Screening Adults With Obesity to Reduce Heart Failure Events

Overview

Rationale: Obesity prevalence in Dutch adults increased to 14.2% in 2020. Obesity is strongly associated with cardiovascular disease, especially heart failure (HF). HF is a serious condition with significant morbidity and mortality. HF in people with obesity often remains undetected for a relatively long time, because symptoms are attributed to the obesity and not to possible HF. As a result, individuals seek help late for already advanced HF. Screening may reveal HF risk factors or a HF diagnosis. Early treatment initiation will improve prognosis, both in terms of quality of life and morbidity and mortality. Objective: To investigate whether active screening for early signs of HF and its risk factors in adults with obesity without known heart disease improves clinical outcome. Study design: Investigator driven, not blinded, randomized controlled superiority trial. Study population: Consecutive individuals with obesity (body mass index ≥30 kg/m2) ≥ 45 years, without known cardiac disease, who sign up to participate in a Combined Lifestyle Intervention program, will be recruited. Intervention: Participant randomized to the intervention will undergo an active screening on HF and its risk factors, using anamnesis, physical examination, an electrocardiogram, blood tests and an echocardiogram. Main study parameters/endpoints: The main study endpoint is a combined endpoint of left ventricular dysfunction and/or HF.

INTRODUCTION AND RATIONALE Obesity (body mass index (BMI) ≥30 kg/m2) prevalence in Dutch adults increased to 14.2% in 2020. In Rotterdam this percentage is even higher. Obesity is strongly associated with cardiovascular disease, especially heart failure (HF). Unpublished data of our research group show that 20% of subjects with obesity without known cardiac disease who are evaluated for bariatric screening, fulfil the diagnostic criteria for HF. HF is a serious condition with significant morbidity and mortality. Dutch health care expenditures for HF are high (currently over half a billion Euros per year), and most of this expenditure is made in hospital care, underscoring the importance of prevention. Apart from the obesity itself, prevalence of other HF risk factors such as diabetes (20%) and hypertension (30%) is high in obese adults. There is no active screening for HF or its risk factors in adults with obesity in the Netherlands. HF in people with obesity often remains undetected for a relatively long time, because symptoms are attributed to the obesity and not to possible HF. As a result, individuals seek help late for already advanced HF. Screening may reveal HF risk factors or a HF diagnosis. Early treatment initiation will improve prognosis, both in terms of quality of life and morbidity and mortality. OBJECTIVES Main research question: does active screening in primary care for HF or its risk factors in adults with obesity improve outcome, defined as a lower prevalence of a combined endpoint of left ventricular (LV) dysfunction and/or HF admission? Hypothesis: Primary care screening for early signs of HF in patients with obesity without known heart disease is effective, both in terms of clinical outcome measures and cost-effectiveness. Screening leads to a lower prevalence of LV dysfunction and HF after one year of follow-up due to initiation of treatment of HF and risk factors for HF identified at the time of screening. Primary Objective: To investigate whether active screening for early signs of HF and its risk factors in adults with obesity without known heart disease improves clinical outcome. Sub-questions: 1. What is the prevalence of undiscovered HF in adults with obesity participating in a combined lifestyle intervention (CLI) program, as identified at baseline by active screening? 2. What is the prevalence of undiscovered risk factors for HF (e.g. hypertension, diabetes) in adults with obesity participating in a CLI program, as identified at baseline by active screening? 3. What are, apart from the obesity itself, risk factors for having HF at baseline in adults with obesity participating in a CLI program? 4. Does active screening for early signs of HF and its risk factors in adults with obesity without known heart disease participating in a CLI program lead to a better of quality of life as compared to standard care at 1 year follow-up? 5. Does active screening for early signs of HF and its risk factors in adults with obesity without known heart disease participating in a CLI program lead to improvement of dyspnea (NYHA classification) at 1 year follow-up in screened patients with HF at the time of screening? 6. Is the effect of the studied screening program driven by treatment of risk factors for HF in screened adults without HF at the time of screening, or by treatment of HF in adults with diagnosed HF at the time of screening? 7. Is the studied screening program in adults with obesity cost-effective as compared to usual care of adults with obesity participating in a CLI program in The Netherlands? STUDY DESIGN SCOR(hf)E is an investigator driven, not blinded, randomized controlled superiority trial. During two years, prospective inclusion of 468 subjects with obesity, aged 45 years and older, without a history of cardiac disease, will take place at 3 organizations of a CLI intervention. Subjects will be randomized 1:1 to either an active screening for HF or a continuation of standard care (a watchful waiting policy; these people will not undergo any investigations at baseline). Subjects randomized to screening for HF will be evaluated using anamnesis, physical examination, an ECG, blood tests and an echocardiogram. Screening may reveal additional HF risk factors such as hypertension or diabetes that will be treated and followed according to guidelines. Patients will be given appropriate lifestyle advice. Patients with a HF diagnosis will be referred to a cardiologist and treated according to the guidelines. The hypothesis of the study is that one year of treatment of the subset of screened patients with HF or HF risk factors identified at the time of screening, will lead to a lower prevalence of LV dysfunction and acute HF admissions (the main study endpoint) in the intervention group (the patients that underwent the HF screening) versus the control group. After one year, echocardiography will be performed in all subjects, both in the intervention and standard care group. Asymptomatic LV dysfunction (assessed with echocardiography) was