- Email: [email protected]
Cardiac Alterations in Furosemide-treated Thiamine-deprived Rats
Journal of Cardiac Failure Vol. 13 No. 9 2007
Basic Science and Experimental Studies
Cardiac Alterations in Furosemide-treated Thiamine-deprived Rats ˜ O BOSCO SALLES, PhD,3 SERGIO DA CUNHA, MD,1,2 JAYME CUNHA BASTOS, PhD,3 JOA 3 ´ MARIA CRISTINA COSTA SILVA, BSc, VERA LUCIA FREIRE CUNHA BASTOS, PhD,3 AND CARLOS ALBERTO MANDARIM-DE-LACERDA, MD, PhD1 Rio de Janeiro, Brazil
ABSTRACT Background: Chronic administration of furosemide may induce thiamine deficiency and cause or aggravate myocardial dysfunction. Methods and Results: Wistar rats were divided into four groups according to food and treatment: (1) thiamine standard chow with intraperitoneal furosemide administration; (2) thiamine standard chow with intraperitoneal saline administration; (3) thiamine-deficient chow with intraperitoneal furosemide administration; and (4) thiamine-deficient chow with intraperitoneal saline administration. Thiamine status was evaluated by high-performance liquid chromatography determination in plasma, erythrocytes, and myocardium, and by erythrocyte transketolase activity and the thiamine pyrophosphate effect to recover transketolase activity. Left ventricular mass index, intramyocardial arteries-to-cardiomyocyte ratio, cardiomyocyte cross-sectional area, and cardiomyocyte nuclei number were estimated. Myocardial structure was also studied by transmission electronic microscopy. Group 3 showed significantly lower blood and myocardial thiamine levels, which was not observed in group 1. Left ventricular mass index, cardiomyocyte cross-sectional area, and intramyocardial arteries-to-cardiomyocyte ratio were smaller in thiaminedeficient and furosemide-treated rats. However, no significant variation was found in the number of cardiomyocyte nuclei among the groups. Transmission electronic microscopy showed mitochondrial alterations in the thiamine-deficient groups. Conclusion: The present results indicate that furosemide administration is not the primary cause of thiamine deficiency in rats with adequate thiamine intake. Furosemide aggravates thiamine deficiency only in situations associated with insufficient thiamine intake, causing cardiac structural alterations, such as myocardial fiber hypotrophy, poor microvascularization, and mitochondrial degeneration. (J Cardiac Fail 2007;13:774e784) Key Words: Diuretic, electronic microscopy, stereology, vitamin B1.
From the 1Laboratory of Morphometry and Cardiovascular Morphology, Medical Clinics Department of Medical Sciences School, and 3Laboratory of Biochemical Toxicology, Department of Biochemistry, Biomedical Center, State University of Rio de Janeiro, Brazil. Manuscript received March 17, 2007; revised manuscript received June 15, 2007; revised manuscript accepted June 19, 2007. Reprint requests: Carlos Alberto Mandarim-de-Lacerda, MD, PhD, Laborato´rio de Morfometria e Morfologia Cardiovascular, Centro Biome´dico, Universidade do Estado do Rio de Janeiro, Av 28 de Setembro 87 (fds)20551-030, Rio de Janeiro, RJ, Brasil. This study was supported by grants from The National Council for Scientific and Technological Development (CNPq, www.cnpq.br) and The Carlos Chagas Filho Foundation for Supporting Research in the Rio de Janeiro State (Faperj, www.faperj.br). 1071-9164/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.cardfail.2007.06.729
Thiamine exists as free thiamine, thiamine monophosphate, thiamine pyrophosphate (TPP), and thiamine triphosphate in animal tissues. The pyrophosphate form, acting as the coenzyme of alpha-keto acid decarboxylation and transketolation,1 represents approximately 80% of the total thiamine. Accordingly, TPP is essential for carbohydrate metabolism, helping to produce cellular energy and ribose, and protecting against tissue oxidative damage by maintaining reduced nicotinamide adenine dinucleotide phosphate. Measurements of TPP, transketolase activity, and the percentage of increase of transketolase activity after TPP addition in vitro, the TPP effect, have been used to identify thiamine deficiency.2 Thiamine deficiency causes
2
774
Furosemide, Thiamine, and Cardiac Lesion
cardiovascular and neurologic damage, which are clinically noticeable in beriberi. The classic cardiovascular picture involves high-output biventricular heart failure, peripheral vasodilatation, volume overload, tachycardia, and wide pulse pressure.3 Furosemide is used for treating congestive heart failure, albeit with some well-known side effects such as electrolytic disturbances and hypovolemia. In the last 25 years, experimental and clinical studies have suggested that furosemide can induce thiamine deficiency4 and possibly cause or aggravate myocardial dysfunction.5e10 Much as cardiac lesions caused by thiamine deficiency have been known for more than 60 years,11 significant myocardial alterations caused by thiamine deficiency occurring with furosemide administration are still under investigation.12e14 Although no connection between furosemide administration and thiamine deficiency has been found,15e17 the present work was undertaken to verify whether long-term furosemide administration can induce myocardial thiamine deficiency with consequential structural alterations. The impairment of thiamine uptake by cardiac cells,18 the reduction of thiamine cellular use secondary to hyponatremia and/or hypomagnesemia,19 and the reduction of thiamine phosphorylation20 have all been associated with thiamine deficiency caused by furosemide. Therefore, thiamine analysis in the present study included measurements of free thiamine and thiamine mono- and pyrophosphate esters, both in the intravascular (plasma and erythrocytes) and tissue compartments (myocardium). Plasma electrolytes and renal function were also studied to eliminate any influence they were exerting on thiamine levels and myocardial structure. Methods Animals and Treatments Twenty-four male 75-day-old Wistar rats were used, and all procedures were carried out in accordance with conventional guidelines for experimentation with animals (National Institutes of Health Pu. N 85-23, revised 1996). The experimental protocols were approved by the local committee. The animals had free access to fresh water and purified rat chow (American Institute of Nutrition [AIN]-93M) containing 5 mg of thiamine per kilogram21 for the first 7 days. After these 7 days of adaptation and thiamine level equilibrium, the animals were anesthetized (intraperitoneal thiopental) and a blood sample was collected by cardiac puncture. Thiamine and its phosphate esters were measured in plasma and erythrocytes, and the erythrocyte transketolase activity (ETA) and TPP effect were analyzed. On the next day, four groups of six animals were randomly separated. This sample size meets guideline recommendations to minimize the number of animals in experiments as is emphasized by the local committee for experimental research. However, a sample size of five or more animals per group satisfies the statistical requirements for such experimental study as reported in the literature.22 Two groups continued to be fed with the AIN-93M standard chow (SC) (Rhoster, Sa˜o Paulo, Brazil; www.rhoster.com.br), and the other two groups were fed AIN-93M thiamine-deficient chow (TD) (0.63 mg/kg, Rhoster). Furosemide (F, Lasix, Aventis,
da Cunha et al
775
Sa˜o Paulo, Brazil), in a daily dose of 30 mg/kg1, was given intraperitoneally to one group of animals fed standard chow (SCF) and to one group of animals fed thiamine-deficient chow (TDF) for 30 days, beginning on day 9 of the experiment. It is noteworthy to say that the same volume of saline (NaCl 0.9% in water) was administered intraperitoneally to the matched untreated groups SC and TD. Daily chow intake was measured for each rat in the treated groups (SCF and TDF), and on the next day an identical amount of chow was offered to the matched untreated groups (SC and TD). The body mass was verified weekly. Given that studies have shown that furosemide causes loss of appetite in rats, paired feeding was used to diminish this effect. Urine volumes were measured on the first day of intraperitoneal injection, and this material was processed for thiamine urinary excretion determination. Light and Electron Microscopy Twenty-four hours after the 30th injection, the animals were deeply anesthetized (intraperitoneal thiopental), the thorax was opened, and the blood samples were collected by right atria puncture. The heart was removed, and small fragments of the left ventricle (LV) were obtained for biochemical analysis and electron microscopy (immersed in fixative for 2 hours, cacodylate 0.1 mol/L, pH 7.2 buffered glutaraldehyde 2.5%, followed by 1 hour in cacodylate buffer and 1% osmium tetroxide). The material was embedded in resin (Polysciences, Niles, Illinois), cut with an ultramicrotome (Ultracult UCT 020 Leica; Wetzlar, Germany), mounted on copper grids (Sigma-Aldrich, St. Louis, Missouri), and counterstained with uranyl acetate23 and lead citrate.24 The specimens were examined and photographed under a LEO 906 transmission electron microscope (Carl Zeiss, Oberkochen, Germany). Other LV fragments from free wall and interventricular septum were kept for 48 hours at room temperature in fixative (freshly prepared 1.27 mol/L formaldehyde in 0.1 mol/L phosphate buffer, pH 7.2),25 embedded in Paraplast plus (Sigma-Aldrich), and sectioned at 3 and 10-mm thicknesses for light microscopy and stereology. Sections were stained with hematoxylin-eosin and PicroSirius red. Five random fields per animal were analyzed. Stereology A video microscopic system composed of a Leica DMRBE microscope, Kappa videocamera (Gleichen, Germany), and Sony Trinitron monitor (Pencoed, United Kingdom) was used. The intramyocardial arteries-to-cardiomyocyte ratio ([ima]/[cmy]), calculated by the ratio between the volume densities (Vv) of arteries and cardiomyocytes, was estimated by point counting with a 36-point test system. The Vv[structure]: 5 Pp[structure]/ PT, where Pp are the points hitting the structure and PT are the total number of points in the test system. The cardiomyocyte mean cross-sectional area as A[cmy]: 5 Vv[cmy]/2.QA[cmy] was also calculated, where QA is the number of cardiomyocyte profiles in the frame. The ‘‘optical disector’’ was used to estimate the number of cardiomyocyte nuclei (cmyn) in the LV. Briefly, the numeric density of cmyn (number of nuclei per cubic millimeter) was estimated from 10 random disector pairs for each animal. For reasons of efficiency, we analyzed the cmyn (one nucleus represented one cmy) in the test area (AT) (Fig. 1). The thickness of the disector (t) was 3 mm, representing one fourth to one third of the height of the cmyn, which in practice is the more important constraint on section thickness.26,27 The total number of cardiomyocyte nuclei
776 Journal of Cardiac Failure Vol. 13 No. 9 November 2007
Fig. 1. Disector method. Upper (A) and lower (B) focal planes of a ‘‘dissector’’ 3 mm distant, respectively. Five cardiomyocyte nuclei (numbered ) and two fibroblast nuclei (arrows) are seen in A, but only nuclei 1 and 2 are in focus. Nuclei 3 to 5 are in focus in B, but nucleus 3 is crossed by ‘‘forbidden line’’ and must not be considered. Therefore, we count nuclei 1 and 2 in reference plane A and nuclei 4 and 5 in reference plane B. See text for more details.
