DR. NG: I’d like to welcome everybody to this roundtable discussion on the topic “Current and Future Strategies for Overcoming Diuretic Resistance in Acute Heart Failure.” I am Dr. Tien Ng, Associate Professor of Clinical Pharmacy at the University of Southern California. The other members of this panel are Dr. Mandeep Mehra, Professor of Medicine at Harvard Medical School and Executive Director of the Center for Advanced Heart Disease at the Brigham and Women’s Hospital, Boston, MA; and Dr. Jo Ellen Rodgers, Clinical Associate Professor at the University of North Carolina Eshelman School of Pharmacy.
It is well recognized that diuretic therapy remains the main approach to volume management in heart failure patients and carries a first-line recommendation according to the American College of Cardiology/American Heart Association (ACC/AHA) guidelines.1 However, in more advanced patients and those who have been on chronic diuretic therapy, many fail to respond to escalating doses of intermittent loop diuretics, a common clinical scenario referred to as “diuretic resistance.”
In patients exhibiting diuretic resistance, several approaches have been utilized. This includes increasing the frequency of administration or initiating a continuous infusion of the loop diuretic, changing to a more potent (on a milligram per milligram basis) loop, or combining a loop diuretic with a thiazide agent. Other approaches to enhancing fluid removal such as ultrafiltration and the addition of other intravenous (IV) vasoactive agents have also been considered. Only a few studies have evaluated the individual efficacy and safety of these different approaches.
There is a paucity of head-to-head studies to guide our approach, with most of the comparative data being limited to the comparison of intermittent bolus to continuous infusion loop diuretic administration. In those studies the results have been somewhat conflicting. In addition, all these data have inherent limitations and are open to interpretation. Thus there continues to be a debate as to the best method for volume management in acute heart failure. In the future we may have new pharmacologic therapies and the role of ultrafiltration and other nonpharmacologic approaches may be clarified.
I thought we could start with the issue of continuous infusion and what is the role of continuous infusion compared to intermittent bolus loop diuretic administration. Perhaps Dr. Mehra could start us off by giving us his impressions on this issue.
DR. MEHRA: Let me first begin by amplifying one important item on the clinical definition of diuretic resistance. We tend to think of a patient with advanced heart failure or heart failure with decompensation as having evidence of diuretic resistance if a certain threshold of exposure to incremental IV diuretic bolus therapy does not evoke an adequate diuretic response. Clinicians consider unresponsiveness to 160 mg or 200 mg of IV furosemide administered as a bolus evokes an anemic urine output response in the subsequent two to three hours as demonstrating clinical diuretic resistance.
It is important to note also that many of these trials did not in fact include just diuretic resistance or diuretic refractory patients. Most of the recent clinical trials, such as Diuretic Optimization Strategies Evaluation (DOSE) looked at patients with congestion in the setting of decompensated heart failure who were being treated with a diuretic regimen with the idea of testing two very specific strategies of diuretic use. First, a low dose compared to a high dose strategy, and the second a continuous infusion strategy compated to a bolus dosing strategy.2 These are clinically useful strategies, which until now have not been subjected to randomized control trials. In that sense the DOSE trial has helped us move forward in a clinical direction.
To summarize the DOSE trial,2 this study suggested that neither of these specific clinical strategies were superior to the other ones, including a low dose/ high dose or bolus dose/continuous dosing with a diuretic. The study found some signals that suggested that the higher dose strategy was associated with less dyspnea but at a risk of a greater incidence of worsening of renal function, at least in the short-term. Thus, the appropriate interpretation of the DOSE trial is that it has shown us evidence of the reasonable safety and effectiveness of all of these strategies when used early in the course of decompensated heart failure. Would you agree with that, Tien and Jo Ellen?
DR. RODGERS: I definitely agree that the DOSE trial demonstrated, in a broad population of patients presenting with acute decompensated heart failure and requiring diuretic therapy ,that there is no difference between IV bolus and continuous diuretic administration, but that there is some benefit with high dose, specifically 2.5 times the oral dose, compared to a low dose. In addition to a reduced area under the curve for dyspnea at 72 hours, patients receiving high dose therapy also experienced a significant reduction in weight and net fluid loss at 72 hours. As Dr. Mehra pointed out, this benefit may be at the expense of temporary worsening of renal function. And thus, in practice, we need to be mindful of patient selection and which patients need more aggressive diuretic response compared to those patients in whom you are not willing to tolerate worsening renal function to gain the additional diuretic response. The former may be the patient who needs more aggressive relief of dyspnea—perhaps the patient in the emergency room with poor oxygen saturation. This patient should be managed with the high dose diuretic strategy. The latter may be the patient with acute or chronic renal failure in whom the risk of worsening renal function does not outweigh the benefit of more aggressive diuresis.
