|Home | About | Journals | Submit | Contact Us | Français|
Strategies to preserve renal function and enhance diuretic responsiveness during therapy for heart failure (HF) are needed. We hypothesized that brain natriuretic peptide (nesiritide), added to standard HF therapy would preserve renal function and enhance diuretic responsiveness.
HF patients with underlying renal dysfunction who were admitted with volume overload were randomized to standard therapy without or with nesiritide (2mcg/kg bolus; 0.01 mcg/kg/min for 48 hours). Patients requiring intravenous vasodilator or inotropic therapy for rapid symptom relief were ineligible. In all patients, diuretics were administered according to a standardized dosing algorithm.
Patients (n=72) had mean creatinine of 1.75±0.59 mg/dl. Nesiritide patients had less increase in creatinine (p=0.048) and blood urea nitrogen (p=0.02), but greater blood pressure reduction (p<0.01). Nesiritide did not enhance diuretic responsiveness (p=0.57) but increased 3’5’ cyclic guanosine monophosphate and decreased endothelin more (p<0.05 for both). There were no differences in the change in atrial natriuretic peptide, N-terminal proBNP, plasma renin activity, angiotensin II, and aldosterone between groups.
When used as adjuvant “renal protective” therapy in HF patients with renal dysfunction, recommended dose nesiritide reduced blood pressure, did not appear to worsen renal function, and suppressed endothelin but did not enhance diuretic responsiveness nor prevent activation of the renin angiotensin aldosterone system.
Worsening renal function (WRF) during treatment for heart failure (HF) is common, frequently associated with diuretic resistance and potently associated with adverse outcomes.(1-5) Strategies to preserve renal function and enhance diuretic responsiveness during therapy for HF are needed.(6)
Nesiritide is recombinant human brain natriuretic peptide (BNP) and is approved for treatment of HF based on its effects to reduce preload and improve dyspnea.(7) While preclinical studies suggested that natriuretic peptides have favorable renal effects in normal animals or humans or in experimental HF(8-12), the potential for deterioration in renal function with variable doses of nesiritide in human HF was previously recognized(7, 9) and recently emphasized.(13) However, use of nesiritide specifically and solely to enhance renal function in HF has not been well studied. Further, while natriuretic peptides can suppress endothelin and the renin-angiotensin-aldosterone system,(9) the effect of nesiritide on counter-regulatory hormones when used in conjunction with diuretic treatment in patients with HF is unclear.
Our primary objective was to determine whether adjuvant therapy with recommended dose nesiritide would improve renal function and enhance diuretic responsiveness in HF patients with renal dysfunction who were admitted for treatment of volume overload but who did not require intravenous vasodilator or inotropic therapy for rapid relief of dyspnea. We also sought to determine its effects on hormone levels, length of hospital stay, and 30-day HF readmission rates.
The trial (Number: NCT00170183, URL: http://www.clinicaltrials.gov) was conducted at a Mayo Clinic affiliated hospital in Rochester, Minnesota. The Mayo Foundation Institutional Review Board (IRB) approved the study, and all patients gave written informed consent. Enrolment started on May 27, 2003, and the last patient was enrolled February 21, 2006. Scios® provided funding for but was not directly involved with the study. The investigators had control of all data at all times, performed all data analysis, and were primarily responsible for publishing the findings.
Consecutive patients with new onset or pre-existing HF age 18 years or older admitted to the Mayo Advanced Heart Failure Progressive Care Unit were considered eligible and were enrolled on admission. Inclusion criteria included a clinical diagnosis of class III-IV HF requiring hospitalization for management of volume overload, an estimated glomerular filtration rate (GFR) < 60 but ≥ 20 ml/min, systolic blood pressure > 90 mmHg, stable rhythm, and no plans for cardiac catheterization. Exclusion criteria included inability to provide informed consent, potential pregnancy, new onset atrial fibrillation with rate > 110 bpm, active ischemia, known or suspected stenotic valve disease, clinical need for intravenous vasodilator or inotropic therapy for rapid symptom relief, or primary reason for admission other than HF.
After consent, patients were randomized to standard therapy or standard therapy plus nesiritide. Randomization was stratified according to severity of renal dysfunction present on admission as assessed by the estimated GFR. Patients were classified as having mild renal dysfunction (GFR = 40-59cc/min/1.73m2) or moderate renal dysfunction (GFR = 20-39cc/min/1.73m2). Both randomization to nesiritide vs standard therapy and the diuretic algorithm (Figure 1) used were based on this stratification. Neither the patients nor the investigators were blinded to treatment.
