The investigation conforms to the principles outlined in the Declaration of Helsinki. The institutional ethics committee approved the study protocol and all patients provided written informed consent.

We performed a single-center analysis of 442 consecutive HF patients referred to our institution between 2009 and 2018 with left ventricular ejection fraction (LVEF) ≤40% and in New York Heart Association (NYHA) class II or III who underwent CPET. All the patients were referred for assessment by the HF team and possible indication for HT or MCS.

Prospective follow-up included initial assessment of each patient within one month with clinical data including HF etiology (ischemic vs. non-ischemic), implanted cardiac devices, medication, comorbidities, NYHA class, laboratory, electrocardiographic and echocardiographic data, CPET data, and Heart Failure Survival Score.

Exclusion criteria were as follows: age <18 years; planned percutaneous coronary revascularization or cardiac surgery; exercise-limiting comorbidities (cerebrovascular disease, musculoskeletal impairment, or severe peripheral vascular disease); or previous HT. Patients with elective HT during the follow-up period (patients who had indication for HT and a heart became available in the first year of follow-up) were censored from the analysis at the time of HT.

All patients were followed for 12 months from the date of completion of the above-mentioned complementary exams.

The primary endpoint was a composite of cardiac death and urgent HT occurring during an unplanned hospitalization for worsening inotrope-dependent HF. Data were obtained from outpatient clinic visits and review of medical charts, and were complemented with a standardized telephone interview with all patients at 12 months of follow-up.

A maximal symptom-limited treadmill CPET was performed using the modified Bruce protocol on a GE Marquette Series 2000 treadmill. Gas analysis was preceded by calibration of the equipment. VE, VO2 and VCO2 were acquired breath-by-breath using a SensorMedics Vmax 229 gas analyzer. Heart rate (HR) was measured by continuous electrocardiogram, blood pressure (BP) was obtained manually with a sphygmomanometer, and oxygen saturation was monitored by pulse oximetry. Standard spirometry was not available in all patients.

Patients were encouraged to perform exercise until the RER was >1.10. A submaximal CPET was defined as one with a peak RER ≤1.10.

pVO2 was defined as the highest 30-s average achieved during exercise and was normalized for body mass. AT was determined by combining the standard methods (V-slope preferentially and ventilatory equivalents). The VE/VCO2 slope was calculated by least squares linear regression, using data acquired throughout the exercise. COP was measured as the minimum value of the ventilatory equivalent for oxygen (minimum VE/VO2). Partial pressure of end-tidal carbon dioxide (PetCO2) was reported before exercise, at AT and at peak exercise in mmHg units, and increase during exercise until AT was achieved (PetCO2DIF=PetCO2 at AT–PetCO2 at rest) was also calculated. Peak oxygen pulse (POP) was calculated by dividing derived pVO2 by maximum HR during exercise and was expressed in ml per beat. Circulatory power was calculated as the product of pVO2 and peak systolic blood pressure and ventilatory power was calculated by dividing peak systolic BP by the VE/VCO2 slope. HR reserve was calculated as the difference between maximum HR achieved with exercise and resting HR. HR recovery in the first minute after exercise was defined as the difference between maximum HR achieved with exercise and HR one minute into recovery.

Various composite CPET parameters were also automatically calculated.

All analyses compare patients with submaximal CPET and patients with maximal CPET. Data were analyzed using IBM SPSS for Windows, version 24.0 (IBM SPSS Inc, Chicago, IL).

Baseline characteristics were summarized as frequencies and percentages for categorical variables, as means and standard deviations for continuous variables when normality was verified, and as medians and interquartile range when normality was not verified by the Kolmogorov-Smirnov test. The Student's t test for independent samples or the Mann-Whitney test when normality was not verified were used for analysis of the variables.

