Author + information
- Received January 30, 2017
- Revision received September 15, 2017
- Accepted September 26, 2017
- Published online March 26, 2018.
- Nicolas Girerd, MD, PhDa,
- Marie-France Seronde, MD, PhDb,
- Stefano Coiro, MDc,
- Tahar Chouihed, MDa,d,
- Pascal Bilbault, MD, PhDe,
- François Braun, MDf,
- David Kenizou, MDg,
- Bruno Maillier, MDh,
- Pierre Nazeyrollas, MD, PhDi,
- Gérard Roul, MD, PhDj,
- Ludivine Fillieux, PharmDk,
- William T. Abraham, MDl,
- James Januzzi Jr., MDm,
- Laurent Sebbag, MD, PhDn,
- Faiez Zannad, MD, PhDa,
- Alexandre Mebazaa, MD, PhDo,
- Patrick Rossignol, MD, PhDa,∗ (, )
- on behalf of INI-CRCT, Great Network, and the EF-HF Group
- aINSERM, Centre d’Investigations Cliniques 1433, Université de Lorraine, CHU de Nancy, Institut Lorrain du Coeur et des Vaisseaux, Nancy, France, Inserm 1116 and INI-CRCT (Cardiovascular and Renal Clinical Trialists) F-CRIN Network, Nancy, France
- bService de cardiologie CHU de Besançon, EA 3920, Unité INSERM 942 CHU Lariboisière, Paris, France
- cDivision of Cardiology, University of Perugia, School of Medicine, Perugia, Italy
- dEmergency Department, CHU de Nancy, France
- eEmergency Department, Hôpitaux Universitaires de Strasbourg, Strasbourg, France and EA 7293 Stress vasculaire, Fédération de Médecine Translationnelle de Strasbourg, Strasbourg, France
- fStructures de Médecine d'Urgence, Centre Hospitalier Régional, Hôpital de Mercy, Metz, France
- gService de cardiologie, Hôpital Emile Muller, Mulhouse, France
- hService de cardiologie, Centre hospitalier de Troyes, Anatole, France
- iCentre de Recherche et d’Investigation Clinique, Service de Cardiologie, CHU de Reims, Reims, France
- jPôle d'activité médico-chirurgicale cardiovasculaire Nouvel Hôpital Civil, Strasbourg, France and Unité d’insuffisance cardiaque, Centre de compétence des cardiomyopathies
- kNovartis Pharma SAS, Rueil-Malmaison, France
- lDavis Heart and Lung Research Institute, Ohio State University, Columbus, Ohio
- mDivision of Cardiology, Massachusetts General Hospital, Boston, Massachusetts
- nHospices Civils de Lyon, Hôpital Louis Pradel, Pôle Médico-Chirurgical de Transplantation Cardiaque Adulte, Bron, France
- oDepartment of Anesthesiology, Critical Care and Burn Unit, St. Louis Hospital, University Paris, UMR-S942, INSERM and INI-CRCT (Cardiovascular and Renal Clinical Trialists) F-CRIN Network Nancy, GREAT Network, Paris, France
- ↵∗Address for correspondence:
Prof. Patrick Rossignol, Centre d’Investigations Cliniques-INSERM CHU de Nancy, Institut lorrain du Cœur et des Vaisseaux Louis Mathieu, 4 rue du Morvan, 54500 Vandoeuvre Lès Nancy, France.
Congestion is one of the main predictors of poor patient outcome in patients with heart failure. However, congestion is difficult to assess, especially when symptoms are mild. Although numerous clinical scores, imaging tools, and biological tests are available to assist physicians in ascertaining and quantifying congestion, not all are appropriate for use in all stages of patient management. In recent years, multidisciplinary management in the community has become increasingly important to prevent heart failure hospitalizations. Electronic alert systems and communication platforms are emerging that could be used to facilitate patient home monitoring that identifies congestion from heart failure decompensation at an earlier stage. This paper describes the role of congestion detection methods at key stages of patient care: pre-admission, admission to the emergency department, in-hospital management, and lastly, discharge and continued monitoring in the community. The multidisciplinary working group, which consisted of cardiologists, emergency physicians, and a nephrologist with both clinical and research backgrounds, reviewed the current literature regarding the various scores, tools, and tests to detect and quantify congestion. This paper describes the role of each tool at key stages of patient care and discusses the advantages of telemedicine as a means of providing true integrated patient care.
