REVIEW For reprint orders, please contact: [email protected]
Cardiothoracic ratio for prediction of left ventricular dilation: a systematic review and pooled analysis Rohit S Loomba*,1, Parinda H Shah2, Karan Nijhawan3, Saurabh Aggarwal4 & Rohit Arora5 ABSTRACT Increased cardiothoracic ratio noted on chest radiographs often prompts concern and further evaluation with additional imaging. This study pools available data assessing the utility of cardiothoracic ratio in predicting left ventricular dilation. A systematic review of the literature was conducted to identify studies comparing cardiothoracic ratio by chest x-ray to left ventricular dilation by echocardiography. Electronic databases were used to identify studies which were then assessed for quality and bias, with those with adequate quality and minimal bias ultimately being included in the pooled analysis. The pooled data were used to determine the sensitivity, specificity, positive predictive value and negative predictive value of cardiomegaly in predicting left ventricular dilation. A total of six studies consisting of 466 patients were included in this analysis. Cardiothoracic ratio had 83.3% sensitivity, 45.4% specificity, 43.5% positive predictive value and 82.7% negative predictive value. When a secondary analysis was conducted with a pediatric study excluded, a total of five studies consisting of 371 patients were included. Cardiothoracic ratio had 86.2% sensitivity, 25.2% specificity, 42.5% positive predictive value and 74.0% negative predictive value. Cardiothoracic ratio as determined by chest radiograph is sensitive but not specific for identifying left ventricular dilation. Cardiothoracic ratio also has a strong negative predictive value for identifying left ventricular dilation. Chest radiographs are a commonly used diagnostic tool, providing objective and subjective data regarding the bony structures, vascular structures and viscera in the thorax. Assessment of the cardiovascular system can be of particular value whether it is in regards to evaluation of pulmonary under circulation or over circulation noted with various types of congenital heart disease, pulmonary vascular congestion and edema noted with heart failure, or heart shape and size that may be altered for various reasons. The standard posteroanterior (PA), anteroposterior (AP) and lateral projections of chest radiographs offer a vast amount of information which can be extracted if the image is obtained effectively. One of the most commonly assessed features of the cardiovascular system is the cardiac silhouette and the surrounding mediastinal contours. The cardiac silhouette is often evaluated to determine if there is cardiac chamber enlargement with cardiothoracic ratios of 50 and 55% being considered normal for PA and AP chest radiographs, respectively [1] . This study aims to provide a polled analysis of published data regarding the utility of frontal chest radiographs in predicting left ventricular dilation.
KEYWORDS
• cardiomegaly • cardiothoracic ratio • chest radiograph • dilation • echocardiography • left ventricular • x-ray
Department of Pediatric Cardiology, Children’s Hospital of Wisconsin/Medical College of Wisconsin, Milwaukee, WI, USA Department of Radiology, Advocate Illinois Masonic Medical Center, Chicago, IL, USA 3 Department of Medicine, Rush Medical Center, Chicago, IL, USA 4 Department of Cardiology, Creighton University Medical Center, Omaha, NE, USA 5 Rosalind Franklin University Medical Center, North Chicago, IL, USA *Author for correspondence: [email protected] 1 2
10.2217/FCA.15.5 © 2015 Future Medicine Ltd
Future Cardiol. (2015) 11(2), 171–175
part of
ISSN 1479-6678
171
Review Loomba, Shah, Nijhawan, Aggarwal & Arora Methods A systematic review of the literature was conducted utilizing the PRISMA guidelines. No established review protocol for this study was established and published for this review before it was conducted. The study was intended to analyze the statistical validity of utilizing cardiothoracic ratio as determined by chest radiography, in the PA or AP projection, to determine cardiac enlargement utilizing left ventricular end diastolic dimension via echocardiography as a reference value. A thorough review of published literature was conducted utilizing searches of the following electronic databases: embase, pubmed and medline. Search terms used included ‘cardiothoracic ratio’, ‘cardiomegaly’, ‘cardiac silhouette’, ‘heart size’, ‘x-ray’, ‘radiography’, ‘plain film’, ‘echocardiography’, ‘echocardiogram’, ‘left ventricular end diastolic dimension’, ‘dilation’ and ‘cardiac enlargement’, along with various combinations of these search terms. The abstracts of resulting manuscripts were screened for relevance and the references of these manuscripts screened for additional references. This was done by both authors individually and manuscripts were identified for review of full text. Full text manuscripts were then assessed for study type with review articles being excluded. Retrospective and prospective studies were then further assessed for availability of relevant data. At this point, studies that did not report cardiothoracic ratio or left ventricular end diastolic dimension were excluded. Next, manuscript quality was assessed and scored by both authors individually. Scores were then compared and general consensus was reached for studies with scores that differed between the two authors. Data were then extracted by both authors individually from studies identified for inclusion. Extracted data were compared and any differences in extracted data prompted review by both authors together. Data points collected included number of patients with normal and abnormal cardiothoracic ratio by chest radiograph and number of patients with normal and abnormal left ventricular end diastolic dimension. Once data were extracted and verified, analysis was conducted which consisted of calculating sensitivity, specificity, positive predictive value and negative predictive value of cardiothoracic ratio by chest radiograph utilizing left ventricular end diastolic dimension by echocardiography as the reference value. These were also calculated with exclusion of Satou et al. which included pediatric patients.
