Cardiac dysfunction is common in patients with COVID-19, but whether myocardial dysfunction in COVID-19 stems from preexisting disease or is the result of acute COVID-19 effects on the heart remains unknown. This study examined the magnitude and prognostic significance of COVID-associated changes in left ventricular (LV) and right ventricular (RV) function in patients with pre-and postinfection echocardiograms. The population comprised inpatients (≥18 years) undergoing echocardiogram during COVID-19 hospitalization between March 25 and December 21, 2020, who also had preexisting echocardiograms ≥1 month and <1 year prior to COVID-19 diagnosis and quantifiable strain on each echocardiogram. All echocardiograms were retrieved from multicenter institutional databases within the New York-Presbyterian Hospital network and transferred to a centralized core laboratory using an established COVID-19 registry protocol.1Kim J. Volodarskiy A. Sultana R. Pollie M.P. Yum B. Nambiar L. et al.Prognostic utility of right ventricular remodeling over conventional risk stratification in patients with COVID-19.J Am Coll Cardiol. 2020; 76: 1965-1977Crossref PubMed Scopus (79) Google Scholar Echocardiographic analyses were performed by experienced study investigators (J.K., R.B.D.) for whom high reproducibility of quantitative cardiac remodeling indices has been documented.2Kim J. Srinivasan A. Seoane T. Di Franco A. Peskin C.S. McQueen D. et al.Echocardiographic linear dimensions for assessment of right ventricular chamber volume as demonstrated by cardiac magnetic resonance.J Am Soc Echocardiogr. 2016; 29: 861-870Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar Echocardiograms were obtained as part of routine clinical care (a median of 8 days after hospitalization; interquartile range, 3-26), were performed using standard commercial equipment, and were analyzed via standardized core laboratory protocols. Left ventricular global longitudinal strain (GLS) was quantified via two-, three-, and four-chamber views, and RV GLS was quantified in RV focused views where RV GLS was calculated as an aggregate measure of RV septal and free wall longitudinal strain. Global longitudinal strain was quantified using commercial software (TomTEC, Unterschleissheim, Germany), for which automated border detection was manually adjusted to ensure optimal tracking; GLS is reported as an absolute value. Mortality status was determined via electronic medical record review blinded to echocardiographic analyses. In total, 114 patients (65 ± 16 years, 60% male, 28% coronary artery disease, 31% heart failure history) with echocardiograms pre- (7 [interquartile range, 3-13] months) and acutely post-COVID-19 infection were studied. As shown in Table 1, Table 2, Table 3, Table 4, LV and RV dysfunction prevalence increased following acute COVID-19 infection (35% to 48%, P = .003; and 12% to 31%; P < .001, respectively). Left ventricular function declined post-COVID-19 infection whether assessed based on LV ejection fraction or LV GLS (53.3% ± 15.1% to 50.6% ± 17.8%, P = .02; and 17.5% ± 3.9% to 13.2% ± 4.3%, P < .001, respectively). Right ventricular function also declined as measured via RV GLS (24.6% ± 10.0% to 20.6% ± 5.7%, P < .001). Supporting the concept that acute COVID-19 affects RV hemodynamics, pulmonary hypertension and RV dilation also increased after acute COVID-19 infection (48% to 64%, P = .04; and 33% to 60%, P < .001, respectively). During follow-up (median, 116 days), all-cause mortality was 23%. While the magnitude of change in RV GLS did not predict mortality (hazard ratio [HR] = 1.02 [CI, 0.97-1.06]; P = .49), abnormal RV GLS following acute COVID-19 infection was associated with an over four-fold higher mortality risk (HR = 4.59 [CI, 1.08-19.5]; P = .039).Table 1Baseline clinical indicesClinical characteristicsValueDemographic indices: Age, years65 ± 16 Gender, male60 (67) Body surface area, m21.8 ± 0.24 Heart rate, bpm76 ± 13 Systolic blood pressure, mm Hg125 ± 23 Diastolic blood pressure, mm Hg71 ± 13 Race/ethnicity:White34 (39)Black24 (27)Asian11 (12)Other24 (27)Hispanic25 (29)Cardiovascular risk factors: Hypertension69 (79) Diabetes mellitus39 (44) Obesity20 (23) Coronary artery disease28 (32) Heart failure31 (35) Tobacco use46 (52) Pulmonary disease:Asthma11 (12)Chronic obstructive pulmonary disease5 (6)Baseline cardiovascular medications: Angiotensin-converting enzyme inhibitors28 (32) Statin55 (63) Beta blockers49 (56) Aspirin41 (47)In-hospital clinical course: ICU admission27 (31) Mechanical intubation22 (25) COVID-19 directed medications:Hydroxychloroquine25 (28)Steroids36 (41)Remdesivir13 (15) Pressor requirement19 (21) Laboratory markers:Troponin I, ng/mL1.4 ± 6.6Ferritin, ng/mL2,075.6 ± 3,329.4D-dimer, ng/mL4,662.5 ± 9,017.9C-reactive protein, mg/dL16.0 ± 17.9White blood cell count, 109/L14.7 ±10.4Data presented as mean ± SD or percentage (n). Open table in a new tab Table 2Echocardiographic indices pre and post COVID-19 infectionEchocardiographic indicesPre COVID-19Post-COVID-19P valueEchocardiogram indications: Shock2.3 (3)6.9 (9).07 Dyspnea54 (71)62 (81).35 Biomarker (troponin) elevation5.3 (7)5.3 (7)>.99 Coronary artery disease21 (28)12 (15).02 Valvular disease21 (29)9.2 (12).01 Bacteremia8.4 (11)9.2 (12)>.99 Arrhythmia14 (18)21 (28).35 Chest pain12 (15)11 (14)>.99LV function/morphology: LV ejection fraction, %53.3 ± 15.150.6 ± 17.8.02LV ejection fraction < 55%35 (40)48 (55).003 LV stroke volume, mL69.0 ± 26.757.8 ± 26.3<.001 LV cardiac output, L/minute6.1 ± 5.65.2 ± 6.0<.001 LV end-diastolic volume, mL/m276.2 ± 31.872.8 ± 30.7.02LV end-diastolic dilation∗LV end-diastolic dilation: female patients, >81.4 mL/m2, male patients, >88.5 mL/m2.32 (31)23 (22).30LV end-systolic volume, mL/m239.5 ± 29.941.8 ± 31.9.31LV end-systolic dilation†LV end-systolic dilation: female patients, >34.9 mL/m2, male patients, >40.3 mL/m2.34 (33)37 (32).79 LVIDd, cm5.3 ± 1.05.2 ± 1.0.06 LVIDs, cm4.1 ± 3.23.9 ± 1.2.26 LV myocardial mass, g/m299.3 ± 36.396.8 ± 36.6.36LV hypertrophy‡LV myocardial mass: female patients, >95 g/m2, male patients >115 g/m2.35 (34)28 (26).45 LV GLS, %17.5 ± 3.913.2 ± 4.3<.001Abnormal LV GLS§Abnormal LV GLS: |LV GLS| < 17%.630 (34)80 (91)<.001Left atrial morphology: LA volume, mL/m239.6 ± 17.239.2 ± 17.1.92 LA diameter, cm4.0 ± 0.83.8 ± 1.0.14RV function/morphology: RV diameter, cm3.9 ± 0.84.4 ± 0.8<.001 RV diameter, indexed (cm/m2)2.2 ± 0.52.5 ± 0.5.32 RV dilation¶RV diameter > 4.1 cm.33 (37)60 (66)<.001 TAPSE, cm2.1 ± 0.62.0 ± 0.7.08 S′, cm/sec12.0 ± 3.412.0 ± 4.3.69 RV dysfunction‖TAPSE < 1.6 cm and S' < 10 cm/sec.12 (12)31 (30)<.001 RV GLS, %24.6 ± 10.020.6 ± 5.7<.001Abnormal RV GLS#RV GLS| < 16%.74 (4)65 (74)<.001Hemodynamic and valvular indices: Pulmonary artery systolic pressure, mm Hg36.8 ± 15.240.8 ± 12.6.18 Pulmonary hypertensionPulmonary artery systolic pressure >35 mm Hg.48 (42)64 (51).04 Mitral regurgitation (≥2+)14 (15)16 (18)>.99 Tricuspid regurgitation (≥2+)18 (20)19 (21)>.99 Central venous pressure, mm Hg8.1 ± 4.17.9 ± 4.3.45LVIDd, left ventricular internal diameter, end-diastolic; LVIDs, left ventricular internal diameter, end-systolic.Data presented as mean ± SD or percentage (n).Bold values denote statistical significance. Two-sided P < .05 was considered statistically significant.∗ LV end-diastolic dilation: female patients, >81.4 mL/m2, male patients, >88.5 mL/m2.† LV end-systolic dilation: female patients, >34.9 mL/m2, male patients, >40.3 mL/m2.‡ LV myocardial mass: female patients, >95 g/m2, male patients >115 g/m2.§ Abnormal LV GLS: |LV GLS| < 17%.6Yingchoncharoen T. Agarwal S. Popović Z.B. Marwick T.H. Normal ranges of left ventricular strain: a meta-analysis.J Am Soc Echocardiogr. 