key: cord-0720194-yv74udud authors: Clavario, Piero; De Marzo, Vincenzo; Lotti, Roberta; Barbara, Cristina; Porcile, Annalisa; Russo, Carmelo; Beccaria, Federica; Bonavia, Marco; Bottaro, Luigi Carlo; Caltabellotta, Marta; Chioni, Flavia; Santangelo, Monica; Hautala, Arto J.; Griffo, Raffaele; Parati, Gianfranco; Corrà, Ugo; Porto, Italo title: Cardiopulmonary exercise testing in COVID-19 patients at 3 months follow-up() date: 2021-07-24 journal: Int J Cardiol DOI: 10.1016/j.ijcard.2021.07.033 sha: 823d8b19b1373ff1f2090fa20139bf1caa5c6f8f doc_id: 720194 cord_uid: yv74udud BACKGROUND: Long-term effects of Coronavirus Disease of 2019 (COVID-19) and their sustainability are of the utmost relevance. We aimed to determine: 1) functional capacity of COVID-19 survivors by cardiopulmonary exercise testing (CPET); 2) characteristics associated with cardiopulmonary exercise testing (CPET) performance; 3) safety and tolerability of CPET. METHODS: We prospectively enrolled consecutive patients with laboratory-confirmed COVID-19 discharged alive at Azienda Sanitaria Locale-3, Genoa. At 3-month from hospital discharge, complete clinical evaluation, trans-thoracic echocardiography, CPET, pulmonary function test, and dominant leg extension (DLE) maximal strength evaluation were performed. RESULTS: From 225 patients discharged from March to November 2020, we excluded 12 incomplete/missing cases, 13 unable to perform CPET leading to a final population of 200. Median percent-predicted peak oxygen uptake (%pVO2) was 88% (78.3–103.1). Ninety-nine(49.5%) patients had %pVO2 below, whereas 101(50.5%) above the 85% predicted value (indicating normality). Of 61/99 patients with reduced %pVO2 but normal anaerobic threshold, 9(14.8%) had respiratory, 21(34.4%) cardiac, and 31(50.8%) non-cardiopulmonary limitation of exercise. One-hundred sixty(80.0%) patients complain at least one symptom, without relationship with pVO2. Multivariate linear regression analysis showed percent-predicted forced expiratory volume in one-second(β = 5.29,p = 0.023), percent-predicted diffusing capacity of lungs for carbon monoxide(β = 6.31,p = 0.001), and DLE maximal strength(β = 14.09,p = 0.008) independently associated with pVO2. None adverse event was reported during/after CPET neither the involved health professionals developed COVID-19. CONCLUSIONS: CPET after COVID-19 is safe and about 1/3rd of COVID-19 survivors show functional capacity limitation mainly explained by muscular impairment, calling for future research to identify patients at higher risk of long-term effects that may benefit from careful surveillance and targeted rehabilitation. To date, the Coronavirus Disease of 2019 (COVID-19) pandemic accounts for more than 115 million confirmed cases and up to 3 million deaths worldwide [1] . COVID-19 is a mainly respiratory disease, but cardiovascular (CV) alterations are associated with worse prognosis [2] [3] [4] [5] [6] . For the chronic phase, the concerns are the development of pulmonary interstitial disease and/or lingering CV involvement [7] and potentially explained by the Severe Acute Respiratory Syndrome-Coronavirus type 2 (SARS-CoV2)-associated endothelitis [8] . How to intercept, assess, and treat these patients with long-term consequences of COVID-19 remains uncertain [9] [10] [11] . Most patients with COVID-19 are not critically ill: among the large number of COVID-19 patients hospitalized in Italy and discharged, those classified as clinically severe or critical represent less than 20% [12] . Every patient received a symptom-limited CPET, according to European Respiratory Society (ERS) standard criteria [14] and COSMED system (Quark CPET, COSMED, Rome, Italy) with an electronically braked cycle ergometer (Excalibur Sport, Lode, Groningen, Germany) was used. The incremental and symptom-limited CPET was conducted under the supervision of an expert physician. The symptom-limited CPET was carried out adding a specific antibacterial/antiviral filter (A-182-300-004 filter, COSMED, Rome, Italy). This could ensure a 99.999% viral filtering without interfering with the gas exchange and respiratory measurement up to ventilations of 200 L/min, almost the double of those reached by our best performing patients [15] . According to the large proportion of female post COVID-19 patients, the normal exercise capacity limit was set at 85% of the percent predicted peak VO2 (%pVO2) for each patient [14] . For every patient we adopted the flowchart approach for CPET interpretation according to Wassermann and ERS criteria to identify the "main" reason causing the exercise limitation [14, 16] . The indication of the main reason for limiting the exercise capacity was done independently by two experienced cardiopulmonary exercise testing researchers; when the interpretation was discordant, a third evaluation by a senior CPET researcher was performed. All patients signed an informed consent to perform complete clinical evaluation as per protocol and to use their anonymized data for scientific purpose. (SD) or median and interquartile range (IQR) according to their distribution. Normally distributed variables were compared by means of unpaired Student's t test. Non-normally distributed variables were compared with the U Mann-Whitney non-parametric test. The main outcome measure was peak VO2 (pVO2), although patients were categorized according to percent predicted (%pVO2) below or above 85% (threshold for normality) [14] . A multivariate linear regression model was used to estimate the beta coefficients with 95% confidence interval (CI) of pVO2. The model was adjusted for time from hospital discharge to CPET and all clinically meaningful covariates with p <0.10 in univariate analysis. A sensitivity analysis was also performed in the following populations: 1) population with and without any significant cardiovascular disease (except for arterial hypertension), respiratory disease, kidney disease, or cancer (indicating patients with and without known previous disease that could impact on CPET results regardless of COVID-19); 2) population with a length of hospital stay ≤ 7 days (indicating patients with rapid recovery from symptomatic COVID-19 in whom muscular deconditioning appears very unlikely). All analyses were performed with R environment 3.6.3 (R. Foundation for Statistical Computing, Vienna, Austria) and packages tableone, finalfit, and ggplot2. Of the 225 patients evaluated, we excluded 12 for missing data on anamnesis and/or CPET/PFT evaluations, and 13 as they were unable to perform CPET (10 because of orthopedic reasons, 2 uncapable of pedaling, and 1 for general frailty). Eighty-five (42.5%) patients were active smokers. Patients with %pVO2 below 85% had lower median 6-minute walking test distance than patients with %pVO2 above 85% (525.0 (450.0-577.5) vs. 540 (480.0-600.0) meters, p=0.064). Except for 10 (4.8%) patients with heart failure with reduced ejection fraction (HFrEF), of whom 4 (2.0%) showing severe reduction in EF (<30%), and for 15 (7.5%) with chronic obstructive pulmonary disease (COPD), of whom 5 with concomitant HFrEF, there were no significant preexisting comorbidities at the time of COVID-19 infection. The reason for maximal CPET interruption was exhaustion/leg fatigue in 93.0% patients, dyspnea in 5, and arrhythmia induced by exercise in 2.0%. Median %pVO2 was 88.0 (78.3-103.1), with a median peak RER 1.1 (1.1-1.2); 49.5% patients achieved a %pVO2 below 85% (reduced exercise tolerance) while 50% attained above this cut-off value of %pVO2 (preserved exercise tolerance). The anaerobic threshold (AT) was identified by V-slope and ventilatory equivalents for O2 methods. Of the 99 patients with reduced %pVO2, 61 had normal AT. Among these, 9/61 (14.8%) had RLE, 21/61 (34.4%) CLE, and 31/61 (50.8%) had non-cardiopulmonary limitation of exercise. The CPET test was well tolerated and only two cases (1.0%) had a mild post exercise symptomatic hypotension. During all the study protocol, lasting for more than nine months, the medical staff was periodically tested for asymptomatic COVID-19 infection: none was observed both before and after all the staff received the COVID vaccines. At PFT, median forced expiratory volume in one second (FEV1), forced vital capacity (FVC), and diffusing capacity of lungs for carbon monoxide (DLCO) were within normal limits. However, those patients with %pVO2 below 85% had percent predicted FEV1 (p<0.001), percent predicted FVC (p=0.001), and percent predicted DLCO (p<0.001) lower than patients with %pVO2 above 85%. At multivariate linear regression analysis, percent predicted FEV1 (β=5.29, 95%CI: 0.73-9.85, p=0.023), percent predicted DLCO (β=6.31, 95%CI: 2.49-10.13, p=0.001), and DLE maximal strength (β=14.09, 95%CI: 3.83-24.35, p=0.008) (Figure 1 ) were independently associated with pVO2 ( Table 2) . At multivariate linear regression analysis in sub-group of patients without any significant cardiovascular disease (except for arterial hypertension), respiratory disease, kidney disease, or cancer, percent predicted FEV1 (p=0.013), percent predicted DLCO (p=0.018) and DLE maximal strength (p=0.041), and 6-minute walking test (p=0.025) were independently associated with pVO2 (Table S1) . Percent predicted DLCO (p=0.018) and DLE maximal strength (p=0.007) remained independently associated with pVO2 in sub-group of patients with significant history of cardiovascular disease (except for arterial hypertension), respiratory disease, kidney disease, or cancer (Table S2) . At multivariate linear regression analysis in sub-group of patients with a length of hospital stay ≤7 days percent predicted FEV1 (p=0.039), percent predicted DLCO (p=0.037) and DLE maximal strength (p=0.029) remained significantly associated with pVO2 (Table S3 ). These main findings deserve mention: 1) about 1/2 nd of COVID-19 survivors had a significant alteration in pVO2 at 3 months after hospital discharge; 2) in nearly 1/3 rd (31/99) of patients with reduced %pVO2, this was probably due to abnormal peripheral oxygen extraction, most likely to some degree of muscle impairment, as DLE maximal strength was independently associated with pVO2; 3) 80% of patients experimented at least one disabling symptom at 3 months after hospital discharge, although there was no relationship between symptoms and abnormal pVO2; 4) cardiopulmonary exercise testing is both well tolerated and safe in post COVID-19; 5) cardiopulmonary exercise testing in post COVID-19 patients is safe for the supervising medical staff, as well. To our knowledge, for the first time, we reported clinical status and exercise capacity (pVO2) of COVID-19 patients performing complete CPET evaluation after hospital discharge. Our results are in agreement of that of Ong et al. [17] , who found a 41% prevalence in reduced pVO2 among 44 SARS 3-month survivors, albeit they included 22.7% patients that had required invasive ventilation. On note, abnormal physical function and performance in COVID-19 survivors have been preliminarily described by Belli et al. [9] , using 1-min sit-to-stand test and Short Physical performance Battery, without