(N[cmyn]) was calculated by multiplying the numeric density of cmyn by the LV volume,28 previously estimated.
Nv½cmyn : 5Q½cmyn =t$AT
and accuracy, was 5 nmol/L. Precision was established by duplicate analyses of the most important thiamine forms in each compartment, which revealed the following correlation coefficients: a) plasma: free thiamine, r 5 0.97 (P ! .0001), thiamine monophosphate, r 5 0.99 (P ! .0001); b) erythrocyte: TPP, r 5 0.97 (P ! .0001); c) heart: TPP, r 5 0.97 (P ! .0001).
Biochemical assays
Electrolytes and Renal Function
Thiamine Status Analysis. Blood samples were conserved under refrigeration and processed on the same day of collection. Plasma was separated from cells by centrifugation at 3000g. Cells were washed three times with chilly saline 0.9%. Erythrocytes were divided into two aliquots, one for ETA and the TPP effect determinations, using the method described by Brin,29 and one for thiamine and its mono and pyrophosphate esters measurements by high-performance liquid chromatography (HPLC). In this last aliquot, trichloroacetic acid (TCA) 40% was used for plasma and erythrocytes aliquots deproteination. TCA was extracted with five volumes of ethylic ether (water saturated). The inferior watery fraction was collected for derivatization by adding 50 mL of potassium ferricyanide (30.4 mmol/L) and 50 mL of sodium hydroxide 0.8 mol/L for each aliquot milliliter to form the thiochrome derivatives. Then, 100 mL of the resulting yellow solution was injected onto the HPLC column (MacheneyNagel, GmbH & Co., Du¨ren, Germany). The HPLC system (Shimadzu Co, Kyoto, Japan) was used, composed of an LC-10AD pump, an FCV-10AL valve, an RF-535 fluorescence detector, a Chromatopac C-R6A integrator, and a Nucleosil 100-5 NH2 (25 cm, 4.6 mm) column. The mobile phase, methanol 25% in potassium phosphate buffer 0.1 mol/L pH 7.5 (isocratic), was pumped at a flow rate of 1.5 mL/min. Thiochrome was detected by fluorimetry at 450-nm emission and 370-nm excitation wavelengths. Standards equivalent to 5, 10, 50, 100, 200, and 300 nM were prepared for calibration curves construction. The retention times were 3.6 minutes for free thiamine, 6.0 minutes for thiamine monophosphate, and 8.9 minutes for the pyrophosphate. The detection limit was 1 nmol/L, defined as the lowest value that could be distinguished from zero. The quantification limit, defined as the lowest value that could be measured with precision
Electrolytes, urea, and creatinine were measured in the Central Laboratory of the University Hospital (State University of Rio de Janeiro) by automated means. Urinary Thiamine Measurement Urine collected on the first day of intraperitoneal injection was kept frozen at 75 C after the addition of HCl 1 mol/L (10% of urine volume). These samples were then thawed and analyzed after centrifugation at 2000g for 5 minutes. TCA 40% (5% of sample volume) was used for deproteination. TCA extraction and inferior watery fraction derivatization were carried out as described for blood samples. Afterward, 100 mL of the resulting yellow solution was injected onto the same HPLC column used for blood and tissue aliquots analysis. Heart Heart fragments from the LV apex were collected for thiamine determination, conserved in ice, and homogenized on the day of euthanasia in a Potter-Elvehjem apparatus for 10 minutes at 5 C and suspended in five volumes of potassium phosphate buffer 0.1 mol/L, pH 7.5, for each gram of wet heart. This procedure was followed by centrifugation at 16,000g for 30 minutes at 5 C. To quantify free thiamine and its phosphate ester, the supernatant was deproteinized with 40% TCA, derivatized, and injected into an HPLC as carried out for blood samples. Thiamine Deficiency Criterion Thiamine deficiency was defined as the occurrence of simultaneous low TPP and transketolase activity levels, and a TPP effect greater than 15%.2 TPP and transketolase activity levels were
Furosemide, Thiamine, and Cardiac Lesion considered low when they were less than 2 standard deviations from the mean value obtained from the initial measurements of all animals. The myocardium TPP level was considered for thiamine deficiency definition in this tissue; the cutoff was a value less than 2 standard deviations than the mean level in the SC group. Data Analysis One-way analysis of variance and the Newman-Keuls post hoc test were used to test differences among the groups for biochemical and biometrical values (pairwise comparisons), and the KruskalWallis nonparametric analysis of variance and Dunn’s post hoc test were used to test differences among the groups for stereologic values. The normality and homoscedasticity of the variances (Bartlett’s test) were analyzed before performing the analysis of variance. Biochemical comparison between matched groups was carried out using the t test for independent variables (Prism version 4.03, GraphPad Software, San Diego, California). The level of significance was fixed at .05.30
da Cunha et al
777
groups. Rats in the TD groups showed a significantly smaller chow intake than the animals in the SC groups from the third week up to the end of the experiment (Fig. 2). The rats in the TD groups showed significant body mass reduction from the third week until the end of the experiment (Fig. 3). However, the final body mass of matched groups was not different. Urinary volumes on the first day of intraperitoneal injection were as follows: saline groups, 13.4 6 2.5 mL/d1; furosemide groups, 28.8 6 3.1 mL/d1 (P 5 .001). Thiamine urinary excretion on the first day of intraperitoneal injection was as follows: saline groups, 1.71 6 0.33 nmol/h1; furosemide groups, 3.05 6 0.42 nmol/h1 (P 5 .02). Blood Assays
The results are shown as mean 6 standard error of the mean. During the experimental period, the average daily chow intake was different between the SC and TD groups (SC group 14.85 6 0.07 g and TD group 11.35 6 0.78 g [P 5 .0002]), but there was no difference between matched
Tables 1 and 2 show plasma and erythrocytes thiamine levels obtained after the 7-day period of AIN-93M standardchow intake (baseline values) and at euthanasia (final values). Baseline values showed that neither of the groups presented the association of low erythrocyte TPP and ETA with high TPP effect, more than 15%, which characterizes thiamine deficiency (Table 2; Fig. 4). At the end of the experiment all thiamine forms were significantly reduced only in the TDF and TD groups, both in plasma and erythrocytes. Erythrocyte TPP and plasma thiamine monophosphate levels were lower in the TDF group
Fig. 2. Chow intake (mean). Thiamine-deficient rats showed significant appetite loss after the fourth experimental week (third week of thiamine-deficient chow intake). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment.
Fig. 3. Body mass evolution (mean). Significant weight loss in thiamine-deficient rats was observed after the fourth experimental week (third week of thiamine-deficient chow intake). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment.
Results
778 Journal of Cardiac Failure Vol. 13 No. 9 November 2007 Table 1. Baseline and Final Plasma Levels of Thiamine and Its Phosphate Esters (Mean 6 Standard Error of the Mean) Free Thiamine (ng/100 mL) Group
Baseline
Final
SC SCF TD TDF
5.8 7.1 8.3 7.3
6 6 6 6
5.1 6 0.7 6.2 6 0.4 nd nd
0.3 0.4 0.8 0.8
Thiamine Monophosphate (ng/100 mL) Baseline 20.9 21.5 19.1 16.7
6 6 6 6
0.6 1.1 1.0 1.7
Final 19.3 22.7 1.5 0.8
6 6 6 6
2.8 0.3 0.1* 0.2*,y
Thiamine Pyrophosphate (ng/100 mL) Baseline
Final
6 6 6 6
nd nd nd nd
0.22 0.32 0.93 0.43
0.03 0.03 0.18 0.13
nd, not detected; SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide administration. Baseline values: blood collected by cardiac puncture after the first 7 days of AIN-93M chow intake. Final values: blood collected at euthanasia. At the end of the experiment, all forms of thiamine analyzed were significantly reduced only in the TDF and TD groups. *Different from baseline values of each thiamine ester (P ! .05). y Lower thiamine monophosphate levels were observed in the TDF group than in the TD group, P 5 .006).
compared with the TD group injected with saline. These results are depicted in Tables 1 and 2. Likewise, the same groups showed the lowest erythrocyte transketolase activities, with TPP effects greater than 15% and even lower enzyme activity in the TDF group (Fig. 4). This observation was associated with a higher TPP effect in the same group (40% vs 90%). Plasma biochemical analysis on the day of euthanasia did not show significant renal or electrolyte alterations (Table 3). Myocardial Thiamine
TPP represented 90% of total myocardial thiamine in the SC group, with no significant difference between the SCF and SC groups. The TD and TDF groups had low myocardial thiamine content. Only TPP was detected in these two groups, with significantly lower levels in furosemidetreated animals (Table 4).
A[cmy] went with LVmi, but TDF rats showed no difference compared with TD rats (Fig. 7). The [ima]/[cmy] ratio measured myocardial relative vascularization and showed smaller ratio in TDF rats than in TD and SCF rats, respectively (Fig. 8). Finally, at the end of the experiment, neither furosemide treatment nor thiamine deficiency had significantly altered the LV number of cardiomyocyte nuclei (Fig. 9). Animals fed with AIN-93M standard chow (SCF and SC groups) did not have any ultrastructural abnormality (Fig. 10A). Thiamine-deprived rats (TD and TDF groups) revealed focal mitochondrial abnormalities. Some mitochondria were shrunken and dense (Fig. 10, B and C). Occasional fibers presented membrane-lined vesicles containing mitochondrial debris in the form of recognizable cristae, and giant mitochondria were also observed, with swelling and crystal disintegration (Fig. 10D). Myofibrillary structure did not show any abnormality.
Heart Structure and Ultrastructure
Discussion
LV mass index (LVmi) showed two main differences among the groups: a) furosemide-treated rats showed smaller LVmi than untreated rats, and b) TD rats showed smaller LVmi than SC rats (Fig. 5). However, there were no great qualitative microscopic differences among the groups, but thiamine-deficient rats showed rare focal scars with inflammatory infiltrate (Fig. 6).