The primary endpoints were changes in renal function (change in creatinine and BUN) and diuretic responsiveness (change in weight and total fluid balance) over time (measured at 24, 48, and 72 hours and/or at discharge). Secondary endpoints were changes in cystatin C and hormone levels from admission to 48 hours, length of hospital stay, and 30-day readmission for HF.
All patients received standard HF therapy at the discretion of the attending cardiologist with the exception of diuretic therapy, which was standardized for all patients based on renal function and adjusted according to an algorithm to achieve a daily weight loss of ≥ 1.5 kg and/or a daily total fluid loss of ≥ 1500 cc (Figure 1). All other diuretics were held except for spironolactone, which was continued but not initiated. After baseline labs, intravenous nesiritide (bolus of 0.2 mcg/kg followed by 0.01 mcg/kg/min) was initiated in the nesiritide group, and the first diuretic dose was initiated subsequently. Hormones and cystatin C were measured on admission and at 48 hours (just prior to discontinuation of nesiritide in the nesiritide group). Patients were restricted to 2 g dietary sodium and fluid intake of 1.5 liters per day. Weights were obtained at the same time each day on the same scale for the first 72 hours of therapy and again at discharge. Intake and output, blood pressure, and heart rate were recorded daily for the first 72 hours of therapy under supervision of the study coordinator. Nesiritide infusion was held and resumed for transient asymptomatic hypotension and discontinued for symptomatic hypotension. Daily electrolyte measurements were obtained. Measurement of BNP was made with both the Biosite® Triage® and Shionogi® assays.(14) N-terminal proBNP (NT-proBNP), atrial natriuretic peptide (ANP), plasma renin activity (PRA), aldosterone, endothelin, angiotensin II, and 3’5’cyclic guanosine monophosphate (cGMP) were measured as previously described.(15-18) GFR was estimated using the simplified modification of diet in renal disease.(19) Cystatin C was measured with a latex-enhanced immuno-nephelometric assay.(20)
Patients were contacted 30 days after discharge. Hospital records were reviewed to determine reason for readmissions.
Details regarding the original statistical design can be viewed at URL: http://www.clinicaltrials.gov. Baseline group comparisons were made with the Student’s t-test or the chi square test or Fisher’s exact test as appropriate. The paired Student’s t-test was used for within-group comparisons. Nonparametric tests were used to examine differences between (nonparametric Kruskal-Wallis test) and within (nonparametric matched pair Wilcoxon Signed Rank test) groups in hormone levels as the distribution of these variables were skewed.
As renal function and hemodynamic parameters were assessed at multiple time points and length of stay varied, data was not complete at all study time points. Thus, we used the Generalized Estimated Equation (GEE) Model(21) which accounts for missing data when comparing changes between groups over time. Similar to a two-factor repeated measures ANOVA test, the GEE determines whether there is an interaction between treatment assignment and time and whether there are time or group differences. If the interaction term is significant, unpaired t-tests must be used to compare data between groups at different times.
Data are reported as mean ± standard deviation. All calculated p-values were two-sided, and p-values < 0.05 were considered statistically significant. All analysis was by intention to treat.
Due to changes in the hospital service structure, fewer patients with HF were admitted to the Advanced HF Unit and enrollment was lower than anticipated despite continuing enrollment for almost twice the anticipated study duration. The decision to terminate enrollment was made solely by the investigators.
The study size was estimated based on a retrospective sample of 60 patients with HF at our institution where the mean change in creatinine at 48 hours was 0.02±0.38 mg/dl. Assuming this variability in creatinine change, a sample size of 104 (52 per group) yielded a power of 80% to detect a difference in creatinine change of 0.21 mg/dl between groups. Recruitment goals were not met, but the standard deviation of the creatinine change in the two study groups was significantly less than that observed in the retrospective sample. Thus, post hoc analysis of statistical power based on a two-sided, two-sample t-test with a type-I error level of 5% and using the observed standard deviation for the creatinine change in the standard group at 48 hours (0.18 mg/dl), the study had 80% power to detect a difference in mean creatinine change between the two groups of 0.12 mg/dl.
Seventy-five patients consented to the study and were randomized. Three patients randomized to standard therapy withdrew from the study prior to initiation of therapy at the request of the attending physician (n = 2) or the investigator (patient did not meet entry criteria; n=1). No data were collected on these three patients. One of these patients crossed over to nesiritide at the discretion of the attending physician.