CPET evaluation. Due to the ongoing and accelerating COVID-19 worldwide pandemic, these observations raise concerns for health systems, as we proved that a substantial number of COVID-19 patients still had objective exercise impairment several months after hospital discharge. Moreover, a cardiopulmonary cause determining the exercise capacity and pVO2 reduction could be found in about 1/2 nd of post COVID-19 patients with normal AT. Interestingly, DLE maximal strength was independently associated with pVO2, suggesting that muscle impairment should be responsible, probably due to bed rest and subsequently muscular deconditioning, but also with a potential role for corticosteroid myopathy [18, 19] We believe that the most important finding is the relationship between pVO2 and maximal strength of the lower limb muscles, maintained even after accounting for cardiopulmonary variables, for Interestingly, these results were also confirmed when we analyzed those patients with short length of hospital stay (≤ 7 days), where the well-known effect of bed rest and subsequent loss of muscle mass could not have played a significant role. It has been known for many years that most critically ill patients face long-lasting functional impairment after discharge [18] ; what is mostly worrying about our data is that we found severe mid-term consequences of COVID-19 also in non-ICU patients. This observation supports the need for targeted management of these patients also during the acute phase (e.g. applying appropriate nutrition and early mobilization plans). Moreover, as there was no relationship with pVO2, symptoms alone should not guide the post-acute management of COVID-19 patients: more objective techniques, such as CPET, should probably be used to rapidly intercept and assess the exercise impairment and its underlying mechanisms and, perhaps, to decide whether to start a physical rehabilitation program. Most of the attentions of the research in the COVID-19 disease have been devoted to its prevention and to the acute phase of the disease, whereas less has been done on the follow-up phase. Our study demonstrates that most of post COVID-19 patients have not reached a full recovery at three months. The vast majority of them complains limiting symptoms and reduced exercise capacity. The cardiopulmonary exercise testing appears the only one that gives unequivocal results about the actual existence of a functional capacity reduction and also insights about the mechanisms of the reduction. From this point of view, we believe that CPET should be proposed to post COVID-19 patients complaining limiting symptoms, until an equally effective but more practical and easy test for functional capacity evaluation will be found. Our study has important limitations. Firstly, all patients came from a single area of the city of Genoa and the generalization of results could be misleading. Secondly, the functional capacity evaluation was conducted three months after hospital discharge, with the patients unsupervised in the meantime and no data available, except for the anamnestic risk factors and comorbidities, about the baseline conditions prior to COVID-19; prior to infection, only 20 patients (10.0%) showed significant baseline comorbidities which could have impact on subsequent evaluation. In addition, it was not possible to achieve a direct comparison with a control group since none of the patients undergone CPET evaluation before the COVID-19 (most of patients were healthy, before) and very few patients with other pathological conditions who usually undergone CPET have similar characteristics of our cohort (e.g. relatively low median age 58.8 years, low rate (5.5%) of significant baseline comorbidities, and very prolonged (median 17.0 days) length of hospital stay). Moreover, no direct structural evaluation at the muscle level was performed, and no data about body composition (except for BMI) was available. Finally, due to the outbreak of COVID-19 that forced to limit the interaction in order to reduce the contagion, we performed a real-world observational protocol not including blood gas analysis and ESC Guidance for the Diagnosis and Management of CV Disease during the COVID-19 Pandemic Cardiovascular aspects of COVID-19 Cardiac complications in patients hospitalised with COVID-19 Potential Effects of Coronaviruses on the Cardiovascular System: A Review Cardiovascular Manifestations of COVID-19 Infection Outcomes of Cardiovascular Magnetic Resonance Imaging in Patients Recently Recovered From Coronavirus Disease 2019 (COVID-19) Endothelial cell infection and endotheliitis in COVID-19 Low physical functioning COVID-19 and pulmonary rehabilitation: preparing for phase three Personal protective equipment for preventing highly infectious diseases due to exposure to contaminated body fluids in healthcare staff ERS statement on standardisation of cardiopulmonary exercise testing in chronic lung diseases Cardiopulmonary exercise testing in the COVID-19 endemic phase Whipp's Principles of Exercise Testing and Interpretation Pulmonary function and exercise capacity in survivors of severe acute respiratory syndrome ICU-acquired weakness and recovery from critical illness. The New England journal of medicine Steroid myopathy: some unresolved issues Mitochondria and Critical Illness The Role of Genetic Sex and Mitochondria in Response to COVID-19 Infection COVID-19, Mitochondria, and Interferon Mitochondria and microbiota dysfunction in COVID-19 pathogenesis The Authors declare that there is no conflict of interest.