Patients with heart failure (HF) often develop congestion that may require urgent hospitalization, especially if pulmonary congestion (PC) is present. Development of congestion leading to HF decompensation is a powerful predictor of poor patient outcome (1–5). Therefore, it may be important to better detect and monitor congestion before it leads to decompensation. However, congestion can be difficult to assess, especially when extrapulmonary signs of congestion are mild, such as in the setting of acute PC due to hypertension or in patients nearing discharge from a HF hospitalization.
Increased intracardiac filling pressures often silently precede the appearance of congestive symptoms by days or weeks (1). Increasing filling pressures are often subtle and difficult to detect, and can be masked by other comorbidities (e.g., infections). Current European Society of Cardiology guidelines recommend treating signs and symptoms of congestion so that patients achieve near-optimal volume status (6,7). Unfortunately, 50% of patients admitted for acute heart failure (AHF) are discharged with residual congestion (2), possibly due to an absence of a clear congestion evaluation strategy. Such residual congestion at discharge is associated with rehospitalization and death within 6 months after discharge, independent of the underlying pathology (2). Importantly, although current guidelines emphasize the importance of aggressively treating congestion, they do not stipulate which congestion targets should be aimed at discharge for AHF hospitalization or in an ambulatory setting.
At each stage of the journey of a patient with HF, specific evaluation tools are used to qualify and quantify congestion to support treatment decisions. However, not all of these tools are appropriate for use at each point of clinical care within this journey (i.e., clinical settings where a patient may be managed).
The objective of this position paper is to outline the 4 types of tools for evaluating and quantifying congestion, namely: 1) clinical tools and scores; 2) biological biomarkers; 3) imaging; and 4) pressure (hemodynamics) and impedance-based tools. The role of each tool is described during key stages of patient care and according to the main points of clinical care.
Clinical Tools for Evaluating Congestion
Signs and symptoms of congestion
Physical assessment can only detect a moderate to high level of congestion. Although many clinical signs and symptoms of congestion have been well characterized and are recognized by published guidelines (8), no single element from clinical history or physical examination can accurately detect the underlying hemodynamic changes that lead to congestion. Dyspnea, orthopnea, systemic edema, jugular venous pressure, and the third heart sound are all important clinical findings to identify decompensated HF. A review of clinical signs and symptoms is detailed in the Online Appendix. However, relying on only 1 single clinical finding to identify decompensated HF has a low sensitivity and poor predictive value (9), whereas the reproducibility of the medical interview, especially to assess dyspnea, is poor (10).
Clinical congestion scores
Clinical scores that combine several clinical indicators have been shown to assess the level of congestion more accurately than any standalone indicator (9). Yet, they are more often used as prognostic rather than diagnostic tools. The Stevenson classification, Lucas Score, and Rhode Score are detailed in the Online Appendix and are reported in Table 1. The EVEREST score, which was developed by Ambrosy et al. in 2013 (2) within the EVEREST (Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with Tolvaptan) trial (Table 1), captures changes in congestion within a hospitalization time frame and is associated with a markedly increased risk of HF mortality in the 15% of patients with overt clinical congestion (2). The Gheorghiade score (4) has been promoted as an integrative congestion score, integrating bedside clinical parameters, dynamic parameters, and natriuretic peptides (NPs) (Table 1).
The respective role of these congestion scores in routine clinical practice still remains to be determined. However, the EVEREST score is the most evidence-based in the current era of AHF management, and it appears to be the best candidate for routine use.
Circulating Biomarkers for Evaluating Congestion
NPs, such as B-type natriuretic peptide (BNP), N-terminal fragment pro-B-type natriuretic peptide (NT-proBNP), and atrial NP, are the most studied circulating biomarkers in HF. The characteristics of NPs are listed in Online Table 1.
Current clinical practice guidelines recommend BNP or NT-proBNP for assisting with the diagnosis of AHF (7,11). Biomarkers are also useful for determining prognosis after hospitalization, as well as in the chronic phase of HF. NT-proBNP is likely to become the standard because of its better risk-stratifying properties in the setting of AHF (12) and its straightforward interpretation in patients treated with sacubitril/valsartan.
Plasma volume, hemoconcentration
Several routinely assessed biological parameters, such as serum protein, albumin, hemoglobin, and hematocrit have been proposed as surrogate markers of (de)congestion and have been found to be associated with cardiovascular endpoints (13).
Several formulas have also been developed to indirectly estimate plasma volume, using hemoglobin and/or hematocrit, which may therefore be valuable for monitoring both acute and chronic (de)congestion (14). One simple formula to assess estimated plasma volume variations is the Strauss formula:
An instantaneous version of this formula (Duarte’s formula ) enables calculating plasma volume without previous hemoglobin and hematocrit data:
Evaluating renal function is part of routine care in AHF and has links with evaluating and managing congestion. Worsening renal function can alternately be a marker of congestion or it may imply adequate decongestion; increased venous congestion has been elegantly shown to be the most important hemodynamic factor driving worsening renal function in decompensated patients (5). However, other causes of renal dysfunction (dehydration following an intensive decongestion, nephroangiosclerosis, and drug adverse effects) are frequently observed in patients with HF. The blood urea nitrogen to creatinine ratio is a useful variable to better identify patients with renal dysfunction due to congestion (15).
