocardiographic, and hemodynamic determinations are presented in Table 1. A total of 254 patients had a follow-up electrolyte determination at 6 months after the diagnostic RHC with a median (IQR) difference between baseline and 6-month serum chloride of 6 (5-8) months.
Table 1.
Patient Characteristics at Time of PAH Diagnosis
| Variable | Value |
|---|---|
| No. | 277 |
| Age (y) | 51.3 ± 18.1 |
| Sex (female) | 203 (73) |
| Race | |
| White | 224 (81) |
| Black | 38 (14) |
| Other | 15 (5) |
| PAH etiology | |
| Idiopathic | 243 (88) |
| Heritable | 34 (12) |
| BMI (kg/m2) | 30.2 ± 8.0 |
| NYHA functional class (III/IV)a | 166 (68) |
| Right ventricular dysfunction (moderate to severe)b | 119 (58) |
| 6MWT distance (m) | 299 ± 120 |
| NT-pro-BNP (pg/mL)c | 1,932 ± 3,824 |
| eGFR (mL/min/1.73 m2) | 75.4 ± 31.9 |
| Hemodynamic measurements | |
| Right atrial pressure (mm Hg) | 10.8 ± 6.4 |
| mPAP (mm Hg) | 53.4 ± 12.9 |
| PAWP (mm Hg) | 9.8 ± 3.7 |
| Transpulmonary gradient (mm Hg) | 43.3 ± 12.7 |
| Cardiac index (L/min/m2) | 2.20 ± 0.61 |
| PVR (Woods unit) | 12.0 ± 6.7 |
Data are presented as mean ± SD or No. (%) unless otherwise indicated. 6MWT = 6-min walk test; eGFR = estimated glomerular filtration rate by Modification of Diet in Renal Disease equation; HR = hazard ratio; mPAP = mean pulmonary artery pressure; NT-pro-BNP = N-terminal pro brain natriuretic peptide; NYHA = New York Heart Association; PAH = pulmonary arterial hypertension, PAWP = pulmonary artery wedge pressure; PVR = pulmonary vascular resistance.
Data available in 226 patients at the time of PAH diagnosis.
Test available in 189 patients at the time of PAH diagnosis.
Test available in 91 patients at the time of PAH diagnosis.
Laboratory Determinations
At baseline, serum chloride, sodium, and creatinine levels were 102.9 ± 3.9 mM/L, 139.2 ± 3.0 mM/L, and 1.03 ± 0.45 mg/dL, respectively (n = 277). Baseline and 6-month laboratory determinations, for subjects who had both, are shown in Table 2 (n = 254). No significant differences were noted between these measurements except for a small decrease in serum potassium (0.07 ± 0.57 mM/L) at 6 months.
Table 2.
Laboratory Determinations at Baseline and 6-Mo Follow-Up
| Determinations | Baseline | 6 Mo | P (Paired t Test, McNemar Test) |
|---|---|---|---|
| No. | 254 | 254 | |
| Sodium (mM/L) | 139.3 ± 2.9 | 139.6 ± 3.4 | .22 |
| Potassium (mM/L) | 4.14 ± 0.54 | 4.07 ± 0.54 | .049 |
| Chloride (mM/L) | 103.1 ± 3.8 | 102.9 ± 3.9 | .38 |
| Bicarbonate (mM/L) | 24.18 ± 3.70 | 24.40 ± 3.39 | .35 |
| BUN (mg/dL) | 18.88 ± 9.41 | 19.87 ± 14.63 | .23 |
| Creatinine (mg/dL) | 1.00 ± 0.38 | 1.03 ± 0.68 | .36 |
| Use of diureticsa | 84 (32.7) | 121 (47.1) | < .001 |
| Use of loop diuretics | 71 (27.6) | 105 (40.9) | < .001 |
Data are presented as mean ± SD or No. (%) unless otherwise indicated.
Diuretics at baseline included furosemide (n = 66), bumetanide (n = 4), hydrochlorothiazide (n = 13), and/or spironolactone (n = 18). At 6 mo, patients received furosemide (n = 100), bumetanide (n = 5), hydrochlorothiazide (n = 14), and/or spironolactone (n = 30). At baseline and 6 months, a total of 19 (7.6%) and 28 (11%) patients received two diuretics.
