Article Text
PDF
Abstract
Objective Pulmonary congestion is the main cause of hospital admission in patients with heart failure (HF). Lung ultrasound (LUS) is a useful tool to identify subclinical pulmonary congestion. We evaluated the usefulness of LUS in addition to physical examination (PE) in the management of outpatients with HF.
Methods In this randomised multicentre unblinded study, patients with chronic HF and optimised medical therapy were randomised in two groups: ‘PE+LUS’ group undergoing PE and LUS and ‘PE only’ group. Diuretic therapy was modified according to LUS findings and PE, respectively. The primary endpoint was the reduction in hospitalisation rate for acute decompensated heart failure (ADHF) at 90-day follow-up. Secondary endpoints were reduction in NT-proBNP, quality-of-life test (QLT) and cardiac mortality at 90-day follow-up.
Results A total of 244 patients with chronic HF and optimised medical therapy were enrolled and randomised in ‘PE+LUS’ group undergoing PE and LUS, and in ‘PE only’ group. Thirty-seven primary outcome events occurred. The hospitalisation for ADHF at 90 day was significantly reduced in ‘PE+LUS’ group (9.4% vs 21.4% in ‘PE only’ group; relative risk=0.44; 95% CI 0.23 to 0.84; p=0.01), with a reduction of risk for hospitalisation for ADHF by 56% (p=0.01) and a number needed to treat of 8.4 patients (95% CI 4.8 to 34.3). At day 90, NT-proBNP and QLT score were significantly reduced in ‘PE+LUS’ group, whereas in ‘PE only’ group both were increased. There were no differences in mortality between the two groups.
Conclusions LUS-guided management reduces hospitalisation for ADHF at mid-term follow-up in outpatients with chronic HF.
- heart failure
- heart failure with reduced ejection fraction
Statistics from Altmetric.com
Introduction
Despite significant improvements in therapies witnessed in the last decades, the prevalence of heart failure (HF) as well as hospitalisation rate for acute decompensated HF (ADHF) continue to raise.1 Pulmonary congestion (PC) is the main cause of hospital admission.2 Clinical signs of HF occur late in the decompensation phase, even though increase in body weight3 and intrathoracic fluid4 can be detected at least 10 days preceding hospitalisation.
Medical history and physical examination (PE) are a mainstay for the assessment of HF. However, lung auscultation shows poor sensitivity and accuracy in detecting mild PC.5 The availability of an additional tool to identify subclinical PC is attractive, since the ensuing treatment may prevent hospitalisation.
Lung ultrasound (LUS) provides a semiquantitative assessment of PC, identifying extravascular lung water (EVLW) as B-lines.6–10 The latter are significantly related with established parameters of decompensation in outpatients with chronic HF11 and portend worse outcome in terms of hospitalisation for ADHF and death.12
We investigated the clinical impact of LUS-guided management in outpatients with chronic HF.
Methods
Study design
A randomised multicentre unblinded study was designed. Patients were considered eligible if on optimal medical therapy for HF for at least 2 months and with left ventricular ejection fraction <45%. At the beginning of cardiology outpatient visits at HF clinic, patients who satisfied enrolment criteria were randomised in two parallel groups (1:1): ‘PE+LUS’ group undergoing PE and LUS and ‘PE only’ group. The study was approved by local ethics committees at each centre. All patients provided written informed consent.
PE was performed by trained physicians according to a validated clinical assessment score.13 The quantification of EVLW was graded on the lung field height where B-lines, defined as echogenic wedge-shaped signal, were visualised9 (details in online supplementary appendix).
Supplemental material
In ‘PE+LUS’ group, baseline loop diuretic dose was modified according to physician’s judgement based on LUS score, considering the extent of B-line distribution (basal, middle, apical fields). There was no fixed scheme of dose modification, rather variable increments ranging from 25% to 150%, considering also vital signs, renal function and ongoing diuretic dose. In ‘PE only’ group, diuretic therapy was optimised according to PE, blood tests, echocardiogram and chest X-ray when available. The investigators were unblinded to group assignment and LUS findings. Medical history, PE, therapeutic changes, blood tests including N-terminal fragment of BNP pro-hormone (NT-proBNP) and left ventricular dysfunction (LVD-36) questionnaire quality-of-life test (QLT)14 were recorded at baseline and at 90-day follow-up. The trial is registered at ClinicalTrials.gov (NCT03262571). There was no patients and public involvement in the design of our research.
