Understanding Exertional Dyspnea and Exercise Intolerance in COVID-19

Purpose and Justification:

In less than one year, the novel coronavirus has infected more than 80 000 000 people
worldwide. Infection causes Corona Virus Disease 2019 (covid-19) and in some cases severe
acute respiratory syndrome. Overall risk of mortality from covid-19 is low, however, risk
radically increases with age and cardiovascular comorbidity. The long-term consequences of
covid-19 are not known. Already a phenotype of survivors with prolonged symptom burden has
presented; this phenotype is characterized by persistent respiratory (cough, sputum, dyspnea,
wheeze) and musculoskeletal (pain, fatigue) symptoms. Preliminary data from clinical exercise
testing conducted at the UofA pulmonary function laboratory suggests covid-19 survivors with
prolonged symptoms have significantly reduced exercise tolerance and increased exertional
dyspnea.

Impaired exercise tolerance measured as peak oxygen uptake (VO2peak) is the strongest
independent predictor of cardiovascular and all-cause mortality. The investigators
preliminary data in seven persistently symptomatic covid-19 survivors (PS-CoV) 3-months post
molecular confirmation of infection shows a mean 30% impairment in VO2peak relative to age-,
sex- and body mass index matched controls. Several facets of covid-19, including treatment
and recovery, may contribute to the range and severity of debilitation and impairment in
VO2peak in PS-CoV. The purpose of this study is to investigate impairments in VO2peak, and
pulmonary, cardiac and peripheral factors contributing to impaired VO2peak, exercise
intolerance, and persistent dyspnea in PS-CoV.

Coronavirus gains cellular entry through binding angiotensin converting enzyme in the lungs,
making the lungs and pulmonary vasculature a logical starting point for investigation of
persistent symptomology. During active infection, pulmonary vascular dysfunction,
microthromboemboli, micro-angiopathy and pulmonary inflammation and/or fibrosis are reported.
Accompanying this is a reduction in diffusion capacity at rest, increased tortuosity of
pulmonary vasculature, and elevated pulmonary vascular resistance. One mechanistic
explanation is that regions downstream of micro-thromboemboli become fibrotic secondary to
reduced blood flow resulting in reduced diffusion capacity. Physiological adaptation through
intussusceptive angiogenesis results in increased tortuosity of pulmonary vasculature, with a
secondary consequence of increased pulmonary vascular resistance. However, evidence of
isolated decreases in diffusion capacity in the absence of pulmonary fibrosis are at odds
with this theory. An alternative explanation is that pulmonary vascular dysfunction precedes
lesions viewed by computed tomography (CT) and changes in lung volumes. Regardless of
incipient damage, for ~1/3 of hospitalized covid-19 patients, the end result is pulmonary
fibrosis, impaired diffusion capacity (measured as the diffusion limitation of carbon
monoxide, DLCO), reduced forced vital capacity (FVC) and proportionately reduced forced
expiratory volume in one second (FEV1).

In PS-CoV, lung impairment at 3-month follow-up is characterized by reduced resting DLCO, FVC
and FEV1, and incomplete normalization of pulmonary CT consolidation and opacities.13 The
investigators preliminary data in PS-CoV show increased respiratory rate and VE/VCO2
(indicative of increased deadspace or excessive ventilatory drive) at peak exercise-
characteristic of parenchymal or restrictive lung disease and consistent with pathology of
covid-19 including parenchymal cell death and pulmonary fibrosis. Despite these findings,
initial data suggest that PS-CoV patients' operating lung volumes during exercise and peak
breathing reserve are relatively preserved. Previous work in COPD has shown that an elevated
VE/VCO2 during exercise is explained by higher deadspace, and this increased VE/VCO2
contributes to increased dyspnea secondary to increased drive to breathe. The investigators
work in COPD has shown that the increase VE/VCO2 is due to hypoperfusion of the pulmonary
capillaries as demonstrated by a reduced DLCO and reduced pulmonary capillary blood volume
during exercise, and that when pulmonary perfusion is improved by using inspired NO, VE/VCO2
and dyspnea are decreased resulting in an increase in VO2peak.

