Abstract
This review summarises various applications of how ventilatory equivalent (ventilatory efficiency or better still ventilatory inefficiency) and the minute ventilation (VʹE)/carbon dioxide production (VʹCO2) slope obtained from cardiopulmonary exercise testing (CPET) can be used in the diagnostic or prognostic workup of patients with congenital heart disease.
The field of congenital heart disease comprises not only a very heterogeneous patient group with various heart diseases, but also various conditions in different stages of repair, as well as the different residuals seen in long-term follow-up. As such, various physiologic disarrangements must be considered in the analysis of increased VʹE/VʹCO2 slope from CPET in patients with congenital heart disease. In addition to congestive heart failure (CHF), cyanosis, unilateral pulmonary stenosis and pulmonary hypertension (PH) provide the background for this finding. The predictive value of increased VʹE/VʹCO2 slope on prognosis seems to be more important in conditions where circulatory failure is associated with failure of the systemic ventricle. In cyanotic patients, those with Fontan circulation, or those with substantial mortality from arrhythmia, the impact of VʹE/VʹCO2 on prognosis is not that important.
Abstract
VʹE/VʹCO2 elevation is a common finding in patients with congenital heart disease. It can be used as a sign for right-to-left shunting, unilateral pulmonary stenosis, pulmonary hypertension and circulatory failure. It is predictive for clinical worsening. https://bit.ly/33gj3NQ
Introduction
The survival of patients with congenital heart defects has improved tremendously in the past decades. Nowadays, most of them reach adulthood despite various residuals limiting exercise capacity and quality of life [1, 2]. Cardiopulmonary exercise testing (CPET) is increasingly used in the follow up of these patients in order to detect these residuals, as well as to quantify their importance. Furthermore, some variables obtained from CPET have turned out to predict survival and/or deterioration of disease.
Cardiology for congenital heart defects comprises very heterogeneous patient groups (with very different heart defects), various conditions in different stages of palliation and “repair”, and different residuals in long-term follow-up. As such, many different kinds of physiologic disarrangement must be considered in the analysis of CPET (including the minute ventilation (VʹE)/carbon dioxide production (VʹCO2) slope).
Physiologic background
The physiologic background of VʹE/VʹCO2 is best described by the Bohr equation (figure 1) that has only a few requirements, as follows: 1. there is no carbon dioxide in the inhaled air; 2. the exhaled air is fully saturated with water; 3. the barometric pressure is 760 mmHg; 4. the alveolar partial pressure of gases is unique throughout the lung and does not change within the ventilatory cycle (such that equation 1 applies).
1Here, VʹE is under body temperature, pressure, fully saturated with water (BTPS) conditions and VʹCO2 is under standard temperature, pressure, dry (STPD) conditions (at 863 mmHg barometric pressure at body temperature, equivalent to 760 mmHg·(310 K/273 K)). Other parameters include mixed expiratory carbon dioxide tension (PĒCO2), alveolar carbon dioxide tension (PACO2), dead space volume (VD) and tidal volume (VT).
Carbon dioxide tension (PCO2) in the exhalation gas of healthy subjects at rest (left) and at exercise (right). PETCO2: end-tidal carbon dioxide tension; PĒCO2: mixed expiratory carbon dioxide tension; PACO2: estimated alveolar carbon dioxide tension; PĀCO2: time-averaged alveolar carbon dioxide tension.
In healthy subjects at rest, PACO2 can be estimated fairly well from end-tidal carbon dioxide tension (PETCO2) or even more easy from arterial carbon dioxide tension (PaCO2). Therefore, Enghoff [3] proposed substituting PACO2 with PaCO2 in the Bohr equation. However, this has tremendous effects on the definition of dead space, especially when the calculation is also used in subjects without uniform ventilation and without uniform alveolar partial pressures. The new calculation of dead space is nowadays referred as physiologic dead space and no longer resembles the anatomic dead space, as it includes all kinds of ventilation wasted in achieving a certain PaCO2. This additional alveolar dead space comprises unperfused alveoli, wasted ventilation from ventilation (Vʹ)–perfusion (Qʹ) heterogeneity, an increased Vʹ/Qʹ ratio (as in circulatory failure), diffusion impairment and right-to-left shunt [4].
Does end-tidal carbon dioxide tension help to differentiate the components of physiological dead space?
