#NephJC Chat
Tuesday, March 22nd, 2022 at 9 pm Eastern Standard Time
Wednesday, March 23rd, 2022 at 9 pm Indian Standard Time and 3:30 pm GMT
Eur Heart J. 2021 Nov 14;42(43):4468-4477
DOI: 10.1093/eurheartj/ehab620.
Compensatory post-diuretic renal sodium reabsorption is not a dominant mechanism of diuretic resistance in acute heart failure
Zachary L Cox, Veena S Rao, Juan B Ivey-Miranda, Julieta Moreno-Villagomez, Devin Mahoney, Piotr Ponikowski, Jan Biegus, Jeffrey M Turner, Christopher Maulion, Lavanya Bellumkonda, Jennifer L Asher, Helen Parise, Perry F Wilson, David H Ellison, Christopher S Wilcox, Jeffrey M Testani
PMID: 34529781
Introduction
We have all seen diuretic resistance on hospital rounds. Everyone has their own strategy, and frequently it is some combination of increasing the loop diuretic, adding a thiazide, and fluid/salt restriction (though not everyone agrees which approach is optimal!). On day 5 of the hospital stay, the patient still hasn't lost any weight. You turn to your medical students and give a quick teaching session about how the patient is up against compensatory post-diuretic sodium reabsorption, just as you were taught by your mentors in the past. But is your teaching about heart failure management correct, or are you just repeating medical myths?
Acute decompensated heart failure (ADHF) is the second most common hospital discharge diagnosis (infection is first), with over 64 million people affected by heart failure worldwide (Groenewegen et al., European Journal of Heart Failure, 2020). We have written about the problem of diuretic resistance in these patients before; check out this great Renal Fellow Network blog by Carlo Trinidad, as well as the cardiorenal region scouting report of this year’s NephMadness. The current understanding of diuretic resistance (DR) in ADHF is extrapolated from healthy subjects and patients with hypertension and chronic kidney disease, with various mechanisms being explored (Cox and Testani, Cardiorenal Syndrome In Heart Failure, 2020), as per Figure 1 below.
Figure 1: Diuretic Resistance Mechanisms in Heart Failure (courtesy of NSMC intern Husam Alzayer).
However, there are very few studies of diuretic resistance mechanisms specifically in patients with heart failure. The Mechanisms of Diuretic Resistance (MDR) study, established a prospective, observational cohort and biorepository of patients hospitalized with ADHF and accompanying DR (Cox et al., ESC Heart Failure, 2020). This specific cohort has the potential to explore research questions from the diagnosis of diuretic resistance to the post-discharge outcomes of different diuretic therapies. It is from this cohort that today’s study question stems: does compensatory post-diuretic sodium reabsorption (CPDSR) play a major role in diuretic resistance in acute heart failure?
But before we begin, what exactly is CPDSR? The theory goes that after loop diuretic induced natriuresis, the kidney creates a period of sodium reabsorption, which alleviates volume loss and produces a near-net-neutral sodium balance. So, even with an adequate initial diuretic response (i.e., large volume urine output), the patient fails to achieve negative fluid balance due to this subsequent sodium reabsorption. In Figure 2 below, you can see that with loop diuretic administration, the patient excretes an increased amount of sodium in the urine. With CPDSR in healthy individuals, the kidney then decreases natriuresis to compensate for the amount of sodium lost. Tonight’s article dives into how patients with heart failure react to high-dose diuretics and how CPDSR applies to their specific pathophysiology.
Figure 2: Concept of compensatory post-diuretic sodium reabsorption. (adapted from Figure 1, Cox et al., Eur Heart J 2021)
The Study
Patients in this study were originally screened for inclusion in the MDR study (designed to investigate mechanisms of diuretic resistance in a prospective cohort of patients hospitalized for ADHF). This study design was an a priori study question (Latin for “from the former”): meaning this study is relating to reasoning or knowledge from theoretical deduction based on the previous MDR study.
Methods
A single-center, prospective, observational cohort of patients hospitalized with ADHF within the Yale-New Haven Hospital system (USA). This cohort followed a strict sodium-controlled diet (3g/day). IV diuretic dosing was dictated by the treating physician. Enrollment was allowed at any time during their hospital stay.
Inclusion criteria
Use of intermittent IV loop diuretic therapy with a projected need of IV diuretics for at least 3 days
A goal of significant fluid removal (> 1L net fluid loss/day)
At least one objective sign of volume overload
Exclusion criteria
Bladder dysfunction
Urinary incontinence
Inability to comply with the urine collection procedure
Received a thiazide within previous 24 hours or did not receive a loop before urine collection
Study visit protocol
Study visits are repeated during the hospital stay while on IV diuretic therapy. A single study visit is considered a 24h period.
