OUP user menu

Diastolic function in Chagas' disease: An echo and tissue Doppler imaging study

M.V.L Barros, F.S Machado, A.L.P Ribeiro, M.O da Costa Rocha
DOI: http://dx.doi.org/10.1016/S1525-2167(03)00078-7 182-188 First published online: 1 June 2004

Abstract

Aims: With the purpose of studying left ventricular filling in Chagas' disease (Chd), we evaluated 169 patients with Chd using echocardiography and Doppler and tissue Doppler imaging (TDI).

Methods and results: The patients were divided into four groups according to the pattern of left ventricular filling: Group 0—normal filling pattern, Group 1—abnormal relaxation, Group 2—pseudonormal flow pattern, and Group 3—restrictive pattern. All patients were submitted to TDI of the basal portion of the left ventricle's walls. Diastolic dysfunction was found in 21.3% of the patients, with a strong correlation between the worsening of diastolic function and ejection fraction (r = 0.78, p<0.001). TDI septal e′ wave measurement was the best method for the detection of any kind of diastolic dysfunction. Considering a cut-off point of 11 cm/s, a reduced e′ wave value has 97% sensitivity, 84% specificity, 62% positive predictive value, and 99% negative predictive value. The septal E/e′ ratio was the best index for the detection of advanced diastolic dysfunction. Considering a cut-point of 7.2, an elevated E/e′ ratio has 100% sensitivity, 88% specificity, 54.2% positive predictive value, and 100% negative predictive value.

Conclusion: This study showed the characterization of the various patterns of left ventricle diastolic function by echocardiography and Doppler in Chagas' disease and the usefulness of TDI in the assessment of diagnosis of diastolic dysfunction in this disease.

Keywords
  • Chagas' disease
  • tissue Doppler imaging
  • diastolic function

1 Introduction

Chagas' disease is an infection caused by the protozoan Trypanosoma cruzi. The number of infected people in Latin American countries is estimated to be about 20 million, with a marked social and economic impact.1 Heart involvement by Chagas' disease is the major mechanism responsible for the high morbidity and mortality rate in this population. Although diastolic dysfunction has been demonstrated to occur even during the early stages of the disease,2 the analysis of the different patterns of progress of left ventricle (LV) diastolic dysfunction by pulsed wave Doppler (PWD) echocardiography has not been well characterized.

Tissue Doppler imaging (TDI) permits a direct quantitative measurement of the contraction and relaxation velocities of the cardiac muscle. We have previously demonstrated the utility of TDI in the evaluation of early and regional diastolic dysfunction in Chagas' disease as well as in the assessment of cases which pose a special challenge for interpretation based only on the usual indexes obtained by conventional PWD.3,4 The goals of this study were to characterize the different patterns of left ventricular filling by PWD in Chagas' disease and to verify if TDI could be useful in the diagnosis and stratification of diastolic function in this disease.

2 Material and methods

From January 1998 to April 2000, 169 Chagas' disease patients with no other cardiac abnormalities were selected at the Reference Center for Treatment of Infectious and Parasitic Diseases (CTR-DIP) of the UFMG after being screened by standardized clinical examination, ECG, chest X-ray, laboratory tests (including serum exams for T. cruzi) and Doppler echocardiography. We regarded Chagas' disease patients as those individuals whose serum tested positive for T. cruzi by at least two of the three different techniques available. The exclusion criteria were systemic arterial hypertension, coronary artery disease, rheumatic diseases, diabetes mellitus or glucose intolerance, thyroid dysfunction, kidney failure, chronic obstructive pulmonary disease, hydroelectrolyte disorders, significant anemia, pregnancy, and situations that could interfere in the diagnosis of diastolic function by Doppler evaluation, e.g., significant mitral or aortic valvular disease, pacemaker and frequent arrhythmias.