2013; 26: 185-191Abstract Full Text Full Text PDF PubMed Scopus (667) Google Scholar¶ RV diameter > 4.1 cm.‖ TAPSE < 1.6 cm and S' < 10 cm/sec.# RV GLS| < 16%.7Park S.J. Park J.H. Lee H.S. Kim M.S. Park Y.K. Park Y. et al.Impaired RV global longitudinal strain is associated with poor long-term clinical outcomes in patients with acute inferior STEMI.JACC Cardiovasc Imaging. 2015; 8: 161-169Crossref PubMed Scopus (85) Google Scholar∗∗ Pulmonary artery systolic pressure >35 mm Hg. Open table in a new tab Table 3Univariate Cox models for all-cause mortalityHR (95% CI)PEchocardiographic measurements: ΔLV GLS∗Post–LV GLS – Pre–LV GLS.1.175 [1.055-1.308].003 ΔRV GLS†Post–RV GLS – Pre–RV GLS.1.016 [0.972-1.062].491Demographic indices: Age, years1.033 [1.005-1.061].018 Gender, male0.666 [0.308-1.443].303 Body surface area, m20.420 [0.074-2.392].323 Heart rate, bpm1.001 [0.996-1.007].630 Systolic blood pressure, mm Hg0.977 [0.954-1.001].062 Diastolic blood pressure, mm Hg0.975 [0.935-1.017].237Cardiovascular risk factors: Hypertension1.295 [0.544-3.083].558 Diabetes mellitus1.303 [0.598-2.841].506 Obesity‡Obesity was defined as body mass index ≥ 30 kg/m2.1.586 [0.666-3.776].298 Coronary artery disease§Coronary artery disease was defined as history of prior myocardial infarction and/or coronary revascularization.1.277 [0.555-2.939].565 Tobacco use¶Tobacco use indicated current and past smoking.1.219 [0.722-2.059].458 Heart failure0.774 [0.325-1.843].561 Pulmonary disease:Asthma1.082 [0.324-3.607].898Chronic obstructive pulmonary disease1.575 [0.372-6.666].538Baseline cardiovascular medications: Angiotensin-converting enzyme inhibitor/angiotensin II receptor blockers1.243 [0.540-2.861].609 Statin0.984 [0.454-2.130].966 Beta blocker1.007 [0.467-2.174].985 Aspirin0.689 [0.307-1.546].366 Diuretic0.540 [0.162-1.803].317 In-hospital clinical course:ICU admission2.709 [1.251-5.866].011Mechanical intubation4.632 [2.138-10.037<.001Pressor requirement4.307 [1.972-9.408]<.001Bold values denote statistical significance. Two-sided P < .05 was considered statistically significant.∗ Post–LV GLS – Pre–LV GLS.† Post–RV GLS – Pre–RV GLS.‡ Obesity was defined as body mass index ≥ 30 kg/m2.§ Coronary artery disease was defined as history of prior myocardial infarction and/or coronary revascularization.¶ Tobacco use indicated current and past smoking. Open table in a new tab Table 4Multivariable Cox models for all-cause mortalityModelHR (95% CI)Model 1:χ2 = 12.667, P = .002 ΔLV GLS1.148 [1.030-1.279].012 Age1.029 [1.000-1.058].054Model 2:χ2 = 15.551, P<.001 ΔLV GLS1.153 [1.039-1.279].007 ICU admission2.405 [1.095-5.283].029Model 3:χ2 = 24.915, P<.001 ΔLV GLS1.124 [1.014-1.245].026 Mechanical intubation3.990 [1.799-8.852]<.001Model 4:χ2 = 23.939, P<.001 ΔLV GLS1.136 [1.027-1.256].014 Pressor requirement3.751 [1.681-8.371].001Bold values denote statistical significance. Two-sided P < .05 was considered statistically significant. Open table in a new tab Data presented as mean ± SD or percentage (n). LVIDd, left ventricular internal diameter, end-diastolic; LVIDs, left ventricular internal diameter, end-systolic. Data presented as mean ± SD or percentage (n). Bold values denote statistical significance. Two-sided P < .05 was considered statistically significant. Bold values denote statistical significance. Two-sided P < .05 was considered statistically significant. Bold values denote statistical significance. Two-sided P < .05 was considered statistically significant. As shown in univariable and multivariable Cox models for all-cause mortality, ΔLV GLS as well as known clinical risk factors for COVID-related mortality including age, intensive care unit (ICU) admission, intubation status, and pressor requirement were all associated with increased mortality risk (P < .05 for all; Tables 4 and 3). Importantly, ΔLV GLS was independently associated with death even after controlling for each clinical