Patients were randomized an average of 4.57±4.61 hours after admission in the nesiritide group and 4.92±5.54 hours after admission in the standard therapy group (p=0.77). Patients who received nesiritide had slightly higher sodium and were less often being treated with spironolactone on admission (Table 1). Patients in the nesiritide group tended to be older and more frequently male and thus, as expected, tended to have higher creatinine. However, GFR was similar between groups.
The change in creatinine was lower in the nesiritide group, although this difference was of nominal significance (Table 2 and Figure 2). The change in BUN was lower in the nesiritide group. However, the change in cystatin C from baseline to 48 hours was similar between groups (Table 3 and Figure 2). The incidence of worsening renal function (increase in Cr > 0.3 mg/dl at 24, 48, 72 hours or at discharge) was 33% in the nesiritide group and 28% in the standard therapy group (p=0.78). When defined as an increase of Cr of > 0.5 mg/dl, worsening renal function occurred in 2 patients (7%) in the nesiritide group and 3 patients (11%) in the standard therapy group (p=0.64).
The cumulative weight loss and the cumulative fluid loss in patients on standard therapy tended to be greater than in nesiritide-treated patients, but this difference was not significant (Figure 2 and Table 2). The cumulative furosemide dose was similar between groups. Fourteen patients were on therapy with spironolactone, 2 in the nesiritide group and 12 in the standard therapy group. Confining the analysis to those patients not on spironolactone (n=58), the weight change at 24 hours (-1.8 ± 1.8 (nesiritide) vs -2.3 ± 2.3 (standard therapy) Kg, p=0.338) and over the entire hospitalization (-4.1 ± 3.0 (nesiritide) vs -6.5 ± 5.1 (standard therapy) Kg, p=0.053) tended to be greater in the standard therapy group, consistent with the overall findings.
In a post hoc analysis, the relationship between volume removal (weight loss) and changes in Cr or BUN at 24, 48, and 72 hours was not strong (Figure 3) but showed a trend for more favorable changes in Cr and BUN in patients with greater diuresis at 48 and 72 hours. Adjusting for weight loss, patients treated with BNP had more of a decline in Cr and BUN at 24 hours, consistent with the primary analysis.
Patients treated with nesiritide had a greater decline in systolic blood pressure at 24 hours than standard therapy patients (Table 2, Figure 2), but there were no other significant differences in change in blood pressure at other time points and no difference in change in heart rate at any time point (data not shown). Six patients on nesiritide and two patients on standard therapy (16% vs 6%, p = 0.20) developed hypotension (systolic blood pressure < 90 mmHg).
There was no difference in the baseline levels of any hormone between the two groups (Table 3). After 48 hours of nesiritide infusion, BNP and cGMP increased in the nesiritide group and declined in the standard therapy group. Endothelin declined in the nesiritide group and showed an insignificant increase in the standard therapy group with a significant difference in the mean change in endothelin between groups. Levels of ANP and NT-proBNP declined to a similar degree in both groups. Plasma renin activity increased in the nesiritide group and showed an insignificant increase in the standard therapy group with no difference in the mean change between groups. There was no significant change in angiotensin II or aldosterone levels within the two groups, and the mean change in both was similar between groups.
Length of stay was similar in the nesiritide (5.1±3.7 days) and standard therapy (5.0±3.9 days, p=0.98) groups. There were 16 readmissions within 30 days of discharge in 13 patients. All-cause 30-day readmissions were similar in the standard therapy (n=10) and nesiritide (n=6) groups. Thirty-day HF readmissions were similar between groups (two HF readmissions in the nesiritide group and one in the standard therapy group).
This study was unique in that it investigated the use of nesiritide specifically and solely to improve renal function during the treatment of volume overload in patients with HF and renal dysfunction. As compared to standard therapy, nesiritide treated patients displayed small improvements in creatinine and BUN despite greater blood pressure reduction. However, these small relative changes were of only borderline statistical significance and the findings may best be summarized as supporting the absence of an adverse renal effect in this patient population. Importantly, this nesiritide dose did not enhance the response to a bolus diuretic dosing strategy, which was stratified according to renal function and titrated by prespecified criteria to obtain adequate volume reduction. Indeed, a non-significant trend towards less diuresis was observed in the nesiritide group. As compared to standard therapy alone, at 48 hours of therapy, adjuvant nesiritide augmented BNP and cGMP levels and suppressed endothelin but did not suppress the renin-angiotensin-aldosterone system and did not produce greater preload reduction as assessed by effects on endogenous natriuretic peptide levels (change in ANP and NT-proBNP).
