Doppler echocardiographic evaluation of left ventricular diastolic function in patients with systemic lupus erythematosus Subclinical myocardial involvement frequently occurs in patients with systemic lupus erythematosus (SLE). In this study, left ventricular diastolic function was assessed in 58 patients (54 female and 4 male; mean age 32 + 11 years) and in 40 sex-matched and age-matched healthy control subjects (37 female and 3 male; mean age 33 + 9 years) by means of pulsed Doppler echocardiography. All subjects had no clinical evidence of overt myocardial disease or abnormal left ventricular systolic function. Compared with the control group, patients with SLE had significantly prolonged isovolumic relaxation time (62 + 12 vs 80 -+ 14 msec; p < 0.011, reduced peak early diastolic flow velocity (peak E) (62 t- 18 vs 76 + 16 cm/set; p < 0.05), increased peak late diastolic flow velocity (peak A) (45 + 7 vs 53 + 8 cm/set; p < O.Ol), reduced E/A ratio (1.81 f 0.32 vs 1.46 + 0.29; p < O.OOl), and lower deceleration rate of early diastolic flow velocity (EF slope) (489 + 151 vs 361 + 185 cm/sec2; p < 0.05). Subgroup analysis according to disease activity revealed that when compared to the inactive disease group, the active disease group had significantly longer lsovolumic relaxation time (74 ? 21 msec vs 92 rt 16 msec; p < O-01), lower peak E (78 + 16 vs 70 ? 16 cm/set; p < O-05), higher peak A (51 * 8 vs 57 I 7 cm/set; p < O.Ol), lower E/A ratio (1.56 ? 0.25 vs 1.26 + 0.27; p < 0.001) and lower EF slope (379 f 169 cm/sec2 vs 324 f 172 cm/sec2; p < 0.05). In conclusion, an abnormal pattern of left ventricular diastolic filling dynamics occurs in patients with SLE. These abnormalities may represent myocardlal involvement in SLE and may be related to disease activity. (AM HEART J 1990; 120:82.)

Wing-Hung Leung, MB, MRCP (UK), Kee-Lam Wong, MB, FRACP, Chu-Pak Lau, MD, MRCP (UK), Cheuk-Kit Wong, MB, MRCP (UK), Chun-Ho Cheng, MB, MRCP (UK), and Yau-Ting Tai, MB, MRCP (UK). Hong Kong

Although myocardial abnormalities are frequently found in autopsy studies of systemic lupus erythematosus (SLE), clinically manifested myocardial dysfunction is uncommon.ly 2 Left ventricular systolic abnormalities have been identified in asymptomatic patients with SLE by means of various noninvasive techniques3-7; however, more invasive hemodynamic studies have suggested that left ventricular diastolic dysfunction may be the predominant feature.8* g Recently, pulsed Doppler echocardiography has been shown to be useful in evaluating left ventricular diastolic filling dynamics,lO-l4 and a close correlation with those measured from cineangiographyll

and radionuclide angiography has been demonstrated.14 This noninvasive method has identified abnormal left ventricular diastolic function in patients with left ventricular hypertrophy due to hypertension or hypertrophic cardiomyopathy,1° ischemic heart disease,15 acute cardiac allograft rejection,16 diabetic cardiomyopathy17 and doxorubicin cardiotoxicity. l8 In these disease states, the diastolic abnormalities often precede systolic dysfunction. Therefore the present study was initiated to evaluate left ventricular filling dynamics by using pulsed Doppler echocardiography in patients with SLE who had no evidence of overt myocardial disease or abnormal systolic function.

From Hong

METHODS

the Department of Medicine, Kong, Hong Kong.

Received

for publication

Reprint requests: Falk Cardiovascular Medical Center,

4/l/20461

82

Sept.

Queen

20, 1989;

Dr. Wing-Hung Leung, Research Center, Stanford, CA 94305.

Mary

accepted

Hospital,

University

Feb. 20, 1990.

MB, MRCP, CVRC 161,

Cardiology Stanford

Division, University

of

Consecutive patients followed up in our Rheumatology Clinic with established SLE who fulfilled the revised American Rheumatism Association criteria1g were considered for study. Patients with valvular dysfunction, systemic hypertension, clinical or electrocardiographic evi-