In healthy subjects at rest, PACO2 can be estimated fairly well from PETCO2 and the calculated Bohr dead space resembles the anatomical dead space of the conducting airways (the upper airways, larynx and the tracheobronchial tree down to but excluding the acini) reasonably well [4]. There is very little shunting by bronchial and Thebesian veins and the arterial–end-tidal carbon dioxide tension difference (Pa−ETCO2) is <5 mmHg. However, Pa−ETCO2 increases as PACO2 starts to oscillate with breathing (even with very slow breathing and a high VT) and PaCO2 represents only mean alveolar and not end-tidal PACO2. On the other hand, very quick and shallow breathing causes expiratory carbon dioxide tension (PCO2) to not reach the alveolar plateau [5].
At exercise there is so much carbon dioxide delivered from the working muscles to the alveoli that the oscillation of PACO2 increases. In exercising healthy subjects, Pa−ETCO2 exceeds PaCO2 with a Pa−ETCO2 value of −4 mmHg (figure 2) [6, 7]. On the other hand, patients with ventilatory restriction have a high ventilatory rate together with a low VT that reduces end-tidal Pa−ETCO2. Otherwise, patients with severely obstructed and little ventilated lung units have a continuous high inflow of additional carbon dioxide to the expiratory gas. At the end of expiration there is still efflux from these poorly ventilated regions, such that PETCO2 is extraordinarily high compared to PĒCO2 [7, 8]. This reduced ratio of PĒCO2/PETCO2 seems to be pathognomonic for severe bronchial constriction, especially at rest [7, 8].
Rough interpretation of an elevated ventilatory equivalent (minute ventilation (VʹE)/carbon dioxide production (VʹCO2)) with decreased mixed expiratory carbon dioxide tension (PĒCO2). The physiologic background is on the left with panel seven of a Wasserman nine-panel plot for a healthy subject on the right. PCO2: carbon dioxide tension; PaCO2: arterial carbon dioxide tension; PETCO2: end-tidal carbon dioxide tension; PETO2: end-tidal oxygen tension; SpO2: oxygen saturation measured by pulse oximetry.
Ventilatory inefficiency
According to the Enghoff modification of the Bohr equation, ventilatory inefficiency (elevated VʹE/VʹCO2) is due either to decreased PaCO2 (such as by voluntary hyperventilation or increased ventilatory drive by metabolic acidosis, enhanced peripheral ergoreceptor reflex, or increased chemoreceptor sensitivity) or to elevated physiological dead space (alveolar or anatomical). However, equation 1 also shows that VʹE/VʹCO2 is just the reciprocal of PĒCO2 and, for a detailed analysis of the reasons for elevated VʹE/VʹCO2, it seems more reasonable to have a close look at PĒCO2 and determine the reasons for the low level of this parameter. PĒCO2 resembles alveolar ventilation and must be analysed together with PETCO2 and PaCO2. The relative contribution of enhanced ventilatory drive, as well as alveolar and anatomic dead space, can be estimated roughly according to figure 2, bearing in mind the difficulties in the interpretation of end-tidal PETCO2 discussed previously.
VʹE/VʹCO2versus VʹE/VʹCO2 slope
The VʹE/VʹCO2 quotient can be calculated from every single breath throughout CPET. Usually, it declines hyperbolically at the beginning of the exercise test and reaches a minimum at the ventilatory compensation point, when lactic acidosis starts to increase ventilation independent of carbon dioxide production. On the other hand, the VʹE/VʹCO2 slope is measured based on the whole exercise dataset, up to the ventilatory compensation point and is mathematically the asymptote of the VʹE/VʹCO2 versus time curve that is almost touched at the ventilatory compensation point. Despite the smallest VʹE/VʹCO2 value throughout the test, the VʹE/VʹCO2 at the ventilatory compensation point can be measured more reliably [9] and most metabolic carts already correct VʹE/VʹCO2 for the dead space of the face mask. The VʹE/VʹCO2 slope is much more extensively studied in cardiology, possibly because the slope can be measured even in incomplete exercise tests where the patient does not reach the ventilatory compensation point.
Some centres measure the VʹE/VʹCO2 slope over the entire exercise dataset, arguing that this excludes inter-observer variability and gives a better correlation to survival in heart failure with reduced ejection fraction [10]. However, it makes no sense physiologically to include such non-linear values after the ventilatory compensation point, as ventilatory drive is changed substantially at this moment by overt lactic acidosis. Furthermore, it makes the VʹE/VʹCO2 slope dependent on the grade of exhaustion at the end of exercise, as these few data points at the end of the test increase the calculated slope [7, 11].
