- Split View
-
Views
-
Cite
Cite
Steven K. Engle, David E. Watson, Natriuretic Peptides as Cardiovascular Safety Biomarkers in Rats: Comparison With Blood Pressure, Heart Rate, and Heart Weight, Toxicological Sciences, Volume 149, Issue 2, February 2016, Pages 458–472, https://doi.org/10.1093/toxsci/kfv240
- Share Icon Share
Abstract
Cardiovascular (CV) toxicity is an important cause of failure during drug development. Blood-based biomarkers can be used to detect CV toxicity during preclinical development and prioritize compounds at lower risk of causing such toxicities. Evidence of myocardial degeneration can be detected by measuring concentrations of biomarkers such as cardiac troponin I and creatine kinase in blood; however, detection of functional changes in the CV system, such as blood pressure, generally requires studies in animals with surgically implanted pressure transducers. This is a significant limitation because sustained changes in blood pressure are often accompanied by changes in heart rate and together can lead to cardiac hypertrophy and myocardial degeneration in animals, and major adverse cardiovascular events (MACE) in humans. Increased concentrations of NPs in blood correlate with higher risk of cardiac mortality, all-cause mortality, and MACE in humans. Their utility as biomarkers of CV function and toxicity in rodents was investigated by exploring the relationships between plasma concentrations of NTproANP and NTproBNP, blood pressure, heart rate, and heart weight in Sprague Dawley rats administered compounds that caused hypotension or hypertension, including nifedipine, fluprostenol, minoxidil, L-NAME, L-thyroxine, or sunitinib for 1–2 weeks. Changes in NTproANP and/or NTproBNP concentrations were inversely correlated with changes in blood pressure. NTproANP and NTproBNP concentrations were inconsistently correlated with relative heart weights. In addition, increased heart rate was associated with increased heart weights. These studies support the use of natriuretic peptides and heart rate to detect changes in blood pressure and cardiac hypertrophy in short-duration rat studies.
Cardiovascular (CV) toxicity is a primary cause of attrition during drug development, addition of black box warnings to drug labels, and withdrawal of drugs from the market (Lasser et al., 2002; Laverty et al., 2011; Olson et al., 2000; Stevens and Baker, 2009). Although chronic changes in blood pressure (BP) of as little as 2 mmHg have been associated with increased morbidity and mortality in humans (Lewington et al., 2002), changes in BP or cardiac mass as a result of test article administration may go undetected in short-duration rat toxicology studies. Blood-based biomarkers of altered cardiac mass or function could be used in preclinical drug development to detect cardiac toxicity and select safer compounds for clinical development.
Atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP) are important regulators of cardiac growth, blood volume, and BP. They are produced and secreted by cardiomyocytes, and are hormonally active in the heart, kidneys, and vasculature, where they inhibit cardiac hypertrophy, cause increased glomerular filtration rate, inhibit renin secretion and reabsorption of sodium and water, and cause vasodilation (Jensen et al., 1998; Levin et al., 1998; Potter et al., 2006).
ANP and BNP are not only constitutively synthesized and secreted in the atria but are also transcriptionally upregulated and secreted by ventricular cardiomyocytes in cases of left ventricular dysfunction, cardiac hypertrophy, and hypotension in both humans and veterinary species (Engle et al., 2010; McGrath and de Bold, 2005; Oyama et al., 2013; Yasue et al., 1994). ANP and BNP are stored as prohormones in vesicles within atrial cardiomyocytes. Upon stimulation, they are proteolytically cleaved to yield amino-terminal pro-atrial natriuretic peptide (NTproANP) and amino-terminal pro-brain natriuretic peptide (NTproBNP) and the active natriuretic peptide (NP) hormones ANP and BNP. The active hormones are secreted in equimolar amounts with their respective N-terminal fragments, bind to transmembrane receptors with guanylate cyclase activity, and are cleared by receptor mediated internalization and degradation or proteolytic degradation by neutral endopeptidase and excretion through kidneys (Potter et al., 2006). The N-terminal fragments have longer half-lives than the active hormones (20–30 s vs 2.5 min for ANP and NTproANP, and 6.4 min vs 15.5 min for BNP and NTproBNP) leading to higher plasma concentrations and making them easier to detect (Semenov et al., 2011; Thibault et al., 1988). ANP and BNP secretion are associated with cardiac wall stretch, and plasma concentrations of both have been reported to correlate with left ventricular end diastolic pressure in humans with mild-to-severe congestive heart failure, but not with mean blood pressure (MBP), suggesting their secretion is more dependent on preload than afterload (Maeda et al., 1998; Thibault et al., 1999). Their diagnostic and prognostic utility in detecting cardiac stress and dysfunction in humans and animals alike make ANP and BNP, along with their N-terminal fragments, candidate biomarkers for detection of CV toxicities in rats and other veterinary species (Greco et al., 2003; Maisel et al., 2008; Oyama et al., 2013; Rienstra et al., 2006).
Previously, we reported a positive correlation between NTproANP concentrations in plasma and increased heart weights, and an inverse relationship between NTproANP concentrations and BP in rats (Engle et al., 2010). We sought to extend these relationships, and to begin characterizing NTproBNP in rats, to more diverse mechanisms of action (MOA) and drug targets in order to support the use of NPs as CV safety biomarkers for the detection of changes in BP or heart weight in rat toxicology studies. Therefore, we investigated the relationships between NTproANP and NTproBNP concentrations in plasma, histopathological alterations, heart rate (HR), and BP in rats administered compounds that cause CV changes in animals and humans. Sunitinib, nifedipine, minoxidil, fluprostenol, NG-nitro-L-arginine methyl ester (L-NAME), and L-thyroxine were chosen for these studies due to their known effects on BP, HR, and cardiac mass.
Sunitinib, used in the treatment of cancer, is known to cause hypertension through antagonism of the vascular endothelial growth factor receptor (Chu et al., 2007; Schmidinger et al., 2008). Nifedipine, an antihypertensive, is an L-type calcium channel inhibitor and causes vasorelaxation and decreased cardiac contractility, resulting in hypotension. Minoxidil, also an antihypertensive, causes hypotension through increased potassium permeability in smooth muscle (Meisheri et al., 1988). Fluprostenol, a prostaglandin F2α (PGF2α) analog, has been reported to cause cardiomyocyte hypertrophy (Adams et al., 1996; Lai et al., 1996), and PGF2α has been reported to stimulate secretion of NPs (Ruskoaho 1992; Thibault et al., 1999). L-NAME causes vasoconstriction by inhibiting the synthesis of nitric oxide, resulting in increased smooth muscle tone and hypertension. L-thyroxine has been reported to cause systolic hypertension in patients and increased relative heart weight in rats (Hu et al., 2005; Prisant et al., 2006).
MATERIALS AND METHODS
Animals
Animal protocols were approved by Eli Lilly and Company’s or Covance’s (Greenfield, Indiana) Institutional Animal Care and Use Committee and studies were conducted in facilities approved by the Association for Assessment and Accreditation of Laboratory Animal Care (Memphis, Tennessee). Male Sprague Dawley (CD/IGS) rats, aged 12–18 weeks, were obtained from Charles River Laboratories (Wilmington, Massachusetts). Rats were housed singly in ventilated stainless-steel racks with access to food (Teklad Global Rodent Diet 2014, Harlan Sprague Dawley, Inc, Indianapolis, Indiana) and water ad libitum, and allowed to acclimate for at least 1 week to caging, feeding, and watering conditions before further handling. Rats were maintained on a 12 h light-dark cycle with lights on from 6 am till 6 pm and administered test article or vehicle at approximately 8 am and again at 8 pm for twice daily dosing (nifedipine and minoxidil only). Time zero on all graphs (Figures 1–6) is approximately 8 am (± 15 min). Blood samples were collected via orbital plexus under isoflurane anesthesia. Rats were euthanized by exsanguination under anesthesia followed by organ removal.