Serum Chloride
A serum chloride ≤ 100 mM/L (based on Youden index) was noted in 65 (24%) and 53 (21%) patients at diagnosis and 6 months, respectively. Hypochloremia (chloride ≤ 96 mM/L) was noted in 15 patients both at diagnosis and 6 months. A total of 84 (32.7%) and 121 (47.6%) patients received diuretics (predominantly loop diuretics) at diagnosis and 6 months, respectively (Table 2). The majority of patients who were taking diuretics at diagnosis continued taking them at 6 months (71/84 [85%]). The use of diuretics, loop diuretics, doses, or the number of diuretics at diagnosis or at 6 months was not associated with the change in chloride during the first 6 months after diagnosis. In fact, the decrease in serum chloride at 6 months was 0.28 mM/L in patients not on diuretics vs 0.12 mM/L in those receiving diuretics at 6 months (P = .76). Patients receiving parenteral prostacyclin analogs at 6 months (n = 72, 28%) had higher serum chloride (103.9 ± 3.6 vs 102.5 ± 3.9 mM/L, P = .008) and a significant increase in this electrolyte during the course of the first 6 months (+0.60 ± 4.02 vs –0.52 ± 4.06 mM/L, P = .049) when compared with subjects not on this treatment.
At diagnosis, we noted a modest association between serum chloride and sodium (r = 0.60, P < .001), creatinine (r = −0.12, P = .045), BUN (r = −0.23, P < .001), and bicarbonate (r = −0.31, P < .001). Serum chloride at diagnosis was inversely associated with the levels of NT-pro-BNP (r = −0.38, P < .001). Serum chloride at diagnosis was associated with the measurement at 6 months (r = 0.43, P < .001) with a median (IQR) difference between measurements of 0 (−3 to +2) mM/L and a range of −16 to +13 mM/L (−16.7 to 14.0% change).
Survival Analysis
The median (95% CI) survival after the initial and 6-month laboratory determination was 105 (61-149) and 99 (55-143) months, respectively. After adjusting for age, sex, use of diuretics at baseline, and PVR, we noted that only BUN and creatinine predicted survival at PAH diagnosis (Table 3). With the same adjustments, serum chloride, BUN, and creatinine measured at 6 months predicted long-term survival (Table 3). Serum chloride at 6 months remained a significant predictor of survival when adjusted for age, sex, PVR, use of diuretics at 6 months, and treatment with parenteral prostacyclin analogs (HR, 0.92; 95% CI, 0.87-0.97), serum sodium (HR, 0.93; 95% CI, 0.87-0.99), or creatinine (HR, 0.92; 95% CI, 0.88-0.97).
Table 3.
Laboratory Determinations and Adjusted Long-Term Survival
| Variables | HR (95% CI) | P |
|---|---|---|
| Baseline (n = 277) | ||
| Sodiuma | 1.05 (0.99-1.13) | .12 |
| Potassiuma | 0.94 (0.67-1.30) | .69 |
| Chloridea | 1.04 (0.99-1.11) | .11 |
| Bicarbonatea | 0.98 (0.94-1.03) | .43 |
| BUNb | 1.03 (1.01-1.05) | .001 |
| Creatinineb | 1.98 (1.38-2.85) | < .001 |
| 6-month follow-up (n = 254) | ||
| Sodiuma | 0.93 (0.87-0.99) | .02 |
| Potassiuma | 0.76 (0.51-1.13) | .17 |
| Chloridea | 0.92 (0.87-0.96) | .001 |
| Bicarbonatea | 0.99 (0.93-1.05) | .68 |
| BUNb | 1.02 (1.01-1.03) | < .001 |
| Creatinineb | 1.70 (1.39-2.08) | < .001 |
| Change in laboratory determinations at 6 mo (n = 254) | ||
| Sodiuma | 0.99 (0.98-0.99) | .02 |
| Potassiuma | 0.91 (0.65-1.26) | .57 |
| Chloridea | 0.90 (0.85-0.94) | < .001 |
| Bicarbonatea | 0.99 (0.94-1.04) | .70 |
| BUNb | 1.01 (0.99-1.03) | .09 |
| Creatinineb | 0.99 (0.87-1.14) | .94 |
Cox survival model adjusted by age, sex, use of diuretics, and pulmonary vascular resistance. See Table 1 legend for expansion of abbreviations.