Liver function markers
Cholestatic liver injury due to congestion is primarily observed when central venous pressure is particularly high. The resulting right atrial pressure transmitted directly to the hepatic veins impairs hepatocyte function (16). Therefore, bilirubin and gamma-glutamyl transpeptidase have been suggested as possible biomarkers for congestion in the appropriate clinical setting.
Imaging Tools for Evaluating Congestion
Radiological findings reflect the anatomical and pathological alterations induced by PC. Traditionally, a chest x-ray is the first-line procedure. X-ray signs (including peri-bronchial cuffing, cardiomegaly, pulmonary venous congestion, or pleural effusion) are reasonably specific but poorly sensitive; negative chest x-ray findings have been reported in a substantial percentage of patients with AHF (17).
Echocardiography is currently the gold standard non invasive tool for detecting and monitoring HF. It is most useful in the assessment of HF etiology and in classifying HF according to the left ventricular (LV) ejection fraction (EF).
Comprehensive echocardiography may serve as a baseline for serial evaluations in patients whose clinical status may change over time as disease progresses or it may be used to follow therapeutic interventions (18). However, it involves a lengthy procedure that is typically not performed at bedside other than in intensive care units. Logistically, it is impractical to repeatedly perform these examinations; thus, comprehensive echocardiography is not usually performed at discharge if already performed during hospitalization. Furthermore, the algorithm used to determine elevated filling pressures can only be performed by an experienced cardiologist skilled in echocardiography. Despite these drawbacks, echocardiography remains the gold standard for evaluating blood volume and LV filling pressures before discharge (18). It is superior to clinical scores alone for predicting readmission over short- to medium-term follow-up (19).
In the emergency department or during clinical points of care, simple echocardiography may be performed relatively easily with portable, pocket-sized ultrasound devices. These devices are capable of measuring important variables such as inferior vena cava diameter, ejection fraction, and the right atrial volume index.
Lung ultrasound (LUS) is highly useful in comprehensive congestion evaluation at the bedside. LUS produces comet-like images, and the number of “comets” viewed is proportional to the severity of congestion (Figure 1). LUS effectively measures the amount of intrapulmonary fluid, indicating PC, which is associated with the level of pulmonary arterial pressure (20). LUS provides useful information for the prognostic stratification of patients admitted to the emergency department with dyspnea and/or chest pain syndrome (21). In patients hospitalized for AHF, residual PC as assessed by LUS is a strong predictor of post-discharge outcome (3). Minimum training is required; reproducible results have been obtained after 30 min to 1 h of training (22). This technique can rapidly detect changes in congestion over a few hours using freely available and easy to use hand-held ultrasound devices (23).
Pressure and Impedance-Based Tools
In some difficult clinical settings, right heart catheterization (details described in the Online Appendix) remains the gold standard assessment to evaluate both right ventricular and LV filling pressures and hemodynamic congestion.
Importantly, and in contrast to the ESCAPE (Evaluation study of congestive heart failure and pulmonary artery catheterization effectiveness) results, the CHAMPION (CardioMEMS Heart Sensor Allows Monitoring of Pressure to Improve Outcomes in NYHA Class III Heart Failure Patients) trial demonstrated that tailoring HF management to achieve protocol-defined pulmonary artery pressures according to a long-term implanted wireless pulmonary artery hemodynamic monitoring system was associated with fewer HF hospitalizations in patients with chronic HF (1). In CHAMPION, a significant reduction in HF hospitalizations was observed in patients with HF reduced EF and patients with HF preserved EF, making pressure-guided HF therapy an option for HF patients, regardless of LVEF (24). The use of these wireless devices may change the management of AHF and chronic HF, especially the management of congestion in the near future.
Bioimpedance vector analysis is a promising noninvasive technique (Online Appendix). Of note, the IMPEDANCE-HF (Outpatient Lung Impedance-Guided Preventive Therapy in Patients With Chronic Heart Failure) trial recently demonstrated that lung bioimpedance-guided management could reduce hospitalization rates (25).
Applying the Right Tool for Each Stage of Patient Management
Based on the preceding literature review, the authors recommend the following parameters for evaluating congestion at each step of patient management until appropriate discharge (Central Illustration, Online Table 2).
