Author + information
- Received September 12, 2016
- Revision received November 1, 2016
- Accepted November 3, 2016
- Published online March 6, 2017.
- Hong Yang, BMeda,
- Kazuaki Negishi, MD, PhDa,
- Ying Wang, BMed, MSa,
- Mark Nolan, MBBSa and
- Thomas H. Marwick, MBBS, PhD, MPHa,b,∗ ()
- aMenzies Institute for Medical Research, Hobart, Tasmania, Australia
- bBaker-IDI Heart and Diabetes Institute, Melbourne, Victoria, Australia
- ↵∗Address for correspondence:
Prof. Thomas H. Marwick, Baker-IDI Heart and Diabetes Institute, 75 Commercial Road, Melbourne, Victoria 3004, Australia.
Objectives The purpose of the study was to evaluate the benefit of care guided by the detection of stage B heart failure (SBHF) using advanced echocardiography for the reduction of new HF in the community.
Background The detection of nonischemic SBHF has been facilitated by advanced echocardiographic imaging modalities. However, improved outcomes have not been proven as they are predicated on benefit of treatment.
Methods Between September 13, 2013 and November 6, 2015, 618 asymptomatic community-based patients with HF risks (age 71 ± 5 years) were randomized to care guided by advanced echocardiography (myocardial deformation and detailed diastolic function) versus usual care. Evidence of SBHF led to advice to the patients and their primary physicians to initiate treatment with angiotensin-converting enzyme inhibition and beta-adrenoceptor blockade. The trial followed the PROBE (Prospective Randomized Open Blinded Endpoint) design. Participants were followed for 1 year for the primary composite endpoint of death from cardiovascular causes and new HF.
Results Advanced echocardiography identified 219 as having SBHF and treatment was advised. Over a mean follow-up of 13 ± 6 months, 67 reached the primary endpoint. The incidence rate of HF was no different between the 2 arms (p = 0.47), likely because only 43% initiated therapy, and only 9% achieved target dose. Among subjects needing therapy on the basis of imaging and adherence to therapy, imaging-guided care showed a 77% lower hazard for the primary outcome (p = 0.04).
Conclusions The detection of SBHF from strain and diastolic function evaluation was associated with a higher incidence of incidence HF and death. The efficacy of pharmacological intervention with angiotensin-converting enzyme inhibition and beta-adrenoceptor blockade is limited by its uptake, and alternative strategies should be considered. (Tasmanian Study of Echocardiographic Detection of Left Ventricular Dysfunction [TAS-ELF]; ACTRN12614000080628)
- community detection
- medication nonadherence
- pharmacological intervention
- randomized controlled trial
- stage B heart failure
Stage B heart failure (SBHF) has been defined in the American College of Cardiology/American Heart Association guidelines to include asymptomatic patients with abnormal left ventricular (LV) structure or function (1). Patients with SBHF are at higher risk for developing overt heart failure (HF) (2). In contrast to stage A HF ([SAHF], i.e., HF risk factors) (3), where therapy is directed toward the control of cardiovascular (CV) risk factors, cardioprotective therapies are recommended in SBHF (1), on the basis of trials showing pharmacological interventions to delay and reduce the burden of HF in the SBHF population (4,5). The difference between SAHF and SBHF is a problem, because although highly prevalent, nonischemic SBHF requires an effective screening strategy for its recognition. Whereas modern imaging techniques may be able to provide this strategy, their incorporation into routine care needs evidence that management will be altered.
Accordingly, we developed a community-based screening program in elderly patients for detection of SBHF using echocardiography and subsequent cardioprotection. Our primary aims were to assess whether imaging-guided cardioprotective therapy, when added to usual care (UC), would reduce the rate of incident heart failure and to assess the feasibility of coupling an effective community screening program with interventional treatment in high-risk asymptomatic elderly individuals with SBHF.
The TAS-ELF (Tasmanian Study of Echocardiographic Detection of Left Ventricular Dysfunction) is a community-based screening and interventional program. The study followed a PROBE (Prospective Randomized Open Blinded Endpoint) design (6). Individuals were randomized to either undergo an imaging-guided screening strategy for early detection and treatment of SBHF or to continue with usual care (7). The data were collected at various clinical sites within the State of Tasmania, Australia.