Hemodynamic Measurements
Data Sciences International (DSI, St. Paul, Minnesota) transmitters and catheters were surgically implanted in the descending aorta as previously described (Engle et al., 2010) and used to measure HR, systolic pressure (SYS), and diastolic pressure (DIA). MBP and pulse pressure (PP) were derived from SYS and DIA. Rate pressure product (RPP), an index of myocardial oxygen demand (Ansari et al., 2012), was derived from SYS and HR (RPP = SYS x HR). Animals were allowed to recover for at least 10 days after surgery prior to being placed on study. Hemodynamic measurements were collected for 72 h prior to first dose administration in studies 2–6. Rats remained naïve during the first 24 h of predose recording and received only vehicle during the next 48 h in order to minimize the effect of animal handling on BP and HR. Predose data were used to detect any significant baseline differences or vehicle effects in studies 2–6. Hemodynamic data were averaged into 1-h blocks for analysis. Perturbations of BP and HR due to handling during dosing or blood collection were excluded from analysis for compound effects. Data were acquired and analyzed using DSI’s Ponemah Physiology Platform (P3) acquisition and analysis system.
Blood Collection and Biomarker Concentrations in Plasma
Blood was collected into tubes containing EDTA (K3) via orbital plexus. Plasma was collected after centrifugation and mixed with aprotinin to a final concentration of 700 KIU/ml prior to freezing and storing at −70°C for 1–2 weeks prior to analysis. NTproANP was measured by an enzyme linked immunosorbent assay (ELISA) developed for human NTproANP (Cat 04-BI-20892, Alpco Diagnostics, Salem, New Hampshire); cross-reactivity with rat NTproANP was confirmed using recombinant rat NTproANP. NTproBNP was measured by a rat specific electrochemiluminescent immunoassay (Cat K153JKD, Meso Scale Discovery, Rockville, Maryland).
Histology and Organ Weights
Hearts, brains, and left kidneys were excised and weighed. Heart weights were expressed as ratios to brain weight (HW/BrW). Brains were discarded after weighing. Mesenteries were collected and laid flat prior to embedding for histopathologic examination. Hearts, left kidneys, and mesenteries were processed for histology as previously described (Engle et al., 2010).
Study 1
Sunitinib was administered by oral gavage (10 ml/kg) in a 1% hydroxyethylcellulose solution at 0, 5, or 50 mg/kg once a day (q.d.) for 3 days to 4 rats per group (n = 4). Hemodynamic data were collected continuously for 96 h. Plasma samples were collected for determination of NP concentrations at 4 and 8 h after the first dose administration, and at 4, 8, and 24 h after the third (final) dose administration. Tissues were not collected in study 1.
Study 2
Nifedipine was purchased from Sigma-Aldrich (St. Louis, Missouri) and suspended in a 5% dextrose solution for oral gavage at 0, 3, 10, or 30 mg/kg twice a day (b.i.d.) for 14 days (n = 6 for blood and tissue collection, n = 4 for hemodynamic measurements). Hemodynamic data were collected in 24-h blocks of time over 5 days. Plasma samples were collected for determination of NP concentrations at 28, 76, 148, 168, and 316 h after first dose administration. These times correspond to 4 h after the 4th, 8th, and 28th doses, 4 h after the 14th dose and 12 h after the 15th. Hearts, brains, left kidneys, and mesenteries were collected at necropsy, 12 h after the last dose.
Study 3
Minoxidil was purchased from Sigma-Aldrich and administered in drinking water for 14 days at 0, 80, 120, or 240 mg/l (n = 6 for blood and tissue collection, n = 4 for hemodynamic measurements). Water consumption was measured daily to calculate dose consumed. Hemodynamic data were collected in 24-h blocks of time over 5 days. Plasma samples were collected for determination of NP concentrations at 28, 76, 148, 168, and 316 h after introduction of minoxidil in drinking water. Hearts, brains, left kidneys, and mesenteries were collected at necropsy, 336 h after compound introduction.
Study 4
Fluprostenol was purchased from Cayman Chemical Company (Ann Arbor, Michigan) and suspended in 0.9% sodium chloride for subcutaneous injection (1 ml/kg) b.i.d. (bis in die) at 0, 0.05, 0.15, or 0.3 mg/kg for 14 days (n = 6 for blood and tissue collection, n = 4 for hemodynamic measurements). Hemodynamic data were collected in 24-h blocks of time over 5 days. Plasma samples were collected for determination of NP concentrations at 28, 76, 148, 168, and 316 h after first dose administration. These times correspond to 4 h after the 4th, 8th, and 28th doses, 4 h after the 14th dose and 12 h after the 15th. Hearts, brains, left kidneys, and mesenteries were collected at necropsy, 12 h after the last dose.
Study 5
L-NAME was purchased from Sigma-Aldrich and suspended in 0.9% sodium chloride for intraperitoneal injection (5 ml/kg) q.d. (quaque die) at 0, 0.3, 3, or 30 mg/kg for 14 days (n = 6 for blood and tissue collection, n = 4 for hemodynamic measurements). Hemodynamic data were collected in 24-h blocks of time over 5 days. Plasma samples were collected for determination of NP concentrations at 28, 76, 148, 168, and 316 h after first dose administration. These times correspond to 4 h after the 2nd, 4th, and 14th dose administration, and 4 and 24 h after the 7th dose administration. Hearts, brains, left kidneys and mesenteries were collected at necropsy, 24 h after the last dose.
Study 6
L-thyroxine was purchased from Sigma-Aldrich and administered by oral gavage in a 1% carboxymethylcellulose solution at 0, 1, 10, or 100 mg/kg q.d. for 7 days (n = 6 for blood and tissue collection, n = 4 for hemodynamic measurements). Hemodynamic data were collected in 24-h blocks of time over 4 days. Plasma samples were collected 4 h after the first, fourth, and seventh (final) dose administrations. Hearts, brains, left kidneys, and mesenteries were collected at necropsy, 24 h after the last dose.
Statistics
Hemodynamic parameters were evaluated by a repeated measures analysis of variance (ANOVA). Analysis phases were assigned according to each light and dark cycle. The least squares mean and standard error were calculated for every treatment group at each time point and overall for all parameters evaluated. Data were analyzed using the MIXED procedure within the SAS/STAT System software (SAS, Cary, North Carolina). Factors in the model included pretreatment baseline, treatment group, time after dose, and the interaction of time after dose and treatment group. The correlation between repeated measurements on subjects was accounted for in the covariance structure which was selected based on the finite-sample corrected Akaike’s Information Criterion (Keselman et al. 1998). Denominator degrees of freedom were computed using the Kenward and Roger (1997) method.
Monotonicity of dose response was examined using a sequential trend test based on ordinal spacing of dose levels (Tukey et al. 1985). The trend test on treatment means was preceded by the interaction test of linear trend in treatment by time. If the interaction was significant, then the linear trend test was performed for each time point, otherwise, it was applied only to the treatment means pooled across all time points in each analysis phase. Nonmonotonic dose response was evaluated whenever no significant linear treatment trends were detected. The mean of each treatment group was compared with that of the vehicle control group using a Bonferroni adjusted t-test (Miller, 1981). This comparison was preceded by the interaction test of treatment by time. If the interaction was significant, the Bonferroni adjusted t-test was applied to the treatment means for each time point, otherwise, it was applied only to the treatment means pooled across all time points in each analysis phase.
Organ weight data were analyzed by ANOVA using GraphPad’s Prism software (La Jolla, California). NP data were analyzed by 2-way ANOVA using time and dose as covariables. Correlations and linear regressions were also analyzed using Prism. Results were considered statistically significant when P < .05. All organ weight, NP, and HR data are presented as group mean ± standard error of the mean (SEM).