Pre-hospital and emergency department
Before hospital arrival, the objective is to accurately identify AHF and direct the patient to the appropriate service. The tools are mainly related to clinical assessment (including patient history, clinical evaluation of breathlessness, and physical examination to assess signs of HF) and may include LUS, BNP, or NT-proBNP measurement. Preliminary data regarding the use of pre-hospital LUS are promising (26). In the emergency department, most validated tools for assessing congestion are routinely available. In particular, our group recommends that LUS should be systematically performed, because it is easy to use, and results are rapid. We also recommend confirming the diagnosis with either BNP or NT-proBNP dosing as recommended by the current clinical guidelines, although the results can take several hours to produce. Because patients with AHF benefit from early initiation of therapeutic interventions (27), congestion should be treated as early as possible based on the LUS results. Therapy can be adjusted later according to the BNP or NT-proBNP results. A chest x-ray should also be systematically performed in the emergency department.
Hospitalization course through discharge
The objective of care in the hospitalized setting is to monitor changes in congestion level and overall condition following treatment in patients admitted for congestive symptoms. Because of the risk of rehospitalization observed in these patients, which is attributable to residual congestion at discharge, we recommend using a repeated multiparameter testing strategy to achieve the best possible decongestion. This would involve clinical examinations (EVEREST score), quantitative dyspnea evaluation (Likert, visual analogic score), biological biomarkers (BNP/NT-proBNP concentrations, plasma volume estimation based on hematologic data, and liver and kidney biomarkers in appropriate patients), and ultrasounds (both echocardiography and LUS).
Although clinical evaluation is typically performed repeatedly within routine care, only refined clinical scoring such as the EVEREST score can identify rapid congestion changes, except in patients in whom the baseline congestion level is mild. However, this clinical score is time-consuming and requires clinical expertise; it is rarely performed with routine care.
NPs could be used to guide decongestion therapy. However, the half-life of NT-proBNP renders it unsuitable to assess rapid congestion changes, and the cost of both BNP and NT-proBNP measurements makes daily quantification uneconomical and not recommended. However, a lack of reduction of at least 30% in the pre-discharge value for either peptide or a discharge BNP <250 pg/ml reveals the lack of significant decongestion and of a high risk for repeat hospitalization and/or death (28). The Strauss formula, based on hemoglobin and hematocrit levels, is a useful, inexpensive, and simple method to assess daily hemoconcentrations. Nonetheless, it is scarcely used despite these practical advantages (29).
Echocardiography can be repeatedly performed during a hospitalization for AHF. However, additional echographic examinations would be impractical to perform. In contrast, LUS and inferior vena cava echo can be performed at the bedside with a hand-held device in several minutes. This simple heart and lung ultrasound captures rapid changes in congestion and provides valuable information to guide choices for decongestion therapy (23,30), and may represent the extension of clinical examination in patients with AHF.
Other multimodal evaluations could be envisioned, including clinical evaluations, with special focus on the determination of dry weight and BNP and/or NT-proBNP quantification at discharge. Measuring in-hospital changes in estimated plasma volume and comprehensive but simple cardiac and lung echographic data could become the tools of choice during hospitalization and could be included in this multimodal evaluation at discharge. LUS- and simplified echocardiography−based in-hospital management of AHF patients has been recently shown to decrease the risk of death or rehospitalization in a proof-of-concept study (30). However, this strategy, as acknowledged previously by our group, has yet to be validated in a randomized clinical trial.
Typically, multimodal congestion assessment could be most useful upon admission, during decongestion therapy, and upon discharge. Ideally, optimal decongestion should be achieved before patients are discharged, especially in wet and warm patients with overt signs of clinical congestion at admission, and in whom congestion is the key driver of hospitalization. A selected set of decongestion targets to achieve before discharge are proposed in Table 2.