Asymptomatic community-based patients (≥65 years of age) with HF risks volunteered for enrolment through local community presentations and media advertising. Patients were eligible for inclusion if they had SAHF risk factors (1), which include the following: hypertension ([HTN], based on blood pressure ≥140/90 mm Hg and self-report of HTN including antihypertensive medication); type 2 diabetes mellitus (based on self-report of diagnosis including medical management); obesity (body mass index ≥30 kg/m2); previous chemotherapy; family history of HF; previous history of heart disease (but not existing HF). We excluded subjects with the following: symptoms or a known history of HF; known coronary artery disease; more than moderate valvular heart disease; reduced LV ejection fraction (EF) (<40%) on baseline echocardiography; already taking both trial medication, beta-blocker (BB) and angiotensin-converting enzyme inhibitor (ACEi)/angiotensin-receptor blocker (ARB), at baseline and contraindication of BB and/or ACEi and ARB. We also excluded those in whom we were unable to acquire interpretable images from baseline echocardiography (inclusion and exclusion in Online Table 1).
The primary composite endpoints were incident HF and death from CV causes. Potential HF symptoms were assessed through regular follow-up phone calls, followed by symptom surveillance questionnaires and clinical visits. New onset HF was adjudicated by a blinded endpoint committee using Framingham criteria at 1 year. Loss to follow-up was defined as not having replied for evaluation in ≥2 months.
Sample size was calculated based on the following: 1) an expected prevalence of abnormal cardiac function in ∼50% using advanced echocardiographic imaging; 2) an expected 7.5% annual rate of loss to follow-up; 3) a previously reported 12% annualized rate of incident HF among patients with evidence of SBHF who are receiving UC (8); 4) assumption of a 50% reduction of events with intervention compared with UC. A sample size of 400 in each randomized group would provide 80% power to document the benefit of therapy at a 2-sided α = 0.05.
The trial was monitored by the Data and Safety Monitoring Board with termination guidelines for futility (if conditional probability of rejecting the null hypothesis was unlikely to achieve statistical significance) and feasibility (if recruitment or other aspects of its conduct were unable to fulfil requirements due to unforeseen circumstances).
Randomization was done using a central web-based program with adaptive allocation stratified by diabetes status. The eligible participants were randomized to advance echocardiographic (AE) imaging (involving measurement of global longitudinal strain [GLS] and diastolic function) versus UC (continuing with their UC treatment for primary risk factors such as HTN and diabetes). The randomization list and intervention list were prepared by assigned persons who were blinded to details of the investigations.
Patient report outcome measures and functional capacity
All participants enrolled in the study underwent a physical examination and standard patient report outcome measures questionnaires relating to health status (EuroQol 5 Dimensions Index), activity status (Duke Activity Score Index), symptom status (Minnesota Living with HF Questionnaire), depression (Patient Health Questionnaire-9), and anxiety (Generalized Anxiety Disorder Questionnaire).
Functional capacity was assessed using a 6-min walk test distance following a standardized protocol (9).
Standard transthoracic 2-dimensional and Doppler echocardiographic studies were performed using standard equipment (Siemens ACUSON SC2000, Siemens Healthcare USA, Mountain View, California) and transducer (4V1c, 1.25 to 4.5 MHz; 4Z1c, 1.5 to 3.5 MHz) in accordance with the American Society of Echocardiography guidelines (10). Diastolic function was assessed by measuring mitral inflow peak early diastolic velocity (E), peak late diastolic velocity (A), E/A ratio, and E-wave deceleration time (11–13). Echocardiographic LV hypertrophy (LV mass index by body surface area >95 g/m2 in women and >115 g/m2 in men), left atrial enlargement was assessed according to recommended cutoffs (10). Diastolic dysfunction (DD) grade was defined as previously described (11,12). Grade I: E/A <0.8, E/e′ <10, pulmonary venous inflow S > D. Grade II: 0.8 < E/A < 1.5, E/e′ >13 or left atrial enlargement, or presence of mid-diastolic forward flow (L-wave), or positive Valsalva (>50% increase of E/A ratio). Grade III: E/A >1.5, E-wave deceleration time <140 ms. LV global peak longitudinal strain (GLS) measurements were obtained from gray scale–recorded images in the apical 4-chamber, 2-chamber, and long-axis views. Strain was analyzed using velocity vector imaging (Syngo VVI, Siemens Medical Solutions, Siemens Healthcare USA). GLS was measured online by averaging strain from the region of interest in 3 apical views. Impaired GLS was defined using cutoff of <18% (14). All echocardiographic measurements were obtained online before participants left the clinic. Echocardiographic definitions of SBHF were impaired GLS (<18%) and/or ≥grade I diastolic dysfunction.