The primary endpoint was hospitalisation for ADHF at 90-day follow-up. Secondary endpoints were mortality, modifications in NT-proBNP values and QLT score at 90-day follow-up. ADHF failure was defined as sudden or gradual onset of signs or symptoms of HF.15
Statistics
We estimated that the 90-day rate of primary endpoint would be 20%, hence the need to follow 440 patients for 3 months to provide the study with a power of 80% to detect a relative reduction of 50% in risk of hospitalisation for ADHF in LUS group, at an overall two-sided alpha level of 0.05. At the third interim analysis, scheduled at 244 enrolled patients, principal investigators verified that prespecified stopping boundary for overwhelming benefit had been crossed and decided to stop the study.
Data are expressed as mean±SD, median (IQR) or percentage when appropriate. For group comparisons, paired and unpaired t-test, χ2 test or Fisher’s exact test and Wilcoxon matched-pairs signed rank test were used as appropriate. The relative risk (RR), number needed to treat (NNT) and 95% CI were defined. Survival free from ADHF events was estimated with Kaplan-Meier method and compared by long-rank test. According to the interim nature of analysis, p<0.027 was considered statistically significant to reject the null hypothesis (details in online supplementary appendix).
Results
From January 2011 to November 2016, 256 patients were screened and 244 enrolled: 127 (52%) in ‘PE+LUS’ group and 117 (48%) in ‘PE only’ group (figure 1). Baseline characteristics are shown in tables 1 and 2. All participants completed follow-up.
Enrolment of participants in the study. LUS, lung ultrasound; PE, physical examination.
Baseline characteristics of the overall study population, group ‘PE+LUS’ and group ‘PE only’
Medical therapy and renal function of the study population
In the ‘PE+LUS’ group, 26.8% of patients presented no B-lines, 25.2% B-lines only in basal fields, 33.9% B-lines extended to middle fields and 14.2% B-lines extended to apical fields. In none of the patients pleural effusion was found.
In ‘PE+LUS’ group, 60 patients (47%) were both LUS and PE positive, 33 patients (26%) were LUS positive and PE negative, 34 (27%) patients resulted LUS-negative. Therefore, 93 (73%) LUS-positive patients were addressed with an increase in diuretic therapy (mean furosemide uptitration 50.7±57.2 mg), whereas in the remaining 34 (27%) ongoing therapy was confirmed.
At 90-day follow-up, 56% of ‘PE+LUS’ patients showed a significant reduction in B-line extension (0.5 (IQR 0–1) vs 1 (IQR 0.75–2), p<0.001).
At 90-day follow-up, hospitalisation for ADHF occurred in 12 patients (9.4%) in ‘PE+LUS’ group and in 25 patients (21.4%) in ‘PE only’ group (relative risk (RR)=0.44; 95% CI 0.23 to 0.84; p=0.01). As compared with ‘PE only’ group, the risk was reduced by 56% in ‘PE+LUS’ group (p=0.01) with an NNT of 8.4 patients (95% CI 4.8 to 34.3).
The survival-free rate from hospitalisation for ADHF was 84.5% (95% CI 81.5 to 87.5) for ‘PE+LUS’ group and 79.7% (95% CI 75.6 to 83.7) for ‘PE only’ group (log-rank 6.5, p=0.01) (figure 2).
Kaplan-Meier curves for hospitalisation-free survival from ADHF in group ‘PE+LUS’ versus group ‘PE only’ patients. ADHF, acute decompensated heart failure; LUS, lung ultrasound; PE, physical examination.
At 90-day follow-up in ‘PE+LUS’ group, the NT-proBNP value was significantly reduced (p=0.026) and QLT score significantly decreased (p=0.001) indicating an improvement in health status, while in ‘PE only’ group, NT-proBNP value (p=0.004) and QLT score (p<0.001) were significantly increased (figure 3A,B).
NT-proBNP and QLT values at baseline and at 90-day follow-up in ‘PE+LUS’ and ‘PE only’ groups. LUS, lung ultrasound; PE, physical examination; QLT, quality-of-life test.
No differences in mortality rate were observed between the groups at 90-day follow-up (table 3).
ADHF and death according to ‘PE+LUS’ group and ‘PE only’ group in the entire study population and subgroups PE positive and PE negative
A post hoc subgroup analysis, further dividing the overall population in negative and positive PE patients was performed. Subgroups were not defined by characteristics known before randomisation, and thus subgroup analyses are a non-randomised unadjusted comparison with potential biases.
In patients with negative PE (n=152, 66 from ‘PE+LUS’ group), primary outcome occurred in 3 (4.5%) patients in ‘PE+LUS’ group and in 19 (22.1%) of the ‘PE only’ group (RR=0.20; 95% CI 0.06 to 0.67; p=0.002). Hence, ‘PE+LUS’ patients with negative PE had an 80% risk reduction for hospitalisation for ADHF as compared with ‘PE only’ group (p=0.002) (table 3). At 90-day follow-up in ‘PE+LUS’ group, the NT-proBNP value was significantly reduced (p=0.01) and QLT score significantly decreased (p=0.02), while in ‘PE only’ NT-proBNP value (p<0.001) and QLT score (p<0.001) were significantly increased (figure 4A,B).
