2016 Volume 80 Issue 9 Pages 1965-1970
Background: The monitoring of tissue hypoperfusion and the subsequent neurohumoral activation (ie, arterial underfilling) during decongestion is important for the management of acute decompensated heart failure (ADHF). The transtubular potassium concentration gradient (TTKG) has been reported to be a marker of renal aldosterone bioactivity. This study tested the hypothesis that TTKG can be a surrogate of arterial underfilling in patients with ADHF.
Methods and Results: We measured TTKG at discharge in 100 ADHF patients. The primary outcome measure was the occurrence of tissue hypoperfusion events (defined according to the “Cold Modified 2014” definition criteria) within 1 month after discharge. The secondary outcome measure was the occurrence of cardiac death or ADHF readmission within 3 months after discharge. On receiver operating characteristic curve analysis, TTKG predicted tissue hypoperfusion events with high accuracy (C-statistic, 0.889) for a cut-off of 6.0. Multivariate Cox regression analyses demonstrated independent relationships between TTKG and both the primary and secondary outcomes.
Conclusions: TTKG has utility as a surrogate of arterial underfilling, and spot TTKG at discharge may be a prognostic marker in ADHF patients. (Circ J 2016; 80: 1965–1970)
Inadequate cardiac output in heart failure (HF) may lead to tissue hypoperfusion and arterial underfilling with compensatory neurohumoral activation, including increased activity in the renin-angiotensin-aldosterone system (RAAS), the sympathetic nervous system, and the release of arginine vasopressin.1 This activation of the neurohumoral axis precedes the retention of salt and water that can ultimately lead to the development of acute decompensated HF (ADHF). For this reason, accurately quantifying the degree of neurohumoral axis activation due to arterial underfilling in each patient is crucial for the management of ADHF.
Transtubular potassium concentration gradient (TTKG) was developed to enable the semi-quantitative assessment of the driving force behind potassium secretion in the cortical collecting duct (CCD), which in turn is indicative of the renal bioactivity of aldosterone.2 The validity of this measure is contingent on the following assumptions: (1) few solutes are reabsorbed in the medullary collecting duct (MCD); and (2) potassium is neither secreted nor reabsorbed in the MCD. Urea, however, was found to be reabsorbed in the MCD, and this process is mainly influenced by vasopressin activity.3 Kamel and Halperin subsequently reported that TTKG is not an appropriate index of aldosterone bioactivity because urea recycling in the MCD aids the excretion of potassium.4 Considering that vasopressin regulates urea recycling as well as accelerates the excretion of potassium, TTKG is likely to reflect the bioactivity of both vasopressin and aldosterone. Due to these characteristics and its ease of quantification, TTKG may therefore have applications as a surrogate marker of arterial underfilling in the clinical setting. The aim of this study was to test the hypothesis that TTKG measurements in ADHF patients can be used as a surrogate marker of tissue hypoperfusion and subsequent recurrence of congestion.
This study was approved by the institutional ethics review committee of Osaka National Hospital. The study sample consisted of 100 consecutive patients admitted with a primary diagnosis of ADHF who had been referred to hospital for intensive medical therapy between March 2013 and May 2014. ADHF was defined as new-onset decompensated HF or decompensation of chronic HF with symptoms that warrant hospitalization. All patients were admitted with clinical and radiographic evidence of congestion.5 Patients with serum creatinine (serum Cre) >3.0 mg/dl and patients who had died during hospitalization were excluded from analysis. We ascertained if each patient had been admitted for ADHF within the past 12 months prior to the index admission.
Clinical ProfileCare was provided to each patient under the direction of the treating physician according to Japanese ADHF guidelines. Baseline characteristics such as blood pressure, heart rate, as well as brain natriuretic peptide, serum bilirubin, urea nitrogen (UN), Cre, sodium, potassium, and hemoglobin were recorded both on admission and immediately prior to discharge. Left ventricular ejection fraction (LVEF) and velocity time integral of the LV outflow tract were also determined on echocardiography on admission and immediately prior to discharge. Medications prescribed both on admission and at discharge, as well as comorbidities during hospitalization (including hypertension, diabetes, and atrial fibrillation) were identified. Patients were discharged after the treating physician confirmed restoration of clinical stability.
