Author + information
- Received November 23, 2013
- Revision received January 4, 2014
- Accepted January 6, 2014
- Published online April 8, 2014.
- Lakhmir S. Chawla, MD∗∗ (, )
- Charles A. Herzog, MD†,
- Maria Rosa Costanzo, MD‡,
- James Tumlin, MD§,
- John A. Kellum, MD‖,
- Peter A. McCullough, MD, MPH¶,#,
- Claudio Ronco, MD∗∗,
- for the ADQI XI Workgroup
- ∗Department of Medicine, Division of Nephrology, Washington DC Veterans Affairs Medical Center, and Department of Anesthesiology and Critical Care Medicine, George Washington University Medical Center, Washington, DC
- †Department of Medicine, Division of Cardiology, Hennepin County Medical Center, University of Minnesota, Minneapolis, Minnesota
- ‡Midwest Heart Specialists–Advocate Medical Group, Naperville, Illinois
- §Southeast Renal Research Institute, Chattanooga, Tennessee
- ‖Center for Critical Care Nephrology, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania
- ¶Baylor University Medical Center, Baylor Heart and Vascular Institute, Baylor Jack and Jane Hamilton Heart and Vascular Hospital, Dallas, Texas
- #Heart Hospital, Plano, Texas
- ∗∗Department of Nephrology, Dialysis, and Transplantation, International Renal Research Institute, San Bortolo Hospital, Vicenza, Italy
- ↵∗Reprint requests and correspondence:
Dr. Lakhmir S. Chawla, Department of Anesthesiology and Critical Care Medicine, George Washington University Medical Center, 900 23rd Street NW, RILF, G-105, Washington, DC 20037.
Structural heart disease is highly prevalent in patients with chronic kidney disease requiring dialysis. More than 80% of patients with end-stage renal disease (ESRD) are reported to have cardiovascular disease. This observation has enormous clinical relevance because the leading causes of death for patients with ESRD are of cardiovascular disease etiology, including heart failure, myocardial infarction, and sudden cardiac death. The 2 systems most commonly used to classify the severity of heart failure are the New York Heart Association (NYHA) functional classification and the American Heart Association (AHA)/American College of Cardiology (ACC) staging system. With rare exceptions, patients with ESRD who do not receive renal replacement therapy (RRT) develop signs and symptoms of heart failure, including dyspnea and edema due to inability of the severely diseased kidneys to excrete sodium and water. Thus, by definition, nearly all patients with ESRD develop a symptomatology consistent with heart failure if fluid removal by RRT is delayed. Neither the AHA/ACC heart failure staging nor the NYHA functional classification system identifies the variable symptomatology that patients with ESRD experience depending upon whether evaluation occurs before or after fluid removal by RRT. Consequently, the incidence, severity, and outcomes of heart failure in patients with ESRD are poorly characterized. The 11th Acute Dialysis Quality Initiative has identified this issue as a critical unmet need for the proper evaluation and treatment of heart failure in patients with ESRD. We propose a classification schema based on patient-reported dyspnea assessed both pre- and post-ultrafiltration, in conjunction with echocardiography.
The literature provides multiple definitions of heart failure (1). A recent attempt to describe this syndrome posited that heart failure is a mechanical and neurohumoral syndrome in which the heart fails as a pump, resulting in stasis of blood in the lungs and venous system, fatigue, effort intolerance, and reduced longevity. From a purely mechanical perspective, heart failure is defined as a clinical syndrome that can result from any structural or functional cardiac disorder impairing ventricular filling or ejection of blood (2). Incidence of structural heart disease is high in patients with advanced chronic kidney disease (CKD), particularly those with end-stage renal disease (ESRD) requiring dialysis. In large observational cohort studies, more than 80% of patients with ESRD are reported to have cardiovascular disease (3,4). This observation is of paramount clinical relevance because the leading causes of death for patients with ESRD are due to cardiovascular disease, including heart failure, myocardial infarction, and sudden cardiac death (3,5–7).
In CKD, and particularly ESRD, 3 major mechanisms induce and exacerbate cardiomyopathy: volume overload, pressure overload, and nonhemodynamic factors associated with CKD. The nonhemodynamic factors are manifold, and include activation of the renin-angiotensin system, catalytic iron-dependent oxidative stress, inflammation, and stimulation of profibrogenic factors (5,7) These pathological mechanisms cause direct and immediate effects and affect long-term disease progression in these patients (5,7).
