- Email: [email protected]
Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model
Journal of Veterinary Cardiology (2018)
-, -e-
www.elsevier.com/locate/jvc
Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model D. Adin, DVM*, C. Atkins, DVM , M.G. Papich, DVM College of Veterinary Medicine, North Carolina State University, 1060 William Moore Dr., Raleigh, NC 27607, USA Received 20 July 2017; received in revised form 15 January 2018; accepted 22 January 2018
KEYWORDS Lasix; Urine; Natriuresis; Aldosterone
Abstract Introduction: Diuretic failure is a potential life-ending event but is unpredictable and poorly understood. The objectives of this study were to evaluate pharmacodynamic markers of furosemide-induced diuresis and to investigate mechanisms of diuretic braking in dogs receiving constant rate infusion (CRI) of furosemide. Animals: Six healthy male dogs. Methods: Raw data and stored samples from one arm of a previously published study were further analyzed to mechanistically investigate causes of diuretic braking in these dogs. Urine volume was recorded hourly during a 5-h furosemide CRI. Urine and blood samples were collected hourly to measure serum and urine electrolytes, urine aldosterone, and plasma and urine furosemide. Serum electrolyte fractional excretion was calculated. Urine sodium concentration was indexed to urine potassium (uNa:uK) and urine furosemide (uNa:uFur) concentrations, plasma furosemide concentration was indexed to urine furosemide concentration (pFur:uFur), and urine aldosterone was indexed to urine creatinine (UAldo:C). Temporal change and the relationship to urine volume were evaluated for these measured and calculated variables. Results: Urine volume was significantly correlated with urine electrolyte amounts and with uNa:uK. The ratio of pFur:uFur decreased during the infusion, whereas furosemide excretion was unchanged. Conclusions: There was a strong relationship between urine volume and absolute urine electrolyte excretion. Urine volume was strongly correlated to uNa:uK, giving it potential as a spot indicator of urine production during diuresis. The decrease in
* Corresponding author. E-mail address: [email protected] (D. Adin). https://doi.org/10.1016/j.jvc.2018.01.003 1760-2734/ª 2018 Elsevier B.V. All rights reserved.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
2
D. Adin et al. uNa:uK over time during the infusion is consistent with mineralocorticoid modification of urinary electrolyte excretion, supporting renineangiotensinealdosterone activation as a cause of diuretic braking in this model. ª 2018 Elsevier B.V. All rights reserved.
Introduction Abbreviations CHF CRI FE pFur:uFur RAAS sCr UAldo:C uCl uCr uFur uK uNa
congestive heart failure constant rate infusion fractional excretion plasma furosemide to urine furosemide ratio renin angiotensin aldosterone system serum creatinine urine aldosterone to creatinine ratio urine chloride urine creatinine urine furosemide urine potassium urine sodium
Furosemide is the most commonly administered diuretic for the treatment of congestive heart failure (CHF) in dogs [1]. Several studies have demonstrated increased urine production after oral and parenteral administration of furosemide, but there are no other specific measures of efficacy reported, and diuretic dosage is not always correlated to diuretic response in the clinical setting [2e8]. Defining diuretic efficacy is important for the identification and management of patients with poor diuretic responsiveness and continued CHF despite appropriate treatment [9]. Multiple factors can affect the clinical response to a diuretic medication, including gastrointestinal absorption of the drug, renal tubular secretion of the drug, neurohormonal activation, and the development of true diuretic resistance [9]. Diuretic resistance is a clinically important problem identified in up to 17% of people with CHF [10]. Early identification is important because this phenomenon is associated with a poor prognosis [9,11]. Although diuretic resistance is suspected to occur in dogs with CHF, it has not been defined using indicators of pharmacodynamic response to furosemide [1,12]. Urine output is an ideal indicator of diuretic response, but it is difficult to quantify in patients, and the finding of decreased urine production does not provide mechanistic
information that might be useful in restoring diuretic responsiveness in refractory CHF. Several metrics, other than urine volume, have been proposed to quantify diuretic responsiveness, including weight loss, resolution of CHF, urinary sodium (uNa) concentrations, and fractional excretion (FE) of sodium [10,13]. The relationship of bloodto-urine furosemide concentrations is also necessary to understand the action of furosemide [14,15]. The knowledge provided by such metrics may aid in the development of therapeutic strategies to restore urine production in patients with refractory CHF. Hypochloremia, reduced spot uNa, low FE of sodium, low urine sodium to urine furosemide (uNa:uFur) and low urine sodium to urine potassium (uNa:uK) have been associated with a clinical diagnosis of diuretic resistance in people but are unexplored in dogs [10,11,13,16,17]. Diuretic braking is the abrupt reduction in urine production shortly after diuretic initiation, contributing to the initial loss of diuretic responsiveness [18]. Our laboratory demonstrated that urine production stimulated by a constant rate infusion (CRI) of furosemide, while initially high in dogs and horses, declines towards baseline after several hours, despite constant urine furosemide concentrations [6,7,19]. These findings are consistent with diuretic braking and the timing is compatible with renineangiotensinealdosterone system (RAAS) activation as a cause [6]. The study described herein evaluated measures of serum and urine electrolytes, urine aldosterone to creatinine ratio (UAldo:C), uNa:uFur and plasma furosemide to urine furosemide ratio (pFur:uFur) for the determination of diuretic efficacy and elucidation of the causes of diuretic braking in dogs. We hypothesized that urinary electrolyte excretion and furosemide excretion would be directly related to urine volume, and that indicators of RAAS activation would increase concurrently or before urine volume began to decline during furosemide CRI.
Animals, materials and methods The raw data utilized for this study were obtained from a previously published study which compared the effects of two diluents used for a furosemide
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy CRI, on urine production in normal dogs [6]. The present study, which addresses an entirely different and mechanistic aspect of diuresis, further evaluated urine and bloodwork values from one arm of the study by assessing the correlation of ratios and metrics with urine volume, furosemide excretion and indicators of RAAS activation during furosemide CRI with dextrose 5% in water as the diluent.
Dogs and protocol summary Six male dogs, determined to be healthy by physical examination, complete blood count, chemistry panel, and urinalysis, were included in the study. The Institutional Animal Care and Use Committee at the North Carolina State University Veterinary Hospital approved this study (protocol # 15-075-O). Dogs were instrumented with a urinary catheter, peripheral intravenous catheter, and jugular catheter under propofol anesthesia as previously reported [6]. After recovery, all dogs received an intravenous furosemide bolus of 0.66 mg/kg followed by 3.3 mg/ kg delivered with a 300-min infusion diluted with dextrose 5% in water to a final concentration of 2.2 mg/mL [6]. Dogs were allowed free access to water during the study period. Urine was collected via closed collection system. Dogs were fed a maintenance canine dieta with a sodium content of 112 mg/100 kcal before and during the study.
Data collection Urine volume was recorded hourly, and aliquots from each dog’s total hourly urine volume were collected during the 60-min baseline period and hourly for 300 min, for the analysis of urine furosemide concentrations, urine electrolyte concentrations (uNa, urine potassium {uK}, urine chloride {uCl}), urine creatinine concentrations (uCr), and urine aldosterone concentrations. Blood samples collected from the jugular catheter were centrifuged to harvest plasma or serum. These blood samples were obtained before furosemide administration (0 min) and hourly for 300 min for analysis of serum electrolytes (sodium, potassium, and chloride), serum creatinine (sCr), and plasma furosemide concentrations [6]. Serum was immediately analyzed for electrolytes and sCr, whereas urine and plasma were frozen at 80 Celsius. Biochemical analysis of serum and urine was performed using an automated analyzerb at the
3 clinical pathology laboratory of the North Carolina State University Veterinary Hospital. Furosemide was measured with reverse-phase high-pressure liquid chromatography and fluorescence detection using a validated method at North Carolina State University as previously described [6,19]. Urine aldosterone concentration was measured with a commercially available radioimmunoassay kitc, run by a veterinary diagnostic laboratoryd following manufacturer’s instructions as previously described [20]. Urine creatinine was measured using a standard colorimetric assay on the same samples at the same laboratory.d
Calculations Urine aldosterone amount (mg) was indexed to uCr amount (grams) to generate UAldo:C, an indicator of RAAS activity [21,22]. The amounts of urine electrolyte and furosemide excreted were calculated by multiplying the concentration (mg/dL or mg/mL) in each sample by the urine volume from the collection period. Electrolyte values reported in mmol/L were converted to mg/dL. Absolute amounts of urinary electrolytes and furosemide in milligrams (mg) were normalized to body weight in kilograms. The uNa:uK is used as a short-term biomarker of mineralocorticoid action and blockade because it reflects aldosterone’s effects on urinary electrolyte balance [23e25]. This indirect indicator of RAAS activation provides insight into the body’s avidity for these electrolytes under the influence of furosemide. The ratio was evaluated using concentrations (mmol/L) reported by the laboratory and also using absolute hourly amounts (mg/kg) of sodium and potassium, as described above. Electrolyte FE were calculated to determine the percent of filtered electrolytes excreted in the urine, thereby accounting for the effects of glomerular filtration rate and urinary flow rate on electrolyte loss [11]. Electrolyte FE was calculated using the following equation: fractional excretion of electrolyte (%) ¼ 100 {(urinary electrolyte sCr)/(serum electrolyte uCr)}. Hourly absolute uNa amounts (mg) were indexed to uFur amounts (mg) after both were normalized for body weight (uNa/kg:uFur/kg ¼ uNa:uFur) as a unitless indicator of furosemide’s pharmacodynamic efficacy [11,26]. Hourly, uNa and uFur amounts were also evaluated separately to aid interpretation of uNa:uFur.