TTKG AssessmentThe TTKG was measured using the first morning urine samples collected immediately prior to discharge from hospital. TTKG was calculated based on the following equation:
TTKG=(KU/KS)×(plasma osmolality/urine osmolality)
where KU is urine potassium concentration and KS is serum potassium concentration. Plasma osmolality was calculated using the following well-established formula:
Plasma osmolality=2×(NaS+KS)+BUN/2.8+BS/18
where NaS is serum sodium concentration, BUN is blood UN concentration, and BS is blood sugar. Urine osmolality was calculated using the following formula, which has been previously validated in ADHF patients:6
Urine osmolality=1.07×(2×NaU+UNU/2.8+CreU×2/3)+16
where NaU is urine sodium concentration, UNU is urine UN concentration, and CreU is urine Cre concentration.
Follow-up and Outcome MeasuresAll patients were examined prior to discharge. Patients and their families also attended comprehensive discharge education sessions. Patients were discharged with guideline-compliant pharmacologic therapy, and follow-up clinic visits were scheduled for every 2–4 weeks.
The primary clinical outcome measure was occurrence of tissue hypoperfusion events. Nohria et al reported that clinical hypoperfusion (cold hemodynamic profile) assessed during physical examination predicts survival in patients with ADHF.7 That definition, however, had low specificity in the prediction of low cardiac output and was affected by high inter-observer variability. Frea et al proposed a new cold profile definition (Cold Modified 2014) in patients with ADHF that had greater accuracy and better prognostic value than the older definition.8 Based on the Frea et al definition, we identified patients as having a cold profile if they presented with at least 2 of the following clusters within 1 month after discharge: (1) hemodynamic cluster, involving at least one of the following: low pulse pressure (proportional pulse pressure [PPP] <25%) or mean blood pressure (<60 mmHg), hyponatremia (serum sodium concentration <135 mEq/L), cool extremities, or impaired mentation; (2) renal cluster, involving worsening renal function (WRF; elevation of serum Cre concentration compared with immediately prior to discharge >0.3 mg/dl or >25%); and (3) hepatic cluster, involving hyperbilirubinemia (total bilirubin >1.2 mg/dl). The secondary clinical outcome measure was the occurrence of cardiac death or ADHF readmission within 3 months after discharge.
Statistical AnalysisData are given as mean±SD. Statistical significance was set at P<0.05. Receiver operating characteristic (ROC) curve analysis of TTKG at discharge was used to determine the optimal cut-off for predicting tissue hypoperfusion events within 1 month after discharge based on the Youden index. To assess the relationship between the primary and secondary clinical outcomes, the patients were categorized into 2 groups according to prior incidence of tissue hypoperfusion events; next, the occurrence of cardiac events for these groups was compared on Kaplan-Meier analysis. The occurrence of both primary and secondary clinical outcomes was also compared using the same method for the 2 groups categorized by the aforementioned Youden index cut-off of TTKG at discharge. Univariate and multivariate Cox regression analysis was used to identify independent relationships between TTKG and the other characteristics at discharge with the outcome measures. In addition, the relationship between TTKG and various medications (β-blockers, angiotensin-converting enzyme inhibitors [ACEI], angiotensin receptor blockers [ARB], mineralocorticoid receptor antagonists, loop diuretics, and tolvaptan) was similarly analyzed. We included all of the factors in the first multivariate model using a forward stepwise variable selection method (selection criterion: P<0.05). The association of each variable with the outcome is given as hazard ratio (HR) and 95% CI. All statistical analysis was performed using MedCalc for Windows, version 13.1.2.0 (MedCalc Software, Ostend, Belgium).
