#NephJC Chat
Tuesday September 20 at 9 pm EDT
Wednesday September 21 at 9 pm IST
N Engl J Med 2022 Aug 27. doi: 10.1056/NEJMoa2203094. Online ahead of print.
Acetazolamide in Acute Decompensated Heart Failure with Volume Overload
Wilfried Mullens, Jeroen Dauw, Pieter Martens, Frederik H Verbrugge, Petra Nijst, Evelyne Meekers, Katrien Tartaglia, Fabien Chenot, Samer Moubayed, Riet Dierckx, Philippe Blouard, Pierre Troisfontaines, David Derthoo, Walter Smolders, Liesbeth Bruckers, Walter Droogne, Jozine M Ter Maaten, Kevin Damman, Johan Lassus, Alexandre Mebazaa, Gerasimos Filippatos, Frank Ruschitzka, Matthias Dupont, ADVOR Study Group
PMID: 36027559
Introduction
Congestion and hypoperfusion in acute decompensated heart failure are the two clinical variables which demand diligent and speedy optimization. Intravenous loop diuretics are the cornerstone therapeutic agents in patients with acute decompensated heart failure used to achieve decongestion. Evidence suggests that adequate diuresis prior to discharge is an important measure of short-term outcomes and early rehospitalizations (Rubio-Gracia et al, J Int Cardiol, 2018).
However, despite using loop diuretics at a dose of 2-2.5 times the oral maintenance doses, volume overload commonly can not be addressed to a satisfactory extent by the time of discharge from the hospital. The DOSE Trial (Felker et al, NEJM, 2011) noted that only 15 % of the decompensated heart failure patients could achieve optimum decongestion after 72 hours of treatment.
Acetazolamide is a heterocyclic derivative of sulphanilamide developed in the 1950s and is typically considered a weak diuretic. Often regarded as a forgotten diuretic, it continues to be used only for a few niche indications like prevention of mountain sickness and glaucoma (Rahim and Sica, Card in Rev 2011). It is a short acting carbonic anhydrase inhibitor, whose primary action is the inhibition of proximal tubular reabsorption of bicarbonate, sodium, and chloride, resulting in both NaCl and bicarbonate losses. Its diuretic effect is ushered by the appearance of alkaline diuresis. The diuretic efficacy of acetazolamide is modest in view of increased distal segment sodium reabsorption that offsets the proximal sodium losses.. Thus, though the first study of it’s effect (against nothing) was promising (Friedberg et al, NEJM 1953), once loop diuretics entered the picture, head:head comparisons showed that loop diuretics were clearly much more effective (Sim et al, J Pharm Pharmacol 1976). Recently, acetazolamide has also been reported to be an inhibitor of pendrin expression (Zahedi et al, PLoS One 2013) which may mediate its metabolic effects as well as net natriuretic effect. This twitter thread from Christos Argyropoulos and this one from Nayan Arora nicely give an overview of the topic.
Even though loop diuretics are more potent natriuretic agents, their diuretic efficacy is decreased over time due to the onset of diuretic resistance. Acute decompensated heart failure is a state of avid sodium retention, and thereby the dose-response curve of the most commonly used agents, loop diuretics, is shifted downwards and to the right. In a retrospective analysis of 1,153 patients with advanced HF, 402 (35%) patients had diuretic resistance, defined in this study (Neuberg et al, Am Heart J 2002) as the requirement of furosemide >80 mg or bumetanide >2 mg daily. Sequential therapy with the usage of two or more diuretics has been advocated as a more effective strategy to attain better decongestion but there is dearth of concrete evidence in this regard. Check out our previous NephJC write up on this topic here, and the NephMadness discussion here.