All patients were submitted to a complete Doppler echocardiographic study using a commercially available ultrasound system equipped with TDI (ATL HDI 5000, Bothell, Washington, DC, USA). The following echocardiographic parameters were assessed: LV end-systolic and end-diastolic diameters, left atrial diameter, wall motion score and LV ejection fraction by the method of Simpson. PWD of transmitral flow was recorded from the apical four chamber view, with the Doppler sample volume being placed at the level of the mitral valve leaflet tips in order to measure peak early (E) wave and atrial contraction (A) wave velocities and early filling deceleration time (DT). Isovolumic relaxation time (IVRT) was measured by placing the Doppler sample volume at the LV outflow tract. When necessary, pulmonary venous flow was recorded by placing the sampling volume 1 cm into the right upper pulmonary vein and measuring the peak systolic wave, diastolic wave, and atrial reversal wave velocities. In order to classify the various patterns of LV filling we used the published criteria,5 and divided the patients into four groups: Group 0 = normal filling pattern, Group 1 = abnormal relaxation, Group 2 = pseudonormal flow pattern, and Group 3 = restrictive flow pattern.

The septal, lateral, inferior, anterior, and posterior LV walls were assessed by TDI through the apical window at apical-4c, apical-2c and long apical views. The sample volume was placed at the basal portion of the mentioned walls in order to determine the longitudinal expansion velocities, which were obtained by measuring the initial (e′) and final (a′) diastolic velocities and the ratio between them from three averaged consecutive beats. The E/e′ ratio was calculated by dividing the peak early wave of transmitral flow (E) to initial diastolic wave (e′) by TDI.

The clinical and echocardiographic data were analyzed using the statistical packages Epiinfo version 6 and MINITAB version 11. In all tests, the level of significance was set at p<0.05. Qualitative variables were compared among the four groups by the 2×k test. Quantitative variables are reported as means of the average and standard deviation, analysis of variance (ANOVA) was carried out and the averages for the four groups were compared by the Fisher test. Pearson's rank correlation test was used to correlate several parameters of systolic and diastolic function related to the diastolic function pattern.

For each e′ wave and E/e′ ratio index, a receiver operating characteristic (ROC) curve was constructed by plotting the sensitivity (true positive rate) against 1-specificity (false positive rate). Cut-off points for the detection of any kind of diastolic dysfunction and advanced diastolic dysfunction were chosen and sensitivity, specificity, predictive values and likelihood ratios were calculated.

3 Results

Table 1 describes the general clinical and echocardiographic parameters obtained for the groups studied. Among the 169 patients evaluated, diastolic dysfunction, as detected by Doppler criteria, was found in 21.3% of the Chd patients. Age was significantly higher in patients with an abnormal LV relaxation flow pattern (Group 1), but did not correlate significantly with TDI measurements. We observed a clear deterioration of ejection fraction (r = 0.75, p<0.001) associated with enlargement of the left atrial (r = 0.68, p<0.001) and ventricular (r = 0.66, p<0.001) cardiac chambers related to the deterioration of diastolic function among the groups. The PWD parameters of transmitral flow demonstrated a bimodal pattern among the groups (Fig. 1).

Figure 1

Pulsed wave transmitral Doppler parameters in Chagas' disease patients showing a non-linear distribution among the diastolic function groups.

View this table:
Table 1

General and echocardiographic parameters in 169 Chagas' disease patients with normal (Group 0) and altered (Groups 1–3) LV diastolic function

VariableGroup 0Group 1Group 2Group 3
N13316155
Age40.9 ± 9.454.4 ± 7.6*44 ± 11.9§41 ± 8.3§
LVDD (mm)50.6 ± 4.955.2 ± 5.9*63.4 ± 7.6*§65.7 ± 3.9*§
LVSD (mm)34.3 ± 5.040.4 ± 7.7*50.7 ± 9.4*§50.0 ± 13.9*§
LA (mm)34.3 ± 2.436.8 ± 3.8*41.8 ± 5.5*§44.0 ± 2.9*§†
EF (%)60.9 ± 6.550.7 ± 11.5*36.2 ± 7.8*§31.5 ± 6.9*§
SWMA (%)4275*100*§100*§
E (cm/s)72.5 ± 10.151.3 ± 14.2*75.5 ± 12.9§78.5 ± 20.2§
A (cm/s)52.4 ± 5.269.9 ± 18.2*53.1 ± 15.2§31.5 ± 7.6*§†
IVRT (ms)90.6 ± 6.7107.4 ± 5.0*88.6 ± 8.7§71.0 ± 10.6*§†
DT (ms)184.9 ± 17.3240.0 ± 36.3*189.8 ± 28.8§146.0 ± 11.2*§†
  • LVDD, left ventricular end-diastolic diameter; LVSD, left ventricular end-systolic diameter; LA, left atrial diameter; EF, left ventricular ejection fraction; SWMA, segmental wall motion abnormality; E, peak early transmitral flow wave velocity; A, atrial contraction velocity; IVRT, isovolumic relaxation time; DT, early filling deceleration time. *p<0.05 versus Group 0; §p<0.05 versus Group 1; †p<0.05 versus Group 2.