Subject clinical characteristics are listed in Table 1. The proportion of patients given initial treatment and medication for ADHF were as follows: intubation, 6%; non-invasive positive-pressure ventilation, 7%; i.v. furosemide infusion, 39%; human atrial natriuretic peptide (carperitide), 40%; nitroglycerine, 11%; dobutamine, 22%; and milrinone, 15%.
| Sex (male) | 51 (51) |
| Age (years) | 74±12 (30–95) |
| Ischemic | 15 (15) |
| SBP (mmHg) | 135±34 |
| DBP (mmHg) | 78±18 |
| Heart rate (beats/min) | 88±25 |
| BNP (pg/ml) | 728±734 |
| Serum bilirubin (mg/dl) | 1.1±0.8 |
| Serum UN (mg/dl) | 26±11 |
| Serum Cre (mg/dl) | 1.25±0.51 |
| Serum sodium (mEq/L) | 141±4 |
| Serum potassium (mEq/L) | 4.0±0.4 |
| LVEF (%) | 41±17 |
| LVOT-VTI (cm) | 12.6±4.5 |
| Comorbidities | |
| Hypertension | 69 (69) |
| Diabetes | 33 (33) |
| Atrial fibrillation | 49 (49) |
| Medications | |
| β-blockers | 57 (57) |
| ACEI/ARB | 63 (63) |
| MRA | 38 (38) |
| Loop diuretics | 75 (75) |
| Tolvaptan | 15 (15) |
Data given as n (%) or mean±SD (range). ACEI/ARB, angiotensin-converting enzyme inhibitors/angiotensin receptor blockers; BNP, brain natriuretic peptide; Cre, creatinine; DBP, diastolic blood pressure; LVEF, left ventricular ejection fraction; LVOT-VTI, velocity time integral of the left ventricular outflow tract; MRA, mineralocorticoid receptor antagonists; SBP, systolic blood pressure; UN, urea nitrogen.
A total of 15 patients had tissue hypoperfusion events within 1 month after discharge (hemodynamic cluster, n=17; renal cluster: n=20; hepatic cluster: n=23). None of these 15 patients had met the criteria at discharge. ROC curve analysis of TTKG at discharge for predicting tissue hypoperfusion events is shown in Figure 1. TTKG at discharge was found to predict tissue hypoperfusion events with high accuracy (C-statistic, 0.889) when using a Youden index TTKG cut-off of 6.0 (sensitivity, 80.0%; specificity, 90.6%).
Receiver operating characteristic curve analysis of transtubular potassium concentration gradient at discharge for predicting tissue hypoperfusion events within 1 month after discharge.
Table 2 lists the clinical parameters at discharge according to TTKG at discharge (<6.0 or ≥6.0). Patients with higher TTKG had paradoxically lower UN and Cre concentration compared with patients with lower TTKG. No significant differences were observed in the other parameters.