The American Heart Association (AHA), American College of Cardiology (ACC), and European Society of Cardiology (ESC) have developed guidelines recommending that heart failure be classified in a manner that reflects its risk factors, severity, and natural history (2,8,9). In the AHA/ACC guidelines, stage C heart failure defines patients with current or prior symptoms of heart failure. The primary form of maintenance renal replacement therapy (RRT) worldwide is hemodialysis, and a smaller proportion of patients use peritoneal dialysis. With rare exceptions, lack of fluid removal by RRT in patients undergoing dialysis leads to dyspnea and congestion due to salt and water retention. Thus, by definition, all patients undergoing dialysis, even those with no structural heart disease, experience New York Heart Association (NYHA) functional class (FC) III to IV symptoms if fluid removal by RRT is delayed. Thus, management and treatment of patients undergoing dialysis with heart failure is difficult due to the dynamic nature of fluid overload.
A “real world” example is illustrated by the following clinical vignette: a patient undergoing dialysis is referred to the cardiologist because of worsening heart failure symptoms. Because of transport to and from the dialysis clinic in addition to the time required for dialysis, which usually exceeds 6 h, patients on dialysis rarely schedule non-nephrology doctor's visits on the same day as dialysis sessions. Thus, visits to the cardiologist almost always occur prior to a dialysis day. During the visit, the cardiologist correctly identifies fluid overload, and the patient is likely to be returned to the nephrologist for “better” dialysis and more ultrafiltration. This is problematic because the nephrologist likely sent the patient to the cardiologist because the patient was already receiving the maximal tolerable ultrafiltration.
The incidence, severity, and outcomes of heart failure in patients undergoing dialysis are poorly characterized and potentially undertreated, because the severity of symptoms relative to the frequent changes in volume status occurring before and after RRT are not witnessed and remain poorly documented (5,7). In addition, none of the existing practice guidelines for the treatment of fluid overload in patients on dialysis comment on whether it is appropriate to apply the AHA/ACC/ESC stages of heart failure or the NYHA FC system (10).
For advancement in the treatment of heart failure in patients on dialysis, a methodology that improves the characterization of heart failure in these patients is essential and urgently required. The implementation of classification systems for any disease syndrome can have a positive impact on disease management (10–12). For example, introduction of the CKD staging system and the Risk, Injury, Failure, Loss, End-stage renal disease (RIFLE) (11,12) staging system for acute kidney injury have allowed clinical investigators to better characterize disease processes, conduct epidemiological studies, guide clinical trial enrollment, standardize practice guidelines, and improve clinicians' ability to track disease progression (13,14).
The 11th Acute Dialysis Quality Initiative (ADQI) meeting was convened to focus on cardiorenal syndrome (15). The working group that focused on cardiorenal syndrome type IV (advanced CKD and progressive heart disease) identified this clinical and research issue as a critical unmet need in the care of patients on dialysis with heart failure. We based this assessment on the fact that accurate risk stratification of patients with any given disease state is essential to improve clinical status and outcomes.
The specific methods for the ADQI conferences were developed and refined over the first 4 international conferences (16). Briefly, our methods comprise: 1) a systematic search for evidence, with review and evaluation of the available literature; 2) establishment of clinical and physiological outcomes and of measures to be used for treatment comparisons; 3) description of current practice and the rationale for use of current techniques; and 4) analysis of areas in which evidence is lacking and future research is required to obtain new information. The topics chosen for each conference are selected on the basis of the following criteria: 1) prevalence of the clinical problem; 2) estimates of variation in clinical practice; 3) potential influence on outcome; 4) potential for development of evidence-based guidelines; and 5) availability of scientific evidence. A detailed report of the ADQI process has been published previously (16).
Before the conference and according to the structure of previous ADQI consensus meetings, we performed a systematic review of the literature (11,17). Specifically, we used the search terms “heart failure,” “peritoneal dialysis,” “hemodialysis,” “renal replacement therapy,” “chronic kidney disease,” “dialysis,” “guidelines,” “consensus,” and “end-stage renal/kidney disease,” combined with “prognosis,” “major adverse cardiovascular events,” “myocardial infarction,” “stroke,” “volume overload,” “fluid overload,” “sudden cardiac death,” “death,” and “mortality.” In view of the volume of retrieved literature, only representative publications are cited in this proposal. Furthermore, we opted to base our staging proposal on clinical scenarios for which data are ample. We also excluded the specific clinical situations pertaining to renal or heart transplant or to mechanical circulatory support.
