Congenital heart diseases are abnormalities in the structure of the heart at birth that may involve the interior surface of the walls of the heart, cardiac valves or blood vessels that carry blood to and from the heart to the body.1 The incidence of congenital heart disease is eight cases per 1,000 live births and the estimated prevalence is more than 1,000,000 of adults in the USA.1^,2 The prevalence of congenital heart disease in adults has increased in the last five decades due to significant progression in the treatment of these conditions during childhood.3 These patients, however, have decreased functional capacity4^-7 and higher morbidity and mortality rates as compared with healthy individuals.8^-14

Congenital heart diseases in adults have characteristics similar to heart failure (HF) caused by other etiologies, such as exercise intolerance,8^-11 ventricular dysfunction (right or left),15 cardiac arrhythmias,16 myocardial fibrosis,17 ventilatory inefficiency,3^,11^,18 increased inflammatory cytokines,19 and neurohormonal activation.20

The most precise method to quantify aerobic capacity is the direct measure of peak exercise oxygen consumption (VO₂ max) obtained exclusively by cardiopulmonary exercise testing (CPET). Reduced values of VO₂ max not only indicate more accentuated functional impairment, but also imply an adverse prognosis especially in HF,21 and also in other clinical conditions including congenital heart diseases in adults.8^-11

The aim of the present study was to describe the level of functional limitation and hemodynamic, ventilatory and metabolic responses to exercise in adults with congenital heart disease referred for CPET in a tertiary cardiology hospital.

This was an observational, cross-sectional study that evaluated adults with congenital heart disease (cyanotic or acyanotic), under medical, surgical or percutaneous treatment. The study was approved by the local ethics committee (approval number 47563315.2.0000.5272). All patients were informed about the aim of the study and signed an informed consent form. The study was conducted in accordance with the World Health Organization recommendations and the Helsinki Declaration (October 2013) and the Brazilian National Health Council resolution number 466/2012.

Patients referred for CPET between April 2016 and August 2017 were included in this study. Exclusion criteria were: age younger than 18 years, unwillingness to sign the informed consent form, contraindications for CPET according to the Guidelines on Exercise Tests of the Brazilian Society of Cardiology.22

Thirty-one (17 female; 54.8%) adults with congenital heart disease, aged 35.7 ± 14.2 years participated in the study (Table 1). Twenty-seven (87.1%) patients had acyanotic heart disease and the most frequent disease was Tetralogy of Fallot (n = 9; 29.1%) (Table 1). Twenty-four patients had a history of surgery (77.4%), one patient (3.2%) had undergone percutaneous treatment, and six patients (19.4%) were under medical treatment with no history of interventional treatment. The most frequent comorbidities were - systemic arterial pressure (n = 3; 9.7%), dyslipidemia (n = 2; 6.4%), hyperuricemia (n = 2; 6.4%), coronary artery disease (n = 1; 3.2%) and hypothyroidism (n = 1; 3.2%). Medications used by the patients are listed in Table 2.

CPET results are described in Table 3. Patients with congenital heart disease evaluated in the present study had reduced aerobic capacity, with a VO₂ max of 44.9% of predicted. AT could not be identified in 31 (25.8%) patients.

In addition, our study population showed ventilatory inefficiency for oxygen consumption, with OUES of 1.49 (61.4% of predicted) and limited SAP increment in response to exercise (∆SAP: 31.4 mmHg). Also, patients showed reduced O₂P (8.7 mL/beat; 58.9% of the predicted value), even when only acyanotic subgroup was considered in the analysis (9.1 mL/beat; 60.9% of the predicted value).

Duration of the ramp exercise test was 9.2 ± 3.6 minutes, which was considered adequate. R value was 1.21 ± 0.26, indicating maximum efficiency of the test.

