This was a single-center retrospective study of patients with stable chronic heart failure who routinely underwent annual CPET to assess functional capacity and determine candidacy for heart transplantation. The study period was from December 2020 through January 2023. Patients were identified from a larger group of individuals initially referred for evaluation of exertional dyspnea and fatigue. The investigation was approved as an exempt study by the institutional review board of the University of Southern California Health Sciences Center (HS-21-00017). As policy, in our laboratory all patients are asked to study and sign a consent form prior to any CPET, including those conducted for routine clinical purposes. Heart failure was identified according to the consensus statement published by Bozkurt et al. [3] . Patients with pneumonia, acute heart and/or respiratory failure, cancer, COVID-19 infection, and chronic respiratory disorders such as obstructive airway, interstitial lung and neuromuscular diseases were excluded from analysis.

During the analysis, we did not separate patients with heart failure with reduced ejection fraction from those with preserved ejection fraction because there were fewer patients in the latter group. For the purpose of this study, we grouped all heart failure patients together and categorized them by mortality status and whether they received heart transplants or not ( Figure 1 ). All testing was completed before any heart transplants took place.

Spirometry was performed in the seated position according to American Thoracic Society/European Respiratory Society (ATS/ERS) guidelines [4] . Reference values for FVC and FEV₁ were from Crapo et al. [5] .

These studies incorporated the following details: 1) clinical and anthropometric characteristics of patients; 2) adherence to international guidelines for methods of CPET [6] ; and 3) the safety of CPET, defined as reported adverse events. The CPETs were performed before any heart transplants took place.

The exercise testing equipment consisted of a stationary cycle ergometer (Med Graphics CPX Ultima system, Medical Graphics Corporation, St. Paul, MN) that was calibrated before and after each test. The mechanical dead space volume, depending on the mouthpiece and connections used, ranged from 45 to 65 mL for this system. Calibration of gas concentrations using primary standard gases and flow was performed using a 3 L syringe prior to each test. Patients were asked not to exercise on the day of the test and to refrain from consuming caffeinated beverages four hours before the test. Following the explanation of each procedure, informed consent was obtained under witness. Prior to beginning the test, subjects were familiarized with the stationary cycle ergometer and mouthpiece and cycled on the ergometer for approximately 10 minutes. They were seated and breathed through a mouthpiece with a nose clip in place. After a minimum of five minutes of resting measurements, they were exercised on the ergometer with increasing workloads at increments of 5 - 15 Watts, based on patients’ tolerability, using the Godfrey protocol [7] . Data collection continued for several minutes post-exercise for gas collection and electrocardiogram monitoring purposes.

The following variables were recorded every 15 seconds: minute ventilation (V’_E), inspired oxygen concentration (FiO₂), end-tidal oxygen tension (PetO₂), inspired oxygen uptake (VO₂), and end-tidal carbon dioxide output (PetCO₂). Peak VO₂ was expressed as the highest 30-second average value recorded during the last stage of the exercise test [8] . Anaerobic threshold (AT) was determined by the V-slope method [9] . Heart rate and rhythm were monitored continuously throughout the study with the 12-lead ECG. In addition, the following variables were derived from the data: tidal volume (V_t), respiratory rate, minute ventilation Ve/maximum voluntary ventilation (Ve/MVV), respiratory exchange ratio (RER), O₂/pulse (VO₂/HR), and ventilatory equivalents for carbon dioxide and oxygen (Ve/VCO₂ and Ve/VO₂, respectively). Peak VO₂ was determined as the highest 20- to 30-second average achieved during exercise and expressed as ml/kg/min. The Ve/VCO₂ slope was computed using least squares linear regression fitting of breath-by-breath values recorded throughout the whole exercise [8] . Normal values for these variables were derived from Sun et al. [9] . Ve/VCO₂ was also corrected for VO₂ (Ve/VCO₂/VO₂) [10] . The chronotropic index (CI) was calculated by dividing the difference between maximal and resting heart rate (in beats/min) by the difference between maximal and resting oxygen uptake (in L/min). Predicted values for CI were computed as follows:

CI = 106.9 + (0.16 × age) + (14.3 × sex) - (0.31 × height) - (0.24 × weight) [11] .

All patients with VO₂max (expressed in mL/min) less than 84% predicted were considered as having functional limitation of cardiovascular, pulmonary vascular, or ventilatory limitation [12] ATS Statement on CPX]. Patients with a Ve/MVV of greater than 75% predicted were considered to have a reduced ventilatory reserve. An O₂/pulse of less than 80% predicted represented a reduced stroke volume related to circulatory limitation [13] . No patients experienced any adverse events during testing.