Based on the literature identified prior to the conference, the following key questions were considered:
1. Are there current heart failure staging systems that can be applied specifically to patients undergoing dialysis?
2. What are the critical features of a staging system that can be easily used by clinicians to establish appropriate diagnostic and therapeutic approaches?
3. How can a new heart failure staging system specifically address the unique nature of nonphysiological periodic volume removal, which characterizes all forms of dialysis?
Current Staging Systems of Heart Failure
Multiple heart failure staging systems and classification schemes have been proposed. In 1977, Wagner and Cohn (18) proposed a classification system based on 3 broad pathophysiological mechanisms: systolic dysfunction, diastolic dysfunction, and volume overload. These 3 conceptual etiologies remain relevant in the current assessment and treatment of heart failure. Since 2005, the ESC/ACC/AHA have developed consensus guidelines that include a staging system for heart failure to reflect its severity as well as its natural history (2). However, the most commonly used system to characterize the severity of heart failure symptoms is the NYHA FC (10) (Table 1), which is based on a grading of dyspnea relative to the intensity of physical activity. The ESC/ACC/AHA and NYHA classification systems have proven to be robust and have been used successfully to assess the epidemiology and severity of heart failure. They have been the basis for establishing inclusion criteria and response to therapy in seminal clinical trials conducted in patients with heart failure (19–21).
However, neither of these staging systems can be used in their current forms to assess the severity of heart failure in patients on dialysis because dyspnea in these patients is not solely attributable to heart failure and its severity changes in relation to the timing of volume removal. For nearly all patients undergoing dialysis, delay in or absence of RRT results in dyspnea due to fluid overload even if the patient's heart is structurally normal. Thus, all patients on dialysis can theoretically be considered in NYHA FC III to IV, which can improve to NYHA FC I after fluid removal. In patients not on dialysis with acutely decompensated heart failure, euvolemia can be restored and maintained for variable periods of time with appropriate medical therapies; however, all patients undergoing dialysis begin to retain salt and water immediately after RRT treatment, and accumulation continues until the next prescribed RRT (this occurs in both peritoneal dialysis and hemodialysis to varying degrees). We concluded that a staging system for heart failure in patients on dialysis must be able to define heart failure symptoms in relation to the response to the periodic nonphysiological volume removal that occurs with all forms of dialysis. Thus, the timing of assessment and the periodicity of dialysis must be taken into account. In addition, we sought to develop a patient-centered staging system that was easily understandable by care providers and therefore easily usable for planning further diagnostic evaluations and treatment for patients with coexisting heart failure and dialysis-requiring ESRD.
Proposed Staging System
We considered a number of possible solutions to the problem outlined in the previous text. To specifically exclude patients with normal hearts on dialysis, the proposed classification system incorporates echocardiographic criteria that indicate the presence of underlying structural heart disease. We modified the NYHA FC of dyspnea severity into a new system with a description of heart failure symptoms and their response to fluid removal by RRT/ultrafiltration. The 3 elements of our proposed staging schema are as follows:
1. Standardized echocardiographic evidence of structural and/or functional heart abnormalities;
2. Dyspnea occurring in the absence of primary lung disease, including isolated pulmonary hypertension; and
3. Response of congestive symptoms to RRT/ultrafiltration.
The standardized echocardiographic criteria are based on the evidence of mild to moderate echocardiographic disease cutoffs that are detailed in the American Association of Echocardiography consensus guidelines (which are consistent with the European Society of Echocardiography guidelines) (22,23). The echocardiographic criteria that suggest or are supportive of cardiac disease are summarized in Table 2.
The classification system that we propose is summarized in Figure 1.
Patients who present with severe dyspnea that is relieved by RRT/ultrafiltration are categorized by their post-therapy status. For example, a patient with echocardiographic evidence of left ventricular hypertrophy who presents to the dialysis clinic with dyspnea at rest, undergoes RRT, and then experiences no symptoms would be classified as ADQI class 2R. In addition, the RRT/ultrafiltration metric does not necessarily suggest a single therapy. Symptomology relieved by RRT/ultrafiltration assumes that the patient is receiving RRT at the appropriate frequency.
