| Summary: | Cardiopulmonary exercise testing (CPX) is established for the investigation of cardiac disease. In patients with heart failure multiple variables have shown prognostic benefit, although peak VO2 remains the most widely used. It is accepted that peak VO2 is affected by respiratory disease as well, and may be highly susceptible to influence from respiratory disease coexistent with cardiac disease. I propose an alternative variable will show significantly greater specificity for cardiovascular disease when compared to peak VO2 (and other variables) and aim to identify this 'ideal' variable through the investigation of patients undergoing isolated cardiac interventions. Patients were recruited to the following groups: undergone/going cardiac resynchronisation therapy (CRT); heart failure; mitral valve surgery; COPD or mixed disease. Each patient underwent echocardiography, pulmonary function tests and blood sampling. They then performed (after an initial familiarisation CPX) a baseline CPX. In patients undergoing intervention (CRT and mitral surgery) they underwent another CPX 2-3 months after intervention, and a further CPX at 6 months in the mitral valve group. I assessed the following characteristics to aid in finding an ideal variable: ability to discriminate between heart and lung disease; high reproducibility; relation to exercise capacity and disease severity; and appropriate changes with physiological interventions. ROC curve analysis showed that breathing reserve, OUES and double product had the greatest areas under curve when differentiating COPD from heart failure. OUES also showed excellent test-retest reproducibility and was strongly correlated to disease severity. Following mitral surgery OUES fell less at 2 months than peak VO2; OUES may therefore be influenced to a lesser degree by muscular maladaptation post-surgery. The ideal cardiopulmonary exercise test variable for cardiac patients has yet to be described. OUES appears to show discriminating properties between heart and lung, is reproducible and is less influenced by peripheral changes when compared to peak VO2.
|
|---|