Figure showing diuretic dose response curve in heart failure, Casu and Merella, Eur Cardiol, 2015
The addition of metolazone, and spironolactone or amiloride are much more common now in the setting of heart failure. Can addition of a proximal tubular diuretic achieve the ‘total nephron blockade’ goal? It would be logical to presume that addition of an otherwise weak diuretic such as acetazolamide to a loop diuretic can augment its efficacy by mitigating one of the important mechanisms of diuretic resistance. A small prospective single center trial (Verbrugge et al, Eur J Heart Fail, 2019) indicated that addition of acetazolamide to loop diuretics enhances the natriuretic response in acute heart failure patients.
The primary objective of ADVOR Trial (Acetazolamide in Decompensated Heart Failure with Volume Overload) was to analyze the therapeutic benefit of adding acetazolamide to a standardized loop diuretic regimen with regard to achieving enhanced decongestion in acute decongestive heart failure.
The Study
Methods
ADVOR trial was a multi-center, randomized, parallel group, double blind, placebo controlled clinical trial. This was an investigator-initiated academic clinical trial and was not industry funded.
Inclusion Criteria
Patients hospitalized with acute decompensated heart failure AND at least one clinical marker of volume overload:
Edema, pleural effusion, or ascites
N-terminal Pro B type natriuretic peptide level >1000 pg/ml or B-type natriuretic peptide level >250 pg/ml) were included in the study.
All those who were included were receiving oral maintenance diuretic therapy (at least furosemide 40 mg or an equivalent dose of other loop diuretics).
Exclusion Criteria
Those with a history of recent acetazolamide maintenance therapy or other proximal tubular diuretic therapy including SGLT2-inhibitors
SBP < 90 mmHg
An estimated GFR of <20 ml/min/1.73m2
Those who were already receiving more than 80 mg of intravenous furosemide during index hospitalization.
Study participants were randomized in 1:1 ratio to receive either intravenous acetazolamide of 500 mg per day or placebo which were administered soon after randomization and for either next two days or until the occurrence of complete decongestion. All patients received intravenous loop diuretics at double the dose of the maintenance therapy as a single bolus dose at the time of randomization (Day 1) and then given as two split doses (separated by ≥6 hours) on each of the next 2 days (Day 2 & 3).
Escalation therapy was mandated in those who were still volume overloaded AND the cumulative urine volume was less than 3.5 liters over a period of 30 - 48 hours after randomization. Escalation therapy consisted of doubling of loop diuretic therapy / addition of oral chlorthalidone 50mg per day / Ultrafiltration or renal replacement therapy.
It is worth noting all participants received an infusion of 3 grams of magnesium sulfate in 500 ml of 5% dextrose, administered over a period of 24 hours. Additionally, potassium at a dose of 40 mmol was added to the infusions once K < 4 mmol/L - and oral potassium was left to the discretion of the treating team. Similarly, 100 ml of 8.4% NaHCO3 was recommended once HCO3 was < 20 mmol/L.
Endpoints
The primary endpoint was successful decongestion, defined by the absence of signs of volume overload (disappearance of edema except for trace edema, absence of residual pleural effusion and/or ascites) within three days of randomization and without the need for escalation of IV loop diuretics
Secondary endpoints were the composite endpoint of death from any cause, rehospitalization for heart failure during 3 months of follow up, and duration of the hospitalization from randomization to the date of discharge
Data about adverse events was recorded which included parameters such as severe metabolic acidosis, hypokalemia, hypotension and renal events.
Doubling of serum creatinine from the base line or loss of at least 50% of the estimated GFR defined the renal events.
Statistical Analysis
Primary and secondary endpoints were analyzed based on the intention-to-treat principle. Generalized linear mixed model (log-link binomial model) was utilized to evaluate the primary end point. Cox proportional hazards model was utilized to assess the composite end point of death and rehospitalizations. Kaplan-Meier survival curves were used to summarize the results. They estimated that 15% of the patients in the placebo group versus 25% of the patients in the acetazolamide group (i.e. 10% higher absolute benefit) would have successful decongestion within 3 days after randomization. Assuming a two-sided alpha of 0.05 and a statistical power of 80%, we calculated the targeted sample size for the trial to be 494, and to account for a potential withdrawal of 5% of the patients, the trial sample size for enrolment was 519 patients.
