- Email: [email protected]
Echocardiographic measurement of cardiac valvular thickness in healthy cows, cows with bacterial endocarditis, and cows with cardiorespiratory diseases
Journal of Veterinary Cardiology (2013) 15, 253e261
www.elsevier.com/locate/jvc
Echocardiographic measurement of cardiac valvular thickness in healthy cows, cows with bacterial endocarditis, and cows with cardiorespiratory diseases ´bastien Buczinski, Dr Ve ´t, MSc a,*, Mohammad Tolouei, DVM, PhD b, Se Ali Rezakhani, DVM, PhD c, Mohamed Tharwat, DVM, PhD d a
Clinique Ambulatoire Bovine, De´partement des sciences cliniques, Faculte´ de Me´decine Ve´te´rinaire, Universite´ de Montre´al, Que´bec, Canada b Department of Clinical Science, University of Tabriz, Iran c Department of Clinical Studies, School of Veterinary Medicine, Shiraz University, Shiraz, Iran d Department of Animal Medicine, Faculty of Veterinary Medicine, Zagazig University, Zagazig, Egypt Received 18 February 2012; received in revised form 4 June 2013; accepted 1 August 2013
KEYWORDS Heart disease; Valvular; Endocardium; Echocardiography
Abstract Objective: To determine valvular thickness in healthy cows, cows with bacterial endocarditis, and cows with various cardiorespiratory diseases. Animals: 40 healthy Holstein adult cows (CONTROL), 6 adult cows with confirmed bacterial endocarditis (BE), and 10 cows with other cardiorespiratory disorders (NONBE). Methods: Prospective study using right transthoracic echocardiographic examination in CONTROL, BE and NONBE cows. The valvular thicknesses of all cows were assessed in four different locations for all cardiac valves, and the maximal value was used for further analysis. Results: The mean [standard deviation (SD)] maximal thicknesses of the tricuspid, mitral, aortic, and pulmonary valves in the CONTROL group were 0.69 0.10 cm, 0.85 0.21 cm, 0.72 0.17 cm, and 0.58 0.12 cm, respectively. The maximal valvular thicknesses were less than 0.97 cm for the tricuspid, less than 0.91 cm for pulmonary, less than 1.05 cm for the aortic, and less than 1.28 cm for the mitral. In BE cows, the maximal valvular thickness of affected valves (median: 4.22 cm; range: 2.52e6.97 cm) and non affected valves (median: 0.75 cm; range: 0.45e1.52 cm) were significantly different (P ¼ 0.0004). The maximal valvular
* Corresponding author. E-mail address: [email protected] (S. Buczinski). 1760-2734/$ - see front matter ª 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jvc.2013.08.001
254
S. Buczinski et al. thicknesses of the NONBE valves as well as the unaffected valves in the BE group were not significantly different compared to the CONTROL group valves. Conclusions: Using the mean 2SD formula for each valve in healthy cows, a thickness of tricuspid, mitral, aortic or pulmonary valves greater than 0.85 cm, 1.27 cm, 1.06 cm or 0.82 cm respectively should raise the suspicion of valvular bacterial endocarditis. ª 2013 Elsevier B.V. All rights reserved.
Abbreviations BE cTnI SD TEE TTE
bacterial endocarditis cardiac troponin I standard deviation transesophageal echocardiography transthoracic echocardiography
Introduction Although bovine heart diseases are not highly prevalent, their clinical impact is important because the majority of these diseases have poor prognoses.1,2 Bacterial endocarditis (BE) is the most frequent valvular disease in cattle.3e5 The diagnosis of BE may be difficult due to the absence of specific clinical findings in many cases. The most common clinical sign is persistent tachycardia alongside evidence of a chronic inflammatory process on routine bloodwork.1,3,6 Heart murmurs are not always detected in cases of BE and may lack specificity.7 The presence of heart failure is uncommon in cases of BE.1e4,6,8 Because of these challenges, ancillary tests are required to make a definitive diagnosis. In other species, echocardiography is the gold standard test used for antemortem diagnosis.9,10 Transthoracic echocardiography (TTE) allows for an assessment of the consequences of valvular regurgitation, specifically cardiac dilation.9,10 The major echocardiographic findings in cases of BE classically include irregular valvular thickening with various degrees of valvular regurgitation and secondary cardiac chamber enlargement.3,11,12 Due to the fairly low sensitivity and specificity of echo findings, a classification system based upon the presence of clinical findings and results of ancillary tests has been proposed in human cardiology. This classification scheme is known as the Duke criteria13 and it is based on the presence of two levels of diagnostic criteria, the major and minor criteria. Echocardiographic findings with evidence of endocardial involvement (such as oscillating structures, abscesses or new valve
dehiscence) are considered to be major criteria in the antemortem diagnosis of BE in humans.13 To our knowledge, these criteria have not been tested or validated for bovine cases of BE. However, “typical echocardiographic” findings are commonly used in the antemortem diagnosis of bovine BE. These include different subjective findings such as valvular thickening or vegetation, valvular irregularity and/or heterogeneous appearance.2e4,8,11,12 In cattle, a recent randomeffect meta-analysis of published studies in which BE was confirmed using necropsy as the gold standard revealed that TTE had a pooled sensitivity of 84.3% (95% confidence interval: 60.4e95.0%) for BE diagnosis.14 Most of the echocardiographic examinations in published cases, however, have been performed by people with extensive experience of bovine echocardiography in hospital settings. The echocardiographic findings in bovine BE consist of various types of valvular thickening. This definition is quite subjective, especially for noncardiology board-certified veterinarians. Bovine ultrasonography is generally performed by nonspecialists, especially in farm settings, and for this reason it is important to have more objective data to define the normal thickness of cardiac valves in healthy cattle vs. cattle with BE when performing TTE. Therefore, the first aim of this study was to determine the maximal valvular thickness of the mitral, tricuspid, pulmonic, and aortic valves in healthy cows. The second objective of this study was to compare those measurements with a group of cattle with confirmed or presumed BE, or other cardiorespiratory diseases, thereby defining what should be considered valvular thickening.
Animals, materials and methods The study protocol was accepted by the ethical committee of the Faculte ´ de Me ´decine Ve ´te ´rinaire ´thique et d’Utilisation des Animaux). (Comite ´ d’E Forty adult Holstein cows, determined to be free of endocardial disease after a thorough physical examination, cardiac auscultation, and echocardiogram, were used for the determination of
Bovine valvular thickness measurement normal valvular thickness. Holstein cows from 0 to 60 days in milk were randomly selected from four tie-stall dairy farms in the area of the bovine ambulatory clinic, St-Hyacinthe, Que ´bec, Canada. They were initially selected on the basis of the absence of any disease during early lactation, and the fact that they had not received medications in the prior 14 days. The cows were then examined, paying special attention to the cardiovascular system. Cows were excluded if they showed clinical signs of illness, including fever, anorexia, decreased milk production, mastitis, or lameness. A blood sample from each cow was collected from a coccygeal vessel into a heparin-lithium tube, and cardiac troponin I (cTnI) levels were immediately assessed using a handheld analyzer.e
Echocardiography Echocardiography in the CONTROL group A standard echocardiographic examination was performed using a 3.2-MHz phased array probe,f as previously described in standing cows using a right parasternal approach.15e17 The valvular thickness measurements were always performed when the valves were closed during ventricular systole for atrioventricular valves and during ventricular diastole for semilunar valves. Particular attention was paid to the quality of the image recorded, and to measuring the closed valves when the image quality was optimal using the cine-loop mode of the ultrasound. The mitral and tricuspid valve thicknesses were measured using the long-axis four-chamber view during systole (Fig. 1A); both the septal and free-wall leaflets were measured. The aortic valve thickness was measured using the right long-axis view of the left ventricular outflow tract during diastole (Fig. 1B); both the left and right coronary cusps were measured. The pulmonic valve thickness was measured using the right long-axis view of the right ventricular outflow tract during diastole (Fig. 1C); both the anterior and right cusps were measured. Four measurements were obtained for each valve. Two measurements (M1 and M4) were taken at the attachment of the valve to the mural endocardium, with special attention to avoid valvular chordae in the measurement (Fig. 2). Two measurements (M2 and M3) were taken at the free extremity of the valve. The maximal valvular thicknesses were determined for each cardiac e f
I-Stat 1, Abaxis, Union City, CA, USA. Logiqbook, General Electric, Wawatosa, WI, USA.