Patients who showed either ventilatory or circulatory limitation were classified based on predominant CPET features; if, for example, they exhibited resting tachycardia (heart rate > 90 beats/min), a normal Ve/MVV (<75%), a decreased O₂/pulse (<80% predicted), and ventilatory equivalents that were increased (>30) and/or failed to decrease with exercise, they were classified as having a circulatory deficit. The slope and intercept for Ve/VCO₂ were computed for all patients according to the relationship y = a + bx, where y was the difference between Ve at rest and peak exercise, x was the difference between VCO₂ at rest and peak exercise, a was the intercept and b was the slope [9] . The oxygen uptake efficiency slope (OUES) was derived from the relationship of VO₂ to the logarithm of Ve during exercise: VO₂ = a log 10 Ve + b, where “a” is the OUES and “b” is the intercept [14] .

Descriptive data are shown as mean and standard deviation. Normality of the data was confirmed using the Shapiro-Wilk test. Comparisons among cohorts were conducted by multivariate analysis of variance (ANOVA), using the Bonferroni correction to account for the large number of compared variables. Associations between physiologic variables were determined by Pearson’s correlation, expressed as r². A p-value of <0.05 was considered statistically significant for intergroup comparisons and for inter-variable associations.

Of 350 patients who underwent evaluation for exertional dyspnea and were screened for heart failure, 166 were diagnosed with heart failure related to structural heart disease ( Figure 1 ). The remaining patients (n = 184) had primary diagnoses of various forms of respiratory disorders or idiopathic pulmonary hypertension. Their overall mean age and standard deviation (±SD) was 52 ± 15.6 years with a distribution of 128 males and 38 females ( Table 1 ). Their mean body mass index (BMI) was 25.9 ± 2.5 kg/m². Eighty (48%) patients were former smokers (more than 2 years previously); none were current smokers. Beta-blockers were used among 75% of study participants. The severity of cardiac dysfunction was classified as New York Heart Association Severity II and III. In 97 patients for whom echocardiographic data were available, the mean (±SD) left ventricular ejection fraction (LVEF) was 38 ± 17%; eighty percent of patients exhibited a reduced ejection fraction (<40%). We found no significant differences in the LVEF among survivors and expired patients, nor among those who underwent heart transplant versus those who did not ( Table 2 and Table 3 ). Seventeen percent of patients were diagnosed with ischemic etiology of heart failure, while 75% were diagnosed with non-ischemic dilated cardiomyopathy. Twenty-eight patients (17%) expired by the end of data collection. Eighteen patients (11%) underwent heart transplantation. Two (11%) of the transplanted and 26 (18%) of the non-transplanted patients expired during the study.

Values are mean ± SD. Abbreviations: BMI, body mass index; LVEF, left ventricular ejection fraction; FVC, forced vital capacity; FEV1, forced expiratory volume in one second; Ve, minute ventilation; MVV, maximum voluntary ventilation; RER, respiratory exchange ratio; VO₂, oxygen output; VCO₂, carbon dioxide output; AT, anaerobic threshold; OUES, oxygen uptake efficiency slope; PetCO₂, end-tidal carbon dioxide tension; PetO₂, end-tidal oxygen tension.

Values are mean ± SD. Abbreviations are the same as in Table 1 .

Values are mean ± SD. Abbreviations are the same as in Table 1 .

Table 2 compares expired and living patients based on cardiopulmonary exercise testing (CPET) results. Expired patients had 19% lower forced vital capacity (FVC) and 18% lower forced expiratory volume in one second (FEV1) compared to living patients (p < 0.001 and p = 0.002, respectively). However, there were no significant differences among % predicted FVC, FEV1/FVC, and Ve/VCO₂ intercept values.

Concerning exercise capacity, expired patients exhibited 27% lower VO₂ max (p < 0.01), 20.7% lower VO₂max % predicted (p < 0.001), and 23% lower values in VO₂ at AT (p < 0.001) compared to living patients. Deceased patients also exhibited a 29% lower VCO₂ max (p < 0.001), 14% higher peak Ve/VCO₂ (p = 0.003), 59% higher Ve/VCO₂/O₂ (p = 0.005), and 19% higher Ve/VCO₂ slope (p = 0.009). The deceased group had a 17% lower O₂/pulse than the surviving group (p = 0.02). The oxygen uptake efficiency slope (OUES) was 25% lower in the deceased group (p < 0.001). While PetCO₂ at rest showed no significant difference among subgroups (p = 0.71), expired patients exhibited a PetCO₂ at peak exercise that was 7.7% less than in living patients (p = 0.01). No statistically significant differences were found between expired and living patients for Ve/VCO₂ at AT, Ve/VO₂ at AT, and ΔVO₂/ΔWork.