| TTKG at discharge | P-value | ||
|---|---|---|---|
| Low (<6.0), n=80 | High (≥6.0), n=20 | ||
| Sex (M/F) | 42/38 | 9/11 | 0.835 |
| Age (years) | 73±12 | 77±12 | 0.190 |
| SBP (mmHg) | 114±15 | 117±11 | 0.398 |
| Pulse pressure (mmHg) | 47±11 | 50±10 | 0.301 |
| PPP (%) | 41.4±7.0 | 42.9±6.5 | 0.404 |
| BNP (pg/ml) | 303±231 | 308±272 | 0.930 |
| Serum bilirubin (mg/dl) | 0.8±0.4 | 0.8±0.4 | 0.998 |
| Serum UN (mg/dl) | 29±12 | 23±6 | 0.002 |
| Serum Cre (mg/dl) | 1.31±0.51 | 1.07±0.32 | 0.011 |
| Serum sodium (mEq/L) | 140±4 | 141±3 | 0.177 |
| Serum potassium (mEq/L) | 4.4±0.5 | 4.2±0.4 | 0.059 |
| LVEF (%) | 42±18 | 40±16 | 0.693 |
| LVOT-VTI (cm) | 14.7±3.5 | 12.9±3.7 | 0.096 |
| Medications | |||
| β-blockers | 58 (73) | 13 (65) | 0.804 |
| ACEI/ARB | 63 (79) | 17 (85) | 0.823 |
| MRA | 54 (68) | 13 (65) | 0.978 |
| Loop diuretics (mg/day) | 39±30 | 46±35 | 0.376 |
| Tolvaptan | 19 (24) | 7 (35) | 0.591 |
Data given as n (%) or mean±SD . Oral loop diuretics dosage was converted into furosemide equivalents based on the following: 40 mg of furosemide=60 mg of azosemide=8 mg of torasemide. PPP, proportional pulse pressure; TTKG, transtubular potassium concentration gradient. Other abbreviations as in Table 1.
The TTKG ≥6.0 at discharge was significantly associated with higher 1-month tissue hypoperfusion event rate (TTKG ≥6.0, 60.0%; TTKG <6.0, 3.75%; HR, 22.0; 95% CI: 5.41–89.5, P<0.001) (Figure 2). On multivariate Cox regression analysis there was an independent relationship between TTKG and tissue hypoperfusion events (Table 3).
Kaplan-Meier cumulative incidence of tissue hypoperfusion events vs. transtubular potassium concentration gradient (TTKG) at discharge: high, TTKG ≥6.0; low, TTKG <6.0. HR, hazard ratio.
| Variables at discharge | HR (95% CI) | P-value |
|---|---|---|
| Univariate | ||
| Age (years) | 1.039 (0.985–1.096) | 0.162 |
| SBP per 10 mmHg | 1.212 (0.859–1.709) | 0.276 |
| Pulse pressure per 10 mmHg | 1.275 (0.815–1.994) | 0.290 |
| PPP per 10% | 0.982 (0.473–2.038) | 0.961 |
| BNP per 10 pg/ml | 1.001 (0.999–1.003) | 0.608 |
| Serum bilirubin (mg/dl) | 1.690 (0.672–4.247) | 0.267 |
| Serum UN (mg/dl) | 0.964 (0.914–1.018) | 0.191 |
| Serum Cre per 0.1 mg/dl | 0.906 (0.796–1.032) | 0.138 |
| Serum sodium per 10 mEq/L | 1.531 (0.312–7.522) | 0.602 |
| Serum potassium (mEq/L) | 0.254 (0.079–0.813) | 0.022 |
| LVEF per 10% | 1.070 (0.806–1.422) | 0.641 |
| LVOT-VTI (cm) | 0.738 (0.557–0.977) | 0.035 |
| TTKG | 1.649 (1.365–1.992) | <0.001 |
| Medications | ||
| β-blockers | 0.678 (0.200–2.301) | 0.535 |
| ACEI/ARB | 1.045 (0.297–3.676) | 0.946 |
| MRA | 3.406 (0.774–14.99) | 0.107 |
| Loop diuretics, 10 mg/day | 1.168 (1.026–1.329) | 0.019 |
| Multivariate | ||
| TTKG | 1.681 (1.275–.217) | <0.001 |
Oral loop diuretics dosage was converted into furosemide equivalents based on the following: 40 mg of furosemide=60 mg of azosemide=8 mg of torasemide. CI, confidence interval; HR, hazard ratio. Other abbreviations as in Tables 1,2.