Cardioprotective treatment for SBHF
The treatment protocol was a combination of ramipril (starting with 1.25 and up-titrated to 10 mg/day) with the addition of metoprolol (start 12.5 and up-titrated to 100 mg/day) (titration plan in Online Table 2). Cardioprotection was initiated by the patients’ general physicians under the guidance of the investigators. The titration plan and guidance for monitoring was proposed to subjects with evidence of SBHF who required treatment in the AE arm and their physicians. As some patients were already on 1 of the 2 agents (at submaximal dose) at recruitment, any pre-existing ACEi/ARB or BB, other than ramipril and metoprolol, were up-titrated to the maximal tolerable dose (Online Table 2).
General guidance regarding risk reduction was shared with all subjects, and a report (including clinical and imaging information, according to randomization arms) was shared with all patients and their physicians.
The primary analysis involved comparison on an intention-to-treat basis, in the belief that this would provide direct evidence of the clinical effectiveness of screening for SBHF.
Those who required treatment on the basis of results of AE were notified of this at the screening visit and recontacted within 4 to 6 weeks to reinforce the need to follow-up with their general physicians.
A process evaluation was performed within a period of 3 months to allow reasonable time for up-titration. Adherence to protocol was defined on the basis of treatment initiation and dose maximization.
Data are presented as mean ± SD after testing for normal distribution (Shapiro-Wilk test). Data deviating from normality are expressed as median (interquartile range). Categorical variables are expressed as percentages. All analyses were performed on an intention-to-treat basis. The comparability of baseline characteristics in the 2 arms was assessed by chi-square tests for categorical variables and the Mann-Whitney U test for continuous variables. Cox proportional hazard regression model with time-dependent covariates was used to assess the relative risk of the primary outcome of individuals under UC versus those under imaging-guided care. Survival analysis was performed using the Kaplan-Meier method, and the differences in survival between groups were assessed by the log-rank test. Statistical analyses were performed using a standard statistical software package (SPSS software version 22.0, SPSS Inc., IBM, Armonk, New York). Statistical significance was defined as p < 0.05.
This report follows the recommendations of the 2010 Consolidated Standards of Reporting Trials Statement (15). This study was performed in accordance with a research protocol approved by the Tasmanian Human Research Ethics Committee and registered with the Australian and New Zealand Clinical Trials Registry (ACTRN12614000080628). Individual written informed consent was obtained from participants after explanation of the nature and purpose, complexity, and level of risk of the study.
Patient selection and characteristics
Between September 13, 2013 and November 6, 2015, a total of 1,026 self-referred community participants ≥65 years of age, and with nonischemic HF risk factors, volunteered in response to advertising through community organizations and local media. After clinical and echocardiographic screening for eligibility, 618 (age 71 ± 5 years, 48% men) met inclusion criteria. The most common HF risk factors were HTN (79%), diabetes mellitus (51%), and obesity (44%); 79% of individuals had ≥2 of the listed risk factors. Of the total, 6.5% of participants were already taking BB and 67% were on ACEi or ARB treatment at baseline; none were taking both at baseline and none were at maximal doses.
Of these 618 individuals who underwent randomization, 308 were assigned to the AE-guided care arm. The remaining 310 were assigned to the UC arm (Figure 1). There were no differences in the demographic, clinical, socioeconomic features, nor were the patient-reported symptoms and activity status between the 2 groups (Table 1). There were 23 patients who were in atrial fibrillation (AF) at baseline: 12 in AE arm and 11 in UC arm. Of the 12 in the AE arm, all had abnormal GLS and treatment was required on the basis of abnormal GLS. None of the 11 AF of the UC arm required treatment for abnormal EF.
