Pediatric Cardiology

, Volume 29, Issue 6, pp 1048–1053

What is the True Frequency of Carditis in Acute Rheumatic Fever? A Prospective Clinical and Doppler Blind Study of 56 Children with up to 60 Months of Follow-Up Evaluation

Authors

  • Álvaro Manuel Caldas
    • Division of Allergy, Clinical Immunology and RheumatologyUniversidade Federal de São Paulo (UNIFESP)
    • Division of Allergy, Clinical Immunology and RheumatologyUniversidade Federal de São Paulo (UNIFESP)
  • Valdir Ambrosio Moises
    • Discipline of Cardiology, Department of MedicineUniversidade Federal de São Paulo (UNIFESP)
  • Célia Maria Camelo Silva
    • Discipline of Cardiology, Department of MedicineUniversidade Federal de São Paulo (UNIFESP)
  • Cláudio Arnaldo Len
    • Division of Allergy, Clinical Immunology and RheumatologyUniversidade Federal de São Paulo (UNIFESP)
  • Antonio Carlos Carvalho
    • Discipline of Cardiology, Department of MedicineUniversidade Federal de São Paulo (UNIFESP)
  • Maria Odete Esteves Hilário
    • Division of Allergy, Clinical Immunology and RheumatologyUniversidade Federal de São Paulo (UNIFESP)
Original Article

DOI: 10.1007/s00246-008-9242-z

Cite this article as:
Caldas, Á.M., Terreri, M.T.R.A., Moises, V.A. et al. Pediatr Cardiol (2008) 29: 1048. doi:10.1007/s00246-008-9242-z

Abstract

Background This study aimed to evaluate prospectively clinical and echocardiographic findings of patients who had rheumatic fever with and without clinical features of cardiac involvement. Methods For this study, 56 consecutive patients (mean age, 11.4 years) with acute rheumatic fever diagnosed according to the 1992 modified Jones criteria were evaluated at diagnosis, after 3 and 6 months, then at 2 and 5 years. All assessments were performed blindly and included physical and cardiac examination, electrocardiogram, chest X-ray, and two-dimensional color-flow Doppler echocardiography. Results Initial clinical carditis was observed for 27 patients (48.2%), all of whom had positive echocardiographic abnormalities. Echocardiographic abnormalities were observed in 11 patients who had arthritis or chorea presentation without initial clinical carditis. Persistence of the abnormalities was observed at a late follow-up evaluation in 72.7% of the cases. Sydenham’s chorea was present in 8 patients with clinical carditis and in 10 without this disorder, 8 of whom had echocardiographic abnormalities. Conclusion Patients who had acute rheumatic fever without clinical signs of carditis showed acute and late follow-up echocardiographic abnormalities suggestive of cardiac involvement. Clinicians should be attentive for the presence of cardiac involvement among patients with chorea.

Keywords

EchocardiographyJones criteriaRheumatic feverSubclinical carditisSydenham’ chorea

Introduction

Acute rheumatic fever (ARF) is a late nonsuppurative inflammatory complication of upper airways infected by Lancefield group A β-hemolytic Streptococcus. Although the incidence of ARF has dramatically decreased during the past 30 years in North America and Western Europe, it still is a challenging disease with high morbidity and mortality in developing countries [32].

Carditis is the most severe clinical feature of ARF, contributing to long-standing sequelae [9]. Literature reports have shown echocardiographic abnormalities suggestive of carditis in patients with normal clinical cardiac examination [1, 2, 7, 10, 1215, 19, 35]. Two-dimensional Doppler echocardiography (2D Echo Doppler) certainly is the most reliable noninvasive technique for diagnosing valvar regurgitation because it is more sensitive than cardiac auscultation, particularly in mild cases. However, due to the elevated frequency of physiologic regurgitation in the healthy population, whether 2D Echo Doppler can accurately distinguish pathologic from physiologic regurgitation has generated uncertainties. These uncertainties are compounded by the fact that few prospective studies are available evaluating valvar functioning late after the initial ARF episode [5, 8, 18, 22, 29, 36].

Considering the need for adequate detection and control of cardiac involvement in ARF, we conducted a prospective study aimed at evaluating 2D Echo Doppler findings for patients with and without clinical carditis at diagnosis and late during a follow-up period that extended to 5 years.