c
a b
Certified Canine Diet 5007, LabDiet, St. Louis, MO, USA. Roche Cobas C501, Indianapolis, IN, USA.
Siemens Medical Diagnostic Solutions, Los Angeles, CA, USA. Diagnostic Center for Population and Animal Health, Michigan State University, Lansing, MI, USA. d
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
4
D. Adin et al.
The ratio of pFur concentrations (mg/mL) to uFur concentrations (mg/mL; pFur:uFur) was calculated as a unitless indicator of diuretic secretion into the renal tubule, as previously described [27]. To aid interpretation of pFur:uFur, values for plasma and urine furosemide concentrations, as well as the absolute urine furosemide amounts (mg/kg) independent of concentration, were separately evaluated.
Statistical analysis Statistical analysis was performed using commercially available software.e Calculations used the mean value of the six dogs at each time point. Data sets were tested for normality using the KolmogoroveSmirnov test. Normally distributed data are expressed as mean and standard deviation. Non-normally distributed data are expressed as median and range. Significance was set at p < 0.05. Correlations between variables were explored using Pearson’s correlation if data were normally distributed, or Spearman’s correlation if data were not normally distributed. Bonferroni’s correction to address potential type I error was applied to correlations by dividing the alpha of 0.05 by the number of variables tested (17) to obtain a lower alpha. Adjusted p-values are shown in Table 2. Changes over time during furosemide infusion were investigated using one-way repeated measures analysis of variance (RM-ANOVA) if data were normally distributed, or the Friedman test if data were not normally distributed. Post hoc multiple comparison tests were utilized when indicated (Tukey’s or Dunn’s). Time 0 min was not included for analysis of variance testing so that changes over time would be reflective only of the effects of the constant infusion of furosemide and not of a change from baseline, thus resulting in slightly different statistical values compared with previous report [6].
Results Hourly sodium (uNa) and chloride (uCl) excretion (mg/kg) were initially high with the furosemide CRI but significantly decreased during furosemide infusion in a pattern similar to the change in urine volume, whereas potassium (uK) excretion (mg/kg) was constant (Table 1). Hourly values for uNa, uNa:uK, electrolyte FE and UAldo:C are shown in Table 2. e
GraphPad Prism 6, La Jolla, CA, USA.
Table 1 Repeated measures analysis of variance testing for hourly changes from 60 to 300 min during the furosemide infusion (ANOVA or Friedman). Variable Urine volume (mL/kg) uNa (mg/kg) uCl (mg/kg) uK (mg/kg) uNa:uKc uNa:uKb UAldo:C (mg/g) uFur (mg/kg) uNa:Ufurc pFur (mg/mL) uFur (mg/mL) pFur:uFurb
Change over time
p-value
Decrease Decrease Decrease NS Decrease Decrease NS NS Decrease NS Increase Decrease
0.005a 0.0001a 0.0001a 0.2 0.0003a <0.0001a 0.5 0.08 0.009a 0.4 0.004a 0.003a
The p-values shown are adjusted after multiple comparison testing (Tukey’s or Dunn’s post hoc tests). Abbreviations: NS, not significant; uNa, urine sodium; uCl, urine chloride; uK, urine potassium; uNa:uK, urine sodium to potassium ratio; UAldo:C, urine aldosterone to creatinine ratio; uFur, urine furosemide; uNa:uFur, urine sodium to furosemide ratio (both mg/kg); pFur, plasma furosemide; pFur:uFur, plasma furosemide to urine furosemide ratio (both mg/mL). a Significance set at p < 0.05. b Unitless ratio with both components expressed as mmol/ L. c Unitless ratio with both components expressed as mg/kg.
Correlations to urine volume Urine volume was significantly and positively correlated with absolute amounts (mg/kg) of urine electrolyte excreted, but, there was no correlation with urine electrolyte FE or urine electrolyte concentrations (Table 3). Figure. 1A shows the strong positive relationship of urine volume to uNa amounts excreted (mg/kg). Urine volume was also significantly and positively correlated to uNa:uK generated as electrolyte amounts indexed to body weight (mg/kg) and also when the ratio was generated using concentrations (mmol/L; Fig. 1B, Table 3). Urine volume was not significantly correlated to sCl or UAldo:C. There was no significant correlation between urine volume and uFur absolute amounts (mg/kg) or indexed to uNa amounts (uNa:uFur; Table 3). A strong positive correlation was found between uNa and uCl absolute amounts (mg/kg), with approximately twice as much uCl as uNa excreted throughout the infusion (Table 3, Fig. 2).
Indicators of RAAS activation The increase in UAldo:C during the infusion, though not statistically significant (Table 1),
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy
5
Table 2 Mean standard deviation at each time point for uNa:uK, uNa, fractional excretion of electrolytes and UAldo:C. 0 min uNa:uK uNa FE Na FE K FE Cl UAldo:C
0.18 3.33 0.18 10.12 0.18 0.97
0.47 1.98 0.03 4.94 0.07 0.27
60 min 6.17 59.54 8.86 54.75 13.49 1.63
120 min
1.44 23.54 2.18 11.79 3.26 1.76
5.24 43.54 10.08 71.88 16.14 1.67
180 min
0.45 16.93 3.82 29.60 6.59 1.44
3.90 28.24 7.72 67.12 12.65 2.90
240 min
0.60 8.61 5.34 36.06 7.90 2.58
3.04 25.88 6.45 80.84 11.42 2.44
0.54 14.52 5.30 62.65 9.34 1.26
300 min 2.35 15.40 3.52 61.14 6.48 2.13
0.75 7.05 1.49 33.05 3.22 1.40
Abbreviations: uNa:uK, urine sodium to urine potassium ratio (generated using mmol/L); uNa, urine sodium (mg/kg), FE Na, fractional excretion of sodium (%), FE K, fractional excretion of potassium (%); FE Cl, fractional excretion of chloride (%); UAldo:C, urine aldosterone to creatinine ratio (mg/g).
temporally paralleled the decline in urine volume (Fig. 3, Table 2). The uNa:uK significantly decreased during the furosemide infusion regardless of whether this was calculated using mg/kg values or mmol/L concentrations (Table 1). The decline in uNa:uK occurred because of unchanged uK and decreased uNa over time (Table 1).