172
Future Cardiol. (2015) 11(2)
Included studies defined cardiomegaly by chest radiographs as a cardiothoracic ratio of greater than 50 or 55% in the PA and AP lateral projections, respectively. Left ventricular dilation was defined as a left ventricular end diastolic dimension of greater than 5.5 cm in all but one study which used a cutoff of 5.0 cm. Risk of bias was identified across studies utilizing the Cochrane Collaboration’s Risk of Bias tool. Studies were assessed for adequate sequence generation, allocation concealment, blinding, how incomplete outcome data were addressed, and if they were free of selective reporting. Collected data were then pooled and used to calculate sensitive, specificity, negative predictive value and positive predictive value. Results Electronic searches and hand search of referenced manuscripts identified 15 manuscripts for full text review. After exclusions were made, as outlined above, seven manuscripts identified for inclusion in the final analysis. Of these, seven studies were included for the final analysis [2–7] . Six of these studies were prospective in nature while one study was retrospective. All of these studies except two included consecutive patients while two of these studies consisted of a particular patient population. Arrieta et al. consisted of patients with systemic hypertension while Kono et al. consisted of patients with dilated cardiomyopathy [2,5] . Median patient age in the pooled population was 52.6 years with a range from 2 days of age to 68 years of age. All the included studies utilized chest radiographs that were taken in the posterioanterior projection. Cardiothoracic ratio was assessed in all these studies and was deemed to be abnormal if greater than 50 % of the diameter of the chest cavity. All the included studies also reported left ventricular end diastolic dimension. Risk of bias was assessed for all studies and is summarized in Table 1. Overall, the level of bias assessed in all the studies was low. A total of 466 patients were included in the final pooled analysis. Data from the included studies were used to calculate sensitivity, specificity, positive predictive value and negative predictive value of cardiothoracic chest radiography in prediction of abnormal left ventricular end diastolic dimension. A total of 135 patients had cardiomegaly by chest radiography and abnormal left ventricular end diastolic dimension by echocardiography, 175 patients had cardiomegaly by
future science group
Cardiothoracic ratio for prediction of left ventricular dilation
Review
Table 1. Characteristics of included studies. Study (year)
Design
n
Age (years)
CXR projection
Satou et al. (2001) Kadhum et al. (2007) Kono et al. (1992) Lupow et al. (2002) Diaz Arrieta et al. (2006) Pereira-Barretto et al. (1983)
Prospective Prospective Prospective Retrospective Prospective Prospective
95 150 12 115 72 22
7.5 ± 5 – 47.7 ± 13.2 – 62.4 ± 10 36.0 ± 6.4
Posterioanterior Posterioanterior Posterioanterior Posterioanterior Posterioanterior Posterioanterior
Ref. [3] [4] [5] [6] [2] [7]
Age is expressed as mean ± standard deviation. CXR: Chest x-ray.
chest radiography but normal left ventricular end diastolic dimension by echocardiography, 27 patients had no cardiomegaly by chest radiography but did have an abnormal left ventricular end diastolic dimension by echocardiography and 129 patients had no cardiomegaly by chest radiography and had normal left ventricular end diastolic dimension by echocardiography. This resulted in a sensitivity of 83.3%, specificity of 42.4%, positive predictive value of 43.5% and negative predictive value of 82.7%. If the pediatric study by Satou et al. is excluded from the analysis then a total of 371 patients were included in the final polled analysis. A total of 125 patients had cardiomegaly by chest radiography and abnormal left ventricular end diastolic dimension by echocardiography, 169 patients had cardiomegaly by chest radiography but normal left ventricular end diastolic dimension by echocardiography, 20 patients had no cardiomegaly by chest radiography but did have an abnormal left ventricular end diastolic dimension by echocardiography and 57 patients had no cardiomegaly by chest radiography and had normal left ventricular end diastolic dimension by echocardiography. This resulted in a sensitivity of 86.2%, specificity of 25.2%, positive predictive value of 42.5% and negative predictive value of 74.0%. Discussion This pooled analysis demonstrates the lack of utility of cardiothoracic ratio in the assessment of left ventricular dilation. The sensitivity of 83.3% and specificity of 42.4% make the cardiothoracic ratio neither valuable as a screening nor a confirmatory test. With a negative predictive value of 82.7%, one can be reassured that a normal cardiothoracic ratio is associated with no left ventricular dilation but with a positive predictive value of 43.5% one cannot be certain that there is left ventricular dilation if the cardiac silhouette appears enlarged.