Patients and Methods

Patients

Patients diagnosed with ARF according to the revised Jones criteria [9] were consecutively selected to participate in the study. Inclusion criteria specified patients younger than 18 with their first episode of ARF. The study enrolled 56 patients (33 boys) with a mean age of 11.4 years (range, 7–14 years). The patients were grouped according to the presence of carditis: group 1 (27 ARF patients with a clinical manifestation of carditis) and group 2 (29 ARF patients without a clinical manifestation of carditis). All parents of the patients gave informed consent for their participation in the study, which had prior approval of the hospital investigational review board.

Methods

All clinical evaluations were performed on the same day by one pediatric rheumatologist and blindly by one pediatric cardiologist. The patients were evaluated at initial presentation, at 3 and 6 months, and then 2 and 5 years after the acute episode. They regularly underwent chest X-rays, electrocardiography, and 2D Echo Doppler.

A comprehensive 2D Echo Doppler was performed blindly by an experienced examiner using a commercially available echocardiographic system with a 2.5-, 3.5-, or 5.0-MHz transducer that had continuous and pulsed Doppler as well as color-flow mapping capabilities. Measurements for the diameters of the left atrium and left ventricle (diastolic and systolic) were obtained, and the ejection fraction was estimated by the cube method. Chambers were considered dilated if the measurements exceeded the reference value according to the weight.

With regard to color-flow mapping, appropriate gain and pulse repetition frequency settings were used, and the following aspects were taken into account. Mitral regurgitation was considered to be present if blue, green, or mosaic signals were seen originating from the mitral valve and spreading into the left atrium during systole. Aortic regurgitation was considered to be present if red, yellow, or mosaic signals were seen originating from the aortic valve and spreading into the left ventricle during diastole [23, 28, 31]. The degree of mitral regurgitation was measured by the relationship between maximal jet areas corrected for the atrial area. A relationship of less than 20% was considered mild, whereas a relationship of 20% to 40% was considered moderate, and more than 40% was considered severe [31].

The degree of aortic regurgitation was measured by the ratio of jet width to outflow tract diameter from the images obtained on the parasternal longitudinal view. A ratio less than 46% was considered mild; 46% to 64% was considered moderate; and more than 64% was considered severe [28]. Physiologic regurgitation was characterized by a small jet of very short duration detected only at the level of the paravalvar region. Isolated tricuspid and pulmonary regurgitations were not considered in this study. All examinations were recorded on videotape (VHS) for future analysis.

The cardiothoracic index was determined from chest X-rays by two different examiners. An index of 0.45 ± 0.05 cm or less was considered to be within normal limits [20].

Electrocardiography was performed using a commercially available electrocardiographic system, and the analysis took into account PR (interval between the beginning of the P wave and the beginning of the R wave of the electrocardiographic trace) interval prolongation and left atrium/ventricle hypertrophy criteria.

To evaluate the concordance among all evaluations (clinical cardiac examination, chest X-rays, electrocardiography, and 2D Echo Doppler), we performed the kappa test, considering the clinical cardiac evaluation as the gold standard. The level of statistical significance was set at a p value less than 0.05.

Results

Of the patients with clinical carditis (group 1, n = 27) 14 presented with associated arthritis, 6 with associated chorea, and 2 with concomitant arthritis and chorea. Five patients had isolated carditis. The demographic, clinical, and laboratory manifestations are shown in Table 1. In the entire sample of 56 patients, 35 (62.5%) had ARF arthritis, 27 (48.2%) had carditis, and 18 (32.1%) had chorea.
Table 1

Demographic, clinical, and laboratory findings for patients with and without clinical carditis

 

Group 1 (n = 27)

Group 2 (n = 29)

Total (n = 56)

Males

13

20

33

Mean age at onset (years)

11.5

11.3

11.4

Major manifestations

    Isolated arthritis

0

19

19

    Isolated chorea

0

10

10

    Isolated carditis

5

0

5

    Carditis/arthritis

14

0

14

    Carditis/chorea

6

0

6

    Carditis/arthritis/chorea

2

0

2

Minor manifestations

    Fever

13

11

24

    Arthralgia

3

1

4

    Elevated acute-phase reactants

27

21

48

    Prolonged PR interval

9

0

9

Evidence of preceding streptococcal infection

27

25

52

PR, interval between the beginning of the P wave and the beginning of the R wave of the electrocardiographic trace

In group 1, the follow-up period was 6 months for all 27 patients, 2 years for 23 patients, and 5 years for 8 patients. In this group, mitral regurgitation murmur was detected in 25 patients, which persisted in 20 patients for at least 90 days. The 2D Echo Doppler findings in this group were mitral regurgitation for 25 patients (92.6%), mitral thickening for 20 patients (74%), and aortic regurgitation for 10 patients (37%). Isolated mitral involvement was experienced by 15 patients and aortic involvement by 2 patients.