There were 26 cases of the secondary outcome measure (cardiac death or readmission due to ADHF within 3 months after discharge); of these, 24 events were readmission for HF and the remaining events were cases of sudden cardiac death. On Kaplan-Meier analysis there was a significant relationship between the occurrence of tissue hypoperfusion events within 1 month and cardiac events within 3 months after discharge (66.7% vs. 18.8%, respectively; HR, 4.97; 95% CI: 1.47–16.8, P<0.001), and also a significant association between TTKG ≥6.0 at discharge and higher 3-month cardiac event rate (TTKG ≥6.0, 55.0%; TTKG <6.0, 18.8%; H, 3.67; 95% CI: 1.31–10.3, P<0.001) (Figure 3). On multivariate Cox regression analysis LVEF, loop diuretic dose, and TTKG were associated with an increased risk of cardiac death or ADHF readmission within 3 months after discharge (Table 4).
Kaplan-Meier cumulative incidence of cardiac death and acute decompensated heart failure readmission. High transtubular potassium concentration gradient (TTKG), TTKG at discharge ≥6.0; low TTKG, TTKG at discharge <6.0. HR, hazard ratio.
| Variables at discharge | Univariate | Multivariate | ||
|---|---|---|---|---|
| HR (95% CI) | P-value | HR (95% CI) | P-value | |
| Age (years) | 1.014 (0.979–1.049) | 0.450 | ||
| Prior ADHF admissions (last 12 months) | 2.630 (1.212–5.708) | 0.015 | ||
| Serum UN (mg/dl) | 0.976 (0.938–1.016) | 0.239 | ||
| Serum Cre per 0.1 mg/dl | 0.994 (0.918–1.078) | 0.891 | ||
| Serum sodium (mEq/L) | 1.011 (0.902–1.133) | 0.854 | ||
| Hemoglobin (g/dl) | 0.890 (0.727–1.089) | 0.260 | ||
| BNP per 10 pg/ml | 1.015 (1.002–1.028) | 0.025 | ||
| LVEF per 10% | 0.621 (0.455–0.847) | 0.003 | 0.587 (0.411–0.839) | 0.004 |
| TTKG | 1.447 (1.224–1.710) | <0.001 | 1.352 (1.145–1.597) | <0.001 |
| Medications | ||||
| β-blockers | 1.439 (0.580–3.566) | 0.435 | ||
| ACEI/ARB | 0.859 (0.346–2.131) | 0.744 | ||
| MRA | 1.378 (0.582–3.264) | 0.468 | ||
| Loop diuretics per 10 mg/day | 1.191 (1.076–1.319) | <0.001 | 1.129 (1.014–1.257) | 0.028 |
Oral loop diuretics dosage was converted into furosemide equivalents based on the following: 40 mg of furosemide=60 mg of azosemide=8 mg of torasemide. ADHF, acute decompensated heart failure. Other abbreviations as in Tables 1–3.
This study indicates that TTKG at discharge is a sensitive marker for predicting clinical manifestation of tissue hypoperfusion immediately following discharge, and is also a novel prognostic marker in ADHF patients.
Although the prognostic implications of plasma aldosterone concentration have been previously reported in ADHF patients,9 it is impractical as a clinical measure due to its high cost, lengthy time requirements, and complicated conditions for specimen collection. TTKG was originally established to indicate the renal bioactivity of aldosterone through the degree of potassium excretion in the CCD, and has been shown to be useful in the differentiation of abnormalities in aldosterone activity in patients with hyperkalemia or hypokalemia.2 Furthermore, it is readily quantified using plasma and urine electrolyte measurements. Kamel and Halperin, however, reported that urea recycling in the MCD, which is mainly regulated by vasopressin, also aids in the excretion of potassium.4 In consideration of that finding, TTKG is likely to reflect not only the bioactivity of aldosterone, but also that of vasopressin.