RESULTS
Survival
One rat, given 0.15 mg/kg fluprostenol b.i.d, died under anesthesia following blood collection at 148 h after the first dose administration. This death was likely a complication of anesthesia, as no adverse behavior or appearance was noted in rats administered fluprostenol, nor loss of body weight or decrease in food consumption; however, necropsy was not performed to confirm. One rat administered L-thyroxine at 100 mg/kg was found dead after 5 daily doses. This death was accompanied by marked loss of body weight and adverse clinical observations (rough haircoat, alopecia and soiling) and was likely a result of administration of L-thyroxine. All other rats survived until their scheduled sacrifice at the completion of studies. Out of 212 rats treated with these 6 compounds, overall survival to study termination was greater than 99%.
Study 1–Sunitinib
Summary
Hypertension and bradycardia occurred after administration of sunitinib at 50 mg/kg (Figs. 1A and B). Decreased plasma concentrations of NTproANP were observed at the high dose, with a trend of decreases at the low dose. NTproBNP was significantly decreased at the high dose at a single time point (Figs. 1C and D).
Hemodynamics
MBP was increased by sunitinib at 50 mg/kg, but not at 5 mg/kg (Figure 1B). Average MBP for the whole study was 101 ± 0.5, 98 ± 0.5, and 114 ± 1 mmHg at 0, 5, and 50 mg/kg, respectively. MBP was 15% greater than control (P = .007) at 22 h after the first administration of sunitinib at 50 mg/kg (94 ± 3, 96 ± 6, and 108 ± 3 mmHg at 0, 5, and 50 mg/kg) and remained significantly increased through the end of the study (48 h after the third dose). BPs were 20%–30% greater in rats administered sunitinib at 50 mg/kg compared with control animals (DIA = 32%, SYS = 23%, MBP = 27%, PP = 11%) at the time of maximum change (MBP = 104 ± 3, 95 ± 5, and 132 ± 2 mmHg at 0, 5 and 50 mg/kg), which was 40 h after the third dose at 50 mg/kg (88 h after first dose).
Heart rates were 10% slower than controls 12 h after the first administration of sunitinib at 50 mg/kg (412 ± 24, 413 ± 11, and 369 ± 7 bpm at 0, 5, and 50 mg/kg, respectively, Figure 1A). Maximum decrease in HR was 19% (P = .004) and occurred with administration of sunitinib at 50 mg/kg 25 h after the first dose (407 ± 29, 410 ± 20, and 329 ± 11 bpm at 0, 5, and 50 mg/kg). Heart rates returned to normal (384 ± 11, 392 ± 4, and 383 ± 7 bpm at 0, 5, and 50 mg/kg) by 80 h after the first dose (32 h after the third and final dose). Mean HR for the study was 383 ± 4, 379 ± 4, and 355 ± 4 bpm at 0, 5, and 50 mg/kg.
Natriuretic peptides
NTproANP concentrations in plasma were 50%–80% lower in the 50 mg/kg group compared with control (P < .0001) starting at 8 h after the first dose (1.4 ± 0.2, 0.9 ± 0.1, and 0.7 ± 0.1 nM at 0, 5, and 50 mg/kg) and continuing throughout the remaining time points (Figure 1C). NTproANP concentrations were also significantly lower than control (P < .04) after 5 mg/kg at 8 h after the first and third doses. Overall average NTproANP concentrations were 1.1 ± 0.1, 0.9 ± 0.1, and 0.5 ± 0.1 nM at 0, 5 and 50 mg/kg, respectively.
NTproBNP concentrations were decreased 20%–60% in the 50 mg/kg group compared with control starting at 8 hours after the first dose (30.0 ± 3.7, 20.1 ± 2.3, and 20.1 ± 2.5 pM at 0, 5, and 50 mg/kg) and throughout the remaining time points (Figure 1D). Decreased NTproBNP was statistically significant (P < .001) after 50 mg/kg 8 h after the third dose (33.1 ± 6.0, 24.6 ± 2.2, and 11.7 ± 1.6 pM at 0, 5, and 50 mg/kg). Average plasma NTproBNP concentrations for the entire study were 27.2 ± 4.5, 22.2 ± 2.5, and 18.3 ± 2.6 pM at 0, 5, and 50 mg/kg.
No treatment-related changes in body weight occurred. Organ weights and tissues for histologic examination were not collected in this study.
Study 2–Nifedipine
Summary
Administration of nifedipine resulted in hypotension and tachycardia after each dose (Figs. 2A and B). Plasma concentrations of NTproANP were initially increased but were not different from control after 1 week of dosing (144 h, Figure 1C). Plasma concentrations of NTproBNP were increased only at the high dose during the first week. Significant changes in relative heart weight were not observed (Table 1, Figs. 2A–D). Prior to each consecutive dose, BP and HR returned to normal, or rebounded into hypertension and bradycardia.
. | Control . | Low . | Mid . | High . |
---|---|---|---|---|
Nifedipine | 0.56 ± 0.02 | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.64 ± 0.03 |
Minoxidil | 0.61 ± 0.03 | 0.69 ± 0.01 | 0.76* ± 0.02 | 0.74* ± 0.01 |
Fluprostenol | 0.62 ± 0.03 | 0.63 ± 0.01 | 0.69 ± 0.02 | 0.61 ± 0.01 |
L-NAME | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.57 ± 0.02 | 0.59 ± 0.01 |
L-Thyroxine | 0.60 ± 0.04 | 0.73 ± 0.03 | 0.75* ± 0.01 | 0.77* ± 0.03 |
. | Control . | Low . | Mid . | High . |
---|---|---|---|---|
Nifedipine | 0.56 ± 0.02 | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.64 ± 0.03 |
Minoxidil | 0.61 ± 0.03 | 0.69 ± 0.01 | 0.76* ± 0.02 | 0.74* ± 0.01 |
Fluprostenol | 0.62 ± 0.03 | 0.63 ± 0.01 | 0.69 ± 0.02 | 0.61 ± 0.01 |
L-NAME | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.57 ± 0.02 | 0.59 ± 0.01 |
L-Thyroxine | 0.60 ± 0.04 | 0.73 ± 0.03 | 0.75* ± 0.01 | 0.77* ± 0.03 |
Mean heart weight to brain weight ratio (HW/BrW).
*P ≤ .05 compared with time-matched, vehicle-treated controls (0 mg/kg). N = 6.
. | Control . | Low . | Mid . | High . |
---|---|---|---|---|
Nifedipine | 0.56 ± 0.02 | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.64 ± 0.03 |
Minoxidil | 0.61 ± 0.03 | 0.69 ± 0.01 | 0.76* ± 0.02 | 0.74* ± 0.01 |
Fluprostenol | 0.62 ± 0.03 | 0.63 ± 0.01 | 0.69 ± 0.02 | 0.61 ± 0.01 |
L-NAME | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.57 ± 0.02 | 0.59 ± 0.01 |
L-Thyroxine | 0.60 ± 0.04 | 0.73 ± 0.03 | 0.75* ± 0.01 | 0.77* ± 0.03 |
. | Control . | Low . | Mid . | High . |
---|---|---|---|---|
Nifedipine | 0.56 ± 0.02 | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.64 ± 0.03 |
Minoxidil | 0.61 ± 0.03 | 0.69 ± 0.01 | 0.76* ± 0.02 | 0.74* ± 0.01 |
Fluprostenol | 0.62 ± 0.03 | 0.63 ± 0.01 | 0.69 ± 0.02 | 0.61 ± 0.01 |
L-NAME | 0.59 ± 0.01 | 0.61 ± 0.03 | 0.57 ± 0.02 | 0.59 ± 0.01 |
L-Thyroxine | 0.60 ± 0.04 | 0.73 ± 0.03 | 0.75* ± 0.01 | 0.77* ± 0.03 |
Mean heart weight to brain weight ratio (HW/BrW).
*P ≤ .05 compared with time-matched, vehicle-treated controls (0 mg/kg). N = 6.