Furosemide excretion The uNa:uFur significantly decreased during the furosemide infusion because of unchanged
furosemide excretion (mg/kg) and decreased sodium excretion (mg/kg) during the infusion (Table 1). Although correlation of uNa:uFur with urine volume did not reach statistical significance, these two variables decreased in parallel (Fig. 4). Plasma furosemide concentrations (mg/mL) did not change over time but urine furosemide concentrations (mg/mL) increased during the furosemide infusion, resulting in a significant decrease in pFur:uFur from 0.13 0.08 at 60 min to 0.06 0.06 at 300 min (Table 1). The amount of furosemide (mg/kg) excreted into the urine remained unchanged during the infusion, despite
Table 3 Correlations of hourly urine volume to hourly urine electrolytes, urine furosemide, uNa:uFur, and indicators of RAAS activation. Variable 1 (hourly) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg)
Variable 2 (hourly)
Correlation coefficient
p value
uNa (mg/kg) uCl (mg/kg) uK (mg/kg) FE Na (%) FE Cl (%) FE K (%) uNa (mmol/L) uCl (mmol/L) uK (mmol/L) uNa:uKb uNa:uKc sCl (mmol/L) uFur (mg/mL) uFur (mg/kg) uNa:uFurc UAldo:C (mg/ g)
0.996 0.998 0.971 0.896 0.862 0.546 0.891 0.772 0.761 0.973 0.982 0.040 0.948 0.837 0.936 0.086
0.002a 0.002a 0.02a 0.3 0.5 1.0 0.3 1.0 1.0 0.02a 0.009a 1.0 0.2 1.0 0.3 1.0
The p-values shown are adjusted after Bonferroni correction. Abbreviations: uNa, urine sodium; uCl, urine chloride; uK, urine potassium; FE Na, fractional excretion of sodium; FE Cl, fractional excretion of chloride; FE K, fractional excretion of potassium; uNa:uK, urine sodium to potassium ratio; sCl, serum chloride concentration; uFur, urine furosemide; uNa:uFur, urine sodium to furosemide ratio (both as mg/kg); UAldo:C, urine aldosterone to creatinine ratio. a Significance set p < 0.05. b Unitless ratio with both components expressed as mmol/L. c Unitless ratio with both components expressed as mg/kg.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
6
D. Adin et al.
100
50
0 30
0 24
0 18
0 12
60
0
Time (minutes) Figure 2 The hourly mean standard deviation urine chloride (uCl; ,) and urine sodium (uNa, ) amounts (mg/kg) significantly decreased during the furosemide infusion (p ¼ 0.0001 uCl, p ¼ 0.0001 uNa) with uCl being approximately double uNa throughout the infusion. *Significantly different from 60 min, #significantly different from 120 min, ^significantly different from 180 min.
Figure 1 A. Hourly urine volume (mL/kg) and urine sodium (uNa; mg/kg) were highly correlated during the infusion (r ¼ 0.996, p < 0.0001). B. Hourly urine volume (mL/kg) and urine sodium to potassium ratio (uNa:uK) (using concentrations) were highly correlated during the infusion (r ¼ 0.976, p ¼ 0.001).
the increase in urine furosemide concentrations because of a significant decrease in urine volume during the latter part of the infusion (Table 1).
Discussion Urine volume was significantly and positively correlated with the amount of sodium, chloride, and potassium excreted in the urine but not with the amount of furosemide or aldosterone excreted in the urine. Urine volume was more strongly correlated to absolute (mg/kg) electrolyte excretion than to FE of these electrolytes. Therefore,
accounting for the effects of glomerular filtration rate and urinary flow rate by calculating FE does not appear to improve the analysis in dogs, as has been shown in people [11]. Other studies in people, however, have shown that FE of sodium <0.2% in CHF is associated with diuretic resistance, therefore, it is possible that we might find different results in clinical patients [10,28]. The close relationship between urine production and electrolyte excretion is explained by the mechanism of action of furosemide, which is to inhibit the renal Naþ/Kþ/2Cl cotransporter [18]. The Naþ/Kþ/2Cl cotransporter has twice the effect on chloride as it does on sodium, explaining the difference in urinary sodium excretion compared with urinary chloride excretion. Natriuresis was equivalent to urine volume as an indicator of diuretic efficacy after administration of furosemide. This finding is logical because water passively follows sodium, which is retained in the loop of Henle and excreted, after cotransporter inhibition by furosemide. However, urine sodium concentration (mmol/L) was not a useful correlate. Calculation of the absolute amount of urine sodium (mg) excreted was significantly correlated to diuretic efficacy but requires a measurement of the amount of urine produced to make an assessment. Urine volume measurement is a challenge in the clinical setting, hence limiting the clinical utility of uNa as a marker of diuretic success or failure.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy
7
Figure 3 The hourly mean urine volume (mL/kg; ▀ solid lines; p ¼ 0.005) is shown on the left y-axis and the hourly urine aldosterone to creatinine ratio (UAldo:C; mg/g; dashed lines; p ¼ 0.5) is shown on the right y-axis. It is noteworthy that urine furosemide (uFur; mg/kg) remained constant during the urine volume decline.
The natriuretic response to furosemide was also assessed by calculating the uNa:uFur ratio. Urine furosemide hourly amounts (mg/kg) did not change during the continuous rate infusion, supporting successful continuous delivery to the loop of Henle, despite a reduction in urine volume. The decrease in uNa:uFur could have occurred as a
result of the decrease in uNa, because furosemide requires the presence of intraluminal sodium to exert its diuretic effect [18]. This finding is in agreement with the mechanism of the Naþ/ Kþ/2Cl cotransporter, however, aldosteronemediated sodium and water reabsorption in the distal tubule could also have affected this ratio,
Figure 4 The hourly mean standard deviation urine sodium to urine furosemide ratio (uNa:uFur; on the right yaxis) significantly decreased during the furosemide infusion (p ¼ 0.009) as did urine volume (, on the left y-axis, p ¼ 0.005), indicating a decrement in furosemide efficacy (diuretic braking). *Significantly different from 60 min, #significantly different from 120 min.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
8 because we sampled urine, not tubular fluid. Urinary sodium has been previously used as a surrogate for renal tubular sodium because of the challenges inherent to collecting loop of Henle filtrate [11]. Direct sampling of intraluminal fluid from the renal tubule would be necessary to further evaluate electrolyte and drug concentrations at their site of action. The uNa:uK ratio, generated either by indexed amounts or by concentrations, was highly correlated with urine volume throughout the range of hourly urine volumes measured in this study. This finding is consistent with a study in people which showed that concentration generated uNa:uK < 1.0 was a characteristic of people with inadequate urine production in CHF and diuretic resistance [10]. Assessment of urine electrolytes in a spot urine sample to calculate uNa:uK using urine electrolyte concentrations is potentially useful as an indicator of urine volume after loop diuretic administration. Importantly, evaluation of uNa:uK is unitless and does not require indexing to body weight or accounting for urine volume. Therefore, this metric deserves additional study in clinical patients receiving different furosemide dosages and methods of administration and may be used to further study the effects of RAAS suppressors in clinical patients [23e25]. The decrease in urine volume after several hours of furosemide infusion was consistent with diuretic braking, and several findings in this study support RAAS activation as a cause. The decrease of uNa:uK during the infusion was highly correlated to urine volume and is explained by a decrease in uNa excretion coupled with unchanged uK excretion. Unopposed loop diuresis would be expected to cause parallel sodium and potassium loss [24]. The declining uNa:uK we found during furosemide infusion is consistent with increasing effects of aldosterone on urinary electrolyte balance in the distal nephron [23]. Aldosterone-mediated resorption of sodium in the distal tubule causes passive water reabsorption distal to the site of furosemide action in the loop of Henle, with the net effect of reduced urine production. Although we did not find a significant increase in UAldo:C over time, the decline in urine volume in this study occurred near the time of rising UAldo:C values, supporting future evaluation of aldosterone (and potentially other RAAS mediators) as a cause of diuretic braking and modified electrolyte excretion. Intravascular volume depletion, produced by furosemide-induced diuresis, is proposed to be the cause of RAAS activation in this model [29].
D. Adin et al. We explored the possibility of reduced furosemide secretion into the renal tubule as a mechanism of diuretic braking by evaluating the relationship between plasma and urine furosemide concentrations. The pFur:uFur is dependent on maintenance of therapeutic pFur and uninhibited furosemide secretion into the renal tubules, and so is an indicator of dosing and renal sensitivity to furosemide [29]. Continued and steady furosemide secretion during the infusion, in conjunction with reduced urine volume, was illustrated by the decline in the pFur:uFur. Although initial pFur:uFur was similar to that provided in previous reports after a single intravenous furosemide dose in dogs and people [27,29], we found a decline in this ratio over time, because uFur concentration increased but pFur concentrations were unchanged. The uFur concentration increased because urine volume decreased and uFur excretion was unchanged. Continuous infusion of furosemide was expected to, and did, maintain steady-state plasma furosemide concentrations. The finding of a falling pFur:uFur ratio is consistent with diuretic braking, whereas a rising ratio would have supported impaired secretion of furosemide into the renal tubule as the cause of falling urine volume [29]. The finding of a decreasing pFur:uFur rules out reduced delivery of furosemide to the site of action as a cause of diuretic braking in this model. This study has several limitations. The acute study was conducted in healthy dogs. Transference of results to patients with CHF receiving different dosages and routes of administration of furosemide, as well as RAAS suppressant agents, will require additional subacute or chronic studies. Specifically, applicability of these findings to patients receiving oral furosemide is unknown. Finally, we did not evaluate other markers of RAAS activation such as plasma angiotensin and plasma aldosterone, which may have provided additional insight into neurohormonal aberrations in this model.