The echocardiographic alterations persisted with 20 patients, and had worsened for 2 of these patients at the 2-year follow-up evaluation. The diameter of the left atrium was altered initially in 6 patients (mean, 33.9 mm), and at the end of the follow-up period in 1 patient (mean, 30.7 mm). The diastolic diameter of the left ventricle was altered initially in 4 patients (mean, 47.1 mm) and at the end of the follow-up period in 3 patients (mean, 46.2 mm). The ejection fraction of the left ventricle was normal in all the patients at the beginning (mean, 0.76) and end (mean, 0.71) of the follow-up period.

Among the patients without clinical carditis (group 2, n = 29), arthritis was present in 19 patients and chorea in 10 patients. The demographic, clinical, and laboratory manifestations are shown in Table 1. In this group, 11 patients (37.9%) presented with echocardiographic valvar alterations, indicating subclinical carditis. Follow-up evaluation involved 29 patients at 3 months, 27 patients at 6 months, 24 patients at 2 years, and 15 patients at 5 years. In this group, 2D Echo Doppler detected valvar abnormalities in 11 cases: mild mitral regurgitation in 8 patients at the first evaluation and in 3 patients at the second evaluation. Minimum to mild mitral thickening also was observed in three patients initially and in one patient during follow-up evaluation. Mild aortic regurgitation was associated with mitral regurgitation in three patients initially and in two patients during follow-up evaluation. Abnormalities persisted for eight patients (72.7%) in the consecutive follow-up assessments, and had worsened in two of these patients at 5 years.

The 2D Echo Doppler features of groups 1 and 2 are shown in Table 2. The anteroposterior diameter of the left atrium was normal in all the patients at the beginning (mean, 28.4 mm) and at the last evaluation (mean, 30.6 mm). The diastolic diameter of the left ventricle was altered in only one patient at the last evaluation (mean values of 42.8 mm at the beginning and 47.6 mm at the last evaluation). The ejection fraction of the left ventricle remained normal in all evaluations (mean values of 0.74 at the beginning and 0.73 at the last evaluation).
Table 2

Two-dimensional color flow Doppler echocardiography findings for the patients with rheumatic fever (with and without clinical carditis) at the initial and final evaluations