Hemodynamics
Decreased BP and increased HR relative to control (Figs. 2A and B) occurred after each dose of nifedipine at 10 or 30 mg/kg, and after most doses of 3 mg/kg. Duration of hypotension increased with dose and initial changes were followed by a return to baseline or increased BP and decreased HR prior to the next dose. Three hours after the first dose, MBP was significantly decreased (P ≤ .04) at all doses (94.5 ± 2.8, 90.2 ± 1.4, 81.4 ± 3.7 and 80.1 ± 2.0 mmHg at 0, 3, 10, and 30 mg/kg, respectively) and HR was significantly increased (P < .001) at 10 and 30 mg/kg (424 ± 14.0, 409 ± 10.5, 459 ± 8.4 and 487 ± 15.1 bpm at 0, 3, 10, and 30 mg/kg, respectively). Hypotension persisted after each dose for 0–3 h at 3 mg/kg, 3–6 h at 10 mg/kg, and 4‐ to 11-h postdose at 30 mg/kg. MBP was significantly decreased compared with control (P < .05) at 20%, 30%, and 50% of all time points at 3, 10, and 30 mg/kg, respectively. Magnitude of hypotension was greatest 2–4 h after each dose and increased with dose from 3 to 10 mg/kg. Maximum decrease in BP ranged from 14% to 25% less than control at 3 mg/kg (DIA = 15%, SYS = 14%, MBP = 14%, PP = 25%), 22%–26% at 10 mg/kg (DIA = 26%, SYS = 22%, MBP = 23%, PP = 22%), and 18%–26% less than control at 30 mg/kg (DIA = 22%, SYS = 20%, MBP = 18%, PP = 26%).
Duration of tachycardia also increased with dose (Figure 2A). Heart rate was significantly increased compared with control (P < .05) at 5%, 14%, and 22% of all time points at 3, 10, and 30 mg/kg, respectively. Maximum increase in HR was observed at 2- to 4-h postdose and was 12%, 11%, and 15% at 3, 10, and 30 mg/kg, respectively. Duration of hypotension and tachycardia diminished with continued administration of nifedipine. For example, MBP was significantly decreased for 11 h after the first administration of nifedipine at 30 mg/kg, but was only decreased for 5 h after the final administration.
Hypotension and tachycardia after dosing were followed by hypertension and bradycardia prior to the next administration of nifedipine. Significant increases in BP (P ≤ .01) first occurred 23 h after the first dose at 10 and 30 mg/kg (96.1 ± 3.1, 101.4 ± 2.0, 102.8 ± 2.5, and 106.7 ± 3.5 mmHg at 0, 3, 10, and 30 mg/kg). Maximum increases in MBP were 9, 11, and 18% greater than control at 3, 10, and 30 mg/kg, respectively. MBP was significantly increased compared with control (P < .05) at 0, 7, and 15% of all time points at 3, 10, and 30 mg/kg, respectively. Following initial increases, HRs were not different or were decreased compared with control for the remainder of each 12-h period prior to the next dose, with maximum decreases of 14%, 13%, and 12% at 3, 10, or 30 mg/kg, respectively. Significant decreases in HR (P < .05) were observed as early as 12 h after the second dose (immediately prior to the third dose) at 10 and 30 mg/kg (452 ± 17.6, 434 ± 16.1, 406 ± 22.3, and 402 ± 13.5 bpm at 0, 3, 10, and 30 mg/kg).
Although nifedipine caused significant decreases in BP at each dose, the net changes in average BP and HR were very small because of the rebound effect. Average MBP and HR across all time points were 102.1 ± 0.3, 101.6 ± 0.4, 100.2 ± 0.7, and 98.2 ± 0.8 mmHg and 449.3 ± 3.6, 429.2 ± 3.7, 429.2 ± 3.9, and 446.5 ± 3.7 bpm and 0, 3, 10, and 30 mg/kg, respectively.
Natriuretic peptides
NTproANP concentrations in plasma were increased after administration of nifedipine (Figure 2C) at 10 and 30 mg/kg 28 h (P < .05) after the first dose (0.67 ± 0.08, 0.77 ± 0.08, 1.25 ± 0.10, and 2.37 ± 0.23 nM at 0, 3, 10 and 30 mg/kg, respectively). NTproANP concentrations were also increased after 30 mg/kg at 76 h (P < .05) after the first dose (0.93 ± 0.14, 0.81 ± 0.06, 1.02 ± 0.09, and 2.03 ± 0.11 nM, at 0, 30, 10, and 30 mg/kg respectively) but were not different from control at subsequent times. NTproANP concentrations were not significantly affected by nifedipine at 3 mg/kg at any time point tested. Average NTproANP concentrations across all time points were 0.80 ± 0.10, 0.70 ± 0.07, 0.89 ± 0.11, and 1.41 ± 0.34 nM at 0, 3, 10, and 30 mg/kg, respectively.
NTproBNP concentrations in plasma (Figure 2D) were significantly increased (P < .05) after administration of nifedipine at 30 mg/kg 28 h after the first dose (15.4 ± 1.36, 17.57 ± 1.66, 23.41 ± 2.24, 42.43 ± 3.37 pM), and again 148 h after the first dose (14.45 ± 2.03, 13.41 ± 1.38, 16.56 ± 2.13, and 24.61 ± 2.28 pM). NTproBNP concentrations were not different from control at any other times or after administration of nifedipine at 3 or 10 mg/kg. Average NTproBNP concentrations across all time points were 14.57 ± 1.00, 14.15 ± 0.99, 16.64 ± 1.79, and 24.82 ± 4.61 pM at 0, 3, 10, and 30 mg/kg, respectively.
Organ weight, body weight, and histologic examination
A trend of increased heart weight to brain weight ratio (HW/BrW) was observed after administration of nifedipine (6%, 10%, and 16% at 3, 10, and 30 mg/kg), but was not statistically significant at any dose (P = .06 at 30 mg/kg, Table 1). No treatment-related changes were observed in kidney or body weights. Increased incidence of minimal basophilia in the tunica media of coronary arteries occurred as dosage increased (1 out of 6 at 0 and 3 mg/kg, 3 out of 6 at 10 mg/kg, and 5 out of 6 at 30 mg/kg). No other histologic alterations were observed.
Study 3–Minoxidil
Summary
Hypotension and tachycardia occurred after administration of minoxidil, along with increases in NTproANP and NTproBNP concentrations in plasma and increased relative heart weights (Table 1, Figs. 3A–D).
Water consumption
Average daily doses of minoxidil were 7, 10.5, and 19.5 mg/kg in the 80, 120, and 240 mg/l groups, respectively, as determined by volume of water consumed. Water consumption was measured for 3 days prior to introduction of minoxidil to drinking water and compared with consumption during the study. Addition of minoxidil to drinking water did not reduce the amount of water consumed during the study, with an average daily consumption of 30.4 ± 0.7, 28.8 ± 0.5, 29.0 ± 0.5, and 27.3 ± 0.57 ml in the 0, 7, 10.5, and 19.5 mg/kg groups, respectively. In the 3 days prior to first administration, average daily water consumption was 29.6 ± 1.1, 27.4 ± 0.8, 27.2 ± 1.1, and 27.1 ± 0.7 ml in the same groups.
Hemodynamics
Administration of minoxidil caused hypotension and tachycardia (Figs. 3A and B), both of which lessened in severity with continued dosing and were resolved by the final day of hemodynamic data collection (day 14). Significant decreases in MBP were observed on the first 4 days of collection (1, 4, 7, and 10). Hypotension began within 3–4 h of the introduction of minoxidil to drinking water, with statistically significant decreases in MBP and increased HR (P < .001) observed at all 3 doses by 9 h (103.7 ± 3.4, 94.3 ± 3.4, 96.3 ± 4.4, and 102.2 ± 2.5 mmHg; 497 ± 18.6, 485 ± 7.3, 508 ± 6.1, and 521 ± 25.3 bpm at 0, 7, 10.5, and 19.5 mg/kg, respectively). Significant decreases in MBP occurred through day 10; at 237 h after introduction of minoxidil, MBP was significantly decreased (P < .03) only in the 10.5 mg/kg group (98.6 ± 4.02, 93.7 ± 0.92, 95.3 ± 4.51, and 97.0 ± 1.70 mmHg at 0, 7, 10.5, and 19.5 mg/kg).