255 valve. The mean maximal valvular thicknesses of the pulmonary, aortic, tricuspid and mitral valves were then determined.
Echocardiography in the BE group Between January 2007 and July 2012, every cow presented to the bovine ambulatory clinic of the Faculte ´ de Me ´decine Ve ´te ´rinaire, St-Hyacinthe, suspected of having cardiac disease underwent a complete right parasternal echocardiographic examination by the same operator (SB) with the same ultrasound unit.f Every echocardiographic examination was stored for further analysis. Echocardiograms of 12 client-owned adult Holstein cows with a diagnosis of valvular BE were reviewed by the same operator (SB). All cows underwent standard right transthoracic echocardiography15e17 during which particular attention was paid to the visualization of both semilunar and atrioventricular valves. The diagnosis of BE was based on the history, physical examination, a serum biochemistry panel compatible with a chronic inflammatory process (e.g., hyperglobulinemia, hyperfibrinogenemia), complete blood cell count (e.g., nonregenerative anemia), echocardiographic, and necropsy findings. The echocardiographic status of each valve was established on the day of the initial examination. For 12 cows in which a diagnosis of BE was made during that period, only 6 underwent necropsy which enabled a confirmation of the diagnosis. The 6 other cows were not included in this study since the diagnosis of BE was only based on echocardiographic findings and was not confirmed by other ancillary tests. The thickness of each valve was measured three times on the echocardiogram. The echocardiographic classification of the valves was not changed after systematic measurements were performed. The determination of the valve status (affected vs. unaffected) was made by reviewing the necropsy findings of macroscopic evidence of endocarditis in the valves.
Echocardiography in the NONBE group The echocardiographic examination of 10 Holstein cows with other cardiorespiratory diseases (NONBE), were also used to assess the valvular thickness in comparison with the CONTROL and BE group. These cows were examined by the same operator (SB), using the same ultrasound unit.f A total of 3 sets of data were obtained for each group of cows and the means of these values were used for further analysis.
256
S. Buczinski et al.
Figure 1 A. Long-axis 4-chambers view. Determination of the mitral (MV) and tricuspid (TV) valve thicknesses when observing the closed valves during early ventricular systole. B. Right long-axis view of the left ventricular outflow tract. Determination of the aortic valvular thickness (Ao) during ventricular diastole. C. Right long-axis view of the right ventricular outflow tract. Determination of the pulmonic valve thickness (PV) during ventricular diastole.
Figure 2 Determination of the valvular thickness at different areas during echocardiography: application to the aortic valve. Two measures (M1 and M4) were taken at the attachment of the valve to the mural endocardium. Two measures (M2 and M3) were taken at the free extremity of the valve. The mean and maximum values were used for further analyses. Ao: aorta, LA: left atrium, LV: left ventricle, RV: right ventricle.
Bovine valvular thickness measurement
257
Table 1 The mean (SD), minimal and maximal values of valvular thicknesses determined by right transthoracic echocardiography in 40 healthy Holstein early lactating cows. Group (no of cases) CONTROL (n ¼ 40)
Valve Tricuspid
Mitral
Aortic
Pulmonary
Measurement
Valve thickness (cm) Mean
SD
Min
Max
0.49 0.48 0.55 0.61 0.67 0.64 0.67 0.80 0.62 0.49 0.53 0.66 0.46 0.37 0.39 0.52
0.16 0.14 0.13 0.13 0.17 0.19 0.17 0.17 0.14 0.16 0.20 0.15 0.14 0.11 0.15 0.14
0.23 0.25 0.25 0.41 0.35 0.28 0.41 0.43 0.34 0.18 0.23 0.39 0.16 0.21 0.16 0.26
0.83 0.84 0.78 0.97 1.01 1.11 1.05 1.28 0.84 0.78 0.91 1.05 0.82 0.74 0.71 0.91
M1 M2 M3 M4 M1 M2 M3 M4 M1 M2 M3 M4 M1 M2 M3 M4
(CONTROL) healthy cows free from heart diseases used as a control group; (M1, M4) valvular thickness near the attachment of the valve to the myocardium; (M2, M3) valvular thickness at free end of the valve.
Data analysis
Results
Descriptive statistics were collected from the CONTROL, BE, and NONBE groups. The agreement between the echocardiographic classification of each valve and the final status of the valve based on necropsy was analyzed using a Kappa (k) Cohen correlation test in the BE group. The AndersoneDarling test confirmed normality in each group (CONTROL, BE and NONBE). A mixed linear model, with valve type as the fixed factor and animal identification number as the random factor (to control for possible clustering effects among individuals that provided multiple measurements), was used to determine differences in mean maximal thickness as a function of valve type in the control group. To compare mean maximal valvular thickness for each valve type across the three groups, a linear model assuming unequal variances, with group as the fixed factor was used, excluding the affected valves. In the previous statistical tests, Tukey’s post hoc tests were used to examine differences between pairs of means. The maximal valvular thicknesses of affected and unaffected valves in the BE group were compared using a ManneWhitney test due to the low numbers of data. Statistical analyses were performed using commercial software.g The level of statistical significance was set at 0.05.
In the CONTROL group, the blood cTnI levels were compatible with the absence of myocardial injury (<0.05 ng/mL). This was confirmed by the echocardiographic examinations, which were normal. The echocardiographic data from the CONTROL, BE and NONBE valves were normally distributed as assessed by the Anderson-Darling test. The mean (SD), minimal, and maximal values of valvular thicknesses for each location (M1-M4) in the CONTROL group are presented in Table 1. The mean maximal valvular thicknesses are presented in Table 2. The mixed linear model indicated that significant differences existed among mean maximal valve thicknesses in the CONTROL group (P < 0.0001). Tukey post-hoc tests indicated that the mean maximal valve thicknesses were significantly different between all valve pairs except for the aortic and tricuspid valves in the CONTROL group (Fig. 3). In the BE group, 3 cows had tricuspid valve BE, 1 had mitral valve BE, 1 had pulmonary valve BE, and 1 had aortic valve BE. The diagnosis of BE was confirmed by gross necropsy with a perfect correlation between echocardiography and necropsy (k ¼ 1.0). Only one endocardial lesion was seen in each cow. The maximal thickness was significantly higher in affected vs. unaffected valves in the BE group (P ¼ 0.0004) (Fig. 4). The NONBE group was composed of 10 cows with various cardiorespiratory diseases including
g
SAS V.9.2, Cary, NC, USA.