future science group
When the pediatric study by Satou et al. was removed from the analysis there was minimal difference in all values except for the specificity which was low, thus not changing the utility of the chest radiographs. Sensitivity was 86.2%, specificity was 25.2%, positive predictive value of 42.5 and negative predictive value of 74.0. It is not particularly surprising that removing the pediatric study does not significantly change these values as the same criteria for cardiothoracic ratio is used in all age groups and cardiothoracic ratio does not change with age. As such it seems arbitrary to expect a change between those that are 16 years of age and 26 years of age compared with those who are 30 years of age compared with 60 years of age. This additional analysis demonstrates this. The change in specificity is of minimal consequence as the specificity with or without these pediatric patients is not high enough to deem chest radiographs as a helpful test in positively identifying those with left ventricular dilation but simply to rule out left ventricular dilation. Change in cardiothoracic ratio with age has been studied but primarily in the elderly population. In a study of 110 elderly women, Inoue et al. demonstrated that cardiothoracic ratio increased 2.0% over 9 years of follow-up. While this finding was statistically significant this is unlikely to be of clinical consequence [8] . Potter et al. demonstrated similar findings in 243 men over 12 years of follow-up. A majority of patients still remained under the cutoff for abnormal [9] . Both studies found that part of this increase is due to not only increase in cardiac size but also a decrease in the thoracic diameter that is associated with aging. Thus while there is a change in cardiothoracic ratio with age, the degree of this has minimal clinical significance. Cardiothoracic ratio can be impacted by and implicated in several physiologic derangements, one of which is heart failure. In a study of patients with heart failure, Fukuta et al. demonstrated that increased cardiothoracic ratio on
www.futuremedicine.com
173
Review Loomba, Shah, Nijhawan, Aggarwal & Arora chest radiographs correlated with right-sided chamber size moreso than left sided chamber size [10] . Some studies have further investigated cardiothoracic ratio in the setting of congestive heart failure, with particular interest in determining whether or not cardiothoracic ratio can be utilized to help determine ejection fraction. Philbin et al. found that a negative correlation does exist between cardiothoracic ratio and ejection fraction but this was a weak correlation and the study concluded that cardiothoracic ratio cannot be utilized in estimating ejection fraction [11] . These findings were further confirmed by Clark et al. who determined that chest radiographs are not a reliable indicator of left ventricular dysfunction [12] . Another study found that cardiothoracic ratio can be a prognostic indicator in those with poor ejection fraction but only in a multivariate setting thus further highlighting the findings by the aforementioned studies that cardiothoracic ratio, in isolation, is of limited value in the evaluation of heart failure [13] . While the cardiothoracic ratio does not appear to be useful in predicting left ventricular dilation this does not mean that chest radiographs cannot be used to assess cardiac structure. More sophisticated techniques have been described to evaluate for left ventricular dilation using both the frontal and lateral projections which are of more utility. These, however, do require more time and are less intuitive. Elliott et al. proposed a systematic approach to interpreting chest radiographs with respect to the cardiovascular system. First, the technical integrity of the image must be evaluated. It is necessary to establish whether the radiograph was obtained in the PA or AP projection as the cardiac silhouette appears larger in the AP projection when compared with the PA projection. Additionally, portable chest radiographs obtained with patients in the supine position may also make the cardiothoracic ratio appear larger. This also occurs with films obtained during expiration. Angulation of the radiograph beam can also impact the cardiothoracic ratio observed. Next, in the system proposed by Elliott et al., assessment of the extracardiac structures is done, as these can be associated with cardiac abnormalities. For instance, situs inversus of the abdominal viscera is often associated with heterotaxy and interruption of the inferior vena cava with azygos continuation. Pulmonary vascularity can be used in neonates to assist with the diagnosis of various congenital
174
Future Cardiol. (2015) 11(2)
heart defects as these may present with varying degree of pulmonary blood flow [14,15] . The final step is then to assess the cardiac structures themselves. Discussion of this in its entirety is beyond the scope of the manuscript but assessment of left ventricular dilation can be done utilizing features other than the cardiothoracic ratio. Displacement of the cardiac apex leftward and toward the diaphragm is often seen with left ventricular dilation. A portion of the apex may even appear to be below the diaphragm even and posterior to the gastric bubble. Dilation of the left ventricle also tends to augment the concavity present in the middle of the left heart border [14,15] . The question then arises as to what should be done if the cardiothoracic ratio is found to be abnormal. Without other concerns in the history or physical examination, it is reasonable to obtain an electrocardiogram as it is relatively inexpensive. Echocardiography, however, should not be avoided if the index of suspicion is high using the overall clinical situation. Conclusion Enlarged cardiothoracic ratio may be noted on chest radiograph but is not of particular value in predicting left ventricular dilation. Suspicion of left ventricular dilation secondary to chest radiograph findings should prompt evaluation by electrocardiogram before imaging by echocardiography or other modalities. Future perspective Currently echocardiography is still a relatively expensive and time-consuming endeavor. Thus it is impractical to obtain an echocardiogram on all patients who demonstrate evidence of cardiomegaly on a chest x-ray. It is likely that developments in portable, handheld ultrasound technology will one day allow for rapid and inexpensive echocardiography to determine the left ventricular dimensions if there are concerns for echocardiography. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.
future science group
Cardiothoracic ratio for prediction of left ventricular dilation
Review
EXECUTIVE SUMMARY Frequency of chest radiographs ●●
Chest radiographs are among the most common radiologic examination ordered today.