Significant increases in HR were not observed past day 4; after 96 h HR was not different from control at any dose (468 ± 17.3, 490 ± 7.9, 479 ± 13.3, and 490 ± 10.0 bpm at 0, 7, 10.5, and 19.5 mg/kg, Figure 3A). SYS decreased more than DIA at all times measured resulting in larger changes in PP than MBP. Maximum decreases in BP occurred 22 h after introduction of minoxidil at 7 mg/kg and ranged from 13% to 36% less than control (DIA = 13%, SYS = 20%, MBP = 16%, PP = 36%). Diastolic pressures were no longer decreased at the study’s end and decreases in SYS were less severe (maximum decrease in SYS on day 14: 13% at 7 mg/kg). The net effect on HR was a slight increase (1%–3%), with HR greater than control at 63%, 84%, and 74% of all time points and average HR for the entire study were 418 ± 3.7, 423 ± 4.3, 433 ± 4.3, and 431 ± 4.6 bpm at 7, 10.5, and 19.5 mg/kg, respectively. The net effect on MBP was a decrease of 1%–7%, with MBP less than control at 98%, 91%, and 59% of all time points and average MBP for the entire study were 100.2 ± 0.4, 92.7 ± 0.4, 94.9 ± 0.5, and 98.8 ± 0.5 mmHg at 7, 10.5, and 19.5 mg/kg, respectively.
Natriuretic peptides
NTproANP concentrations in plasma were increased in rats administered minoxidil (Figure 3C). Significant increases (P ≤ .0003) occurred at 28 and 76 h after administration of minoxidil at 10.5 and 19.5 mg/kg, and at 28 h after 7 mg/kg (P < .05). The maximum increase in NTproANP concentration occurred in the 19.5 mg/kg group at 76 h (0.95 ± 0.23, 1.94 ± 0.30, 2.30 ± 0.39, and 3.37 ± 0.27 nM at 0, 7, 10.5, and 19.5 mg/kg). Average NTproANP concentrations across all time points were 1.11 ± 0.29, 1.91 ± 0.18, 2.20 ± 0.26, and 2.52 ± 0.32 nM at 0, 7, 10.5, and 19.5 mg/kg, respectively.
Statistically significant increases in NTproBNP concentrations in plasma (P ≤ .02) were observed in all dose groups at all times tested (Figure 3D). The maximum increase in NTproBNP occurred in the 10.5 mg/kg group at the 168-h time point (8.50 ± 1.41, 36.34 ± 4.52, 53.49 ± 5.87, and 52.64 ± 4.48 pM, at 0, 7, 10.5, and 19.5 mg/kg). Notably, plasma NTproBNP concentrations remained increased at 316 h, after BP had normalized, possibly reflecting the presence of increased heart weights. Average NTproBNP concentrations across all time points were 12.35 ± 1.34, 44.75 ± 3.81, 56.51 ± 4.38, and 57.27 ± 2.28 pM.
Organ weight, body weight, and histologic examination
Minimal to slight increases in HW/BrW (13%, 25%, and 21%) occurred after administration of minoxidil at 7, 10.5, and 19.5 mg/kg for 14 days, respectively (P ≤ .02 compared with control at 10.5 and 19.5 mg/kg, Table 1). No treatment-related changes were observed in kidney or body weights. Minimal to slight degeneration and/or regeneration of renal tubules occurred in 2 of 6 rats at the mid-dose (10.5 mg/kg) and 3 of 6 rats at the high-dose (19.5 mg/kg).
Study 4–Fluprostenol
Summary
Hypotension and bradycardia occurred after administration of fluprostenol and were accompanied by increased NTproANP and NTproBNP concentrations without changes in relative heart weights (Table 1, Figs. 4A–D).
Hemodynamics
Administration of fluprostenol caused hypotension at all doses and bradycardia at 0.15 and 0.3 mg/kg throughout the study (Figs. 4A and B). MBPs were significantly decreased (P < .05 compared with control) by 0–2 h following each twice daily injection and persisted for 4–9 h at 0.05 mg/kg, 5–12 h at 0.15 mg/kg and for 8–12 h at 0.3 mg/kg. The magnitude of hypotension increased gradually after each dose. Maximum decrease in MBP after the first dose was 13.5% at 7 h after administration of fluprostenol (105.2 ± 0.7, 94.2 ± 2.2, 92.5 ± 1.5, and 91.0 ± 2.5 mmHg at 0, 0.05, 0.15, and 0.3 mg/kg, P ≤ .005 compared with control at all 3 dose levels) while the maximum decrease after the final dose was 17% at 325 h (108.4 ± 2.8, 91.0 ± 2.9, 90.2 ± 1.8, and 90.0 ± 2.7 mmHg at 0, 0.05, 0.15, and 0.3 mg/kg, P < .001 compared with control at all doses). Decreases in BP compared with control ranged from 16%–26% at 0.05 mg/kg (DIA = 14%, SYS = 18%, MBP = 16%, PP = 26%), from 15% to 6% at 0.15 mg/kg (DIA = 15%, SYS = 18%, MBP = 17%, PP = 26%), and from 20% to 36% at 0.3 mg/kg (DIA = 15%, SYS = 18%, MBP = 17%, PP = 25%) at 325 h. Average MBP across all time points was 107.3 ± 0.4, 97.2 ± 0.5, 94.8 ± 0.5, and 93.1 ± 0.5 mmHg at 0, 0.05, 0.15, and 0.3 mg/kg.
Heart rates (Figure 4A) were briefly increased (for 1–2 hours) after the first of 2 daily administrations of fluprostenol on the first 3 days of hemodynamic data collection (days 1, 2, and 4). Thereafter, HR was decreased throughout the study at 0.15 and 0.3 mg/kg, with maximum decrease from control occurring at 318 h on the final day (367 ± 8.1, 327 ± 12.5, 296 ± 2.4, and 310 ± 3.2 bpm at 0, 0.05 0.15, and 0.3 mg/kg, P ≤ .04 at 0.15 and 0.3 mg/kg vs control). Net effect on HR at 0.05 mg/kg was a slight increase (2.6%) and a decrease at 0.15 and 0.3 mg/kg (−5.8, −5.9%) with HR less than control at 83% and 86% of all time points (0.15 and 0.3 mg/kg, respectively). Average HR across all time points was 421 ± 3.9, 432 ± 4.3, 396 ± 4.1, and 396 ± 3.7 bpm at 0, 0.05, 0.15, and 0.3 mg/kg.
Natriuretic peptides
NTproANP concentrations in plasma were increased relative to control after administration of fluprostenol at 0.05 mg/kg (Figure 4C, P = .0004) at 76 hours (0.76 ± 0.07, 1.58 ± 0.07, 1.51 ± 0.07, and 1.49 ± 0.12 nM at 0, 0.05, 0.15, and 0.3 mg/kg). NTproANP concentrations were also increased at 28, 76, and 148 h in the 0.15 mg/kg group (P < .01), and at all time points in the 0.3 mg/kg group (P < .02). Average NTproANP concentrations across all time points were 0.72 ± 0.12, 1.11 ± 0.33, 1.29 ± 0.24, and 1.48 ± 0.21 nM at 0, 0.05, 0.15, and 0.3 mg/kg, respectively.
NTproBNP concentrations in plasma were increased (Figure 4D, P ≤ .0002) at 76 (40.15 ± 3.17, 85.57 ± 7.63, 77.33 ± 5.04, and 83.47 ± 7.74 pM at 0, 0.05, 0.15, and 0.3 mg/kg) and 316 h in the 0.05 mg/kg group, at 76–316 h at 0.15 mg/kg (P < .03), and at 28, 76, and 316 h at 0.3 mg/kg (P < .05). Average NTproBNP concentrations across all time points were 28.81 ± 9.24, 52.66 ± 23.77, 65.05 ± 12.04, and 68.72 ± 24.09 pM.
Organ weight, body weight, and histologic examination
No treatment-related changes in organ weights, body weights, or histological alterations were observed at any dose of fluprostenol.
