Can You Deprescribe After Ejection Fraction Improvement?

This transcript has been edited for clarity. 
Ileana L. Piña, MD, MPH: Hello. I’m Ileana Piña, the quality chief for the cardiovascular line at Thomas Jefferson University, and a lifelong heart failure transplant doc. 
I’m here at the ESC. With me today is a researcher from Sao Paulo, Brazil, from the Institute of the Heart, which is a wonderful institution in Sao Paulo. I have visited them. She is here to tell us about their work with reverse remodeling. 
Deborah Belfort, MD: Thank you very much. I’m Deborah. I’m from Heart Institute. It’s a pleasure to be here talking about our research, which is, ” Carvedilol as single maintenance therapy for heart failure with improved ejection fraction: the CATHEDRAL-HF trial. ”
Piña: How did you come up with this idea? How did you start? 
Belfort: In clinical practice, we always wonder if we have to maintain all medications for heart failure once the patient improves ejection fraction. 
Piña: If they improve. 
Belfort: If they improve. Current guidelines recommend we keep all medications since the TRED-HF trial, in which 44% of patients experienced relapse of dilated cardiomyopathy after suspension of all medications, signaling that probably these patients are in remission rather than this being a cure of a disease. We have to keep some degree of neurohormonal blockade to prevent them from relapsing. 
Piña: If you know the etiology, if it happens maybe from hypertension or something else? 
Belfort: For some etiologies, we have to give medications for other reasons. For example, ischemic cardiomyopathy or hypertensive, they require antihypertensive. Maybe for those etiologies it doesn’t change much, but we have patients with idiopathic dilated cardiomyopathy or myocarditis or peripartum cardiomyopathy, which are etiologies that may make us think we can suspend or reduce medications lifelong.
Piña: Tell me about the structure of your study. 
Belfort: We selected patients with reverse remodeling, and we considered patients with previous HFrEF — so previously their left ventricular ejection fraction (LVEF) was lower than 40%, and their current LVEF was greater or equal to 50% with normal NT-proBNP (< 250 pg/mL) — without symptoms of heart failure, and on carvedilol in optimized doses, as well as ACE inhibitors or ARBs, and with or without spironolactone. 
Then we randomized patients in two groups, one to maintain usual treatment and the other one to withdraw heart failure treatment, maintaining carvedilol. We suspended ACE inhibitors, spironolactone, and furosemide. Then we followed both groups initially for 24 weeks, and then we extended follow-up for 52 weeks.
Piña: It was still a highly selected group. 
Belfort: Yes, we excluded some etiologies of heart failure such as ischemic cardiomyopathy, hypertensive, valvular, infiltrative cardiomyopathy, and patients with Chagas, which is common in Brazil. 
Piña: Yes, that’s pretty common.
Belfort: Patients with creatinine clearance less than 30 mg/dL were also excluded.
Piña: You didn’t want the CKD sick patients in there? 
Belfort: No, especially because these patients have indication for ACE inhibitors or ARBs because they have proteinuria. 
Piña: It’s protective of the kidney.
Belfort: We excluded these patients. We had a population with a median age of 55 years old, and the most common etiology for heart failure was idiopathic dilated cardiomyopathy in 60% of cases. The median LVEF at diagnosis was 30% at diagnosis and the current LVEF was 58% in both groups, with NT-proBNP less than 100 pg/mL in both groups, and a slightly reduced global longitudinal strain.
Piña: Those are low NT-proBNPs.
Belfort: We included below 250, and the two groups had around 80 pg/mL. 
Piña: The NT-proBNPs that I see are in the thousands. 
Belfort: We selected patients that were asymptomatic. 
Piña: They had improved, obviously, at some point. 
Belfort: The primary outcome was the recurrence of ventricular dysfunction, but we used one of four criteria: (1) a reduction in LVEF >10%; (2) an increase in left ventricular end diastolic volume > 10%; (3) an increase in NT-proBNP to above 400 pg/mL, or a twofold increase; and (4) clinical evidence of heart failure. If the patient had any one of these, that would count toward the primary outcome. 
What we found at 24 weeks was that three patients in the intervention group and one in the control group had an increase in NT-proBNP, at 10% and 3.7%. One patient in the intervention group, one of these three patients, had also an increase in left ventricular end-diastolic volume.
Piña: Do you find that the NT-proBNP tells you that the patient is remodeling reversely?