The association among furosemide treatment, thiamine deficiency, and cardiac failure has been investigated extensively in the last 27 years,4,6,10,15,17,20,31,32 but the jury is still out. Furosemide is a well-known potent diuretic. It acts primarily to inhibit chloride and sodium absorption in the
Table 2. Baseline and Final Erythrocyte Levels of Thiamine and Its Phosphate Esters (Mean 6 Standard Error of the Mean) Free Thiamine (ng/100 mL) Group SC SCF TD TDF
Baseline
Final
nd nd nd nd
nd nd nd nd
Thiamine Monophosphate (ng/100 mL) Baseline
Final
6 6 6 6
1.45 6 0.18* 1.14 6 0.11* nd nd
0.52 0.60 0.32 0.09
0.04 0.13 0.08 0.06
Thiamine Pyrophosphate (ng/100 mL) Baseline 14.84 12.33 14.02 13.57
6 6 6 6
1.08 0.40 0.77 0.42
Final 16.25 13.67 0.94 0.24
6 6 6 6
1.17 0.88 0.07* 0.07*,y
nd, not detected; SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide administration. Baseline values: blood collected by cardiac puncture after the first 7 days of AIN-93M chow intake. Final values: blood collected at euthanasia. Despite the difference found in final thiamine monophosphate between the SC and SCF groups, this is irrelevant because the thiamine ester has minor importance in the erythrocyte. *Different from baseline values of each thiamine ester (P ! .05). y Lower thiamine pyrophosphate levels were observed in the TDF group than in the TD group (P ! .0001).
Furosemide, Thiamine, and Cardiac Lesion
da Cunha et al
779
Fig. 4. ETA measurement at the beginning and end of the experiment. Bar 1 is ETA, and bar 2 is ETA after TPP addition in vitro. At baseline ETA, no group showed thiamine deficiency; there was no ETA increase more than 15% after TPP addition. At final ETA, groups TDF and TD showed ETA 1 reduction that was associated with a significant (15%) increase in activity after TPP addition in vitro (ETA 2), biochemical alterations that characterized thiamine deficiency, observed only in these groups. Greater thiamine deficiency was observed in the TDF group than in the TD group (increase of ETA after TPP addition in vitro: 90%). When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD), [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF), and [c] when compared ETA 1 versus ETA 2 from each group. SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; ETA, erythrocyte transketolase activity.
ascending limb of the renal loop of Henle and, as a consequence, causes intense water and salt excretion.33 The resultant electrolyte deficiency, secondary to urinary flow increment, is also well known and promptly investigated and treated by the majority of physicians. Nevertheless, physicians do not usually consider water-soluble vitamin depletion in this context. Thiamine is a water-soluble vitamin that is present in many sorts of foods. Body stores of thiamine are relatively low, and a regular intake is required because large doses are poorly absorbed. The human body stores a maximum of 30 mg of thiamine, or 30 times the daily requirement.1 Therefore, thiamine deficiency can develop after only 1 month of a thiamine-free diet. The importance of thiamine to cardiac function was stated many decades ago, when beriberi was described as a clinical condition, characterized by cardiac dysfunction and neurologic symptoms.3 Thiamine
deficiency associated with furosemide treatment has mainly been explained by urinary flow increment.7,8 There is concern about the impact of persisting subclinical or borderline thiamine deficiency on the health status of a population1 and on the aggravation of cardiac failure stemmed from other causes.20 Despite all those observations, thiamine supplementation for chronically furosemide-treated patients still raises questions. Many different reasons have been mentioned: a) difficulty to determine thiamine level consistently;15e17 b) difficulty in drawing definitive conclusions based on studies that lacked randomization involving limited numbers of patients;9 and c) experiments on young rats4,5 prone to thiamine deficiency because growth needs higher thiamine levels.1 Because thiamine deficiency induces well-known cardiovascular clinical conditions associated with important
Table 3. Plasma Biochemical Analysis at Euthanasia (Mean 6 Standard Error of the Mean) Urea
Creatinine
Naþ
Kþ
Caþþ
P-
Mgþþ
Cl-
Group
mg/dL
mg/dL
mEq/L
mEq/L
mEq/L
mEq/L
mEq/L
mEq/L
SC SCF TD TDF
31 41 30 75
6 6 6 6
6 8 8 20
0.5 0.5 0.5 0.5
6 6 6 6
0.05 0.04 0.05 0.08
146 144 146 134
6 6 6 6
2 2 1 5
4.3 4.1 3.5 3.5
6 6 6 6
0.3 0.9 0.2 0.8
9.5 9.5 8.9 9.6
6 6 6 6
0.7 0.2 0.5 0.6
4.7 5.9 4.2 4.6
6 6 6 6
0.8 0.7 0.4 0.6
1.9 2.4 2.2 2.5
6 6 6 6
0.2 0.3 0.2 0.1
107 107 105 85
6 6 6 6
2 3 2 4
SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide administration. Renal function was not affected by the diuretic; there were no significant electrolyte abnormalities.
780 Journal of Cardiac Failure Vol. 13 No. 9 November 2007 Table 4. Myocardial Thiamine and Its Phosphate Esters (Mean 6 Standard Error of the Mean)
Group SC SCF TD TDF
Free Thiamine (ng/100 mL)
Thiamine Monophosphate (ng/100 mL)
0.8 6 0.1 0.9 6 0.1 nd nd
6.6 6 1.0 13.5 6 3.6 nd nd
Thiamine Pyrophosphate (ng/100 mL) 160.2 164.4 27.6 2.0
6 6 6 6
24.8 17.6 6.4* 1.1*,y
nd, not detected; SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiaminedeficient chow with furosemide administration. *TD versus SC or TDF versus SCF, P ! .05. y TDF versus TD, P ! .05.
cardiac structural alterations,34 we decided to analyze the biochemical aspects involved with furosemide treatment using validated biochemical techniques for thiamine determination.1,35,36 Also, erythrocyte and myocardium TPP measurements, ETA, and the effect of thiamine on the recovering of transketolase activity were used to define thiamine deficiency. Structural alterations were also investigated by microscopy. Three-month-old rats fed with thiamine-deficient chow and submitted to furosemide treatment had significantly lower blood and myocardium thiamine levels, as well as ETA, with an increased thiamine effect on transketolase activity recovery. Rats treated with furosemide, but fed with enough thiamine, did not show thiamine deficiency.
Fig. 5. LV mass index. When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD) and [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; LV, left ventricular; SEM, standard error of the mean.
Previous studies with rats demonstrated that only furosemide doses greater than 20 mg$kg1$d1 or 30 mg$kg1$d1 induce thiamine deficiency, which was evaluated by erythrocyte TPP effect.17,31 We decided to use the largest of these doses to be sure that the possible absence of effect was not related to low doses. Furosemide-treated rats had significantly greater urine volumes and urinary thiamine excretion. This observation was also observed in healthy volunteers.8 Assaying free thiamine and its phosphate esters helped us to establish whether thiamine phosphorylation was impaired by furosemide or not. Where there is impairment, the free thiamine should be accumulated in the intracellular compartment with simultaneous low levels of the phosphorylated forms; however, this is not what we found. Thiamine uptake impairment did not take place in cardiomyocytes because there was significant intracellular thiamine; nonetheless, our present ongoing studies are investigating this question. The erythrocyte thiamine level resembled myocardium values, which was described decades ago.4,37 Likewise, our confirmation is of particular importance because it does not support a recent description18 that furosemide induced cardiomyocyte thiamine uptake impairment. We found that if there was thiamine uptake blockage, there would be low intracellular thiamine, which was not the case. Also, thiamine uptake blockage could occur with different intensity in erythrocyte. This is of major significance because ETA is still an important parameter for a diagnosis of thiamine deficiency in clinical practice. Electrolytes and renal function alterations interfere with thiamine organic content; electrolytes interfere with intestinal absorption, cellular uptake, and phosphorylation; and renal function interferes with thiamine excretion.7,38 However, the absence of significant alterations in these parameters in the studied groups ruled out their influence on the biochemical and structural changes observed in the present study. There was significant reduction in the thiamine level and transketolase activity in rat heart after furosemide treatment in animals with and without thiamine deprivation, according to Yui et al.4 However, the TPP effect was in the range that characterizes thiamine deficiency (O15%) only in the thiamine-deprived group. Moreover, in the previous article the animals studied were younger, which may partly explain the difference in the results of our study with mature animals. Appetite and weight loss have been associated with thiamine deficiency.4,12 In the present study, both the TDF and TD groups showed appetite and weight loss after the third week of experimentation. Although chow intake in the TDF group was greater than in the TD group, there was no difference in body masses at the end of the experiment. Even with greater chow intake, the low myocardial thiamine level found in the TDF group was significantly lower. Thiamine-deficient rats treated with furosemide showed smaller LVmi than the control groups. Because of thiamine importance for cellular energy, nucleic acids and reduced
Furosemide, Thiamine, and Cardiac Lesion
da Cunha et al
781
Fig. 6. Myocardium photomicrographs: groups SC (A), SCF (B), TD (C), and TDF (D). Hematoxylin-eosin stained, bar 5 50 mm. No significant microscopic differences are noted among the groups, but ischemic/fibrotic focal areas with inflammatory infiltrate can be identified in thiamine-deficient rats (C).
Fig. 7. Cardiomyocyte cross-sectional area. When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD) and [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; SEM, standard error of the mean; cmy, cardiomyocyte.
Fig. 8. Intramyocardial arteries-to-cardiomyocytes ratio. When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD) and [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; SEM, standard error of the mean; [ima]/[cmy], intramyocardial arteries-to-cardiomyocyte ratio.