Study cohorts
Two groups: healthy volunteers AND patients with ADHF
Two assessments: diuretic-induced natriuresis using the six-hour cumulative urine collection AND post-diuretic spontaneous natriuresis using the 18-hour cumulative urine collection or a spot urine collection
Healthy volunteers
The purpose of this group was to replicate previous CPDSR studies and establish an “expected” post-diuretic natriuresis measurement to compare to the heart failure group. Twenty healthy, euvolemic participants with no history of loop diuretic therapy or heart failure were recruited. Each participant received a single dose of furosemide 40mg IV. These participants were able to have clear, sodium-free liquids and low-sodium snacks during the first six hours of the study visit. During the next 18 hours (post-diuretic time period), they were allowed to eat salty foods and drink water to replenish volume status as needed.
Patients with acute heart failure
The MDR study cohort included a total of 462 total diuretic administrations from 285 unique patients (remember, some patients received more than one dose of diuretic during their hospital stay, so there may be more doses than patients). These patients were further subdivided into three groups (see Figure 3 below):
Measured Cohort
Calculated Cohort
Randomized Intervention Cohort
Measured Cohort:
This cohort (94 unique patients with a total of 117 diuretic administrations) served as the primary cohort, as these patients only received one dose of diuretics in a 24 hour period. This cohort was considered less generalizable as most hospitalized patients with ADHF typically receive more than one diuretic dose in a day.
Calculated Cohort:
This cohort served as the validation cohort to the Measured Cohort. In this group, 75% of patients received more than one diuretic dose during a study visit (i.e., at least one additional dose of diuretic during hours 6 through 24 of a single hospital day). 130 patients were included once (28%), 284 patients were included twice (62%), and 48 patients were included three times (10%) in this cohort. Patients supplied a random urine sample before the second diuretic dose (this dose is any time after hour 6 but before the next study day’s hour 0). 18-hour cumulative sodium excretion was calculated from this random urine sample using a validated equation.
The Calculated Cohort lacks any “carryover effect” of diuretic-induced natriuresis during the post-diuretic phase (i.e., hours 6-24). Instead of separate six-hour and 18-hour urine collections (i.e., the Measured Cohort), this cohort had a spot urine sodium level measured about 15 hours after the last diuretic dose, rather than a new post-diuretic six-hour timed collection. This cohort is considered more generalizable than the Measured Cohort, as hospitalized patients with ADHF typically receive more than one dose of diuretics in a 24 hour period.
Randomized Intervention Cohort:
This cohort (43 unique patients) measured the intra-individual change in post-diuretic spontaneous natriuresis after a randomized intervention to increase diuretic-induced natriuresis. Patients in this cohort had a poor diuretic response in the first study visit (defined by a six-hour cumulative sodium excretion < 100 mmol). These patients were randomized into a controlled study the next day to investigate different diuretic strategies to help increase natriuresis:
2.5x the previous IV loop diuretic dose
Same IV loop diuretic dose plus IV chlorthalidone
This cohort provided insight into post-diuretic natriuresis by observing how diuretics affect CPDSR and sodium avidity, regardless of diuretic response.
Figure 3: Patient cohort flow diagram. The number of diuretic administrations, unique patients, and purpose for each patient cohort are shown in a flow diagram. (Figure 2 from Cox et al., Eur Heart J 2021)
Measurement of natriuresis
To determine the diuretic-induced expected natriuresis, ADHF patients’ cumulative natriuresis was compared to that of the healthy volunteers' average cumulative natriuresis.
To determine the post-diuretic expected spontaneous natriuresis, the following equation below was used. Values are based on the typical daily sodium intake of 130 mmol with approximately 95% renal excretion of 124 mmol in a 24 hour period. The expected renal sodium excretion of the daily 124 mmol in 18 hours is 93 mmol.
The 24-hour sodium balance is the difference between 124 mmol and the collective six-hour and 18-hour sodium output.
Statistical analysis
Linear mixed models were used to analyze associations of continuous variables and for comparisons between loop diuretic administrations (considered at repeated observations). The 18-hour natriuresis was compared to the expected mean of 93 mmol using a one-sample t-test. Paired t-test was used in the Randomized Intervention Cohort comparing the 18-hour natriuresis of the first visit and the intervention visit. Statistical significance was defined as a 2-tailed P < 0.05. A detailed description of the statistical analysis is found in the supplemental material online.
Results
To confirm the reproducibility of the CPDSR theory, a healthy volunteer cohort consisting of 20 volunteers with a mean age of 26 (90% white, 50% female) was recruited. This cohort had a basal pre-diuretic 24-hour natriuresis of 149 ± 65, a 6-hour post-diuretic of 199 ± 49, and an 18-hour spontaneous natriuresis of 50 ± 23. These findings replicate the previously established compensatory response in healthy individuals and indicate that the CPDSR effect is present in the post-diuretic period in healthy individuals.