Funding
The trial was supported by the Belgian Health Care Knowledge Center under the KCE Trials Program, and no industry support was involved.
Results
The study was conducted at 27 sites, all in Belgium, between November 2018 to January 2022. Adult patients (>18 years) admitted to the hospital on an emergency or elective basis, with a diagnosis of decompensated heart failure and clinical hypervolemia, 2915 were screened. Out of these, 519 patients met criteria, consented, and were randomized(259 to the acetazolamide arm and 260 to the placebo arm). (See Figure S1, supplementary material)
Figure S1 from Mullens et al, NEJM, 2022
Three patients in the intervention arm and 1 patient in the control arm did not receive the intended intervention (reasons mentioned in Figure S1). Follow-up at 3 months was not available for 2 patients in each group.
Table 1 from Mullens et al, NEJM, 2022
Table 1 shows the baseline characteristics of the study population. The cohort was composed of elderly individuals (mean age- 78.2 +/- 8.9 years) with a predominance of males (approximately two-thirds). Also, the population was homogenous for race, nearly all (99%) individuals were White. Co-existing conditions in the total cohort were hypertension (75%), historical evidence of atrial fibrillation (72.4%) and diabetes (47.2%). These patients were on 60 mg (interquartile range 40 to100 mg) median maintenance dose of oral furosemide equivalent. The majority (57%) belonged to NYHA Class III, and 43.2% had an LVEF of <40%. Most patients (81.7%) had an eGFR <60 ml/min/1.73m2.
Primary Outcome
Figure 1 of Mullens et al, NEJM, 2022
Successful decongestion, referring to absence of signs of volume overload (as assessed by a cardiologist within 3 days of randomization) was achieved in 42.2% of acetazolamide arm vs 30.5% of control arm, and the benefit continued over consecutive days (see Figure 1).
Mean urine output on the second morning after randomization was 4.6 (+/- 1.7) liters in those receiving acetazolamide, and 4.1 (+/- 1.8) liters in those receiving placebo. This was consistent with a higher natriuresis as expected (mean 468 mmol versus 369 mmol).
The primary endpoint was consistent across subgroups (Figure 2).
Figure 2 from Mullens et al, NEJM, 2022
Secondary Outcomes
Key secondary outcomes of death from any cause, or re-hospitalisation for heart failure were the same in both groups. Mean duration of index hospitalization was 8.8 days in the acetazolamide arm and 9.9 days in the control arm.
Adverse effects
Severe metabolic acidosis was not observed and incidence of other adverse events, such as hypokalemia and hypotension, were similar in both groups (Table 2).
Table 2 from Mullens et al, NEJM, 2022
The actual numbers are also available here:
Discussion
This study has demonstrated that when acetazolamide is combined with standardized intravenous loop diuretic therapy in patients with acute decompensated heart failure, a higher incidence of successful and faster decongestion within 3 days of randomization was achieved. In addition, the acetazolamide group attained more natriuresis, shorter hospital stay and more chances of getting discharged without the attributes of residual volume overload.
Like for the proverbial parachute, the use of diuretics in heart failure is supported by a low quality of evidence. There is a woeful shortage of studies for agents other than loop diuretics and comparison of different diuretic regimes as strategies for additive efficacy. Acetazolamide, as cited earlier, was evaluated previously with promising results, though further studies confirming this effect are warranted in other settings and different populations. This study fulfills a long standing lacuna surrounding the question, and points towards an uplifting strategy in heart failure. Indeed there is evidence in favor of early effective diuresis, (Matsue et al, J Am Coll Cardiol, 2017) suggests that earlier the time to diuresis, lesser is the observed mortality.
