The characteristics and outcomes in patients with acute heart failure who used tolvaptan: from KCHF registry

Study population

The Kyoto Congestive Heart Failure (KCHF) registry is a physician-initiated, prospective, observational, multicentre cohort study enrolling consecutive patients who were hospitalized due to AHF for the first time between 1 October 2014 and 31 March 2016 in the 19 participating secondary and tertiary hospitals in Japan. The overall designs of the KCHF registry have been previously described in detail.^{15, 16} In brief, we enrolled consecutive patients with AHF, as defined by the modified Framingham criteria, who were admitted to the participating centres and who underwent HF-specific treatment involving intravenous drugs administered within 24 h of hospital presentation.

Among consecutive 4056 patients enrolled in the KCHF registry, the current study population consisted of 3802 patients after excluding 44 patients on dialysis, 169 patients with tolvaptan use prior to admission, 19 patients with unknown timing for tolvaptan use, and 22 patients with unknown tolvaptan use (Figure 1). There were 773 patients in the tolvaptan use group and 3029 patients in the no tolvaptan use group. Based on the median time for starting tolvaptan, we defined early tolvaptan use as that starting within 48 h from hospital presentation, whereas we defined late tolvaptan use as that starting beyond 48 h. There were 378 patients in the early tolvaptan use group and 395 patients in the late tolvaptan use group.

In the present study, we sought to compare the baseline characteristics, discharge status, and clinical outcomes at 90 days after admission between the two groups with and without tolvaptan use. We also sought to identify the factors associated with tolvaptan use. Moreover, we sought to compare the baseline characteristics between the two groups with early and late tolvaptan use. Finally, we sought to evaluate the inter-hospital variation for tolvaptan use and for the timing of tolvaptan use.

Ethics

The investigation conformed to the principles outlined in the Declaration of Helsinki. The study protocol was approved by the ethics committee in Kyoto University Hospital (local identifier: E2311) and each participating hospital. A waiver of written informed consent was granted by the institutional review boards of Kyoto University and each participating centre, as the study met the conditions outlined in the Japanese ethical guidelines for medical and health research involving human subjects.¹⁷ We disclosed the details of the present study to the public as an opt-out method and informed the patients of their right to refuse enrolment. Details were described in the supporting information. This study was registered with UMIN (UMIN identifier: UMIN000015238).

Data collection and definitions

We collected data on patient demographics, medical history, underlying heart disease, pre-hospital activities, socioeconomic status, signs, symptoms, medications, laboratory tests at hospital presentation, electrocardiogram, echocardiography, and clinical events during the index hospitalization by the attending physicians or research assistants at each participating hospital. The definitions for these data were previously reported and summarized in the supporting information.¹⁵

The attending physicians or research assistants at each participating hospital collected data regarding clinical events that occurred during 90 days of follow-up after admission from the hospital charts. The outcome measures for the current study were all-cause death, cardiovascular death, and non-cardiovascular death. The causes of death were classified according to the VARC (Valve Academic Research Consortium) definitions.¹⁸ A clinical event committee adjudicated all the endpoint events.^{15, 16} We also assessed worsening HF (WHF) and WRF. WHF during the index hospitalization was defined as additional intravenous drug treatment for HF, haemodialysis, or mechanical circulatory or respiratory support, occurring >24 h after therapy initiation, while WRF was defined as >0.3 mg/dL increase in serum creatinine during the index hospitalization.^{19, 20} Valvular heart disease was classified as moderate to severe aortic stenosis, aortic regurgitation, mitral stenosis, mitral regurgitation (excluding functional mitral regurgitation), tricuspid regurgitation (TR), and prosthetic valve dysfunction.

Statistical analyses

The categorical variables were presented as numbers and percentages and were compared using a χ² test or Fisher's exact test. The continuous variables were expressed as means (standard deviations) or medians [interquartile ranges (IQRs)]. On the basis of their distributions, the continuous variables were compared using the Student's t-test or the Wilcoxon rank sum test.

We used a multivariable logistic regression model to explore the factors associated with tolvaptan use. We included the following 19 variables based on the clinical relevance including those factors related to the volume status: age (over 80 years old), prior HF hospitalization, diabetes mellitus, poor medical adherence, high-dose loop diuretics, mineralocorticoid receptor antagonists (MRAs), New York Heart Association (NYHA) classification III or IV, oedema, pleural effusion, hypotension (systolic blood pressure below 90 mmHg), anaemia, hypoalbuminaemia (albumin level < 3.0 g/dL), hyponatraemia (sodium level < 135 mEq/L), estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73 m², left ventricular ejection fraction (LVEF) < 40% by echocardiography, moderate or severe TR, infusion therapy of carperitide within 24 h, infusion therapy of dobutamine within 24 h, and infusion therapy of additional inotropes 24 h after admission. The odds ratios (ORs) and 95% confidence intervals (CIs) were calculated.

To estimate the risk of tolvaptan use relative to no tolvaptan use for the clinical outcome measures at 90 days after admission, we used Cox proportional hazard models incorporating tolvaptan use as a time-updated covariate accounting for the time to start tolvaptan. We included the following clinically relevant 23 risk-adjusting variables shown in Table 1 into the model consistent with previous studies.²¹ The adjusted hazard ratios (HRs) and 95% CIs were calculated.