According to the Brazilian Society of Cardiology criteria,22 the prevalence of parasympathetic autonomic dysfunction was 22.2%, and the small increase in SAP during incremental exercise indicated depressed response of this parameter. The prevalence of chronotropic incompetence was 44.4% based on different cut-off points for beta-blocker users26 and non-users22 (0.62 and 0.80, respectively).

Given the impossibility of HR measurement, patients with a permanent pacemaker (n = 3) and patients with atrial fibrillation (n = 1) were excluded from O₂P analysis, chronotropic response (HRR and chronotropic index) and parasympathetic autonomous modulation (HRrec).

In the present study, we showed that adult patients with congenital heart disease have reduced aerobic capacity, with a decrease not only in maximum aerobic power (indicated by VO₂ max), but also in aerobic capacity for submaximal activities (indicated by the AT). In addition, results of the CPET revealed ventilatory inefficiency for oxygen consumption and limited inotropic response to exercise.

Previous studies have reported lower values of VO₂ max4^-6^,9^,27 and AT5 in adults with congenital heart disease as compared with healthy adults.5 Reduced aerobic capacity is a common condition in this population, and 80% of these patients have a VO₂ max lower than predicted for age and sex.4 It is worth mentioning that many adults with congenital heart disease overestimate their own clinical conditions in light of the long period of exercise restriction. The level of exercise intolerance is more accurately assessed by measurement of the VO₂max, and even asymptomatic patients considered as New York Heart Association class 1 have lower VO₂max when compared with healthy individuals of the same age,9 suggesting a discrepancy between a subjective and an objective approach of exercise capacity in this population.

The decrease in maximal aerobic capacity in congenital heart disease adults is so important that VO₂ max in these patients is comparable to that in patients with HF caused by other conditions.9 VO₂ max is a traditional marker of an unfavorable prognosis of HF and has a central role in the evaluation of eligibility for heart transplantation in this group.28 Similarly, a reduced VO₂ max is associated with higher morbidity and mortality in adults with congenital heart disease.8^,9

We found a mean VO₂ max of 19.1 mL.kg^-1.min^-1. Diller et al.,9 evaluating a large group of patients with congenital heart disease, suggested a VO₂ max of 15.5 mL.kg^-1.min^-1 as a cut-off for predicting cardiac events. Patients with a VO₂ max lower than this had a three-time higher risk of death and hospitalization. Based on the studies by Diller et al.,9^,10 our patients had a good prognosis. However, this was a young population (< 35.7 years of age) and, in fact, VO₂ max found in these patients was only 44.9% of predicted, i.e., considerably lower than that expected for healthy individuals of the same sex and age. Inuzuka et al.,8 also evaluated adults with congenital heart disease and suggested a cut-off value of 64% of predicted VO₂ max to identify patients with a low or high 5-year survival mortality risk.8 Therefore, according to these authors, our study group could be, in fact, at risk of a poor prognosis.

Modern concepts of HF classify this syndrome in progressive stages. Patients with cardiac structural abnormalities but no signs of HF would be designated as stage B.29^,30 Although this classification did not include congenital heart disease, in a scientific statement published in 2016, the American Heart Association recommended that patients with congenital structural heart disease should be classified as being at least stage B of HF.31 In agreement with this position, our results demonstrated a depressed response of SAP (∆SAP: 31.4 mmHg) concomitantly with reduced O₂P in absolute (8.7 mL/beats) and relative (58.9% of predicted) values, suggesting a limited inotropism during exercise. Although O₂P response cannot be attributed only to the systolic volume behavior during incremental exercise (since it also depends on the arteriovenous oxygen difference, which in turn is altered in cyanotic and hypoxemic patients), the higher frequency of acyanotic patients in our sample (87.1%) allow us to infer that the decreased O₂P was probably a consequence of a limited increase in systolic volume during exercise. We corroborate this hypothesis, as we found low O₂P values even when cyanotic patients (9.1mL/beat) were excluded from analysis. As previously mentioned, the decreased response of SAP reinforces the hypothesis of a modest inotropic response during exercise.