Volume Number

120 1

dence of coronary heart disease, histories of previous cardiac disease,alcoholism, diabetes mellitus, and echocardiographic evidence of global or regional systolic dysfunction were excluded from study. Sixty-three patients were considered,but five were not enrolled becauseof unsatisfactory echocardiographic recordings. There were 54 female and 4 male subjects.The agesranged from 16 to 63 years (mean 32 + 11 years). Forty sex-matched and agematched healthy subjects(37 female and 3 male; mean age 33 f 9 years; range 16 to 62 years) were usedasthe control group. Lupus diseasewasconsideredto be active if the patients had at least three of the following features: fever with no apparent infective cause;serositis; recent skin lesionsor exacerbation of old ones; recent onset or acceleration of alopecia; nasobuccal ulcerations; active central nervous system involvement; increasinglymphadenopathy, leukopenia, or thrombocytopenia; an erythrocyte sedimentation rate exceeding50 mm per hour in the absenceof infection; hypocomplementemia; and features of active nephritis.20 Each patient underwent the following diagnostic procedures: a thorough physical examination, standard 12-lead electrocardiogram, antinuclear and anti-DNA antibody titers, and complement measurement. Echocardiography. Patients were studied in the left lateral recumbent position. The examination was completed in 10 to 15 minutes, during which there wasno obvious change in hemodynamic state. M-mode and pulsed Doppler studiesunder two-dimensional echocardiographic guidance were obtained by an ATL MK600 recorder (Advanced Technology Laboratories, Inc., Bothell, Wash.) with a 3 MHz transducer with simultaneouselectrocardiogramsand phonocardiograms.Particular care wastaken to rule out significant valvular abnormalities that are not infrequent in lupus patients21,22and that have been demonstrated to affect mitral inflow velocities. M-mode and pulsedDoppler imageswererecorded on paper with a speed of 100mm/set for subsequentmeasurement.The following M-mode variables were measuredand averagedin five cardiaccycles:left ventricular end-diastolicdimension(LVEDD) and left ventricular end-systolic(LVESD) dimension(from which fractional shortening[FS] wasderived); interventricular septalthickness(IVST) and left ventricular posteriorwall thickness(LVPWT) at end diastole. PulsedDoppler studiesof left ventricular diastolic filling were performed through an apical four-chamber view with the Doppler cursor oriented parallel to the long axis plane of the left ventricle and the samplevolume placedjust below the level of the mitral annulus. According to established methods,12B l3 the following measurements were made on five cardiac cycles: (1) isovolumic relaxation time (IVRT)-taken asthe time interval from the first positive deflection of the aortic component of the second heart soundto the onset of the diastolic flow velocity waveform; (2) peak early diastolic flow velocity (peak E); (3) peak late diastolic flow velocity that is due to atria1contraction (peak A); (4) the ratio between peak early and late diastolic flow velocity (E/A ratio); (5) deceleration rate of early diastolic flow velocity (EF slope). All the M-mode and Doppler echocardiographicdata were measuredby another physi-

LV diastolic function

in SLE

83

cian without knowledgeof the clinical status of the subject. Informed consent was obtained from all patients. Statistical analysis. Data were expressed as mean * standard deviation. Statistical analysiswasperformed with the Student’s unpaired t test and linear regressionanalysis. Statistical significancewas assumedwhen the p value was lessthan 0.05. RESULTS

The SLE and control groups were similar in age, sex, resting heart rate, systolic and diastolic blood pressure, and M-mode echocardiographic findings (Table I). Active disease was present in 19 patients. Subdivision of the SLE group into active and inactive disease showed comparable clinical and M-mode echocardiographic parameters (Table I). However, the active disease group had shorter disease duration (38 -t 36 months vs 74 + 48 months; p < 0.01) than the inactive disease group. Left ventricular diastolic filling indices (Table II and Fig. 1). Compared with the control group, patients with SLE had significantly prolonged IVRT (62 -t 12 msec vs 80 & 14 msec; p < O.Ol), reduced peak E (82 t- 18 cmhec vs 76 t- 16 cm/set; p < 0.05), increased peak A (45 * 7 vs 53 + 8 cm/set; p < O.Ol), reduced E/A ratio (1.81 + 0.32 vs 1.46 ? 0.29; p < O.OOl), and lower EF slope (489 f 151 vs 361 t 185 cm/sec2; p < 0.05). Subgroup analysis according to disease activity revealed that when compared to the inactive disease group, the active disease group had significantly longer IVRT (74 + 21 msec vs 92 & 18 msec; p < O.Ol), lower peak E (78 -t 16 cm/set vs 70 ? 16 cm/set; p < 0.05), higher peak A (51 + 8 cm/set vs 57 f 7 cm/set; p < O.Ol), lower E/A ratio (1.56 + 0.25 vs 1.26 + 0.27; p < 0.001) and lower EF slope (379 +- 169 cm/sec2 vs 324 + 172 cm/ sec2; p < 0.05) (Fig. 1). The E/A ratio of the control, active, and inactive disease groups showed a linear relationship with age (Fig. 2). The ratio in all groups decreased with age. However, the ratio was lower in the SLE group when compared with normal age-matched controls, and patients with active disease had the lowest values. There was no significant correlation of left ventricular diastolic filling indices with disease duration, LVEDD, LVESD, IVST, LVPWT, and FS. Further subdivision of patients with SLE according to the existence of renal disease and corticosteroid/immunosuppressive therapy failed to show any significant difference in these indices between the groups. DISCUSSION

Myocardial involvement in SLE is common in postmortem studies with an overall incidence of about 40 % . lp 2 However, clinically evident myocar-

July

84

Leung et al.

American

*

pco.05

**

***

pio.01 p