Per 1 mM/L increase.
Per 1 mg/dL increase.
Of the 254 patients who had a laboratory determination at 6 months, 235 were followed for at least 2 years; during this interval, 59 (23.2%) subjects died. Serum chloride at 6 months showed an AUC (95% CI) of 0.69 (0.63-0.75) for discriminating patients who died during the 2-year follow-up. A serum chloride at 6 months ≤ 100 mM/L (Youden index) carried a sensitivity of 45.8%, specificity of 87.5%, and positive likelihood ratio of 3.7 for predicting 2-year mortality. The difference in serum chloride between baseline and 6 months showed an AUC of 0.69 (0.61-0.78); meanwhile, serum sodium, BUN, and creatinine at 6 months revealed an AUC (95% CI) of 0.60 (0.50-0.69), 0.67 (0.59-0.75), and 0.61 (0.53-0.70), respectively, for predicting 2-year mortality.
A serum chloride level ≤ 100 mM/L at 6 months tracked with increase mortality in a Cox survival model adjusted by age, sex, PVR, and use of diuretics at 6 months (HR, 2.36; 95% CI, 1.52-3.68) (Fig 1A). Serum chloride ≤ 100 mM/L at 6 months remained a significant predictor of mortality even when adding to the model treatment with parenteral prostacyclin analogues, serum creatinine, and sodium measured at 6 months (HR, 1.83; 95% CI, 1.11-3.00) (Fig 1B). Furthermore, these results remained significant even when using estimated glomerular filtration rate instead of creatinine (HR, 2.07; 95% CI, 1.24-3.44) or including in the model number of PH-specific therapies (0-3) taken at 6 months (HR, 1.72; 95% CI, 1.04-2.85). The negative impact on long-term survival of a serum chloride level ≤ 100 mM/L was sustained when only considering the subgroup of patients who were not receiving diuretics at the time of the 6-month laboratory determination (data not shown). A serum chloride level ≤ 96 mM/L at 6 months also tracked with increased mortality in a Cox survival model adjusted by age, sex, PVR, and use of diuretics at 6 months (HR, 2.95; 95% CI, 1.51-5.78). Serum chloride ≤ 96 mM/L at 6 months remained a significant predictor of mortality after adding to the model treatment with parenteral prostacyclin analogs, serum creatinine, and sodium measured at 6 months (HR, 2.03; 95% CI, 1.00-4.33).
Figure 1.
Cox survival analysis stratified on a serum chloride level ≤ 100 or > 100 mM/L 6 months from the diagnostic right heart catherization. A, Serum chloride ≤ 100 mM/L at 6 months tracked with increase mortality when adjusted by age, sex, pulmonary vascular resistance, and use of diuretics at 6 months. B, This significant difference was sustained when including in the model treatment with parenteral prostacyclin analog, serum creatinine, and sodium measured at 6 months. Cum = cumulative; HR = hazard ratio.
We created four subgroups of patients based on serum chloride concentration > 100 or ≤ 100 mM/L at diagnosis and 6 months. A total 171 (67%) patients had serum chloride levels > 100 mM/L at diagnosis and 6 months, 30 (12%) had chloride levels ≤ 100 mM/L at diagnosis only, 27 (11%) had chloride levels ≤ 100 mM/L at 6 months only, whereas the rest (n = 26, 10%) had serum chloride levels ≤100 mM/L at both time points. In the Kaplan-Meier survival analysis, we noted a separation (log rank test, P < .001) between individuals with Cl ≤ 100 mM/L at 6 months compared with those with Cl > 100 mM/L both at diagnosis and 6 months or Cl ≤ 100 mM/L at diagnosis only (Fig 2A). In a Cox survival analysis adjusted for age, sex, PVR, and use of diuretics at diagnosis and 6 months, we also observed that individuals with serum chloride ≤ 100 mM/L at 6 months did worse, irrespective of the chloride level at baseline (Fig 2B).
Figure 2.