Recently, there have also been publications on the VʹE/VʹCO2 intercept, which is the extrapolation of the VʹE/VʹCO2 line to the y-axis. This intercept resembles the grade of VʹE/VʹCO2 improvement at low to moderate exercise and his in turn translates to the improvement in anatomic dead space due to the rise in VT. In some patients with severe pulmonary hypertension (PH), the intercept is at zero or even has a negative value, which means that there is no improvement or even a worsening in alveolar dead space at exercise [12]. This might be due to early hyperventilation, early failure of the circulation to maintain adequate lung perfusion or increasing local ventilation perfusion heterogeneity in these patients.
VʹE/VʹCO2 to detect diagnostic details in congenital heart disease
Right-to-left shunt
Only a small number of all congenital heart defects are cyanotic at birth. In one group, this cyanosis is due to a shunt defect combined with a blood flow obstruction downstream to (or in) the lungs. A second group has a malconnection on the veno-atrial level (e.g. total anomalous pulmonary venous return), the atrioventricular level (e.g. a double-inlet left ventricle), or the ventriculo-arterial level (e.g. transposition of the great arteries) that directs venous blood into the aorta. A third group consists of those patients with a univentricular heart, where one ventricle is serving both the pulmonary and systemic circulation (e.g. hypoplastic left-heart syndrome) unless Fontan circulation is established. A fourth group has intrapulmonary shunts (e.g. Osler's disease). Later in life a fifth group appears, namely Eisenmenger patients (who are born with a left-to-right shunt). The initial hyperperfusion of the lung leads to pulmonary vascular disease and finally shunt reversal. All these patients are cyanotic at rest. Oxygen saturation measured by pulse oximetry (SpO2) declines further at exercise [12, 13] when mixed venous blood (and shunting blood) becomes more desaturated and/or the pulmonary obstruction reaches a flow limitation. The nadir of SpO2 is about 30 s after exercise (personal observation).
VʹE/VʹCO2 slope is strongly and inversely associated with resting SpO2 in all congenital heart disease [11], as well as in cyanotic patients with and without PH [12, 13]. (Cyanosis has an additional effect on the VʹE/VʹCO2 slope in patients with PH [11]). As cyanosis is nowadays detected by pulse oximetry, the sensitivity of which has improved tremendously following the introducing of the forehead sensor and software eliminating movement artifacts, pulse oximetry should never be omitted on CPET of patients with congenital heart defects.
Exercise-induced right-to-left shunt
Shunts at the venous, atrial, ventricular or arterial level are usually left-to-right shunts. When the right heart fails or there is an extraordinary rise in pulmonary arterial pressure (PAP) (e.g. PH), a fixed right-ventricular outflow tract stenosis (e.g. pulmonary valve stenosis), or an increase in severe tricuspid regurgitation (sometimes seen in the Ebstein anomaly), these shunts flip to a right-to-left configuration. Even if there is no shunt at rest, these conditions can also open a foramen ovale (PFO) whenever right-atrial pressure exceeds left-atrial pressure [14].
In patients with Fontan circulation (e.g. a univentricular heart with direct connection of the caval veins to the pulmonary artery without the right ventricle), even small increases in PAP can lead to veno-venous fistulae with increasing right-to-left shunt.
When the exercise-induced shunt appears on CPET, there is a sudden rise in VʹE/VʹCO2, VʹE/oxygen uptake (VʹO2), respiratory exchange ratio (RER) and end-tidal oxygen tension (PETO2), combined with a sudden decrease in PETCO2 and SpO2 [14]. In patients with an early-appearing shunt the findings may, with the exception of the decrease in SpO2, be similar to anxiety-induced hyperventilation [14]. In patients with a shunt only at high workload, the anaerobic threshold and ventilatory compensation point can no longer be determined and it appears, very pronouncedly, as if these two thresholds are reached at the same time point combined with the beginning of decline in SpO2. In such patients, at the end of the exercise test, all typical findings for the exercise-induced shunt return quickly to the values seen before the opening of the shunt [14].
Finally, it should be mentioned that, in young healthy subjects, heavy exercise with very high pulmonary blood flow and elevated pulmonary blood pressure itself causes alveolar capillary dilatation with a diffusion–perfusion mismatch that ends up in a decline of SpO2 at the peak of exercise [15–17]. Again, cyanosis is nowadays primarily detected by pulse oximetry, but should be considered in the analysis of elevated VʹE/VʹCO2 slope with all the other changes with cyanosis.
