Echocardiographic Comparison of COVID-19 Patients with or without Prior Biochemical Evidence of Cardiac Injury after Recovery.

Although COVID-19 involves primarily the respiratory system, evidence of pathological involvement of the heart during the acute phase of illness is rapidly growing. Cardiac injury has been demonstrated in COVID-19 patients by detection of changes in cardiac function and increased markers of myocardial damage. Because most COVID-19 patients recover from illness, knowledge of the cardiac outcome of infection is important. The goal of this study was to compare the cardiac structure and function of patients who recovered from COVID-19 with or without prior biochemical evidence of cardiac injury.

We examined 105 patients (age, 57 ± 14 years; 56 male and 49 female patients) who had been hospitalized for COVID-19 at a university hospital and who were recruited after hospital discharge. From March 15 to April 15, 2020, patients had presented with fever and/or respiratory symptoms and tested positive for the SARS-CoV-2 on a nasopharyngeal swab. Depending on clinical severity, patients were admitted either to COVID-19 isolation units for clinical monitoring and oxygen support with high-flow nasal cannula or subintensive/intensive care units for mechanical ventilation. After discharge, echocardiography was done in consecutive patients who attended follow-up visits from April 15 to May 15, 2020. Sixty-one percent of the patients who were discharged from March 15 to April 15, 2002, attended follow-up visits during this period, and therefore a selection bias cannot be excluded. Echocardiography was performed with the use of appropriate personal protective equipment by investigators who were blinded to the clinical and laboratory data. The structure of cardiac chambers and left and right ventricular systolic and diastolic function were assessed. The date of COVID-19 diagnosis was considered as the index date to calculate the timing of cardiac ultrasound examination, which was performed after a median of 41 days (interquartile range, 37-44 days). The study received approval from the Internal Review Board, and all patients gave their written informed consent. Analysis was a comparison of cardiac data that were measured in patients with or without biochemical evidence of cardiac injury during hospitalization, which was defined by the presence of measurable serum troponin I levels.

During hospitalization, 87 (83%) of 105 COVID-19 patients had troponin I levels below the lower limit of detection of 0.015 ng/mL, and the remaining 18 (17%) had measurable values (median, 0.028 ng/mL; interquartile range, 0.015- 0.115). At hospital admission, 13 (72%) of 18 patients with measurable serum troponin I had severe respiratory failure (arterial oxygen saturation < 93%; PaO2/FIO2 < 300) that required mechanical ventilation with either continuous positive airway pressure (n = 5) or orotracheal intubation (n = 8). Table 1 summarizes the results of laboratory tests of COVID-19 patients during hospitalization. Patients with serum troponin I ≥ 0.015 ng/mL had significantly lower lymphocyte count and significantly higher lactate dehydrogenase, C-reactive protein, procalcitonin, D-dimer, ferritin, and interleukin-6 serum levels than patients with troponin I < 0.015 ng/mL (all P < .001).

Table 1

Laboratory variables of patients with COVID-19 with and without increased serum troponin I level during hospitalization

Variable	Troponin I <0.015 ng/mL (n = 87)	Troponin I ≥0.015 ng/mL (n = 18)	Difference (95% CI)	P value
Hemoglobin, g/dL	14.0 ± 1.3	14.2 ± 1.0	–0.2 [–0.9 to 0.5]	.582
White cell count, n/mm3	5,676 ± 2,628	6,395 ± 3,085	–719 [–3.335 to 1.023]	.137
Lymphocyte count, n/mm3	1,019 ± 412	638 ± 214	381 [169 to 592]	<.001
Platelet count, n × 1,000/mm3	205 ± 80	195 ± 59	10 [–31 to 53]	.609
Creatinine, mg/dL	0.96 ± 0.28	1.08 ± 0.27	–0.12 [–0.27 to 0.03]	.114
Alanine aminotransferase, U/L	20 [14-31]	27 [21-42]	–7 [–14 to –1]	.049
Aspartate aminotransferase, U/L	24 [20-32]	36 [27-45]	–12 [–17 to –2]	.007
Total bilirubin, mg/dL	0.49 [0.38-0.65]	0.47 [0.40-0.63]	0.02 [–0.11 to 0.12]	.924
Creatine kinase, U/L	87 [56-133]	114 [77-220]	–27 [–76 to 11]	.150
Lactate dehydrogenase, U/L	437 [354-609]	731 [487-889]	–294 [–379 to –105]	<.001
C-reactive protein, mg/L	41 [10-82]	153 [106-223]	–112 [–145 to –52]	<.001
Procalcitonin, ng/mL	0.05 [0.03-0.12]	0.24 [0.13-0.60]	–0.19 [–0.34 to –0.07]	<.001
D-dimer, ng/mL	465 [305-801]	1,127 [849-2,134]	–662 [–1,350 to 329]	<.001
Ferritin, ng/mL	443 [235-841]	1,430 [1,268-1,786]	–988 [–1,308 to –598]	<.001
Interleukin-6, pg/mL	20 [11-36]	90 [30-275]	–70 [–135 to –17]	<.001