Group 1

Group 2

Patients

Initial

Final

Patients

Initial

Final

1

Normal

M MR, M AoR

1

Normal

M MR

M AoR

2

Mo MR, Mo AoR

Mo AoR, Mi MT

2

Normal

M MR

Mi MT

Mi AoT

PE

3

Mo MR, S AoR

M MR, M MT

3

M MR, M AoR

S MR

M MT, Mi AoT

Mi AoT

M MT,

M MT

Mi AoT

Mi AoT

4

Mo MR, Mi MT

M MR, M AoR

4

M MR

Normal

Mi AoT

Mi MT, Mi AoT

5

Mo MR, M MT

M MR, M MT

5

M MR, M AoR

M AoR

Mi MT

6

S MR, M MT

S MR, M MT

6

Normal

M MR

7

M MR, M MT

M MT

7

M MR

Normal

M AoR

8

Mo MR, Mi MT

M MR, Mi MT

8

M MR

M MR, M AoR

M AoR

9

M MR

Normal

9

M MR, M AoR

Normal

M MT

10

M MR, M AoR

M MR, M AoR

10

M MR

M MR, Mi MT

Mi MT

Mi MT, M AoT

11

Mo MR, M AoR

Normal

11

M MR

Normal

12

M MR, M MT

M MR, M MT

12

Normal

Normal

M AoT

13

Mo AoR

M AoR

13

Normal

Normal

14

Mo MR, M MT

M MR, M MT

14

Normal

Normal

15

S MR, M MT

Mo MR, M MT

15

Normal

Normal

16

M MR, M MT

Normal

16

Normal

Normal

17

S MR, M MT

Mo MR, S MT

17

Normal

Normal

18

Mo MR/DP

S MR, M MT

18

Normal

Normal

19

M MR, M MT

M MR, M MT

19

Normal

Normal

20

S MR, M MT

Mo MR, M MT

20

Normal

Normal

Mo AoR, M AoT

21

Mo MR, M MT

Normal

21

Normal

Normal

22

Mo MR, M MT

Mo MR, Mo MT

22

Normal

Normal

23

M AoR

Normal

23

Normal

Normal

24

S MR, M MT

Mo MR, M MT

24

Normal

Normal

25

Mo MR, M AoR

Normal

25

Normal

Normal

26

Mo MR, M MT

Mi MT

26

Normal

Normal

27

Mo MR, Mi MT

Normal

27

Normal

Normal

28

Normal

Normal

29

Normal

Normal

M, mild; MR, mitral regurgitation; AoR, aortic regurgitation; Mo, moderate; Mi, minimum; MT, mitral thickening; AoT, aortic thickening; PE, pericardic effusion; S, severe

Chorea was present in 8 group 1 patients and 10 group 2 patients. Of these 10 patients, 8 had 2D Echo Doppler abnormalities. Therefore, 16 (88.9%) of 18 patients with chorea had abnormal echocardiographic findings.

Chest X-ray alterations were present in 11 group 1 patients (40.7%) and 3 group 2 patients (10.3%), resulting in mean cardiothoracic indexes of 0.54 and 0.53 cm, respectively. Electrocardiographic abnormalities were present in 34.6% of the group 1 patients and 20% of the group 2 patients.

We observed a statistically significant concordance of abnormalities between the physical cardiac examination performed by the “blind” cardiologist and the 2D Echo Doppler assessment for the group 1 patients (p < 0.05, concordance analysis). We also found a significant concordance between the chest X-ray and the electrocardiography at the first evaluation for the group 1 patients (p < 0.05, concordance analysis).

Discussion

As an established technique for the evaluation and management of rheumatic cardiac disease, 2D Echo Doppler currently is accepted as more sensitive than cardiac auscultation for detecting cardiac abnormalities. Furthermore, with 2D Echo Doppler, minimal degrees of valvar compromise can be identified and measured. However, the help it may give in the specific diagnosis of subclinical rheumatic carditis (i.e., carditis not evidenced by physical examination) remains controversial and a subject of debate [11].

This dilemma arises from the fact that physiologic regurgitation can be detected by 2D Echo Doppler in healthy individuals [25, 26, 34]. Thus, questions have been raised regarding its accuracy in distinguishing subclinical pathologic from physiologic minimal regurgitation and, in turn, regarding the clinical implications of such findings for the patient [25, 26, 34].

With the aim of contributing to a better understanding of the 2D Echo Doppler’s potential to establish an adequate diagnosis of pathologic valvar regurgitation in ARF, we prospectively and blindly evaluated both clinically and echocardiographically 56 patients with a first acute episode of the disease and followed them up to 60 months. Care was taken to ensure (1) that the cardiologist performing auscultation was blinded to the history, electrocardiography, chest X-ray, and 2D Echo Doppler of each case every time, (2) that the physical examination performed by the cardiologist always took place on the same day as the evaluation by the rheumatologist and the laboratory proceedings, and (3) that the auscultation, echocardiograms, and rheumatologic and laboratory evaluation always were performed by the senior physicians involved in the study. As described in the literature, arthritis was the most frequent manifestation of ARF in our cases, followed by carditis and chorea [30, 32].

We observed a highly significant concordance in detected abnormalities between the cardiac examination and the 2D Echo Doppler assessment among patients with clinical carditis in almost all the analyzed periods, but especially in the first evaluation. Chest X-ray and electrocardiography also showed a significant concordance. The low frequency of the abnormalities probably was due to the absence of congestive heart failure for most of our patients.

With regard to echocardiographic abnormalities in our group 1 patients, we observed a predominance of mitral regurgitation (92.6%), followed by mitral thickening (74%) and aortic regurgitation (37%). Our findings, including aortic insufficiency, corroborate other reports in the literature [3, 21, 33]. A few patients in this group showed an increase in left atrium and left ventricle dimensions, as assessed by echocardiography.