Study 5–L-NAME
Summary
Administration of L-NAME resulted in a pattern of changes in NPs that was different in its relationship to BP than studies 1–4. Persistent hypertension occurred after administration of L-NAME, along with decreased HR and no changes in relative heart weights (Table 1, Figs. 5A and B). No significant changes in NTproANP concentrations occurred in plasma, but NTproBNP concentrations were increased at all times measured in the 30 mg/kg group (Figs. 5C and D).
Hemodynamics
Administration of L-NAME caused increased MBP throughout the study (Figure 5B). MBP was increased within 1 h of the first administration of L-NAME at 3 and 30 mg/kg (107.7 ± 3.6, 106.1 ± 2.0, 114.7 ± 2.7, and 117.5 ± 1.1 mmHg at 0, 0.3, 3, and 30 mg/kg, respectively, P ≤ .005 compared with control at 3 and 30 mg/kg) and remained significantly increased for the rest of the study at 30 mg/kg (P < .001) and at most times at 3 mg/kg (P < .05). MBP was significantly increased at 0.3 mg/kg (P ≤ .02) at 75, and 227–240 h. Maximum increases compared with control ranged from 7% to 30% at 0.3 mg/kg (DIA = 7%, SYS = 14%, MBP = 11%, PP = 30%), from 16% to 65% at 3 mg/kg (DIA = 17%, SYS = 18%, MBP = 16%, PP = 65%), and from 37% to 59% at 30 mg/kg (DIA = 39%, SYS = 37%, MBP = 37%, PP = 59%). Average MBP across all time points was 102.7 ± 0.3, 104.6 ± 0.4, 111.0 ± 0.3, and 121.5 ± 0.5 mmHg at 0, 0.3, 3, and 30 mg/kg.
Heart rates in rats administered L-NAME showed a trend of being decreased from control. Statistically significant decreases were sporadic (Figure 5A) and were observed most often at 30 mg/kg. Heart rates were less than control at 79%, 98%, 92% of all time points, by an average of 3%, 7%, and 7% at 0.3, 3, and 30 mg/kg, respectively. Net effects on HRs were decreases of 3%, 7%, and 7% at 0.3, 3, and 30 mg/kg, respectively. Average HR across all time points was 420 ± 3.3, 408 ± 3.3, 389 ± 2.5, and 391 ± 3.2 bpm at 0, 0.3, 3, and 30 mg/kg.
Natriuretic peptides
NTproANP concentrations in plasma were not significantly different than control after administration of L-NAME (Figure 5C). Average NTproANP concentrations across all time points were 0.61 ± 0.09, 0.61 ± 0.08, 0.48 ± 0.04, and 0.77 ± 0.08 nM at 0, 0.3, 3, and 30 mg/kg.
NTproBNP concentrations in plasma (Figure 5D) were increased in rats administered L-NAME at 30 mg/kg (P < .03) starting at 28 hours after the first dose (18.1 ± 1.5, 21.2 ± 1.8, 24.6 ± 3.1, and 48.0 ± 11.4 pM at 0, 0.3, 3, and 30 mg/kg) and continuing at all times except 168 h. NTproBNP was not increased at 0.3 or 3 mg/kg. Average NTproBNP concentrations across all time points were 20.3 ± 1.1, 21.5 ± 1.6, 25.5 ± 1.7, and 43.2 ± 3.5 pM at 0, 0.3, 3, and 30 mg/kg.
Organ weight, body weight, and histologic examination
No changes occurred in organ or body weights after administration of L-NAME. Minimal to slight, multifocal, histiocytic infiltration, consisting of aggregated macrophages scattered in the myocardium, often with fibrous connective tissue replacing myocytes, and occasionally infiltrating necrotic myocytes, was observed in the hearts of 3 of 6 rats administered L-NAME at 30 mg/kg and was considered compound related. Compound-related renal injury was observed in the cortical tubules and consisted of epithelial degeneration and regeneration in 1 rat given 3 mg/kg and all 6 rats given 30 mg/kg and increased incidence of basophilic tubules in 5 rats each at 3 and 30 mg/kg. Basophilic tubules were also observed in 2 control rats.
Study 6–L-thyroxine
Summary
Administration of L-thyroxine caused systolic hypertension after the second dose, along with tachycardia, and increased relative heart weights without detectable changes in NP concentrations (Table 1, Figs. 6A–D).
Hemodynamics
Hypotension occurred 1 h following the first administration of L-thyroxine and persisted for 10 h, with significant decreases in MBP (2%–9%, Figure 6B) and DIA (2%–10%) at 10 and 100 mg/kg (P < .02). SYSs were increased at all doses by the second dose administration, 25 h after the first dose (SYS = 120.9 ± 0.5, 127.8 ± 2.0, 130.3 ± 2.4, and 129.5 ± 2.5, P < .001 vs control at all doses), and were significantly increased at all subsequent time points through the end of the study (P ≤ .003). Maximum observed increases in SYS were 12%, 15%, and 13% at 1, 10, and 100 mg/kg, respectively.
MBPs were significantly increased at all doses after the second dose, beginning at 35 h (102.7 ± 2.0, 108.0 ± 1.5, 105.1 ± 3.4, and 108.2 ± 2.4 mmHg at 0, 1, 10, and 100 mg/kg, P < .04 at all doses) and continuing through 96 h. Significant increases in MBP continued at 10 and 100 mg/kg from 145 to 154 h, but were no longer different from control from 155–168 h. Average MBP across all time points was 102.5 ± 0.5, 105.4 ± 0.5, 104.5 ± 0.6, and 105.7 ± 0.6 mmHg at 0, 1, 10, and 100 mg/kg, respectively
Heart rates were increased by 11 h after the first dose at 10 and 100 mg/kg (Figure 6A, P < .001) and remained significantly higher than control for the rest of the study. Heart rates were increased by 1 h after the second dose at 1 mg/kg (418 ± 15.8, 424 ± 15.4, 482 ± 7.9, and 458 ± 11.0 bpm at 0, 1, 10, and 100 mg/kg, P ≤ .004 at all doses) and remained increased at all remaining time points. Average HR across all time points was 417.4 ± 3.6, 443.0 ± 4.6, 481.1 ± 5.0, and 481.5 ± 5.0 bpm at 0, 1, 10, and 100 mg/kg.
Natriuretic peptides
NTproANP and NTproBNP concentrations in plasma from rats administered L-thyroxine were not significantly different from controls at any time (Figs. 6C and D). An increase in NTproBNP in the 100 mg/kg group at the final time point (31.8 ± 10.8, 24.8 ± 3.4, 34.4 ± 9.2, and 63.0 ± 29.8 pM at 0, 1, 10, and 100 mg/kg) was due to the influence of a single animal and was not statistically significant at the group level (P = .06). Average NTproANP and NTproBNP concentrations in plasma across all time points were 1.1 ± 0.1, 1.0 ± 0.1, 1.6 ± 0.2, and 1.4 ± 0.2 nM, and 22.0 ± 3.9, 21.2 ± 1.9, 24.2 ± 4.0, and 35.4 ± 9.9 at 0, 1, 10, and 100 mg/kg, respectively.
Organ weight, body weight, and histologic examination
Relative heart weights were increased in rats given L-thyroxine at 10 and 100 mg/kg (P < .03 compared with control, Table 1). Heart weight to brain weight ratio was increased 23% after 1 mg/kg, 25% after 10 mg/kg, and 29% after 100 mg/kg. Minimal to slight, multifocal, myocardial necrosis and inflammation occurred in 2 of 6 rats at 1 mg/kg, in 6 of 6 rats at 10 mg/kg, in 4 of 5 rats examined in the 100 mg/kg group and were considered compound related. In these rats, individual or small groups of cardiomyocytes were necrotic and infiltrated and/or surrounded by macrophages and lymphocytes. Kidney weights were also increased (16, 10 and 23% at 1, 10 and 100 mg/kg, respectively) and epithelial hypertrophy was observed in scattered cortical tubules in 3 of 6 rats given 1 mg/kg, 5 of 6 rats given 10 mg/kg, and in 5 of 5 rats given 100 mg/kg.