Normally distributed values are shown as means ± SD. Variables with skewed distribution are shown as medians [interquartile range]. Serum troponin I level of 0.015 ng/mL was the inferior limit of detection in our assay. Normality of variables distribution was assessed with the Shapiro-Wilk test. Independent-sample Student's t test was used for comparison between groups of normally distributed variables, and the Mann-Whitney U test was used for comparison of skewed variables.

At follow-up, all COVID-19 patients had troponin I levels below 0.015 ng/mL, and 25 (22 with prior troponin I < 0.015 ng/mL, three with prior troponin I ≥ 0.015 ng/mL) of 105 COVID-19 patients still had positive test results for SARS-CoV-2 on the nasopharyngeal swab. The echocardiographic variables of patients with or without increased serum troponin I during hospitalization are summarized in Table 2 . Mean values of left and right atrial volumes, left and right ventricular dimensions, and left ventricular mass and geometry indexes did not differ in COVID-19 patients with or without prior biochemical evidence of cardiac injury. In addition, no significant differences in left and right ventricular systolic and diastolic function were observed with use of two-dimensional, Doppler, and tissue Doppler imaging cardiac ultrasound between COVID-19 patients with or without prior serum troponin I increase. The mean value of systolic pulmonary artery pressure was 23 mm Hg in patients without prior troponin I increase and 21 mm Hg in patients with prior troponin I increase.

Table 2

Follow-up echocardiographic variables of COVID-19 patients according to serum troponin I levels that were measured during hospitalization