782 Journal of Cardiac Failure Vol. 13 No. 9 November 2007
Fig. 9. LV cardiomyocyte nuclei number. There was no difference between studied groups. SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; SEM, standard error of the mean; cmyn, cardiomyocyte nuclei.
nicotinamide adenine dinucleotide phosphate synthesis, there was cell hypotrophy in the thiamine-deficient groups, reflected by low LVmi; thus, thiamine deficiency and furosemide treatment have acted synergistically in the induction of this alteration. Also, there was lower myocardial vascularization in the thiamine-deficient group treated with furosemide, reflecting the same synergism. Given that thiamine-deficient groups also showed smaller cross-sectional cardiomyocyte area, we cannot attribute this fact to myocardial fiber size. We used adult animals in our analysis of furosemide influence on myocardial thiamine status and structure to eliminate growth influences on both stereologic parameters and thiamine requirements.12,39 There was no significant difference in the myocardial ultrastructure when comparing furosemide-treated rats with untreated rats, but mitochondrial alterations were observed in both thiamine-deficient groups. Nevertheless, there was no difference in myocardial ultrastructure in very young thiamine-deprived rats.12 Our study presents a new perspective about the association between furosemide treatment and thiamine deficiency. All other studies that have analyzed this subject have considered only biochemical furosemide treatment consequences, suggesting that furosemide causes thiamine deficiency. However, none of them analyzed myocardial
Fig. 10. Myocardium electron micrographs (same magnification, bar 5 0.5 mm). A, Animals in SC group with normal myofibrillary and mitochondria structures. Animals in TD group (B) and TDF group (C) showing shrunken and dense mitochondria and normal myofibrillary structure. D, Animals in TDF group with membrane-lined vesicles containing mitochondrial debris in the form of recognizable cristae and some giant mitochondria, with swelling and crystal disintegration.
Furosemide, Thiamine, and Cardiac Lesion
structure, neither by light nor electronic microscopy, after furosemide treatment. This is the first time that myocardial structural alterations are described in association with furosemide treatment and the aggravation of thiamine deficiency. In addition, the present study shows that this happens only when thiamine ingestion is poor. We also compared blood and myocardial thiamine levels, and this allowed us to exclude possible furosemide-caused cellular thiamine uptake impairment described by others.18 We perceive some limitations to extrapolation of our results to medical clinical practice. Experimental studies have to consider practical aspects, such as animal model availability at a reasonable cost and ease of maintenance and handling. The production of animals induced to model chronic human conditions generally takes weeks or months rather than years. For these reasons, rats have been used in many cardiovascular studies. However, significantly greater drug doses are used in experimental studies with rats when compared with those used in human treatment, possibly as a consequence of different drug metabolisms. This was the case of this study. Also, we did not use a cardiac failure model, and we cannot be sure that we would have observed the same results in a different model. Even if we had done so, cardiac failure models do not mimic the natural evolution of human cardiovascular diseases. Nevertheless, experimental studies make possible the absolute control of many variables that are difficult to control in clinical experiments, such as food intake, pair feeding, body mass, and age of the involved individuals in the experiment. Furthermore, experimental studies make tissue analysis possible, which was one of the aims of this study. Because the kidneys rapidly clear excessive amounts of ingested thiamine and no evidence exists of thiamine toxicity by oral administration, we believe that all physicians who prescribe furosemide must consider our results. Cardiac failure may be associated with poor food intake, especially in the advanced stages of the disease and, considering low human thiamine storages, this could be a clinical example in which furosemide treatment might be associated with significant thiamine deficiency and aggravation of myocardial dysfunction. Conclusion Our results indicate that furosemide administration is not the primary cause of thiamine deficiency in rats with adequate thiamine intake. Furosemide aggravates thiamine deficiency only in situations associated with insufficient thiamine intake, causing cardiac structural alterations, such as myocardial fiber hypotrophy, poor microvascularization, and mitochondrial degeneration. Acknowledgments We thank Tathiany Marinho, Cristiane Martins Cardoso Salles, Lin Machado de Lima, and Roosevelt Aguiar Dias
da Cunha et al
783
for their skillful assistance and the comments raised by the referees that helped us to improve the article.
References 1. Tanphaichitr V. Thiamin. In: Shils ME, Olson JA, Sihe M, Ross AC, eds. Modern nutrition in health and disease. 9th ed. Baltimore: Williams & Wilkins; 1999:381e9. 2. Tanphaichitr V, Vimokesant SL, Dhanamitta S, Valyasevi A. Clinical and biochemical studies of adult beriberi. Am J Clin Nutr 1970;23: 1017e26. 3. McIntyre N, Stanley NN. Cardiac beriberi: two modes of presentation. Br Med J 1971;3:567e9. 4. Yui Y, Itokawa Y, Kawai C. Furosemide-induced thiamine deficiency. Cardiovasc Res 1980;14:537e40. 5. Ali BH, Al-Qarawi AA. Thiamin status in furosemide-treated rats. Pharmacol Res 2000;42:21e4. 6. da Cunha S, Albanesi Filho FM, da Cunha Bastos VL, Antelo DS, Souza MM. Thiamin, selenium, and copper levels in patients with idiopathic dilated cardiomyopathy taking diuretics. Arq Bras Cardiol 2002;79:454e65. 7. Lubetsky A, Winaver J, Seligmann H, et al. Urinary thiamine excretion in the rat: effects of furosemide, other diuretics, and volume load. J Lab Clin Med 1999;134:232e7. 8. Rieck J, Halkin H, Almog S, et al. Urinary loss of thiamine is increased by low doses of furosemide in healthy volunteers. J Lab Clin Med 1999;134:238e43. 9. Zenuk C, Healey J, Donnelly J, Vaillancourt R, Almalki Y, Smith S. Thiamine deficiency in congestive heart failure patients receiving long term furosemide therapy. Can J Clin Pharmacol 2003;10:184e8. 10. Hanninen SA, Darling PB, Sole MJ, Barr A, Keith ME. The prevalence of thiamin deficiency in hospitalized patients with congestive heart failure. J Am Coll Cardiol 2006;47:354e61. 11. Ashburn LL, Loury JV. Development of cardiac lesions in thiamine deficient rats. Arch Pathol 1944;37:27. 12. Mall G, Mattfeldt T, Mobius HJ, Leonhard R. Stereological study on the rat heart in chronic alimentary thiamine deficiencydabsence of myocardial changes despite starvation. J Mol Cell Cardiol 1986;18:635e43. 13. Blanc P, Boussuges A. [Is thiamine supplementation necessary in patient with cardiac insufficiency?]. Ann Cardiol Angeiol (Paris) 2001;50:160e8. 14. Constant J. Pearls and pitfalls in the use and abuse of diuretics for chronic congestive heart failure. Cardiology 1999;92:156e61. 15. Leevy WC, Soine LA, Huth MM, Fishbein DP. Thiamine deficiency in congestive heart failure. Am J Med 1992;93:705e6. 16. Yue QY, Beermann B, Lindstrom B, Nyquist O. No difference in blood thiamine diphosphate levels between Swedish Caucasian patients with congestive heart failure treated with furosemide and patients without heart failure. J Intern Med 1997;242:491e5. 17. Kwok T, Falconer-Smith JF, Potter JF, Ives DR. Thiamine status of elderly patients with cardiac failure. Age Ageing 1992;21:67e71. 18. Zangen A, Botzer D, Zangen R, Shainberg A. Furosemide and digoxin inhibit thiamine uptake in cardiac cells. Eur J Pharmacol 1998;361: 151e5. 19. Friedrich W. Thiamine, vitamin B1, aneurin. In: Friedrich W, ed. Vitamins. New York: Walter de Gruyter; 1988:338e41. 20. Seligmann H, Halkin H, Rauchfleisch S, et al. Thiamine deficiency in patients with congestive heart failure receiving long-term furosemide therapy: a pilot study. Am J Med 1991;91:151e5. 21. Reeves PG, Nielsen FH, Fahey GC Jr. AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet. J Nutr 1993;123:1939e51. 22. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey. Am J Physiol 1990;258:L148e56.
784 Journal of Cardiac Failure Vol. 13 No. 9 November 2007 23. Watson ML. Staining of tissue sections for electron microscopy with heavy metals. J Biophys Biochem Cytol 1958;4:475e8. 24. Reynolds ES. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J Cell Biol 1963;17:208e12. 25. Carson FL, Martin JH, Lynn JA. Formalin fixation for electron microscopy: a re-evaluation. Am J Clin Pathol 1973;59:365e73. 26. Aguila MB, Mandarim-de-Lacerda CA, Apfel MI. [Stereology of the myocardium in young and aged rats]. Arq Bras Cardiol 1998;70: 105e9. 27. Martini VC, Mandarim-de-Lacerda CA. Offspring myocardium alteration from dams submitted to nitric oxide synthesis inhibition during pregnancy. Int J Cardiol 2005;100:377e82. 28. Mandarim-de-Lacerda C. Stereological tools in biomedical research. Ann Brazil Acad Sci 2003;75:469e86. 29. Brin M. Transketolase (sedoheptulose-7-phosphate: d-glyceraldehyde3-phosphate dihydroxyacetonetransferase, EC 2.2.1.1.) and the TPP effect in assessing thiamine adequacy. In: McCormic DB WL, ed. Methods in enzymology, vitamins, and coenzymes. New York: Academic Press; 1970:125e33. 30. Zar J. Biostatistical analysis. 4th ed. Upper Saddle River: Prentice Hall; 1999. 31. Shimon I, Almog S, Vered Z, et al. Improved left ventricular function after thiamine supplementation in patients with congestive heart
32.
33.
34. 35.
36. 37. 38.
39.
failure receiving long-term furosemide therapy. Am J Med 1995;98: 485e90. Leslie D, Gheorghiade M. Is there a role for thiamine supplementation in the management of heart failure? Am Heart J 1996;131: 1248e50. Goodman LS, Hardman JG, Limbird LE, Gilman AG. Goodman & Gilman’s the pharmacological basis of therapeutics. 10th ed. New York: McGraw-Hill; 2001. Kawai C. The heart in nutritional deficiencies. Philadelphia: Imago Productions; 1995. Iwata H, Matsuda T, Tonomura H. Improved high-performance liquid chromatographic determination of thiamine and its phosphate esters in animal tissues. J Chromatogr 1988;450:317e23. Lynch PL, Young IS. Determination of thiamine by high-performance liquid chromatography. J Chromatogr A 2000;881:267e84. Brin M. Erythrocyte as a biopsy tissue for functional evaluation of thiamine adequacy. JAMA 1964;187:762e6. Gastaldi G, Cova E, Verri A, Laforenza U, Faelli A, Rindi G. Transport of thiamin in rat renal brush border membrane vesicles. Kidney Int 2000;57:2043e54. Clubb FJ Jr, Bishop SP. Formation of binucleated myocardial cells in the neonatal rat. An index for growth hypertrophy. Lab Invest 1984;50: 571e7.