The issue that is critical to emphasize is that the DOSE trial assessed diuretic effectiveness of various diuretic strategies in patients who were requiring diuretic therapy, but may or may not be truly diuretic resistant. And thus, the patients enrolled were simply in decompensated heart failure secondary to fluid overload. Looking at the demographic characteristics of the trial population including baseline systolic blood pressure, serum urea nitrogen and serum creatinine as well as serum sodium, it would be anticipated that these patients would have fairly low in-hospital mortality rates compared to patients with systolic blood pressures less than 115 mm Hg, serum urea nitrogen greater than 43, and serum creatinine >3.75 mg/dL, as demonstrated by prior Acute Decompensated Heart Failure National Registry (ADHERE) data. Patients who are developing diuretic resistance are more apt to be in this latter group. Such patients may be more prone to renal dysfunction and other adverse outcomes with aggressive diuretic therapy. Diuretic response in this patient population has not been assessed.
DR. NG: Yes, I have to agree with both of you. Overall, the trial does give us some sense of the safety and the efficacy of the high versus the low dose. However, one important caveat when it comes to the interpretation of the continuous infusion compared to the intermittent infusion results was that the dosing was fixed in these two arms for the first 48 hours of the study, but then from 48 to 72 hours it became open label. There definitely was a suggestion that the continuous infusion arm had a lower dose requirement overall and required less dose escalation.
In terms of the interpretation of the DOSE trial for the continuous infusion versus intermittent infusion comparison, the lack of difference in overall urine output is in agreement with the recently published study by Dr. O’Connor’s group showing absolutely no difference between continuous infusion and intermittent bolus infusions using the same daily dose.3 On the surface it seems in contrast to the Thompson study,4 which found a trend toward an increase in urine output with continuous infusion. However, it is consistent if one considers that in the latter study the actual dose received over 24 hours was not the same. I still have some issue with the amount of diuretic that was received in the DOSE study for the intermittent versus continuous infusion comparison when the take home message of the study was that there was no difference between the two regimens.
DR. MEHRA: The fundamental issue relates to the application of these findings to clinical practice. Does it not affect clinical practice at all, or do we take comfort in the finding from the DOSE trial that, hey, the continuous high-dose diuretic strategy is reasonably safe and one should not run away from it if you deem as a clinician that that is the appropriate strategy to use by a particular patient? For example, someone with chronic long-standing heart failure with exposure to chronic high dose diuretic therapy, and some renal dysfunction at the time of inception of continuous IV diuretic therapy may be more suitable for chronic infusion therapy. At that point in time I think that my clinician gestalt would be to use a higher dose continuous infusion strategy, not simply because I am worried about the amount of urine output, but perhaps also because of the toxicity associated with the higher dose diuretic bolus regimens that one may encounter. The DOSE trial did not look at toxicity, particularly oto-toxicity; it mainly focused its safety endpoints on renal function.
DR. NG: I think the issue of other safety signals not addressed in the DOSE trial have come up in some of the other studies. A Cochrane Review,5 which was completed prior to the three latest trials, looked at all the smaller studies and found continuous infusion was associated with a lower incidence of ototoxicity compared to intermittent bolus therapy.
DR. MEHRA: That’s correct. In fact, much of that information comes from the chronic renal failure literature as well, where continuous infusions with drugs like bumetanide have been shown to result in a lower incidence of cochlear toxicity.
DR. NG: Okay, so Dr. Rodgers, I think we’re all in agreement that the DOSE trial allows us to feel better about using higher doses of loop diuretics in select acute heart failure patients. But based on the data from the most recent trials by Thompson, Allen, and the DOSE trial, do you believe there is still a strong rationale for using continuous infusion as opposed to intermittent bolus.
DR. RODGERS: I think taking all three of those trials in total, there is no difference in outcomes efficacy between continuous infusion and intermittent bolus in a broad population of acute decompensated heart failure patients. But what about those patients who are starting to exhibit refractoriness as the clinician continues advancing doses of either continuous or intermittent bolus loop diuretics, well beyond what was high dose in the DOSE trial? In such scenarios, clinicians oftentimes start to consider additional agents to combine with loop diuretic therapy, such as metolazone or hydrochlorothiazide, diuretics with a different mechanism of action.
Any time a clinician is starting to push to higher diuretic doses with no response or is starting to consider additional therapies, one should be cognizant of the risk of additional electrolyte wasting and worsening renal function and assured that these safety parameters are closely monitored. This is when patient selection for more aggressive diuretic regimens must be carefully considered. In addition, there remain many unanswered questions about appropriate alternative therapies and the role of such in further increasing or perhaps mitigating these risks. Alternative therapies studied to date include low dose dopamine, low dose nesiritide, and other vasoactive agents.