Post-discharge and long-term management
Patients with AHF remain at high risk of death and rehospitalization in the months after discharge; as many as 35% of patients are readmitted for AHF within a month after the index AHF hospitalization discharge (31). With each rehospitalization, the risk of mortality rises. Accordingly, once discharged, a multidisciplinary disease management team focused on HF, which includes a cardiologist, a general practitioner, and a HF nurse, should follow the patient to reduce the risk of rehospitalization (7). Because of the frequency of early rehospitalizations, the first post-discharge visit with general practitioners is advisable within 7 days after the initial discharge and with the hospital cardiology team within 2 weeks (7). During this phase, simplified clinical, biological and imaging resources should be used. Patient-reported clinical status and BNP and/or NT-proBNP or plasma volume, because of their scalable nature, would be particularly useful in this setting (Central Illustration, Online Table 2). In agreement with others (32), we also strongly recommend going beyond the clinical evaluation because subclinical congestion is frequent. Using biological biomarkers, implantable hemodynamic monitoring, and/or lung bioimpedance monitoring in selected patients could easily identify patients with subclinical residual congestion who might benefit greatly from intensive treatment optimization. Importantly, identifying congestion could trigger an increase in dosages of life-saving therapy (angiotensin-converting enzyme inhibitors, beta-blockers, mineralocorticoid receptor antagonists), diuretics, and/or nitrates. Higher diuretic doses needed to alleviate congestion might activate the renin-angiotensin-aldosterone system, which could contribute to the progression of HF. However, analyses adjusted for congestion variables usually result in a neutral association between diuretic dose and outcome (33), which suggests that a higher degree of congestion, rather than the diuretic dose itself, is harmful to HF patients.
The GUIDE-IT (Guiding Evidence Based Therapy Using Biomarker Intensified Treatment) trial, which also prioritized an increase of titration of neurohormonal antagonists over diuretics (i.e., it was not primarily focused on congestion relief) in patients with high levels of NPs, provided neutral results (34). The results of GUIDE-IT did not end the discussion regarding NT-proBNP-guided care, because the trial demonstrated no significant difference in HF therapies in the NT-proBNP-guided arm, whereas those in the usual care arm were seen far more frequently than usual (10 visits in 15 months). In total, both study arms had comparable reductions in NT-proBNP during follow-up, which explained the lack of differences in outcome. Future studies that focus on stricter adherence to NT-proBNP–mandated drug therapy adjustment, together with a more rational usual care approach are needed. Importantly, in the successful CHAMPION trial, hemodynamic congestion in the intervention group was treated with diuretics and/or nitrates (i.e., primarily focused on congestion relief), which eventually alleviated congestion and was translated into a lower risk of readmission, whereas diuretics and nitrates treatment were less frequently changed in the control group. In addition, treatment changes were unrelated to pulmonary pressure in the control group, which elicited a poor discrimination of clinical examination alone (32,35).
Integrating These Tools into Heart Failure Networks: The Role of Telemedicine
Although the validated clinical, radiological, and biological tools discussed in the first section of this review might improve patient care both during pre-hospital admission and in the hospital setting, implementing these tools into community-based disease management programs could also provide tremendous improvement in patient outcome.
Clinical signs and symptoms that can be easily monitored in the community setting include heart rate, blood pressure, weight, dyspnea, and edema. Biomarkers such as BNP and/or NT-proBNP and plasma volume status could also be obtained. Once collected, these data could then be incorporated into either existing or new electronic disease management platforms. The value of home-based fingerstick testing for BNP and/or NT-proBNP has been suggested, although more prospective studies are needed.
Telemedicine is likely to be the most useful and scalable solution to implement network management. However, several randomized trials have reported neutral association of remote monitoring with outcome, including the BEAT-HF (Better Effectiveness After Transition-Heart Failure) trial (36). Importantly, a Cochrane meta-analysis recently provided a comprehensive overview of the current literature and reported that “telehealth leads to similar health outcomes as face-to-face or telephone delivery of care” (37). Nonetheless, randomized evidence is still insufficient to formally recommend the systematic use of telemedicine in this setting.
Telemedicine may be a solution to reduce the gap between the increase of the older adult population living with complex, multimorbid conditions and the decreasing amount of available health services. Thus, in the future, patient-centered telemedicine could probably play a key role in HF patient care, particularly by avoiding (re)hospitalizations. Biomarkers (and possibly volume biomarkers) could be of great interest in this field, especially if point-of-care services are developed and used at home by the patients themselves (38). Likewise, pulmonary pressure data derived from wireless devices could be incorporated into these telemedicine modalities. A computer-based algorithm that incorporates the patient-reported signs and symptoms and point-of-care biomarker evaluation could greatly improve patient care through a telemedicine loop, which would trigger alarms and lead to optimization of treatment. This could be managed by trained nurses and physicians possibly via a call platform that is manned for 24 h/7 days by emergency callout services.
The economical viability of this approach remains to be evaluated. Focusing on high-risk populations would be key to achieve the highest possible treatment effect from this strategy. However, most of the cost of HF care is currently focused of the cost of hospitalizations; hence, a dedicated strategy to avoid HF hospitalization could consequently be highly cost-effective. Our group recently quantified the cost-effectiveness of a classical HF management program in a real-world, population-based setting and accordingly found that such a program could decrease health costs through a greater decrease in HF hospitalization despite the extra cost related to the program itself (39). The expected cost-efficiency of a telemedicine loop, which decreases human costs, is likely to be similar.