Basic Science and Experimental Studies
Cardiac Alterations in Furosemide-treated Thiamine-deprived Rats ˜ O BOSCO SALLES, PhD,3 SERGIO DA CUNHA, MD,1,2 JAYME CUNHA BASTOS, PhD,3 JOA 3 ´ MARIA CRISTINA COSTA SILVA, BSc, VERA LUCIA FREIRE CUNHA BASTOS, PhD,3 AND CARLOS ALBERTO MANDARIM-DE-LACERDA, MD, PhD1 Rio de Janeiro, Brazil
ABSTRACT Background: Chronic administration of furosemide may induce thiamine deficiency and cause or aggravate myocardial dysfunction. Methods and Results: Wistar rats were divided into four groups according to food and treatment: (1) thiamine standard chow with intraperitoneal furosemide administration; (2) thiamine standard chow with intraperitoneal saline administration; (3) thiamine-deficient chow with intraperitoneal furosemide administration; and (4) thiamine-deficient chow with intraperitoneal saline administration. Thiamine status was evaluated by high-performance liquid chromatography determination in plasma, erythrocytes, and myocardium, and by erythrocyte transketolase activity and the thiamine pyrophosphate effect to recover transketolase activity. Left ventricular mass index, intramyocardial arteries-to-cardiomyocyte ratio, cardiomyocyte cross-sectional area, and cardiomyocyte nuclei number were estimated. Myocardial structure was also studied by transmission electronic microscopy. Group 3 showed significantly lower blood and myocardial thiamine levels, which was not observed in group 1. Left ventricular mass index, cardiomyocyte cross-sectional area, and intramyocardial arteries-to-cardiomyocyte ratio were smaller in thiaminedeficient and furosemide-treated rats. However, no significant variation was found in the number of cardiomyocyte nuclei among the groups. Transmission electronic microscopy showed mitochondrial alterations in the thiamine-deficient groups. Conclusion: The present results indicate that furosemide administration is not the primary cause of thiamine deficiency in rats with adequate thiamine intake. Furosemide aggravates thiamine deficiency only in situations associated with insufficient thiamine intake, causing cardiac structural alterations, such as myocardial fiber hypotrophy, poor microvascularization, and mitochondrial degeneration. (J Cardiac Fail 2007;13:774e784) Key Words: Diuretic, electronic microscopy, stereology, vitamin B1.
From the 1Laboratory of Morphometry and Cardiovascular Morphology, Medical Clinics Department of Medical Sciences School, and 3Laboratory of Biochemical Toxicology, Department of Biochemistry, Biomedical Center, State University of Rio de Janeiro, Brazil. Manuscript received March 17, 2007; revised manuscript received June 15, 2007; revised manuscript accepted June 19, 2007. Reprint requests: Carlos Alberto Mandarim-de-Lacerda, MD, PhD, Laborato´rio de Morfometria e Morfologia Cardiovascular, Centro Biome´dico, Universidade do Estado do Rio de Janeiro, Av 28 de Setembro 87 (fds)20551-030, Rio de Janeiro, RJ, Brasil. This study was supported by grants from The National Council for Scientific and Technological Development (CNPq, www.cnpq.br) and The Carlos Chagas Filho Foundation for Supporting Research in the Rio de Janeiro State (Faperj, www.faperj.br). 1071-9164/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.cardfail.2007.06.729
Thiamine exists as free thiamine, thiamine monophosphate, thiamine pyrophosphate (TPP), and thiamine triphosphate in animal tissues. The pyrophosphate form, acting as the coenzyme of alpha-keto acid decarboxylation and transketolation,1 represents approximately 80% of the total thiamine. Accordingly, TPP is essential for carbohydrate metabolism, helping to produce cellular energy and ribose, and protecting against tissue oxidative damage by maintaining reduced nicotinamide adenine dinucleotide phosphate. Measurements of TPP, transketolase activity, and the percentage of increase of transketolase activity after TPP addition in vitro, the TPP effect, have been used to identify thiamine deficiency.2 Thiamine deficiency causes
2
774
Furosemide, Thiamine, and Cardiac Lesion
cardiovascular and neurologic damage, which are clinically noticeable in beriberi. The classic cardiovascular picture involves high-output biventricular heart failure, peripheral vasodilatation, volume overload, tachycardia, and wide pulse pressure.3 Furosemide is used for treating congestive heart failure, albeit with some well-known side effects such as electrolytic disturbances and hypovolemia. In the last 25 years, experimental and clinical studies have suggested that furosemide can induce thiamine deficiency4 and possibly cause or aggravate myocardial dysfunction.5e10 Much as cardiac lesions caused by thiamine deficiency have been known for more than 60 years,11 significant myocardial alterations caused by thiamine deficiency occurring with furosemide administration are still under investigation.12e14 Although no connection between furosemide administration and thiamine deficiency has been found,15e17 the present work was undertaken to verify whether long-term furosemide administration can induce myocardial thiamine deficiency with consequential structural alterations. The impairment of thiamine uptake by cardiac cells,18 the reduction of thiamine cellular use secondary to hyponatremia and/or hypomagnesemia,19 and the reduction of thiamine phosphorylation20 have all been associated with thiamine deficiency caused by furosemide. Therefore, thiamine analysis in the present study included measurements of free thiamine and thiamine mono- and pyrophosphate esters, both in the intravascular (plasma and erythrocytes) and tissue compartments (myocardium). Plasma electrolytes and renal function were also studied to eliminate any influence they were exerting on thiamine levels and myocardial structure. Methods Animals and Treatments Twenty-four male 75-day-old Wistar rats were used, and all procedures were carried out in accordance with conventional guidelines for experimentation with animals (National Institutes of Health Pu. N 85-23, revised 1996). The experimental protocols were approved by the local committee. The animals had free access to fresh water and purified rat chow (American Institute of Nutrition [AIN]-93M) containing 5 mg of thiamine per kilogram21 for the first 7 days. After these 7 days of adaptation and thiamine level equilibrium, the animals were anesthetized (intraperitoneal thiopental) and a blood sample was collected by cardiac puncture. Thiamine and its phosphate esters were measured in plasma and erythrocytes, and the erythrocyte transketolase activity (ETA) and TPP effect were analyzed. On the next day, four groups of six animals were randomly separated. This sample size meets guideline recommendations to minimize the number of animals in experiments as is emphasized by the local committee for experimental research. However, a sample size of five or more animals per group satisfies the statistical requirements for such experimental study as reported in the literature.22 Two groups continued to be fed with the AIN-93M standard chow (SC) (Rhoster, Sa˜o Paulo, Brazil; www.rhoster.com.br), and the other two groups were fed AIN-93M thiamine-deficient chow (TD) (0.63 mg/kg, Rhoster). Furosemide (F, Lasix, Aventis,
da Cunha et al
775
Sa˜o Paulo, Brazil), in a daily dose of 30 mg/kg1, was given intraperitoneally to one group of animals fed standard chow (SCF) and to one group of animals fed thiamine-deficient chow (TDF) for 30 days, beginning on day 9 of the experiment. It is noteworthy to say that the same volume of saline (NaCl 0.9% in water) was administered intraperitoneally to the matched untreated groups SC and TD. Daily chow intake was measured for each rat in the treated groups (SCF and TDF), and on the next day an identical amount of chow was offered to the matched untreated groups (SC and TD). The body mass was verified weekly. Given that studies have shown that furosemide causes loss of appetite in rats, paired feeding was used to diminish this effect. Urine volumes were measured on the first day of intraperitoneal injection, and this material was processed for thiamine urinary excretion determination. Light and Electron Microscopy Twenty-four hours after the 30th injection, the animals were deeply anesthetized (intraperitoneal thiopental), the thorax was opened, and the blood samples were collected by right atria puncture. The heart was removed, and small fragments of the left ventricle (LV) were obtained for biochemical analysis and electron microscopy (immersed in fixative for 2 hours, cacodylate 0.1 mol/L, pH 7.2 buffered glutaraldehyde 2.5%, followed by 1 hour in cacodylate buffer and 1% osmium tetroxide). The material was embedded in resin (Polysciences, Niles, Illinois), cut with an ultramicrotome (Ultracult UCT 020 Leica; Wetzlar, Germany), mounted on copper grids (Sigma-Aldrich, St. Louis, Missouri), and counterstained with uranyl acetate23 and lead citrate.24 The specimens were examined and photographed under a LEO 906 transmission electron microscope (Carl Zeiss, Oberkochen, Germany). Other LV fragments from free wall and interventricular septum were kept for 48 hours at room temperature in fixative (freshly prepared 1.27 mol/L formaldehyde in 0.1 mol/L phosphate buffer, pH 7.2),25 embedded in Paraplast plus (Sigma-Aldrich), and sectioned at 3 and 10-mm thicknesses for light microscopy and stereology. Sections were stained with hematoxylin-eosin and PicroSirius red. Five random fields per animal were analyzed. Stereology A video microscopic system composed of a Leica DMRBE microscope, Kappa videocamera (Gleichen, Germany), and Sony Trinitron monitor (Pencoed, United Kingdom) was used. The intramyocardial arteries-to-cardiomyocyte ratio ([ima]/[cmy]), calculated by the ratio between the volume densities (Vv) of arteries and cardiomyocytes, was estimated by point counting with a 36-point test system. The Vv[structure]: 5 Pp[structure]/ PT, where Pp are the points hitting the structure and PT are the total number of points in the test system. The cardiomyocyte mean cross-sectional area as A[cmy]: 5 Vv[cmy]/2.QA[cmy] was also calculated, where QA is the number of cardiomyocyte profiles in the frame. The ‘‘optical disector’’ was used to estimate the number of cardiomyocyte nuclei (cmyn) in the LV. Briefly, the numeric density of cmyn (number of nuclei per cubic millimeter) was estimated from 10 random disector pairs for each animal. For reasons of efficiency, we analyzed the cmyn (one nucleus represented one cmy) in the test area (AT) (Fig. 1). The thickness of the disector (t) was 3 mm, representing one fourth to one third of the height of the cmyn, which in practice is the more important constraint on section thickness.26,27 The total number of cardiomyocyte nuclei
776 Journal of Cardiac Failure Vol. 13 No. 9 November 2007
Fig. 1. Disector method. Upper (A) and lower (B) focal planes of a ‘‘dissector’’ 3 mm distant, respectively. Five cardiomyocyte nuclei (numbered ) and two fibroblast nuclei (arrows) are seen in A, but only nuclei 1 and 2 are in focus. Nuclei 3 to 5 are in focus in B, but nucleus 3 is crossed by ‘‘forbidden line’’ and must not be considered. Therefore, we count nuclei 1 and 2 in reference plane A and nuclei 4 and 5 in reference plane B. See text for more details.