Cardiomegaly noted by chest radiograph ●●
Cardiomegaly is often appreciated on chest radiographs, leading to concern for left ventricular dilation.
Evaluation of left ventricular dilation ●●
Echocardiography is the definitive diagnostic tool which determines whether there is left ventricular dilation but requires neither time nor cost–effectiveness in all cases.
Additional diagnostic testing for assessment of left ventricular dilation ●●
Electrocardiograms should be obtained as an additional screening when there is concern for cardiomegaly based on chest radiographs. Electrocardiogram abnormalities make subsequent echocardiography reasonable to obtain.
References Papers of special note have been highlighted as: •• of considerable interest 1
2
size in comparison to echocardiography. Med. J. Basrah Univ. 25(2), 48–51 (2007). 5
Yoo S-J. Chest Radiographic Interpretation in Pediatric Cardiac Patients. Yoo S-J, Macdonald C, Babyn PS (Eds). Thieme, NY, USA; Toronto, Canada (2010). Díaz Arrieta G, Mendoza Hernández ME, Hernández Cabrera J et al. Correlacion entre la radiograofia de torax y el ecocardiograma para la valoracion de cardiomegalia en pacientes con hipertension arterial sistemica. Arch. Cardiol. Mex. 76(2), 179–184 (2006).
6
7
•• Large study that investigates the relationship between cardiothoracic ratio and left ventricular dilation as noted by echocardiogram. 3
Satou GM, Lacro RV, Chung T, Gauvreau K, Jenkins KJ. Heart size on chest x-ray as a predictor of cardiac enlargement by echocardiography in children. Pediatr. Cardiol. 22(3), 218–222 (2001).
•• Only pediatric study that investigates the relationship between cardiothoracic ratio and left ventricular dilation as noted by echocardiogram. 4
Kadhum AA, Sharif S, Abd-Hazaa M. Validity of chest x-ray in estimation of cardiac
future science group
8
9
Kono T, Suwa M, Hanada H, Hirota Y, Kawamura K. Clinical significance of normal cardiac silhouette in dilated cardiomyopathy: evaluation based upon echocardiography and magnetic resonance imaging. Jpn. Circ. J. 56(4), 359–365 (1992). Lupow JB, Sivak SL, Boss D. The accuracy of the cardiothoracic ratio as a predictor of cardiac enlargement and dysfunction. Acad. Emerg. Med. 9(5), 462 (2002). Pereira-Barretto AC, Mady C, ArteagaFernandez E et al. Valor do indice cardiotoracico na avaliacao do comprometimento miocardico, correlacao com avaliacao ecocardiografica. Rev. Hosp. Clin. Fac. Med. Sao Paulo 38(1), 43–45 (1983). Inoue K, Yoshii K, Ito H. Effect of aging on cardiothoracic ratio in women: a longitudinal study. Gerontology 45(1), 53–58 (1999). Potter JF, Elahi D, Tobin JD, Andres R. Effect of aging on the cardiothoracic ratio of men. J. Am. Geriatr. Soc. 30(6), 404–409 (1982).
10 Fukuta H, Ohte N, Brucks S, Carr JJ, Little
WC. Contribution of right-sided heart enlargement to cardiomegaly on chest
roentgenogram in diastolic and systolic heart failure. Am. J. Cardiol. 99(1), 62–67 (2007). •• Describes the contribution of right-sided chamber size on the cardiothoracic ratio in the setting of heart failure. 11 Philbin EF, Garg R, Danisa K et al. The
relationship between cardiothoracic ratio and left ventricular ejection fraction in congestive heart failure. Digitalis Investigation Group. Arch. Intern. Med. 158(5), 501–506 (1998). 12 Clark AL, Coats AJ. Unreliability of
cardiothoracic ratio as a marker of left ventricular impairment: comparison with radionuclide ventriculography and echocardiography. Postgrad. Med. J. 76(895), 289–291 (2000). 13 Hammermeister KE, Chikos PM, Fisher L,
Dodge HT. Relationship of cardiothoracic ratio and plain film heart volume to late survival. Circulation 59(1), 89–95 (1979). •• Describes the prognostic value of cardiothroacic ratio in the setting of heart failure. 14 Roberts WC. Adult Congenital Heart
Disease. Davis, PA, USA (1987). 15 Elliott JA, Cowan MD. Exercises in Chest
X-Ray Diagnosis. Butterworths, MA, USA (1987).