How is this relevant to nephrologists?
With the increasing prevalence of chronic kidney disease worldwide and co-existing risk factors like age, diabetes, hypertension and cardiovascular disease, heart failure is a frequently encountered situation in clinical practice, probably more so than before. Patients may report a concomitant use of acetazolamide for ocular indications, and CNS toxicity has been noted in the elderly and those on dialysis. CKD is an independent risk factor for heart failure. The present study reflects this interaction with CKD, in that the included patients had a median eGFR of ~40 ml/min/1.73m2, and 81.7% had an eGFR <60 ml/min/1.73m2. It is not known of course, whether acetazolamide will retain its efficacy at lower GFRs, and whether the risk of adverse effects (especially acidemia and electrolyte derangements) will be greater.
Strengths
This is a large well done RCT in an area in which there is a paucity to guide combination diuretic strategies. The time period of 3 days for ascertaining the compound diuretic effect is well considered, since theoretically, tolerance to the action of acetazolamide occurs early in the first few days of its use itself. Congestion score used in this study, may be a suitable index since these same parameters tend to be used clinically for deciding success or failure of therapy and institute change in dosing. The therapeutic effect of the addition of acetazolamide can be said to be substantiated by documentation of increased volume of diuresis and natriuresis in this study. Safety profile of acetazolamide was also established here. Metabolic acidosis was not observed in this study, in contrast to many others.
Limitations
The results of this study might not be extrapolated to patients with newly detected heart failure, as majority of the study population were patients with chronic heart failure on at least 40 mg of oral furosemide (or equivalent).
It might be possible that therapeutic results similar to this study could have been replicated with other dosage regimen of loop diuretics or other diuretics combination, e.g. metolazone.
The measures taken by the investigators to mitigate adverse events are not easy to replicate in day to day clinical practise (such as administering 100 ml of sodium bicarbonate when serum bicarbonate goes < 20 meq/L, supplementing 40 meq of IV potassium when serum potassium is < 4 meq/L, and daily magnesium infusions) - or should be considered as necessary co-interventions if we are contemplating adding acetazolamide to our diuretic armamentarium.
Mean hospital stay, though shorter with acetazolamide, was still ~8 days, which is in contrast to data which says that length of hospital stay for heart failure is declining worldwide.
Though there was a poor correlation between weight loss during stay and outcomes of heart failure, the difference in weight loss between groups here was ~0.5 kg. It leaves us wondering whether such a difference can actually lead to the observed reduction in hospital stay with addition of acetazolamide.
Congestion scoring system used in this study used variables which are reflective of extracellular dehydration but not intracellular dehydration.
Limitation with regard to racial and ethnic generalizability as almost all study subjects were white patients.
This study excluded patients on SGLT-2 inhibitors. We are in an era in which flozins should become ubiquitous therapeutic agents even in acute decompensated heart failure (see EMPULSE Voors et al, Nature Med 2022) and they have evidence of benefit in clinical outcomes. Flozins also act on the proximal tubule, would acetazolamide have similar benefits when added to someone already on a flozin? Perhaps we need a flozin compared to acetazolamide trial?
Conclusions
Overall this study is a welcome addition to the sparse evidence for diuretic combinations in heart failure. Further research can be directed to replicating efficacy demonstrated here, and preferably with multiple measurements like quantum of natriuresis, diuretic efficiency, 6 hour cumulative sodium excretion or the natriuretic response prediction equation (NRPE). Until then, no harm in letting the carbonic anhydrase inhibitor lend a helping hand? We say aye!
Summary prepared by
Dr. PS Vali
Nephrologist,
Asian Institute Of Nephrology & Urology,
Hyderabad
Sayali Thakare
Nephrologist,
KEM Hospital, Mumbai
NSMC Interns Class of 2022
Reviewed by Jamie Willows, Jade Teakell, Elena Cervantes,
Joel Topf and Swapnil Hiremath