Screening results and pharmacological intervention
Comprehensive echocardiograms were performed in all subjects. Of the 308 individuals in the AE arm, evidence of SBHF was found in 219 (71%) defined by abnormal GLS or DD, with abnormal GLS present in 113 (37%) and DD in 211 (69%); 83 individuals (27%) had both. Patients with abnormal function were older and had significantly higher ARIC (Atherosclerosis Risk In Communities) HF risk score and Charlson comorbidity score compared with the remaining AE patients (p < 0.002). Treatment advice with titration plan was proposed to these 219 patients as well as to their general physicians. There were 18 patients in the AE group who had a mid-range EF (40% to 53%), of whom 17 required treatment on the basis of abnormal GLS or DD. Of the 310 individuals in the UC arm, no treatment was advised based on GLS and DD criteria. However, 5 individuals were started on treatment on the basis of mild reduction of LVEF (40% to 53%); none had LVEF <40%.
Side effects were self-reported in 31 individuals, but there was no reported syncope or other severe adverse effects. The most common side effects were dry cough from ramipril and bradycardia and lethargy from BB. Side effects led to discontinuation of the medication in 4% in the AE arm and none in the UC arm, with most continuing with minor adjustment of medication. Side effects were also reported in individuals in the UC group. These are summarized in Table 2.
At the last follow-up clinic (March 18, 2016), 23 participants (11 from AE) were contactable but unable to attend the follow-up assessment. Contact was lost with 8 participants (1.3%): 4 from the AE arm. At 13 ± 6 months, a total of 22 (18 AE and 4 UC) participants (4%) had withdrawn from the study because of reluctance to attend further follow-up, unwillingness to take additional medication, but in only 3 cases due to side effects. Therefore, follow-up for this outcome assessment was 91% complete (275 of 308 vs. 290 of 310) (Figure 1).
There were 5 deaths (2 of CV causes). New HF symptoms developed in 65 patients. The primary composite endpoint of new HF and CV death occurred in 67 patients (11.8%) with an annualized incidence rate of 10.9%. The 3 deaths from non-CV causes were in both arms, whereas the 2 CV deaths were both in the AE group. The incidence of the primary endpoint was 13% (37 of 275) in the AE arm and 10.3% (30 of 290) in the UC arm (p = 0.25). There was no difference in event-free survival using Kaplan-Meier analysis between 2 arms (log-rank chi-square = 0.53; p = 0.47) (Figure 2).
Initiation and up-titration of medication
Medication initiation and up-titration was assessed using self-reporting questionnaires and telephone calls. Almost all patients—307 of 308 in the AE arm and 309 of 310 from the UC arm—responded to our evaluation. Figure 3 summarizes medication status in the entire cohort. An increment of medical therapy (determined by initiation or up-titration) was achieved in 147 individuals across both arms. In the AE arm, an adjustment of medication was achieved in 94 of the 219 SBHF patients (43%) with abnormal imaging (Figure 3). Of these 219 subjects, 20 (9%) had achieved recommended target dose, and 64 had initiated 1 or both medications without dose increments because of concern about reduction of blood pressure or other potential sources of intolerance. For other reasons (generally blood pressure control), an adjustment of medication was also achieved in 15% of the UC arm, and in 7 subjects in the AE arm without evidence of SBHF. The Data and Safety Monitoring Board stopped the trial in November 2015, on the basis of poor therapeutic adherence (∼40%) in both initiation and up-titration of medication.
The potential clinical benefit of screening guided care was evaluated in Cox proportional hazard model and with the adjustment of the number of subjects who required treatment. Given the low medication adherence (43%), the model was also adjusted for subjects who were actually adherent to treatment (treatment up-titration). Echocardiographic imaging-guided care was associated with 77% lower risk for primary outcome at 1 year (hazard ratio: 0.23; 95% confidence interval: 0.06 to 0.98; p = 0.047) (Table 3).