Impact on survival of serum chloride concentration at pulmonary arterial hypertension diagnosis and 6-month follow-up. A, Kaplan-Meier survival analysis showing worse survival in individuals with serum chloride ≤ 100 mM/L at 6 months (log rank test, P < .001). B, Cox survival analysis adjusted by age, sex, pulmonary vascular resistance, and use of diuretics at diagnosis and 6 months, revealing a lower survival in individuals with serum chloride ≤ 100 mM/L at 6 months (HR, 2.35; 95% CI, 1.28-4.27 for the comparison between patients with chloride ≤ 100 mM/L only at 6 months vs those with serum chloride > 100 mM/L at diagnosis and 6 months or serum chloride ≤ 100 mM/L at diagnosis only). bl = baseline. See Figure 1 legend for expansion of other abbreviations.
Factors Associated With Serum Chloride Levels 6 Months After PAH Diagnosis
Patients with serum chloride level ≤ 100 mM/L at 6 months were older, had higher prevalence of diabetes and hypertension, walked less in the 6-min walk test, had worse renal function, took more diuretics, and had larger left atrium and higher PAWP but lower mPAP, transpulmonary gradient, and PVR (Table 4).
Table 4.
Characteristics of Patients With Serum Chloride ≤ 100 mM/L or > 100 mM/L at 6 Months From the Diagnostic RHC
| Variables | Cl ≤ 100 mM/L | Cl > 100 mM/L | P (t Test or χ2) |
|---|---|---|---|
| No. | 53 | 201 | |
| Age (y) | 57.6 ± 17.6 | 49.1 ± 18.1 | .002 |
| Sex (female) | 36 (68) | 150 (75) | .33 |
| Race | |||
| White | 40 (75) | 161 (80) | .70 |
| Black | 9 (17) | 28 (14) | |
| Others | 4 (8) | 12 (6) | |
| PAH etiology | |||
| Idiopathic | 50 (94) | 172 (86) | .09 |
| Heritable | 3 (6) | 29 (14) | |
| Diagnosis of diabetes mellitus (yes) | 18 (34) | 35 (17) | .008 |
| Diagnosis of hypertension (yes) | 28 (53) | 58 (29) | .001 |
| Diagnosis of dyslipidemia (yes) | 13 (25) | 46 (23) | .94 |
| Diagnosis of OSA (yes) | 11 (21) | 37 (18) | .85 |
| BMI at diagnosis (kg/m2) | 30.8 ± 7.8 | 30.0 ± 8.1 | .53 |
| NYHA functional class III/IV at diagnosisa | 32 (76) | 122 (66) | .22 |
| Left atrium dilation (mild or more) | 16 (30) | 33 (16) | .03 |
| Left atrium area (cm2) | 19.8 ± 6.6 | 16.7 ± 4.6 | .003 |
| RV function (moderate to severe) at diagnosisb | 18 (51) | 87 (57) | .59 |
| 6MWT distance at diagnosis (m) | 263 ± 108 | 315 ± 118 | .004 |
| NT-pro-BNP at diagnosis (pg/mL)c | 1,591 ± 2,370 | 1,399 ± 1,911 | .73 |
| Parenteral prostacyclin analogs at 6 mo | 8 (15) | 64 (32) | .02 |
| Hemodynamic measurements at diagnosis | |||
| RA pressure (mm Hg) | 11.4 ± 6.9 | 10.6 ± 6.3 | .46 |
| mPAP (mm Hg) | 48.8 ± 9.7 | 54.6 ± 13.0 | .003 |
| PAWP (mm Hg) | 11.0 ± 3.9 | 9.6 ± 3.7 | .02 |
| Transpulmonary gradient (mm Hg) | 37.4 ± 10.3 | 44.7 ±12.4 | < .001 |
| Cardiac index (L/min/m2) | 2.32 ± 0.57 | 2.18 ± 0.59 | .15 |
| PVR (Woods unit) | 9.2 ± 4.2 | 12.5 ± 6.7 | .001 |
| eGFR (mL/min/1.73 m2)d | |||
| Baseline | 69.1 ± 24.3 | 78.5 ± 33.4 | .03 |
| 6 mo | 70.4 ± 29.2 | 79.5 ± 30.0 | .05 |
| Creatinine at 6 mo (mg/dL)b | 1.31 ± 1.23 | 0.96 ± 0.42 | .04 |
| Diuretics | |||
| Baseline | 30 (57) | 52 (26) | < .001 |
| 6 mo | 34 (64) | 87 (44) | .007 |
| Loop diuretics | |||
| Baseline | 26 (49) | 44 (22) | < .001 |
| 6 mo | 29 (55) | 76 (38) | .03 |
| Furosemide equivalent dose (mg/d) | |||
| Baseline | 46.5 ± 22.4 | 44.9 ± 30.8 | .81 |
| 6 mo | 43.6 ± 26.1 | 37.8 ± 31.1 | .37 |
| More than one diuretic at 6 mo | 10 (20) | 16 (8) | .01 |
Data are presented as mean ± SD or No. (%) unless otherwise indicated. Of the patients with Cl ≤ 100 mM/L, 43% were on diuretics continued from the time of the diagnostic RHC; 17% started and 11% discontinued this medication during the first 6 months. Meanwhile, 30% did not receive diuretics during this interval. Diabetes mellitus was defined as a fasting plasma glucose ≥ 126 mg/dL or drug treatment for the condition. Hypertension was defined as BP ≥ 130/85 mm Hg or drug treatment for the condition. Dyslipidemia was defined as a fasting high-density lipoprotein cholesterol < 40 mg/dL in men or < 50 mg/dL in women or triglycerides ≥ 150 mg/dL or drug treatment for the condition. Cl = chloride; RA = right atrial; RHC = right heart catheterization; RV: right ventricular. See Table 1 legend for expansion of other abbreviations.