Strengths and limitations
Like previously published staging systems, the proposed classification system will require prospective testing and validation in appropriate patient cohorts. In addition, because echocardiographic findings are a prerequisite for entry into the classification system, standardization of the collection and documentation of these echocardiographic data is imperative.
Studies in patients undergoing dialysis that deployed continuous hemodynamic monitoring before, during, and after RRT with ultrafiltration have demonstrated that interdialytic volume accumulation results in increased pressures in the pulmonary artery and right ventricle (Fig. 2) (24,25). The proposed staging system accommodates the “tidal nature” of nonphysiological volume removal. Another strength of this proposed system is that symptom grading is similar to grading in other heart failure symptom scales (e.g., the NYHA FC system) that are familiar to clinicians. In addition, the proposed staging system is based on patient-reported symptomology and is thus patient centered.
Patients with heart failure on dialysis pose unique diagnostic and therapeutic challenges; for example, a patient with mild left ventricular dysfunction undergoing dialysis who becomes dyspneic may be deemed to have fluid overload and referred back to the nephrologist for standard thrice-weekly hemodialysis and additional ultrafiltration. Thus, RRT is a form of treatment for ESRD with heart failure. However, in some cases, this therapy fails to improve or resolve the symptoms of dyspnea. This clinical scenario raises multiple questions regarding presence and severity of heart failure, type of monitoring required, and therapeutic approaches needed to treat the persistent symptoms. A validated classification system would prove useful here and could potentially be useful for uniform assessment of heart failure in patients undergoing dialysis. In addition, surveillance of these patients will allow better understanding of the temporal aspects of heart failure and ESRD. Specifically, does heart failure usually precede ESRD, is it the reverse, or is this a cardiorenal interaction?
Validation of the proposed system would start with classification of a large cohort of patients undergoing dialysis. The initial assessment would include collection and documentation of appropriate echocardiographic data. Echocardiography is recommended for all patients on dialysis within 1 to 3 months of starting RRT (5,7). In addition, patients would be queried about the presence of dyspnea pre- and post-RRT/ultrafiltration. Simple tools such as the Likert scale and visual analogue scale have been shown to be useful in the assessment of dyspnea in patients with heart failure and could be used as appropriate (26,27). Assessment of patient symptoms pre- and post-RRT is routine and considered standard of care. However, these data may not always be gathered in standard database data fields, and standardization may be required. Once the prevalence of the various stages of the new classification system is determined, follow-up to assess outcomes would be needed. Our operating hypothesis is that mortality risk would increase through each of the proposed stages. After initial longitudinal assessment, the proposed staging system might be useful for clinical trial entry and design. For example, does more frequent hemodialysis improve ADQI class function? Are outcomes for patients with advanced heart failure and ESRD improved with peritoneal dialysis versus hemodialysis? In addition, certain standard heart failure medications (e.g., angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, beta-blockers) might be better suited for certain ADQI heart failure stages than for others. Recent data suggest that treating patients with ESRD with spironolactone may improve structural heart disease and improve survival, and the proposed classification schema should be deployed in conjunction with future studies of aldosterone blockade in patients with ESRD (28,29).
In addition to future research concerning the descriptive epidemiology and prognosis of patients on dialysis classified with this new scheme, future studies may use novel technologies such as advanced imaging and biomarkers to aid in support of the classification or in the traditional domains of testing (screening, diagnosis, prognosis, and management) (30,31).
Important therapies for patients with heart failure on dialysis may include more frequent hemodialysis, preference for peritoneal dialysis, or polyelectrolytes that remove sodium via the gut (32). For patients who are dyspneic even with good volume control, standard heart failure therapies may require additional adjustments. However, before treatment advances can be made, it is essential to be able to differentiate patients with fluid overload alone in the absence of structural heart disease from patients in whom fluid overload and dyspnea can be directly attributed to underlying cardiac disease.