(N[cmyn]) was calculated by multiplying the numeric density of cmyn by the LV volume,28 previously estimated.
Nv½cmyn : 5Q½cmyn =t$AT
and accuracy, was 5 nmol/L. Precision was established by duplicate analyses of the most important thiamine forms in each compartment, which revealed the following correlation coefficients: a) plasma: free thiamine, r 5 0.97 (P ! .0001), thiamine monophosphate, r 5 0.99 (P ! .0001); b) erythrocyte: TPP, r 5 0.97 (P ! .0001); c) heart: TPP, r 5 0.97 (P ! .0001).
Biochemical assays
Electrolytes and Renal Function
Thiamine Status Analysis. Blood samples were conserved under refrigeration and processed on the same day of collection. Plasma was separated from cells by centrifugation at 3000g. Cells were washed three times with chilly saline 0.9%. Erythrocytes were divided into two aliquots, one for ETA and the TPP effect determinations, using the method described by Brin,29 and one for thiamine and its mono and pyrophosphate esters measurements by high-performance liquid chromatography (HPLC). In this last aliquot, trichloroacetic acid (TCA) 40% was used for plasma and erythrocytes aliquots deproteination. TCA was extracted with five volumes of ethylic ether (water saturated). The inferior watery fraction was collected for derivatization by adding 50 mL of potassium ferricyanide (30.4 mmol/L) and 50 mL of sodium hydroxide 0.8 mol/L for each aliquot milliliter to form the thiochrome derivatives. Then, 100 mL of the resulting yellow solution was injected onto the HPLC column (MacheneyNagel, GmbH & Co., Du¨ren, Germany). The HPLC system (Shimadzu Co, Kyoto, Japan) was used, composed of an LC-10AD pump, an FCV-10AL valve, an RF-535 fluorescence detector, a Chromatopac C-R6A integrator, and a Nucleosil 100-5 NH2 (25 cm, 4.6 mm) column. The mobile phase, methanol 25% in potassium phosphate buffer 0.1 mol/L pH 7.5 (isocratic), was pumped at a flow rate of 1.5 mL/min. Thiochrome was detected by fluorimetry at 450-nm emission and 370-nm excitation wavelengths. Standards equivalent to 5, 10, 50, 100, 200, and 300 nM were prepared for calibration curves construction. The retention times were 3.6 minutes for free thiamine, 6.0 minutes for thiamine monophosphate, and 8.9 minutes for the pyrophosphate. The detection limit was 1 nmol/L, defined as the lowest value that could be distinguished from zero. The quantification limit, defined as the lowest value that could be measured with precision
Electrolytes, urea, and creatinine were measured in the Central Laboratory of the University Hospital (State University of Rio de Janeiro) by automated means. Urinary Thiamine Measurement Urine collected on the first day of intraperitoneal injection was kept frozen at 75 C after the addition of HCl 1 mol/L (10% of urine volume). These samples were then thawed and analyzed after centrifugation at 2000g for 5 minutes. TCA 40% (5% of sample volume) was used for deproteination. TCA extraction and inferior watery fraction derivatization were carried out as described for blood samples. Afterward, 100 mL of the resulting yellow solution was injected onto the same HPLC column used for blood and tissue aliquots analysis. Heart Heart fragments from the LV apex were collected for thiamine determination, conserved in ice, and homogenized on the day of euthanasia in a Potter-Elvehjem apparatus for 10 minutes at 5 C and suspended in five volumes of potassium phosphate buffer 0.1 mol/L, pH 7.5, for each gram of wet heart. This procedure was followed by centrifugation at 16,000g for 30 minutes at 5 C. To quantify free thiamine and its phosphate ester, the supernatant was deproteinized with 40% TCA, derivatized, and injected into an HPLC as carried out for blood samples. Thiamine Deficiency Criterion Thiamine deficiency was defined as the occurrence of simultaneous low TPP and transketolase activity levels, and a TPP effect greater than 15%.2 TPP and transketolase activity levels were
Furosemide, Thiamine, and Cardiac Lesion considered low when they were less than 2 standard deviations from the mean value obtained from the initial measurements of all animals. The myocardium TPP level was considered for thiamine deficiency definition in this tissue; the cutoff was a value less than 2 standard deviations than the mean level in the SC group. Data Analysis One-way analysis of variance and the Newman-Keuls post hoc test were used to test differences among the groups for biochemical and biometrical values (pairwise comparisons), and the KruskalWallis nonparametric analysis of variance and Dunn’s post hoc test were used to test differences among the groups for stereologic values. The normality and homoscedasticity of the variances (Bartlett’s test) were analyzed before performing the analysis of variance. Biochemical comparison between matched groups was carried out using the t test for independent variables (Prism version 4.03, GraphPad Software, San Diego, California). The level of significance was fixed at .05.30
da Cunha et al
777
groups. Rats in the TD groups showed a significantly smaller chow intake than the animals in the SC groups from the third week up to the end of the experiment (Fig. 2). The rats in the TD groups showed significant body mass reduction from the third week until the end of the experiment (Fig. 3). However, the final body mass of matched groups was not different. Urinary volumes on the first day of intraperitoneal injection were as follows: saline groups, 13.4 6 2.5 mL/d1; furosemide groups, 28.8 6 3.1 mL/d1 (P 5 .001). Thiamine urinary excretion on the first day of intraperitoneal injection was as follows: saline groups, 1.71 6 0.33 nmol/h1; furosemide groups, 3.05 6 0.42 nmol/h1 (P 5 .02). Blood Assays
The results are shown as mean 6 standard error of the mean. During the experimental period, the average daily chow intake was different between the SC and TD groups (SC group 14.85 6 0.07 g and TD group 11.35 6 0.78 g [P 5 .0002]), but there was no difference between matched
Tables 1 and 2 show plasma and erythrocytes thiamine levels obtained after the 7-day period of AIN-93M standardchow intake (baseline values) and at euthanasia (final values). Baseline values showed that neither of the groups presented the association of low erythrocyte TPP and ETA with high TPP effect, more than 15%, which characterizes thiamine deficiency (Table 2; Fig. 4). At the end of the experiment all thiamine forms were significantly reduced only in the TDF and TD groups, both in plasma and erythrocytes. Erythrocyte TPP and plasma thiamine monophosphate levels were lower in the TDF group
Fig. 2. Chow intake (mean). Thiamine-deficient rats showed significant appetite loss after the fourth experimental week (third week of thiamine-deficient chow intake). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment.
Fig. 3. Body mass evolution (mean). Significant weight loss in thiamine-deficient rats was observed after the fourth experimental week (third week of thiamine-deficient chow intake). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment.
Results
778 Journal of Cardiac Failure Vol. 13 No. 9 November 2007 Table 1. Baseline and Final Plasma Levels of Thiamine and Its Phosphate Esters (Mean 6 Standard Error of the Mean) Free Thiamine (ng/100 mL) Group
Baseline
Final
SC SCF TD TDF
5.8 7.1 8.3 7.3
6 6 6 6
5.1 6 0.7 6.2 6 0.4 nd nd
0.3 0.4 0.8 0.8
Thiamine Monophosphate (ng/100 mL) Baseline 20.9 21.5 19.1 16.7
6 6 6 6
0.6 1.1 1.0 1.7
Final 19.3 22.7 1.5 0.8
6 6 6 6
2.8 0.3 0.1* 0.2*,y
Thiamine Pyrophosphate (ng/100 mL) Baseline
Final
6 6 6 6
nd nd nd nd
0.22 0.32 0.93 0.43
0.03 0.03 0.18 0.13
nd, not detected; SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide administration. Baseline values: blood collected by cardiac puncture after the first 7 days of AIN-93M chow intake. Final values: blood collected at euthanasia. At the end of the experiment, all forms of thiamine analyzed were significantly reduced only in the TDF and TD groups. *Different from baseline values of each thiamine ester (P ! .05). y Lower thiamine monophosphate levels were observed in the TDF group than in the TD group, P 5 .006).
compared with the TD group injected with saline. These results are depicted in Tables 1 and 2. Likewise, the same groups showed the lowest erythrocyte transketolase activities, with TPP effects greater than 15% and even lower enzyme activity in the TDF group (Fig. 4). This observation was associated with a higher TPP effect in the same group (40% vs 90%). Plasma biochemical analysis on the day of euthanasia did not show significant renal or electrolyte alterations (Table 3). Myocardial Thiamine
TPP represented 90% of total myocardial thiamine in the SC group, with no significant difference between the SCF and SC groups. The TD and TDF groups had low myocardial thiamine content. Only TPP was detected in these two groups, with significantly lower levels in furosemidetreated animals (Table 4).
A[cmy] went with LVmi, but TDF rats showed no difference compared with TD rats (Fig. 7). The [ima]/[cmy] ratio measured myocardial relative vascularization and showed smaller ratio in TDF rats than in TD and SCF rats, respectively (Fig. 8). Finally, at the end of the experiment, neither furosemide treatment nor thiamine deficiency had significantly altered the LV number of cardiomyocyte nuclei (Fig. 9). Animals fed with AIN-93M standard chow (SCF and SC groups) did not have any ultrastructural abnormality (Fig. 10A). Thiamine-deprived rats (TD and TDF groups) revealed focal mitochondrial abnormalities. Some mitochondria were shrunken and dense (Fig. 10, B and C). Occasional fibers presented membrane-lined vesicles containing mitochondrial debris in the form of recognizable cristae, and giant mitochondria were also observed, with swelling and crystal disintegration (Fig. 10D). Myofibrillary structure did not show any abnormality.
Heart Structure and Ultrastructure
Discussion
LV mass index (LVmi) showed two main differences among the groups: a) furosemide-treated rats showed smaller LVmi than untreated rats, and b) TD rats showed smaller LVmi than SC rats (Fig. 5). However, there were no great qualitative microscopic differences among the groups, but thiamine-deficient rats showed rare focal scars with inflammatory infiltrate (Fig. 6).