Unilateral pulmonary stenosis
Many congenital heart defects are associated with unifocal or multifocal pulmonary artery branch stenoses. Patients with Fallot tetralogy are prone to a stenosis at the former origin of the arterial duct slightly distal to the origin of the left pulmonary artery. This is usually patched at surgical repair but might also appear later, even after retreatment by interventional balloon dilatation or stent implantation. In patients with pulmonary atresia with ventricular septal defect and multiple aorto-pulmonary collateral arteries, these aorto-pulmonary collaterals tend to become stenotic even when they are unifocalised and connected to the right-ventricular outflow tract. They often end up in multifocal central and peripheral stenoses and are difficult to treat because of their high recurrence rate.
There are also post-surgical conditions that might end up with pulmonary branch stenosis. Surgical shunts onto the pulmonary arteries are performed in many patients with complex congenital defects to stabilise pulmonary blood flow until surgical repair or to progress to a univentricular circulation according to Fontan. When these shunts are connected to one of the pulmonary arteries, there is the risk of stenosis after surgical closure of the shunt.
Another important example is transposition of the great arteries that is nowadays treated by an early arterial switch operation with the Lecompt manoeuvre. This is a translocation of the pulmonary bifurcation anterior to the ascending aorta with the risk of future pulmonary stenosis in one or both pulmonary arteries.
In Fallot patients [18], as well as in patients subsequent to arterial switch repair for transposition [19], it has been shown that an abnormal right-to-left ratio of pulmonary blood flow (obtained by cardiac magnetic resonance) has an increased VʹE/VʹCO2 slope compared to patients with balanced pulmonary blood flow. This increased VʹE/VʹCO2 slope improves after ballooning the stenosis [20] or after stenting [21]. The decline of the VʹE/VʹCO2 slope after treatment is directly associated with improvement in pulmonary blood flow ratio [20, 21]. A stenosis of the main pulmonary artery does not cause any perfusion mismatch and, therefore, shows no elevation in VʹE/VʹCO2 slope [22].
Pulmonary hypertension
All classes of PH can occur in patients with congenital heart defects (table 1), and both diagnosis and treatment have been extensively reported upon recently [24]. From idiopathic pulmonary arterial hypertension (PAH) [25] and left-heart failure PH [26, 27], we know that VʹE/VʹCO2 and PETCO2 are closely associated with PAH and this also holds true in congenital heart disease [28]. After vasodilator therapy, VʹE/VʹCO2 improves in Eisenmenger patients [29]; however, specificity is too low to use CPET and especially VʹE/VʹCO2 slope as a screening tool for PH in patients with congenital heart defects [28].
Types of pulmonary hypertension (PH) according to European Society of Cardiology (ESC)/European Respiratory Society (ERS) guidelines [23]; those associated with congenital heart defects are marked in bold
Acyanotic patients with congenital heart disease and PH are rare, and it should be assumed that VʹE/VʹCO2 reacts similarly to elevated pulmonary vascular resistance or left-heart failure. Furthermore, it should be expected that as the number of elderly patients with congenital heart disease continues to rise, the number of patients with left-heart failure especially should also increase over the coming decades [30, 31].
Enhanced ergoreceptor/chemoreceptor reflex
In patients with acquired congestive heart failure (CHF) there are often changes in ventilation pattern. At night there might be sleep apnoea, while at exercise one can detect exercise-induced oscillatory ventilation as well as an enhanced ventilatory drive measured as low PaCO2. In congenital cardiology however, cyanotic patients without heart failure show minimal nocturnal dips in oxygen saturation (so no sleep apnoea) [32]. This might be in contrast to patients with circulatory failure, as periodic breathing is only described in two acyanotic patients with right-heart failure [33]. There are only a few descriptions of exercise oscillatory ventilation in Fontan patients [34, 35]; however, they are contradictory concerning prognostic capabilities.
Concerning the impact of ventilatory drive on elevated VʹE/VʹCO2, there are only a few studies analysing exactly what is contributing to alveolar dead space or enhanced ventilation and most of these studies do not report an arterial blood gas analysis (i.e. PaCO2). Even those few that do are not consistent, as some report slightly reduced PaCO2 at rest which normalises on exercise [13, 36], while others report usually lowered PaCO2 in Eisenmenger patients [37], both with [36] and without [37] an impact on survival.
VʹE/VʹCO2 to estimate prognosis in congenital heart disease
Early studies looked for risk factors from CPET in the whole cohort of congenital heart disease. VʹE/VʹCO2 is associated with functional class [38] and can predict survival [38, 39] independently from New York Heart Association (NYHA) class [38], although the importance of the risk factor is low compared to resting SpO2, peak VʹO2, age, or heart rate reserve [39, 40]. However, these types of study do not help too much as, firstly, we already know that the patient group with the worst cyanosis, most severe PH and the least exercise capacity, which is the Eisenmenger group, has the worst prognosis compared to the other less disabled patients with congenital heart disease. Secondly, the mixture of various different diagnoses and conditions might dilute the results [40] and, as such, studies on isolated patient groups with similar conditions are needed to evaluate the additional prognostic importance of certain biomarkers (such as CPET variables) beyond simple parameters like diagnosis, age, body mass index, resting SpO2 and functional class.