No data are currently available examining symptoms of dyspnea, pulmonary mechanics, VE/VCO2,
and impaired VO2peak in PS-CoV. Moreover, no data are available examining diffusion capacity
or pulmonary capillary blood volume responses during exercise, which may contribute to
increased VE/VCO2, pulmonary inefficiency, perceived dyspnea, and secondary cardiac
consequences.

Cardiac complications of covid-19 have been demonstrated and may contribute to impaired
VO2peak through a reduction in peak cardiac output (Qpeak). Limited data are available, but,
cardiac effects appear to be (mal)adaptation secondary to pulmonary vascular dysfunction,
angiopathy and increased pulmonary vascular resistance. Importantly, pulmonary vascular
dysfunction may impose a cardiac limitation to exercise in the absence of or preceding
structural cardiac changes as in early pulmonary hypertension (exercise induced pulmonary
hypertension). Complications mimic those observed in pulmonary hypertension whereby the thin
walled right ventricle insidiously adapts to and eventually fails against chronically
increased pulmonary artery pressure. This includes right ventricular hypertrophy, dilation
and hypokinesis, and in failure, uncoupling of tricuspid annular plane systolic excursion
(TAPSE) and pulmonary artery systolic pressure (PASP). In a study of 100 consecutive covid-19
patients at rest, 39% of patients had right ventricular dilation and dysfunction and 16% of
patients had left ventricular diastolic dysfunction. No reports of cardiac function during
exercise or cardiac mechanics in response to stress are available following covid-19, and it
is unknown whether cardiac consequences of covid-19 limit VO2peak or contribute to symptom
persistence in PS-CoV.

Detrimental changes in body composition occur in hospitalized covid-19 patients. During
active infection, frailty (in part characterized by muscle loss) is associated with increased
covid-19 severity and mortality. Reduced lean tissue mass and increased adiposity,
particularly in the thigh, are reported following bedrest and are known to impair VO2peak.
Reductions in VO2peak are twofold: absolute VO2peak is reduced due to loss of muscle mass,
and relative VO2peak (ml/kg/min) is reduced due to a combination of reduced absolute VO2peak
and a decrease in the ratio of muscle mass to total body mass. Moreover, bedrest is
associated with reduced mitochondrial density and oxidative enzymatic activity. No data are
available linking increased adiposity, reduced thigh muscle, or impaired muscle quality to
VO2peak or symptom persistence in PS-CoV.

The investigators preliminary data indicate VO2peak is impaired in PS-CoV survivors. The
magnitude of VO2peak, pulmonary, cardiac and peripheral impairment is not known in PS-CoV or
symptom free covid-19 survivors. Through this proposed study, the investigators aim to
comprehensively test VO2peak impairment in PS-CoV survivors and link physiology to symptom
persistence in covid-19.

Objectives:

There are 3 objectives of this study: 1) to evaluate VO2peak in PS-CoV and recovered covid-19
survivors (no longer symptomatic) compared to covid-19 naïve controls matched for age, sex
and body mass index; 2) to evaluate DLCO and pulmonary capillary blood volume at rest and
during exercise in these three groups; and 3) evaluate cardiac structure and function at rest
and during exercise in the three groups.

Hypotheses:

The investigators hypothesize that:

1. VO2peak will be impaired in PS-CoV relative to recovered (symptom free) covid-19
survivors and covid naïve controls, and that recovered covid-19 survivors will have
impaired VO2peak relative to covid naïve controls;

2. Relative to covid-19 naïve controls, PS-CoV will have reduced rest and exercise
pulmonary capillary blood volume and diffusion capacity, which will be correlated with
exercise VE/VCO2.

3. PS-CoV will have reduced peak cardiac output, increased PASP, and uncoupling of
PASP:TAPSE.