The precise criteria that define echocardiographic evidence of structural heart disease are an important component of any proposed classification, specifically, the differentiation of heart failure with preserved ejection fraction versus volume overload alone. Both clinical entities can present with clinical and echocardiographic features of “diastolic dysfunction.” For the subset of patients in whom this distinction is required, we propose a diagnostic pathway to ascertain the correct diagnosis (Online Appendix). The distinction can be particularly important in patients undergoing dialysis due to the high prevalence of left ventricular hypertrophy. Schwartzenberg et al. (33) demonstrated that 35% of patients with heart failure with preserved ejection fraction, despite elevated left ventricular end-diastolic pressure values, experienced a drop in stroke volume with intravenous nitroprusside. This appears to be due to excessive reduction in filling pressure and thus filling volume (end-diastolic volume) due to diastolic dysfunction. Many patients on dialysis would be expected to experience something similar with volume removal during RRT.
The dynamic assessment (i.e., before and after ultrafiltration) is particularly important given that asymptomatic diastolic dysfunction is prevalent in the community and may be more prevalent in the ESRD population (34). As suggested in the Kidney Disease: Improving Global Outcomes consensus guideline, echocardiography should be performed in all incident patients on dialysis (5). Novel and known cardiac biomarkers, along with dynamic exercise testing, may also be useful in identifying important subgroups (35). More invasive hemodynamic monitoring may be of benefit for some patients and for research purposes (24). In the development of a new classification schema, the epidemiology and phenotype of this population should be well characterized. Other important considerations include the standardization of dyspnea assessment and timing of echocardiographic assessment (36).
At the end of 2010, the patient population on dialysis was more than 300,000 in Europe and 414,000 in the United States; these numbers are projected to increase as the population ages (37). Patients undergoing dialysis are characterized by a tremendous burden of structural heart disease and are uniquely susceptible to the consequences of nonphysiological changes in volume status (5). Therefore, the standardization of heart failure diagnosis, prognosis, and treatment is an urgent concern for these patients and for the clinicians charged with delivery of their medical care.
The authors thank Kiran Khush for reviewing this manuscript and Nan Booth, MSW, MPH, ELS, of the Minneapolis Medical Research Foundation, for manuscript editing. The authors thank the AARVI Onlus Foundation of Vicenza for hosting the ADQI Consensus Conference.
A complete list of the ADQI Workgroup members is in the Online Appendix. The 11th international ADQI Consensus Conference was funded in part by generous support in the form of unrestricted educational grants from Alere Medical, Astute Medical, Fresenius Medical Care, and Gambro Medical. The sponsors had no input on the content of this manuscript. Dr. Costanzo has served as a consultant and speaker for Gambro Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- American College of Cardiology
- Acute Dialysis Quality Initiative
- American Heart Association
- chronic kidney disease
- European Society of Cardiology
- end-stage renal disease
- functional class
- New York Heart Association
- renal replacement therapy
- Received November 23, 2013.
- Revision received January 4, 2014.
- Accepted January 6, 2014.
- 2014 American College of Cardiology Foundation
- Hunt S.A.
- McMurray J.J.,
- Adamopoulos S.,
- Anker S.D.,
- et al.
- Yancy C.W.,
- Jessup M.,
- Bozkurt B.,
- et al.
- The Criteria Committee of the New York Heart Association
- Bellomo R.,
- Ronco C.,
- Kellum J.A.,
- Mehta R.L.,
- Palevsky P.
- Weiner D.E.,
- Tighiouart H.,
- Amin M.G.,
- et al.
- Ronco C.,
- Haapio M.,
- House A.A.,
- Anavekar N.,
- Bellomo R.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Allen L.A.,
- Metra M.,
- Milo-Cotter O.,
- et al.
- Ito Y.,
- Mizuno M.,
- Suzuki Y.,
- et al.
- Schwartzenberg S.,
- Redfield M.M.,
- From A.M.,
- Sorajja P.,
- Nishimura R.A.,
- Borlaug B.A.
- Borlaug B.A.,
- Nishimura R.A.,
- Sorajja P.,
- Lam C.S.,
- Redfield M.M.
- ↵2012 Atlas of CKD & ESRD. United States Renal Data System. Available at: http://www.usrds.org/atlas12.aspx. Accessed February 3, 2014.
- de Simone G.,
- Devereux R.B.,
- Daniels S.R.,
- Koren M.J.,
- Meyer R.A.,
- Laragh J.H.