Conclusions In conclusion, the results of this study support RAAS activation as a cause of diuretic braking in healthy dogs receiving a 5-h furosemide infusion. We did not find evidence of impaired renal furosemide secretion in this model of diuretic braking. Urine production was strongly correlated with urinary electrolyte excretion. The ratio of uNa:uK using concentrations may be clinically useful as a
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy spot test for estimating urine production after diuretic administration, and as an indicator of aldosterone influence.
Conflict of Interest Statement The authors do not have any conflicts of interest to disclose.
Acknowledgments This study was partially funded by an Intramural grant award from the North Carolina State University College of Veterinary Medicine and the Jane Lewis seeks endowment.
References [1] Atkins C, Bonagura J, Ettinger S, Fox P, Gordon S, Haggstrom J, Hamlin R, Keene B, Luis-Fuentes V, Stepien R. Guidelines for the diagnosis and treatment of canine chronic valvular heart disease. J Vet Intern Med 2009;23:1142e50. [2] Hori Y, Takusagawa F, Ikadai H, Uechi M, Hoshi F, Higuchi S. Effects of oral administration of furosemide and torsemide in healthy dogs. Am J Vet Res 2007;68:1058e63. [3] Hori Y, Ohshima N, Kanai K, Hoshi F, Itoh N, Higuchi S-I. Differences in the duration of diuretic effects and impact on the renineangiotensinealdosterone system of furosemide in healthy dogs. J Vet Med Sci 2010;72:13e8. [4] Uechi M, Matsuoka M, Kuwajima E, Kaneko T, Yamashita K, Fukushima U, Ishikawa Y. The effects of the loop diuretics furosemide and torasemide on diuresis in dogs and cats. J Vet Med Sci 2003;65:1057e61. [5] Harada K, Ukai Y, Kanakubo K, Yamano S, Lee J, Kurosawa TA, Uechi M. Comparison of the diuretic effect of furosemide by different methods of administration in healthy dogs. J Vet Emerg Crit Care 2015;25: 364e71. [6] Adin D, Atkins C, Papich M, DeFrancesco T, Griffiths E, Penteado M, Kurtz K, Klein A. Furosemide continuous rate infusion diluted with 5% dextrose in water or hypertonic saline in normal adult dogs: a pilot study. J Vet Cardiol 2016;19:44e56. [7] Adin DB, Taylor AW, Hill RC, Scott KC, Martin FG. Intermittent bolus injection versus continuous infusion of furosemide in normal adult greyhound dogs. J Vet Intern Med 2003;17:632e6. [8] Testani JM, Brisco MA, Turner JM, Spatz ES, Bellumkonda L, Parikh CR, Tang WW. Loop diuretic efficiency: a metric of diuretic responsiveness with prognostic importance in acute decompensated heart failure. Circ Hear Fail 2014;7: 261e70. [9] ter Maaten JM, Valente MA, Damman K, Hillege HL, Navis G, Voors AA. Diuretic response in acute heart failurepathophysiology, evaluation, and therapy. Nat Rev Cardiol 2015;12:184e92. [10] Doering A, Jenkins CA, Storrow AB, Lindenfeld J, Fermann GJ, Miller KF, Sperling M, Collins SP. Markers of diuretic resistance in emergency department patients with acute heart failure. Int J Emerg Med 2017;10:17.
9 [11] Singh D, Shrestha K, Testani J, Verbrugge F, Dupont M, Mullens W, Tan W. Insufficient natiuretic response to continuous intravenous furosemide is associated with poor long-term outcomes in acute decompensated heart failure. J Card Fail 2014;20:392e9. [12] Oyama MA, Peddle GD, Reynolds CA, Singletary GE. Use of the loop diuretic torsemide in three dogs with advanced heart failure. J Vet Cardiol 2011;13:287e92. [13] ter Maaten JM, Valente MA, Metra M, Bruno N, O’Connor CM, Ponikowski P, Teerlink JR, Cotter G, Davison B, Cleland JG, Givertz MM, Bloomfield DM, Dittrich HC, van Veldhuisen DJ, Hillege HL, Damman K, Voors AA. A combined clinical and biomarker approach to predict diuretic response in acute heart failure. Clin Res Cardiol 2016;105:145e53. [14] Kron B, Sjostrom P, Karlberg B, Odlind B. Acute tolerance to furosemide pretreatment with captopril or prazosin does not influence diuresis and natriuresis. Scand J Urol Nephrol 1994;28:337e44. [15] Sjo ¨stro ¨m P. Mechanisms of reduced effects of loop diuretics in healthy volunteers and in patients with renal disease. Scand J Urol Nephrol Suppl 1988;111:1e66. [16] White Martin G, van Gelder James, Easter G. The effect of loop diuretics on the excretion of Naþ, Ca2þ, Mg2þ and Cl. J Clin Pharmacol 1981;21:610e4. [17] Ferreira JP, Girerd N, Medeiros PB, Santos M, Carvalho HC, Bettencourt P, Ke ´nizou D, Butler J, Zannad F, Rossignol P. Spot urine sodium excretion as prognostic marker in acutely decompensated heart failure: the spironolactone effect. Clin Res Cardiol 2016;105:489e507. [18] Huang X, Mees ED, Vos P, Hamza S, Braam B. Everything we always wanted to know about furosemide but were afraid to ask. Am J Physiol Ren Physiol 2016;310:F958e71. [19] Johansson AM, Gardner SY, Levine JF, Papich MG, LaFevers DH, Fuquay LR, Reagan VH, Atkins CE. Furosemide continuous rate infusion in the horse: evaluation of enhanced efficacy and reduced side effects. J Vet Intern Med 2003;17:887e95. [20] Ames MK, Atkins CE, Lee S, Lantis AC, zumBrunnen JR. Effects of high doses of enalapril and benazepril on the pharmacologically activated renineangiotensine aldosterone system in clinically normal dogs. Am J Vet Res 2015;76:1041e50. [21] Ames MK, Atkins CE, Lantis AC, Zum Brunnen J. Evaluation of subacute change in RAAS activity (as indicated by urinary aldosterone:creatinine, after pharmacologic provocation) and the response to ACE inhibition. J Renin Angiotensin Aldosterone Syst JRAAS 2016;17:1e12. [22] Gardner SY, Atkins CE, Rausch WP, DeFrancesco TC, Chandler DW, Keene BW. Estimation of 24-h aldosterone secretion in the dog using the urine aldosterone: Creatinine ratio. J Vet Cardiol 2007;9:1e7. [23] Eudy RJ, Sahasrabudhe V, Sweeney K, Tugnait M, KingAhmad A, Near K, Loria P, Banker M, Piotrowski DW, Boustany-Kari CM. The use of plasma aldosterone and urinary sodium to potassium ratio as translatable quantitative biomarkers of mineralocorticoid receptor antagonism. J Transl Med 2011;9:1e11. [24] Brandish PE, Chen H, Szczerba P, Hershey JC. Development of a simplified assay for determination of the antimineralocorticoid activity of compounds dosed in rats. J Pharmacol Toxicol Methods 2008;57:155e60. [25] Guyonnet J, Elliott J, Kaltsatos V. A preclinical pharmacokinetic and pharmacodynamic approach to determine a dose of spironolactone for treatment of congestive heart failure in dog. J Vet Pharmacol Ther 2010;33:260e7.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
10
D. Adin et al.
[26] Rose J, Pruit A, Dayton P, JL M. Relationship of urinary furosemide excretion rate to natriuretic effect in experimental azotemia. J Pharmacol Exp Ther 1976;199:490e7. [27] Hirai J, Miyazaki H, Taneike T. The pharmacokinetics and pharmacodynamics of furosemide in the anesthetized dog. J Vet Pharmacol Ther 1992;15:231e9.
[28] Kumar D, Bagarhatta R. Fractional excretion of sodium and its association with prognosis of decompensated heart failure patients. J Clin Diagn Res 2015;9:OC01e3. [29] Sjo ¨stro ¨m PA, Odlind BG, Beermann BA, HammarlundUdenaes M. On the mechanism of acute tolerance to furosemide diuresis. Scand J Urol Nephrol 1988;22:133e40.
