Monitoring of ventilation during the early part of cardiopulmonary exercise testing: the first step to detect central sleep apnoea in chronic heart failure.

BACKGROUND AND PURPOSE To evaluate the prediction of nocturnal central sleep apnoea (CSA) syndrome from the presence of periodic breathing (PB) on diurnal monitoring of pre-exercise (cardiopulmonary exercise test [CPX]) parameters. CSA syndrome is commonly found in congestive heart failure (CHF) patients and has several prognostic and therapeutic implications but is frequently undiagnosed. Awake PB pattern is sometimes observed during the CPX cardiopulmonary monitoring period of gas exchanges in CHF patients referred to the stress test laboratory for routine peak VO2 determination. PATIENTS AND METHODS Forty-five consecutive ambulatory patients (2 women/43 men; 60.2+/-11.7 years old) with clinically moderate to severe CHF (New York Heart Association [NYHA] class II/III: 22/23; mean+/-standard deviation left ventricular ejection fraction [LVEF]: 30.5+/-6.6%) underwent a classical maximal CPX test including a 2-min period (pre-test) of gas exchange monitoring and nocturnal ambulatory polygraphic monitoring. PB was defined when a cyclical pattern of VE, VO2, VCO2, was visually noted during the pre-exercise period and/or during the first 4 min of the CPX. CSA syndrome was retained as a central apnoea-plus-hypopnea index (cAHI) equal to or more than 10/h. The sleep study scoring procedure was done independently of the knowledge of the CPX results. Sensitivity, specificity and predictive values were calculated and receiver operating characteristic (ROC) curve analysis was constructed. RESULTS Peak VO2 reached 16.4+/-5.2 mL kg(-1)min(-1) (55% of the theoretical value adjusted for gender and age). The polygraphy was completed and validated (at least five consecutive hours of sleep) in all cases. CSA syndrome was found in 28 (62%) patients (mean cAHI: 19.3+/-8.6/h). Sensitivity for the prediction of CSA syndrome reached 92.9% (two false-negative patients with a cAHI of nine) and specificity 94.1% with a predictive accuracy of 93.3%. The only false-positive patient suffered a moderate but significant obstructive sleep apnoea syndrome. Using ROC curve analysis, the W value reached 0.99 for the prediction of CSA from the presence of PB. The presence of CSA syndrome, using logistic regression analysis, is associated with a more severe functional status (NYHA: p<0.01, peak VO2: p<0.002), a lower basal and peak end-expiratory CO2 pressure (PETCO2, all p<0.03), a worse LVEF (p<0.01) and age equal to or more than 60 years (p<0.03). CONCLUSIONS The observation of PB in the preliminary period of the CPX test in CHF patients appeared highly predictive of the presence of CSA syndrome during sleep and could prompt the use of polygraphic monitoring in severe CHF patients.