With regard specifically to ototoxicity, this is an adverse effect that is probably very underappreciated and perhaps it’s because clinicians rarely monitor it. Most clinicians are very aware that if you administer a large, single, IV bolus dose of 180 mg of furosemide too rapidly (for example, greater than 4 mg per minute), the risk of ototoxicity is increased, but I’m less certain at what precise dose of a furosemide continuous infusion a clinician should avoid or more assertively monitor for ototoxicity. Perhaps the focus of future research should include assessing the role of these additional or alternative therapies in minimizing toxicities including ototoxicity. Do you have any suggestions, Dr. Mehra?
DR. MEHRA: I think that the data suggest that cochlear toxicity or ototoxicity is related to the peak effects and not the trough effect, so the higher the bolus dosing, the more the chance that one could have cochlear toxicity, particularly if there are other ototoxic agents being administered concomitantly. For example, if someone is receiving a drug like a gentamycin-type analog, for instance, along with a diuretic infusion or a diuretic bolus, I think there’s magnification of that ototoxicity. Frankly that isn’t the overriding concern of most clinicians, and in fact we do not systematically study it, as you’ve pointed out. I often am more concerned in the diuretic resistant patients with the best dosing strategy. I know that, Tien, you have done some work in that regard and I wonder if we could turn the conversation toward discussing what kind of combination diuretic strategies really work in reducing fluid overload.
DR. NG: Clearly another option which is commonly used in practice is the combination of diuretic agents, specifically a loop diuretic added to a thiazide agent, a thiazide which is still effective even in the setting of poor renal perfusion. There are a lot of observational and small studies that have evaluated whether or not the addition of a thiazide, such as metolazone, to a loop diuretic, such as furosemide or bumetanide, augments sodium excretion and urine output. A paper that was published in the Journal of the American College of Cardiology summarized some of these data found a signal for increased urine output.6 Our group recently evaluated the addition of metolazone to furosemide, compared to just a continuous infusion of furosemide, and found a statistically significant greater increase in mean hourly urine output with the combination compared to furosemide used by itself.
What was interesting about our data, which are consistent with some of the existing data, is there was actually no signal for worsening renal function in our patients who received the combination, although there was a signal for an increased risk of electrolyte disturbances. So the conclusion of our study, since it was an observational study, was because of potential differences in clinical efficacy and safety, there is a great need for further study in prospective trials of these alternative strategies, such as combining agents, to truly understand whether these therapies are more effective while remaining relatively safe.
DR. RODGERS: These studies have made important contributions to the literature, but as Tien explains, are not optimally designed trials. Thus, there is need for prospective evaluations to address these unanswered questions. When conducting these future prospective studies, it will be important to include endpoints such as 30-day rehospitalization. If these alternative regimens more effectively diurese, then length of hospital stay may even be reduced by a day or two. With that said, I would be concerned that patients might subsequently suffer other adverse consequences such as continued unwanted diuresis and related adverse effects post-discharge. This is an important question to address with agents such as metolazone, given the long half-life and considerable intra- and inter-patient variability in onset of action with this agent. Future prospective trials need to address outcomes, such as length of stay and rehospitalization at 30 days, for these very reasons.
There are many reasons these trials have not been prospectively conducted in the past, and one major rate-limiting factor is the uniqueness of the population of patients with refractory fluid overload. When clinicians are faced with these patients, they often start to inquire about the appropriateness of a right heart catheter and inotropic therapy simply because diuretic refractoriness may be an indicator of low cardiac output. Patient characteristics such as concomitant inotropic therapy was most often an exclusion criteria for the previously discussed trials.
DR. MEHRA: So let me respond to that, Jo Ellen. First of all, I was taught to use a lot of metolazone, and frankly, as I matured clinically, I became less enamored by that drug as I realized how tough a drug it was on the kidneys and on electrolyte balance. There is this notion of what is the appropriate clinical endpoint, and I think you alluded to it. The endpoint needs to be clinically relevant. The endpoint should focus on the patient and not be one where the clinician feels better. Essentially we have been a bit swayed by how much urine output can be evoked and how much fluid can be removed. We have a large number of papers and data emerging that tell us that there may be a significant disconnect between the amount of fluid removal and clinical outcomes. As an example, studies like the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with Tolvaptan (EVEREST)7 with a vasopressin receptor antagonist showed effectiveness of the drug tolvaptan in improving hyponatremia, and in reducing weight in patients significantly, compared to placebo, but was no better than placebo in influencing 24 month outcomes from a cardiovascular endpoint. So we have to fundamentally ask the question, do any of these strategies actually make a real difference in patients?