The analysis of TDI e′ wave showed a linear reduction related to the deterioration of diastolic dysfunction in all the LV walls evaluated (Table 2, p<0.001). This pattern was most evident between Chagas' disease patients with normal LV filling and the group with an abnormal LV relaxation flow pattern. The E/e′ ratio showed a close correlation with LV diastolic dysfunction in Chd patients (Fig. 2), allowing discrimination among all the patterns of diastolic function by Fisher's pairwise comparison test, except between Groups 0 and 1 at the septal and inferior walls (Table 3).

Figure 2

E/e′ ratio boxplots in diastolic function groups of Chagas' disease patients showing a clear discrimination among the groups evaluated.

View this table:
Table 2

TDI e′ wave in left ventricular walls and diastolic function groups in Chagas' disease

Wall/Group0123
Septal13.8 ± 3.18.3 ± 1.9*8.0 ± 1.2*6.5 ± 0.5*
Lateral16.6 ± 3.99.3 ± 2.3*8.9 ± 1.6*7.0 ± 0.1*
Inferior14.5 ± 3.68.9 ± 2.9*8.1 ± 1.5*5.5 ± 1.0*
Anterior16.3 ± 3.910.3 ± 2.8*9.2 ± 1.7*7.2 ± 0.5*
Posterior15.5 ± 3.98.6 ± 2.8*8.2 ± 1.8*5.7 ± 0.5*
  • *p<0.05 versus Group 0; §p<0.05 versus Group 1; †p<0.05 versus Group 2.

View this table:
Table 3

E/e′ ratio in the left ventricular walls and diastolic function groups in Chagas' disease

Wall/Group0123
Septal5.5 ± 1.36.3 ± 2.09.7 ± 2.3*§11.9 ± 2.3*§†
Lateral4.7 ± 1.56.2 ± 2.5*8.7 ± 1.9*§11.2 ± 2.8*§†
Inferior5.8 ± 1.96.5 ± 2.69.4 ± 2.2*§15.3 ± 6.1*§†
Posterior5.2 ± 2.06.7 ± 2.1*9.4 ± 2.4*§13.5 ± 2.3*§†
Anterior4.7 ± 1.35.6 ± 1.9*8.4 ± 1.8*§10.7 ± 2.0*§†
  • *p<0.05 versus Group 0; §p<0.05 versus Group 1; †p<0.05 versus Group 2.

Septal e′ wave measurement was the best index for the recognition of any kind of diastolic dysfunction (Table 4), with an area under the ROC curve of 0.96 (CI 95% 0.92–0.99). Considering the cut-off point of 11 cm/s, a reduced e′ wave value has 97% sensitivity, 84% specificity, 62% positive predictive value, and 99% negative predictive value. The positive likelihood ratio, an effective measure of the capacity of a test to diagnose a condition, was 5.9 (3.9–8.9) and the negative likelihood ratio was 0.03 (0.01–0.25). At variance, the septal E/e′ ratio was the best index in the detection of advanced diastolic dysfunction, with an area under the ROC curve of 0.96 (CI 95% 0.94–1.00). Considering the cut-off point of 7.2, an elevated E/e′ ratio has 100% sensitivity, 88% specificity, 54.2% positive predictive value and 100% negative predictive value. The positive likelihood ratio was 8.1 (5.1–12.8) and the negative likelihood ratio was 0 (Table 5).

View this table:
Table 4

Results of the diagnostic tests of e′ TDI wave and E/e′ index in the diagnosis of diastolic dysfunction in Chagas' disease