Cyanotic congenital heart disease and Eisenmenger's disease
In cyanotic patients with congenital heart disease, death was not only attributed to heart failure but also to sudden death (e.g. arrhythmia, lung haemorrhage, dissection of ascending aorta and cerebral accidents) [41, 42] and other organ failure [24]. Despite peak VʹO2 having an important impact on survival, VʹE/VʹCO2 was not related to outcome [39].
Univentricular heart with Fontan circulation
Exercise parameters have an important prognostic value in patients with univentricular heart and Fontan circulation [43]. Despite a substantial elevation of both VʹE/VʹCO2 slope and VʹE/VʹCO2, these values were not predictive for survival or transplantation [44]. VʹE/VʹCO2 only predicts hospitalisation and various morbidities [44, 45].
Fallot tetralogy
In this patient group, VʹE/VʹCO2 is elevated mainly because of pulmonary branch stenosis [18] and a failing left ventricle. Possibly due to a high rate of sudden arrhythmic deaths, VʹE/VʹCO2 is not an additive risk factor for QRS duration and peak VʹO2 in predicting survival [46]. However, it does predict event-free survival independently from QRS duration and peak VʹO2 [46–48].
Transposition of the great arteries after atrial redirection according to Mustard or Senning
In patients with transposition of the great arteries, survival after atrial redirection depends mainly on arrythmia or right-ventricular (systemic) failure, whereas reinterventions were performed for baffle stenosis or leakage [49]. The VʹE/VʹCO2 slope is predictive for survival without emergency hospital admissions [50].
Conclusions
It is inappropriate to summarise over various congenital heart diseases and the various conditions of such patients throughout life. The predictive value of increased VʹE/VʹCO2 slope on prognosis seems to be more important in conditions where circulatory failure is associated with failure of the systemic ventricle. In cyanotic patients, patients with Fontan circulation, or those with substantial mortality from arrhythmia, the impact of VʹE/VʹCO2 on prognosis is not that important. However, the VʹE/VʹCO2 slope is an excellent predictor of clinical deterioration, unscheduled hospitalisation or other non-fatal clinical events both independently and in addition to peak VʹO2.
Footnotes
Provenance: Commissioned article, peer reviewed.
Number 9 in the Series “Ventilatory efficiency and its clinical prognostic value in cardiorespiratory disorders” Edited by Pierantonio Laveneziana and Paolo Palange
Previous articles in this series: No. 1: Laveneziana P, Di Paolo M, Palange P. The clinical value of cardiopulmonary exercise testing in the modern era. Eur Respir Rev 2021; 30: 200187. No. 2: Agnostoni P, Sciomer S, Palermo P, et al. Minute ventilation/carbon dioxide production in chronic heart failure. Eur Respir Rev 2021; 30: 200141. No. 3: Watson M, Ionescu MF, Sylvester K, et al. Minute ventilation/carbon dioxide production in patients with dysfunctional breathing. Eur Respir Rev 2021; 30: 200182. No. 4: Ward SA. Ventilation/carbon dioxide output relationships during exercise in health. Eur Respir Rev 2021; 30: 200160. No. 5: Collins SÉ, Phillips DB, Brotto AR, et al. Ventilatory efficiency in athletes, asthma and obesity. Eur Respir Rev 2021; 30: 200206. No. 6: Schaegger MR, Guenette JA, Jensen D. Impact of ageing and pregnancy on the minute ventilation/carbon dioxide production response to exercise. Eur Respir Rev 2021; 30: 200225. No. 7: Weatherald J, Philipenko B, Montani D, et al., Ventilatory efficiency in pulmonary vascular diseases. Eur Respir Rev 2021; 30: 200214. No. 8: Neder JA, Berton DC, Phillips DB, et al. Exertional ventilation/carbon dioxide output relationship in COPD: from physiological mechanisms to clinical applications. Eur Respir Rev 2021; 30: 200190.
Conflict of interest: A. Hager reports personal fees from Actelion Pharmaceuticals and Bayer AG, outside the submitted work.
- Received June 2, 2020.
- Accepted November 17, 2020.
- Copyright ©ERS 2021
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org