www.futuremedicine.com
175
Cardiothoracic ratio for prediction of left ventricular dilation: a systematic review and pooled analysis Rohit S Loomba*,1, Parinda H Shah2, Karan Nijhawan3, Saurabh Aggarwal4 & Rohit Arora5 ABSTRACT Increased cardiothoracic ratio noted on chest radiographs often prompts concern and further evaluation with additional imaging. This study pools available data assessing the utility of cardiothoracic ratio in predicting left ventricular dilation. A systematic review of the literature was conducted to identify studies comparing cardiothoracic ratio by chest x-ray to left ventricular dilation by echocardiography. Electronic databases were used to identify studies which were then assessed for quality and bias, with those with adequate quality and minimal bias ultimately being included in the pooled analysis. The pooled data were used to determine the sensitivity, specificity, positive predictive value and negative predictive value of cardiomegaly in predicting left ventricular dilation. A total of six studies consisting of 466 patients were included in this analysis. Cardiothoracic ratio had 83.3% sensitivity, 45.4% specificity, 43.5% positive predictive value and 82.7% negative predictive value. When a secondary analysis was conducted with a pediatric study excluded, a total of five studies consisting of 371 patients were included. Cardiothoracic ratio had 86.2% sensitivity, 25.2% specificity, 42.5% positive predictive value and 74.0% negative predictive value. Cardiothoracic ratio as determined by chest radiograph is sensitive but not specific for identifying left ventricular dilation. Cardiothoracic ratio also has a strong negative predictive value for identifying left ventricular dilation. Chest radiographs are a commonly used diagnostic tool, providing objective and subjective data regarding the bony structures, vascular structures and viscera in the thorax. Assessment of the cardiovascular system can be of particular value whether it is in regards to evaluation of pulmonary under circulation or over circulation noted with various types of congenital heart disease, pulmonary vascular congestion and edema noted with heart failure, or heart shape and size that may be altered for various reasons. The standard posteroanterior (PA), anteroposterior (AP) and lateral projections of chest radiographs offer a vast amount of information which can be extracted if the image is obtained effectively. One of the most commonly assessed features of the cardiovascular system is the cardiac silhouette and the surrounding mediastinal contours. The cardiac silhouette is often evaluated to determine if there is cardiac chamber enlargement with cardiothoracic ratios of 50 and 55% being considered normal for PA and AP chest radiographs, respectively [1] . This study aims to provide a polled analysis of published data regarding the utility of frontal chest radiographs in predicting left ventricular dilation.
KEYWORDS
• cardiomegaly • cardiothoracic ratio • chest radiograph • dilation • echocardiography • left ventricular • x-ray
Department of Pediatric Cardiology, Children’s Hospital of Wisconsin/Medical College of Wisconsin, Milwaukee, WI, USA Department of Radiology, Advocate Illinois Masonic Medical Center, Chicago, IL, USA 3 Department of Medicine, Rush Medical Center, Chicago, IL, USA 4 Department of Cardiology, Creighton University Medical Center, Omaha, NE, USA 5 Rosalind Franklin University Medical Center, North Chicago, IL, USA *Author for correspondence: [email protected] 1 2
10.2217/FCA.15.5 © 2015 Future Medicine Ltd
Future Cardiol. (2015) 11(2), 171–175
part of
ISSN 1479-6678
171
Review Loomba, Shah, Nijhawan, Aggarwal & Arora Methods A systematic review of the literature was conducted utilizing the PRISMA guidelines. No established review protocol for this study was established and published for this review before it was conducted. The study was intended to analyze the statistical validity of utilizing cardiothoracic ratio as determined by chest radiography, in the PA or AP projection, to determine cardiac enlargement utilizing left ventricular end diastolic dimension via echocardiography as a reference value. A thorough review of published literature was conducted utilizing searches of the following electronic databases: embase, pubmed and medline. Search terms used included ‘cardiothoracic ratio’, ‘cardiomegaly’, ‘cardiac silhouette’, ‘heart size’, ‘x-ray’, ‘radiography’, ‘plain film’, ‘echocardiography’, ‘echocardiogram’, ‘left ventricular end diastolic dimension’, ‘dilation’ and ‘cardiac enlargement’, along with various combinations of these search terms. The abstracts of resulting manuscripts were screened for relevance and the references of these manuscripts screened for additional references. This was done by both authors individually and manuscripts were identified for review of full text. Full text manuscripts were then assessed for study type with review articles being excluded. Retrospective and prospective studies were then further assessed for availability of relevant data. At this point, studies that did not report cardiothoracic ratio or left ventricular end diastolic dimension were excluded. Next, manuscript quality was assessed and scored by both authors individually. Scores were then compared and general consensus was reached for studies with scores that differed between the two authors. Data were then extracted by both authors individually from studies identified for inclusion. Extracted data were compared and any differences in extracted data prompted review by both authors together. Data points collected included number of patients with normal and abnormal cardiothoracic ratio by chest radiograph and number of patients with normal and abnormal left ventricular end diastolic dimension. Once data were extracted and verified, analysis was conducted which consisted of calculating sensitivity, specificity, positive predictive value and negative predictive value of cardiothoracic ratio by chest radiograph utilizing left ventricular end diastolic dimension by echocardiography as the reference value. These were also calculated with exclusion of Satou et al. which included pediatric patients.