risk factor (P < .05 for all). Figure 1 provides Kaplan-Meier survival curves for patients stratified by post-COVID-19 LV and RV GLS. As shown, COVID-19-associated impairments in LV and RV GLS are associated with increased all-cause mortality risk. To test whether the association between magnitude of change in GLS on mortality is due to preexisting cardiac dysfunction versus acute decline in cardiac function, subgroup analyses were performed in patients with preserved LV and RV function (LV ejection fraction ≥ 55% [n = 74]; and tricuspid annular plane systolic excursion [TAPSE] ≥ 1.6 cm and S' ≥10 cm/sec [n = 102], respectively). Among patients with preserved LV function prior to COVID infection, ΔLV GLS was strongly associated with increased mortality risk (HR = 1.202 [1.055-1.370]; P = .006), paralleled by a similar augmented mortality risk in patients with preserved RV function preinfection (HR = 1.159 [1.038-1.293]; P = .009). These findings support the concept that acute decrements in LV function, rather than preexisting cardiac dysfunction, contribute to increased mortality risk among patients with COVID-19 infection. This is the first longitudinal echocardiographic study to our knowledge to examine the impact of COVID-19 on biventricular contractile function in relation to pre-COVID-19 infection status. Recent work from our group and others has shown the prognostic utility of RV function for COVID-19 risk stratification.1Kim J. Volodarskiy A. Sultana R. Pollie M.P. Yum B. Nambiar L. et al.Prognostic utility of right ventricular remodeling over conventional risk stratification in patients with COVID-19.J Am Coll Cardiol. 2020; 76: 1965-1977Crossref PubMed Scopus (79) Google Scholar,3Li Y. Li H. Zhu S. Xie Y. Wang B. He L. et al.Prognostic value of right ventricular longitudinal strain in patients with COVID-19.JACC Cardiovasc Imaging. 2020; 13: 2287-2299Crossref PubMed Scopus (323) Google Scholar,4Beyls C. Bohbot Y. Huette P. Abou-Arab O. Mahjoub Y. Tricuspid longitudinal annular displacement for the assessment of right ventricular systolic dysfunction during prone positioning in patients with COVID-19.J Am Soc Echocardiogr. 2020; 33: 1055-1057Abstract Full Text Full Text PDF PubMed Scopus (10) Google Scholar Echocardiogram-quantified adverse RV remodeling conferred a more than twofold increased risk for death, after controlling for age and biomarker elevations.1Kim J. Volodarskiy A. Sultana R. Pollie M.P. Yum B. Nambiar L. et al.Prognostic utility of right ventricular remodeling over conventional risk stratification in patients with COVID-19.J Am Coll Cardiol. 2020; 76: 1965-1977Crossref PubMed Scopus (79) Google Scholar Left ventricular abnormalities in patients with COVID-19 have also been shown to be associated with increased risk of in-hospital mortality with progressively greater risk related to worsening LV function.5Faridi K.F. Hennessey K.C. Shah N. Soufer A. Wang Y. Sugeng L. et al.Left ventricular systolic function and inpatient mortality in patients hospitalized with coronavirus disease 2019 (COVID-19).J Am Soc Echocardiogr. 2020; 33: 1414-1415Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar Whereas these prior studies examined the prognostic role of echocardiographic abnormalities in COVID-19 infection, absence of pre-COVID-19 contractile assessment precluded determination as to whether cardiac remodeling and contractile abnormalities in COVID-19 are due to preexisting conditions or to acute infection and systemic illness. In conclusion, our findings demonstrate decrements in LV and RV function following COVID-19 infection to be closely linked to increased mortality risk, even among patients without preexisting cardiac dysfunction. Larger studies are needed to assess other clinical factors that may impact outcomes as well as the long-term consequences of these changes in COVID-19 survivors.