258
S. Buczinski et al.
Table 2 Valvular thickness in 40 healthy Holstein adult cows, cows with bacterial endocarditis (BE) and cows with cardiorespiratory diseases not having BE. Groups Control (n ¼ 40) BE (n ¼ 6) Affected valvesb BE (n ¼ 6) Normal valves NONBE (n ¼ 10)
Valve thickness (cm) a
Mean SD Min Max Meana SD Min Max Meana SD Min Max Meana SD Min Max
TV
MV
Ao
PV
0.69 0.10 0.23 0.97 4.50 1.25 0.25 6.97 0.91 0.33 0.25 1.37 0.76 0.15 0.23 1.11
0.85 0.21 0.28 1.28 4.04 1.32 0.33 4.83 1.12 0.50 0.35 2.28 0.84 0.13 0.34 1.06
0.72 0.17 0.18 1.05 5.16 e 2.68 5.16 0.77 0.26 0.21 1.12 0.67 0.25 0.25 1.27
0.58 0.12 0.16 0.91 2.80 1.24 0.38 3.67 0.65 0.10 0.21 0.76 0.59 0.14 0.16 0.80
a
Mean of the maximal valvular thickness. Tricuspid valve (3), mitral valve (1), pulmonary valve (1), aortic valve (1); (Ao) aortic valve; (BE) bacterial endocarditis; (MV) mitral valve; (NONBE) non-bacterial endocarditis; (PV) pulmonary valve; (TV) tricuspid valve. b
cardiac forms of enzootic lymphoma (n ¼ 2), traumatic pericarditis (n ¼ 2), bronchopneumonia (n ¼ 2), idiopathic hemorrhagic pericarditis (n ¼ 1), ventricular septal defect (n ¼ 1), pleuritis (n ¼ 1), and cor pulmonale (n ¼ 1). The
Figure 3 Box and whiskers plot of the maximal valvular thickness in 40 healthy Holstein cows determined by right transthoracic echocardiography. There were significant differences in the maximal valve thicknesses (P < 0.0001). The number indicated between the different plots is the P value from the post-hoc Tukey tests to assess the differences between pairs of valves. * Significant P value.
Figure 4 Box and whiskers plot of the maximal thickness of endocarditic valves and non endocarditic valves using transthoracic echocardiography in cows with confirmed diagnosis of bacterial endocarditis. The 6 endocarditic valves (3 tricuspid, 1 aortic, 1 pulmonary and 1 mitral) were compared with the 18 unaffected valves in the 6 cows with a confirmed diagnosis of bacterial endocarditis. * Significant P value.
echocardiographic data of the BE and NONBE groups are presented in Table 2. The mean maximal tricuspid, mitral, pulmonary, and aortic thicknesses of unaffected valves of the BE group did not differ from the CONTROL or NONBE groups.
Discussion The most typical echocardiographic sign of BE is valvular thickening.3,11,12 In light of the low prevalence of noninfectious valvular disorders in cattle,5 echocardiographic evidence of thickening of the endocardium should be highly suggestive of BE and should therefore be considered a possible major criterion for the clinical diagnosis of BE. A major problem with this criterion is that the definition of thickening is mostly based on operator experience. In cattle, it is of particular importance to have a more objective definition of thickening since echocardiography is generally performed by people who are not board-certified cardiologists. These clinicians need to know what valvular thickness qualifies as normal and abnormal. This will have an important impact on case management. The cardiac valves are composed of a thin (typically less than 1 mm) fibroconnective lamina that is enveloped in endocardial cells.18 They are attached to the myocardium with the tendinous chordae apparatus for the mitral and tricuspid valves, or the pulmonary or aortic trunk for the
Bovine valvular thickness measurement semilunar valves.19 The high valvular thicknesses that were obtained in this study for healthy cows without heart disease may be due to several factors. First, the absence of concomitant ECG recording during the echocardiographic examination and the method of measuring the closed valves using the cine-loop mode to obtain the best view may have introduced some variability to our measurements. Second, the low frequency phased array probe that was used in this study may have had an impact on the values that were reported. Due to the depth of the measured structures and the potential interference of chordae tendinae and accessory leaflets when the measurement was performed, we may have overestimated valvular thickness. This overestimation is more likely to have occurred for the mitral valve measurement, since the mitral valve is the deepest valve when performing right sided TTE. The depth required to perform TTE in adult Holstein cows (25e30 cm) and the presence of distorsion may have had an impact on the high mitral valve thickness found in this study. The lower resolution of the low frequency phased array probe, as well as side-lobe artifacts which are commonly encountered with phased array probes,18 may have also limited our ability to accurately measure the true mitral valve thickness. However, this right-sided technique is the most practical in cattle due to the relatively good acoustic window with the lack of lung tissue at this location.12,15e17 The left-sided TTE approach would certainly have improved the measurement of the mitral valve, likely leading to lower thickness values than from right-sided TTE as this valve is more superficial on the left, limiting distortion.12,15,16 Post mortem identification of infected valves represents the gold standard for defining the exact localization and extension of BE. In our study, we used this criterion for inclusion so that the pathologic findings could be compared with echocardiographic findings. Cases with presumed BE based on clinical signs and echocardiographic examination were not included in the study, although they represented 50% of the BE cases diagnosed during this period at our institution. These cases may represent more accurately the typical presumptive BE diagnosis in a field context because once the diagnosis is made, and depending on the gravity of the clinical signs, the producer might choose to slaughter the animal before any attempt at therapy is made, despite the higher risk of carcass condemnation without a postmortem examination.1 This study is therefore of importance to establish the normal variation of valvular thickness in healthy cows and cows
259 without valvular problems, in order to be able to rule-in or rule-out BE in cows using a more objective assessment of valvular thickness. Although TTE is an important diagnostic tool for detecting BE in cattle, the use of echocardiography for the diagnosis of valvular BE in humans has limitations.20 The sensitivity of TTE in the diagnosis of BE is 60e70%,10 but the negative predictive value of normal TTE in the evaluation of possible valve endocarditis is more than 90% in humans.21 Transesophageal echocardiography (TEE) is a more sensitive tool in humans since it allows for a more precise evaluation of slight changes in valvular morphology. It is especially useful for detecting small endocardial vegetations.22 However, TEE has never been used for the assessment of bovine endocarditis and has only been used for describing mediastinal structures in cattle.23 The apparent higher sensitivity of TTE in cattle for the diagnosis of BE (84.3%, 95% confidence interval: 60.4e95.0%)14 when compared to TTE in humans may be due to the more advanced stage of the disease when cows are examined. Most of the reports have been based on referred clinical cases and sensitivity has been assessed using necropsy as a gold standard; this may overestimate diagnostic accuracy because it represents a subpopulation of BE cases with advanced stage of the disease. It therefore is difficult to extrapolate this sensitivity for practical purposes since it may also be influenced by performance of the ultrasound unit as well as by the experience of the operators performing the study, especially in the absence of objective criteria for the echocardiographic characteristics of BE. For these reasons, the present study focused on establishing normal values for the thickness of bovine cardiac valves. Due to the high prevalence of BE in cows with valvular thickening, the positive predictive value of valvular thickening should be high for BE diagnosis. Our evaluation of a group of cows with various cardiopulmonary diseases was also interesting. From a clinical standpoint, these cows had diseases for which BE is a potential differential diagnosis. In other species, as well as in cattle, valvular thickening can be observed in patients with myxomatous processes, degenerative changes, tumors, blood cysts, thrombi, or imaging artifacts.10,24e27 For these reasons, it is important to keep in mind that echocardiographic findings alone are not sufficient to establish a diagnosis; they need to be assessed in conjunction with clinical findings and other ancillary tests (e.g., hematology, biochemistry and blood culture). In humans and small animals, multiple clinical and clinicopathological findings are needed to make a
260 clinical diagnosis of endocarditis. The Duke criteria consist of various major and minor criteria. The presence of two major criteria or one major criterion with three minor criteria are required to raise a clinical suspicion of BE. These criteria have been rigorously tested in humans but not in small animals. This type of approach has never been applied to the study of bovine BE since most of the available studies consist of case reports or case series, without including any cows in which BE was a differential diagnosis which was subsequently excluded by further diagnostic tests. In light of the low prevalence of noninfectious valvular disease in cows,5 the results of our study suggest that valvular thickening should be considered as a major criterion for a diagnosis of BE in cows. Data concerning bovine cardiac diseases are still scant due to the low prevalence of these diseases and the advanced stage of the diseases when diagnoses are made. For these reasons, most cows diagnosed with heart disease are culled or euthanized without being sent to a referral center or without echocardiographic confirmation of the diagnosis. With the increasing availability of good quality portable ultrasound units and their increasing use outside of reproduction, it is important to collect specific data concerning bovine echocardiography.