The association among furosemide treatment, thiamine deficiency, and cardiac failure has been investigated extensively in the last 27 years,4,6,10,15,17,20,31,32 but the jury is still out. Furosemide is a well-known potent diuretic. It acts primarily to inhibit chloride and sodium absorption in the
Table 2. Baseline and Final Erythrocyte Levels of Thiamine and Its Phosphate Esters (Mean 6 Standard Error of the Mean) Free Thiamine (ng/100 mL) Group SC SCF TD TDF
Baseline
Final
nd nd nd nd
nd nd nd nd
Thiamine Monophosphate (ng/100 mL) Baseline
Final
6 6 6 6
1.45 6 0.18* 1.14 6 0.11* nd nd
0.52 0.60 0.32 0.09
0.04 0.13 0.08 0.06
Thiamine Pyrophosphate (ng/100 mL) Baseline 14.84 12.33 14.02 13.57
6 6 6 6
1.08 0.40 0.77 0.42
Final 16.25 13.67 0.94 0.24
6 6 6 6
1.17 0.88 0.07* 0.07*,y
nd, not detected; SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide administration. Baseline values: blood collected by cardiac puncture after the first 7 days of AIN-93M chow intake. Final values: blood collected at euthanasia. Despite the difference found in final thiamine monophosphate between the SC and SCF groups, this is irrelevant because the thiamine ester has minor importance in the erythrocyte. *Different from baseline values of each thiamine ester (P ! .05). y Lower thiamine pyrophosphate levels were observed in the TDF group than in the TD group (P ! .0001).
Furosemide, Thiamine, and Cardiac Lesion
da Cunha et al
779
Fig. 4. ETA measurement at the beginning and end of the experiment. Bar 1 is ETA, and bar 2 is ETA after TPP addition in vitro. At baseline ETA, no group showed thiamine deficiency; there was no ETA increase more than 15% after TPP addition. At final ETA, groups TDF and TD showed ETA 1 reduction that was associated with a significant (15%) increase in activity after TPP addition in vitro (ETA 2), biochemical alterations that characterized thiamine deficiency, observed only in these groups. Greater thiamine deficiency was observed in the TDF group than in the TD group (increase of ETA after TPP addition in vitro: 90%). When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD), [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF), and [c] when compared ETA 1 versus ETA 2 from each group. SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; ETA, erythrocyte transketolase activity.
ascending limb of the renal loop of Henle and, as a consequence, causes intense water and salt excretion.33 The resultant electrolyte deficiency, secondary to urinary flow increment, is also well known and promptly investigated and treated by the majority of physicians. Nevertheless, physicians do not usually consider water-soluble vitamin depletion in this context. Thiamine is a water-soluble vitamin that is present in many sorts of foods. Body stores of thiamine are relatively low, and a regular intake is required because large doses are poorly absorbed. The human body stores a maximum of 30 mg of thiamine, or 30 times the daily requirement.1 Therefore, thiamine deficiency can develop after only 1 month of a thiamine-free diet. The importance of thiamine to cardiac function was stated many decades ago, when beriberi was described as a clinical condition, characterized by cardiac dysfunction and neurologic symptoms.3 Thiamine
deficiency associated with furosemide treatment has mainly been explained by urinary flow increment.7,8 There is concern about the impact of persisting subclinical or borderline thiamine deficiency on the health status of a population1 and on the aggravation of cardiac failure stemmed from other causes.20 Despite all those observations, thiamine supplementation for chronically furosemide-treated patients still raises questions. Many different reasons have been mentioned: a) difficulty to determine thiamine level consistently;15e17 b) difficulty in drawing definitive conclusions based on studies that lacked randomization involving limited numbers of patients;9 and c) experiments on young rats4,5 prone to thiamine deficiency because growth needs higher thiamine levels.1 Because thiamine deficiency induces well-known cardiovascular clinical conditions associated with important
Table 3. Plasma Biochemical Analysis at Euthanasia (Mean 6 Standard Error of the Mean) Urea
Creatinine
Naþ
Kþ
Caþþ
P-
Mgþþ
Cl-
Group
mg/dL
mg/dL
mEq/L
mEq/L
mEq/L
mEq/L
mEq/L
mEq/L
SC SCF TD TDF
31 41 30 75
6 6 6 6
6 8 8 20
0.5 0.5 0.5 0.5
6 6 6 6
0.05 0.04 0.05 0.08
146 144 146 134
6 6 6 6
2 2 1 5
4.3 4.1 3.5 3.5
6 6 6 6
0.3 0.9 0.2 0.8
9.5 9.5 8.9 9.6
6 6 6 6
0.7 0.2 0.5 0.6
4.7 5.9 4.2 4.6
6 6 6 6
0.8 0.7 0.4 0.6
1.9 2.4 2.2 2.5
6 6 6 6
0.2 0.3 0.2 0.1
107 107 105 85
6 6 6 6
2 3 2 4
SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide administration. Renal function was not affected by the diuretic; there were no significant electrolyte abnormalities.
780 Journal of Cardiac Failure Vol. 13 No. 9 November 2007 Table 4. Myocardial Thiamine and Its Phosphate Esters (Mean 6 Standard Error of the Mean)
Group SC SCF TD TDF
Free Thiamine (ng/100 mL)
Thiamine Monophosphate (ng/100 mL)
0.8 6 0.1 0.9 6 0.1 nd nd
6.6 6 1.0 13.5 6 3.6 nd nd
Thiamine Pyrophosphate (ng/100 mL) 160.2 164.4 27.6 2.0
6 6 6 6
24.8 17.6 6.4* 1.1*,y
nd, not detected; SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiaminedeficient chow with furosemide administration. *TD versus SC or TDF versus SCF, P ! .05. y TDF versus TD, P ! .05.
cardiac structural alterations,34 we decided to analyze the biochemical aspects involved with furosemide treatment using validated biochemical techniques for thiamine determination.1,35,36 Also, erythrocyte and myocardium TPP measurements, ETA, and the effect of thiamine on the recovering of transketolase activity were used to define thiamine deficiency. Structural alterations were also investigated by microscopy. Three-month-old rats fed with thiamine-deficient chow and submitted to furosemide treatment had significantly lower blood and myocardium thiamine levels, as well as ETA, with an increased thiamine effect on transketolase activity recovery. Rats treated with furosemide, but fed with enough thiamine, did not show thiamine deficiency.
Fig. 5. LV mass index. When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD) and [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; LV, left ventricular; SEM, standard error of the mean.
Previous studies with rats demonstrated that only furosemide doses greater than 20 mg$kg1$d1 or 30 mg$kg1$d1 induce thiamine deficiency, which was evaluated by erythrocyte TPP effect.17,31 We decided to use the largest of these doses to be sure that the possible absence of effect was not related to low doses. Furosemide-treated rats had significantly greater urine volumes and urinary thiamine excretion. This observation was also observed in healthy volunteers.8 Assaying free thiamine and its phosphate esters helped us to establish whether thiamine phosphorylation was impaired by furosemide or not. Where there is impairment, the free thiamine should be accumulated in the intracellular compartment with simultaneous low levels of the phosphorylated forms; however, this is not what we found. Thiamine uptake impairment did not take place in cardiomyocytes because there was significant intracellular thiamine; nonetheless, our present ongoing studies are investigating this question. The erythrocyte thiamine level resembled myocardium values, which was described decades ago.4,37 Likewise, our confirmation is of particular importance because it does not support a recent description18 that furosemide induced cardiomyocyte thiamine uptake impairment. We found that if there was thiamine uptake blockage, there would be low intracellular thiamine, which was not the case. Also, thiamine uptake blockage could occur with different intensity in erythrocyte. This is of major significance because ETA is still an important parameter for a diagnosis of thiamine deficiency in clinical practice. Electrolytes and renal function alterations interfere with thiamine organic content; electrolytes interfere with intestinal absorption, cellular uptake, and phosphorylation; and renal function interferes with thiamine excretion.7,38 However, the absence of significant alterations in these parameters in the studied groups ruled out their influence on the biochemical and structural changes observed in the present study. There was significant reduction in the thiamine level and transketolase activity in rat heart after furosemide treatment in animals with and without thiamine deprivation, according to Yui et al.4 However, the TPP effect was in the range that characterizes thiamine deficiency (O15%) only in the thiamine-deprived group. Moreover, in the previous article the animals studied were younger, which may partly explain the difference in the results of our study with mature animals. Appetite and weight loss have been associated with thiamine deficiency.4,12 In the present study, both the TDF and TD groups showed appetite and weight loss after the third week of experimentation. Although chow intake in the TDF group was greater than in the TD group, there was no difference in body masses at the end of the experiment. Even with greater chow intake, the low myocardial thiamine level found in the TDF group was significantly lower. Thiamine-deficient rats treated with furosemide showed smaller LVmi than the control groups. Because of thiamine importance for cellular energy, nucleic acids and reduced
Furosemide, Thiamine, and Cardiac Lesion
da Cunha et al
781
Fig. 6. Myocardium photomicrographs: groups SC (A), SCF (B), TD (C), and TDF (D). Hematoxylin-eosin stained, bar 5 50 mm. No significant microscopic differences are noted among the groups, but ischemic/fibrotic focal areas with inflammatory infiltrate can be identified in thiamine-deficient rats (C).
Fig. 7. Cardiomyocyte cross-sectional area. When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD) and [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; SEM, standard error of the mean; cmy, cardiomyocyte.
Fig. 8. Intramyocardial arteries-to-cardiomyocytes ratio. When groups were compared and the P value was less than .05, the following letters were used to identify the compared groups: [a] when compared with its counterpart group (SCF vs SC and TDF vs TD) and [b] when compared with its SC counterpart group (TD vs SC and TDF vs SCF). SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; SEM, standard error of the mean; [ima]/[cmy], intramyocardial arteries-to-cardiomyocyte ratio.