chosen as a component of the primary endpoint as it reflects a heightened risk status, specifically to the later development of HF. Presence of HF will be defined as HF symptoms that have required an emergency hospital admission. SAMPLE SIZE CALCULATION Sample size calculation Follow-up time and sample size calculation are partly based on findings from the STOP-HF trial. This trial met its primary endpoint, showing a reduction of LV dysfunction and/or HF admissions from 8.7% to 5.3% by a screening program using B-type natriuretic peptide. Estimated effect size: The investigators will use global longitudinal strain (GLS), an echocardiographic parameter clearly more sensitive for LV dysfunction than the parameters used in the aforementioned STOP-HF trial. In our CARdiac Dysfunction In Obesity - Early Signs Evaluation (CARDIOBESE) study, 26% of patients with obesity without known HF ánd not fulfilling the criteria of LV dysfunction of STOP-HF, had significantly decreased GLS (currently used as a criterion for HF diagnosis). A conservative estimate would be that a year after screening, this proportion will be reduced to 19% due to the effect of treatment of patients with HF or risk factors for HF. Also, since the investigators can use improved treatment options developed during the last decade, the effect of the screening is expected to be larger than in STOP-HF. In STOP-HF, acute HF admission occurred in 2.1% of the participants in the control group versus 1.0% of the intervention group. Sample size calculation (two tailed, using α=0.05 and a power level of 80%) based on the aforementioned estimated effect size: proportion of the primary endpoint in the control group 36.8% (26% (decreased GLS) + 8.7% (fulfilling the criteria of LV dysfunction of STOP-HF) + 2.1% (acute HF admissions)), proportion of the primary endpoint in the intervention group 24.0% (19% (decreased GLS) +4.0% (fulfilling the criteria of LV dysfunction of STOP-HF; effect of the screening is expected to be larger than in STOP-HF since the investigators can use improved treatment options) + 1.0% (acute HF admissions): sample size of 199 per group. Withdrawal of study participation is expected to be maximal 5%. Therefore, the investigators will include 420 subjects. RANDOMISATION, BLINDING AND TREATMENT ALLOCATION Participants will be randomly assigned to a screening for HF or usual care (1:1). For allocation, a computer-generated block randomization sequence will be used, stratified by participating CLI practice. An independent physician from Franciscus Gasthuis \& Vlietland will extract the randomization schedule from an electronic randomization tool and send the codes in sealed, sequentially numbered, opaque envelopes to the participating practices. STUDY PROCEDURES Participants randomized to standard care will not undergo any tests related to the study at baseline. Participants randomized to a screening for HF will receive an appointment at either the Franciscus Gasthuis \& Vlietland hospital or Star-shl diagnostic center, according to preference of the participant. Here, anamnesis and physical examination will be performed, blood samples will be taken and an ECG and echocardiogram will be acquired. Diagnosis of HF will be based on the current guidelines. When HF is diagnosed, the patient will be referred to a cardiologist of the Franciscus Gasthuis \& Vlietland. When there are incidental findings, such as increased blood glucose or hypertension, patients will be referred to their general practitioner for follow-up and treatment when necessary. After 1 year follow-up, both participants randomized to standard care and to screening for HF will be approached to assess the combined primary endpoint of LV dysfunction and/or an acute HF admission. Participants will be asked whether they have had an unplanned hospital admission for HF. If this is the case, information will be requested from the hospital in question. A joint panel consisting of three cardiologists will evaluate the findings and results of all diagnostic tests that took place during the admission. Adjudication of diagnoses by a panel of physicians is an accepted reference standard for the evaluation of HF, for which a true reference standard is lacking. In case the panel does not reach consensus, a majority decision will count. Also, in all participants an echocardiogram will be acquired, again in either the Franciscus Gasthuis \& Vlietland hospital or Star-shl diagnostic center. Finally, participants will be asked to fill-out the Health Related - Quality of Life (HR-QoL) with the EQ-5D-5L questionnaire. STATISTICAL ANALYSIS All analyses will be carried out on an intention-to-treat basis. The distributions of variables will be tested for normality by the Shapiro-Wilk test. The significance threshold will be defined as P \<0.05 and all tests are 2-sided. Missing data will be handled using multiple imputation with chained equations with 10 imputations and 10 iterations per imputation. Predictive mean matching will be used for all variables. Primary study parameter The difference in prevalence of LV dysfunction and HF between the intervention group and the control group will be analysed using logistic regression modelling. Secondary study parameters * (subquestion 1 and 2) The prevalence of HF and/or undiscovered risk factors for HF as identified by active screening will be described as a percentage of the total screened population (the intervention group). * (subquestion 3) HF prevalence will be presented stratified by sex, age (tertiles), BMI (tertiles), waist circumference (tertiles), diabetes, hypertension and obstructive sleep apnea. Also, the investigators will provide descriptive statistics (parameters mentioned in the previous sentence) of the subjects in the intervention group with HF versus without HF at baseline. Differences between the groups in age, BMI and waist circumference will be tested using independent t-tests in case of normal distribution or Mann-Whitney-U tests in case of non-normality. Differences in % male, diabetes, hypertension and obstructive sleep apnea will be tested with Chi-square tests. * (subquestion 4) Findings of the HR-QoL questionnaire will be compared between the intervention and standard care groups, using independent t-test or Mann-Whitney-U test for the total scores, and Chi square tests for the subscales. * (subquestion 5) The Chi Square test will be used in screened patients with HF at the time of screening, to compare NYHA classification at baseline and follow-up.