Conclusions From a practical standpoint, a valve thickness in cows greater than 0.85 cm (tricuspid), 1.27 cm (mitral), 1.06 cm (aortic) or 0.82 cm (pulmonic) should raise the possibility of BE when right TTE is performed with a phased array probe. Future studies in cows with confirmed BE and cows in which BE was an initial differential diagnosis but was then excluded, need to be performed in order to improve the diagnostic accuracy of TTE, which represents a clinical challenge for the internist as well as for the bovine practitioner, especially in the early stages of disease.
Conflict of interest The authors have no conflicts of interest.
Acknowledgments The authors would like to acknowledge Mrs. Lucie Lapre ´vote and Mrs. Katia Juste for their technical
S. Buczinski et al. help during the echocardiographic study in healthy cows. The assistance of Mr. Guy Beauchamp in the statistical analysis of the data is greatly appreciated.
References 1. Bexiga R, Mateus A, Philbey AW, Ellis K, Barrett DC, Mellor DJ. Clinicopathological presentation of cardiac diseases in cattle and its impact on decision making. Vet Rec 2008;162:575e580. 2. Buczinski S, Francoz D, Fecteau G, DiFruscia R. Congestive heart failure in cattle: 59 cases (1990e2005). Can Vet J 2010;51:1123e1129. 3. Healy AM. Endocarditis in cattle: a review of 22 cases. Irish Vet J 1996;49:43e48. 4. Buczinski S, Rezakhani A, Boerboom D. Heart disease in cattle: a review of the current data regarding diagnosis, therapeutic approaches and prognosis. Vet J 2010;184: 258e263. 5. Reef VB, McGuirk SM. Valvular heart disease. In: Smith BP, editor. Large animal internal medicine. 4th ed. Philadelphia: Elsevier Saunders; 2009. p. 463e468. 6. Power HT, Rebhun WC. Bacterial endocarditis in adult dairy cattle. J Am Vet Med Assoc 1983;182:806e808. 7. Rezakhani A, Zarifi M. Auscultatory findings of cardiac murmurs in clinically healthy cattle. Online J Vet Res 2007; 11:62e66. 8. Buczinski S, Francoz D, Fecteau G, DiFruscia R. A study of heart diseases without clinical signs of heart failure in 47 cattle. Can Vet J 2010;51:1239e1246. 9. Kittleson MD, Kienle RD. Infective endocarditis (and annuloaortic ectasia). In: Kittleson MD, Kienle RD, editors. Small animal cardiovascular medicine. St. Louis: Mosby; 1998. p. 402e412. 10. Feigenbaum H, Armstrong WF, Ryan T. Infective endocarditis. In: Feigenbaum H, Armstrong WF, Ayan T, editors. Feigenbaum’s echocardiography. 6th ed. Philadelphia: Lippincott William and Wilkins; 2005. p. 375e398. 11. Yamaga Y, Too K. Diagnostic ultrasound imaging of vegetative valvular endocarditis in cattle. Jpn J Vet Res 1987;35: 49e63. 12. Buczinski S. Cardiovascular ultrasonography in cattle. Vet Clin North Am Food Anim Pract 2009;25:611e632. 13. Durack DT, Lukes AS, Bright DK. Duke endocarditis service: new criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994;96:200e209. 14. Buczinski S, Tsuka T, Tharwat M. The diagnostic criteria used in bovine bacterial endocarditis: a meta-analysis of 460 published cases from 1973 to 2011. Vet J 2012;193: 349e357. 15. Braun U, Schweizer T, Pusterla N. Echocardiography of the normal bovine heart: technique and ultrasonographic appearance. Vet Rec 2001;148:47e51. 16. Hallowell GD, Potter TJ, Bowen IM. Methods and normal values for echocardiography in adult dairy cattle. J Vet Cardiol 2007;9:91e98. 17. Varga A, Schober KE, Holloman CH, Stromberg PC, Lakritz J, Rings DM. Correlation of serum cardiac troponin I and myocardial damage in cattle with monensin toxicosis. J Vet Intern Med 2009;23:1108e1116. 18. Barone R. Coeur. In: Barone R, editor. Anatomie compare ´e des mammife `res domestiques, Tome 5: Angiologie. Paris: Vigot; 1996. p. 5e102.
Bovine valvular thickness measurement
261
19. Feigenbaum H, Armstrong WF, Ryan T. Artifacts. In: Feigenbaum H, Armstrong WF, Ryan T, editors. Feigenbaum’s echocardiography. 6th ed. Philadelphia: Lippincott William and Wilkins; 2005. p. 29e32. 20. Erbel R, Rohmann S, Drexler M, Mohr-Kahaly S, Gerharz CD, Iversen S, Oelert H, Meyer J. Improved diagnostic value of echocardiography in patients with infective endocarditis by transesophageal approach. Eur Heart J 1988;9:43e53. 21. McDermott BP, Cunha BA, Choi D, Cohen J, Hage J. Transthoracic echocardiography (TTE): sufficiently sensitive screening test for native valve infective endocarditis (IE). Heart Lung 2011;40:358e360. 22. Kuhl HP, Hanrath P. The impact of transesophageal echocardiography in daily practice. Eur J Echocardiogr 2004;5: 455e458.
23. Braun U, Sicher D, Pusterla N. Ultrasonography of the lungs, pleura, and mediastinum in healthy cows. Am J Vet Res 1996;57:432e438. 24. Borgarelli M, Ha ¨ggstrom J. Canine degenerative myxomatous mitral valve disease: natural history, clinical presentation and therapy. Vet Clin North Am Small Anim Pract 2010;40:651e663. 25. Buczinski S, Badillo M, Be ´langer AM, Rezakhani A. Echocardiographic findings of tricuspid blood cyst mimicking bacterial endocarditis finding in a dairy cow. J Vet Cardiol 2011;13:283e286. 26. Brisville AC, Buczinski S, Chenier S, Francoz D. A cardiac vascular hamartoma in a calf: clinical, echocardiographic and pathologic findings. J Vet Cardiol 2012;14:377e380.