782 Journal of Cardiac Failure Vol. 13 No. 9 November 2007
Fig. 9. LV cardiomyocyte nuclei number. There was no difference between studied groups. SC, standard chow; SCF, standard chow with furosemide administration; TD, thiamine-deficient chow; TDF, thiamine-deficient chow with furosemide treatment; SEM, standard error of the mean; cmyn, cardiomyocyte nuclei.
nicotinamide adenine dinucleotide phosphate synthesis, there was cell hypotrophy in the thiamine-deficient groups, reflected by low LVmi; thus, thiamine deficiency and furosemide treatment have acted synergistically in the induction of this alteration. Also, there was lower myocardial vascularization in the thiamine-deficient group treated with furosemide, reflecting the same synergism. Given that thiamine-deficient groups also showed smaller cross-sectional cardiomyocyte area, we cannot attribute this fact to myocardial fiber size. We used adult animals in our analysis of furosemide influence on myocardial thiamine status and structure to eliminate growth influences on both stereologic parameters and thiamine requirements.12,39 There was no significant difference in the myocardial ultrastructure when comparing furosemide-treated rats with untreated rats, but mitochondrial alterations were observed in both thiamine-deficient groups. Nevertheless, there was no difference in myocardial ultrastructure in very young thiamine-deprived rats.12 Our study presents a new perspective about the association between furosemide treatment and thiamine deficiency. All other studies that have analyzed this subject have considered only biochemical furosemide treatment consequences, suggesting that furosemide causes thiamine deficiency. However, none of them analyzed myocardial
Fig. 10. Myocardium electron micrographs (same magnification, bar 5 0.5 mm). A, Animals in SC group with normal myofibrillary and mitochondria structures. Animals in TD group (B) and TDF group (C) showing shrunken and dense mitochondria and normal myofibrillary structure. D, Animals in TDF group with membrane-lined vesicles containing mitochondrial debris in the form of recognizable cristae and some giant mitochondria, with swelling and crystal disintegration.
Furosemide, Thiamine, and Cardiac Lesion
structure, neither by light nor electronic microscopy, after furosemide treatment. This is the first time that myocardial structural alterations are described in association with furosemide treatment and the aggravation of thiamine deficiency. In addition, the present study shows that this happens only when thiamine ingestion is poor. We also compared blood and myocardial thiamine levels, and this allowed us to exclude possible furosemide-caused cellular thiamine uptake impairment described by others.18 We perceive some limitations to extrapolation of our results to medical clinical practice. Experimental studies have to consider practical aspects, such as animal model availability at a reasonable cost and ease of maintenance and handling. The production of animals induced to model chronic human conditions generally takes weeks or months rather than years. For these reasons, rats have been used in many cardiovascular studies. However, significantly greater drug doses are used in experimental studies with rats when compared with those used in human treatment, possibly as a consequence of different drug metabolisms. This was the case of this study. Also, we did not use a cardiac failure model, and we cannot be sure that we would have observed the same results in a different model. Even if we had done so, cardiac failure models do not mimic the natural evolution of human cardiovascular diseases. Nevertheless, experimental studies make possible the absolute control of many variables that are difficult to control in clinical experiments, such as food intake, pair feeding, body mass, and age of the involved individuals in the experiment. Furthermore, experimental studies make tissue analysis possible, which was one of the aims of this study. Because the kidneys rapidly clear excessive amounts of ingested thiamine and no evidence exists of thiamine toxicity by oral administration, we believe that all physicians who prescribe furosemide must consider our results. Cardiac failure may be associated with poor food intake, especially in the advanced stages of the disease and, considering low human thiamine storages, this could be a clinical example in which furosemide treatment might be associated with significant thiamine deficiency and aggravation of myocardial dysfunction. Conclusion Our results indicate that furosemide administration is not the primary cause of thiamine deficiency in rats with adequate thiamine intake. Furosemide aggravates thiamine deficiency only in situations associated with insufficient thiamine intake, causing cardiac structural alterations, such as myocardial fiber hypotrophy, poor microvascularization, and mitochondrial degeneration. Acknowledgments We thank Tathiany Marinho, Cristiane Martins Cardoso Salles, Lin Machado de Lima, and Roosevelt Aguiar Dias
da Cunha et al
783
for their skillful assistance and the comments raised by the referees that helped us to improve the article.
References 1. Tanphaichitr V. Thiamin. In: Shils ME, Olson JA, Sihe M, Ross AC, eds. Modern nutrition in health and disease. 9th ed. Baltimore: Williams & Wilkins; 1999:381e9. 2. Tanphaichitr V, Vimokesant SL, Dhanamitta S, Valyasevi A. Clinical and biochemical studies of adult beriberi. Am J Clin Nutr 1970;23: 1017e26. 3. McIntyre N, Stanley NN. Cardiac beriberi: two modes of presentation. Br Med J 1971;3:567e9. 4. Yui Y, Itokawa Y, Kawai C. Furosemide-induced thiamine deficiency. Cardiovasc Res 1980;14:537e40. 5. Ali BH, Al-Qarawi AA. Thiamin status in furosemide-treated rats. Pharmacol Res 2000;42:21e4. 6. da Cunha S, Albanesi Filho FM, da Cunha Bastos VL, Antelo DS, Souza MM. Thiamin, selenium, and copper levels in patients with idiopathic dilated cardiomyopathy taking diuretics. Arq Bras Cardiol 2002;79:454e65. 7. Lubetsky A, Winaver J, Seligmann H, et al. Urinary thiamine excretion in the rat: effects of furosemide, other diuretics, and volume load. J Lab Clin Med 1999;134:232e7. 8. Rieck J, Halkin H, Almog S, et al. Urinary loss of thiamine is increased by low doses of furosemide in healthy volunteers. J Lab Clin Med 1999;134:238e43. 9. Zenuk C, Healey J, Donnelly J, Vaillancourt R, Almalki Y, Smith S. Thiamine deficiency in congestive heart failure patients receiving long term furosemide therapy. Can J Clin Pharmacol 2003;10:184e8. 10. Hanninen SA, Darling PB, Sole MJ, Barr A, Keith ME. The prevalence of thiamin deficiency in hospitalized patients with congestive heart failure. J Am Coll Cardiol 2006;47:354e61. 11. Ashburn LL, Loury JV. Development of cardiac lesions in thiamine deficient rats. Arch Pathol 1944;37:27. 12. Mall G, Mattfeldt T, Mobius HJ, Leonhard R. Stereological study on the rat heart in chronic alimentary thiamine deficiencydabsence of myocardial changes despite starvation. J Mol Cell Cardiol 1986;18:635e43. 13. Blanc P, Boussuges A. [Is thiamine supplementation necessary in patient with cardiac insufficiency?]. Ann Cardiol Angeiol (Paris) 2001;50:160e8. 14. Constant J. Pearls and pitfalls in the use and abuse of diuretics for chronic congestive heart failure. Cardiology 1999;92:156e61. 15. Leevy WC, Soine LA, Huth MM, Fishbein DP. Thiamine deficiency in congestive heart failure. Am J Med 1992;93:705e6. 16. Yue QY, Beermann B, Lindstrom B, Nyquist O. No difference in blood thiamine diphosphate levels between Swedish Caucasian patients with congestive heart failure treated with furosemide and patients without heart failure. J Intern Med 1997;242:491e5. 17. Kwok T, Falconer-Smith JF, Potter JF, Ives DR. Thiamine status of elderly patients with cardiac failure. Age Ageing 1992;21:67e71. 18. Zangen A, Botzer D, Zangen R, Shainberg A. Furosemide and digoxin inhibit thiamine uptake in cardiac cells. Eur J Pharmacol 1998;361: 151e5. 19. Friedrich W. Thiamine, vitamin B1, aneurin. In: Friedrich W, ed. Vitamins. New York: Walter de Gruyter; 1988:338e41. 20. Seligmann H, Halkin H, Rauchfleisch S, et al. Thiamine deficiency in patients with congestive heart failure receiving long-term furosemide therapy: a pilot study. Am J Med 1991;91:151e5. 21. Reeves PG, Nielsen FH, Fahey GC Jr. AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet. J Nutr 1993;123:1939e51. 22. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell biology: a brief survey. Am J Physiol 1990;258:L148e56.
784 Journal of Cardiac Failure Vol. 13 No. 9 November 2007 23. Watson ML. Staining of tissue sections for electron microscopy with heavy metals. J Biophys Biochem Cytol 1958;4:475e8. 24. Reynolds ES. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J Cell Biol 1963;17:208e12. 25. Carson FL, Martin JH, Lynn JA. Formalin fixation for electron microscopy: a re-evaluation. Am J Clin Pathol 1973;59:365e73. 26. Aguila MB, Mandarim-de-Lacerda CA, Apfel MI. [Stereology of the myocardium in young and aged rats]. Arq Bras Cardiol 1998;70: 105e9. 27. Martini VC, Mandarim-de-Lacerda CA. Offspring myocardium alteration from dams submitted to nitric oxide synthesis inhibition during pregnancy. Int J Cardiol 2005;100:377e82. 28. Mandarim-de-Lacerda C. Stereological tools in biomedical research. Ann Brazil Acad Sci 2003;75:469e86. 29. Brin M. Transketolase (sedoheptulose-7-phosphate: d-glyceraldehyde3-phosphate dihydroxyacetonetransferase, EC 2.2.1.1.) and the TPP effect in assessing thiamine adequacy. In: McCormic DB WL, ed. Methods in enzymology, vitamins, and coenzymes. New York: Academic Press; 1970:125e33. 30. Zar J. Biostatistical analysis. 4th ed. Upper Saddle River: Prentice Hall; 1999. 31. Shimon I, Almog S, Vered Z, et al. Improved left ventricular function after thiamine supplementation in patients with congestive heart
32.
33.
34. 35.
36. 37. 38.
39.
failure receiving long-term furosemide therapy. Am J Med 1995;98: 485e90. Leslie D, Gheorghiade M. Is there a role for thiamine supplementation in the management of heart failure? Am Heart J 1996;131: 1248e50. Goodman LS, Hardman JG, Limbird LE, Gilman AG. Goodman & Gilman’s the pharmacological basis of therapeutics. 10th ed. New York: McGraw-Hill; 2001. Kawai C. The heart in nutritional deficiencies. Philadelphia: Imago Productions; 1995. Iwata H, Matsuda T, Tonomura H. Improved high-performance liquid chromatographic determination of thiamine and its phosphate esters in animal tissues. J Chromatogr 1988;450:317e23. Lynch PL, Young IS. Determination of thiamine by high-performance liquid chromatography. J Chromatogr A 2000;881:267e84. Brin M. Erythrocyte as a biopsy tissue for functional evaluation of thiamine adequacy. JAMA 1964;187:762e6. Gastaldi G, Cova E, Verri A, Laforenza U, Faelli A, Rindi G. Transport of thiamin in rat renal brush border membrane vesicles. Kidney Int 2000;57:2043e54. Clubb FJ Jr, Bishop SP. Formation of binucleated myocardial cells in the neonatal rat. An index for growth hypertrophy. Lab Invest 1984;50: 571e7.