Several studies importantly affecting the treatment of patients with heart failure
were reported at the 2018 American Heart Association Scientific Sessions.
In this article, we will review three of them.
Among the most important and unique studies was TRED-HF (Withdrawal of
pharmacological treatment for heart failure in patients with recovered dilated
cardiomyopathy). This small (51 patients) pilot study featured an open-label, randomized,
phased withdrawal (ie, standard drug therapy was withdrawn according
to a prespecified drug-by-drug schedule, not all at once) of guideline-directed
medical therapy in patients who were devoid of heart failure symptoms, were
receiving standard multidrug, guideline-based therapy (including loop diuretic,
β-blocker, ACE inhibitor, ARB, or MRA, or a combination of these drugs), and who
had manifested LVEF ≤40%, abnormal left ventricular end diastolic volume, and
an NT-proBNP concentration ≥250 ng/L (ie, who had had HFREF) an average of
4.9 years from the time that the diagnosis was made and therapy begun and who
then improved on therapy to a non–heart failure, noncardiomyopathic state with
LVEF that had increased at least 10% to >50%, with normal left ventricular end diastolic
volume and NT-proBNP concentration <250 ng/L (average=72 ng/L). At the
end of 6 months after randomized withdrawal, a comparison was made of the
number of patients in each group that remained “recovered.” At that point, therapy
was reestablished among those from whom it had been withdrawn and was
withdrawn among 25 of the 26 in whom it had been continued and follow-up continued
in this cohort.
The results were relatively dramatic. During the initial 6-month follow-up interval,
11 of the 25 patients who were randomized to drug withdrawal (44%) relapsed to
their pretherapy state. Medication withdrawal resulted in a mean 9.5% decrease in
LVEF versus baseline, a 15 bpm increase in heart rate, a 7 mm Hg rise in diastolic blood pressure, and a 5.1 point decrease in the Kansas City Cardiomyopathy
Questionnaire score, the latter a substantial and clinically meaningful deterioration.
Of the 26 patients who continued their prerandomization therapy, none
relapsed (between-group difference P<0.0001). No patient in either group died,
but 3 serious adverse events, none cardiovascular, occurred in the withdrawal
group. However, among the 25 patients who underwent withdrawal after completing
therapy during the initial 6-month on therapy follow-up, 9 relapsed (a
Kaplan-Meier event rate of 36%).
None of the data enabled prediction of those patients who would relapse.
Although the study was small, the results were so very consistent that reasonably
firm conclusions can be drawn. Specifically, among patients with HFREF, withdrawal
of therapy should not usually be attempted, at least until we can predict
who will relapse. As the study’s presenter, John Halliday, PhD, of Imperial College,
London, stated, “Improvement in function represents remission rather than permanent
recovery for many patients.”
For clinicians, patients, and drug regulatory professionals, the study is highly illuminating.
For several distinct cardiovascular indications, multiple drugs now are
FDA- and EMA-approved. Once one or several drugs are thus approved, a new drug
of another group (so-called “class”) for the same indication usually must be tested
on a background that includes the already approved drug(s). This is true primarily
for ethical reasons, so that all study subjects can be expected to realize the
benefits for which the earlier drugs were approved and administered. Therefore,
application of the new drug after approval requires simultaneous treatment with
the drugs that were used as background drugs during the pivotal clinical trials.
To justify stopping one or several of the background drugs, new appropriately designed
clinical trials would be needed. Such trials are complex and expensive and
generally have not been performed for most indications. Therefore, as new therapies
are introduced, the “cocktail” administered for some common conditions, like
HFREF, becomes progressively larger, more expensive, and potentially associated
with increasing adverse events, implications for pregnancy, or deleterious interactions
with drugs given for other indications. TRED-HF is the first trial to assess
the risk of discontinuing background drugs for HFREF once clinical and objective
indications of reversal of the underlying pathophysiology have been identified.
The results of this pilot study strongly suggest that, at least for HFREF, the current
multidrug standard therapy should continue indefinitely. This study can be
expected to serve as a template for the study of background removal for other
Another important and potentially practice-changing study was PIONEER-HF
(Comparison of Sacubitril–Valsartan versus Enalapril on Effect on NT-proBNP in
Patients Stabilized from an Acute Heart Failure Episode). This study compared
the effects of the sacubitril-valsartan combination versus enalapril on NT-proBNP
change and adverse events among 881 patients with HFREF and acute decompensated
heart failure randomized between the two treatments. The data were
presented by Eric J. Velazquez, MD of Yale University, US (Yale was one of the institutions
participating in this TIMI-organized study).