NT-proBNP and QLT values at baseline and at 90-day follow-up in the subgroup of patients with negative PE. LUS, lung ultrasound; PE, physical examination; QLT, quality-of-life test.
In patients with a positive PE (n=92, 60 from ‘PE+LUS’ group) no significant differences in hospitalisation for ADHF between the two groups were observed (table 3). Neither significant differences in NT-proBNP at 3 months were observed (figure 5A), while QLT scores were significantly decreased in ‘PE+LUS’ group (p=0.01) and increased in ‘PE only’ group (p=0.02) at 3 months (figure 5B).
NT-proBNP and QLT values at baseline and at 90-day follow-up in the subgroup of patients with positive PE. LUS, lung ultrasound; PE, physical examination; QLT, quality-of-life test.
In these subgroups, no difference in mortality between ‘PE+LUS’ and ‘PE only’ at 90-day follow-up was observed (table 3).
We found no safety issues in the group PE-LUS with no significant difference in serum creatinine (p=0.5) and creatinine clearance (p=0.66), while a trend of worsening of renal function in ‘PE only’ group at 90-day follow-up was identified (further details in online supplementary appendix).
Discussion
The presence of PC at PE in patients with chronic HF portends a high risk of ADHF and death.16 LUS is an attractive tool, since it provides a fast and low-cost bedside examination to detect subclinical PC, overcoming the limited sensitivity and specificity of lung auscultation.14
Prior studies have extensively demonstrated the strong correlation between LUS and EVLW.17 Platz et al asserted that PC is frequently detected with LUS examination in ambulatory patients with chronic HF and associated with a worse prognosis in terms of hospitalisation for ADHF and death.12 Similar results have been suggested by Pellicori and colleagues, who demonstrated that as many as 58% of 342 patients with chronic HF outpatients, and 47% of subjects clinically free of congestion, were LUS-positive.18 A recent randomised clinical trial from Rivas-Lasarte et al showed that LUS-guided strategy significantly improved the combined endpoint of urgent visit, hospitalisation for ADHF and death at 6 month after an ADHF episode in 123 patients.19
Our study demonstrates that LUS improves the management of patients with chronic HF. Indeed, LUS accuracy in detecting subclinical PC provides the possibility to tackle an early decompensation phase, reducing admissions for ADHF and HF biomarkers and improving quality of life at 90-day follow-up. The usefulness of LUS was also confirmed by the better trend in renal function when therapeutic management was LUS guided.
The study has some limitations. First, the midterm follow-up could limit the prognostic yield in predicting outcome events. Second, the lack of blinding is a potential source of bias. Additionally, we did not use the stratified randomisation method to control the influence of covariates. However, the two groups of the study population were homogeneous in terms of baseline characteristics, minimising the possible influence of covariates.
Conclusions
LUS is a low-cost, radiation-free and rapid diagnostics. The detection of EVLW by LUS in patients with chronic HF, especially in subjects with negative PE, might help physician to optimise medical treatment, preventing hospitalisation for ADHF.
Key messages
What is already known on this subject?
Lung ultrasound provides a validated semiquantitative assessment of pulmonary congestion, but it is not known the prognostic impact on the management of patients with chronic heart failure.
What might this study add?
This study shows that the detection of congestion with lung ultrasound during the ambulatory evaluation of patients with chronic heart failure, especially in patients with no overt signs of heart failure detected by physical examination, leads to a reduction of hospitalisation for acute decompensated heart failure with respect to the sole physical examination at midterm follow-up (9.4% vs 21.4% at 90-day follow-up, respectively, RR=0.44; 95% CI 0.23 to 0.84; p=0.01), reducing the risk for hospitalisation by 56%.
How might this impact on clinical practice?
Lung ultrasound is a low-cost, radiation-free and no time-consuming method with steep learning curve. The integration of lung ultrasound in heart failure clinics might help to improve the treatment of heart failure targeting early signs of decompensation.
References
Footnotes
Contributors EA, GF, AM: study planning, study analysis and responsible for the overall content as guarantors. CM, GF, LI, HS, VT, GI, SS, FA, FL, PI, MFC, LS, SG, MA, AM, EA conducted the study and performed the examinations. EA, CM, LI performed the statistical analysis and wrote the manuscript. All authors have read and approved the final version of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article and uploaded as supplementary information, and further data are available on request.