The measured cohort in this study consisted of 94 patients with ADHF with a mean age of 65 (62% white, 61% male). Mean eGFR was 56, median home furosemide equivalent dose 80 mg, and 20% were already receiving spironolactone. They were subdivided into those with a positive 24-hour sodium balance (n = 33) and a negative 24-hour sodium balance (n = 61) as shown below in Table 1.
Table 1 from Cox et al., Eur Heart J 2021
Comparing several characteristics between the positive and negative sodium balance patients, there were no significant differences based upon EF, eGFR, or serum markers such as creatinine, albumin, NT-proBNP, among others. Notably, systolic blood pressure and BMI were significantly different between the groups, where those with a negative sodium balance had greater BMI and systolic blood pressure, which could be related to the level of RAAS activity, some other yet to be elucidated mechanism, or by chance.
Patients that exhibited diuretic resistance, defined as 6-hour post-diuretic natriuresis of less than 100 mmol, were then randomized in the “intervention” cohort with a goal of allowing evaluation of intra-individual response to increased diuretics. This cohort consisted of 43 individuals (67% male, 74% white) that were either randomized to receive 2.5 times the previous diuretic dose (n = 18) or loop diuretic plus chlorothiazide (n = 25).
Both interventions, increased dose of loop or addition of a thiazide, were successful at increasing the 6-hour post-diuretic natriuresis, but the 18-hour spontaneous diuresis remained unchanged, contrary to the decreased natriuresis noted in the healthy volunteer cohort during this window. There was no head-to-head comparison between the response to the addition of thiazide versus increase in loop diuretic dose. Patients with net negative 24-hour sodium balance had both greater diuretic-induced and spontaneous natriuresis, while those with positive sodium balance had both poorer post-diuretic and spontaneous natriuresis. These findings point towards an overall sodium avid state in some patients that is likely contributory to their volume overload, and that CPDSR is not a strong contributor to the post-diuretic response. Thus, further investigation into identifying which patients will have this overall sodium avid state and how to overcome this presents a clinically important question in the management of patients with ADHF
Discussion
The study design elegantly highlights how the CPDSR theory, although widely accepted and present in healthy adults, does not generalize to patients that are volume-overloaded in ADHF in this mechanistic study setup. Rather than a compensatory response, there seems to be an underlying basal sodium avidity in patients with heart failure. The authors argue that the strongest evidence to support this viewpoint is the response noted in the intervention cohort in which there was a positive correlation between diuretic-induced natriuresis and spontaneous natriuresis in the following 18 hours (contrary to CPDSR, which would have a negative correlation between the two).
The intervention group of this study took things a step further to show evidence of success in two of the popular mechanisms of overcoming diuretic resistance highlighted in this year’s NephMadness cardiorenal region scouting report, written by Caitlyn Vlasschaert: sequential nephron blockade (through the addition of a thiazide) and maximizing loop efficiency (by increasing the dosage). This same post highlights some other potential medications to modulate sodium affinity at other parts of the nephron including acetazolamide, mineralocorticoid receptor antagonists (although several patients in this study were already on MRAs), and everyone’s favorite: SGLT2-inhibitors. If we were able to more readily identify patients that will have this underlying sodium avidity, we could more quickly reach into our armamentarium of nephron-modulating medications.
Limitations
The gold standard of sodium balance is holding patients in a specialized metabolic ward with a controlled diet consisting of a known sodium load; however, in practice, this is, at best, time-intensive and at worst, unsafe for patients that are acutely volume overloaded. This seems to be a minor concern in the context of this study because it is unlikely that more meticulous tracking of sodium intake would have a meaningful effect on CPDSR. The time window for which the study evaluates CPDSR is 18-hours post-diuretic administration, and thus it is also possible that there could be a response that takes place outside of this window; however, in practice, it is unlikely that this patient population would go longer than that without receiving another diuretic dose, and thus does not seem to be a clinically relevant concern. The authors also highlight the complementary nature of the measured and calculated cohorts and how potential biases in each cohort are addressed by the other.
Other limitations include that the study population consists of patients from a single site that tended to have better kidney function (mean eGFR 56 ±19) than many of the patients with ADHF we may treat. They were also all hospitalized, so it is unclear whether this could be generalized to patients with chronic HF who are being managed outpatient. There was also no description of what exact objective measures of volume were utilized for patient inclusion and no data collection on the fluid intake of patients. Finally, the patients in this study received 3 grams of sodium per day, which is higher than the 2g/d that is commonly used in patients with heart failure.
Conclusion
Overall, this work seems to be strong evidence that we should question our current theories of diuretic resistance mechanisms in patients with ADHF. This was an exciting study as it indicates our classical teaching on CPDSR is incorrect, but it raises more questions going forwards regarding how we identify and treat patients who have this increased sodium avidity.