RESUMEN
AIMS: Hypertonic saline solution (HSS) plus intravenous (IV) loop diuretic appears to enhance the diuretic response in patients hospitalized for heart failure (HF). The efficacy and safety of this therapy in the ambulatory setting have not been evaluated. We aimed to describe the design and baseline characteristics of the SALT-HF trial participants. METHODS AND RESULTS: 'Efficacy of Saline Hypertonic Therapy in Ambulatory Patients with HF' (SALT-HF) trial was a multicenter, double-blinded, and randomized study involving ambulatory patients who experienced worsening heart failure (WHF) without criteria for hospitalization. Enrolled patients had to present at least two signs of volume overload, use ≥ 80 mg of oral furosemide daily, and have elevated natriuretic peptides. Patients were randomized 1:1 to treatment with a 1-h infusion of IV furosemide plus HSS (2.6-3.4% NaCl depending on plasmatic sodium levels) versus a 1-h infusion of IV furosemide at the same dose (125-250 mg, depending on basal loop diuretic dose). Clinical, laboratory, and imaging parameters were collected at baseline and after 7 days, and a telephone visit was planned after 30 days. The primary endpoint was 3-h diuresis after treatment started. Secondary endpoints included (a) 7-day changes in congestion data, (b) 7-day changes in kidney function and electrolytes, (c) 30-day clinical events (need of IV diuretic, HF hospitalization, cardiovascular mortality, all-cause mortality or HF-hospitalization). RESULTS: A total of 167 participants [median age, 81 years; interquartile range (IQR), 73-87, 30.5% females] were randomized across 13 sites between December 2020 and March 2023. Half of the participants (n = 82) had an ejection fraction >50%. Most patients showed a high burden of comorbidities, with a median Charlson index of 3 (IQR: 2-4). Common co-morbidities included diabetes mellitus (41%, n = 69), atrial fibrillation (80%, n = 134), and chronic kidney disease (64%, n = 107). Patients exhibited a poor functional NYHA class (69% presenting NYHA III) and several signs of congestion. The mean composite congestion score was 4.3 (standard deviation: 1.7). Ninety per cent of the patients (n = 151) presented oedema and jugular engorgement, and 71% (n = 118) showed lung B lines assessed by ultrasound. Median inferior vena cava diameter was 23 mm, (IQR: 21-25), and plasmatic levels of N-terminal-pro-B-type natriuretic peptide (NTproBNP) and antigen carbohydrate 125 (CA125) were increased (median NT-proBNP 4969 pg/mL, IQR: 2508-9328; median CA125 46 U/L, IQR: 20-114). CONCLUSIONS: SALT-HF trial randomized 167 ambulatory patients with WHF and will determine whether an infusion of hypertonic saline therapy plus furosemide increases diuresis and improves decongestion compared to equivalent furosemide administration alone.
RESUMEN
Introducción y objetivos HeartLogic es un algoritmo multiparamétrico incorporado a desfibriladores automáticos implantables (DAI). La alerta asociada predice descompensaciones de insuficiencia cardiaca (IC). Nuestro objetivo es analizar la asociación entre alertas y eventos clínicos bajo un protocolo de seguimiento común en un registro multicéntrico. Métodos Se evaluaron la fase 1 (investigadores ciegos al estado de la alerta) y las fases 2 y 3 (tras la activación de HeartLogic, según práctica local y un protocolo común respectivamente). Resultados Se incluyó a 288 pacientes en 15 centros. En fase 1, tras una media de observación de 10 meses, hubo 73 alertas (0,72 alertas/paciente-año), con 8 hospitalizaciones y 2 visitas a urgencias por IC (0,10 eventos/año-paciente). No hubo hospitalizaciones fuera del periodo de alerta. Las fases activas tuvieron una media de seguimiento de 16 (IC95%, 15-22) meses, con 277 alertas (0,89 alertas/año-paciente); 33 se asociaron con hospitalizaciones o muerte por IC, 46 con descompensaciones menores y 78 con otros eventos. La tasa de alertas inexplicables fue 0,39/año-paciente. Fuera del estado de alerta solo hubo una hospitalización y una descompensación menor. La mayoría de las alertas (el 82% en fase 2 y el 81% en fase 3; p=0,861) se gestionaron a distancia. La mediana de NT-proBNP fue superior en estado de alerta que fuera de él (7.378 frente a 1.210 pg/ml; p <0,001). Conclusiones El índice HeartLogic se asoció con descompensaciones de IC y otros eventos relevantes, con baja tasa de alertas inexplicables. Un protocolo estandarizado permitió detectar y actuar a distancia con seguridad sobre las alertas (AU)
Introduction and objectives HeartLogic is a multiparametric algorithm incorporated into implantable cardioverter-defibrillators (ICD). The associated alerts predict impending heart failure (HF) decompensations. Our objective was to analyze the association between alerts and clinical events and to describe the implementation of a protocol for remote management in a multicenter registry. Methods We evaluated study phase 1 (the investigators were blinded to the alert state) and phases 2 and 3 (after HeartLogic activation, managed as per local practice and with a standardized protocol, respectively). Results We included 288 patients from 15 centers. In phase 1, the median observation period was 10 months and there were 73 alerts (0.72 alerts/patient-y), with 8 hospitalizations and 2 emergency room admissions for HF (0.10 events/patient-y). There were no HF hospitalizations outside the alert period. In the active phases, the median follow-up was 16 (95%CI, 15-22) months and there were 277 alerts (0.89 alerts/patient-y); 33 were associated with HF hospitalizations or HF death (n=6), 46 with minor decompensations, and 78 with other events. The unexplained alert rate was 0.39 alerts/patient-y. Outside the alert state, there was only 1 HF hospitalization and 1 minor HF decompensation. Most alerts (82% in phase 2 and 81% in phase 3; P=.861) were remotely managed. The median NT-proBNP value was higher within than outside the alert state (7378 vs 1210 pg/mL; P <.001). Conclusions The HeartLogic index was frequently associated with HF-related events and other clinically relevant situations, with a low rate of unexplained events. A standardized protocol allowed alerts to be safely and remotely detected and appropriate action to be taken on them (AU)