Available online at www.sciencedirect.com
ScienceDirect
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
-, -e-
www.elsevier.com/locate/jvc
Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model D. Adin, DVM*, C. Atkins, DVM , M.G. Papich, DVM College of Veterinary Medicine, North Carolina State University, 1060 William Moore Dr., Raleigh, NC 27607, USA Received 20 July 2017; received in revised form 15 January 2018; accepted 22 January 2018
KEYWORDS Lasix; Urine; Natriuresis; Aldosterone
Abstract Introduction: Diuretic failure is a potential life-ending event but is unpredictable and poorly understood. The objectives of this study were to evaluate pharmacodynamic markers of furosemide-induced diuresis and to investigate mechanisms of diuretic braking in dogs receiving constant rate infusion (CRI) of furosemide. Animals: Six healthy male dogs. Methods: Raw data and stored samples from one arm of a previously published study were further analyzed to mechanistically investigate causes of diuretic braking in these dogs. Urine volume was recorded hourly during a 5-h furosemide CRI. Urine and blood samples were collected hourly to measure serum and urine electrolytes, urine aldosterone, and plasma and urine furosemide. Serum electrolyte fractional excretion was calculated. Urine sodium concentration was indexed to urine potassium (uNa:uK) and urine furosemide (uNa:uFur) concentrations, plasma furosemide concentration was indexed to urine furosemide concentration (pFur:uFur), and urine aldosterone was indexed to urine creatinine (UAldo:C). Temporal change and the relationship to urine volume were evaluated for these measured and calculated variables. Results: Urine volume was significantly correlated with urine electrolyte amounts and with uNa:uK. The ratio of pFur:uFur decreased during the infusion, whereas furosemide excretion was unchanged. Conclusions: There was a strong relationship between urine volume and absolute urine electrolyte excretion. Urine volume was strongly correlated to uNa:uK, giving it potential as a spot indicator of urine production during diuresis. The decrease in
* Corresponding author. E-mail address: [email protected] (D. Adin). https://doi.org/10.1016/j.jvc.2018.01.003 1760-2734/ª 2018 Elsevier B.V. All rights reserved.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
2
D. Adin et al. uNa:uK over time during the infusion is consistent with mineralocorticoid modification of urinary electrolyte excretion, supporting renineangiotensinealdosterone activation as a cause of diuretic braking in this model. ª 2018 Elsevier B.V. All rights reserved.
Introduction Abbreviations CHF CRI FE pFur:uFur RAAS sCr UAldo:C uCl uCr uFur uK uNa
congestive heart failure constant rate infusion fractional excretion plasma furosemide to urine furosemide ratio renin angiotensin aldosterone system serum creatinine urine aldosterone to creatinine ratio urine chloride urine creatinine urine furosemide urine potassium urine sodium
Furosemide is the most commonly administered diuretic for the treatment of congestive heart failure (CHF) in dogs [1]. Several studies have demonstrated increased urine production after oral and parenteral administration of furosemide, but there are no other specific measures of efficacy reported, and diuretic dosage is not always correlated to diuretic response in the clinical setting [2e8]. Defining diuretic efficacy is important for the identification and management of patients with poor diuretic responsiveness and continued CHF despite appropriate treatment [9]. Multiple factors can affect the clinical response to a diuretic medication, including gastrointestinal absorption of the drug, renal tubular secretion of the drug, neurohormonal activation, and the development of true diuretic resistance [9]. Diuretic resistance is a clinically important problem identified in up to 17% of people with CHF [10]. Early identification is important because this phenomenon is associated with a poor prognosis [9,11]. Although diuretic resistance is suspected to occur in dogs with CHF, it has not been defined using indicators of pharmacodynamic response to furosemide [1,12]. Urine output is an ideal indicator of diuretic response, but it is difficult to quantify in patients, and the finding of decreased urine production does not provide mechanistic
information that might be useful in restoring diuretic responsiveness in refractory CHF. Several metrics, other than urine volume, have been proposed to quantify diuretic responsiveness, including weight loss, resolution of CHF, urinary sodium (uNa) concentrations, and fractional excretion (FE) of sodium [10,13]. The relationship of bloodto-urine furosemide concentrations is also necessary to understand the action of furosemide [14,15]. The knowledge provided by such metrics may aid in the development of therapeutic strategies to restore urine production in patients with refractory CHF. Hypochloremia, reduced spot uNa, low FE of sodium, low urine sodium to urine furosemide (uNa:uFur) and low urine sodium to urine potassium (uNa:uK) have been associated with a clinical diagnosis of diuretic resistance in people but are unexplored in dogs [10,11,13,16,17]. Diuretic braking is the abrupt reduction in urine production shortly after diuretic initiation, contributing to the initial loss of diuretic responsiveness [18]. Our laboratory demonstrated that urine production stimulated by a constant rate infusion (CRI) of furosemide, while initially high in dogs and horses, declines towards baseline after several hours, despite constant urine furosemide concentrations [6,7,19]. These findings are consistent with diuretic braking and the timing is compatible with renineangiotensinealdosterone system (RAAS) activation as a cause [6]. The study described herein evaluated measures of serum and urine electrolytes, urine aldosterone to creatinine ratio (UAldo:C), uNa:uFur and plasma furosemide to urine furosemide ratio (pFur:uFur) for the determination of diuretic efficacy and elucidation of the causes of diuretic braking in dogs. We hypothesized that urinary electrolyte excretion and furosemide excretion would be directly related to urine volume, and that indicators of RAAS activation would increase concurrently or before urine volume began to decline during furosemide CRI.
Animals, materials and methods The raw data utilized for this study were obtained from a previously published study which compared the effects of two diluents used for a furosemide
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy CRI, on urine production in normal dogs [6]. The present study, which addresses an entirely different and mechanistic aspect of diuresis, further evaluated urine and bloodwork values from one arm of the study by assessing the correlation of ratios and metrics with urine volume, furosemide excretion and indicators of RAAS activation during furosemide CRI with dextrose 5% in water as the diluent.
Dogs and protocol summary Six male dogs, determined to be healthy by physical examination, complete blood count, chemistry panel, and urinalysis, were included in the study. The Institutional Animal Care and Use Committee at the North Carolina State University Veterinary Hospital approved this study (protocol # 15-075-O). Dogs were instrumented with a urinary catheter, peripheral intravenous catheter, and jugular catheter under propofol anesthesia as previously reported [6]. After recovery, all dogs received an intravenous furosemide bolus of 0.66 mg/kg followed by 3.3 mg/ kg delivered with a 300-min infusion diluted with dextrose 5% in water to a final concentration of 2.2 mg/mL [6]. Dogs were allowed free access to water during the study period. Urine was collected via closed collection system. Dogs were fed a maintenance canine dieta with a sodium content of 112 mg/100 kcal before and during the study.
Data collection Urine volume was recorded hourly, and aliquots from each dog’s total hourly urine volume were collected during the 60-min baseline period and hourly for 300 min, for the analysis of urine furosemide concentrations, urine electrolyte concentrations (uNa, urine potassium {uK}, urine chloride {uCl}), urine creatinine concentrations (uCr), and urine aldosterone concentrations. Blood samples collected from the jugular catheter were centrifuged to harvest plasma or serum. These blood samples were obtained before furosemide administration (0 min) and hourly for 300 min for analysis of serum electrolytes (sodium, potassium, and chloride), serum creatinine (sCr), and plasma furosemide concentrations [6]. Serum was immediately analyzed for electrolytes and sCr, whereas urine and plasma were frozen at 80 Celsius. Biochemical analysis of serum and urine was performed using an automated analyzerb at the
3 clinical pathology laboratory of the North Carolina State University Veterinary Hospital. Furosemide was measured with reverse-phase high-pressure liquid chromatography and fluorescence detection using a validated method at North Carolina State University as previously described [6,19]. Urine aldosterone concentration was measured with a commercially available radioimmunoassay kitc, run by a veterinary diagnostic laboratoryd following manufacturer’s instructions as previously described [20]. Urine creatinine was measured using a standard colorimetric assay on the same samples at the same laboratory.d
Calculations Urine aldosterone amount (mg) was indexed to uCr amount (grams) to generate UAldo:C, an indicator of RAAS activity [21,22]. The amounts of urine electrolyte and furosemide excreted were calculated by multiplying the concentration (mg/dL or mg/mL) in each sample by the urine volume from the collection period. Electrolyte values reported in mmol/L were converted to mg/dL. Absolute amounts of urinary electrolytes and furosemide in milligrams (mg) were normalized to body weight in kilograms. The uNa:uK is used as a short-term biomarker of mineralocorticoid action and blockade because it reflects aldosterone’s effects on urinary electrolyte balance [23e25]. This indirect indicator of RAAS activation provides insight into the body’s avidity for these electrolytes under the influence of furosemide. The ratio was evaluated using concentrations (mmol/L) reported by the laboratory and also using absolute hourly amounts (mg/kg) of sodium and potassium, as described above. Electrolyte FE were calculated to determine the percent of filtered electrolytes excreted in the urine, thereby accounting for the effects of glomerular filtration rate and urinary flow rate on electrolyte loss [11]. Electrolyte FE was calculated using the following equation: fractional excretion of electrolyte (%) ¼ 100 {(urinary electrolyte sCr)/(serum electrolyte uCr)}. Hourly absolute uNa amounts (mg) were indexed to uFur amounts (mg) after both were normalized for body weight (uNa/kg:uFur/kg ¼ uNa:uFur) as a unitless indicator of furosemide’s pharmacodynamic efficacy [11,26]. Hourly, uNa and uFur amounts were also evaluated separately to aid interpretation of uNa:uFur.
c
a b
Certified Canine Diet 5007, LabDiet, St. Louis, MO, USA. Roche Cobas C501, Indianapolis, IN, USA.
