P231 PULMONARY ARTERY WEDGE PRESSURE VS LEFT VENTRICULAR END–DIASTOLIC PRESSURE DURING EXERCISE IN PATIENTS WITH UNEXPLAINED DYSPNEA

Abstract Background Left ventricular end–diastolic pressure (LVEDP) approximates LV stiffness. Pulmonary artery wedge pressure (PAWP) reflects the hemodynamic load imposed by the left heart on the pulmonary circulation, with the interposition of the left atrium (LA). PAWP is frequently seen as a surrogate measure for LVEDP, but the two measures may not coincide. The imprecision of PAWP estimates for LVEDP and the disagreement between the two measures have been widely described for cardiac catheterization at rest. PAWP measurement during exercise has been advocated to discriminate heart failure with preserved ejection fraction (HFpEF) from non–cardiac dyspnea (NCD), with an end–expiratory pathologic threshold ≥25 mmHg. However, a formal comparison of PAWP vs LVEDP during exercise has never been performed. Aim. We sought to compare LVEDP and PAWP during exercise. Methods We retrospectively analyzed consecutive patients with unexplained dyspnea and a normal LV ejection fraction, who performed right and left heart catheterization at rest and during exercise. Results Forty–six consecutive patients were included in the analysis (80% with a peak PAWP ≥25 mmHg). We found a good correlation between both end–diastolic and mean PAWP on one side, and LVEDP on the other side (R2>0.55). End–expiratory, end–diastolic (mid–A) peak PAWP had no bias as compared with LVEDP, while mean PAWP slightly overestimated LVEDP by 1–2 mmHg. However, confidence intervals were quite large (Figure 1), suggesting imprecision of PAWP estimates for LVEDP. A disagreement between PAWP and LVEDP, using a threshold of ≥ 25 mmHg for both variables at peak exercise, was found in 11% of patients. In 4%, PAWP was ≥25 but LVEDP <25 mmHg, due to the appearance of tall V waves in the PAWP position (increased LA stiffness), increasing PAWP above LVEDP. In 7%, LVEDP was ≥25 but PAWP <25 mmHg. Those patients would have not been diagnosed as HFpEF based on peak PAWP alone. However, they had a PAWP increase during exercise relative to cardiac output changes (PAWP/CO slope) >2 mmHg/L/min, as an alternative parameter confirming HFpEF. Conclusions PAWP estimates for LVEDP during exercise bear a good correlation and a minimal bias, albeit being burdened by a relevant imprecision. In patients with unexplained exertional dyspnea and peak PAWP <25 mmHg, additional measures (including LVEDP or PAWP/CO slope) may be indicated in order to maximize the diagnostic yield of exercise cardiac catheterization.