Variable	Troponin I <0.015 ng/mL (n = 87)	Troponin I ≥0.015 ng/mL (n = 18)	Difference (95% CI)	P value
Left heart two-dimensional and Doppler ultrasound:
Left atrial volume, mL	42 ± 19	39 ± 17	3 (–7 to 13)	.579
Left ventricular end-diastolic diameter, mm	47 ± 6	48 ± 7	–1 (–5 to 2)	.388
Left ventricular end-systolic diameter, mm	29 ± 6	28 ± 6	1 (–2 to 4)	.524
Interventricular septum, mm	9.1 ± 1.9	9.5 ± 1.9	–0.4 (–2.0 to 0.1)	.052
Posterior wall, mm	9.0 ± 1.9	9.4 ± 1.3	–0.3 (–1.7 to 0.2)	.107
Left ventricular mass index, g/m2.7	33.9 ± 10.4	38.9 ± 11.4	–5.0 (–10.6 to 0.6)	.079
Relative wall thickness, %	0.391 ± 0.100	0.418 ± 0.086	–0.027 (–0.079 to 0.024)	.295
Left ventricular ejection fraction, %	67 ± 10	70 ± 10	–3 (–8 to 2)	.248
Left ventricular fractional shortening, %	38.5 ± 6.9	42.3 ± 7.2	–3.8 (–7.7 to 0.1)	.053
Mitral annular plane systolic excursion, mm	12.9 ± 2.5	12.8 ± 1.8	0.1 (–1.1 to 1.4)	.823
Peak E wave velocity, cm/sec∗	76 ± 20	68 ± 16	8 (–3 to 18)	.136
E/A ratio∗	1.28 ± 0.56	1.10 ± 0.43	0.18 (–0.11 to 0.47)	.220
E wave deceleration time, msec∗	206 ± 57	220 ± 28	–14 (–43 to14)	.311
Right heart two-dimensional and Doppler ultrasound:
Right atrial volume, mL	42 ± 20	45 ± 16	–3 (–15 to 6)	.362
Right ventricular diastolic area, cm2	21.4 ± 4.9	23.7 ± 4.9	–2.3 (–4.8 to 0.2)	.072
Right ventricular systolic area, cm2	13.0 ± 4.0	14.2 (3.2)	–1.2 (–3.2 to 0.7)	.214
Right ventricular fractional area change, %	0.390 ± 0.121	0.384 ± 0.147	0.006 (–0.058 to 0.072)	.843
Tricuspid annular plane systolic excursion, mm	21.4 ± 3.8	21.5 ± 5.1	–0.1 (–2.2 to 2.0)	.922
Tricuspid regurgitation peak velocity, m/sec†	2.33 ± 0.42	2.24 ± 0.48	0.09 (–0.17 to 0.36)	.477
Systolic pulmonary artery pressure, mm Hg†	23 ± 8	21 ± 8	2 (–3 to 7)	.502
Left heart tissue Doppler imaging‡:
Mitral annular peak systolic S′ velocity, cm/sec	9.1 ± 1.8	9.9 ± 2.1	–0.8 (–1.7 to 0.2)	.127
Mitral annular e′ velocity, cm/sec	11.3 ± 3.7	10.4 ± 2.8	0.9 (–0.9 to 2.8)	.301
Mitral annular e′/a′ ratio	1.13 ± 0.54	0.94 ± 0.48	0.19 (–0.08 to 0.46)	.170
Mitral E/e′ ratio	6.74 ± 2.27	6.97 ± 2.19	–0.23 (–1.43 to 0.96)	.700
Right heart tissue Doppler imaging‡:
Tricuspid annular peak systolic S′ velocity, cm/sec	13.5 ± 2.9	14.7 ± 3.2	–1.2 (–2.7 to 0.4)	.141
Tricuspid annular e′ velocity, cm/sec	14.4 ± 3.7	13.9 ± 3.8	0.5 (–1.4 to 2.4)	.614
Tricuspid annular e′/a′ ratio	1.08 ± 0.57	0.87 ± 0.32	0.21 (–0.07 to 0.50)	.142

Values are means ± SD. The 95% CIs have not been adjusted for multiple testing and should not be used to infer definitive effects. Serum troponin I level of 0.015 ng/mL was the inferior limit of detection in our assay. Independent-sample Student's t test was used for comparison between groups.

E, early-wave transmitral peak velocity; A, late-wave transmitral peak velocity.

Measurement of tricuspid regurgitation peak velocity and calculation of systolic arterial pulmonary pressure could be performed in 85.1% of COVID-19 patients with troponin I < 0.015 ng/mL and 83.3% of COVID-19 patients with troponin I ≥ 0.015 ng/mL, and the mean value was imputed in the remaining patients.

S′, peak systolic velocities were measured at either the mitral or tricuspid annulus with the Doppler beam directed parallel to the myocardial walls; e′, early-diastolic myocardial velocity was measured at either the mitral or tricuspid annulus; a′ late-diastolic myocardial velocity was measured at either the mitral or tricuspid annulus.

This study was done in survivors of COVID-19 soon after hospital discharge and shows that even patients with prior detection of increased troponin do not have any evidence of persistent cardiac dysfunction. Findings should be generalized with caution, because of limitations of the study that include lack of echocardiographic measurements obtained during hospitalization, lack of measurement of left ventricular global longitudinal strain that could detect more subtle cardiac changes, possible inclusion of patients at relatively low risk of persistent cardiac injury as suggested by the small percentage of those who had measurable troponin I, significant differences in drug treatments of COVID-19 in the hospital, and a relatively short period of recruitment (1 month) that might have missed COVID-19 survivors with detectable cardiac dysfunction. Retrospective analyses , reported right ventricular dilatation and systolic dysfunction in a small percentage of COVID-19 patients who had elevated serum troponin, whereas other studies have reported functional involvement of the left ventricle as well that might persist after recovery. In this study, we have not observed structural or functional sequelae in the heart of survivors of COVID-19 more than 1 month after first detection of infection, including those with prior evidence of myocardial injury.