172
Future Cardiol. (2015) 11(2)
Included studies defined cardiomegaly by chest radiographs as a cardiothoracic ratio of greater than 50 or 55% in the PA and AP lateral projections, respectively. Left ventricular dilation was defined as a left ventricular end diastolic dimension of greater than 5.5 cm in all but one study which used a cutoff of 5.0 cm. Risk of bias was identified across studies utilizing the Cochrane Collaboration’s Risk of Bias tool. Studies were assessed for adequate sequence generation, allocation concealment, blinding, how incomplete outcome data were addressed, and if they were free of selective reporting. Collected data were then pooled and used to calculate sensitive, specificity, negative predictive value and positive predictive value. Results Electronic searches and hand search of referenced manuscripts identified 15 manuscripts for full text review. After exclusions were made, as outlined above, seven manuscripts identified for inclusion in the final analysis. Of these, seven studies were included for the final analysis [2–7] . Six of these studies were prospective in nature while one study was retrospective. All of these studies except two included consecutive patients while two of these studies consisted of a particular patient population. Arrieta et al. consisted of patients with systemic hypertension while Kono et al. consisted of patients with dilated cardiomyopathy [2,5] . Median patient age in the pooled population was 52.6 years with a range from 2 days of age to 68 years of age. All the included studies utilized chest radiographs that were taken in the posterioanterior projection. Cardiothoracic ratio was assessed in all these studies and was deemed to be abnormal if greater than 50 % of the diameter of the chest cavity. All the included studies also reported left ventricular end diastolic dimension. Risk of bias was assessed for all studies and is summarized in Table 1. Overall, the level of bias assessed in all the studies was low. A total of 466 patients were included in the final pooled analysis. Data from the included studies were used to calculate sensitivity, specificity, positive predictive value and negative predictive value of cardiothoracic chest radiography in prediction of abnormal left ventricular end diastolic dimension. A total of 135 patients had cardiomegaly by chest radiography and abnormal left ventricular end diastolic dimension by echocardiography, 175 patients had cardiomegaly by
future science group
Cardiothoracic ratio for prediction of left ventricular dilation
Review
Table 1. Characteristics of included studies. Study (year)
Design
n
Age (years)
CXR projection
Satou et al. (2001) Kadhum et al. (2007) Kono et al. (1992) Lupow et al. (2002) Diaz Arrieta et al. (2006) Pereira-Barretto et al. (1983)
Prospective Prospective Prospective Retrospective Prospective Prospective
95 150 12 115 72 22
7.5 ± 5 – 47.7 ± 13.2 – 62.4 ± 10 36.0 ± 6.4
Posterioanterior Posterioanterior Posterioanterior Posterioanterior Posterioanterior Posterioanterior
Ref. [3] [4] [5] [6] [2] [7]
Age is expressed as mean ± standard deviation. CXR: Chest x-ray.
chest radiography but normal left ventricular end diastolic dimension by echocardiography, 27 patients had no cardiomegaly by chest radiography but did have an abnormal left ventricular end diastolic dimension by echocardiography and 129 patients had no cardiomegaly by chest radiography and had normal left ventricular end diastolic dimension by echocardiography. This resulted in a sensitivity of 83.3%, specificity of 42.4%, positive predictive value of 43.5% and negative predictive value of 82.7%. If the pediatric study by Satou et al. is excluded from the analysis then a total of 371 patients were included in the final polled analysis. A total of 125 patients had cardiomegaly by chest radiography and abnormal left ventricular end diastolic dimension by echocardiography, 169 patients had cardiomegaly by chest radiography but normal left ventricular end diastolic dimension by echocardiography, 20 patients had no cardiomegaly by chest radiography but did have an abnormal left ventricular end diastolic dimension by echocardiography and 57 patients had no cardiomegaly by chest radiography and had normal left ventricular end diastolic dimension by echocardiography. This resulted in a sensitivity of 86.2%, specificity of 25.2%, positive predictive value of 42.5% and negative predictive value of 74.0%. Discussion This pooled analysis demonstrates the lack of utility of cardiothoracic ratio in the assessment of left ventricular dilation. The sensitivity of 83.3% and specificity of 42.4% make the cardiothoracic ratio neither valuable as a screening nor a confirmatory test. With a negative predictive value of 82.7%, one can be reassured that a normal cardiothoracic ratio is associated with no left ventricular dilation but with a positive predictive value of 43.5% one cannot be certain that there is left ventricular dilation if the cardiac silhouette appears enlarged.