Siemens Medical Diagnostic Solutions, Los Angeles, CA, USA. Diagnostic Center for Population and Animal Health, Michigan State University, Lansing, MI, USA. d
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
4
D. Adin et al.
The ratio of pFur concentrations (mg/mL) to uFur concentrations (mg/mL; pFur:uFur) was calculated as a unitless indicator of diuretic secretion into the renal tubule, as previously described [27]. To aid interpretation of pFur:uFur, values for plasma and urine furosemide concentrations, as well as the absolute urine furosemide amounts (mg/kg) independent of concentration, were separately evaluated.
Statistical analysis Statistical analysis was performed using commercially available software.e Calculations used the mean value of the six dogs at each time point. Data sets were tested for normality using the KolmogoroveSmirnov test. Normally distributed data are expressed as mean and standard deviation. Non-normally distributed data are expressed as median and range. Significance was set at p < 0.05. Correlations between variables were explored using Pearson’s correlation if data were normally distributed, or Spearman’s correlation if data were not normally distributed. Bonferroni’s correction to address potential type I error was applied to correlations by dividing the alpha of 0.05 by the number of variables tested (17) to obtain a lower alpha. Adjusted p-values are shown in Table 2. Changes over time during furosemide infusion were investigated using one-way repeated measures analysis of variance (RM-ANOVA) if data were normally distributed, or the Friedman test if data were not normally distributed. Post hoc multiple comparison tests were utilized when indicated (Tukey’s or Dunn’s). Time 0 min was not included for analysis of variance testing so that changes over time would be reflective only of the effects of the constant infusion of furosemide and not of a change from baseline, thus resulting in slightly different statistical values compared with previous report [6].
Results Hourly sodium (uNa) and chloride (uCl) excretion (mg/kg) were initially high with the furosemide CRI but significantly decreased during furosemide infusion in a pattern similar to the change in urine volume, whereas potassium (uK) excretion (mg/kg) was constant (Table 1). Hourly values for uNa, uNa:uK, electrolyte FE and UAldo:C are shown in Table 2. e
GraphPad Prism 6, La Jolla, CA, USA.
Table 1 Repeated measures analysis of variance testing for hourly changes from 60 to 300 min during the furosemide infusion (ANOVA or Friedman). Variable Urine volume (mL/kg) uNa (mg/kg) uCl (mg/kg) uK (mg/kg) uNa:uKc uNa:uKb UAldo:C (mg/g) uFur (mg/kg) uNa:Ufurc pFur (mg/mL) uFur (mg/mL) pFur:uFurb
Change over time
p-value
Decrease Decrease Decrease NS Decrease Decrease NS NS Decrease NS Increase Decrease
0.005a 0.0001a 0.0001a 0.2 0.0003a <0.0001a 0.5 0.08 0.009a 0.4 0.004a 0.003a
The p-values shown are adjusted after multiple comparison testing (Tukey’s or Dunn’s post hoc tests). Abbreviations: NS, not significant; uNa, urine sodium; uCl, urine chloride; uK, urine potassium; uNa:uK, urine sodium to potassium ratio; UAldo:C, urine aldosterone to creatinine ratio; uFur, urine furosemide; uNa:uFur, urine sodium to furosemide ratio (both mg/kg); pFur, plasma furosemide; pFur:uFur, plasma furosemide to urine furosemide ratio (both mg/mL). a Significance set at p < 0.05. b Unitless ratio with both components expressed as mmol/ L. c Unitless ratio with both components expressed as mg/kg.
Correlations to urine volume Urine volume was significantly and positively correlated with absolute amounts (mg/kg) of urine electrolyte excreted, but, there was no correlation with urine electrolyte FE or urine electrolyte concentrations (Table 3). Figure. 1A shows the strong positive relationship of urine volume to uNa amounts excreted (mg/kg). Urine volume was also significantly and positively correlated to uNa:uK generated as electrolyte amounts indexed to body weight (mg/kg) and also when the ratio was generated using concentrations (mmol/L; Fig. 1B, Table 3). Urine volume was not significantly correlated to sCl or UAldo:C. There was no significant correlation between urine volume and uFur absolute amounts (mg/kg) or indexed to uNa amounts (uNa:uFur; Table 3). A strong positive correlation was found between uNa and uCl absolute amounts (mg/kg), with approximately twice as much uCl as uNa excreted throughout the infusion (Table 3, Fig. 2).
Indicators of RAAS activation The increase in UAldo:C during the infusion, though not statistically significant (Table 1),
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy
5
Table 2 Mean standard deviation at each time point for uNa:uK, uNa, fractional excretion of electrolytes and UAldo:C. 0 min uNa:uK uNa FE Na FE K FE Cl UAldo:C
0.18 3.33 0.18 10.12 0.18 0.97
0.47 1.98 0.03 4.94 0.07 0.27
60 min 6.17 59.54 8.86 54.75 13.49 1.63
120 min
1.44 23.54 2.18 11.79 3.26 1.76
5.24 43.54 10.08 71.88 16.14 1.67
180 min
0.45 16.93 3.82 29.60 6.59 1.44
3.90 28.24 7.72 67.12 12.65 2.90
240 min
0.60 8.61 5.34 36.06 7.90 2.58
3.04 25.88 6.45 80.84 11.42 2.44
0.54 14.52 5.30 62.65 9.34 1.26
300 min 2.35 15.40 3.52 61.14 6.48 2.13
0.75 7.05 1.49 33.05 3.22 1.40
Abbreviations: uNa:uK, urine sodium to urine potassium ratio (generated using mmol/L); uNa, urine sodium (mg/kg), FE Na, fractional excretion of sodium (%), FE K, fractional excretion of potassium (%); FE Cl, fractional excretion of chloride (%); UAldo:C, urine aldosterone to creatinine ratio (mg/g).
temporally paralleled the decline in urine volume (Fig. 3, Table 2). The uNa:uK significantly decreased during the furosemide infusion regardless of whether this was calculated using mg/kg values or mmol/L concentrations (Table 1). The decline in uNa:uK occurred because of unchanged uK and decreased uNa over time (Table 1).
Furosemide excretion The uNa:uFur significantly decreased during the furosemide infusion because of unchanged
furosemide excretion (mg/kg) and decreased sodium excretion (mg/kg) during the infusion (Table 1). Although correlation of uNa:uFur with urine volume did not reach statistical significance, these two variables decreased in parallel (Fig. 4). Plasma furosemide concentrations (mg/mL) did not change over time but urine furosemide concentrations (mg/mL) increased during the furosemide infusion, resulting in a significant decrease in pFur:uFur from 0.13 0.08 at 60 min to 0.06 0.06 at 300 min (Table 1). The amount of furosemide (mg/kg) excreted into the urine remained unchanged during the infusion, despite
Table 3 Correlations of hourly urine volume to hourly urine electrolytes, urine furosemide, uNa:uFur, and indicators of RAAS activation. Variable 1 (hourly) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg) Urine volume (mL/kg)
Variable 2 (hourly)
Correlation coefficient
p value
uNa (mg/kg) uCl (mg/kg) uK (mg/kg) FE Na (%) FE Cl (%) FE K (%) uNa (mmol/L) uCl (mmol/L) uK (mmol/L) uNa:uKb uNa:uKc sCl (mmol/L) uFur (mg/mL) uFur (mg/kg) uNa:uFurc UAldo:C (mg/ g)
0.996 0.998 0.971 0.896 0.862 0.546 0.891 0.772 0.761 0.973 0.982 0.040 0.948 0.837 0.936 0.086
0.002a 0.002a 0.02a 0.3 0.5 1.0 0.3 1.0 1.0 0.02a 0.009a 1.0 0.2 1.0 0.3 1.0
The p-values shown are adjusted after Bonferroni correction. Abbreviations: uNa, urine sodium; uCl, urine chloride; uK, urine potassium; FE Na, fractional excretion of sodium; FE Cl, fractional excretion of chloride; FE K, fractional excretion of potassium; uNa:uK, urine sodium to potassium ratio; sCl, serum chloride concentration; uFur, urine furosemide; uNa:uFur, urine sodium to furosemide ratio (both as mg/kg); UAldo:C, urine aldosterone to creatinine ratio. a Significance set p < 0.05. b Unitless ratio with both components expressed as mmol/L. c Unitless ratio with both components expressed as mg/kg.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