Available online at www.sciencedirect.com
ScienceDirect
www.elsevier.com/locate/jvc
Echocardiographic measurement of cardiac valvular thickness in healthy cows, cows with bacterial endocarditis, and cows with cardiorespiratory diseases ´bastien Buczinski, Dr Ve ´t, MSc a,*, Mohammad Tolouei, DVM, PhD b, Se Ali Rezakhani, DVM, PhD c, Mohamed Tharwat, DVM, PhD d a
Clinique Ambulatoire Bovine, De´partement des sciences cliniques, Faculte´ de Me´decine Ve´te´rinaire, Universite´ de Montre´al, Que´bec, Canada b Department of Clinical Science, University of Tabriz, Iran c Department of Clinical Studies, School of Veterinary Medicine, Shiraz University, Shiraz, Iran d Department of Animal Medicine, Faculty of Veterinary Medicine, Zagazig University, Zagazig, Egypt Received 18 February 2012; received in revised form 4 June 2013; accepted 1 August 2013
KEYWORDS Heart disease; Valvular; Endocardium; Echocardiography
Abstract Objective: To determine valvular thickness in healthy cows, cows with bacterial endocarditis, and cows with various cardiorespiratory diseases. Animals: 40 healthy Holstein adult cows (CONTROL), 6 adult cows with confirmed bacterial endocarditis (BE), and 10 cows with other cardiorespiratory disorders (NONBE). Methods: Prospective study using right transthoracic echocardiographic examination in CONTROL, BE and NONBE cows. The valvular thicknesses of all cows were assessed in four different locations for all cardiac valves, and the maximal value was used for further analysis. Results: The mean [standard deviation (SD)] maximal thicknesses of the tricuspid, mitral, aortic, and pulmonary valves in the CONTROL group were 0.69 0.10 cm, 0.85 0.21 cm, 0.72 0.17 cm, and 0.58 0.12 cm, respectively. The maximal valvular thicknesses were less than 0.97 cm for the tricuspid, less than 0.91 cm for pulmonary, less than 1.05 cm for the aortic, and less than 1.28 cm for the mitral. In BE cows, the maximal valvular thickness of affected valves (median: 4.22 cm; range: 2.52e6.97 cm) and non affected valves (median: 0.75 cm; range: 0.45e1.52 cm) were significantly different (P ¼ 0.0004). The maximal valvular
* Corresponding author. E-mail address: [email protected] (S. Buczinski). 1760-2734/$ - see front matter ª 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jvc.2013.08.001
254
S. Buczinski et al. thicknesses of the NONBE valves as well as the unaffected valves in the BE group were not significantly different compared to the CONTROL group valves. Conclusions: Using the mean 2SD formula for each valve in healthy cows, a thickness of tricuspid, mitral, aortic or pulmonary valves greater than 0.85 cm, 1.27 cm, 1.06 cm or 0.82 cm respectively should raise the suspicion of valvular bacterial endocarditis. ª 2013 Elsevier B.V. All rights reserved.
Abbreviations BE cTnI SD TEE TTE
bacterial endocarditis cardiac troponin I standard deviation transesophageal echocardiography transthoracic echocardiography
Introduction Although bovine heart diseases are not highly prevalent, their clinical impact is important because the majority of these diseases have poor prognoses.1,2 Bacterial endocarditis (BE) is the most frequent valvular disease in cattle.3e5 The diagnosis of BE may be difficult due to the absence of specific clinical findings in many cases. The most common clinical sign is persistent tachycardia alongside evidence of a chronic inflammatory process on routine bloodwork.1,3,6 Heart murmurs are not always detected in cases of BE and may lack specificity.7 The presence of heart failure is uncommon in cases of BE.1e4,6,8 Because of these challenges, ancillary tests are required to make a definitive diagnosis. In other species, echocardiography is the gold standard test used for antemortem diagnosis.9,10 Transthoracic echocardiography (TTE) allows for an assessment of the consequences of valvular regurgitation, specifically cardiac dilation.9,10 The major echocardiographic findings in cases of BE classically include irregular valvular thickening with various degrees of valvular regurgitation and secondary cardiac chamber enlargement.3,11,12 Due to the fairly low sensitivity and specificity of echo findings, a classification system based upon the presence of clinical findings and results of ancillary tests has been proposed in human cardiology. This classification scheme is known as the Duke criteria13 and it is based on the presence of two levels of diagnostic criteria, the major and minor criteria. Echocardiographic findings with evidence of endocardial involvement (such as oscillating structures, abscesses or new valve
dehiscence) are considered to be major criteria in the antemortem diagnosis of BE in humans.13 To our knowledge, these criteria have not been tested or validated for bovine cases of BE. However, “typical echocardiographic” findings are commonly used in the antemortem diagnosis of bovine BE. These include different subjective findings such as valvular thickening or vegetation, valvular irregularity and/or heterogeneous appearance.2e4,8,11,12 In cattle, a recent randomeffect meta-analysis of published studies in which BE was confirmed using necropsy as the gold standard revealed that TTE had a pooled sensitivity of 84.3% (95% confidence interval: 60.4e95.0%) for BE diagnosis.14 Most of the echocardiographic examinations in published cases, however, have been performed by people with extensive experience of bovine echocardiography in hospital settings. The echocardiographic findings in bovine BE consist of various types of valvular thickening. This definition is quite subjective, especially for noncardiology board-certified veterinarians. Bovine ultrasonography is generally performed by nonspecialists, especially in farm settings, and for this reason it is important to have more objective data to define the normal thickness of cardiac valves in healthy cattle vs. cattle with BE when performing TTE. Therefore, the first aim of this study was to determine the maximal valvular thickness of the mitral, tricuspid, pulmonic, and aortic valves in healthy cows. The second objective of this study was to compare those measurements with a group of cattle with confirmed or presumed BE, or other cardiorespiratory diseases, thereby defining what should be considered valvular thickening.
Animals, materials and methods The study protocol was accepted by the ethical committee of the Faculte ´ de Me ´decine Ve ´te ´rinaire ´thique et d’Utilisation des Animaux). (Comite ´ d’E Forty adult Holstein cows, determined to be free of endocardial disease after a thorough physical examination, cardiac auscultation, and echocardiogram, were used for the determination of
Bovine valvular thickness measurement normal valvular thickness. Holstein cows from 0 to 60 days in milk were randomly selected from four tie-stall dairy farms in the area of the bovine ambulatory clinic, St-Hyacinthe, Que ´bec, Canada. They were initially selected on the basis of the absence of any disease during early lactation, and the fact that they had not received medications in the prior 14 days. The cows were then examined, paying special attention to the cardiovascular system. Cows were excluded if they showed clinical signs of illness, including fever, anorexia, decreased milk production, mastitis, or lameness. A blood sample from each cow was collected from a coccygeal vessel into a heparin-lithium tube, and cardiac troponin I (cTnI) levels were immediately assessed using a handheld analyzer.e
Echocardiography Echocardiography in the CONTROL group A standard echocardiographic examination was performed using a 3.2-MHz phased array probe,f as previously described in standing cows using a right parasternal approach.15e17 The valvular thickness measurements were always performed when the valves were closed during ventricular systole for atrioventricular valves and during ventricular diastole for semilunar valves. Particular attention was paid to the quality of the image recorded, and to measuring the closed valves when the image quality was optimal using the cine-loop mode of the ultrasound. The mitral and tricuspid valve thicknesses were measured using the long-axis four-chamber view during systole (Fig. 1A); both the septal and free-wall leaflets were measured. The aortic valve thickness was measured using the right long-axis view of the left ventricular outflow tract during diastole (Fig. 1B); both the left and right coronary cusps were measured. The pulmonic valve thickness was measured using the right long-axis view of the right ventricular outflow tract during diastole (Fig. 1C); both the anterior and right cusps were measured. Four measurements were obtained for each valve. Two measurements (M1 and M4) were taken at the attachment of the valve to the mural endocardium, with special attention to avoid valvular chordae in the measurement (Fig. 2). Two measurements (M2 and M3) were taken at the free extremity of the valve. The maximal valvular thicknesses were determined for each cardiac e f
I-Stat 1, Abaxis, Union City, CA, USA. Logiqbook, General Electric, Wawatosa, WI, USA.