To explore the inter-hospital variation of tolvaptan use, we evaluated the prevalence of early and late tolvaptan use across the participating centres. We also assessed the correlation between the prevalence of tolvaptan use and the ratio of early to late tolvaptan use in each participating centre using Spearman's rank correlation coefficient weighting the number of cases at the participating centres.

All statistical analyses were conducted by a physician (R. N.) and a statistician (T. M.) using JMP 16.0 (SAS Institute Inc., Cary, NC, USA) and EZR (Easy R). All the reported P values were two-tailed, and P values < 0.05 were considered statistically significant.

Characteristics of the study population with and without tolvaptan use

Of the 3802 patients without tolvaptan use before the index hospitalization included in this current study, 773 patients (20%) used tolvaptan (Figure 1). Characteristics of the two groups with and without tolvaptan use are shown in Table 1. Patients who used tolvaptan as compared with those who did not use tolvaptan were older and had higher B-type natriuretic peptide or N-terminal pro-brain-type natriuretic peptide and serum creatinine levels. They also had a higher prevalence of valvular heart disease aetiology, diabetes, chronic kidney disease (CKD), cognitive dysfunction, symptoms of congestion like orthopnoea, dyspnoea on exertion, oedema, jugular vein distension, pleural effusion, NYHA class III or IV, and moderate to severe TR and mitral regurgitation, while they had a lower prevalence of ambulatory status, poor medical adherence, employed status, and public assistance and had a lower heart rate, systolic blood pressure, serum albumin, sodium, and haemoglobin level.

Patients who used tolvaptan as compared with those who did not use tolvaptan were more frequently treated with MRAs and loop diuretics at admission but were less frequently treated with angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker and beta-blockers at discharge. Patients who used tolvaptan were more frequently treated with dobutamine within 24 h after hospital presentation and were more frequently treated with furosemide, carperitide, nitrates, and inotropes beyond 24 h after hospital presentation than those who did not use tolvaptan (Table 1).

Factors associated with tolvaptan use during hospitalization

By the multivariable logistic regression analysis, the factors independently associated with tolvaptan use were diabetes (OR = 1.36, 95% CI = 1.05–1.76, P = 0.02), poor medical adherence (OR = 0.66, 95% CI = 0.46–0.95, P = 0.02), oedema (OR = 1.54, 95% CI = 1.18–2.01, P = 0.002), pleural effusion (OR = 1.53, 95% CI = 1.18–1.98, P = 0.001), hyponatraemia (sodium level < 135 mEq/L) (OR = 1.51, 95% CI = 1.07–2.13, P = 0.02), eGFR < 30 mL/min/1.73 m² (OR = 1.66, 95% CI = 1.27–2.16, P = 0.0002), moderate or severe TR (OR = 1.56, 95% CI = 1.20–2.02, P = 0.0008), infusion therapy of dobutamine within 24 h (OR = 1.71, 95% CI = 1.18–2.47, P = 0.005), and infusion therapy of additional inotropes 24 h after admission (OR = 2.49, 95% CI = 1.58–3.94, P < 0.0001) (Table 2).

Clinical outcomes after 90 days of follow-up from the index hospitalization and discharge status: the tolvaptan use vs. no tolvaptan use groups

The incidences of all-cause death and cardiovascular death at 90 days of follow-up from the index hospitalization were significantly higher in the tolvaptan use group than in the no tolvaptan use group (14.3% vs. 8.6%, P = 0.049 and 10.0% vs. 6.2%, P = 0.03, respectively). After adjusting for confounders incorporating tolvaptan use as a time-updated covariate accounting for the time to start tolvaptan, the excess risk of tolvaptan use relative to no tolvaptan use was no longer significant for all-cause death, cardiovascular death, and non-cardiovascular death at 90 days of follow-up from the index hospitalization (adjusted HR = 1.53, 95% CI = 0.77–3.02, P = 0.22; adjusted HR = 2.00, 95% CI = 0.85–4.72, P = 0.11; adjusted HR = 0.82, 95% CI = 0.24–2.74, P = 0.74) (Table 3). Patients who used tolvaptan had a longer hospital stay [median 22 (15–34) days vs. 15 (11–21) days, P < 0.0001] and a higher prevalence of WRF (47.0% vs. 31.8%, P < 0.0001) and WHF (24.8% vs. 14.4%, P < 0.0001). Also, they more often had high-dose loop diuretic, dyspnoea on exertion, and oedema at discharge (Supporting Information, Table S1).

Early vs. late use of tolvaptan

Of the 773 patients who used tolvaptan, 378 patients used tolvaptan early and 395 patients used tolvaptan late. The median interval from admission to the day of starting tolvaptan was 2 (IQR: 0–6) days. The histogram for the days of starting tolvaptan was shown in Supporting Information, Figure S1. Characteristics of the early and late use of tolvaptan groups were shown in Supporting Information, Table S2. Patients with late tolvaptan use had higher levels of Troponin T, serum sodium, and haemoglobin at presentation than patients with early tolvaptan use. Patients with late tolvaptan use were more frequently treated with carperitide and noradrenaline and less frequently treated with nicardipine within 24 h after hospital presentation and were more frequently treated with furosemide and inotropes beyond 24 h after hospital presentation than patients with early tolvaptan use (Supporting Information, Table S2).

Tolvaptan use across the participating centres

The prevalence of tolvaptan use varied from 48.7% to 0% across the participating centres (Figure 2). There was a significant positive correlation between the prevalence of tolvaptan use and the ratio of early to late tolvaptan use in each participating centre (R² = 0.256, P = 0.03) (Figure 3).