6
D. Adin et al.
100
50
0 30
0 24
0 18
0 12
60
0
Time (minutes) Figure 2 The hourly mean standard deviation urine chloride (uCl; ,) and urine sodium (uNa, ) amounts (mg/kg) significantly decreased during the furosemide infusion (p ¼ 0.0001 uCl, p ¼ 0.0001 uNa) with uCl being approximately double uNa throughout the infusion. *Significantly different from 60 min, #significantly different from 120 min, ^significantly different from 180 min.
Figure 1 A. Hourly urine volume (mL/kg) and urine sodium (uNa; mg/kg) were highly correlated during the infusion (r ¼ 0.996, p < 0.0001). B. Hourly urine volume (mL/kg) and urine sodium to potassium ratio (uNa:uK) (using concentrations) were highly correlated during the infusion (r ¼ 0.976, p ¼ 0.001).
the increase in urine furosemide concentrations because of a significant decrease in urine volume during the latter part of the infusion (Table 1).
Discussion Urine volume was significantly and positively correlated with the amount of sodium, chloride, and potassium excreted in the urine but not with the amount of furosemide or aldosterone excreted in the urine. Urine volume was more strongly correlated to absolute (mg/kg) electrolyte excretion than to FE of these electrolytes. Therefore,
accounting for the effects of glomerular filtration rate and urinary flow rate by calculating FE does not appear to improve the analysis in dogs, as has been shown in people [11]. Other studies in people, however, have shown that FE of sodium <0.2% in CHF is associated with diuretic resistance, therefore, it is possible that we might find different results in clinical patients [10,28]. The close relationship between urine production and electrolyte excretion is explained by the mechanism of action of furosemide, which is to inhibit the renal Naþ/Kþ/2Cl cotransporter [18]. The Naþ/Kþ/2Cl cotransporter has twice the effect on chloride as it does on sodium, explaining the difference in urinary sodium excretion compared with urinary chloride excretion. Natriuresis was equivalent to urine volume as an indicator of diuretic efficacy after administration of furosemide. This finding is logical because water passively follows sodium, which is retained in the loop of Henle and excreted, after cotransporter inhibition by furosemide. However, urine sodium concentration (mmol/L) was not a useful correlate. Calculation of the absolute amount of urine sodium (mg) excreted was significantly correlated to diuretic efficacy but requires a measurement of the amount of urine produced to make an assessment. Urine volume measurement is a challenge in the clinical setting, hence limiting the clinical utility of uNa as a marker of diuretic success or failure.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy
7
Figure 3 The hourly mean urine volume (mL/kg; ▀ solid lines; p ¼ 0.005) is shown on the left y-axis and the hourly urine aldosterone to creatinine ratio (UAldo:C; mg/g; dashed lines; p ¼ 0.5) is shown on the right y-axis. It is noteworthy that urine furosemide (uFur; mg/kg) remained constant during the urine volume decline.
The natriuretic response to furosemide was also assessed by calculating the uNa:uFur ratio. Urine furosemide hourly amounts (mg/kg) did not change during the continuous rate infusion, supporting successful continuous delivery to the loop of Henle, despite a reduction in urine volume. The decrease in uNa:uFur could have occurred as a
result of the decrease in uNa, because furosemide requires the presence of intraluminal sodium to exert its diuretic effect [18]. This finding is in agreement with the mechanism of the Naþ/ Kþ/2Cl cotransporter, however, aldosteronemediated sodium and water reabsorption in the distal tubule could also have affected this ratio,
Figure 4 The hourly mean standard deviation urine sodium to urine furosemide ratio (uNa:uFur; on the right yaxis) significantly decreased during the furosemide infusion (p ¼ 0.009) as did urine volume (, on the left y-axis, p ¼ 0.005), indicating a decrement in furosemide efficacy (diuretic braking). *Significantly different from 60 min, #significantly different from 120 min.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
8 because we sampled urine, not tubular fluid. Urinary sodium has been previously used as a surrogate for renal tubular sodium because of the challenges inherent to collecting loop of Henle filtrate [11]. Direct sampling of intraluminal fluid from the renal tubule would be necessary to further evaluate electrolyte and drug concentrations at their site of action. The uNa:uK ratio, generated either by indexed amounts or by concentrations, was highly correlated with urine volume throughout the range of hourly urine volumes measured in this study. This finding is consistent with a study in people which showed that concentration generated uNa:uK < 1.0 was a characteristic of people with inadequate urine production in CHF and diuretic resistance [10]. Assessment of urine electrolytes in a spot urine sample to calculate uNa:uK using urine electrolyte concentrations is potentially useful as an indicator of urine volume after loop diuretic administration. Importantly, evaluation of uNa:uK is unitless and does not require indexing to body weight or accounting for urine volume. Therefore, this metric deserves additional study in clinical patients receiving different furosemide dosages and methods of administration and may be used to further study the effects of RAAS suppressors in clinical patients [23e25]. The decrease in urine volume after several hours of furosemide infusion was consistent with diuretic braking, and several findings in this study support RAAS activation as a cause. The decrease of uNa:uK during the infusion was highly correlated to urine volume and is explained by a decrease in uNa excretion coupled with unchanged uK excretion. Unopposed loop diuresis would be expected to cause parallel sodium and potassium loss [24]. The declining uNa:uK we found during furosemide infusion is consistent with increasing effects of aldosterone on urinary electrolyte balance in the distal nephron [23]. Aldosterone-mediated resorption of sodium in the distal tubule causes passive water reabsorption distal to the site of furosemide action in the loop of Henle, with the net effect of reduced urine production. Although we did not find a significant increase in UAldo:C over time, the decline in urine volume in this study occurred near the time of rising UAldo:C values, supporting future evaluation of aldosterone (and potentially other RAAS mediators) as a cause of diuretic braking and modified electrolyte excretion. Intravascular volume depletion, produced by furosemide-induced diuresis, is proposed to be the cause of RAAS activation in this model [29].
D. Adin et al. We explored the possibility of reduced furosemide secretion into the renal tubule as a mechanism of diuretic braking by evaluating the relationship between plasma and urine furosemide concentrations. The pFur:uFur is dependent on maintenance of therapeutic pFur and uninhibited furosemide secretion into the renal tubules, and so is an indicator of dosing and renal sensitivity to furosemide [29]. Continued and steady furosemide secretion during the infusion, in conjunction with reduced urine volume, was illustrated by the decline in the pFur:uFur. Although initial pFur:uFur was similar to that provided in previous reports after a single intravenous furosemide dose in dogs and people [27,29], we found a decline in this ratio over time, because uFur concentration increased but pFur concentrations were unchanged. The uFur concentration increased because urine volume decreased and uFur excretion was unchanged. Continuous infusion of furosemide was expected to, and did, maintain steady-state plasma furosemide concentrations. The finding of a falling pFur:uFur ratio is consistent with diuretic braking, whereas a rising ratio would have supported impaired secretion of furosemide into the renal tubule as the cause of falling urine volume [29]. The finding of a decreasing pFur:uFur rules out reduced delivery of furosemide to the site of action as a cause of diuretic braking in this model. This study has several limitations. The acute study was conducted in healthy dogs. Transference of results to patients with CHF receiving different dosages and routes of administration of furosemide, as well as RAAS suppressant agents, will require additional subacute or chronic studies. Specifically, applicability of these findings to patients receiving oral furosemide is unknown. Finally, we did not evaluate other markers of RAAS activation such as plasma angiotensin and plasma aldosterone, which may have provided additional insight into neurohormonal aberrations in this model.