255 valve. The mean maximal valvular thicknesses of the pulmonary, aortic, tricuspid and mitral valves were then determined.
Echocardiography in the BE group Between January 2007 and July 2012, every cow presented to the bovine ambulatory clinic of the Faculte ´ de Me ´decine Ve ´te ´rinaire, St-Hyacinthe, suspected of having cardiac disease underwent a complete right parasternal echocardiographic examination by the same operator (SB) with the same ultrasound unit.f Every echocardiographic examination was stored for further analysis. Echocardiograms of 12 client-owned adult Holstein cows with a diagnosis of valvular BE were reviewed by the same operator (SB). All cows underwent standard right transthoracic echocardiography15e17 during which particular attention was paid to the visualization of both semilunar and atrioventricular valves. The diagnosis of BE was based on the history, physical examination, a serum biochemistry panel compatible with a chronic inflammatory process (e.g., hyperglobulinemia, hyperfibrinogenemia), complete blood cell count (e.g., nonregenerative anemia), echocardiographic, and necropsy findings. The echocardiographic status of each valve was established on the day of the initial examination. For 12 cows in which a diagnosis of BE was made during that period, only 6 underwent necropsy which enabled a confirmation of the diagnosis. The 6 other cows were not included in this study since the diagnosis of BE was only based on echocardiographic findings and was not confirmed by other ancillary tests. The thickness of each valve was measured three times on the echocardiogram. The echocardiographic classification of the valves was not changed after systematic measurements were performed. The determination of the valve status (affected vs. unaffected) was made by reviewing the necropsy findings of macroscopic evidence of endocarditis in the valves.
Echocardiography in the NONBE group The echocardiographic examination of 10 Holstein cows with other cardiorespiratory diseases (NONBE), were also used to assess the valvular thickness in comparison with the CONTROL and BE group. These cows were examined by the same operator (SB), using the same ultrasound unit.f A total of 3 sets of data were obtained for each group of cows and the means of these values were used for further analysis.
256
S. Buczinski et al.
Figure 1 A. Long-axis 4-chambers view. Determination of the mitral (MV) and tricuspid (TV) valve thicknesses when observing the closed valves during early ventricular systole. B. Right long-axis view of the left ventricular outflow tract. Determination of the aortic valvular thickness (Ao) during ventricular diastole. C. Right long-axis view of the right ventricular outflow tract. Determination of the pulmonic valve thickness (PV) during ventricular diastole.
Figure 2 Determination of the valvular thickness at different areas during echocardiography: application to the aortic valve. Two measures (M1 and M4) were taken at the attachment of the valve to the mural endocardium. Two measures (M2 and M3) were taken at the free extremity of the valve. The mean and maximum values were used for further analyses. Ao: aorta, LA: left atrium, LV: left ventricle, RV: right ventricle.
Bovine valvular thickness measurement
257
Table 1 The mean (SD), minimal and maximal values of valvular thicknesses determined by right transthoracic echocardiography in 40 healthy Holstein early lactating cows. Group (no of cases) CONTROL (n ¼ 40)
Valve Tricuspid
Mitral
Aortic
Pulmonary
Measurement
Valve thickness (cm) Mean
SD
Min
Max
0.49 0.48 0.55 0.61 0.67 0.64 0.67 0.80 0.62 0.49 0.53 0.66 0.46 0.37 0.39 0.52
0.16 0.14 0.13 0.13 0.17 0.19 0.17 0.17 0.14 0.16 0.20 0.15 0.14 0.11 0.15 0.14
0.23 0.25 0.25 0.41 0.35 0.28 0.41 0.43 0.34 0.18 0.23 0.39 0.16 0.21 0.16 0.26
0.83 0.84 0.78 0.97 1.01 1.11 1.05 1.28 0.84 0.78 0.91 1.05 0.82 0.74 0.71 0.91
M1 M2 M3 M4 M1 M2 M3 M4 M1 M2 M3 M4 M1 M2 M3 M4
(CONTROL) healthy cows free from heart diseases used as a control group; (M1, M4) valvular thickness near the attachment of the valve to the myocardium; (M2, M3) valvular thickness at free end of the valve.
Data analysis
Results
Descriptive statistics were collected from the CONTROL, BE, and NONBE groups. The agreement between the echocardiographic classification of each valve and the final status of the valve based on necropsy was analyzed using a Kappa (k) Cohen correlation test in the BE group. The AndersoneDarling test confirmed normality in each group (CONTROL, BE and NONBE). A mixed linear model, with valve type as the fixed factor and animal identification number as the random factor (to control for possible clustering effects among individuals that provided multiple measurements), was used to determine differences in mean maximal thickness as a function of valve type in the control group. To compare mean maximal valvular thickness for each valve type across the three groups, a linear model assuming unequal variances, with group as the fixed factor was used, excluding the affected valves. In the previous statistical tests, Tukey’s post hoc tests were used to examine differences between pairs of means. The maximal valvular thicknesses of affected and unaffected valves in the BE group were compared using a ManneWhitney test due to the low numbers of data. Statistical analyses were performed using commercial software.g The level of statistical significance was set at 0.05.
In the CONTROL group, the blood cTnI levels were compatible with the absence of myocardial injury (<0.05 ng/mL). This was confirmed by the echocardiographic examinations, which were normal. The echocardiographic data from the CONTROL, BE and NONBE valves were normally distributed as assessed by the Anderson-Darling test. The mean (SD), minimal, and maximal values of valvular thicknesses for each location (M1-M4) in the CONTROL group are presented in Table 1. The mean maximal valvular thicknesses are presented in Table 2. The mixed linear model indicated that significant differences existed among mean maximal valve thicknesses in the CONTROL group (P < 0.0001). Tukey post-hoc tests indicated that the mean maximal valve thicknesses were significantly different between all valve pairs except for the aortic and tricuspid valves in the CONTROL group (Fig. 3). In the BE group, 3 cows had tricuspid valve BE, 1 had mitral valve BE, 1 had pulmonary valve BE, and 1 had aortic valve BE. The diagnosis of BE was confirmed by gross necropsy with a perfect correlation between echocardiography and necropsy (k ¼ 1.0). Only one endocardial lesion was seen in each cow. The maximal thickness was significantly higher in affected vs. unaffected valves in the BE group (P ¼ 0.0004) (Fig. 4). The NONBE group was composed of 10 cows with various cardiorespiratory diseases including
g
SAS V.9.2, Cary, NC, USA.