future science group
When the pediatric study by Satou et al. was removed from the analysis there was minimal difference in all values except for the specificity which was low, thus not changing the utility of the chest radiographs. Sensitivity was 86.2%, specificity was 25.2%, positive predictive value of 42.5 and negative predictive value of 74.0. It is not particularly surprising that removing the pediatric study does not significantly change these values as the same criteria for cardiothoracic ratio is used in all age groups and cardiothoracic ratio does not change with age. As such it seems arbitrary to expect a change between those that are 16 years of age and 26 years of age compared with those who are 30 years of age compared with 60 years of age. This additional analysis demonstrates this. The change in specificity is of minimal consequence as the specificity with or without these pediatric patients is not high enough to deem chest radiographs as a helpful test in positively identifying those with left ventricular dilation but simply to rule out left ventricular dilation. Change in cardiothoracic ratio with age has been studied but primarily in the elderly population. In a study of 110 elderly women, Inoue et al. demonstrated that cardiothoracic ratio increased 2.0% over 9 years of follow-up. While this finding was statistically significant this is unlikely to be of clinical consequence [8] . Potter et al. demonstrated similar findings in 243 men over 12 years of follow-up. A majority of patients still remained under the cutoff for abnormal [9] . Both studies found that part of this increase is due to not only increase in cardiac size but also a decrease in the thoracic diameter that is associated with aging. Thus while there is a change in cardiothoracic ratio with age, the degree of this has minimal clinical significance. Cardiothoracic ratio can be impacted by and implicated in several physiologic derangements, one of which is heart failure. In a study of patients with heart failure, Fukuta et al. demonstrated that increased cardiothoracic ratio on
www.futuremedicine.com
173
Review Loomba, Shah, Nijhawan, Aggarwal & Arora chest radiographs correlated with right-sided chamber size moreso than left sided chamber size [10] . Some studies have further investigated cardiothoracic ratio in the setting of congestive heart failure, with particular interest in determining whether or not cardiothoracic ratio can be utilized to help determine ejection fraction. Philbin et al. found that a negative correlation does exist between cardiothoracic ratio and ejection fraction but this was a weak correlation and the study concluded that cardiothoracic ratio cannot be utilized in estimating ejection fraction [11] . These findings were further confirmed by Clark et al. who determined that chest radiographs are not a reliable indicator of left ventricular dysfunction [12] . Another study found that cardiothoracic ratio can be a prognostic indicator in those with poor ejection fraction but only in a multivariate setting thus further highlighting the findings by the aforementioned studies that cardiothoracic ratio, in isolation, is of limited value in the evaluation of heart failure [13] . While the cardiothoracic ratio does not appear to be useful in predicting left ventricular dilation this does not mean that chest radiographs cannot be used to assess cardiac structure. More sophisticated techniques have been described to evaluate for left ventricular dilation using both the frontal and lateral projections which are of more utility. These, however, do require more time and are less intuitive. Elliott et al. proposed a systematic approach to interpreting chest radiographs with respect to the cardiovascular system. First, the technical integrity of the image must be evaluated. It is necessary to establish whether the radiograph was obtained in the PA or AP projection as the cardiac silhouette appears larger in the AP projection when compared with the PA projection. Additionally, portable chest radiographs obtained with patients in the supine position may also make the cardiothoracic ratio appear larger. This also occurs with films obtained during expiration. Angulation of the radiograph beam can also impact the cardiothoracic ratio observed. Next, in the system proposed by Elliott et al., assessment of the extracardiac structures is done, as these can be associated with cardiac abnormalities. For instance, situs inversus of the abdominal viscera is often associated with heterotaxy and interruption of the inferior vena cava with azygos continuation. Pulmonary vascularity can be used in neonates to assist with the diagnosis of various congenital
174
Future Cardiol. (2015) 11(2)
heart defects as these may present with varying degree of pulmonary blood flow [14,15] . The final step is then to assess the cardiac structures themselves. Discussion of this in its entirety is beyond the scope of the manuscript but assessment of left ventricular dilation can be done utilizing features other than the cardiothoracic ratio. Displacement of the cardiac apex leftward and toward the diaphragm is often seen with left ventricular dilation. A portion of the apex may even appear to be below the diaphragm even and posterior to the gastric bubble. Dilation of the left ventricle also tends to augment the concavity present in the middle of the left heart border [14,15] . The question then arises as to what should be done if the cardiothoracic ratio is found to be abnormal. Without other concerns in the history or physical examination, it is reasonable to obtain an electrocardiogram as it is relatively inexpensive. Echocardiography, however, should not be avoided if the index of suspicion is high using the overall clinical situation. Conclusion Enlarged cardiothoracic ratio may be noted on chest radiograph but is not of particular value in predicting left ventricular dilation. Suspicion of left ventricular dilation secondary to chest radiograph findings should prompt evaluation by electrocardiogram before imaging by echocardiography or other modalities. Future perspective Currently echocardiography is still a relatively expensive and time-consuming endeavor. Thus it is impractical to obtain an echocardiogram on all patients who demonstrate evidence of cardiomegaly on a chest x-ray. It is likely that developments in portable, handheld ultrasound technology will one day allow for rapid and inexpensive echocardiography to determine the left ventricular dimensions if there are concerns for echocardiography. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.
future science group
Cardiothoracic ratio for prediction of left ventricular dilation
Review
EXECUTIVE SUMMARY Frequency of chest radiographs ●●
Chest radiographs are among the most common radiologic examination ordered today.