Gaps in Evidence and Unmet Needs
Congestion is strongly associated with HF prognosis, especially during and following AHF hospitalization. Detection, dynamic monitoring, and management of congestion could help improve HF management at all stages of the journey of the patient.
Congestion quantification using a standardized quantitative approach (e.g., the EVEREST clinical score and B-lines count) could represent a valuable management tool upon admission and during an AHF hospitalization. In our view, LUS is a strong candidate to test these congestion-guided treatments because of its strong association with outcome (3,23,30). One major aspect that will undoubtedly promote LUS expansion is the need for early treatment (as early as 1 h) after admission to the emergency department as emphasized in the current European Society of Cardiology guidelines for the management of AHF (6,27). The strong association of early treatment using intravenous loop diuretics with lower in-hospital mortality supports this recommendation (27). LUS could represent the key practice-changing tool that would enable physicians to provide the right treatment (i.e., vasodilators and/or diuretics) to the right patients (i.e., patients with AHF) as quickly as possible. In addition, the swift AHF diagnosis that can be achieved by LUS could allow patients to be triaged to the most appropriate facilities (cardiology ward or intensive care unit with cardiovascular focus) almost immediately after the first medical contact. During the hospital stay, daily LUS could also guide diuretic and/or vasodilator therapy and optimize the timing of discharge (30). The impact of LUS-guided AHF management at every step of the AHF hospitalization will ultimately need to be tested in specific randomized clinical trials.
At post-discharge and in ambulatory HF patients, aside from the randomized evidence from the CHAMPION trial (1), there is little evidence that HF management guided by standardized congestion assessment strategies is associated with better prognosis. Thus, pulmonary artery pressure–guided HF management should be considered in selected patients, similar to those studied in the CHAMPION trial. NPs provide information relative to HF severity as well as congestion, and BNP- and/or NT-proBNP–guided HF care (which in large part is dependent on treatment and prevention of congestion) seems to be appear promising (40). However, the GUIDE-IT trial, which was expected to provide definitive evidence of the impact of NP-guided therapy on clinical outcome in high-risk patients with HF, was terminated for futility and retrieved neutral results for all considered outcomes (34). Importantly, titration of neurohormonal antagonists over diuretics was prioritized in GUIDE-IT; thus, this study did not primarily target decongestion in patients with high NP levels. In addition, although there were more adjustments to therapy in the biomarker-guided group, neither doses of HF therapy nor the achieved NT-proBNP concentrations were significantly different between the treatment groups. Although this would likely decrease our confidence on the efficiency of management mainly or solely based on repeated NP measurements, it did not, however, solve the question regarding the use of NP measurements within a multimodality strategy or the impact of a strategy based primarily on decongestion (as in the CHAMPION trial) guided with NT-proBNP.
There is strong evidence to support the risk-stratifying properties of residual congestion at discharge for AHF hospitalizations. However, congestion assessed at the beginning (21) or during (19) hospitalization for AHF is also strongly associated with outcome. To strengthen the case of congestion as a therapeutic target, it would be useful to assess whether residual congestion is more, less, or equally prognostic than initial congestion. Such an analysis would support that residual congestion is not simply a marker for the severity of illness.
A Roadmap for the Clinical Validation of Congestion Variables as Actionable Biomarkers for HF Management
As already emphasized, established tools such as biomarker measurement and echocardiography, which are currently used to guide treatment decisions, did not require evidence of improvement in outcomes for their use to be approved (41). Concurrently, none of the congestion variables were sufficiently validated to be considered as a Class I, Level of Evidence: A, actionable biomarker to guide congestion treatment. However, NPs do have a class I recommendation for HF identification, mostly for their ability to detect left-sided congestion. Although NPs provide the strongest evidence of congestion variables, there is uncertainty regarding the best strategy to use in NP-guided therapy in clinical practice. Overall, much of the literature on congestion consists of observational data.
One limitation of some of the congestion trials was the absence of a strictly defined therapeutic strategy that was applied according to biomarker results. In contrast with most trials in this field, the CHAMPION trial used a strict therapeutic algorithmic strategy according to congestion assessments (1), which might be one of the factors that led to the success of this trial. Interinvestigator management practice variability could be a major source of “noise” within trials that could be reduced by implementing protocol-specified management strategies for treating congestion.