258
S. Buczinski et al.
Table 2 Valvular thickness in 40 healthy Holstein adult cows, cows with bacterial endocarditis (BE) and cows with cardiorespiratory diseases not having BE. Groups Control (n ¼ 40) BE (n ¼ 6) Affected valvesb BE (n ¼ 6) Normal valves NONBE (n ¼ 10)
Valve thickness (cm) a
Mean SD Min Max Meana SD Min Max Meana SD Min Max Meana SD Min Max
TV
MV
Ao
PV
0.69 0.10 0.23 0.97 4.50 1.25 0.25 6.97 0.91 0.33 0.25 1.37 0.76 0.15 0.23 1.11
0.85 0.21 0.28 1.28 4.04 1.32 0.33 4.83 1.12 0.50 0.35 2.28 0.84 0.13 0.34 1.06
0.72 0.17 0.18 1.05 5.16 e 2.68 5.16 0.77 0.26 0.21 1.12 0.67 0.25 0.25 1.27
0.58 0.12 0.16 0.91 2.80 1.24 0.38 3.67 0.65 0.10 0.21 0.76 0.59 0.14 0.16 0.80
a
Mean of the maximal valvular thickness. Tricuspid valve (3), mitral valve (1), pulmonary valve (1), aortic valve (1); (Ao) aortic valve; (BE) bacterial endocarditis; (MV) mitral valve; (NONBE) non-bacterial endocarditis; (PV) pulmonary valve; (TV) tricuspid valve. b
cardiac forms of enzootic lymphoma (n ¼ 2), traumatic pericarditis (n ¼ 2), bronchopneumonia (n ¼ 2), idiopathic hemorrhagic pericarditis (n ¼ 1), ventricular septal defect (n ¼ 1), pleuritis (n ¼ 1), and cor pulmonale (n ¼ 1). The
Figure 3 Box and whiskers plot of the maximal valvular thickness in 40 healthy Holstein cows determined by right transthoracic echocardiography. There were significant differences in the maximal valve thicknesses (P < 0.0001). The number indicated between the different plots is the P value from the post-hoc Tukey tests to assess the differences between pairs of valves. * Significant P value.
Figure 4 Box and whiskers plot of the maximal thickness of endocarditic valves and non endocarditic valves using transthoracic echocardiography in cows with confirmed diagnosis of bacterial endocarditis. The 6 endocarditic valves (3 tricuspid, 1 aortic, 1 pulmonary and 1 mitral) were compared with the 18 unaffected valves in the 6 cows with a confirmed diagnosis of bacterial endocarditis. * Significant P value.
echocardiographic data of the BE and NONBE groups are presented in Table 2. The mean maximal tricuspid, mitral, pulmonary, and aortic thicknesses of unaffected valves of the BE group did not differ from the CONTROL or NONBE groups.
Discussion The most typical echocardiographic sign of BE is valvular thickening.3,11,12 In light of the low prevalence of noninfectious valvular disorders in cattle,5 echocardiographic evidence of thickening of the endocardium should be highly suggestive of BE and should therefore be considered a possible major criterion for the clinical diagnosis of BE. A major problem with this criterion is that the definition of thickening is mostly based on operator experience. In cattle, it is of particular importance to have a more objective definition of thickening since echocardiography is generally performed by people who are not board-certified cardiologists. These clinicians need to know what valvular thickness qualifies as normal and abnormal. This will have an important impact on case management. The cardiac valves are composed of a thin (typically less than 1 mm) fibroconnective lamina that is enveloped in endocardial cells.18 They are attached to the myocardium with the tendinous chordae apparatus for the mitral and tricuspid valves, or the pulmonary or aortic trunk for the
Bovine valvular thickness measurement semilunar valves.19 The high valvular thicknesses that were obtained in this study for healthy cows without heart disease may be due to several factors. First, the absence of concomitant ECG recording during the echocardiographic examination and the method of measuring the closed valves using the cine-loop mode to obtain the best view may have introduced some variability to our measurements. Second, the low frequency phased array probe that was used in this study may have had an impact on the values that were reported. Due to the depth of the measured structures and the potential interference of chordae tendinae and accessory leaflets when the measurement was performed, we may have overestimated valvular thickness. This overestimation is more likely to have occurred for the mitral valve measurement, since the mitral valve is the deepest valve when performing right sided TTE. The depth required to perform TTE in adult Holstein cows (25e30 cm) and the presence of distorsion may have had an impact on the high mitral valve thickness found in this study. The lower resolution of the low frequency phased array probe, as well as side-lobe artifacts which are commonly encountered with phased array probes,18 may have also limited our ability to accurately measure the true mitral valve thickness. However, this right-sided technique is the most practical in cattle due to the relatively good acoustic window with the lack of lung tissue at this location.12,15e17 The left-sided TTE approach would certainly have improved the measurement of the mitral valve, likely leading to lower thickness values than from right-sided TTE as this valve is more superficial on the left, limiting distortion.12,15,16 Post mortem identification of infected valves represents the gold standard for defining the exact localization and extension of BE. In our study, we used this criterion for inclusion so that the pathologic findings could be compared with echocardiographic findings. Cases with presumed BE based on clinical signs and echocardiographic examination were not included in the study, although they represented 50% of the BE cases diagnosed during this period at our institution. These cases may represent more accurately the typical presumptive BE diagnosis in a field context because once the diagnosis is made, and depending on the gravity of the clinical signs, the producer might choose to slaughter the animal before any attempt at therapy is made, despite the higher risk of carcass condemnation without a postmortem examination.1 This study is therefore of importance to establish the normal variation of valvular thickness in healthy cows and cows
259 without valvular problems, in order to be able to rule-in or rule-out BE in cows using a more objective assessment of valvular thickness. Although TTE is an important diagnostic tool for detecting BE in cattle, the use of echocardiography for the diagnosis of valvular BE in humans has limitations.20 The sensitivity of TTE in the diagnosis of BE is 60e70%,10 but the negative predictive value of normal TTE in the evaluation of possible valve endocarditis is more than 90% in humans.21 Transesophageal echocardiography (TEE) is a more sensitive tool in humans since it allows for a more precise evaluation of slight changes in valvular morphology. It is especially useful for detecting small endocardial vegetations.22 However, TEE has never been used for the assessment of bovine endocarditis and has only been used for describing mediastinal structures in cattle.23 The apparent higher sensitivity of TTE in cattle for the diagnosis of BE (84.3%, 95% confidence interval: 60.4e95.0%)14 when compared to TTE in humans may be due to the more advanced stage of the disease when cows are examined. Most of the reports have been based on referred clinical cases and sensitivity has been assessed using necropsy as a gold standard; this may overestimate diagnostic accuracy because it represents a subpopulation of BE cases with advanced stage of the disease. It therefore is difficult to extrapolate this sensitivity for practical purposes since it may also be influenced by performance of the ultrasound unit as well as by the experience of the operators performing the study, especially in the absence of objective criteria for the echocardiographic characteristics of BE. For these reasons, the present study focused on establishing normal values for the thickness of bovine cardiac valves. Due to the high prevalence of BE in cows with valvular thickening, the positive predictive value of valvular thickening should be high for BE diagnosis. Our evaluation of a group of cows with various cardiopulmonary diseases was also interesting. From a clinical standpoint, these cows had diseases for which BE is a potential differential diagnosis. In other species, as well as in cattle, valvular thickening can be observed in patients with myxomatous processes, degenerative changes, tumors, blood cysts, thrombi, or imaging artifacts.10,24e27 For these reasons, it is important to keep in mind that echocardiographic findings alone are not sufficient to establish a diagnosis; they need to be assessed in conjunction with clinical findings and other ancillary tests (e.g., hematology, biochemistry and blood culture). In humans and small animals, multiple clinical and clinicopathological findings are needed to make a
260 clinical diagnosis of endocarditis. The Duke criteria consist of various major and minor criteria. The presence of two major criteria or one major criterion with three minor criteria are required to raise a clinical suspicion of BE. These criteria have been rigorously tested in humans but not in small animals. This type of approach has never been applied to the study of bovine BE since most of the available studies consist of case reports or case series, without including any cows in which BE was a differential diagnosis which was subsequently excluded by further diagnostic tests. In light of the low prevalence of noninfectious valvular disease in cows,5 the results of our study suggest that valvular thickening should be considered as a major criterion for a diagnosis of BE in cows. Data concerning bovine cardiac diseases are still scant due to the low prevalence of these diseases and the advanced stage of the diseases when diagnoses are made. For these reasons, most cows diagnosed with heart disease are culled or euthanized without being sent to a referral center or without echocardiographic confirmation of the diagnosis. With the increasing availability of good quality portable ultrasound units and their increasing use outside of reproduction, it is important to collect specific data concerning bovine echocardiography.