Statistical Correlations and Baseline Concentrations
Prior to correlation analysis, distribution of NP data was found to be nonnormal. A log transformation was therefore conducted on the group mean fold change from control. Hemodynamic data (DIA, SYS, MBP, PP, RPP, and HR) were analyzed as change from predose baseline (1 h prior to first dose administration) in order to normalize diurnal variation corresponding to the light and dark cycles and differences in baselines. Changes in NTproANP concentrations were best correlated (inversely) with MBP and SYS and changes in NTproBNP concentrations were best correlated (inversely) with changes in PP in the cases of sunitinib, nifedipine, minoxidil, fluprostenol, and L-thyroxine, but neither NTproANP nor NTproBNP were well correlated with RPP or HR (Table 2). Changes in NTproBNP were positively correlated with DIA, SYS, MBP, PP and RPP after administration of L-NAME, but not with HR (Table 3). Correlations with NTproANP after L-NAME were not evaluated due to a lack of significant change in NTproANP concentrations.
. | NTproANP . | NTproBNP . | ||
---|---|---|---|---|
. | r2 . | P . | r2 . | P . |
DIA | 0.41 | <.0001 | 0.14 | .0004 |
SYS | 0.53 | <.0001 | 0.36 | <.0001 |
MBP | 0.5 | <.0001 | 0.3 | <.0001 |
PP | 0.32a | <.0001 | 0.49b | <.0001 |
RPP | 0.13 | .0007 | 0.13 | .0005 |
HR | 0.002 | .62 | 0.003 | .62 |
. | NTproANP . | NTproBNP . | ||
---|---|---|---|---|
. | r2 . | P . | r2 . | P . |
DIA | 0.41 | <.0001 | 0.14 | .0004 |
SYS | 0.53 | <.0001 | 0.36 | <.0001 |
MBP | 0.5 | <.0001 | 0.3 | <.0001 |
PP | 0.32a | <.0001 | 0.49b | <.0001 |
RPP | 0.13 | .0007 | 0.13 | .0005 |
HR | 0.002 | .62 | 0.003 | .62 |
R2 and P values for inverse correlations between log fold change from control in NTproANP and NTproBNP concentrations in plasma and change from predose (1 h prior to first dose) in diastolic, systolic, mean and pulse pressures, as well as rate pressure product and heart rate.
a0.52 if L-thyroxine is not included.
b0.68 if L-thyroxine is not included.
. | NTproANP . | NTproBNP . | ||
---|---|---|---|---|
. | r2 . | P . | r2 . | P . |
DIA | 0.41 | <.0001 | 0.14 | .0004 |
SYS | 0.53 | <.0001 | 0.36 | <.0001 |
MBP | 0.5 | <.0001 | 0.3 | <.0001 |
PP | 0.32a | <.0001 | 0.49b | <.0001 |
RPP | 0.13 | .0007 | 0.13 | .0005 |
HR | 0.002 | .62 | 0.003 | .62 |
. | NTproANP . | NTproBNP . | ||
---|---|---|---|---|
. | r2 . | P . | r2 . | P . |
DIA | 0.41 | <.0001 | 0.14 | .0004 |
SYS | 0.53 | <.0001 | 0.36 | <.0001 |
MBP | 0.5 | <.0001 | 0.3 | <.0001 |
PP | 0.32a | <.0001 | 0.49b | <.0001 |
RPP | 0.13 | .0007 | 0.13 | .0005 |
HR | 0.002 | .62 | 0.003 | .62 |
R2 and P values for inverse correlations between log fold change from control in NTproANP and NTproBNP concentrations in plasma and change from predose (1 h prior to first dose) in diastolic, systolic, mean and pulse pressures, as well as rate pressure product and heart rate.
a0.52 if L-thyroxine is not included.
b0.68 if L-thyroxine is not included.
. | NTproBNP . | |
---|---|---|
. | r2 . | P . |
DIA | 0.75 | <.0001 |
SYS | 0.73 | <.0001 |
MBP | 0.81 | <.0001 |
PP | 0.57 | .0001 |
RPP | 0.59 | <.0001 |
HR | 0.02 | .52 |
. | NTproBNP . | |
---|---|---|
. | r2 . | P . |
DIA | 0.75 | <.0001 |
SYS | 0.73 | <.0001 |
MBP | 0.81 | <.0001 |
PP | 0.57 | .0001 |
RPP | 0.59 | <.0001 |
HR | 0.02 | .52 |
R2 and P values for positive correlations between log fold change from control in NTproBNP concentrations in plasma and change from predose (1 h prior to first dose) in diastolic, systolic, mean and pulse pressures, as well as rate pressure product and heart rate.
. | NTproBNP . | |
---|---|---|
. | r2 . | P . |
DIA | 0.75 | <.0001 |
SYS | 0.73 | <.0001 |
MBP | 0.81 | <.0001 |
PP | 0.57 | .0001 |
RPP | 0.59 | <.0001 |
HR | 0.02 | .52 |
. | NTproBNP . | |
---|---|---|
. | r2 . | P . |
DIA | 0.75 | <.0001 |
SYS | 0.73 | <.0001 |
MBP | 0.81 | <.0001 |
PP | 0.57 | .0001 |
RPP | 0.59 | <.0001 |
HR | 0.02 | .52 |
R2 and P values for positive correlations between log fold change from control in NTproBNP concentrations in plasma and change from predose (1 h prior to first dose) in diastolic, systolic, mean and pulse pressures, as well as rate pressure product and heart rate.
In the cases of fluprostenol, nifedipine, minoxidil, and L-NAME there was a positive correlation between NTproANP concentrations at interim and terminal time points and relative heart weights measured at necropsy (Figure 7B). NTproBNP concentrations were not significantly correlated with HW/BrW at earlier timepoints (r2 = 0.21, P = .07 at 76 h) but were positively correlated at later times (r2 = 0.63, P = .0003 at 148 h). Change in HR was also positively correlated with relative heart weights (Figure 7C) for all compounds measured.
NTproANP and NTproBNP concentrations were pooled from all vehicle-treated control animals at all time points in order to establish expected baseline concentrations in male rats. NTproANP concentrations were 0.88 ± 0.04 nM and NTproBNP concentrations were 20.38 ± 0.88 pM in 158 samples.
DISCUSSION
Changes in NTproANP (Figure 7A) and NTproBNP concentrations in plasma were correlated with changes in BP (Tables 2 and 3). We previously observed a similar relationship between BP and plasma NTproANP concentrations in rats treated with a PPAR α/γ agonist (Engle et al., 2010). This correlation, along with the ease of detection of plasma NPs, supports their use as biomarkers of compound-related changes in BP in rats during general drug discovery toxicology studies to aid in the selection of safer compounds for development. Indeed, we have found NPs useful for identifying CV safety liabilities of compounds (including ion channel inhibitors, metabolic enzyme inhibitors, phosphodiesterase inhibitors, kinase inhibitors, and nuclear hormone receptor modulators) tested in rat toxicology studies (Figure 7D), much earlier in drug development than a conventional CV safety pharmacology study would be conducted.
The inverse relationships between plasma NTproANP and NTproBNP concentrations and BP suggests that increased plasma concentrations of NPs after administration of nifedipine, minoxidil, and fluprostenol may contribute to hypotension resulting from their administration. Increased secretion of ANP has been observed as a result of increased plasma volume in humans (Kamoi et al., 1988; Lieberman et al., 1991). Administration of nifedipine has resulted in increased plasma volume after 14 days in rats (Fekete et al., 2011) and increased plasma volumes in humans and rats have been observed after administration of minoxidil (Bryan et al., 1977; Grim et al., 1979; Sanz et al., 1990). Increased plasma volume may account for increased secretion of ANP and NTproANP in rats administered nifedipine or minoxidil, which in turn causes vasodilation and contributes to decreased BP measured in the descending aorta. The effect of sunitinib on plasma volume is not known, but decreased plasma volume has been proposed as part of the mechanism leading to erythrocytosis after administration of sunitinib (van der Veldt et al., 2009). Decreased plasma volume may account for the decreased NPs observed after administration of sunitinib and additional studies measuring plasma volume may help confirm this. The seemingly unlikely scenario of increased plasma volume and decreased BP has been previously reported (Ueda et al., 1986) and further studies in rats given nifedipine or minoxidil are required to directly measure their effects on plasma volume and whether manipulation of plasma volume through administration of diuretics may normalize NTproANP concentrations.