Study patients were ≥18 years of age with an LVEF ≤40% and a NT-proBNP concentration
≥1600 pg/mL or a BNP concentration ≥400 pg/mL and who had received a
primary clinical diagnosis of acute decompensated heart failure, including signs
and symptoms of fluid overload. Patients were enrolled ≥24 hours and ≤10 days
after the initial presentation to the hospital and were still hospitalized when
initiated into the study. Prior to randomization, patients needed to be hemodynamically
stable (as indicated by systolic blood pressure ≥100 mm Hg for the
preceding 6 hours, no increase in intravenous diuretic dose and no intravenous
vasodilators during the preceding 6 hours and no intravenous inotropic drugs for
the preceding 24 hours).
Compared with baseline values, NT-proBNP concentration decreased significantly
(P<0.001) more (46.7%) at 4 and 8 weeks of randomized therapy in the sacubitril–
valsartan group than in the enalapril group (which manifested a 25.3% decrease).
Although NT-proBNP is a nonvalidated surrogate for clinical alterations, it does
suggest that reduction in fluid overloading was highly unlikely to be less effective
on the combination drug than on enalapril. Importantly, the rates of worsening
renal function, hyperkalemia, symptomatic hypotension, and angioedema did not
differ significantly between the two treatments.
The importance of the results of PIONEER-HF is that the use of the sacubitril-valsartan
combination, found to be superior to enalapril in patients with chronic
stable HFREF and not to have unacceptable safety problems in that setting in the
PARADIGM-HF trial, can also be administered in the setting of acute decompensated
heart failure without loss of efficacy and without incurring an adversity
penalty. Without these data, therapy-naive patients with acute decompensated
heart failure would need to begin with enalapril or some other ACE inhibitor or
ARB and, when stabilized, transition to the combination drug. Thus, PIONEER-HF
allows more efficient management of patients with acute decompensated heart
failure than had been possible previously, with the likelihood that the mortality
benefits of sacubitril-valsartan can be realized earlier than they might be with the
current staged-therapy protocol.
OPTIMIZE HEART FAILURE CARE: EFFECT OF ADHERENCE TO GUIDELINES VS
Several drugs have been tested for their effects on hospitalization for HFREF. The
SHIFT study also assessed the effect of therapy on rehospitalization rates, demonstrating
a significant drug-mediated reduction in rehospitalization when that drug
was employed on a background of the remainder of guideline-based HFREF therapy.
However, the effect of adherence to guideline-recommended therapy has not
been systematically assessed from prospectively collected data. In a study mounted
by collaborating investigators from institutions within the Russian Federation,
data from an international prospective multicenter Optimize Heart Failure Care
program that were collected over 12 months from 635 patients hospitalized with
worsening heart failure were analyzed to assess the effects of physicians’ and
patients’ adherence to guideline-based therapy on rehospitalization. Physicians’
prescribing adherence was classified as good (use of drugs from all guideline-recommended
heart failure medication groups), moderate (use of more than half
the recommended medication groups), or poor (use of less than or equal to half
the guideline-recommended medication groups). Patients’ adherence was assessed
by patient-reported compliance via a purpose-designed questionnaire.
Again, three groups were defined: good (always took all prescribed medications at
target doses), moderate (sometimes failed to take all prescribed drugs), or poor
(patients failed to take all prescribed medications). As explained by Yuri Lopatin,
MD of the Volgograd State Medical University, Volgograd, Russian Federation, who
presented the study, after 12 months of follow-up, physicians’ adherence did not
reveal a clear pattern of hospitalization reduction compared with baseline, but
rehospitalization was significantly higher than baseline (P<0.0001) when patients’
adherence was poor in any physicians’ adherence group. The authors concluded
that educational initiatives are necessary for both physicians and patients if optimal
impact of current drug therapy is to be achieved.
Many other important studies relating to heart failure were presented, but these
three have particularly important implications for current care.