Cardiomegaly noted by chest radiograph ●●
Cardiomegaly is often appreciated on chest radiographs, leading to concern for left ventricular dilation.
Evaluation of left ventricular dilation ●●
Echocardiography is the definitive diagnostic tool which determines whether there is left ventricular dilation but requires neither time nor cost–effectiveness in all cases.
Additional diagnostic testing for assessment of left ventricular dilation ●●
Electrocardiograms should be obtained as an additional screening when there is concern for cardiomegaly based on chest radiographs. Electrocardiogram abnormalities make subsequent echocardiography reasonable to obtain.
References Papers of special note have been highlighted as: •• of considerable interest 1
2
size in comparison to echocardiography. Med. J. Basrah Univ. 25(2), 48–51 (2007). 5
Yoo S-J. Chest Radiographic Interpretation in Pediatric Cardiac Patients. Yoo S-J, Macdonald C, Babyn PS (Eds). Thieme, NY, USA; Toronto, Canada (2010). Díaz Arrieta G, Mendoza Hernández ME, Hernández Cabrera J et al. Correlacion entre la radiograofia de torax y el ecocardiograma para la valoracion de cardiomegalia en pacientes con hipertension arterial sistemica. Arch. Cardiol. Mex. 76(2), 179–184 (2006).
6
7
•• Large study that investigates the relationship between cardiothoracic ratio and left ventricular dilation as noted by echocardiogram. 3
Satou GM, Lacro RV, Chung T, Gauvreau K, Jenkins KJ. Heart size on chest x-ray as a predictor of cardiac enlargement by echocardiography in children. Pediatr. Cardiol. 22(3), 218–222 (2001).
•• Only pediatric study that investigates the relationship between cardiothoracic ratio and left ventricular dilation as noted by echocardiogram. 4
Kadhum AA, Sharif S, Abd-Hazaa M. Validity of chest x-ray in estimation of cardiac
future science group
8
9
Kono T, Suwa M, Hanada H, Hirota Y, Kawamura K. Clinical significance of normal cardiac silhouette in dilated cardiomyopathy: evaluation based upon echocardiography and magnetic resonance imaging. Jpn. Circ. J. 56(4), 359–365 (1992). Lupow JB, Sivak SL, Boss D. The accuracy of the cardiothoracic ratio as a predictor of cardiac enlargement and dysfunction. Acad. Emerg. Med. 9(5), 462 (2002). Pereira-Barretto AC, Mady C, ArteagaFernandez E et al. Valor do indice cardiotoracico na avaliacao do comprometimento miocardico, correlacao com avaliacao ecocardiografica. Rev. Hosp. Clin. Fac. Med. Sao Paulo 38(1), 43–45 (1983). Inoue K, Yoshii K, Ito H. Effect of aging on cardiothoracic ratio in women: a longitudinal study. Gerontology 45(1), 53–58 (1999). Potter JF, Elahi D, Tobin JD, Andres R. Effect of aging on the cardiothoracic ratio of men. J. Am. Geriatr. Soc. 30(6), 404–409 (1982).
10 Fukuta H, Ohte N, Brucks S, Carr JJ, Little
WC. Contribution of right-sided heart enlargement to cardiomegaly on chest
roentgenogram in diastolic and systolic heart failure. Am. J. Cardiol. 99(1), 62–67 (2007). •• Describes the contribution of right-sided chamber size on the cardiothoracic ratio in the setting of heart failure. 11 Philbin EF, Garg R, Danisa K et al. The
relationship between cardiothoracic ratio and left ventricular ejection fraction in congestive heart failure. Digitalis Investigation Group. Arch. Intern. Med. 158(5), 501–506 (1998). 12 Clark AL, Coats AJ. Unreliability of
cardiothoracic ratio as a marker of left ventricular impairment: comparison with radionuclide ventriculography and echocardiography. Postgrad. Med. J. 76(895), 289–291 (2000). 13 Hammermeister KE, Chikos PM, Fisher L,
Dodge HT. Relationship of cardiothoracic ratio and plain film heart volume to late survival. Circulation 59(1), 89–95 (1979). •• Describes the prognostic value of cardiothroacic ratio in the setting of heart failure. 14 Roberts WC. Adult Congenital Heart
Disease. Davis, PA, USA (1987). 15 Elliott JA, Cowan MD. Exercises in Chest
X-Ray Diagnosis. Butterworths, MA, USA (1987).
www.futuremedicine.com
175