Conclusions From a practical standpoint, a valve thickness in cows greater than 0.85 cm (tricuspid), 1.27 cm (mitral), 1.06 cm (aortic) or 0.82 cm (pulmonic) should raise the possibility of BE when right TTE is performed with a phased array probe. Future studies in cows with confirmed BE and cows in which BE was an initial differential diagnosis but was then excluded, need to be performed in order to improve the diagnostic accuracy of TTE, which represents a clinical challenge for the internist as well as for the bovine practitioner, especially in the early stages of disease.
Conflict of interest The authors have no conflicts of interest.
Acknowledgments The authors would like to acknowledge Mrs. Lucie Lapre ´vote and Mrs. Katia Juste for their technical
S. Buczinski et al. help during the echocardiographic study in healthy cows. The assistance of Mr. Guy Beauchamp in the statistical analysis of the data is greatly appreciated.
References 1. Bexiga R, Mateus A, Philbey AW, Ellis K, Barrett DC, Mellor DJ. Clinicopathological presentation of cardiac diseases in cattle and its impact on decision making. Vet Rec 2008;162:575e580. 2. Buczinski S, Francoz D, Fecteau G, DiFruscia R. Congestive heart failure in cattle: 59 cases (1990e2005). Can Vet J 2010;51:1123e1129. 3. Healy AM. Endocarditis in cattle: a review of 22 cases. Irish Vet J 1996;49:43e48. 4. Buczinski S, Rezakhani A, Boerboom D. Heart disease in cattle: a review of the current data regarding diagnosis, therapeutic approaches and prognosis. Vet J 2010;184: 258e263. 5. Reef VB, McGuirk SM. Valvular heart disease. In: Smith BP, editor. Large animal internal medicine. 4th ed. Philadelphia: Elsevier Saunders; 2009. p. 463e468. 6. Power HT, Rebhun WC. Bacterial endocarditis in adult dairy cattle. J Am Vet Med Assoc 1983;182:806e808. 7. Rezakhani A, Zarifi M. Auscultatory findings of cardiac murmurs in clinically healthy cattle. Online J Vet Res 2007; 11:62e66. 8. Buczinski S, Francoz D, Fecteau G, DiFruscia R. A study of heart diseases without clinical signs of heart failure in 47 cattle. Can Vet J 2010;51:1239e1246. 9. Kittleson MD, Kienle RD. Infective endocarditis (and annuloaortic ectasia). In: Kittleson MD, Kienle RD, editors. Small animal cardiovascular medicine. St. Louis: Mosby; 1998. p. 402e412. 10. Feigenbaum H, Armstrong WF, Ryan T. Infective endocarditis. In: Feigenbaum H, Armstrong WF, Ayan T, editors. Feigenbaum’s echocardiography. 6th ed. Philadelphia: Lippincott William and Wilkins; 2005. p. 375e398. 11. Yamaga Y, Too K. Diagnostic ultrasound imaging of vegetative valvular endocarditis in cattle. Jpn J Vet Res 1987;35: 49e63. 12. Buczinski S. Cardiovascular ultrasonography in cattle. Vet Clin North Am Food Anim Pract 2009;25:611e632. 13. Durack DT, Lukes AS, Bright DK. Duke endocarditis service: new criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994;96:200e209. 14. Buczinski S, Tsuka T, Tharwat M. The diagnostic criteria used in bovine bacterial endocarditis: a meta-analysis of 460 published cases from 1973 to 2011. Vet J 2012;193: 349e357. 15. Braun U, Schweizer T, Pusterla N. Echocardiography of the normal bovine heart: technique and ultrasonographic appearance. Vet Rec 2001;148:47e51. 16. Hallowell GD, Potter TJ, Bowen IM. Methods and normal values for echocardiography in adult dairy cattle. J Vet Cardiol 2007;9:91e98. 17. Varga A, Schober KE, Holloman CH, Stromberg PC, Lakritz J, Rings DM. Correlation of serum cardiac troponin I and myocardial damage in cattle with monensin toxicosis. J Vet Intern Med 2009;23:1108e1116. 18. Barone R. Coeur. In: Barone R, editor. Anatomie compare ´e des mammife `res domestiques, Tome 5: Angiologie. Paris: Vigot; 1996. p. 5e102.
Bovine valvular thickness measurement
261
19. Feigenbaum H, Armstrong WF, Ryan T. Artifacts. In: Feigenbaum H, Armstrong WF, Ryan T, editors. Feigenbaum’s echocardiography. 6th ed. Philadelphia: Lippincott William and Wilkins; 2005. p. 29e32. 20. Erbel R, Rohmann S, Drexler M, Mohr-Kahaly S, Gerharz CD, Iversen S, Oelert H, Meyer J. Improved diagnostic value of echocardiography in patients with infective endocarditis by transesophageal approach. Eur Heart J 1988;9:43e53. 21. McDermott BP, Cunha BA, Choi D, Cohen J, Hage J. Transthoracic echocardiography (TTE): sufficiently sensitive screening test for native valve infective endocarditis (IE). Heart Lung 2011;40:358e360. 22. Kuhl HP, Hanrath P. The impact of transesophageal echocardiography in daily practice. Eur J Echocardiogr 2004;5: 455e458.
23. Braun U, Sicher D, Pusterla N. Ultrasonography of the lungs, pleura, and mediastinum in healthy cows. Am J Vet Res 1996;57:432e438. 24. Borgarelli M, Ha ¨ggstrom J. Canine degenerative myxomatous mitral valve disease: natural history, clinical presentation and therapy. Vet Clin North Am Small Anim Pract 2010;40:651e663. 25. Buczinski S, Badillo M, Be ´langer AM, Rezakhani A. Echocardiographic findings of tricuspid blood cyst mimicking bacterial endocarditis finding in a dairy cow. J Vet Cardiol 2011;13:283e286. 26. Brisville AC, Buczinski S, Chenier S, Francoz D. A cardiac vascular hamartoma in a calf: clinical, echocardiographic and pathologic findings. J Vet Cardiol 2012;14:377e380.
Available online at www.sciencedirect.com
ScienceDirect