Importantly, 2 recently published trials that investigated the effect of vasodilator treatment, the TRUE-AHF (Efficacy and Safety of Ularitide for the Treatment of Acute Decompensated Heart Failure) (42) and RELAX-AHF 2 (Efficacy, Safety and Tolerability of Serelaxin When Added to Standard Therapy in AHF) (43) trials, provided neutral results. These results greatly hampered the concept of reaching better mid-term outcome by a better and/or smarter treatment of congestion during the early hospitalization period. Notwithstanding the latter, it should be acknowledged that in both trials, interventions lasted for 48 h, and treatment was not tailored to the congestion phenotype (including hemodynamic congestion) of patients admitted with AHF. In addition, the CHAMPION trial, which monitored and subsequently acted on congestion during a longer period of time, was successful. From our point of view, it is likely that future successful AHF trials would not use decongestion therapies equally, regardless of the congestion profile, but would instead constitute genuine congestion-guided treatment trials that use specifically defined individualized therapeutic algorithms triggered by individualized biomarker or clinical assessment targets according to the clinical context. This hypothesis is appealing, although extensive work is needed to fill in the gaps in evidence with regard to optimal decongestion therapy, especially in the aftermath of the results of TRUE-AHF and RELAX-AHF2. From our standpoint, there is a need for a trial that tests a personalized algorithm, in contrast to the strategy used in TRUE-AHF and RELAX-AHF 2, which were based on LUS and echographic biomarkers of congestion during the hospital stay in patients with AHF. In these trials, as in the CHAMPION trial, a management algorithm personalized to the clinical setting would guide investigator intervention, thus decreasing interinvestigator variability, and ultimately ensuring that the information leads to appropriate interventions. These individualized novel strategies have the potential to answer many lingering questions related to congestion-based treatment of acute and post-acute worsening HF.
From our standpoint, a need for a trial at the initial stage of AHF management still remains, when diagnostic uncertainty is often still present. The use of LUS has been successfully explored for triage in pre-hospital pilot studies (44), but its use within the first minutes of acute dyspnea management is not yet commonplace, at least in Europe. We advocate for a randomized trial to investigate the impact of early LUS on clinical outcome in patients with acute dyspnea. Improving the certainty in early diagnosis of congestion due to AHF would enable rapid implementation of treatments for AHF worldwide, and potentially translate into improved outcomes that are direly needed in this group of patients.
Evaluating congestion in HF is as complex as it is crucial, and its timely and effective treatment can improve outcome (35). Because of the importance of congestion in HF, we and others (32), recommend that assessments extend beyond basic clinical evaluations. The currently available various tools need to be applied in a coherent and effective manner within each stage of the patient management cycle.
We recommend the use of the EVEREST clinical score in most nonemergency situations. NPs (already widely used), together with estimated plasma volume variables (almost never used), can be useful to repeatedly assess congestion throughout the patient management cycle. LUS is easy to use, provides rapid results, and may be a practice-changing tool in the pre- and in-hospital management of patients with AHF.
Telemedicine, although currently underdeveloped, could represent the cornerstone for out-of-hospital monitoring of patients with HF by mostly using a combination of self-reported clinical congestion variables and point-of-care biological data.
The CHAMPION trial convincingly showed that treating hemodynamic congestion, mostly by optimizing diuretics and nitrates, results in better patient outcomes (1,35). In patients who do not have wireless pulmonary artery hemodynamic monitoring (as used in the CHAMPION trial), it appears suboptimal to guide treatment based only on clinical considerations. These patients could benefit from a multiparameter approach to detect signs of congestion, including clinical evaluation, biological biomarkers, and ultrasound, to improve outcome and reduce re-hospitalizations. The framework in which these tools could operate in detecting congestion is shown in Figure 2.
Medical writing support was provided by Amy Whereat for Matrix Consultants and funded by an unrestricted grant from Novartis Pharmaceuticals (Bâle, Switzerland). The authors also thank Wendy Gattis Stough and Erwan Bozec for the editing of the paper.