Belfort: The NT-proBNP increased before any possible changes in echocardiogram, which is different from literature. Usually, we see echocardiographic changes before the NT-proBNP.
Piña: I see the NT-proBNP go first. 
Belfort: Really? 
Piña: Yes. 
Belfort: The interesting thing is that, when the patient met the primary endpoint, we restarted medications as a safety measure, and then we followed the patients for 1 year.
Those patients who increased NT-proBNP were not the ones who had worsened left ventricular ejection fracture in the future. We don’t know if maybe this initial increase in NT-proBNP was due to suspension of diuretics if they required some degree of diuretic medications.
We’re not sure. What we saw was that, from the four patients that met the primary endpoint at 24 weeks, only one really decreased LVEF in the future.
Piña: That’s pretty good. If you were going to tell clinicians what to do, what would you tell them? 
Belfort: I think the answer right now is to individualize. I think it’s still too early to recommend suspension of medication. I think the point of this study is to open up the area for future researchers.
Piña: We have some work to do. 
Belfort: Yes, I would say to keep the medications, and in individual cases maybe suspend and perform a close follow-up to see if there are any changes in biomarkers or in echocardiographic parameters. I would be cautious about telling everyone to suspend medications. 
Piña: You got the NT-proBNP at baseline, and then you got it when?
Belfort: At 12 weeks, then 24 weeks, and then 52 weeks. 
Piña: What would be the time period that you should see the lowest of the proBNP? Would it be a baseline?
Belfort: We evaluated NT-proBNP at different time points between groups. 
Piña: That’s why I’m asking.
Belfort: The median value was similar between groups. 
Piña: It didn’t really change much. 
Belfort: It didn’t really change much. These three patients, in 12 weeks, they had increased NT-proBNP. A fourth patient had increased NT-proBNP at 24 weeks. Another two patients with worsened ventricular function at 1 year had increased NT-proBNP. The median value was similar between the groups through follow-up as well as the left ventricular end-diastolic volume. The LVEF was similar between groups, but at 52 weeks, it was slightly lower than at baseline.
Piña: You didn’t take anybody with that HFmrEF, 40%-45% ejection fraction? 
Belfort: No. 
Piña: So that’s been excluded. 
Belfort: Yes. 
Piña: Because you know, some of the SGLT2 trials have enriched the population with LVEFs of 40%-49%. 
Belfort: Yes, what we did was try to select the patients that had really improved. 
Piña: You got the real HFrEF.
Belfort: Yes.
Piña: Your longest patient now without the drugs is how long? 
Belfort: It’s 1 year and a few months. 
Piña: What’s your next step? What are you going to do next? 
Belfort: We’re going to follow these patients for longer periods, but we intend to perform a larger study with longer follow-up and using cardiac sensors and maybe genetic tests to better understand this population to see if we can find any predictor of worsening. 
We know that a few patients will worsen in the future despite pharmacologic treatment. Maybe if we can identify these patients, we would not suspend medications in this group, and maybe the rest can use fewer medications in a safe way.
Piña: I have found in my own practice that, if the patients start doing something like drinking, it can raise their NT-proBNPs again and make them sick. You may want to collect that.
Belfort: That’s interesting. In fact, two of the four patients that increased NT-proBNP at 24 weeks started drinking. I don’t know if it was for that, but it happens.
Piña: Yeah, it happens.
Belfort: They had alcoholic cardiomyopathy, but I can’t say it was that for certain because it was too small a number.
Piña: I congratulate you. You did a study that was sequential — you know, one point after the other so that you can really make a good reassessment. Thank you for being with me. 
Belfort: Thank you very much. 
Piña: That’s it for me with this group. I think it’s so fascinating that we don’t talk much about the improvement that we can make in pump function, and we can. I think we need to continue to say that to the clinicians. Don’t miss the opportunity to make your patients better because some of them will change. We have researchers that are working on how to do it, how to diagnose it, and how to tell who isn’t going to get better.
Thank you for joining me today from the ESC. Have a great day.
Ileana L. Piña, MD, MPH, is a heart failure and cardiac transplantation expert. She serves as an advisor/consultant to the FDA’s Center for Devices and Radiological Health and has been a volunteer for the American Heart Association since 1982. Originally from Havana, Cuba, she is passionate about enrolling more women and minorities in clinical trials. She also enjoys cooking and taking spin classes.