Cut-off point (cm/s)Area (95% CI)S (95% CI)Sp (95% CI)+PV (95% CI)−PV (95% CI)+LR (95% CI)−LR (95% CI)
e′ TDI wave
Septal110.96 (0.92–0.99)97 (82–99)84 (75–90)62 (47–75)99 (94–100)5.9 (3.9–9.0)0.03 (0.005–0.25)
Lateral140.95 (0.91–0.98)96 (82–99)82 (73–88)60 (45–73)99 (94–100)5.4 (3.6–7.9)0.04 (0.006–0.26)
Inferior110.91 (0.86–0.98)91 (74–97)80 (72–87)56 (41–79)97 (90–99)4.6 (3.1–6.7)0.11 (0.04–0.34)
Posterior120.93 (0.89–0.97)94 (78–99)82 (73–88)59 (44–72)98 (92–100)5.2 (3.5–7.7)0.07 (0.02–0.29)
Anterior130.95 (0.90–0.98)94 (78–99)85 (76–90)62 (47–76)98 (92–100)6.0 (3.9–9.3)0.07 (0.02–0.28)
E/e′ index
Septal6.00.86 (0.76–0.95)81 (63–92)78 (70–85)51 (37–65)93 (86–97)3.8 (2.6–5.5)0.23 (0.12–0.49)
Lateral5.40.87 (0.78–0.95)84 (66–94)83 (74–89)57 (42–71)95 (88–98)4.9 (3.2–7.5)0.05 (0.02–0.12)
Inferior5.90.82 (0.73–0.92)84 (66–94)74 (65–81)47 (34–61)94 (87–98)3.3 (2.3–4.6)0.21 (0.09–0.47)
Posterior5.40.87 (0.80–0.95)88 (70–96)76 (67–83)50 (36–63)96 (89–99)3.6 (2.6–5.1)0.16 (0.07–0.41)
Anterior5.20.86 (0.77–0.95)81 (63–92)79 (71–86)52 (38–66)94 (87–97)3.9 (2.7–5.8)0.24 (0.11–0.49)
  • Area, area under ROC curve; S, sensitivity (%); Sp, specificity (%); +PV, positive predictive value (%); −PV, negative predictive value (%); +LR, positive likelihood ratio; −LR, negative likelihood ratio.

View this table:
Table 5

Results of the diagnostic tests of e′ wave by TDI and E/e′ index in the diagnosis of advanced diastolic dysfunction (pseudonormal or restrictive pattern) in Chagas' disease

Cut-off point (cm/s)Area (95% CI)S (95% CI)Sp (95% CI)+PV (95% CI)−PV(95% CI)+LR (95% CI)−LR (95% CI)
e′ TDI wave
Septal110.93 (0.89–0.97)100 (79–100)76 (67–83)38 (25–53)100 (95–100)4.2 (3.1–5.7)0
Lateral140.91 (0.86–0.96)100 (79–100)74 (66–81)37 (24–51)100 (95–100)3.9 (2.9–5.2)0
Inferior110.89 (0.83–0.94)95 (72–100)74 (65–81)35 (22–49)99 (94–100)3.6 (2.6–4.9)0.07 (0.01–0.48)
Posterior120.90 (0.85–0.96)95 (72–100)74 (66–82)35 (23–50)99 (94–100)3.7 (2.7–5.1)0.07 (0.01–0.47)
Anterior130.92 (0.88–0.97)94 (72–100)77 (68–84)38 (24–53)99 (94–100)4.1 (2.9–5.7)0.07 (0.01–0.46)
E/e′ index
Septal7.20.97 (0.94–1)100 (79–100)88 (80–93)54 (37–71)100 (96–100)8.1 (5.1–12.7)0
Lateral5.40.95 (0.93–0.99)100 (79–100)78 (70–85)40 (27–56)100 (95–100)4.6 (3.3–6.4)0
Inferior6.20.93 (0.88–0.97)100 (79–100)77 (68–84)39 (26–54)100 (95–100)4.3 (3.1–5.9)0
Posterior6.50.94 (0.90–0.98)95 (72–100)81 (73–87)42 (27–58)99 (94–100)4.9 (3.4–7.1)0.06 (0.01–0.44)
Anterior5.30.96 (0.93–0.99)100 (79–100)81 (73–87)43 (29–59)100 (96–100)5.2 (3.6–7.3)0
  • Area, area under ROC curve; S, sensitivity (%); Sp, specificity (%); +PV, positive predictive value (%); −PV, negative predictive value (%); +LR, positive likelihood ratio; −LR, negative likelihood ratio.