Data available in 226 patients.
Test available in 189 patients.
Test available in 91 patients at the time of diagnosis.
eGFR by Modification of Diet in Renal Disease equation.
Discussion
In a relatively large cohort of patients with well-characterized idiopathic or heritable PAH, we evaluated whether the serum chloride concentration either at the diagnostic RHC or at 6-month follow-up predicts long-term mortality. We noted that the prevalence of hypochloremia (chloride ≤ 96 mM/L) at PAH diagnosis and 6-month follow-up was low (∼6%). At the time of PAH diagnosis, serum chloride levels did not significantly affect long-term survival; however, a lower serum chloride 6 months from PAH diagnosis or a decrease in serum chloride during the first 6 months after diagnosis was a significant predictor of long-term mortality, even when adjusted for the use of diuretics or parenteral prostacyclin analogs, serum creatinine, and sodium. A lower serum chloride at 6 months was associated with older age, worse functional capacity and renal function, and a larger left atrium and higher PAWP.
Chloride is an essential anion that helps maintain electrical neutrality, osmotic tone, and acid-base balance.15 Serum chloride determinations are broadly available and a low level may reflect dilution (water gain), depletion (urine loss from renal wasting or diuretic effect), or a response to an acid base disturbance (eg, conditions leading to high serum bicarbonate). Serum chloride plays a role in (1) neurohormonal activation by suppressing plasma renin activity and (2) sodium homeostasis by participating in renal salt sensing, tubuloglomerular feedback, and regulation of sodium transporter pathways through phosphorylation of serine-threonine kinases.16, 17, 18, 19, 20
The value of a low serum chloride remained unnoticed until a series of studies in patients with left heart failure showed that a lower serum chloride is associated with increased mortality and that this determination performed better than serum sodium in predicting outcomes.9, 10, 11, 21 Hypochloremia in left heart failure is attributed to renal wasting of chloride relative to sodium, diuretic resistance, and maladaptive neurohormonal activation (increase in arginine vasopressin and renin-angiotensin-aldosterone system).18
Up to this point, there have been no investigations testing whether lower serum chloride carries poor prognosis in PAH. Patients with right heart failure commonly have hypervolemia in part related to neurohormonal activation22 and increased sympathetic activity.23, 24, 25 In our study, we noted that at the time of PAH diagnosis, serum chloride was associated directly with serum sodium and inversely with serum creatinine, BUN, bicarbonate, and NT-pro-BNP, as seen in patients with left heart failure.9, 11, 18, 21, 26, 27 In our cohort, a serum chloride level ≤ 96 mM/L was uncommon; however, chloride levels above this cutoff appear to offer valuable information. In fact, subjects with serum chloride levels ≤ 100 mM/L, 6 months after the diagnostic RHC, were older, had decreased functional capacity, worse renal function, and higher use of diuretics. These conditions suggest a congestive state with a more pronounced neurohormonal activation and cardiorenal dysfunction.10, 11, 21, 27 Furthermore, we noted that patients with serum chloride ≤ 100 mM/L at 6 months had a higher baseline PAWP, larger left atrium, and higher prevalence of diabetes mellitus and hypertension, conditions that may suggest a degree of overlap with group 2 PH in these subjects.