In the absence of treatment protection, individuals with abnormal GLS, DD, and the presence of both were associated with worse outcome (Figure 4). No treatment benefit was observed in subgroups with abnormal GLS, DD, and the presence of both (p > 0.4).
The results of this study show that among individuals in the community who were ≥65 years of age, with mainly multiple HF risk factors, evidence of SBHF was identified in 71% with the use of AE, and these showed an annualized event rate of 10.9%. Despite the prognostic information from imaging, the trial emphasizes the difficulties in providing cardioprotection in asymptomatic elderly individuals at risk of HF. Pharmacological intervention was unsuccessful in intention-to-treat analysis, although there may have been potential benefits from imaging-guided care in adherent patients.
Definition of new HF
HF is a clinical diagnosis, and patients may minimize or deny their symptoms in the early phases, making clinical recognition difficult. Prevalence estimates may vary broadly depending on the diagnostic criteria; a recent meta-analysis reported that incident HF diagnosis in 8 of 15 included studies was based on a nonstandardized clinical description (16).
Differences in the diagnostic criteria for HF may also affect the outcome assessment in these studies. Among 4 commonly used HF diagnostic criteria (Framingham, Boston, Gothenburg, and European Society of Cardiology criteria) (17), there were significant differences in predicting clinically relevant outcomes including incident hospital admission. As disturbances of cardiac structural and functional characteristics at baseline were best predicted using the Framingham HF criteria (17), we selected the Framingham HF criteria to adjudicate events in our study. However, although the absolute 3-year risk of hospital admission following a Framingham HF diagnosis was 6.1% (odds ratio: 6.9; 95% confidence interval: 1.3 to 36.1; p = 0.022) (17), the annualized rate of incident HF in our study was 10.9%. This likely reflects the high clinical risk of these subjects, most of whom had multiple HF risk factors. In addition, although we excluded symptomatic HF at entry, Framingham criteria may not capture mild or early HF (stage C1) in individuals with atypical symptoms (2). Moreover, a high rate of incident HF was observed in a similar community study of a cohort with combined diabetes and HTN (8), in whom E/e′ > 15 (detected in 23%) was used to categorized SBHF. In our cohort, the prevalence of increased E/e′ was similar in those with both HTN and type 2 diabetes mellitus (20%).
Diagnosis and management of SBHF
Imaging screening for SBHF is based on the concept that this stage is a precursor of overt clinical HF, and has been advocated in practice guidelines (18) because SBHF is associated with a 5-fold greater risk of symptomatic HF (19). Imaging by echocardiography is generally believed to be safe and accurate for the diagnosis of SBHF. Although EF is the only functional feature in the conventional criteria for SBHF (20), both strain and DD are associated with adverse outcome and impaired functional capacity (21).
The accuracy and prognostic value of diagnostic tests have no direct effect on patient outcomes unless indirectly by influencing the therapeutic pathway. The evidence to date for pharmacological therapy in SBHF is largely based on EF. This supports ACEi as the foundation of management of SBHF (22), with the improvements in all-cause mortality, recurrent CV events, and progression to HF (23,24). In the HOPE (Heart Outcomes Prevention Evaluation) study (25), ACEi also significantly reduced the rate of death, myocardial infarction, and stroke, and reduction of HF admission and death was reported after treatment in a similar population with SAHF (26). Evidence also supports the use of BB in asymptomatic patients (22). In the SAVE (Survival and Ventricular Enlargement) and SOLVD (Studies of Left Ventricular Dysfunction) trials, concomitant use of BB therapy was associated with reductions in CV death and HF (23,24).
In this study, we sought to gather evidence on the benefit of early detection of SBHF using new functional markers to guide management. We assessed this by randomizing subjects to either screening strategy or UC and comparing their outcomes, using an intention-to-treat approach (7). This design allows assessment of the benefit of early detection, as well as recognition of potential risks associated with diagnostic procedures, including false positives resulting in over treatment. In addition, our intention was also to determine whether an intervention works among a community screening population. Unfortunately, the negative results of the current study were attributable to lack of effectiveness (27) in medication delivery following imaging screening. Side effects were uncommon, and the main causes were reluctance of patients and physicians to initiate and up-titrate therapy in the presence of a normal EF and the absence of symptoms.