Natriuretic peptides may influence renal function via direct effects on glomerular arterioles, mesangial cells, and renal tubules as well as by indirect effects on renal sympathetic nerve activity, systemic hemodynamics, and humoral function.(8-12) Studies which have characterized the effect of natriuretic peptides on renal function were diverse in terms of dose, methodology, and study population. Nonetheless, these previous studies suggested that natriuretic peptide infusion had the potential to enhance or maintain renal function despite reduction in blood pressure, suppressed renin, aldosterone and endothelin, and improved cardiac output, variably depending on peptide, dose, and subject characteristics.(8-12) While these studies were often performed in normal animals or humans or in experimental HF without underlying renal dysfunction, they engendered expectation that nesiritide therapy would have beneficial effects on renal function in patients with HF. However, clinical studies leading to Food and Drug Administration (FDA) labeling of nesiritide as a vasoactive treatment for HF focused on the hemodynamic effects of nesiritide, did not extensively address effects on renal function and included higher doses than currently recommended.(22, 23) Recommended dose nesiritide did not improve renal function in a small but elegant study of patients who had already developed WRF during treatment of HF.(24) In the current study, adjuvant therapy with recommended dose nesiritide initiated on admission had no adverse effect on changes in creatinine or cystatin C despite its effect to lower blood pressure more than standard therapy. This effect did not appear related to the fact that nesiritide patients tended to have less of a diuresis. While there was no significant difference in changes in cystatin C between groups, this parameter was only assessed at one time point. As animal studies have demonstrated that the renal effects of natriuretic peptides are sensitive to renal perfusion pressure,(25) excessive vasodilatation can limit renal effects and, as observed here, can also activate renin, which may further oppose renal protective effects.
To date, clinical studies of nesiritide have focused on symptom relief or hemodynamic endpoints and established a dose related adverse effect of nesiritide on renal function in acute decompensated HF. (7, 9, 13) However, in a case control study, Riter et al. recently reported that much lower, non-hypotensive nesiritide doses were associated with improved renal function and enhanced diuretic responsiveness in patients with HF and severe renal dysfunction.(26) Clinical studies exploring intra-renal nesiritide administration with novel catheter delivery systems to maximize renal effects while limiting hemodynamic effects in HF are ongoing. Such studies are supported by a recent study (NAPA Trial) suggesting renal protective effects of nesiritide in cardiac surgery(27) where the recommended infusion dose was given without a bolus. We speculate that use of nesiritide without a bolus may have minimized detrimental hypotensive effects and enhanced the renal protective effects in the NAPA trial. While the current study provides reassurance that the recommended bolus and infusion dose of nesiritide does not worsen renal function, it suggests that the dose currently recommended for hemodynamic efficacy may not be the ideal dose to produce sustained benefit on renal function.
While natriuretic peptide infusion in normal animals is potently diuretic and natriuretic, these effects are blunted in animals with experimental HF(28, 29). While, a previous study in experimental HF indicated that a lower dose of nesiritide augmented the diuretic response to continuous furosemide infusion.(16), most studies in human HF have not demonstrated significant diuretic effects with nesiritide.(24, 30, 31) Here, we did not observe enhanced diuretic responsiveness with recommended dose nesiritide and bolus furosemide dosing in patients with HF and concomitant renal dysfunction. Indeed, some trend in the opposite direction was observed. As recently described, natriuretic peptides have potent effects to enhance vascular permeability.(32) Such effects may contribute to the preload reducing effects of nesiritide and may limit diuresis. Further, the reduction in blood pressure, and thus renal perfusion pressure associated with this dose of nesiritide, may also limit the diuretic effects of both furosemide(33) and nesiritide as mentioned above. Based on the potent hypotensive effect of this dose of nesiritide and the preliminary findings with lower dose nesiritide mentioned above,(16, 26, 27) we speculate that alternate natriuretic peptide/furosemide dosing strategies may optimize diuresis while minimizing renal dysfunction in HF.