We found that a lower chloride level at 6 months but not at the time of PAH diagnosis predicted long-term mortality. Of the serum determinations recorded at the time of PAH diagnosis, only serum creatinine and BUN significantly affected survival. It is possible that in this treatment naïve cohort, the initiation of PAH-specific therapies mitigates its prognostic value. However, over the course of the first 6 months, the change in serum chloride may reflect variations in the volume status, renal function, and neurohormonal axis. In fact, serum chloride increased at 6 months in patients who received parenteral prostacyclin analogues. A low serum chloride may indicate a more pronounced neurohormonal activation and renal dysfunction that track with the severity of cardiopulmonary derangements and affect prognosis.22 Depletion in serum chloride from renal losses in patients taking diuretics is certainly possible; however, in our analysis, we did not find a significant association between the use of diuretics, the type, doses, or number of diuretics and the change in serum chloride at 6 months. A study in patients with acute decompensated left heart failure and volume overload showed that a new or persistent hypochloremia at 2 weeks was associated with reduced survival in contrast with subjects in whom the hypochloremia resolved.19 Similarly, in our study, individuals in whom a lower serum chloride improved over the course of 6 months had a better survival than those with lower serum chloride that persisted at 6 months or appeared de novo.
Limitations of the present study include: (1) given the post hoc nature, limitations inherent to retrospective data analysis are therefore applicable, (2) a total of 38 (13%) did not have serum chloride levels tested at 6 months, (3) NT-pro-BNP was not used as a covariate in our models given that this determination was available in only 91 (35%) patients at the time of PAH diagnosis; and (4) results may not be pertinent to other PH groups or causes of PAH beside the idiopathic or heritable form.13 In the present study, we used a serum chloride cutoff of 100 mM/L (based on Youden index) instead of 96 mM/L (commonly used in the literature to define hypochloremia) given the limited number of patients (n = 15) with serum chloride levels ≤ 96 mM/L. In spite of these limitations, this is the first study to assess the predictive value of serum chloride at the time of diagnosis and at 6 months in a cohort of well-characterized patients with idiopathic or heritable PAH. In fact, a serum chloride level ≤ 100 mM/L at 6 months showed a good specificity (87.5%) for identifying patients that died over the course of 2 years. Although we showed that a lower serum chloride levels at 6 months from PAH diagnosis are associated with worse survival, it remains critical to better understand the underpinnings of this electrolyte abnormality to be able to intervene with safe and effective treatments (eg, sodium-free chloride salts such as choline chloride or lysine chloride18).
Conclusions
Serum chloride at 6 months from PAH diagnosis is a strong and independent predictor of mortality in patients with idiopathic or heritable PAH, even when adjusted by serum sodium levels, renal function, and diuretic or parenteral prostacyclin analog use. Consistent with studies in left heart failure, a lower serum chloride level tracked with worse functional class and renal function as well as more frequent use of diuretics.
Acknowledgments
Author contributions: T. N. participated in the design of the study, data collection, statistical analysis, interpretation of the results, writing and critical revision of the manuscript for important intellectual content, and final approval of the manuscript submitted. B. A. participated in the data collection, interpretation of the results, writing and critical revision of the manuscript for important intellectual content, and final approval of the manuscript submitted. G. K. participated in the data collection, interpretation of the results, writing and critical revision of the manuscript for important intellectual content, and final approval of the manuscript submitted. R. A. D. participated in the interpretation of the results and critical revision of the manuscript for important intellectual content and final approval of the manuscript submitted. W. H. W. T. participated in the interpretation of the results and critical revision of the manuscript for important intellectual content and final approval of the manuscript submitted. A. R. T. participated in the design of the study, statistical analysis, and interpretation of the results; writing and critical revision of the manuscript for important intellectual content; and final approval of the manuscript submitted. A. R. T. is the guarantor of the paper, taking responsibility for the integrity of the work as a whole, from inception to published article.