The follow-up time was only 1 year. The method adopted for assessment of medication use was self-report rather than pill counts and rate of prescription refills. However, there is no optimal means of achieving this (28), and self-report is simple and feasible in the context of community screening and has been strongly associated with adverse cardiac events (29).
We excluded patients with EF <40% because previous clinical trials showed a benefit from therapy (22). The study did not incorporate the analysis of HF with mid-range EF, a category that has been recently proposed in the guidelines (30), and only 1 patient was in this category in the absence of abnormal GLS.
AF was not excluded in the initial design, because the optimal management of this important subgroup in nonischemic SBHF is unclear. However, the numbers of patients with AF was small, and they were equally distributed among the groups. Of the 12 in the AE arm, all had abnormal GLS and treatment was required based on abnormal GLS. None of the 11 AF patients in the UC arm required treatment for abnormal EF.
The application of advanced echocardiographic markers is feasible in the community diagnosis of SBHF and predicts incident HF. Despite their risk levels, the adherence to intervention in this asymptomatic group with a normal EF was poor. Subsequent attempts with this study design might include a disease management program to ensure adherence with therapy.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients >65 years old with HF risk factors have a high prevalence of SBHF, evidenced by LV hypertrophy and LV diastolic or systolic dysfunction. These patients are at risk of early development of clinical HF.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: The appropriate therapeutic response to the detection of SBHF is undefined. This study showed no benefit from screening for SBHF, but uptake of cardioprotective therapy and dose titration were both inadequate.
TRANSLATIONAL OUTLOOK: A strategy to identify SBHF seems unlikely to alter the progression to HF in the absence of an integrated disease management program.
The authors gratefully acknowledge the contribution of our tireless volunteer coordinators, Diane Binns, Jasmine Prichard, and Jane Mitchell.
For a supplemental table, please see the online version of this article.
This study was partially supported by Tasmanian Community Fund, Hobart and Diabetes Australia, Melbourne and Siemens Healthcare Australia, Melbourne, Australia. None of the funding agencies had any role in design, analysis, or interpretation of this study. Dr. Yang is supported by a Health Professional Scholarship from the National Heart Foundation of Australia (100307). Dr. Negishi is supported by an award from the Select Foundation. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Pamela Douglas, MD, served as the Guest Editor for this article.
- Abbreviations and Acronyms
- peak late diastolic velocity
- angiotensin-converting enzyme inhibitor
- advanced echocardiography arm
- atrial fibrillation
- angiotensin-receptor blocker
- diastolic dysfunction
- mitral inflow peak early diastolic velocity
- peak early mitral annular tissue Doppler velocity
- ejection fraction
- global longitudinal strain
- heart failure
- left ventricular
- stage A heart failure
- stage B heart failure
- usual care arm
- Received September 12, 2016.
- Revision received November 1, 2016.
- Accepted November 3, 2016.
- American College of Cardiology Foundation
- Hunt S.A.,
- Abraham W.T.,
- Chin M.H.,
- et al.
- Ammar K.A.,
- Jacobsen S.J.,
- Mahoney D.W.,
- et al.
- Cheng J.,
- Zhang W.,
- Zhang X.,
- et al.
- From A.M.,
- Scott C.G.,
- Chen H.H.
- Abhayaratna W.P.,
- Marwick T.H.,
- Smith W.T.,
- Becker N.G.
- Nagueh S.F.,
- Smiseth O.A.,
- Appleton C.P.,
- et al.
- Yang H.,
- Negishi K.,
- Otahal P.,
- Marwick T.H.
- Di Bari M.,
- Pozzi C.,
- Cavallini M.C.,
- et al.
- Yancy C.W.,
- Jessup M.,
- Bozkurt B.,
- et al.
- Kosmala W.,
- Jellis C.L.,
- Marwick T.H.
- Reed B.N.,
- Sueta C.A.
- Pfeffer M.A.,
- Braunwald E.,
- Moye L.A.,
- et al.,
- for the SAVE Investigators
- Ho P.M.,
- Bryson C.L.,
- Rumsfeld J.S.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.