Oxygen pulse during cardiopulmonary exercise testing in cardiac patients as a surrogate of stroke volume: fact or fiction?

Abstract Background According to the Fick equation, oxygen uptake (VO2) equals the product of cardiac output (CO) and arteriovenous oxygen difference (a-vO2Diff). Cardiopulmonary exercise testing (CPET) provides the oxygen pulse (O2pulse) as the ratio of VO2 and heart rate. The O2pulse is often considered a surrogate for stroke volume (SV) during effort, ignoring the AVO2Diff, to assess the hemodynamic response during exercise. Purpose To compare O2pulse and SV measured independently by simultaneous CPET and exercise echocardiography (CPETecho). Methods Observation study of CPETecho exams performed between July 2015 and May 2024. The O2pulse was obtained by CPET at rest, intermediate and peak loads, while the SV was calculated by the ultrasound velocity time integral of the left ventricular outflow tract and the estimated a-vO2Diff by the Fick equation. Variables were compared by one-way ANOVA and Tukey test. Differences in O2pulse and SV between intermediate and peak (ΔO2pulse and ΔSV) were used to analyze the agreement of effort patterns: ‘ascending’ with values increasing from intermediate to peak, or ‘descending’ with the opposite response. The pattern agreement was evaluated by Cohen kappa. A p 0.05 was considered significant. Results A total of 1,866 exams were included in the study (54.1% males, age: 65.4 ± 13.9 y). The indications for the exams were heart failure with preserved ejection fraction (39.7%), valvular heart diseases (35.5%), investigation of myocardial ischemia (6.3%) and others (18.5%). The VO2 at intermediate level was 66±16% of the peak values. The O2pulse significantly raised from rest to intermediate and peak (4.0±1.9, 9.0±3.3 and 10.9±3.9 mL/beat; p 0.0001; Figure 1). The SV during rest, intermediate and peak were 70.2±18.1, 84.5±20.9 and 85.1±21.3 mL/beat, respectively, without significant difference between intermediate and peak (p = 0.074). In contrast, the a-vO2Diff progressively increased from rest, intermediate and peak (p 0.001). Notably, the ΔO2pulse and ΔSV exhibited a very weak correlation (r = 0.11, Figure 2). An ascending pattern was identified in 86.4% when considering ΔO2pulse and in 48% for ΔSV, with specific agreement of 43.4%. Conversely, a descending pattern was identified in 13.6% for ΔO2pulse and in 52.0% for ΔSV, with a specific agreement of 8.9%. The overall agreement was 52.3% (Cohen's kappa: 0.07), indicating almost no agreement between the patterns. Conclusion This study reveals a very low agreement between changes in O2pulse and SV during exercise in cardiac patients. The rise in O2pulse from intermediate to peak exercise was more closely related to the increase in a-vO2Diff whereas SV remained constant. Consequently, in cardiac patients, O2pulse changes cannot be assumed to be a reliable surrogate for SV changes, but rather reflect the combined effects of SV and muscle oxygen extraction, underscoring the importance of CPETecho for assessing exercise hemodynamic trends accurately.