Conclusions In conclusion, the results of this study support RAAS activation as a cause of diuretic braking in healthy dogs receiving a 5-h furosemide infusion. We did not find evidence of impaired renal furosemide secretion in this model of diuretic braking. Urine production was strongly correlated with urinary electrolyte excretion. The ratio of uNa:uK using concentrations may be clinically useful as a
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
Assessment of diuretic efficacy spot test for estimating urine production after diuretic administration, and as an indicator of aldosterone influence.
Conflict of Interest Statement The authors do not have any conflicts of interest to disclose.
Acknowledgments This study was partially funded by an Intramural grant award from the North Carolina State University College of Veterinary Medicine and the Jane Lewis seeks endowment.
References [1] Atkins C, Bonagura J, Ettinger S, Fox P, Gordon S, Haggstrom J, Hamlin R, Keene B, Luis-Fuentes V, Stepien R. Guidelines for the diagnosis and treatment of canine chronic valvular heart disease. J Vet Intern Med 2009;23:1142e50. [2] Hori Y, Takusagawa F, Ikadai H, Uechi M, Hoshi F, Higuchi S. Effects of oral administration of furosemide and torsemide in healthy dogs. Am J Vet Res 2007;68:1058e63. [3] Hori Y, Ohshima N, Kanai K, Hoshi F, Itoh N, Higuchi S-I. Differences in the duration of diuretic effects and impact on the renineangiotensinealdosterone system of furosemide in healthy dogs. J Vet Med Sci 2010;72:13e8. [4] Uechi M, Matsuoka M, Kuwajima E, Kaneko T, Yamashita K, Fukushima U, Ishikawa Y. The effects of the loop diuretics furosemide and torasemide on diuresis in dogs and cats. J Vet Med Sci 2003;65:1057e61. [5] Harada K, Ukai Y, Kanakubo K, Yamano S, Lee J, Kurosawa TA, Uechi M. Comparison of the diuretic effect of furosemide by different methods of administration in healthy dogs. J Vet Emerg Crit Care 2015;25: 364e71. [6] Adin D, Atkins C, Papich M, DeFrancesco T, Griffiths E, Penteado M, Kurtz K, Klein A. Furosemide continuous rate infusion diluted with 5% dextrose in water or hypertonic saline in normal adult dogs: a pilot study. J Vet Cardiol 2016;19:44e56. [7] Adin DB, Taylor AW, Hill RC, Scott KC, Martin FG. Intermittent bolus injection versus continuous infusion of furosemide in normal adult greyhound dogs. J Vet Intern Med 2003;17:632e6. [8] Testani JM, Brisco MA, Turner JM, Spatz ES, Bellumkonda L, Parikh CR, Tang WW. Loop diuretic efficiency: a metric of diuretic responsiveness with prognostic importance in acute decompensated heart failure. Circ Hear Fail 2014;7: 261e70. [9] ter Maaten JM, Valente MA, Damman K, Hillege HL, Navis G, Voors AA. Diuretic response in acute heart failurepathophysiology, evaluation, and therapy. Nat Rev Cardiol 2015;12:184e92. [10] Doering A, Jenkins CA, Storrow AB, Lindenfeld J, Fermann GJ, Miller KF, Sperling M, Collins SP. Markers of diuretic resistance in emergency department patients with acute heart failure. Int J Emerg Med 2017;10:17.
9 [11] Singh D, Shrestha K, Testani J, Verbrugge F, Dupont M, Mullens W, Tan W. Insufficient natiuretic response to continuous intravenous furosemide is associated with poor long-term outcomes in acute decompensated heart failure. J Card Fail 2014;20:392e9. [12] Oyama MA, Peddle GD, Reynolds CA, Singletary GE. Use of the loop diuretic torsemide in three dogs with advanced heart failure. J Vet Cardiol 2011;13:287e92. [13] ter Maaten JM, Valente MA, Metra M, Bruno N, O’Connor CM, Ponikowski P, Teerlink JR, Cotter G, Davison B, Cleland JG, Givertz MM, Bloomfield DM, Dittrich HC, van Veldhuisen DJ, Hillege HL, Damman K, Voors AA. A combined clinical and biomarker approach to predict diuretic response in acute heart failure. Clin Res Cardiol 2016;105:145e53. [14] Kron B, Sjostrom P, Karlberg B, Odlind B. Acute tolerance to furosemide pretreatment with captopril or prazosin does not influence diuresis and natriuresis. Scand J Urol Nephrol 1994;28:337e44. [15] Sjo ¨stro ¨m P. Mechanisms of reduced effects of loop diuretics in healthy volunteers and in patients with renal disease. Scand J Urol Nephrol Suppl 1988;111:1e66. [16] White Martin G, van Gelder James, Easter G. The effect of loop diuretics on the excretion of Naþ, Ca2þ, Mg2þ and Cl. J Clin Pharmacol 1981;21:610e4. [17] Ferreira JP, Girerd N, Medeiros PB, Santos M, Carvalho HC, Bettencourt P, Ke ´nizou D, Butler J, Zannad F, Rossignol P. Spot urine sodium excretion as prognostic marker in acutely decompensated heart failure: the spironolactone effect. Clin Res Cardiol 2016;105:489e507. [18] Huang X, Mees ED, Vos P, Hamza S, Braam B. Everything we always wanted to know about furosemide but were afraid to ask. Am J Physiol Ren Physiol 2016;310:F958e71. [19] Johansson AM, Gardner SY, Levine JF, Papich MG, LaFevers DH, Fuquay LR, Reagan VH, Atkins CE. Furosemide continuous rate infusion in the horse: evaluation of enhanced efficacy and reduced side effects. J Vet Intern Med 2003;17:887e95. [20] Ames MK, Atkins CE, Lee S, Lantis AC, zumBrunnen JR. Effects of high doses of enalapril and benazepril on the pharmacologically activated renineangiotensine aldosterone system in clinically normal dogs. Am J Vet Res 2015;76:1041e50. [21] Ames MK, Atkins CE, Lantis AC, Zum Brunnen J. Evaluation of subacute change in RAAS activity (as indicated by urinary aldosterone:creatinine, after pharmacologic provocation) and the response to ACE inhibition. J Renin Angiotensin Aldosterone Syst JRAAS 2016;17:1e12. [22] Gardner SY, Atkins CE, Rausch WP, DeFrancesco TC, Chandler DW, Keene BW. Estimation of 24-h aldosterone secretion in the dog using the urine aldosterone: Creatinine ratio. J Vet Cardiol 2007;9:1e7. [23] Eudy RJ, Sahasrabudhe V, Sweeney K, Tugnait M, KingAhmad A, Near K, Loria P, Banker M, Piotrowski DW, Boustany-Kari CM. The use of plasma aldosterone and urinary sodium to potassium ratio as translatable quantitative biomarkers of mineralocorticoid receptor antagonism. J Transl Med 2011;9:1e11. [24] Brandish PE, Chen H, Szczerba P, Hershey JC. Development of a simplified assay for determination of the antimineralocorticoid activity of compounds dosed in rats. J Pharmacol Toxicol Methods 2008;57:155e60. [25] Guyonnet J, Elliott J, Kaltsatos V. A preclinical pharmacokinetic and pharmacodynamic approach to determine a dose of spironolactone for treatment of congestive heart failure in dog. J Vet Pharmacol Ther 2010;33:260e7.
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003
10
D. Adin et al.
[26] Rose J, Pruit A, Dayton P, JL M. Relationship of urinary furosemide excretion rate to natriuretic effect in experimental azotemia. J Pharmacol Exp Ther 1976;199:490e7. [27] Hirai J, Miyazaki H, Taneike T. The pharmacokinetics and pharmacodynamics of furosemide in the anesthetized dog. J Vet Pharmacol Ther 1992;15:231e9.
[28] Kumar D, Bagarhatta R. Fractional excretion of sodium and its association with prognosis of decompensated heart failure patients. J Clin Diagn Res 2015;9:OC01e3. [29] Sjo ¨stro ¨m PA, Odlind BG, Beermann BA, HammarlundUdenaes M. On the mechanism of acute tolerance to furosemide diuresis. Scand J Urol Nephrol 1988;22:133e40.
Available online at www.sciencedirect.com
ScienceDirect
Please cite this article in press as: Adin D, et al., Pharmacodynamic assessment of diuretic efficacy and braking in a furosemide continuous infusion model, Journal of Veterinary Cardiology (2018), https://doi.org/10.1016/j.jvc.2018.01.003