Natriuretic peptides have well-described effects to suppress endothelin and the renin-angiotensin-aldosterone system as well as catecholamine activation.(9) However, circulating levels of these counter-regulatory hormones are also influenced by the HF state, hemodynamic factors, and by other medications. Thus, favorable effects on hormone levels may not be consistently observed with nesiritide therapy in the clinical setting. As expected, BNP and cGMP levels increased in the nesiritide but not in the standard therapy group. Endogenous natriuretic peptide production decreased to a similar degree at 48 hours in both groups suggesting equivalent preload reduction as discussed above. Effects at 24 hours (when blood pressures differed between groups) were not assessed. We speculate that the mechanism of preload reduction may have differed between groups, with a trend towards more diuresis in the standard therapy and more of an effect on venous tone or vascular permeability in the nesiritide group as mentioned above. A decrease in endothelin levels was apparent only in the nesiritide group consistent with previous studies. However, we did not observe suppression of the renin-angiotensin-aldosterone system with nesiritide; indeed, plasma renin activity increased significantly at 48 hours in the nesiritide group but not in the standard therapy group, although the mean change in plasma renin activity was similar in the two groups. A higher dose of nesiritide decreased aldosterone levels at six hours in HF in a previous study, which prohibited concomitant diuretic therapy(23) and a lower dose inhibited aldosterone activation associated with continuous furosemide infusion in experimental HF,(16) but effects of nesiritide at recommended dose on humoral activation in the clinical setting have not been well described. Bolus diuretic dosing is a potent activator of the renin-angiotensin-aldosterone system, and while angiotensin antagonism can attenuate such activation in controlled settings,(34) complete suppression may not be apparent in clinical settings.
Due to the plethora of factors which influence dismissal planning, large numbers of patients are needed to test differences in length of stay with HF therapy.(35) Nonetheless, we saw no favorable trends in length of stay in the nesiritide group. Further, we saw no benefit in diuretic responsiveness, which might have translated into shorter length of stay in a larger study. A recent report suggested the potential for shorter length of stay when nesiritide therapy was started in the emergency department in patients needing intravenous therapy for symptom relief.(36) Here, the rate of HF readmissions at 30 days was low in both groups and may reflect the adequacy of diuretic dosing, which was adjusted according to renal function and titrated to reach a minimum daily fluid removal goal.
Inability to reach the target sample size exposed the study findings to a type-2 error, although post hoc analysis suggested adequate power to detect a clinically relevant difference in creatinine change. This is a small study with multiple endpoints and cannot be considered adequate to support use of nesiritide to enhance renal function in acute HF. We did not directly measure sodium excretion, GFR, or renal blood flow. Our study was conducted at a single tertiary institution and findings may not generalize to the community, more ethnically diverse populations, or to patients not managed by HF specialists. Patients with severe renal dysfunction and those who required intravenous vasodilator or inotropic therapy for rapid symptom relief were excluded, and effects of nesiritide on renal parameters may be different in such patients. Patients and physicians were not blinded to treatment and this could have influenced the study findings. However, a standardized diuretic algorithm was used, and the primary endpoints were objective. More patients in the standard therapy arm were treated with spironolactone, which could have enhanced diuretic responsiveness. However, when excluding patients on spironolactone, nesiritide therapy still did not enhance diuretic responsiveness as compared to standard therapy. We did not assess daily doses of other medications which may have affected renal function. Further, we did not assess renal function beyond discharge.
When administered solely to preserve renal function during the treatment of volume overload in patients with HF and renal dysfunction, nesiritide produced greater reduction in blood pressure and did not worsen renal dysfunction. Rather there were small improvements in creatinine and BUN of only borderline statistical significance. Importantly, this nesiritide dose did not enhance the response to a bolus diuretic dosing strategy, which was stratified according to renal function and titrated by prespecified criteria to obtain adequate volume reduction. Indeed, a non-significant trend towards less diuresis was observed in the nesiritide group. Recognizing the favorable effects of lower doses on renal function as demonstrated in recent studies, we speculate that alternate natriuretic peptide and/or diuretic dosing regimens or routes of administration may yet hold promise as renal protective therapy in HF. Further studies will be needed to test this hypothesis and define the role of this or other natriuretic peptides as renal enhancing therapy in HF.
The authors wish to thank Damita J. Carryer for her role as the study coordinator for this trial.
Grant Support: This study was funded in part by a grant from Scios ®. Funding for Dr. Owan (T32-HL07111) and Drs. Burnett, Chen, and Redfield (P01 HL 76611) is provided in part by the National Institutes of Health.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.