The other issue is that we need to do a better job of discriminating among the distinct patient groups that present with diuretic resistance syndrome. So, for example, it would be clinically less useful to start using incremental diuretic strategies and multiple drug regimens in a patient who has a clear low output state causing diuretic resistance. In such a patient, perhaps, a strategy that improves cardiac output with or without the need for mechanical circulatory support might be the appropriate one to overcome diuretic resistance.
In another situation where you have acute pulmonary edema, for instance, in the setting of severe hypertension or acute ischemic syndromes, I think the treatment ought to be very different. In such situations, step number one from a clinical standpoint is to segregate the patients into classifications of why the diuretic resistance has occurred and what is the relationship of that diuretic resistance to pump function. As an example, most of the multiple drug use strategies that work do so in the setting where true renal causes of diuretic resistance have occurred. This is manifested pathologically by the development of tubular hypertrophy, which determines classical diuretic resistance. In those situations, even considering a “renal rest” such as with ultrafiltration for a few days in some series has shown to be of benefit. It is in those situations where using multiple agents may be beneficial from a fluid removal standpoint. You’re right, Jo Ellen, we have to ask ourselves to what end we are actually making patients better. And our definitions of how we make patients better are equally critical.
Tien, do you have some thoughts about that?
DR. NG: Yes, I have to agree. I think an issue has always been, what is the appropriate endpoint for our acute heart failure patients? This has been debated for years and years now, and when you look at the safety signals or the safety endpoints that are currently being used and the efficacy endpoints, we really don’t know whether these are the optimal safety and efficacy endpoints. In one sense we’re using urine output, and as you’ve pointed out there are clearly data showing that urine output isn’t necessarily the greatest indicator of an improved prognosis in these patients. At the same time, looking at serum creatinine changes as a measure of renal function obviously has its limitations as well. With serum creatinine one wouldn’t expect large changes in these relatively short-term trials, as it is a delayed marker of renal injury. So in the future, to look at the renal safety of these agents it will be important to incorporate some novel, more accurate and early markers of acute kidney injury. I think that will help us understand whether or not we are actually doing something beneficial for these patients or whether we are just treating ourselves.
DR. RODGERS: I completely agree, Dr. Ng. I think that many of our future trials are focusing on either the right population or the right endpoint. Some good examples, Dr. Mehra just mentioned ultrafiltration and the Cardiorenal Rescue Study in Acute Decompensated Heart Failure (CARRESS) trial,8 which is currently recruiting patients, and this study is specifically enrolling patients who are congested and fluid overloaded, but who have experienced renal insufficiency. The primary endpoint is a combination of weight and renal function changes. I am optimistic that any follow-up studies to CARRESS will assess endpoints such as mortality and rehospitalization. Other trials such as the Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in Diastolic Heart Failure (RELAX) trial9 with the agent relaxin, also are enrolling patients with mild to moderate renal insufficiency. This trial is designed to assess clinically meaningful endpoints such as worsening heart failure associated with longer length of stay, as well as 60-day outcomes. It is great that these trials are addressing how to manage the patients that clinicians most struggle with.
DR. NG: Would either of you like to comment on some of the novel pharmacologic approaches which are currently being evaluated? This includes low dose nesiritide compared to low dose dopamine, the Renal Optimization Strategies Evaluation in Acute Heart Failure (ROSE) trial,10 designer natriuretic peptides such as CDNP or CUNP. Thus far agents such as vasopressin antagonists and adenosine receptor antagonists haven’t been associated with positive results from a cardiorenal standpoint; however, I was wondering if you believe these agents may still have a potential role?
DR. RODGERS: The ROSE trial, again alludes to what we previously discussed regarding the addition of other agents in an attempt to minimize toxicity of diuretic therapy. This trial is assessing our ability to effectively manage the fluid overload without causing renal injury. Similar to RELAX,9 the ROSE trial is enrolling patients with impaired renal function as defined by a creatinine clearance between 15 and 60 mL/min. As Dr. Ng had suggested, they are utilizing a new marker for renal dysfunction in this trial. Their primary endpoint is change in cystatin C. Cystatin C is an excellent marker of kidney function because it is devoid of some of the limitations or inaccuracies of measuring serum creatinine, specifically it may more readily detect milder forms of renal impairment. Also, it is more reliable in patients with varying muscle mass or protein intake, which is commonly a problem in our heart failure patients who have cachexia due to low cardiac output or fluid overload.