4 Discussion

The assessment of LV filling plays an important role in our understanding of the severity of the physiological damage caused by a cardiopathy. The development of LV diastolic dysfunction carries a negative prognosis for the outcome of patients with several pathological heart conditions, and sometimes the proper assessment of LV filling physiology may represent one of the most important aspects of the global clinical decision making and therapy strategy.6–9

Chagas' cardiopathy may lead to impairment of both phases of LV filling. Initially, it alters ventricular relaxation and progressively causes disorders related to the compliance of the chamber. The mechanisms involved may result from alterations affecting different components of the heart, and not only the myocardial fibers, but also the support system, interstice, conductor system, autonomic nervous system and vascular integrity.10

Diastolic function parameters have already been studied thoroughly by PWD echocardiography and is considered useful to assess several physiological parameters related to diastolic function such as relaxation, compliance and ventricular filling pressure.11–14 However, the diversity of the conditions that may, singly or jointly, affect diastolic function impairs the determination of the degree of involvement of any given functional aspect of the heart, and it is frequently necessary to use clinical data, maneuvers, and correlation of complementary echocardiograph parameters for a more precisely defined diagnosis.15,16

Few studies in the literature have used echocardiography and Doppler in the evaluation of diastolic function in Chagas' disease. Generally speaking, these studies focused on aspects related to early abnormalities in LV relaxation found in Chagas' disease patients.2,17,18 The present work shows the characterization of the various patterns of LV diastolic function with the use of echocardiography and Doppler, showing a strong correlation between systolic dysfunction and impairment of LV filling in Chagas' disease.

TDI has proved to be very promising in the evaluation of ventricular relaxation in several pathologies, including Chagas' disease.3,4,19–22 The characteristics of the motion of the myocardium walls along the cardiac cycle can be analyzed quantitatively, providing a more accurate assessment of myocardium behavior. Furthermore, the expansion waves from the ventricular walls assessed by TDI are less influenced by load alterations and are primarily related to ventricular relaxation.23,24

The study of LV diastolic function by TDI shows a peculiar characteristic in patients with Chd. In contrast to the parameters available by PWD, the e′ wave by TDI showed a satisfactory correlation with the different patterns of diastolic LV function, with an inverse relationship between e′ wave values and the worsening of diastolic dysfunction (Table 2). This characteristic pattern has been previously demonstrated by Farias et al.25 Moreover, the use of the E/e′ ratio provided excellent discrimination among the various patterns of diastolic function. This ratio has been demonstrated to be useful in the diagnosis of elevated filling pressures in the LV.26

The diagnostic performance of TDI in the identification of advanced (Table 5) or any form of diastolic dysfunction (Table 4) makes it a method of choice in the detection and stratification of diastolic dysfunction in Chagas' disease patients. Measurement of the e′ wave is particularly well suited to the recognition of any kind of diastolic dysfunction, although the E/e′ ratio was the best index in the detection of advanced diastolic dysfunction. The analysis is highly effective in all ventricular walls, but is slightly superior in the septal wall.

Among the limitations of this study, it is important to mention the noninvasive nature of the methods used here to assess LV diastolic function.27 The measurements obtained by TDI are not related exclusively to the expansion of the myocardial fibers but may also be influenced by the translation movement of the heart even though the use of the longitudinal axis for the analyses and measurements done at the cardiac base minimized these effects. In our sample, segmental wall motion is strongly correlated with systolic and diastolic function and might have influenced the measurement of TDI diastolic indexes. However, we argue that segmental wall motion is not responsible for the different TDI patterns observed among diastolic function groups. We found a similar behavior of E/e′ index and e′ value in all the walls we evaluated, independent of the presence or absence of wall motion abnormalities. Indeed, this progressive reduction was observed both in the anterior wall, where there was a low prevalence in the segmental wall abnormality in Chagas' disease, and in the posterior wall, where it was described as a high prevalence of segmental wall dysfunction. We suggest that this evaluation should be performed in a wall free of systolic segmental abnormality.

To conclude, this study shows, the characterization of the different patterns of LV diastolic function by eco-Doppler cardiography and the close correlation with parameters of LV systolic function in Chagas' disease. Moreover, by using the TDI e′ wave and E/e′ ratio we were able to discriminate all the patterns of diastolic function with statistically significant differences, thus demonstrating the usefulness of TDI in the assessment of diagnosis and progression of diastolic dysfunction in this disease. The prognostic impact of different patterns of LV filling in Chagas' disease and the use of TDI needs to be evaluated in prospective studies in order to determine the importance of the above findings in the risk stratification of Chagas' disease patients.

Acknowledgements

The study was partly financed by CNPQ and PRPq.

References

View Abstract