Prostaglandin F2α has caused secretion of ANP from isolated rat hearts and may also stimulate secretion of BNP (Clerico et al., 2011; Rayner et al., 1993). Fluprostenol, a PGF2α analog may directly stimulate secretion of ANP and BNP from cardiomyocytes, which in turn cause vasorelaxation and decreased BP. Studies of cultured cardiomyocytes treated with fluprostenol are needed to investigate this mechanism.
The range of BP effects in rats detectable with significant changes in plasma NP concentrations is important to estimate. Our data indicate that NPs are useful at detecting drug-related changes in MBP of not less than 10%–15% (equivalent to 10–15 mmHg), as with sunitinib, nifedipine, minoxidil, and fluprostenol. Changes in MBP of <10% (eg, L-thyroxine study) were not associated with changes in plasma NP concentrations. Such low magnitude changes in MBP in rats are associated with high biological variability such that changes in MBP are of short duration and may be spurious or associated with diurnal variation (±5–10 mmHg).
Drugs with MOAs that interfere with intracellular signaling of NPs or regulation of their secretion may affect the diagnostic utility of NPs and L-NAME may be an example of this phenomenon. L-NAME increased MBP in rats in a dose-related manner in the present studies, and has been reported to cause decreased plasma volume in rats (Filep, 1997; Qiu et al., 1998). Based on the inverse relationships between BP and NPs described above, and the effects of plasma volume on ANP secretion, a decrease in NP concentrations in plasma might be expected. Instead, we observed no change or a slight increase in NTproANP, and increased NTproBNP. We attribute these findings to the pharmacology of L-NAME: inhibition of nitric oxide (NO) synthesis. Nitric oxide inhibits secretion of ANP (Ruskoaho, 1992; Thibault et al., 1999). The inhibition of nitric oxide synthases by L-NAME therefore creates a pharmacological stimulus for ANP secretion by decreasing NO, which opposes the effect of hemodynamics in regulating plasma NTproANP concentrations. NTproBNP, being less sensitive to changes in plasma volume (Wambach and Koch, 1995), was increased, likely as a result of myocardial stretch resulting from increased afterload and arterial pressure. Understanding the pharmacological effects of drugs and their relationship to NP synthesis, release, and clearance, is important in deciding when to use and how to interpret data for plasma NPs. When the effects of a compound on NP regulation are not known, care must be taken when interpreting negative NP results.
The predictive value of plasma NPs for increased heart weight (cardiac hypertrophy) is also of interest. We previously reported a positive correlation between increased plasma NTproANP concentrations after 2–28 days and increased HW/BrW after 14 or 28 days in rats administered a PPARα/γ dual-agonist (Engle et al., 2010). Increased NTproANP concentrations and increased heart weights in rats have also been reported in models of aortic banding (Colton et al., 2011). In the present studies, there was a positive correlation between interim and terminal NP concentrations and final relative heart weights after administration of nifedipine, fluprostenol, minoxidil, and L-NAME (Figure 7B). However, when compounds were considered individually, minoxidil alone caused statistically significant increases in both relative heart weights and NP concentrations. In contrast, nifedipine and fluprostenol caused significant changes in both NPs measured and L-NAME in NTproBNP without significant increases in HW/BrW after 14 days of treatment. It is possible that longer durations of treatment would cause increased heart weights, or that measurement of cardiomyocyte diameter would reveal cardiac hypertrophy without significant increases in gross organ weight, but the data collected in these studies suggest an inconsistent relationship between plasma NPs and heart weight after 1 to 2 weeks. The lack of a clear threshold concentration makes prediction of significantly increased heart weights using NP data as a stand-alone marker problematic in studies of this duration. We conclude that increased plasma NPs are helpful as triggers for further investigation of CV safety during drug development, or as a screen for increased CV risk, but insufficient as stand-alone markers to predict which treatments will produce increased heart weight in rats. Further investigations of this relationship would be helpful, including imaging, such as echocardiography to investigate cardiac function (eg, ejection fraction) and measure chamber and wall dimensions within the heart.
Another example of the importance of MOA on utility of NPs as biomarkers is illustrated by L-thyroxine, which increased relative heart weights without changing NP concentrations. Cardiac hypertrophy in rats caused by excess thyroid hormones has been previously characterized as adaptive or physiological, that is, with enhanced cardiac function and proportional increases in wall thickness and chamber volume (Chilian et al., 1985; Kuzman et al., 2005). These results are consistent with our observation that changes in MBP of less than 10 mmHg were not associated with changes in NPs in rats and also with reports of normal BNP and NTproBNP concentrations and increased cardiac mass in endurance athletes, also characterized as physiological hypertrophy, or “athlete’s heart” (Almeida et al., 2002; Pagourelias et al., 2010; Scharhag et al., 2004). These data suggest that NPs may facilitate differentiation between adaptive and maladaptive changes in cardiac mass in rats as well as humans. Further studies combining imaging and measurement of circulating NP levels will help to understand this relationship.
Another important correlation observed in these studies was that increased HR was associated with increased HW/BrW for all compounds assessed (Figure 7C). This was observed in rats following administration of minoxidil, L-thyroxine, and previously with a PPAR α/γ agonist (Engle et al., 2010). Nifedipine, fluprostenol, and L-NAME were associated with decreased HR and no change in HW/BrW. We were unable to replicate previously reported (Lai et al., 1996) increases in heart weight in rats administered fluprostenol.
These studies support measurement of plasma NTproANP and NTproBNP concentrations in rats to detect undesirable changes in BP during general toxicology studies that historically lack endpoints to assess compound effects on CV function. Addition of NP measurements in rodent toxicology studies may help in the selection of safer compounds for continued development and reduced attrition due to CV safety liabilities. We have applied NTproANP as a CV safety biomarker in drug development studies with a diverse array of mechanisms and targets, including ion channel inhibitors, metabolic enzyme inhibitors, phosphodiesterase inhibitors, kinase inhibitors, and nuclear hormone receptor modulators and found an inverse relationship between changes in NTproANP and MBP consistent with the present studies (Figure 7D). When applied in short-duration toxicology studies, ie, 2–4 days, as previously described for use of cardiac troponin I (Engle et al., 2009), addition of NP measurements allow a more comprehensive CV safety assessment earlier in drug development and in a higher throughput paradigm than possible in instrumented rat studies with direct measurement of BP. Whether a result of changes in plasma volume, direct stimulation of secretion from the heart, myocardial stretch, or other unknown mechanisms, changes in either NTproANP or NTproBNP, or both, have proven a reliable indicator of hemodynamic changes within the first 48 h of compound administration. When used as a trigger for direct measurement of HR in rats, NPs can help in the prediction of maladaptive increases in HW/BrW in chronic studies and promote the selection of molecules with better CV safety profiles.
FUNDING
This work was supported by Eli Lilly and Company.
ACKNOWLEDGMENTS
The authors wish to thank John Sullivan, whose support made these studies possible. The authors also wish to thank Jacqueline Akunda, Derek Douglas Best, Alan Chiang, Les Freshwater, Hsiu-Yung C Lee, Derek Leishman, Aimin Lin, Brad Main, Matthew Renninger, Daniel Sall, Albert Eric Schultze, Phil Solter, and Yuewei Qian for their assistance in study design, synthesis of compounds, conduct of studies, and manuscript review.
REFERENCES
Comments