Financial/nonfinancial disclosure: None declared.
Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript.
Footnotes
FUNDING/SUPPORT: Dr Tonelli is supported by National Institutes of Health [Grant R01HL130307].
References
- 1.Tonelli A.R., Arelli V., Minai O.A. Causes and circumstances of death in pulmonary arterial hypertension. Am J Respir Crit Care Med. 2013;188(3):365–369. doi: 10.1164/rccm.201209-1640OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Galie N., Humbert M., Vachiery J.L. 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension: The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS): Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC), International Society for Heart and Lung Transplantation (ISHLT) Eur Heart J. 2016;37(1):67–119. doi: 10.1093/eurheartj/ehv317. [DOI] [PubMed] [Google Scholar]
- 3.McLaughlin V.V., Gaine S.P., Howard L.S. Treatment goals of pulmonary hypertension. J Am Coll Cardiol. 2013;62(25 Suppl):D73–D81. doi: 10.1016/j.jacc.2013.10.034. [DOI] [PubMed] [Google Scholar]
- 4.Shah S.J., Thenappan T., Rich S., Tian L., Archer S.L., Gomberg-Maitland M. Association of serum creatinine with abnormal hemodynamics and mortality in pulmonary arterial hypertension. Circulation. 2008;117(19):2475–2483. doi: 10.1161/CIRCULATIONAHA.107.719500. [DOI] [PubMed] [Google Scholar]
- 5.Forfia P.R., Mathai S.C., Fisher M.R. Hyponatremia predicts right heart failure and poor survival in pulmonary arterial hypertension. Am J Respir Crit Care Med. 2008;177(12):1364–1369. doi: 10.1164/rccm.200712-1876OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rabinovitz A., Raiszadeh F., Zolty R. Association of hyponatremia and outcomes in pulmonary hypertension. J Card Fail. 2013;19(8):550–556. doi: 10.1016/j.cardfail.2013.05.014. [DOI] [PubMed] [Google Scholar]
- 7.Yunos N.M., Bellomo R., Story D., Kellum J. Bench-to-bedside review: chloride in critical illness. Crit Care. 2010;14(4):226. doi: 10.1186/cc9052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Vaduganathan M., Pallais J.C., Fenves A.Z., Butler J., Gheorghiade M. Serum chloride in heart failure: a salty prognosis. Eur J Heart Fail. 2016;18(6):669–671. doi: 10.1002/ejhf.546. [DOI] [PubMed] [Google Scholar]
- 9.Grodin J.L., Simon J., Hachamovitch R. Prognostic role of serum chloride levels in acute decompensated heart failure. J Am Coll Cardiol. 2015;66(6):659–666. doi: 10.1016/j.jacc.2015.06.007. [DOI] [PubMed] [Google Scholar]
- 10.Grodin J.L., Verbrugge F.H., Ellis S.G., Mullens W., Testani J.M., Tang W.H. Importance of abnormal chloride homeostasis in stable chronic heart failure. Circ Heart Fail. 2016;9(1):e002453. doi: 10.1161/CIRCHEARTFAILURE.115.002453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Testani J.M., Hanberg J.S., Arroyo J.P. Hypochloraemia is strongly and independently associated with mortality in patients with chronic heart failure. Eur J Heart Fail. 2016;18(6):660–668. doi: 10.1002/ejhf.477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hoeper M.M., Bogaard H.J., Condliffe R. Definitions and diagnosis of pulmonary hypertension. J Am Coll Cardiol. 2013;62(25 Suppl):D42–D50. doi: 10.1016/j.jacc.2013.10.032. [DOI] [PubMed] [Google Scholar]
- 13.Simonneau G., Gatzoulis M.A., Adatia I. Updated clinical classification of pulmonary hypertension. J Am Coll Cardiol. 2013;62(25 Suppl):D34–D41. doi: 10.1016/j.jacc.2013.10.029. [DOI] [PubMed] [Google Scholar]
