C73 PICK YOUR THRESHOLD! HOW TO CALCULATE THE ANAEROBIC THRESHOLD TO STRATIFY HEART FAILURE PROGNOSIS: A COMPARISON BETWEEN ABSOLUTE VALUE, PERCENTAGE OF PEAK VO2 OR PERCENTAGE OF PREDICTED MAXIMUM VO2 IN A LARGE MULTICENTER COHORT OF HFREF PATIENTS WHO UNDER

Abstract

Abstract Background The anaerobic threshold (AT), identifies the moment during a maximal exercise when hyperventilation occurs in response to the introduction of an anaerobic metabolism. Its value is indicative of the subject‘s training and/or health, it can be used to guide training, rehabilitation or to define appropriateness to undergo major thoracic or abdominal surgery, and it is related to heart failure (HF) prognosis. AT can be expressed as absolute value or as the percentage of predicted maximum VO2 (VO2AT%pred). However, it is not uncommon to find papers that refer AT to the peak VO2 value achieved (VO2AT%peak), rather than its predicted value, but a direct comparison of the prognostic power of these different variables is missing. In this work, we aim to compare the risk–identifying ability of the AT value when expressed in these three different ways in a large population of HF patients. This will help identify which is more correct to use in assessing patient prognosis, especially when peakVO2 is not reached appropriately. Methods The population analyzed counts 7746 patients with HF with history of reduced EF (<40%), recruited between 1998 and 2020 within the MECKI score project. All patients underwent a maximal cardiopulmonary exercise test (CPET), executed in using a ramp protocol on an electronically braked cycle ergometer. Results The present study considered 6157 HF patients with identified AT during the CPET, with a median follow up of 1528 days (689–1826). The main characteristics are reported in Figure 1. Figure 2 shows stratification of patients according to these 3 variables divided in tertiles, considering cardiovascular death (combination of cardiovascular death, urgent transplant or LVAD implantation) as an end point. Comparing the AUC of the three variables considered, we found similar values between VO2AT and VO2AT%pred, while the peak VO2AT% value was significantly lower (p < 0.001), as shown in Figure 3A. Moreover VO2AT%pred is the only variable to maintain a comparable ROC to the peakVO2 one, with the others being significantly lower (Figure 3B). Conclusions VO2 at AT should always be expressed as % of predicted maximal VO2 to be reliable in predicting prognosis in HF patients. Moreover, evaluating a sub–maximal exercise, VO2AT%pred is the only variable to maintain a comparable prognostic power to the peakVO2 one.