VO₂ max has been recognized as the most important marker of morbidity and mortality in HF. Recent evidences, however, have suggested other parameters related to ventilatory efficiency, notably VE/VCO₂ slope and OUES, as better prognostic predictors in HF.32 Similar phenomenon was seen in congenial heart disease in adults. Dimopoulos et al.,13 reported that adults with congenial heart disease had higher VE/VCO₂ slope than healthy individuals, and such difference was observed in all types of congenital heart diseases and was directly proportional to the functional class. According to the authors, cyanosis was the main predictor of an increased VE/VCO₂ slope. This parameter was the most important marker of mortality in adults with congenial heart disease in acyanotic patients (cut-off point of 38).13 Similarly, Inuzuka et al.,8 reported that a VE/VCO₂ slope > 39 was a predictor of mortality only in acyanotic congenital heart disease patients. In our study, mean VE/VCO₂ slope was 27, which is considered normal. This may be explained, at least in part, by the small frequency of cyanotic patients in our sample.

In addition to ventilatory inefficiency for carbon dioxide, adults with congenital heart disease commonly have a ventilatory inefficiency for oxygen consumption also, represented by low OUES values.18 We found a mean OUES of 1.49, corresponding to only 61.4% of predicted. It is worth pointing out, for sake of comparison, that an OUES lower than 1.47 is associated with a poor prognosis of HF,33 which indicates that our study population had a ventilatory efficiency for oxygen similar to that in patients with more severe HF.

HR responses during incremental exercise, particularly in the presence of chronotropic incompetence, were equally important for risk stratification of adults with congenital heart disease. Diller et al.,14 reported that a HRR lower than 51 bpm was a predictor of lower survival in this population, especially when associated with a VO₂ max lower than 16.7 mL.kg^-1.min^-1, which increased the mortality risk by 3.8 times.14 The authors also identified the chronotropic index and reduction in HRrec as parameters associated with unfavorable prognosis. Similarly, Inuzuka et al.,8 identified the VO₂max combined with HRR as the main marker of mortality in cyanotic and acyanotic congenital heart disease. However, the cut-off point was 71 bpm.8 Since we found a mean HRR of 69 bpm, our patients would be at increased risk according to these results reported by Inuzuka et al.,8 but not according to those reported by Diller et al.14

The prevalence of chronotropic incompetence in our study was 44.4%, lower than that reported by Diller et al.,14 (62%); chronotropic index, however, were similar in both studies (0.64 and 0.70, respectively). We found a mean HRrec of 23.5 bpm, which was considered normal and suggestive of adequate parasympathetic autonomous modulation. Only 22.2% of our population met criteria for parasympathetric dysautonomia.

Recently, CP and VP have emerged as important markers of adverse events in HF,34 by using the product of SAP with VO₂ max, and the quotient of PAS by VE/VCO₂ slope, respectively. The cut-off points for CP and VP in HF were 1,750 mmHg.mLO₂.kg^-1.min^-1 and 3.5 mmHg, respectively.34

In adults with congenital heart disease, Giardini et al.,12 found that low CP values (lower than 1,476 mmHg.mLO₂.kg^-1.min^-1) were associated with a 15.4-fold increase in the 4-year risk of death. The authors described that despite an inverse relationship between CP and functional class, even asymptomatic patients had a lower CP than healthy individuals. We found an adequate CP (2,890.4 mmHg.mLO₂.kg^-1.min^-1), as compared with that in HF patients and also with the findings by Giardini et al.,12 In addition, we found adequate VP values taking into account HF patients, although we did not find other studies evaluating this parameter in adults with congenital heart disease in the literature.

Adults with congenital heart disease had similar responses to HF patients during a cardiopulmonary exercise test, indicating an impaired aerobic capacity, ventilatory inefficiency for oxygen uptake and reduced inotropic response to exercise.