- 14.Levey A.S., Bosch J.P., Lewis J.B., Greene T., Rogers N., Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130(6):461–470. doi: 10.7326/0003-4819-130-6-199903160-00002. [DOI] [PubMed] [Google Scholar]
- 15.Berend K., van Hulsteijn L.H., Gans R.O. Chloride: the queen of electrolytes? Eur J Intern Med. 2012;23(3):203–211. doi: 10.1016/j.ejim.2011.11.013. [DOI] [PubMed] [Google Scholar]
- 16.Packer M., Lee W.H., Kessler P.D., Gottlieb S.S., Bernstein J.L., Kukin M.L. Role of neurohormonal mechanisms in determining survival in patients with severe chronic heart failure. Circulation. 1987;75(5 Pt 2):IV80–IV92. [PubMed] [Google Scholar]
- 17.Ponce-Coria J., San-Cristobal P., Kahle K.T. Regulation of NKCC2 by a chloride-sensing mechanism involving the WNK3 and SPAK kinases. Proc Natl Acad Sci U S A. 2008;105(24):8458–8463. doi: 10.1073/pnas.0802966105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hanberg J.S., Rao V., Ter Maaten J.M. Hypochloremia and diuretic resistance in heart failure: mechanistic insights. Circ Heart Fail. 2016;9(8):e003180. doi: 10.1161/CIRCHEARTFAILURE.116.003180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kotchen T.A., Luke R.G., Ott C.E., Galla J.H., Whitescarver S. Effect of chloride on renin and blood pressure responses to sodium chloride. Ann Intern Med. 1983;98(5 Pt 2):817–822. doi: 10.7326/0003-4819-98-5-817. [DOI] [PubMed] [Google Scholar]
- 20.Piala A.T., Moon T.M., Akella R., He H., Cobb M.H., Goldsmith E.J. Chloride sensing by WNK1 involves inhibition of autophosphorylation. Sci Signal. 2014;7(324):ra41. doi: 10.1126/scisignal.2005050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ferreira J.P., Girerd N., Duarte K. Serum chloride and sodium interplay in patients with acute myocardial infarction and heart failure with reduced ejection fraction: an analysis from the High-Risk Myocardial Infarction Database Initiative. Circ Heart Fail. 2017;10(2):e003500. doi: 10.1161/CIRCHEARTFAILURE.116.003500. [DOI] [PubMed] [Google Scholar]
- 22.Nootens M., Kaufmann E., Rector T. Neurohormonal activation in patients with right ventricular failure from pulmonary hypertension: relation to hemodynamic variables and endothelin levels. J Am Coll Cardiol. 1995;26(7):1581–1585. doi: 10.1016/0735-1097(95)00399-1. [DOI] [PubMed] [Google Scholar]
- 23.Ciarka A., Doan V., Velez-Roa S., Naeije R., van de Borne P. Prognostic significance of sympathetic nervous system activation in pulmonary arterial hypertension. Am J Respir Crit Care Med. 2010;181(11):1269–1275. doi: 10.1164/rccm.200912-1856OC. [DOI] [PubMed] [Google Scholar]
- 24.Velez-Roa S., Ciarka A., Najem B., Vachiery J.L., Naeije R., van de Borne P. Increased sympathetic nerve activity in pulmonary artery hypertension. Circulation. 2004;110(10):1308–1312. doi: 10.1161/01.CIR.0000140724.90898.D3. [DOI] [PubMed] [Google Scholar]
- 25.Damman K., Navis G., Smilde T.D. Decreased cardiac output, venous congestion and the association with renal impairment in patients with cardiac dysfunction. Eur J Heart Fail. 2007;9(9):872–878. doi: 10.1016/j.ejheart.2007.05.010. [DOI] [PubMed] [Google Scholar]
- 26.Grodin J.L., Sun J.L., Anstrom K.J. Implications of serum chloride homeostasis in acute heart failure (from ROSE-AHF) Am J Cardiol. 2017;119(1):78–83. doi: 10.1016/j.amjcard.2016.09.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ter Maaten J.M., Damman K., Hanberg J.S. Hypochloremia, diuretic resistance, and outcome in patients with acute heart failure. Circ Heart Fail. 2016;9(8):e003109. doi: 10.1161/CIRCHEARTFAILURE.116.003109. [DOI] [PubMed] [Google Scholar]