The predictive power of various CPET parameters regarding the primary endpoint was analyzed for the highest area under the curve (AUC) in follow-up. Cut-off values for variables were determined from receiver operating characteristic (ROC) curves so that the sum of sensitivity and specificity was maximized. The Hanley-McNeil test was used to compare two correlated ROC curves.16

Survival curves were determined using the Kaplan-Meier method for subjects with submaximal CPET.

A total of 442 patients were enrolled in the study. The mean age of the overall population was 56.2±12.5 years, 80% were male and mean LVEF was 28.6±6.9%. A submaximal CPET (peak RER ≤1.10) was recorded in 66% of cases (290 patients). The baseline characteristics of both groups are presented in Table 1.

At one year, the primary endpoint had occurred in 30 (6.8%) patients, with 18 patients experiencing cardiac death and 12 patient undergoing urgent HT (Table 2). No patients required urgent MCS. A similar proportion of patients met the primary endpoint in both groups (5.9% vs. 8.6%, p=0.285), which also applied to the individual components of the primary endpoint (cardiac mortality: 3.1% vs. 5.9%, p=0.155; urgent HT: 2.8% vs. 2.6%, p=0.872).

All CPET parameters were analyzed for the highest AUC in the follow-up period (Table 3). In the submaximal CPET group, COP had the highest AUC value (0.989), followed by PetCO2 at AT (AUC=0.835) and VE/VCO2 slope (AUC=0.815). PP presented the lowest predictive power in this group (AUC=0.694). The Hanley-McNeil test (Table 4) revealed that COP's predictive power was significantly higher than that of other variables (p<0.05 for all), including pVO2 (p=0.004) and VE/VCO2 slope (p=0.026). Despite the high AUC value of both PetCO2 at AT and VE/VCO2 slope in comparison to pVO2, no statistically significant difference was found when the Hanley-McNeil test was applied. COP retained a high predictive ability for the study's primary endpoint, irrespective of patient age (≤50 vs. >50 years) and gender, HF etiology (ischemic vs. non-ischemic), heart rhythm (sinus rhythm vs. atrial fibrillation) or the presence of chronic kidney disease (CKD).

In the maximal CPET arm, COP presented a low predictive power: AUC=0.632, with a significant difference in its predictive power between the groups (p=0.017) (Table 3). OUES and PP were the variables with the highest AUC values, 0.853 and 0.810, respectively. No statistically significant difference between submaximal and maximal CPET was found regarding other variables, including pVO2 (p=0.970) and VE/VCO2 slope (p=0.532).

Cut-off values for variables were determined from the ROC curves so that the sum of sensitivity and specificity was maximized. A COP of 36 had a sensitivity of 100% and a specificity of 89% for the primary endpoint. Patients with peak RER ≤1.10 presenting a COP value of less than 36 had a significantly lower 12-month survival free of events (p<0.001) (Figure 1).

Figure 1.

Survival analysis of patients in the submaximal cardiopulmonary exercise test group presenting a cardiorespiratory optimal point ≥36.

(0,1MB).

There are limitations to our study that should be noted. This was a single-center experience, and therefore the results may reflect local practice and results may not be applicable to other HF centers.

There is no consensus about the optimal peak RER cut-off to define maximal effort, especially in HF patients. The ISHLT guidelines10 use a peak RER of 1.05 as a cut-off but other documents propose different cut-offs, ranging between 1.0 and 1.10.10–13 In this study we chose a peak RER of 1.10 to define a maximal CPET.

Although there have been few studies assessing the performance of submaximal CPET parameters, our study's population has a similar clinical profile to that of other studies in this area (including age, prevalence of ischemic heart disease, diabetes and CKD and device implantation rates), with high rates of neurohormonal blockade drug therapy. However, HF may have been less severe in our population than in other studies, as LVEF and pVO2 were higher, and there was a low rate of events (particularly urgent HT), which may compromise the reproducibility of our results.31

Since subjects were referred for assessment by the HF team and possible indication for advanced therapies, they may not be representative of older HF patients or those with more comorbidities.

Finally, a long-term follow-up is crucial for effective validation of our results.