DR. MEHRA: I think that the future is clearly in many of these trials that are being undertaken. I’m particularly optimistic about the combined use of diuretic infusions with low dose dopamine, for instance. That has been a strategy that’s been used in the past, but it’s very controversial. I understand that one of the components of the ROSE trial that is testing these various strategies includes low dose dopamine. It should be an interesting facet to investigate to see what is the role of an old drug that’s easily available and won’t cost us incremental dollars in a cost-restrained system.
I actually do think that agents need to continue to be investigated, but I have been very disappointed with the way the current field of agents looks. As you look at the vasopressin receptor antagonists, they’ve all come out fairly short in enhancing clinical outcomes. My suspicion is that there will be sub-populations that benefit and sub-populations that are harmed by these therapies. That is the uniform outcome in almost all situations where you have high morbidity and mortality, such as in acute decompensated heart failure. My personal feeling is that we need to do a better job of finding which populations are most likely to benefit and our attention ought to be focused there. I still think that there will be selective roles for these agents in the appropriately defined groups of patients.
The problem with any such study that looks either at surrogate biomarkers like cystatin C or looks at just patient derived global assessment scales or renal function, is the very small study size, often with a small number of patients in each group. It is therefore very difficult to pull out the groups of patients in whom harm occurred and distinguish them very clearly from the groups of patients in whom clear benefit was observed. I suspect that one of the studies from which we ought to get more information in this regard is the large Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure Trial (ASCEND), with more than 7000 patients.11 I think the value of that particular trial is probably confined to developing a very clear interface of investigations that will guide us to pick out those particular subgroups of patients in whom we should use targeted approaches for incremental diuretic or fluid removing strategies. What do you guys think?
DR. NG: I fully agree with both of your previous sentiments. Clearly the general sense is loop diuretics, regardless of how they’re administered or dosed, represents a suboptimal method to affect fluid removal from our volume overloaded patients. There remains hope that in the future new therapies that are more sparing from a pathophysiology standpoint will be developed so that we can effectively manage volume status in our patients while also having a positive effect on their outcomes.
So, yes, I hold hope that different combinations of pharmacologic agents as well as device therapies will be developed, and we’ll gain a greater appreciation of how to optimally select between these therapeutic options for individual patients.
Before concluding, I want to get your thoughts on one last issue. In the past we’ve approached diuresis as being a consequence of hypoperfusion. More recently, data focusing on venous congestion and how that influences diuretic responsiveness has emerged. My feeling is that as we understand the contribution of venous congestion to renal function in acute heart failure, new physiologic targets may be identified.
DR. MEHRA: Let me make a few comments about that. If you just think about the person who is most likely to develop diuretic resistance, it would be a person who has low blood pressure, chronic renal insufficiency, and very high right sided pressures. I think we’re learning now that the frequency of cardiorenal syndrome in patients with chronic heart failure is singularly linked to the severity of right heart function. And right heart failure is probably the closest correlate. Small observational studies have addressed the issue of intra-abdominal pressure as a surrogate marker for worsening renal function and improvements following release of intra-abdominal pressure, by removal of ascitic fluid or by decongesting these patients to reduce right atrial pressure, these have all shown some signals towards reduction in serum creatinine and recovery from cardiorenal syndrome.
So I think we are beginning to move in the direction that congestion is bad, that right-sided heart failure is probably the focus of our therapy, and that one surrogate for right heart failure other than the jugular venous pressure is intra-abdominal pressure, which can be very easily measured by placing a Foley in the bladders of these patients and attaching that to a pressure manometer.
DR. NG: Jo Ellen, did you have anything to add there?
DR. RODGERS: No.
DR. NG: Okay, in that case I would like to thank you all for your participation. To conclude I’ll provide a brief summary of what we’ve discussed. We began with a discussion of the role of continuous infusion and high-dose compared to low-dose loop diuretics as a method for overcoming clinical diuretic resistance. We all agree that, although the data are not airtight, they do suggest that there doesn’t seem to be large differences in terms of efficacy and in safety using high versus low or intermittent versus continuous infusion. The actual selection of the strategy should take into account the clinical status and characteristics of the individual patient.
We then moved to the issue of different diuretic strategies. We are hopeful that the addition of other agents which help augment overall diuresis will allow for a loop diuretic sparing effect, and may represent a better approach. Some of the current trials will help clarify that issue. We ended with just a very brief discussion on the need in future studies to identify optimal patients and optimal endpoints, as well as looking at the overall model of congestion, right-sided failure, and its contribution to the cardiorenal syndrome that we manage daily in our acute heart failure patients.
FoxP2 Media LLC is the publisher of The Medical Roundtable.