In seven patients (25.9%) with initial clinical carditis, we found a disappearance of clinical and valvar abnormalities during a follow-up period of 3 to 6 months. A variable index of 20% to 40% acute lesion resolution, described in the literature, has been attributed to both a mild degree of valvulitis and maintenance of good compliance with secondary prophylaxis [3, 6, 16, 17].

Echocardiographic findings suggestive of valvar abnormalities in the absence of clinical cardiologic manifestations have been observed by several authors [1, 2, 5, 7, 10, 1215, 19, 24, 27]. In this prospective blinded study, we found 2D Echo Doppler abnormalities in 11 group 2 patients (37.9%) (i.e., without apparent clinical carditis). The literature contains similar findings, but our study, conducted in a completely blinded fashion, observed 38 cases (67.8%) of definitive valvar lesions (echocardiographically) in the whole series, and not just 48.2%, as it would appear by taking only clinical auscultation into account [7, 12]. We also observed that in three group 2 patients, the valvar lesion was detected by 2D Echo Doppler only at 3 months.

The most common valve echocardiographic lesion among the group 2 patients was mitral regurgitation, followed by mitral thickening and aortic regurgitation. For eight patients, the lesions persisted throughout the whole follow-up period and worsened for two patients, underscoring the importance of detecting these lesions. Figueroa et al. [12] also observed persistence of valvar illness in at least 60% of patients with subclinical carditis after 5 years of follow-up evaluation despite continuous prophylaxis with penicillin and no evidence of recurrent attacks. Our data support these authors in their statement that these valvar lesions are not a transitory entity. However, the frequency of valve lesion disappearance for those without clinical carditis is variable in the literature, with rates of 30% to 50% for the long-term follow-up assessment [27, 30]. In addition, only one patient had mild left ventricular dilation during the follow-up assessment, simultaneously with the appearance of mitral regurgitation, which may indicate that the valvar lesions in this group were mild.

There are warnings against the risk of misinterpreting 2D Echo Doppler and consequently exposing individuals without ARF carditis to an unnecessary and prolonged secondary prophylaxis [11]. The extremely varied frequency of valvar insufficiency detected by 2D Echo Doppler in healthy individuals probably is due to the use of differing techniques and divergent concepts of valve regurgitation in Doppler evaluation [5, 8, 18]. Many studies have attempted to describe the characteristics and differences of benign valve regurgitation and organic rheumatic valve dysfunction [4, 29, 36]. We emphasize, however, that 2D Echo Doppler is a subsidiary test and as such should be considered in the context of additional signs and symptoms of ARF.

The indication for steroid treatment of valvar lesions, as demonstrated by 2D Echo Doppler, for patients without clinical carditis, has not been well established. Nor has this been the standard for extended secondary prophylaxis in such cases. The therapeutic approach for these patients should be based on the prevalence of ARF in each specific region [24]. Oskutlu et al. [27] believe that valvar lesions in patients without clinical carditis should be treated with steroids because in their experience, such treatment improved resolution of valvar lesions.

It also is worth mentioning that Elevli et al. [10] reported clinical carditis manifestation in 23% of patients with chorea compared with valvar lesions on 2D Echo Doppler without clinical auscultation in 41%. Thus, 64% of their 22 patients with chorea presented some degree of valvar abnormality. In our study, we observed valvar involvement in 16 (88.9%) of 18 patients with chorea. We conclude, therefore, that chorea patients should have regular initial and follow-up 2D Echo Doppler to detect valvar regurgitation or thickening compatible with carditis.

In conclusion, significant agreement existed between abnormal physical findings by a blinded cardiologist and abnormal 2D Echo Doppler findings during an initial episode of ARF. However, 37.9% of cases without apparent cardiac lesion had valvar abnormalities on 2D Echo Doppler, and the majority of these lesions persisted up to the 5-year follow-up assessment. Chorea seems especially prone to frequently determined valvar lesions without significant clinical findings.

Acknowledgments

This study was funded by Fapesp–Fundação de Amparo à Pesquisa do Estado de São Paulo.

Copyright information

© Springer Science+Business Media, LLC 2008