Drs. Girerd, Seronde, Coiro, Chouihed, Bilbaut, Braun, Kenizou, Mailleri, Nazeyrollas, and Rossignol received board fees from Novartis. Drs. Rossignol, Zannad, Mebazaa, Chouihed, and Girerd are supported by a public grant overseen by the French National Research Agency (ANR) as part of the second “Investissements d'Avenir” programme (ANR-15-RHU-0004). Dr. Fillieux is an employee of Novartis. Dr. Abraham is the co-principal investigator (PI) of the CHAMPION trial; a member of Steering Committees of the REDUCEhf and COMPASS-HF trials; is a PI of HOMEOSTASIS trial; and has received consulting fees from St. Jude Medical and Medtronic. Dr. Januzzi has received grant support from Siemens, Singulex, and Prevencio; consulting income from Roche Diagnostics, Critical Diagnostics, Sphingotec, Phillips, and Novartis; and participates in clinical endpoint committees/data safety monitoring boards for Novartis, Amgen, Janssen, and Boehringer Ingelheim. Dr. Zannad is a compensated board member for Boston Scientific; a consultant for Boston Scientific, CVRx, LivaNova, Janssen, Bayer, Pfizer, Novartis, Resmed, Amgen, Quantum Genomics, Takeda, General Electric, Boehringer, Relypsa, ZS Pharma, AstraZeneca, and Roche Diagnostics; and is a compensated speaker with Pfizer and AstraZeneca. Dr. Mebaza has received speaker honoraria from The Medicines Company, Novartis, Orion, Roche, Servier, and Vifor Pharma; and has received fees as member of advisory board and/or steering committee from Cardiorentis, The Medicine Company, Adrenomed, MyCartis, and Critical Diagnostics. Dr. Rossignol is a consultant for Bayer, Relypsa, AstraZeneca, Stealth Peptides, Fresenius, Vifor Fresenius Medical Care Renal Pharma, and CTMA; has received lecture fees from CVRx from Relypsa; and is the cofounder of CardioRenal Diagnostics. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
John R. Teerlink, MD, served as Guest Editor for this paper.
- Acronyms and Abbreviations
- acute heart failure
- B-type natriuretic peptide
- ejection fraction
- heart failure
- lung ultrasound
- left ventricular
- natriuretic peptide
- N-terminal fragment pro- B-type natriuretic peptide
- pulmonary congestion
- Received January 30, 2017.
- Revision received September 15, 2017.
- Accepted September 26, 2017.
- Ambrosy A.P.,
- Pang P.S.,
- Khan S.,
- et al.
- Gheorghiade M.,
- Follath F.,
- Ponikowski P.,
- et al.
- Mullens W.,
- Abrahams Z.,
- Francis G.S.,
- et al.
- Mebazaa A.,
- Yilmaz M.B.,
- Levy P.,
- et al.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- McMurray J.J.,
- Adamopoulos S.,
- Anker S.D.,
- et al.
- Raphael C.,
- Briscoe C.,
- Davies J.,
- et al.
- Yancy C.W.,
- Jessup M.,
- Bozkurt B.,
- et al.
- Duarte K.,
- Monnez J.M.,
- Albuisson E.,
- Pitt B.,
- Zannad F.,
- Rossignol P.
- Brisco M.A.,
- Coca S.G.,
- Chen J.,
- et al.
- Kirkpatrick J.N.,
- Vannan M.A.,
- Narula J.,
- Lang R.M.
- Ohman J.,
- Harjola V.P.,
- Karjalainen P.,
- Lassus J.
- Adamson P.B.,
- Abraham W.T.,
- Bourge R.C.,
- et al.
- Shochat M.K.,
- Shotan A.,
- Blondheim D.S.,
- et al.
- Prosen G.,
- Klemen P.,
- Strnad M.,
- Grmec S.
- Matsue Y.,
- Damman K.,
- Voors A.A.,
- et al.
- McQuade C.N.,
- Mizus M.,
- Wald J.W.,
- Goldberg L.,
- Jessup M.,
- Umscheid C.A.
- Vaduganathan M.,
- Greene S.J.,
- Fonarow G.C.,
- Voors A.A.,
- Butler J.,
- Gheorghiade M.
- Öhman J.,
- Harjola V.P.,
- Karjalainen P.,
- Lassus J.
- Klein L.
- Pellicori P.,
- Cleland J.G.,
- Zhang J.,
- et al.
- Felker G.M.,
- Anstrom K.J.,
- Adams K.F.,
- et al.
- Costanzo M.R.,
- Stevenson L.W.,
- Adamson P.B.,
- et al.
- Ong M.K.,
- Romano P.S.,
- Edgington S.,
- et al.
- Flodgren G.,
- Rachas A.,
- Farmer A.J.,
- Inzitari M.,
- Shepperd S.
- Maisel A.,
- Barnard D.,
- Jaski B.,
- et al.
- Packer M.,
- O'Connor C.,
- McMurray J.J.V.,
- et al.
- Central Illustration
- Clinical Tools for Evaluating Congestion
- Circulating Biomarkers for Evaluating Congestion
- Imaging Tools for Evaluating Congestion
- Pressure and Impedance-Based Tools
- Applying the Right Tool for Each Stage of Patient Management
- Integrating These Tools into Heart Failure Networks: The Role of Telemedicine
- Gaps in Evidence and Unmet Needs
- A Roadmap for the Clinical Validation of Congestion Variables as Actionable Biomarkers for HF Management