Table 3 compares physiologic variables in patients who were eventually transplanted with those who were not. Patients who underwent transplantation exhibited 18% lower FEV1 (L) values compared to those who did not receive a transplant (p = 0.033). No statistically significant differences were found between the two groups for FVC, FVC % predicted, FEV1/FVC, Ve/VCO₂ intercept, and Ve/VO₂ at AT.

With respect to cardiopulmonary exercise capacity, the to-be transplanted group exhibited a VO₂ max that was 28% less than that of the non-transplant group (p = 0.002). Similarly, the VO₂ at anaerobic threshold (VO₂ at AT) was 25% lower in the transplant group than in the non-transplanted group (p = 0.003). Additionally, the transplant group demonstrated a 35% lower value in peak carbon dioxide production (VCO₂ max) compared to the non-transplanted group (p = 0.001). Ve/VCO₂/VO₂ in the expired group was 65% higher than in living patients (p = 0.001).

Transplanted patients exhibited an oxygen uptake efficiency slope (OUES) that was 33% lower (p < 0.001) than in the non-transplanted group. While PetCO₂ at rest exhibited no significant differences (p = 0.35), the transplant group demonstrated a relative decrease in PetCO₂ at peak exercise, 11% lower (p = 0.007) than in the non-transplanted group. Transplanted patients exhibited a chronotropic index that was 94% higher than those who did not undergo transplant (p = 0.005), indicating a notable decrease in heart rate response during exercise. This difference was not observed between deceased and living patients. The O₂/pulse in the to-be transplanted group was 15% less than in the non-transplanted group, a non-significant difference (p = 0.06).

Our findings highlight differences in exercise capacity, gas exchange, and circulatory response associated with heart transplantation compared to patients who did not receive a transplant. Key findings in deceased and transplanted patients (recorded prior to transplant) included reduced lung function, VO₂max, O₂/pulse, OUES, and increased Ve/VCO₂max and Ve/VO₂ as compared to living and non-transplanted patients, respectively.

The chronotropic index among deceased patients and in those who later underwent heart transplant was increased as compared to survivors and patients who did not undergo transplant. The index considers age, resting heart rate, and functional capacity and is independent of the stage of exercise or the protocol used [15] [16] . After adjusting for a number of demographic and physiological factors, Robbins et al. [17] found that a high Ve/VCO₂ and a low chronotropic index remained independent predictors of death due to any cause. Others have described the Ve/VCO₂ slope as a key determinant of mortality [8] [18] - [22] . Lin et al. [23] found that patients with heart failure and with an OUES < 1.3 and Ve/VCO₂ > 38 exhibited a higher risk for cardiac events, particularly those with COPD. We also found lower mean OUES values in patients who underwent transplant, indicating a higher degree of dead space breathing contributing to the decrease in cardiopulmonary reserve. In this connection, the peak PetCO₂ in patients who underwent transplant was reduced, a reflection of the inability of cardiac output to keep up with ventilation, resulting in the overall ventilation/perfusion ratio exceeding 1.

Chronotropic incompetence may be considered a major limiting factor in the exercise capacity of patients with heart failure. It is an important cause of exercise intolerance and an independent predictor of adverse cardiovascular events and mortality [24] . Although the underlying mechanisms for chronotropic incompetence in HF are not fully understood, the imbalance of the autonomic nervous system that is shifted toward the sympathetic pathway decreases β-adrenergic responsiveness, resulting in a reduced heart rate response to exercise [25] . Witte et al. [26] found a linear correlation between change in HR and peak VO₂. CI may be considered a limiting factor in the exercise capacity of patients with HF, but it has not been a consistent finding. In a study of 195 patients, of whom 90 had severe left ventricular systolic dysfunction, Jamil et al. [27] showed that increasing heart rate in unselected patients with heart failure did not improve exercise tolerance or improve symptoms, and conversely, that lowering HR did not worsen exercise tolerance or exercise-related symptoms. There has been a lack of a standardized approach to diagnosing the disease, further complicated by changes in HR dynamics in the HF population, which render reference values derived from a normal population invalid.