Schrier and Abraham proposed the concept of arterial underfilling to describe decreased effective arterial blood volume.10 In cases of arterial underfilling, the sympathetic nervous system and the RAAS work together with the release of non-osmotic vasopressin to maintain arterial perfusion, and this physiological compensatory mechanism can lead to both tissue hypoperfusion and congestion in patients with HF. The careful monitoring and appropriate management of arterial underfilling can prevent deterioration into a clinical state of tissue hypoperfusion. Nohria et al and Stevenson conducted novel assessments of the “cold” state using vital signs, physical findings, and blood tests.7,11 These criteria, however, include several qualitative indices, rendering it difficult to use in the serial assessment of tissue perfusion. If arterial underfilling persists for a prolonged period, congestion can advance and lead to the development of ADHF.
The study has shown that TTKG at discharge can predict tissue hypoperfusion events with a high degree of sensitivity when using a cut-off of 6.0. When the subjects were categorized into 2 groups according to TTKG at discharge (<6.0 or ≥6.0), patients with higher TTKG had paradoxically lower UN and Cre concentration than patients with lower TTKG. No significant differences were observed in the other parameters at discharge. Because the elevation of TTKG reflects preserved tubular function, enabling sufficient potassium excretion in response to the activation of various neurohumoral factors, the basic glomerular function in patients with higher TTKG may have been relatively preserved. Prolonged subclinical tissue hypoperfusion, however, may induce a decline in glomerular function after discharge.
In order to minimize the pharmacological effects of β-blockers, ACEI, ARB, mineralocorticoid receptor antagonists, loop diuretics, and tolvaptan on TTKG without the influence of a hemodynamic mechanism, we measured TTKG immediately prior to morning medication, when the effective concentration of these medications is considered to be the lowest of the day. There were no significant differences in the number of patients of medication between the 2 groups according to TTKG at discharge (<6.0 or ≥6.0). In addition, there were no significant differences in TTKG at discharge between patients with and without medication.
On multivariate Cox regression analysis there was an independent relationship between TTKG at discharge and tissue hypoperfusion events within 1 month after discharge. In contrast, conventional quantitative indices of the degree of tissue perfusion (such as PPP, serum Cre, sodium, and total bilirubin concentration) were not associated with tissue hypoperfusion events. This indicates that TTKG is a useful marker of degree of tissue perfusion, which can be used in the early detection of tissue hypoperfusion events, and that TTKG is therefore also expected to predict the need for inotropic agents during the course of decongestion.
TTKG, together with LVEF and loop diuretic dose, was also found to have an independent relationship with cardiac death and readmission due to ADHF within 3 months after discharge. The observed relationship between 1-month tissue hypoperfusion events and the later cardiac events suggests that subclinical tissue hypoperfusion activates neurohumoral factors by arterial underfilling, subsequently leading to the initiation of hemodynamic congestion. We consider TTKG to be a sensitive predictor of clinical congestion secondary to tissue hypoperfusion.
The rate of rehospitalization during 90 days after discharge in patients with HF with preserved ejection fraction (HFpEF) was similar to that in those with HF with reduced ejection fraction (HFrEF).12 In the present study, however, short-term cardiac events were more frequent in patients with lower LVEF. This difference may be due to the smaller sample size of this study or to the different baseline characteristics of the subjects. Cardiovascular death or hospitalization for HF in Japanese patients with HFrEF is more frequent than in patients with HFpEF, according to a recent CHART-2 study.13 The present study also identified an independent relationship between higher dose of loop diuretics and cardiac event occurrence, which corroborates previous findings of a significant dose-dependent association between chronic use of loop diuretics and worsening prognosis in Japanese patients with chronic HF.14
The present findings should be interpreted in consideration of the following limitation. Because this is a retrospective observational study, the determination of tissue hypoperfusion was dependent on the quality and accuracy of the information in the medical records. Prospective studies are required to verify the validity of TTKG as a marker of arterial underfilling.
Measurement of TTKG, which possibly reflects the combined effects of renal bioactivity of the RAAS and vasopressin, may help to identify ADHF patients at risk of further tissue hypoperfusion and subsequent congestion-related clinical events.
The authors received no external funding for this study.
None.