Deceased patients exhibited lower O₂/pulse, OUES, VO₂max, VO₂ at AT, VCO₂max, and PETCO₂ at peak exercise, and higher VE/VCO₂max and VE/VCO₂ slope (as well corrected for VO₂) compared to living patients. The mean Ve/VCO₂ in our deceased patients is virtually the same as that reported by Sarullo et al. (41.8) [18] . In general, our findings align with those of Brawner’s study [1] , suggesting that even when cardiac transplantation is removed as an endpoint, these values still exhibit a high association with heart failure prognosis.

The difference in the O₂/pulse between deceased and surviving patients indicate a similar predictive feature of outcome [1] [10] .

We found that the OUES in surviving patients was 1.05 L/min/L/min and, on average, 33% higher than in individuals who expired. Guazzi et al. [10] reported an OUES cutoff of 1.05 for high-risk cardiac outcomes, similar to the findings of Lin et al. [23] who reported a threshold of 1.3. OUES is influenced by the onset of lactic acid production and therefore incorporates circulatory, ventilatory, and musculoskeletal function. The log transformation of Ve creates high linearity in relation to VO₂ [14] , thus making the OUES theoretically effort-independent, enabling its calculation also in the case of a submaximal CPET. Yet, OUES has not been considered a necessary variable to be described in a CPET report. In a study of over 2000 patients with heart failure with reduced ejection fraction, Gordon et al. [28] concluded that while OUES was associated with clinical outcomes independent of the VE/VCO₂ slope, its prognostic utility was inferior to that of peak VO₂, even when measured at submaximal effort.

Deceased patients exhibited a mean Ve/VCO₂/VO₂ of 3.5, 59% higher than in surviving individuals. Using multivariate analysis, Guazzi et al. [10] showed that the Ve/VCO₂/VO₂ index retained a prognostic power greater than that of both VE/VCO₂ slope and peak VO₂. A Ve/VCO₂/VO₂ ≥2.4 indicated a higher risk for cardiac mortality.

The mean CI among deceased patients was 17% higher than in survivors (34.4/L vs 29.3/L), although this was not a statistically significant difference due to variability. An increase in the CI has been associated with decreased exercise tolerance and worse outcomes in individuals with heart failure [11] . Spiro et al. [29] have suggested reference values for CI for healthy men (42 - 43/L) and women (63 - 71/L). Given that 77% of our subjects were men, a mean CI of 53.6/L is well above the normal range (for men). This finding indicates a steep relationship between change in heart rate vs change in VO₂ from rest to peak exercise and is typical in patients with heart failure, ischemic heart disease and valvular heart disease, severe deconditioning and certain myopathies [11] .

Deceased patients had lower values for FVC and FEV1. These findings can be attributed to several factors associated with heart failure, including pulmonary congestion, pleural effusion, and/or respiratory muscle weakness.

Overall, transplanted patients showed significantly lower FEV1 values, peak VO₂max, VO₂ at AT, and VCO₂ max compared to non-transplanted individuals. Our findings are similar to those of Mancini et al. [30] who reported that patients listed for heart transplant were more likely to have a VO₂max < 14 ml/kg/min. Our study’s VO₂max range of 11.4 to 15.9 ml/kg/min aligns with this observation. Similarly, Garcia Bras et al. [31] identified a Ve/VCO₂ slope closer to 40 as indicative of a higher likelihood of transplantation. Our study aligns with this observation and deviates slightly from the traditional Ve/VCO₂ slope of 35 as defined by the ACC/AHA guidelines [2] . There has been controversy over how the slope is computed; we computed the Ve/VCO₂ slope from the very beginning to the end of the Ve vs VCO₂ plot, which results in a higher value than when computing the slope from the beginning up to the respiratory compensation point (RT), as the slope increases from RT to VO₂max [22] [32] - [36] . The overall slope also has a greater association with mortality [8] [9] [22] [33] [36] [37] .

Of the ventilatory equivalents, only Ve/VCO₂ discriminated between transplanted and non-transplanted patients. The finding that Ve/VO₂ did not may be related to less variability in Ve/VCO₂ than Ve/VO₂ during moderate intensity exercise because of the sensitivity of ventilatory control to PaCO₂ and pH in the physiologic ranges [ 9 ] [ 32 ] . Transplanted patients may also have reduced sensitivity to blood chemical changes from renal and neural effects of immunosuppressive agents. In addition, the OUES in patients who eventually underwent heart transplantation was, on average, 35% lower than in those who did not undergo transplant (p < 0.001), findings similar to those of Guazzi et al. [10] and Lin et al. [23] , and predictive of worse clinical outcomes (had they not undergone transplant). In addition, patients who eventually underwent heart transplant exhibited a CI that was 94% greater (p < 0.005) than those who did not undergo transplant, with similar implications regarding their potential outcomes.